Mediator Engineering of Saccharomyces cerevisiae To Improve Multidimensional Stress Tolerance

ABSTRACT

Saccharomyces cerevisiae is a well-performing workhorse in chemical production, which encounters complex environmental stresses during industrial processes. We constructed a multiple stress tolerance mutant, Med15^V76R/R84K, that was obtained by engineering the KIX domain of Mediator tail subunit Med15. Med15^V76R/R84K interacted with transcription factor Hap5 to improve ARV1 expression for sterol homeostasis for decreasing membrane fluidity and thereby enhancing acid tolerance. Med15^V76R/R84K interacted with transcription factor Mga2 to improve GIT1 expression for phospholipid biosynthesis for increasing membrane integrity and thereby improving oxidative tolerance. Med15^V76R/R84K interacted with transcription factor Aft1 to improve NFT1 expression for inorganic ion transport for reducing membrane permeability and thereby enhancing osmotic tolerance. Based on this Med15 mutation, Med15^V76R/R84K, the engineered S. cerevisiae strain, showed a 28.1% increase in pyruvate production in a 1.0-L bioreactor compared to that of S. cerevisiae with its native Med15. These results indicated that Mediator engineering provides a potential alternative for improving multidimensional stress tolerance in S. cerevisiae.

IMPORTANCE This study identified the role of the KIX domain of Mediator tail subunit Med15 in response to acetic acid, H₂O₂, and NaCl in S. cerevisiae. Engineered KIX domain by protein engineering, the mutant strain Med15^V76R/R84K, increased multidimensional stress tolerance and pyruvate production compared with that of S. cerevisiae with its native Med15. The Med15^V76R/R84K could increase membrane related genes expression possibly by enhancing interaction with transcription factor to improve membrane physiological functions under stress conditions.

INTRODUCTION

Microbial fermentation of renewable feedstock represents an efficient and realistic solution to produce the desirable chemicals. During the industrial bioprocess, an industrial strain may encounter complex environmental stresses, such as temperature (1), acid (2), oxidative (3), and osmotic stress (4), which can negatively impact cell growth and inhibit metabolite production. Thus, stress tolerance is a critical physiological parameter for industrial strains. For this, several strategies have been built for evolving and achieving a robust industrial strain, such as enhancing the membrane barrier function by constructing lipid and cell wall (5), scavenging toxics by engineering transport system (6), regulating the general or specific transcription factors or coactivators (7), and activating signal transduction cascades (8). However, the achievement of single or multigene modifications is limited due to the complexity and multigenic control of stress response. To address this problem, engineering the general and specific transcription regulators has been developed to regulate the transcriptional regulatory network (2, 9).

The transcriptional regulatory network could be regulated through manipulating and engineering native, artificial, and exogenous regulators (10, 11). First, the native regulators mainly contain general transcription factors in eukaryotic cells (e.g., Saccharomyces cerevisiae) (7), global regulator cyclic AMP receptor protein (CRP) (12) and specific transcription factors (13). Engineering the native regulators provides a promising strategy to enhance cell growth under stress condition. For example, when transcription factor Haa1 was overexpressed in S. cerevisiae, the intracellular acetic acid was decreased by about 31.8%, resulting in a large increase in acetate tolerance (14). Second, the rational or semirational design of artificial transcription factors (ATFs), based on the understanding of structure and function of zinc finger proteins, is a viable strategy to enhance stress tolerance (15). For example, using zinc-finger-based ATF, the maximum optical density at 620 nm (OD₆₂₀) of S. cerevisiae was increased by about 60% compared to that of wild-type strain in the presence of 5 g/L acetic acid (16). Third, exogenous regulators, such as Rky17, Sut2, and IrrE, have been introduced and modified to enhance stress tolerance (17–20). For example, when the heterologous expression of IrrE from Deinococcus radiodurans in Arthrobacter simplex, cell viability was increased by ∼7-fold compared to that of wild-type strain under 16% (vol/vol) ethanol or 20% (vol/vol) methanol (21).

Mediator complex is a multisubunit conserved protein complex that is involved in nearly all gene expression in eukaryotic cells, which organizes into four modules: head, middle, tail, and cyclin-dependent kinase 8 (CDK8) (22–24). The tail module serves to recruit Mediator to sequence-specific transcription factors to active gene expression required for the response to environmental stress (including oxidative, salt, acid, and drug) (2, 25, 26). The Mediator tail subunit Med15 was first as Gal11 in S. cerevisiae (27, 28), which contained four structured domains: a KIX domain and activator-binding domains (ABDs) 1, 2, and 3 (29). The transcription factor-targeted three-helix bundle KIX domain in Med15 subunit was structurally conserved between mammals and yeast (30, 31). Here, to increase the multidimensional stress tolerance of S. cerevisiae, the KIX domains of Mediator tail subunit Med15 was identified as essential regulator and engineered by protein engineering. The final mutation strain, Med15^V76R/R84K, could enhance interaction between Med15 and transcription factors to change membrane fluidity, integrity, and permeability, thus increasing cell tolerance to acid, oxidative, and osmotic stresses.

RESULTS

The Med15 KIX domain was required for cell stress tolerance.

To investigate the effect of environmental stress on the growth of BY4742, the half-maximal inhibitory concentrations (IC₅₀s) for acetic acid, H₂O₂, and NaCl were calculated. We found that its IC₅₀ values were 0.25% (vol/vol), 2.8 mM, and 0.9 M, respectively (Fig. 1a). Thus, we selected 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl for further study due to the fact that cell growth was substantially inhibited, but sufficient biomass concentrations were still obtained under these conditions. In order to identify genes that are differentially regulated and therefore could contribute to the stress tolerance, transcriptome sequencing (RNA-seq) was conducted in BY4742 to compare global gene expression under YNB (0.67% yeast nitrogen base without amino acids, 2% glucose, containing histidine [20 mg/L], leucine [60 mg/L], lysine [200 mg/L], and uracil [20 mg/L]) medium and 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. The restrictive thresholds [|log₂(fold change)| ≥ 1.9; false discovery rate < 0.05] of differentially expressed genes were used for further analysis.

FIG 1.

Med15 KIX domain was required for stress tolerance in S. cerevisiae. (a) Effect of acetic acid, H₂O₂, and NaCl on the IC₅₀ of BY4742. (b) Gene expression overlap in BY4742 through transcription profile analysis under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. (c) The most differentially expressed genes were involved in membrane transport and lipid metabolism under acetic acid, H₂O₂, and osmotic stress. (d to f) Effect of gene deletion and overexpression on cell growth (OD₆₀₀) under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions, respectively. (g) Spot assays of BY4742, med15Δ, med15KIXΔ, med15Δ/KIX, and med15KIXΔ/KIX strains under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions, respectively. The experiments were performed in biological triplicates. Error bars indicate the standard deviations (*, P < 0.05 compared to that of the wild-type strain, as determined by a Student t test).

Gene expression was analyzed. While many genes were up- or downregulated under only one stress condition, a considerable number changed expression for multiple stress conditions, with 87 genes differentially expressed under all the stress conditions (Fig. 1b; see List S1 in the supplemental material). In addition, Gene Ontology (GO) term enrichment analysis revealed that these 87 genes were involved in membrane transport, lipid metabolism, DNA replication and repair, the cell cycle, amino acid transport and metabolism, the translation signal transduction pathway, and secondary metabolite biosynthesis (see Fig. S1). We concluded that membrane transport and lipid metabolism were the most notable differentially regulated pathways under all the stress conditions. Thus, membrane-related genes were proposed to respond to multidimensional stress conditions. The most significant differentially expressed genes involved in membrane transport and lipid metabolism (Fig. 1c; see also List S1 in the supplemental material)—especially for (i) eight genes under 0.3% (vol/vol) acetic acid, such as INO1, ARV1, PRY1, QDR3, GIT1, MCD4, ELO3, and CHO2, (ii) eight genes under 4 mM H₂O₂, such as GIT1, ARV1, INO1, ERG11, ELO3, PRY1, CHO2, and MCD4, and (iii) eight genes under 1.2 M NaCl, such as LCB3, INO1, GIT1, NFT1, ELO3, PRY1, MCD4, and CHO2—were selected, and the corresponding mutants were constructed and extensively characterized.

To test whether these genes were required for acid tolerance, we deleted or overexpressed INO1, ARV1, PRY1, QDR3, GIT1, MCD4, ELO3, and CHO2 in BY4742. Compared to BY4742, the final biomasses of arv1Δ, pry1Δ, and mcd4Δ strains were decreased by 68.3, 81.5, and 70.1%, respectively, under 0.3% (vol/vol) acetic acid conditions, but the final biomass of BY4742/ARV1 was increased by 37.8% (Fig. 1d). Then, to test whether these genes were required for oxidative tolerance, the genes GIT1, ARV1, INO1, ERG11, ELO3, PRY1, CHO2, and MCD4 were deleted or overexpressed. Compared to that of BY4742, the final biomasses of git1Δ, arv1Δ, pry1Δ, and mcd4Δ strains were decreased by 78.9, 77.5, 29.8, and 57.3%, respectively, under 4 mM H₂O₂ condition, but the final biomass of BY4742/GIT1 was increased by 43.7% (Fig. 1e). Finally, to test whether these genes were required for osmotic tolerance, the genes LCB3, INO1, GIT1, NFT1, ELO3, PRY1, MCD4, and CHO2 were deleted or overexpressed. Compared to BY4742, the deletion of NFT1, PRY1, and MCD4 led to 75.2, 58.9, and 67.3% decreases, respectively, in the final biomass under 1.2 M NaCl condition, but overexpression of NFT1 showed a 48.5% increase (Fig. 1f). These results strongly demonstrated that ARV1, GIT1, and NFT1 played important role in response to acid, oxidative, and osmotic stress, respectively.

In our previous studies, we found Mediator complex play an important role in the expression of genes in response to environmental stress (2, 32). To further explore the key Mediator complex subunit that can regulate acid, oxidative, and osmotic stress simultaneously, we examined the effect of 13 Mediator subunits on the tolerance of S. cerevisiae to acetic acid, H₂O₂, and NaCl stress. Among them, the subunit Med15, located at the tail module, was found to be essential for yeast stress tolerance (Fig. 2). Med15 consists of one KIX domain, three activator-binding domains (ABD1/2/3), and Q-rich linkers between the domains (29). The Med15 KIX domain can interact with various transcription factors to regulate several processes, such as stress response and multidrug resistance (30, 33). Thus, we further investigated whether the KIX domain deletion caused a similar phenotype with Med15 deletion. We found that deletion of the KIX domain decreased cell tolerance to acetic acid, H₂O₂, and NaCl, which was similar to the phenotype of the med15Δ strain (Fig. 1g). Importantly, cell growth was restored under acetic acid, H₂O₂, and osmotic stress by complementing med15Δ and KIXΔ strains with the KIX domain (Fig. 1g). Finally, we tested the expression level of ARV1, GIT1, and NFT1 in med15Δ and KIXΔ strains and found that these genes were strongly downregulated under acetic acid, H₂O₂, and NaCl conditions, while the expression level of these genes was restored when complementing med15Δ and KIXΔ with the KIX domain (see Fig. S2). These results demonstrated that the Med15 KIX domain could help S. cerevisiae to resist acetic acid, H₂O₂, and NaCl stress.

FIG 2.

Med15 was required for S. cerevisiae tolerance to acetic acid, H₂O₂, and osmotic stress. Spot assays of S. cerevisiae with mediator subunit deletion under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions were performed.

Engineering the Med15 KIX domain to enhance cell stress tolerance.

The activation domain (AD) of transcription factors could target the large hydrophobic groove harbored by the three helices of the Med15 KIX domain in yeast (Fig. 3a) (26, 31, 34), suggesting that the AD-KIX interaction interface might serve as a promising target for transcription factor ADs. Thus, we analyzed the interaction interfaces of the Med15 KIX domain with ADs of four transcription factors (Crz1, Hal9, Msn2, and Asg1 [13, 35, 36]) by Z-Dock analysis (26). These four transcription factors ADs targeted the same hydrophobic groove of the Med15 KIX domain, and the residues of the Med15 KIX domain interacted with four transcription factors ADs overlap strongly (Fig. 3b; see also Fig. S3).

FIG 3.

Engineering Med15 KIX domain to enhance stress tolerance in S. cerevisiae. (a) Mediator complex builds a bridge between specific transcription factors (STFs) and RNA polymerase II (Pol II). UAS, upstream active sequence; AD, activation domain. (b) Z-DOCK analysis showed that transcription factors ADs could bind to the same hydrophobic groove of the Med15 KIX domain. The overlap residues of the Med15 KIX domain (in red) interacted with four transcription factors ADs. (c) Effect of Med15^V76R/R84K on cell growth (OD₆₀₀) under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. (d) Effect of Med15^V76R/R84K on the IC₅₀ of S. cerevisiae under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. (e) Effect of Med15^V76R/R84K on the intracellular ROS level under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. (f) Effect of Med15^V76R/R84K on cell viability under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. (g) Effect of Med15^V76R/R84K on cell morphology under 0.3% (vol/vol) acetic acid, 4 mM H₂O₂, or 1.2 M NaCl conditions. Scale bars, 1 μm. The experiments were performed in biological triplicates. Error bars indicate the standard deviations (*, P < 0.05 compared to that of the wild-type strain, as determined by a Student t test).

Based on this, to test the effect of these overlap residues between Med15 and transcription factors on KIX-dependent transcriptional responses and stress tolerance, six residues (N22, L24, L28, I31, N32, and N35) in helix 1, six residues (K44, I45, H48, N51, F52, and L56) in helix 2, and eight residues (E65, M67, S69, E72, V76, M77, Y81, and R84) in helix 3 from the interaction interface were chosen and then substituted by alanine. Among these mutations, the final biomasses of the mutant strains Med15^I31A, Med15^H48A, Med15^N51A, Med15^E72A, Med15^V76A, Med15^Y81A, and Med15^R84A were all higher than that of BY4742-13 (BY4742 with plasmid pY13) under acetic acid, H₂O₂, and NaCl stress (see Fig. S4). Thus, residues I31, H48, N51, E72, V76, Y81, and R84 were selected for further saturation mutagenesis, resulting in 672 candidate mutants. Among these mutations, the final biomasses of mutant strains Med15^V76R and Med15^R84K were all better than those of Med15^V76A and Med15^R84A under acetic acid, H₂O₂, and NaCl stress (see Fig. S4 and Table S2). Under the 0.3% (vol/vol) acetic acid condition, the final biomasses of mutant strains Med15^V76R and Med15^R84K were 0.99 and 1.03, which was increased by 11.2 and 9.6% compared those of Med15^V76A (0.89) and Med15^R84A (0.94), respectively. Under the 4 mM H₂O₂ condition, the final biomasses of mutant strains Med15^V76R and Med15^R84K were 1.04 and 1.02, which was increased by 5.8 and 7.5% compared to those of Med15^V76A (0.98) and Med15^R84A (0.95), respectively. Under the 1.2 M NaCl condition, the final biomasses of mutant strains Med15^V76R and Med15^R84K were 0.89 and 0.83, which was increased by 8.5 and 6.4% compared those of Med15^V76A (0.82) and Med15^R84A (0.78), respectively. Moreover, we investigated whether the double mutants further improve stress tolerance. Mutant strain Med15^V76R/R84K was constructed and extensively characterized under stress conditions. Under the 0.3% (vol/vol) acetic acid condition, the final biomass of double mutant strain Med15^V76R/R84K was 1.19, which was increased by 20.2, 15.5, and 67.6% compared to the biomasses of the Med15^V76R, Med15^R84K, and BY4742-13, respectively. Under the 4 mM H₂O₂ condition, the final biomass of double mutant strain Med15^V76R/R84K was 1.10, which was increased by 5.8, 7.8, and 44.7% compared to the biomasses of Med15^V76R, Med15^R84K, and BY4742-13, respectively. Under the 1.2 M NaCl condition, the final biomass of double mutant strain Med15^V76R/R84K was 0.96, which was increased by 7.9, 15.7, and 52.4% compared to the biomasses of Med15^V76R, Med15^R84K, and BY4742-13, respectively (Fig. 3c).

Finally, mutant strain Med15^V76R/R84K led to 28.0, 21.4, and 27.8% increases in the IC₅₀ for acetic acid, H₂O₂, and NaCl, respectively, compared to those of BY4742-13 (Fig. 3d). The intracellular reactive oxygen species (ROS) level was affected by all the stress conditions, and we found that the ROS levels in mutant strain Med15^V76R/R84K were decreased by 40.2, 48.7, and 30.8%, respectively, compared to those of BY4742-13 under acetic acid, H₂O₂, and NaCl (Fig. 3e). Moreover, the cell viabilities of mutant strain Med15^V76R/R84K were increased by 18.4, 36.7, and 22.5% under acetic acid, H₂O₂, and NaCl conditions (Fig. 3f), which was also confirmed by field emission scanning electron microscopy (FE-SEM) (Fig. 3g). Notably, BY4742-13 and mutant strain Med15^V76R/R84K showed similarly healthy membrane morphology in YNB medium without stress conditions. BY4742-13 showed rough and broken membrane morphology under acetic acid, H₂O₂, and NaCl conditions, whereas mutant strain Med15^V76R/R84K barely displayed any sign of membrane damage. These results demonstrated that Med15^V76R/R84K could enhance yeast tolerance to acid, oxidative, and osmotic stress simultaneously.

Med15^V76R/R84K enhanced acid tolerance by decreasing membrane fluidity.

To investigate the effect of Med15^V76R/R84K on ARV1 expression on acid tolerance, the expression level of ARV1 was determined. The expression level of ARV1 was increased by 4.3-fold in mutant strain Med15^V76R/R84K compared to that of BY4742-13 under the 0.3% (vol/vol) acetic acid condition (Fig. 4a). To further confirm this result, ARV1 promoter was fused with enhanced green fluorescent protein (eGFP; P_ARV1-eGFP) and then introduced into med15Δ-13 and mutant strain Med15^V76R/R84K. eGFP fluorescence in med15Δ-13 could not be observed by fluorescence microscopy but could be observed in mutant strain Med15^V76R/R84K (Fig. 4b). In addition, when P_ARV1-eGFP was introduced into BY4742-13 and mutant strain Med15^V76R/R84K, eGFP fluorescence was increased by 79.5% in mutant strain Med15^V76R/R84K compared to that of BY4742-13 under the 0.3% (vol/vol) acetic acid condition (Fig. 4c). These results indicated that mutant strain Med15^V76R/R84K could improve ARV1 expression under acetic acid stress.

FIG 4.

Med15^V76R/R84K enhanced acid tolerance by decreasing membrane fluidity. (a) Effect of Med15^V76R/R84K on the mRNA level of ARV1 expression in S. cerevisiae under the 0.3% (vol/vol) acetic acid condition. (b) Effect of Med15^V76R/R84K on ARV1 promoter to drive the expression of eGFP reporter (P_ARV1-eGFP) in med15Δ-13 and mutant strain Med15^V76R/R84K under the 0.3% (vol/vol) acetic acid condition. Scale bars, 5 μm. (c) Effect of Med15^V76R/R84K on the green fluorescent of P_ARV1-eGFP in BY4742-13 and mutant strain Med15^V76R/R84K under the 0.3% (vol/vol) acetic acid condition. (d) Growth performance of S. cerevisiae with transcription factor (Gcn4, Hap5, Ino2, Rtg1, Flo8, and Aft1) deletion under the 0.3% (vol/vol) acetic acid condition. (e) Effect of Hap5 deletion on cell growth (OD₆₀₀) under the 0.3% (vol/vol) acetic acid condition. (f) Effect of Hap5 deletion on the mRNA level of ARV1 expression in the hap5Δ strain under the 0.3% (vol/vol) acetic acid condition. (g) Yeast two-hybrid assays were used to confirm the interaction between Med15 (or Med15^V76R/R84K) and Hap5. (h) Effect of Med15^V76R/R84K and Arv1 on sterol homeostasis under the 0.3% (vol/vol) acetic acid condition. (i) Effect of Med15^V76R/R84K and Arv1 on membrane fluidity under the 0.3% (vol/vol) acetic acid condition. The experiments were performed in biological triplicates. Error bars indicate the standard deviations (*, P < 0.05 compared to that of the wild-type strain, as determined by a Student t test).

To investigate transcription factors involved in Med15^V76R/R84K-dependent ARV1 expression under acetic acid stress, six transcription factors (Gcn4, Hap5, Ino2, Rtg1, Flo8, and Aft1) were selected according to transcriptome analysis (see Table S1 in the supplemental material). When they were respectively deleted, the hap5Δ strain showed a significant decrease in growth performance compared to those of other mutants under the 0.3% (vol/vol) acetic acid condition (Fig. 4d). In addition, ARV1 expression and cell growth in hap5Δ were decreased by 2.7-fold and 72.9% compared to those of BY4742 (Fig. 4e and f). Further, a yeast two-hybrid (Y2H) analysis was carried out to determine whether Med15 (or Med15^V76R/R84K) interacted with Hap5 in response to acid stress (Fig. 4g). Cells expressing Gal4_BD-Med15 (or Gal4_BD-Med15^V76R/R84K) and Gal4_AD-Hap5 could grow on the medium without histidine, which indicated that Gal4_BD-Med15 (or Gal4_BD-Med15^V76R/R84K) could interact with Gal4_AD-Hap5 to form the active Gal4. These results suggested that Med15^V76R/R84K could interacted with Hap5 to activate ARV1 expression under acid stress.

Arv1 plays an important role in sterol homeostasis, which is involved in sterol uptake, trafficking and distribution into membranes (37). To test the effect of Med15^V76R/R84K mutation and altered ARV1 expression on membrane sterol composition in response to acid stress, the sterol composition of BY4742-13, mutant strain Med15^V76R/R84K, arv1Δ, and BY4742/ARV1 strains in YNB medium with or without 0.3% (vol/vol) acetic acid were determined. Under the 0.3% (vol/vol) acetic acid condition, the proportion of ergosterol of arv1Δ was decreased by 23.4% compared to that of BY4742-13, whereas the proportion of ergosterol of BY4742/ARV1 and mutant strain Med15^V76R/R84K were increased by 37.9 and 64.4%, respectively (Fig. 4h). Because sterol distribution contributes to the fluidity of lipid membrane (38), which was further determined by steady-state anisotropy of diphenylhexatriene (DPH). As a result, the membrane fluidity of the arv1Δ strain was increased by 16.6% compared to that of BY4742-13 under the 0.3% (vol/vol) acetic acid condition, whereas the membrane fluidity of BY4742/ARV1 and mutant strain Med15^V76R/R84K were decreased by 8.3 and 14.4%, respectively (Fig. 4i). It is possible that mutant strain Med15^V76R/R84K could decrease membrane fluidity by enhancing interaction between Med15 and Hap5 to improve ARV1 expression for sterol homeostasis, thus enhancing acid tolerance.

Med15^V76R/R84K improved oxidative tolerance by increasing membrane integrity.

To investigate the effect of Med15^V76R/R84K on GIT1 expression on oxidative tolerance, the expression level of GIT1 was determined. GIT1 expression was increased by 3.6-fold in mutant strain Med15^V76R/R84K compared to that of BY4742-13 under 4 mM H₂O₂ (Fig. 5a). To further confirm this result, GIT1 promoter was fused with eGFP (P_GIT1-eGFP) and then introduced into the med15Δ-13 strain and the mutant strain Med15^V76R/R84K. eGFP fluorescence in strain med15Δ-13 could not be observed by fluorescence microscopy but could be observed in mutant strain Med15^V76R/R84K (Fig. 5b). In addition, when P_GIT1-eGFP was introduced into BY4742-13 and mutant strain Med15^V76R/R84K, eGFP fluorescence was increased by 58.1% in mutant strain Med15^V76R/R84K compared to that of BY4742-13 under 4 mM H₂O₂ (Fig. 5c). These results suggested that mutant strain Med15^V76R/R84K could increase GIT1 expression under oxidative stress.

FIG 5.

Med15^V76R/R84K improved oxidative tolerance by increasing membrane integrity. (a) Effect of Med15^V76R/R84K on the mRNA level of GIT1 expression in S. cerevisiae under the 4 mM H₂O₂ condition. (b) Effect of Med15^V76R/R84K on GIT1 promoter to drive the expression of eGFP reporter (P_GIT1-eGFP) in med15Δ-13 and mutant strain Med15^V76R/R84K under the 4 mM H₂O₂ condition. Scale bars, 5 μm. (c) Effect of Med15^V76R/R84K on the green fluorescent of P_GIT1-eGFP in BY4742-13 and mutant strain Med15^V76R/R84K under 4 mM H₂O₂ stress. (d) Growth performance of S. cerevisiae with transcription factor (Flo8, Yap1, Mig1, Ste12, Mga2, and Flo1) deletion under the 4 mM H₂O₂ condition. (e) Effect of Yap1 or Mga2 deletion on cell growth (OD₆₀₀) under the 4 mM H₂O₂ condition. (f) Effect of Yap1 or Mga2 deletion on the mRNA level of GIT1 expression in yap1Δ and mga2Δ strains under the 4 mM H₂O₂ condition. (g) Yeast two-hybrid assays were used to confirm the interaction between Med15 (or Med15^V76R/R84K) and Mga2. (h) Effect of Med15^V76R/R84K and GIT1 on phospholipid biosynthesis under the 4 mM H₂O₂ condition. (i) Effect of Med15^V76R/R84K and Git1 on membrane integrity analyzed by flow cytometry under the 4 mM H₂O₂ condition. (j) Effect of Med15^V76R/R84K and Git1 on membrane integrity confirmed by LSCM under the 4 mM H₂O₂ condition. Scale bars, 15 μm. The experiments were performed in biological triplicates. Error bars indicate the standard deviations (*, P < 0.05 compared to that of the wild-type strain, as determined by a Student t test).

To investigate transcription factors involved in Med15^V76R/R84K-dependent GIT1 expression under oxidative stress, six transcription factors (Flo8, Yap1, Mig1, Ste12, Mga2, and Flo1) were selected according to transcriptome analysis (see Table S1). When these genes were deleted respectively, the yap1Δ and mga2Δ mutant strains showed a significant decrease in growth performance compared to those of other mutants under 4 mM H₂O₂ (Fig. 5d). In addition, the final biomasses of yap1Δ and mga2Δ strains were decreased by 48.8 and 58.6%, respectively (Fig. 5e). Then, GIT1 expression in yap1Δ and mga2Δ strains were reduced by 1.4- and 3.6-fold compared to those of BY4742 under 4 mM H₂O₂ condition (Fig. 5f), suggesting that Mga2 plays a major role in regulating GIT1 expression. Further, a gene-specific interaction between Med15 (Med15^V76R/R84K) and Mga2 was observed under 4 mM H₂O₂ by Y2H assay (Fig. 5g). Cells expressing Gal4_BD-Med15 (or Gal4_BD-Med15^V76R/R84K) and Gal4_AD-Mga2 could grow on the medium without histidine, which indicated that Gal4_BD-Med15 (or Gal4_BD-Med15^V76R/R84K) could interact with Gal4_AD-Mga2 to form the active Gal4. These results suggested that Med15^V76R/R84K could interacted with Mga2 to activate GIT1 expression under oxidative stress.

Git1 has significant effect on phospholipid biosynthesis, which mediates uptake of both glycerophosphoinositol (GroPIns) and glycerophosphocholine (GroPCho) for further cellular metabolism (39). To test the effect of Med15^V76R/R84K mutation and altered GIT1 expression on the membrane phospholipid composition in response to oxidative stress, the phospholipid compositions of BY4742-13, mutant Med15^V76R/R84K, git1Δ, and BY4742/GIT1 strains were measured in YNB medium with or without 4 mM H₂O₂. Under the 4 mM H₂O₂ condition, the proportion of phosphatidylinositol (PI) of the git1Δ strain was decreased by 48.1% compared to that of BY4742-13, whereas the proportions of PI of BY4742/GIT1 and mutant strain Med15^V76R/R84K were increased by 28.1 and 43.5%, respectively (Fig. 5h). Because phospholipid head distribution could affect membrane integrity (40), the membrane integrities of BY4742-13, mutant Med15^V76R/R84K, git1Δ, and BY4742/GIT1 strains were further tested by SYTOX green staining with flow cytometry. As a result, the membrane integrity of git1Δ was decreased by 46.3% compared to that of BY4742-13 under the 4 mM H₂O₂ condition, whereas the membrane integrities of BY4742/GIT1 and mutant strain Med15^V76R/R84K were increased by 38.1 and 59.2%, respectively (Fig. 5i), and this result was also confirmed by laser scanning confocal microscopy (LSCM) (Fig. 5j). It is possible that Med15^V76R/R84K could increase membrane integrity by enhancing interaction between Med15 and Mga2 to improve GIT1 expression for regulating phospholipid biosynthesis, thus improving oxidative tolerance.

Med15^V76R/R84K increased osmotic tolerance by reducing membrane permeability.

To analyze the effect of Med15^V76R/R84K on NFT1 expression on osmotic tolerance, the expression level of NFT1 was measured. NFT1 expression was increased by 5.1-fold in mutant strain Med15^V76R/R84K compared to that of BY4742-13 under 1.2 M NaCl condition (Fig. 6a). To further confirm this result, NFT1 promoter was fused with eGFP (P_NFT1-eGFP), and then introduced into med15Δ-13 and mutant Med15^V76R/R84K strains. eGFP fluorescence in med15Δ-13 could not be observed by fluorescence microscopy but could be observed in mutant strain Med15^V76R/R84K (Fig. 6b). In addition, when P_NFT1-GFP was introduced into BY4742-13 and mutant strain Med15^V76R/R84K, eGFP fluorescence was increased by 86.2% in mutant strain Med15^V76R/R84K compared to that of BY4742-13 under the 1.2 M NaCl condition (Fig. 6c). These results suggested that Med15^V76R/R84K could increase NFT1 expression under osmotic stress.

FIG 6.

Med15^V76R/R84K increased osmotic tolerance by reducing membrane permeability. (a) Effect of Med15^V76R/R84K on the mRNA level of NFT1 expression in S. cerevisiae under the 1.2 M NaCl condition. (b) Effect of Med15^V76R/R84K on NFT1 promoter to drive the expression of eGFP reporter (P_NFT1-eGFP) in med15Δ-13 and mutant strain Med15^V76R/R84K under the 1.2 M NaCl condition. Scale bars, 5 μm. (c) Effect of Med15^V76R/R84K on the green fluorescent of P_NFT1-eGFP in BY4742-13 and Med15^V76R/R84K under the 1.2 M NaCl condition. (d) Growth performance of S. cerevisiae with transcription factor (Aft1, Cst6, Pdr3, Ste12, Hap5, and Flo8) deletion under the 1.2 M NaCl condition. (e) Effect of Aft1 or Cst6 deletion on cell growth (OD₆₀₀) under the 1.2 M NaCl condition. (f) Effect of Aft1 or Cst6 deletion on the mRNA level of NFT1 expression in aft1Δ and cst6Δ strains under the 1.2 M NaCl condition. (g) Yeast two-hybrid assays were used to confirm the interaction between Med15 (or Med15^V76R/R84K) and Aft1. (h) Effect of Med15^V76R/R84K on Nft1 activity under the 4 mM H₂O₂ condition. (i) Effect of Med15^V76R/R84K and Nft1 on the ratio of intracellular Na⁺ concentration to K⁺ concentration under 4 mM H₂O₂ stress. (j) Effect of Med15^V76R/R84K and Nft1 on membrane permeability under the 1.2 M NaCl condition. The experiments were performed in biological triplicates. Error bars indicate the standard deviations (*, P < 0.05 compared to that of the wild-type strain, as determined by a Student t test).

To investigate transcription factors involved in Med15^V76R/R84K-dependent NFT1 expression, six transcription factors (Aft1, Cst6, Pdr3, Ste12, Hap5, and Flo8) were selected according to transcriptome analysis (see Table S1). When these genes were deleted respectively, the aft1Δ and cst6Δ strains showed significant decreases in growth performance compared to those of other mutants under the 1.2 M NaCl condition (Fig. 6d). In addition, the final biomasses of aft1Δ and cst6Δ strains were decreased by 63.9 and 47.2%, respectively (Fig. 6e). Then, NFT1 expression levels in aft1Δ and cst6Δ strains were decreased by 3.8- and 1.4-fold compared to that of BY4742 under the 1.2 M NaCl condition (Fig. 6f), suggesting that Aft1 plays a major role in regulating NFT1 expression. Further, a gene-specific interaction between Med15 (or Med15^V76R/R84K) and Nft1 was observed under 1.2 M NaCl by Y2H assay (Fig. 6g). Cells expressing Gal4_BD-Med15 (or Gal4_BD-Med15^V76R/R84K) and Gal4_AD-Aft1 could grow on the medium without histidine, which indicated that Gal4_BD-Med15 (or Gal4_BD-Med15^V76R/R84K) could interact with Gal4_AD-Aft1 to form the active Gal4. These results indicated that Med15^V76R/R84K could interacted with Aft1 to activate NFT1 expression under osmotic stress.

NFT1, encoding an ATP-binding cassette (ABC) transporter, plays an essential role in inorganic ion transport, which provides a translocation channel for inorganic ion to take part in further cellular metabolism (41). To test the effect of Med15^V76R/R84K mutation and altered NFT1 expression on the Nft1 activity and intracellular Na⁺/K⁺ ratio in response to osmotic stress, the Nft1 activities of BY4742-13, mutant Med15^V76R/R84K, nft1Δ, and BY4742/NFT1 strains were measured in YNB medium with or without 1.2 M NaCl. Under the 1.2 M NaCl condition, the Nft1 activities of BY4742/NFT1 and mutant strain Med15^V76R/R84K were increased by 26.1 and 34.2% compared to that of BY4742-13, respectively, whereas the Nft1 activity could not be observed in the nft1Δ strain (Fig. 6h). As a result, after incubation in YNB medium with 1.2 M NaCl for 6 h, the intracellular Na⁺/K⁺ ratios of BY4742/NFT1 and Med15^V76R/R84K strains were decreased by 11.9 and 16.7% compared to that of BY4742-13, respectively, whereas the intracellular Na⁺/K⁺ ratio of the nft1Δ strain was increased by 16.7% (Fig. 6i). Because ABC transporters could affect membrane integrity, the membrane integrity were determined by SYTOX green staining with flow cytometry. The membrane integrity of BY4742/NFT1 and mutant strain Med15^V76R/R84K were increased by 28.5 and 47.6% compared to that of BY4742-13, respectively, whereas the membrane integrity of the nft1Δ strain was decreased by 15.9% under the 1.2 M NaCl condition (Fig. 6j). It is possible that Med15^V76R/R84K could reduce membrane permeability by enhancing interaction between Med15 and Aft1 to improve NFT1 expression for increasing inorganic ion transport, thus enhancing osmotic tolerance.

Med15^V76R/R84K increased pyruvate production by enhancing stress tolerance.

To investigate the applications of Mediator engineering in improving organic acid production, we tested whether Med15^V76R/R84K could improve the performance of pyruvate production. The plasmid containing Med15^V76R/R84K was introduced into pyruvate-producing strain S. cerevisiae TAM (TAM), obtained TAM Med15^V76R/R84K. Then, the pyruvate titers of TAM Med15^V76R/R84K and TAM-13 (TAM with plasmid pY13) were determined without pH buffering. It was found that the final biomass and pyruvate titers of strain TAM Med15^V76R/R84K were 6.8 and 31.2 g L⁻¹ in a shaken flask, values which were increased by 25.9 and 17.7% compared to the corresponding values in TAM-13 (5.4 and 26.5 g L⁻¹) (see Fig. S5). To quantify the effect of Med15^V76R/R84K on pyruvate production, the performances of TAM Med15^V76R/R84K and TAM-13 were investigated in a 1.0-L bioreactor. The final biomass and highest pyruvate titer of TAM Med15^V76R/R84K were achieved at 10.4 and 38.4 g L⁻¹, values which were increased by 26.8 and 28.1% compared to the corresponding values in TAM-13 (8.2 and 30.2 g L⁻¹) (Fig. 7a and b). In addition, the intracellular pH was determined, and we found that the intracellular pH of TAM Med15^V76R/R84K was 6.5, which was higher than that of TAM-13 (5.8) (Fig. 7c). The intracellular ROS level was also determined, and we found that the ROS level in TAM Med15^V76R/R84K was decreased by 24.7% compared to that of TAM-13 (Fig. 7d). Moreover, the cell morphology of TAM Med15^V76R/R84K and TAM-13 were analyzed by FE-SEM, and we found that TAM Med15^V76R/R84K could improve cell morphology (Fig. 7e). These results indicated that Med15^V76R/R84K could increase pyruvate production by improving intracellular environment and cell morphology during fermentation process.

FIG 7.

Med15^V76R/R84K increased pyruvate production in bioreactor culture. (a) Effect of Med15^V76R/R84K on cell growth (OD₆₀₀) during pyruvate production in a 1.0-L bioreactor. (b) Effect of Med15^V76R/R84K on pyruvate titer in a 1.0-L bioreactor. (c) Effect of Med15^V76R/R84K on the intracellular pH during pyruvate production in a 1.0-L bioreactor. (d) Effect of Med15^V76R/R84K on the intracellular ROS level during pyruvate production in a 1.0-L bioreactor. (e) Effect of Med15^V76R/R84K on cell morphology during pyruvate production in a 1.0-L bioreactor. Scale bars, 1 μm. The experiments were performed in biological triplicates. Error bars indicate the standard deviations (*, P < 0.05 compared to that of the wild-type strain, as determined by a Student t test).

DISCUSSION

In this study, we found that Mediator tail subunit Med15 was required for stress tolerance, and its KIX domain was engineered to enhance the tolerance of environmental stress by protein engineering. The final Med15 mutation, Med15^V76R/R84K, could interact with transcription factors Hap5, Mga2 and Aft1 to increase the expression level of genes ARV1, GIT1, and NFT1 under acid, oxidative, and osmotic conditions, respectively. As a result, the membrane fluidity, integrity, and permeability were changed to improve cell tolerance to acid, oxidative, and osmotic stress, leading to an increase in pyruvate production. These results open a new window for enhancing multidimensional stress tolerance to enhance chemical production by engineering mediator Med15 to improve membrane physiological function.

The Med15 KIX domain is successfully identified and engineered to enhance the tolerance of S. cerevisiae to complex environmental stress, such as acid, oxidative, and osmotic stress. Recently, a serial of strategies have been developed to enhance cell stress tolerance, including adaptive laboratory evolution, artificial mutagenesis, and stress response regulation (42–44). However, these strategies were not efficient for improving the tolerance of environmental stress due to the complexity process of stress response (42). Thus, researchers have to pay more attention to global transcription regulators such as H-NS (45), Spt15 (7), and cyclic AMP receptor protein (CRP) (46) and to specific transcription factors such as Cst6 (47) and Haa1 (48). Based on this idea, a series of methods have been developed to enhance the physiological functions of global transcription regulators, including gene deletion, gene overexpression, random mutagenesis, and DNA shuffling (9, 49, 50). For example, when Spt15^K31Q was expressed in Kluyveromyces marxianus, its ethanol tolerance was increased compared to that of wild-type strain in the presence of 6% (vol/vol) ethanol (7). Similarly, as an important global transcription regulator, the Med15 KIX domain could interact with the active domain (AD) of the transcription factor to change the expression level of Med15-dependent genes in the stress response process. For example, in C. glabrata (2, 25, 26), the Med15 KIX domain interacted with the Pdr1 AD to increase the expression level of efflux pump Pdr5, thus enhancing its multidrug resistance. In the present study, the Med15 KIX domain was rationally engineered to increase yeast tolerant to acetic acid, H₂O₂, and osmotic stress. This strategy provides a potential method to increase the multidimensional stress tolerance of industrial strains.

Med15^V76R/R84K could interact with the transcription factors Hap5, Mga2, and Aft1 to activate the membrane-related gene ARV1, GIT1, and NFT1 expression in response to acid, oxidative, and osmotic stress, respectively. Here, we found that deletion or overexpression of most of significant differentially expressed genes did not affect the growth rate of the cell under the tested stress conditions. According to previous research, genetic stability could account for why this happens. Genetic stability can be mediated by multiple mechanisms: (i) genetic redundancy could replace inactivated genes to perform normal functions, or (ii) rewiring genetic network could deal with change/disturbance, or (iii) adaptive mutations to cope with lethal mutations and survive (51). In addition, the genetic compensation response (GCR) was recently proposed also as a possible explanation for the phenotypic discrepancies between gene knockout and gene knockdown (52, 53). Nonsense-mediated mRNA decay caused by gene deletion mutation could guide the transcription complex to activate homologous gene transcription (52). We found that ARV1, GIT1, and NFT1 played important role in response to acid, oxidative, and osmotic stress, respectively. Two general ways have been developed to alter membrane functions through engineering global transcription factors to enhance the stress tolerance of an industrial strain. First, the Mediator complex directly regulates the expression level of membrane related genes to alter membrane functions (2, 32). For example, membrane integrity in C. glabrata was enhanced by overexpressing Med3 to upregulate the corresponding genes for fatty acid biosynthesis, leading to a 1.9-fold increase in cell viability at pH 2.0 (32). Second, the Mediator complex indirectly activates the membrane-related genes to alter membrane functions through transcription factors, such as Asg1, Rds2, and Crz1 (13, 35, 36). A good example is that Med15 is required for decanoic acid tolerance in S. cerevisiae by mediating interaction between Med15 and Oaf1 to activate gene expression, such as the genes EEB1, FAA1, and ELO1, in fatty acid metabolism (54). In the present study, Med15^V76R/R84K was engineered and could enhance interaction between Med15 and Hap5, Mga2, and Aft1 to increase the expression levels of the genes ARV1, GIT1, and NFT1 under acid, oxidative, and osmotic stress, respectively. These findings are not only important for enhancing multidimensional stress tolerance but also provide a better understanding of the mechanisms underlying the yeast response to multidimensional stress.

The robustness of industrial strains can be improved by changing membrane functions, such as membrane fluidity, integrity, and permeability. Membrane fluidity, integrity, and permeability can be affected by the type and content of sterols, the length and extent of unsaturation of membrane straight-chain fatty acids, the composition and distribution of phospholipids, and the activity of membrane proteins and the intracellular energy system (38, 55). Thus, a series of potential strategies, including metabolic, biochemical, and genetic engineering, were developed to change the functions and biophysical properties of microbial membranes through regulating their biosynthesis pathways and composition. First, membrane fluidity can be affected by the length and unsaturation of fatty acid chains and the composition of phospholipids and sterols (1, 56). For example, the membrane fluidity was changed by regulating desaturase (Ole1) to increase unsaturated fatty acids in S. cerevisiae (57) or expressing heterologous cis-trans isomerase (Cti) to introduce transunsaturated fatty acids in Escherichia coli (58), thus improving its cell tolerance to toxic products (such as octanoic acid and alcohols) and the adverse conditions (such as low pH and osmotic stress). In this study, the membrane fluidity in mutant strain Med15^V76R/R84K was decreased to enhance stress tolerance to acetic acid, due to the increase of fecosterol and ergosterol content, which was consisted with the fact that a high ergosterol content could decrease membrane fluidity (59). Second, membrane integrity can be affected by membrane lipid composition (40, 43, 60, 61), membrane protein (such as fatty acid transporter OmpF and FadL) (62), and transcription regulators (such as Upc2, Ecm22, Rds2, Crz1, Med3, and Med15) (2, 32, 35, 63, 64). For example, overexpression of CDS1, encoding phosphatidate cytidylyltransferase, and CHO1, encoding phosphatidylserine synthase in S. cerevisiae, showed found that its IC₅₀ value for NaCl was increased by 17.8% compared to that of the wild-type strain (43). In the present study, the membrane integrity in mutant strain Med15^V76R/R84K was increased to enhance oxidative tolerance due to the increase in PI proportion, which was consistent with the fact that phospholipid head distribution could be altered to increase membrane integrity (40). Third, membrane permeability is tightly related to membrane fluidity and integrity, which also can be regulated by engineering membrane lipids and proteins (such as transporters) (55). For example, membrane permeability in S. cerevisiae was reduced by deleting acetate transporter Ady2 to downregulate the intracellular pH and ROS level, thus enhancing acetic acid to improve acetic acid, ethanol, and H₂O₂ tolerance (65). In this study, the membrane permeability in mutant strain Med15^V76R/R84K was improved to enhance stress tolerance to NaCl due to the increase in NFT1 (ion transporter) expression. To sum up, mutant strain Med15^V76R/R84K improved membrane fluidity, integrity, and permeability to enhance yeast tolerance to acid, oxidative, and osmotic stress simultaneously. These findings demonstrated that Mediator engineering provides an easy and convenient choice for improving membrane functions to enhance multidimensional stress tolerance simultaneously, an approach which may be widely used in industrial biotechnology.

An improvement in stress tolerance by Mediator engineering is crucial for increasing chemical production (40). In the past decade, some strategies have been developed to improve membrane physiological functions for enhancing stress tolerance of industrial strain, resulting in an increase of chemical production, such as transcription factor engineering, transporter engineering, metabolic engineering, and adaptive laboratory evolution (7, 40, 62). These strategies could enhance tolerance to some alcohol, short-chain fatty acid, and lipophilic compounds and increase their titers. For example, using metabolic engineering strategies to genetically modulate the proportion of phosphoethanolamine (PE) in E. coli was associated with a significant increase in membrane integrity, which not only enhanced tolerance to exogenous octanoic acid (C₈) but also remarkably increased the C₈ titer (40). However, due to the complexity of regulatory networks involved in stress response and chemical production, enhanced stress tolerance does not always lead to an increase in chemical production (63, 66). Here, membrane physiological functions were improved by rational engineered Med15 KIX domain, resulting in an increase in multidimensional stress tolerance and pyruvate production.

In conclusion, Mediator engineering provides a potential way to change membrane function and biophysical properties, thus increasing yeast tolerance to acid, oxidative, and osmotic stress. With the development of biotechnology, it is possible to efficiently and rationally manipulate multidimensional stress tolerance based on a better understanding of mechanisms for global transcription regulator-mediated stress tolerance. Furthermore, Mediator engineering may provide an easy and convenient approach for improving microbial cell factories for desirable chemical production in the future.

MATERIALS AND METHODS

Strains, plasmids, and growth conditions.

The strains and plasmids used in this study are listed in Table 1 and Table 2, respectively. E. coli JM109 was used as the host to construct all plasmids. S. cerevisiae BY4742 (MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0) was used as the wild-type strain to construct all the mutant strains. The deletion and replacement strains were constructed by CRISPR/Cas9 technology with plasmid pML104 (67). For the overexpression strains, the target genes were amplified from the genome of BY4742 and cloned into pY26 driving with the GPD promoter. The recombinant plasmids were then transformed into BY4742 generated the corresponding overexpression strains. Site-specific mutations were performed by a PCR-based method using the mutagenic primers. All of the primers used in this study are listed in Table 3.

TABLE 1.

Strains used in this study

Strain	Genotype	Source or reference
E. coli JM109		Our laboratory
S. cerevisiae
BY4742	MATα his3Δ1leu2Δ0lys2Δ0ura3Δ0	Our laboratory
med18Δ	BY4742 med18Δ::KanMX4	73
med19Δ	BY4742 med19Δ::KanMX4	73
med20Δ	BY4742 med20Δ::KanMX4	73
med1Δ	BY4742 med1Δ::KanMX4	73
med5Δ	BY4742 med5Δ::KanMX4	73
med9Δ	BY4742 med9Δ::KanMX4	73
med31Δ	BY4742 med31Δ::KanMX4	73
med2Δ	BY4742 med2Δ::KanMX4	73
med3Δ	BY4742 med3Δ::KanMX4	73
med15Δ	BY4742 med15Δ::KanMX4	73
med16Δ	BY4742 med16Δ::KanMX4	73
cdk8Δ	BY4742 cdk8Δ::KanMX4	73
cyccΔ	BY4742 cyccΔ::KanMX4	73
med15 KIXΔ	BY4742 med15 KIXΔ::KanMX4	This study
med15Δ/KIX	BY4742 med15Δ::HIS3 pY13-PMED15/MED15 KIX	This study
med15 KIXΔ/KIX	BY4742 med15 KIXΔ::HIS3 pY13-PMED15/MED15 KIX	This study
BY4742-13	BY4742::HIS3 pY13	This study
Med15V76R/R84K	BY4742 med15Δ::HIS3 pY13-PMED15/Med15V76R/R84K	This study
ino1Δ	BY4742 ino1Δ::KanMX4	This study
arv1Δ	BY4742 arv1Δ::KanMX4	This study
pry1Δ	BY4742 pry1Δ::KanMX4	This study
elo3Δ	BY4742 elo3Δ::KanMX4	This study
qdr2Δ	BY4742 qdr2Δ::KanMX4	This study
cho2Δ	BY4742 cho2Δ::KanMX4	This study
git1Δ	BY4742 git1Δ::KanMX4	This study
mcd4Δ	BY4742 mcd4Δ::KanMX4	This study
erg11Δ	BY4742 erg11Δ::KanMX4	This study
plb1Δ	BY4742 plb1Δ::KanMX4	This study
lcb1Δ	BY4742 lcb1Δ::KanMX4	This study
nft1Δ	BY4742 nft1Δ::KanMX4	This study
BY4742-26	BY4742::URA3 pY26	This study
BY4742/INO1	BY4742::URA3 pY26-PGPD/INO1	This study
BY4742/ARV1	BY4742::URA3 pY26-PGPD/ARV1	This study
BY4742/PRY1	BY4742::URA3 pY26-PGPD/PRY1	This study
BY4742/ELO3	BY4742::URA3 pY26-PGPD/ELO3	This study
BY4742/QDR2	BY4742::URA3 pY26-PGPD/QDR2	This study
BY4742/CHO2	BY4742::URA3 pY26-PGPD/CHO2	This study
BY4742/GIT1	BY4742::URA3 pY26-PGPD/GIT1	This study
BY4742/MCD4	BY4742::URA3 pY26-PGPD/MCD4	This study
BY4742/ERG11	BY4742::URA3 pY26-PGPD/ERG11	This study
BY4742/PLB1	BY4742::URA3 pY26-PGPD/PLB1	This study
BY4742/LCB1	BY4742::URA3 pY26-PGPD/LCB1	This study
BY4742/NFT1	BY4742::URA3 pY26-PGPD/NFT1	This study
NFT1-HA	BY4742::URA3 pY26-PGPD/NFT1-HA	This study
Med15V76R/R84K/NFT1-HA	BY4742 med15Δ::HIS3 pY13-PMED15/Med15V76R/R84K/NFT1-HA	This study
BY4742-13/NFT1-HA	BY4742::HIS3 pY13/NFT1-HA	This study
med15Δ 13	BY4742 med15Δ::KanMX4::HIS3 pY13	This study
med15Δ-13/PARV1-eGFP	BY4742 med15Δ::KanMX4::HIS3 pY13, ::LEU2 YEplac181-PARV1/eGFP	This study
med15Δ-13/PGIT1-eGFP	BY4742 med15Δ::KanMX4::HIS3 pY13, ::LEU2 YEplac181-PGIT1/eGFP	This study
med15Δ-13/PNFT1-eGFP	BY4742 med15Δ::KanMX4::HIS3 pY13, ::LEU2 YEplac181-PNFT1/eGFP	This study
BY4742-13/PARV1-eGFP	BY4742::HIS3 pY13, ::LEU2 YEplac181-PARV1/eGFP	This study
BY4742-13/PGIT1-eGFP	BY4742::HIS3 pY13, ::LEU2 YEplac181-PGIT1/eGFP	This study
BY4742-13/PNFT1-eGFP	BY4742::HIS3 pY13, ::LEU2 YEplac181-PNFT1/eGFP	This study
Med15V76R/R84K /PARV1-eGFP	Med15V76R/R84K::LEU2 YEplac181-PARV1/eGFP	This study
Med15V76R/R84K /PGIT1-eGFP	Med15V76R/R84K::LEU2 YEplac181-PGIT1/eGFP	This study
Med15V76R/R84K /PNFT1-eGFP	Med15V76R/R84K::LEU2 YEplac181-PNFT/eGFP	This study
gcn4Δ	BY4742 gcn4Δ::KanMX4	This study
hap5Δ	BY4742 hap5Δ::KanMX4	This study
ino2Δ	BY4742 ino2Δ::KanMX4	This study
rtg1Δ	BY4742 rtg1Δ::KanMX4	This study
flo8Δ	BY4742 flo8Δ::KanMX4	This study
yap1Δ	BY4742 yap1Δ::KanMX4	This study
mig1Δ	BY4742 mig1Δ::KanMX4	This study
ste12Δ	BY4742 ste12Δ::KanMX4	This study
mga2Δ	BY4742 mga2Δ::KanMX4	This study
flo1Δ	BY4742 flo1Δ::KanMX4	This study
aft1Δ	BY4742 aft1Δ::KanMX4	This study
cst6Δ	BY4742 cst6Δ::KanMX4	This study
pdr3Δ	BY4742 pdr3Δ::KanMX4	This study
AH109	trp1Δ leu2-ura3Δ his3Δ gal4Δ gal80Δ LYS2::GAL1UAS-GAL1TATA-HIS3 GAL2UASGAL2TATA-ADE2 URA3::MEL1UAS-MEL1TATA-LacZ MEL1	Our laboratory
TAM-13	TAM::HIS3 pY13	This study
TAM Med15V76R/R84K	TAM med15Δ::HIS3 pY13-PMED15/Med15V76R/R84K	This study

TABLE 2.

Plasmids used in this study

Plasmid	Genotype	Source
pY26	2μm, Ampr, URA3, PGPD, PTEF	Our laboratory
pY13	CENARS, Ampr, HIS3, PMED15	Our laboratory
YEplac181	2μm, Ampr, LEU2	Our laboratory
pML104	2μm, Ampr, URA3, PSNR52-empty sgRNA cassette, PTDH3-CAS9	Our laboratory
pGBKT7	Kanr, TRP1, GAL4 DNA-binding domain fusion	Our laboratory
pGADT7	Ampr, LEU2, GAL4 transcription-activating domain fusion	Our laboratory

TABLE 3.

Primers used in this study

Function and primer	Sequence (5′–3′)
Deletion
KIX-L-F1	CTAACAAGCAATTACATATTCCC
KIX-L-F2	GTATGGAAACTTCAAATGTTCAAGCAGTAACGGCGTTTTTC
KIX-R-F1	AAAACGCCGTTACTGCTTGAACATTTGAAGTTTCCATACTTT
KIX-R-F2	ATCCTTGGAAGGTGGATAGTG
ARV1-L-F1	CTTCGTCCTACAGTTTGCATAT
ARV1-L-F2	TCTGGGCCTCCATGTCGCTGCGTTTATCACAGCTAATATCTACTTT
ARV1/KanMX4-F1	GATATTAGCTGTGATAAACGCAGCGACATGGAGGCC
ARV1/KanMX4-F2	CTGGATATTTTTTTATTTGCTCGACACTGGATGGCGG
ARV1-R-F1	ACGCCGCCATCCAGTGTCGAGCAAATAAAAAAATATCCAGTATTAC
ARV1-R-F2	ACTTCAGGGCTGGCGTAC
GIT1-L-F1	GTAGTAATAGCGGCGTAGAATG
GIT1-L-F2	TCTGGGCCTCCATGTCGCTGTTTTTATCCTATTCTATTTTTTTGAT
GIT1/KanMX4-F1	AAAATAGAATAGGATAAAAACAGCGACATGGAGGCC
GIT1/KanMX4-F2	GGAAGAAATCGATACCAATTTCGACACTGGATGGCGG
GIT1-R-F1	ACGCCGCCATCCAGTGTCGAAATTGGTATCGATTTCTTCCTC
GIT1-R-F2	TCCAAAGCACTTATTTCTAGGTTA
PRY1-L-F1	ACGCACAGAGTCATTTCCC
PRY1-L-F2	TCTGGGCCTCCATGTCGCTGGATTTATCAATTACGGGATATTAAA
PRY1/KanMX4-F1	TATCCCGTAATTGATAAATCCAGCGACATGGAGGCC
PRY1/KanMX4-F2	ATGAAAACTTTTCTTGAAAGTCGACACTGGATGGCGG
PRY1-R-F1	ACGCCGCCATCCAGTGTCGACTTTCAAGAAAAGTTTTCATTGAT
PRY1-R-F2	AAGGAAGTTCTGTACGAAATGTAA
MCD4-L-F1	TAGTGTAACGGCTATCACATCAC
MCD4-L-F2	TCTGGGCCTCCATGTCGCTGTTTTTACAGATTTCAGTTGTATGAGT
MCD4/KanMX4-F1	ACAACTGAAATCTGTAAAAACAGCGACATGGAGGCC
MCD4/KanMX4-F2	ATAGCGGTGGGTATGTGAATTCGACACTGGATGGCGG
MCD4-R-F1	ACGCCGCCATCCAGTGTCGAATTCACATACCCACCGCTAT
MCD4-R-F2	AATGCCTCTTTTGTTTGTTAGTT
NFT1-L-F1	GGTCACGGAACATCACTTTC
NFT1-L-F2	TCTGGGCCTCCATGTCGCTGTCCTAAAAATTAATACTGTGAATCG
NFT1/KanMX4-F1	CACAGTATTAATTTTTAGGACAGCGACATGGAGGCC
NFT1/KanMX4-F2	CCCGCTAGCCCCGCATCCAATCGACACTGGATGGCGG
NFT1-R-F1	ACGCCGCCATCCAGTGTCGATTGGATGCGGGGCTAG
NFT1-R-F2	CCTGTGGGATGCAGCTTAT
CST6-L-F1	CTATTGGCGAAAATTAAGATAAG
CST6-L-F2	TCTGGGCCTCCATGTCGCTGTATCCTACCAAAAAGGTGTGG
CST6/KanMX4-F1	CACACCTTTTTGGTAGGATACAGCGACATGGAGGCC
CST6/KanMX4-F2	TTGTGCTCACCAAAACTTTGTCGACACTGGATGGCGG
CST6-R-F1	ACGCCGCCATCCAGTGTCGACAAAGTTTTGGTGAGCACAA
CST6-R-F2	CTGCTAAGGAAGCGATGGGACAC
HAP5-L-F1	TATGCAGTATTAAGATCCGTTTT
HAP5-L-F2	TCTGGGCCTCCATGTCGCTGTATGCGAGTAAACAATCCTGAT
HAP5/KanMX4-F1	CAGGATTGTTTACTCGCATACAGCGACATGGAGGCC
HAP5/KanMX4-F2	CTGTATAACCATTAACTCTCTCGACACTGGATGGCGG
HAP5-R-F1	ACGCCGCCATCCAGTGTCGAGAGAGTTAATGGTTATACAGCTGC
HAP5-R-F2	TTATGGCAGAAGATTGTGGTG
YAP1-L-F1	AAATAAGTACGGGAACGAGGTA
YAP1-L-F2	TCTGGGCCTCCATGTCGCTGGGTTTAAGAAACAACTTTTCCTTC
YAP1/KanMX4-F1	GAAAAGTTGTTTCTTAAACCCAGCGACATGGAGGCC
YAP1/KanMX4-F2	TTTTCCATAAAGTTCCCGCTTCGACACTGGATGGCGG
YAP1-R-F1	ACGCCGCCATCCAGTGTCGAAGCGGGAACTTTATGGAAA
YAP1-R-F2	GCCACTAACAAGGATAGAAAGC
AFT1-L-F1	ATTGACTGTTGGATGAAAGGGTA
AFT1-L-F2	TCTGGGCCTCCATGTCGCTGTGTCGTAGATTTTTCTGTTATTTTT
AFT1/KanMX4-F1	TAACAGAAAAATCTACGACACAGCGACATGGAGGCC
AFT1/KanMX4-F2	AGTTTGATTTCATCTATATGTCGACACTGGATGGCGG
AFT1-R-F1	ACGCCGCCATCCAGTGTCGACATATAGATGAAATCAAACTTAGACG
AFT1-R-F2	TCAATCACAACAAAGAAGAAAGG
STE12-L-F1	AACAACTCTTCGCGGTCAG
STE12-L-F2	TCTGGGCCTCCATGTCGCTGCCTTGGTGAACAAGACAATTC
STE12/KanMX4-F1	AATTGTCTTGTTCACCAAGGCAGCGACATGGAGGCC
STE12/KanMX4-F2	AATTCAAAAATTATATTATATCGACACTGGATGGCGG
STE12-R-F1	ACGCCGCCATCCAGTGTCGATATAATATAATTTTTGAATTTATGATACAAG
STE12-R-F2	CTAAGCGATCATGTAGTTTTGGA
MGA2-L-F1	AGAGCGATTGGATGACAGTTAC
MGA2-L-F2	TCTGGGCCTCCATGTCGCTGAACGAAATGTTCTGTTCGCC
MGA2/KanMX4-F1	GGCGAACAGAACATTTCGTTCAGCGACATGGAGGCC
MGA2/KanMX4-F2	TATATACGTAAAAAAGCAGATCGACACTGGATGGCGG
MGA2-R-F1	ACGCCGCCATCCAGTGTCGATCTGCTTTTTTACGTATATATATATATATATG
MGA2-R-F2	AGACACTACCAACCCTCACAAC
qRT-PCR
ARV1-F1	ATCGCCTTTGGCTTCTACTA
ARV1-F2	TATTGGTGGGCTCCAGGTA
GIT1-F1	ATTTTAGATGGAAGACTGCTACTAC
GIT1-F2	GGTCCTTGATAACGGAACTG
NFT1-F1	CTAGTATGCCAATGGATAACAAA
NFT1-F2	GAAGACGGAACGAAGAGGTAA
ACT1-F1	CACTATTGGTAACGAAAGATTCAG
ACT1-F2	TAGAACCACCAATCCAGACG
Point mutation
pY13/PMED15-MED15-F1	ACTAAAGGGAACAAAAGCTGGAGCTCGGAGAACCGTGTTTGGAATT
pY13/PMED15-MED15-F2	GTATCGATAAGCTTGATATCGAATTCTCAAGTAGCACTTGTCCAATTATT
pY13/PMED15-MED15 KIX-F2	GTATCGATAAGCTTGATATCGAATTCTCAAGCAGTAACGGCGTTTT
MED15N22A-F1	GCGAAGAACGTCGCCGGGTTGCTTCAG
MED15N22A-F2	GCGAAGAACGTCGCCGGGTTGCTTCAG
MED15L24A-F1	AACGTCAACGGGGCGCTTCAGGTGCTC
MED15L24A-F2	GAGCACCTGAAGCGCCCCGTTGACGTT
MED15L28A-F1	TTGCTTCAGGTGGCCATGGACATTAAC
MED15L28A-F2	GTTAATGTCCATGGCCACCTGAAGCAA
MED15I31A-F1	GTGCTCATGGACGCTAACACTCTGAAC
MED15I31A-F2	GTTCAGAGTGTTAGCGTCCATGAGCAC
MED15N32A-F1	CTCATGGACATTGCCACTCTGAACGGA
MED15N32A-F2	TCCGTTCAGAGTGGCAATGTCCATGAG
MED15N35A-F1	ATTAACACTCTGGCCGGAGGGAGCTCC
MED15N35A-F2	GGAGCTCCCTCCGGCCAGAGTGTTAAT
MED15K44A-F1	GACACTGCTGATGCGATAAGAATTCAT
MED15K44A-F2	ATGAATTCTTATCGCATCAGCAGTGTC
MED15I45A-F1	ACTGCTGATAAGGCAAGAATTCATGCC
MED15I45A-F2	GGCATGAATTCTTGCCTTATCAGCAGT
MED15H48A-F1	AAGATAAGAATTGCTGCCAAAAACTTC
MED15H48A-F2	GAAGTTTTTGGCAGCAATTCTTATCTT
MED15N51A-F1	ATTCATGCCAAAGCCTTCGAGGCAGCT
MED15N51A-F2	AGCTGCCTCGAAGGCTTTGGCATGAAT
MED15F52A-F1	CATGCCAAAAACGCCGAGGCAGCTTTG
MED15F52A-F2	CAAAGCTGCCTCGGCGTTTTTGGCATG
MED15L56A-F1	TTCGAGGCAGCTGCGTTCGCAAAGAGC
MED15L56A-F2	GCTCTTTGCGAACGCAGCTGCCTCGAA
MED15E65A-F1	TCTTCAAAGAAAGCATACATGGACAGC
MED15E65A-F2	GCTGTCCATGTATGCTTTCTTTGAAGA
MED15M67A-F1	AAGAAAGCATACGCGGCCAGCATGAAC
MED15M67A-F2	GTTCATGCTGGCCGCGTATGCTTTCTT
MED15S69A-F1	GAATACATGGACGCCATGAACGAAAAA
MED15S69A-F2	TTTTTCGTTCATGGCGTCCATGTATTC
MED15E72A-F1	GACAGCATGAACGCAAAAGTTGCTGTC
MED15E72A-F2	GACAGCAACTTTTGCGTTCATGCTGTC
MED15V76A-F1	GAAAAAGTTGCTGCCATGCGCAACACG
MED15V76A-F2	CGTGTTGCGCATGGCAGCAACTTTTTC
MED15M77A-F1	AAAGTTGCTGTCGCGCGCAACACGTAC
MED15M77A-F2	GTACGTGTTGCGCGCGACAGCAACTTT
MED15Y81A-F1	ATGCGCAACACGGCCAATACGAGGAAA
MED15Y81A-F2	TTTCCTCGTATTGGCCGTGTTGCGCAT
MED15R84A-F1	ACGTACAATACGGCGAAAAACGCCGTT
MED15R84A-F2	AACGGCGTTTTTCGCCGTATTGTACGT
MED15V76R-F1	GAAAAAGTTGCTGCGATGCGCAACACG
MED15V76R-F2	CGTGTTGCGCATCGCAGCAACTTTTTC
MED15R84R-F1	ACGTACAATACGAAGAAAAACGCCGTT
MED15R84R-F2	AACGGCGTTTTTCTTCGTATTGTACGT
eGFP expression
ARV1/eGFP-F1	GCTATGACCATGATTACGCCAAGCTTCATACGATAATATGGTTTCTATACTGT
ARV1/eGFP-F2	TCCTCGCCCTTGCTCACCATTCTAGACGTTTATCACAGCTAATATCTACTTT
GIT1/eGFP-F1	GCTATGACCATGATTACGCCAAGCTTAAGATTCGAGTCTGGGTGC
GIT1/eGFP-F2	TCCTCGCCCTTGCTCACCATTCTAGATTTTTATCCTATTCTATTTTTTTGAT
NFT1/eGFP-F1	GCTATGACCATGATTACGCCAAGCTTAATTTCCTCCGTCCTACAAGTA
NFT1/eGFP-F2	TCCTCGCCCTTGCTCACCATTCTAGATCCTAAAAATTAATACTGTGAATCG
Yeast two-hybrid assay
BD-MED15-F1	CGCCCGGGCCTCGAGCCCGGGTCGACATGTCTGCTGCTCCTGTCCAAG
BD-MED15-F2	TTCGCCCGGAATTAGCTTGGCTGCAGCTAAGTAGCACTTGTCCAATTATT
BD-MED15V76R/R84K-F1	CGCCCGGGCCTCGAGCCCGGGTCGACATGTCTGCTGCTCCTGTCCAAG
BD-MED15V76R/R84K-F2	TTCGCCCGGAATTAGCTTGGCTGCAGCTAAGTAGCACTTGTCCAATTATT
AD-HAP5-F1	TAGGATCCTCTGCTAGCAGAGAATTCATGACTGATAGGAATTTCTCACC
AD-HAP5-F2	CTCTAGAGTCGACTAATACTCTCGAGCTATTGTGGAAGAGGTCTTCTAG
AD-MGA2-F1	TAGGATCCTCTGCTAGCAGAGAATTCATGCAGCAGAACAGTGAGT
AD-MGA2-F2	CTCTAGAGTCGACTAATACTCTCGAGCTAACTGACAATTAAATCGTTCAACA
AD-AFT1-F1	TAGGATCCTCTGCTAGCAGAGAATTCATGGAAGGCTTCAATCCGG
AD-AFT1-F2	CTCTAGAGTCGACTAATACTCTCGAGCTAATCTTCTGGCTTCACATACT
Coimmunoprecipitation
pY26/PGPD-MED15-F1	GATTCTAGAACTAGTGGATCCATGTCTGCTGCTCCTGTCCAAG
pY26/PGPD-MED15-F2	GTCGACGGTATCGATAAGCTTCTATACCCATACGACGTCCCAGACTACGCTAGTAGCACTTGTCCAATTATTCC
pY26/PGPD-MED15V76R/R84K-F1	GATTCTAGAACTAGTGGATCCATGTCTGCTGCTCCTGTCCAAG
pY26/PGPD-MED15V76R/R84K-F2	GTCGACGGTATCGATAAGCTTCTATACCCATACGACGTCCCAGACTACGCTAGTAGCACTTGTCCAATTATTCC
pY26/PTEF-HAP5-F1	AAGTTTTCTAGAACTAGCGCGGCCGCATGACTGATAGGAATTTCTCACC
pY26/PTEF-HAP5-F2	GGCGAAGAATTGTTAATTAAAGATCTCAGATCCTCTTCAGAGATGAGTTTCTGCTCTTGTGGAAGAGGTCTTCTAGGC
pY26/PTEF-MGA2-F1	AAGTTTTCTAGAACTAGCGCGGCCGCATGCAGCAGAACAGTGAGT
pY26/PTEF-MGA2-F2	GGCGAAGAATTGTTAATTAAAGATCTCAGATCCTCTTCAGAGATGAGTTTCTGCTCACTGACAATTAAATCGTTCAACATTC
pY26/PTEF-AFT1-F1	AAGTTTTCTAGAACTAGCGCGGCCGCATGGAAGGCTTCAATCCGG
pY26/PTEF-AFT1-F2	GGCGAAGAATTGTTAATTAAAGATCTCAGATCCTCTTCAGAGATGAGTTTCTGCTCATCTTCTGGCTTCACATACTTCA

Yeast cells were cultivated in YPD medium (1% yeast extract, 2% tryptone, and 2% glucose) and YNB medium (0.67% yeast nitrogen base without amino acids, 2% glucose, containing histidine [20 mg/L], leucine [60 mg/L], lysine [200 mg/L], and uracil [20 mg/L]). Yeast cells were incubated at 30°C with shaking at 200 rpm.

Spot assay, growth, and viability analysis.

Log-phase cells were inoculated into liquid YNB medium supplemented with 0.05 to 0.60% (vol/vol) acetic acid, 0.5 to 6 mM H₂O₂, or 0.1 to 1.8 M NaCl with an initiation OD₆₀₀ of 0.1. The OD₆₀₀ values were recorded every 2 h until reaching stationary phase and used to draw a growth curve. The maximum specific growth rate was calculated for acetic acid (H₂O₂ or NaCl) concentration, yielding a half-maximal inhibitory concentration (IC₅₀). For the spot assay, cells were diluted in sterile water to an OD₆₀₀ of 1.0. Serial dilutions (10-fold) of 4 μL were spotted on YNB plates with or without 0.3% (vol/vol) acetic acid (4 mM H₂O₂ or 1.2 M NaCl) and then incubated at 30°C for 3 to 5 days. For viability analysis, cells were cultured in YNB medium with or without 0.3% (vol/vol) acetic acid (4 mM H₂O₂ or 1.2 M NaCl) for 8 h, plated on YNB medium, and then incubated at 30°C for 3 to 5 days. The colonies were counted to analyze cell viability.

Measurement of intracellular reactive oxygen species.

The intracellular ROS level was measured using the cell permeable probe (2′,7′-dichloro-dihydro-fluorescein diacetate [DCFH-DA]; Sigma-Aldrich) (68). In brief, after incubation in YNB medium with or without 0.3% (vol/vol) acetic acid (4 mM H₂O₂ or 1.2 M NaCl) for 2 h, the cells were collected, diluted in phosphate-buffered saline (PBS) to an OD₆₀₀ of 1.0, and then treated with dithiothreitol and snailase to weaken the cell wall. Then, 10 μL of 1 mM DCFH-DA was added to 1-mL samples and subsequently mixed and incubated at 37°C for 40 min. The cells were washed three times to remove excess probe and then resuspended in PBS. The fluorescence intensity was measured at an excitation wavelength of 488 nm and an emission wavelength of 525 nm using flow cytometry (FACSAria; Becton Dickinson, USA).

Total RNA extraction and RNA-seq.

Log-phase cells were incubated in YNB medium with or without 0.3% (vol/vol) acetic acid (4 mM H₂O₂ or 1.2 M NaCl) for 8 h. Total RNA was isolated using a MiniBEST universal RNA extraction kit (TaKaRa Bio, Shiga, Japan). The concentration and quality of total RNA were determined by microspectrophotometry using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA). Frozen samples were sent to the Majorbio Institute for global transcriptome analysis.

qRT-PCR.

Total RNA extraction was carried out as described above. First, 1 μg of total RNA was taken to synthesize cDNA by using a PrimeScript II 1st-Strand cDNA synthesis kit (TaKaRa). The cDNA mixture was diluted to about 100 ng/μL and then used as the template for analyzing the gene expression level by quantitative reverse transcription-PCR (qRT-PCR) with SYBR Premix Ex Taq (TaKaRa, Japan) using an iQ5 continuous fluorescence detector system (Bio-Rad, Hercules, CA). Data were normalized to the β-actin gene ACT1. The primer sequences used for the qRT-PCR are listed in Table 3.

Yeast two-hybrid assays.

All Y2H plasmids were based on either pGBKT7 (Gal4BD) as the DNA-binding domain (BD) plasmid or pGADT7 (Gal4AD) as the transcription-activating domain (AD) plasmid. The Gal4AD and Gal4BD plasmids (pGBKT7-MED15, pGBKT7-MED15^V76R/R84K, pGADT7-HAP5, pGADT7-MGA2, and pGADT7-AFT1) were cotransformed into the yeast AH109 reporter strain. Positive clones were selected, grown in YNB medium, and then spotted onto synthetic dextrose (SD)-Leu-Trp plates and SD-Leu-Trp-His selective plates with the histidine biosynthesis inhibitor 1,2,4-aminotrizole (3-AT) at 30°C for 3 to 5 days.

Sterol and phospholipid measurement.

Log-phase cells were incubated in YNB medium with or without 0.3% (vol/vol) acetic acid (4 mM H₂O₂ or 1.2 M NaCl) for 8 h and then collected and washed with PBS. Membrane sterols were extracted by a modified saponification method (13) and then analyzed by gas chromatography-mass spectrometry (69). The phospholipids were extracted as described previously (55) and then analyzed by electrospray ionization mass spectrometry (35).

Cell membrane integrity, fluidity, and permeability analysis.

Log-phase cells were incubated in YNB medium with or without 0.3% (vol/vol) acetic acid (4 mM H₂O₂ or 1.2 M NaCl) for 8 h and then collected, washed, and resuspended with PBS. For membrane integrity and permeability analysis, samples were incubated with SYTOX (Sigma-Aldrich) at 30°C in dark for 5 min and then used for flow cytometry analysis (55). SYTOX green is a non-cell-membrane-permeable fluorescent dye that can only pass through the disordered area of the cell membrane and stain dead cells. The membrane integrity decreases when the proportion of the SYTOX-stained cells increases. Samples were also incubated with FM4-64 dye and SYTOX dye for laser scanning confocal microscopy analysis (LSCM). For membrane fluidity analysis, samples were incubated with 1 μL of 1 mmol/L 1,6-diphenyl-1,3,5-hexatriene (DPH; Sigma-Aldrich). The fluorescence intensity was measured by using a spectrofluorimeter (Photon Technology International, USA) with excitation at 360 nm and emission at 450 nm. The fluorescence anisotropy value (r) was calculated as described previously (2). Fluorescence anisotropy values and the membrane fluidity showed a negative correlation (13).

Measurement of intracellular Na⁺ and K⁺ concentrations.

Log-phase cells were incubated in YNB medium with or without 1.2 M NaCl. First, 50-mL samples were collected every 2 h and then washed twice in ice-cold 10 mM MgCl₂, 10 mM CaCl₂, and 1 mM HEPES buffer. They were then resuspended in the same buffer. The intracellular Na⁺ and K⁺ contents were measured using a flame-graphite furnace atomic absorption spectrometer, as described previously (70).

Nft1 activity analysis.

Log-phase cells expressing Nft1-Myc were grown in YNB medium with or without 1.2 M NaCl for 4 h and collected as described above. Vacuolar membranes were purified by Ficoll density gradient centrifugation, as described previously (71). The protein Nft1-HA were purified by incubation with anti-myc-conjugated magnetic beads (Bio-Rad) (72). The Nft1 activity was determined according to the amount of inorganic phosphate (P_i) released from ATP at 750 nm in a reaction mixture, as described previously (13). The ATPase activity is expressed as micromoles of P_i released per minute per milligram of the total membrane protein.

Pyruvate production and analysis.

The strains were cultivated at 30°C in flasks or bioreactor containing medium A containing (per L) the following: 80 g of glucose, 3 g of KH₂PO₄, 1.0 g of MgSO₄·7H₂O, 3 g of sodium acetate, 18 μg of thiamine-HCl, 4.0 μg of biotin, 40 μg pyridoxine-HCl, and 0.8 μg of nicotinic acid with an initial OD₆₀₀ biomass of 1.0. Fermentation was performed at 30°C. An HPLC system (Dionex UltiMate 3000 Series; Thermo Scientific, Waltham, MA) was used to determine the concentrations of glucose and organic acid. This involved an Aminex HPX-87H column (7.8 × 300 mm; Bio-Rad Laboratories, Inc., Hercules, CA) at 35°C with 0.05 mM sulfuric acid as the mobile phase. The injection volume was 10 μL, and the flow rate was 0.6 mL/min.