Thioredoxin Deficiency Causes the Constitutive Activation of Yap1, an AP-1-like Transcription Factor in Saccharomyces cerevisiae *

Yap1 is a transcription factor that responds to oxidative stress in Saccharomyces cerevisiae. The activity of Yap1 is regulated at the level of its intracellular localization, and a cysteine-rich domain at the C terminus of Yap1 is involved in this regulation. We investigated the effects of redox-regulatory proteins, thioredoxin and glutaredoxin, on the regulation of Yap1, using the deficient mutants of these thiol-disulfide oxidoreductases. In the thioredoxin-deficient mutant (trx1Δ/trx2Δ), Yap1 was constitutively concentrated in the nucleus and the level of expression of the Yap1 target genes was high under normal conditions, while this was not the case for the glutaredoxin-deficient mutant (grx1Δ/grx2Δ). No distinct difference was observed in the levels of Yap1 protein between the wild type andtrx1Δ/trx2Δ. The constitutive activation of Yap1 in trxΔ/trx2Δ was observed under aerobic conditions but not under anaerobic conditions. These findings suggest that thioredoxin has negative effects on this regulation via the redox states. We also show the synthetic lethality betweenyap1Δ and trx1Δ/trx2Δ mutation, but theyap1Δ/grx1Δ/grx2Δ triple mutant was viable, suggesting a difference of the functions between thioredoxin and glutaredoxin and a genetic interaction between Yap1 and thioredoxin in vivo.

Yap1 is a transcription factor that responds to oxidative stress in Saccharomyces cerevisiae. The activity of Yap1 is regulated at the level of its intracellular localization, and a cysteine-rich domain at the C terminus of Yap1 is involved in this regulation. We investigated the effects of redox-regulatory proteins, thioredoxin and glutaredoxin, on the regulation of Yap1, using the deficient mutants of these thiol-disulfide oxidoreductases. In the thioredoxin-deficient mutant (trx1⌬/trx2⌬), Yap1 was constitutively concentrated in the nucleus and the level of expression of the Yap1 target genes was high under normal conditions, while this was not the case for the glutaredoxin-deficient mutant (grx1⌬/grx2⌬). No distinct difference was observed in the levels of Yap1 protein between the wild type and trx1⌬/trx2⌬. The constitutive activation of Yap1 in trx⌬/trx2⌬ was observed under aerobic conditions but not under anaerobic conditions. These findings suggest that thioredoxin has negative effects on this regulation via the redox states. We also show the synthetic lethality between yap1⌬ and trx1⌬/trx2⌬ mutation, but the yap1⌬/grx1⌬/grx2⌬ triple mutant was viable, suggesting a difference of the functions between thioredoxin and glutaredoxin and a genetic interaction between Yap1 and thioredoxin in vivo.
Many aerobic organisms show adaptive responses to oxidative stress by increasing the levels of antioxidant enzymes. A portion of the adaptive response is regulated at the transcriptional level, and several transcription factors that regulate the expression of antioxidant genes have been reported. In Escherichia coli, OxyR, SoxR, and SoxS are key transcription factors for the adaptive response to oxidative stress. OxyR regulates the expression of genes encoding H 2 O 2 -inducible proteins, and SoxR and SoxS regulate the expression of genes encoding superoxide-inducible proteins (1). In mammals, two transcription factors, NF-B and AP-1, have been strongly implicated in the oxidative stress response (2). The activities of these transcription factors are reversibly controlled through redox states and modulated by thiol-disulfide oxidoreductases such as thioredoxin (TRX) 1 and glutaredoxin (GRX) (2,3). OxyR is activated through the formation of a disulfide bond between Cys 199 and Cys 208 , and deactivated by the enzymatic reduction of the bond with GRX (3). The activities of NF-B and AP-1 are also regulated by the redox modification of cysteine residues. After the dissociation of NF-B/IB complex, TRX augments the DNA binding and transcriptional activities of NF-B by reducing the Cys 62 residue in its DNA-binding loop (4 -6). Similarly, redox modification of AP-1 is regulated by a nuclear redox factor, Ref-1, and the Ref-1 activity is also modulated by TRX (7,8). TRX can associate directly with Ref-1 in the nucleus (9,10).
The contribution of TRX and GRX toward the maintenance of the intracellular environments in reduced states is comparable to glutathione (11,12). These thiol-disulfide oxidoreductases may participate in the control of redox-regulated transcription factors via the intracellular redox states (3,13).
Yap1 (yeast AP-1) is a transcription factor crucial for oxidative stress response in Saccharomyces cerevisiae, and it regulates the expression of several genes whose gene products play major roles in the oxidative stress tolerance. The null mutant of the YAP1 gene displayed hypersensitivity to oxidative stress, and the overexpression of the YAP1 gene confers stress resistance (14 -18). Several target genes for Yap1 have been identified, such as GSH1 encoding ␥-glutamylcysteine synthetase (19), GLR1 encoding glutathione reductase (20), TRX2 encoding TRX (21), and TRR1 encoding thioredoxin reductase (22). These gene products are involved in the adaptive response to oxidative stress.
Yap1 was originally identified as a functional homologue of mammalian AP-1 on the basis of its ability to bind to an AP-1 recognition element (23). Yap1 binds to the Yap1 recognition element (YRE: 5Ј-TTAGT(C/A)A-3Ј) in the promoter region of target genes (24). The N terminus of Yap1 contains a bZip domain, which is conserved among the AP-1 family, including mammalian Jun, Fos, and S. cerevisiae Gcn4 (23). A cysteinerich domain (CRD) at the C terminus of Yap1, containing three Cys-Ser-Glu sequence motifs, plays an important role in the control of Yap1, especially for the intracellular localization of Yap1 (25,26). In the response to oxidative stress, the localization of Yap1 changes dramatically, while the increase in the DNA binding activity is modest and the levels of Yap1 do not increase (26 -28). Yap1 exists both in the cytoplasm and the nucleus under non-stressed conditions, while it is concentrated in the nucleus under oxidative conditions (26). Recently, it has been reported that the localization of Yap1 is controlled by Crm1-mediated nuclear export and that Yap1 has an nuclear export sequence (NES) embedded within the CRD (29,30). Mutational analysis suggested that cysteine residues within the CRD serve as redox sensors, which regulate the availability of the NES (30). Therefore, removal of the CRD causes constitutive accumulation of Yap1 in the nucleus, which results in the increase of transcription of the Yap1 target genes (26). Additionally, the CRD is required for Yap1 to discriminate the stresses elicited by H 2 O 2 , diamide, and CdCl 2 (27,28). * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The importance of the CRD raised the possibility that TRX and/or GRX modulate the activity of Yap1 by the redox modification of cysteine residues, like OxyR, NF-B, and AP-1. To examine whether TRX and/or GRX attend the redox regulation of Yap1, we analyzed the phenotypes of the TRX-and GRXdeficient mutants with respect to the Yap1-mediated gene expression and the localization of Yap1. Here we show that TRX deficiency (trx1⌬/trx2⌬) caused the constitutive activation of Yap1 under aerobic conditions but not under anaerobic conditions, and that GRX deficiency (grx1⌬/grx2⌬) did not affect the Yap1 activity. We also provide the evidence that the yap1⌬/ trx1⌬/trx2⌬ triple mutant shows the synthetic lethality while the yap1⌬/grx1⌬/grx2⌬ triple mutant was viable.

EXPERIMENTAL PROCEDURES
Yeast Strains and Medium-S. cerevisiae YPH250 and YPH252 were obtained from the Yeast Genetic Stock Center, University of California, Berkeley, CA. Cells were cultured in 50 ml of SD minimal medium (2% glucose, 0.67% yeast nitrogen base without amino acids, pH 5.5) with appropriate amino acids and bases at 30°C with reciprocal shaking in 200-ml Erlenmeyer flasks. For anaerobic cultivation, the medium was flushed with nitrogen gas, sealed up, and incubated without shaking at 30°C. Exponentially growing cells were harvested at A 610 ϭ 0.5.
S. cerevisiae has two cytosolic TRX genes, TRX1 and TRX2 (31); one mitochondrial TRX gene, TRX3 (32); and two GRX genes, GRX1 and GRX2 (TTR1) (33,34). To disrupt the TRX1 gene, the TRX1 gene was amplified using the following oligonucleotide primers: 5Ј-GATCAGAA-TGATTGAAATCA-3Ј and 5Ј-GACGAGCTATAGGATGATGA-3Ј. The amplicon (2.0 kb) was treated by Klenow fragment, and then cloned, respectively, to the HincII site of pUC19 (pTR-1) and to the BamHI site of YEp13, which was also treated by Klenow fragment (YEpTRX1). The URA3 gene (1.2 kb) isolated from YEp24 was treated with Klenow fragment and inserted between the MunI/MunI sites internal to the TRX1 gene in pTR-1, which was also treated by Klenow fragment, to yield pTRD1. pTRD1 was digested with ApaLI and DraI, and the trx1⌬::URA3 fragment was transformed to YPH250 and YA-1␣ to yield the trx1⌬ and YT-1␣, respectively. Disruption of the TRX2 gene was done using the trx2::HIS3 disruption plasmid, kindly provided by Drs. S. Kuge and N. Jones (21). The trx2::HIS3 disruption plasmid was digested with SphI, and the trx2⌬::HIS3 fragment was transformed to YPH250 and the trx1⌬ mutant to yield the trx2⌬ mutant and the trx1⌬/trx2⌬ mutant, respectively. Additionally, the trx2::LEU2 disruption plasmid was constructed. The TRX2 gene was amplified by using the following oligonucleotide primers: 5Ј-GATCAGCATAACTTGAGTG-C-3Ј and 5Ј-GATCGCATGGAACGCCAAGC-3Ј. The amplicon (0.8 kb) was treated with Klenow fragment, and then cloned to the HincII site of pUC19 (pTR-2), and to the BamHI site of YEp13 (YEpTRX2), respectively. The LEU2 gene (2.2 kb) isolated from YEp13 was treated with Klenow fragment and inserted between the HincII/EcoO65I sites internal to the TRX2 gene in pTR-2, which was also treated with Klenow fragment, to yield pTRD2. pTRD2 was digested with PstI and SmaI, and the trx2⌬::LEU2 fragment was transformed to the yap1⌬ mutant to yield YT-2a.
The amplicon (1.2 kb) was digested with BamHI and cloned to the BamHI site of YEp13 (YEpGRX2), and also to the BamHI site of pUC19 (pGR-2). The URA3 gene (1.2 kb) isolated from YEp24 was treated with Klenow fragment and inserted between the BsaAI/BsaAI sites internal to the GRX2 gene (pGRD2). pGRD2 was digested with SmaI and SacI, and the grx2⌬::URA3 fragment was transformed to YPH250 and the grx1⌬ to yield the grx2⌬ mutant and the grx1⌬/grx2⌬ mutant, respectively. Disruption of each gene was verified by polymerase chain reaction. The trx1⌬/trx2⌬ mutant showed the methionine auxotrophy, the increase of cell size, and elongation of generation time as reported by Muller (35).
Construction of the GSH1-lacZ Fusion Gene-The GSH1 promoter fragment containing the region from Ϫ800 to ϩ33 (ϩ1 representing the start of translation) was generated by polymerase chain reaction using the primers: 5Ј-TAATCTTATGAATCCCGGGGATTTTATCGG-3Ј and 5Ј-CTAGACTCAAACCCGGGCAAAGGCGTGCCC-3Ј. The amplicon was digested with SmaI and cloned into the SmaI site of pMC1871 containing the coding region of lacZ (Amersham Pharmacia Biotech), to yield pMC-GSH1lacZ. As a result of this construction, the first 11 amino acid residues of Gsh1 were fused to ␤-galactosidase whose first 8 amino acids were deleted. The GSH1-lacZ in pMC-GSH1lacZ was isolated by digestion with SalI and cloned into the SalI site of pRS414 (pRS-GSH1lacZ). A Yap1-dependent lacZ reporter gene containing three SV40 AP-1 sites and the TATA element of the CYC1 promoter was kindly gifted by Dr. Kuge (21,26,29). ␤-Galactosidase activity was measured as described by Miller (36). One unit of the activity was defined as the amount of enzyme increasing A 420 per hour at 30°C. Protein was determined by the method of Lowry et al. (37).
Northern Blotting Analysis-Northern hybridization was performed using 25 g of total cellular RNA isolated from yeast cells by the method of Schmitt et al. (38). The probes were generated by random primed labeling of the 0.4-kb EcoRV-BamHI fragment of GSH1 gene, 0.7-kb PvuII-HincII fragment of TRR1 gene, and 1.1-kb EcoRI-PstI fragment of ACT1 gene, respectively, with [␣-32 P]dCTP using a kit (Random Primer DNA labeling kit version 2, Takara).
Production of Anti-TRX Antibody-A NdeI/BamHI fragment encoding the whole coding region of the TRX2 gene was inserted into the NdeI/BamHI site of pET15-b (Novagen) to construct pET-TRX2. The resulting protein contained six histidine tag residues fused in-frame with the TRX2 coding region. pET-TRX2 was transferred into E. coli strain BL21(DE3). Histidine-tagged Trx2 was purified using histidine affinity column chromatography (His-trap, Amersham Pharmacia Biotech), followed by gel filtration chromatography (Superdex 75, Amersham Pharmacia Biotech). Purified fusion protein was then used for the production of anti-yeast TRX antibody using New Zealand White rabbits. Immunization and purification of anti-Trx antibody were accomplished by Sawady Technology Co., Ltd., Tokyo, Japan. This anti-TRX antibody was specific to yeast TRX and was able to detect both Trx1 and Trx2.
Immunofluorescence Technique-Immunofluorescence microscopic observation of yeast was performed by the methods of Rose et al. (39). Anti-rabbit IgG (HϩL)-fluorescein isothiocyanate (FI-1000, Vector Laboratory Inc.) was used as the secondary antibody.
Western Blotting Analysis-Cell extracts were prepared according to Wemmie et al. (28), and 170 g of total protein of each sample was run on 10% SDS-polyacrylamide gels. Proteins were electrically transferred to polyvinylidene difluoride membrane (Immobilon, Millipore). Anti-Yap1 antiserum raised in rabbit was kindly gifted by Dr. Moye-Rowley. Horseradish peroxidase-conjugated secondary antibody (Jackson Immunoresearch Laboratory, Inc.) and diaminobenzidine were used to visualize immunoreactive protein.
Glutathione Reductase, Total Glutathione, and H 2 O 2 Stress Treatment-The activity of glutathione reductase, total glutathione, and the susceptibility and adaptation to H 2 O 2 were determined as described previously (40,41).
Measurement of Intracellular Oxidation-Intracellular oxidation level of yeast was measured using the oxidant-sensitive probe 2Ј,7Јdichlorofluorescin diacetate (DCFH-DA) purchased from Molecular Probes (42). Cells growing exponentially in SD medium were collected and resuspended in fresh SD medium containing 0.1 mM DCFH-DA and incubated at 30°C for 20 min. After the incubation, cells were washed, resuspended in distilled water, and disrupted by vortexing with glass beads. Cell extracts (70 l) were mixed in 500 l of distilled water, and fluorescence was measured with ex ϭ 490 nm and em ϭ 524 nm using a Hitachi F-3000 spectrofluorometer.
Genetic Manipulation of Yeast-Mating, sporulation, dissection, and tetrad analysis of yeast were done as described by Rose et al. (39).

RESULTS
The TRX-deficient Mutant Shows the Constitutive Activation of Yap1-In order to see the effect of TRX and GRX on the activity of Yap1, we investigated the levels of cellular glutathione and glutathione reductase (GR) activity in the TRX-and GRX-deficient mutants. The GSH1 and GLR1 genes, which are target genes for Yap1 (19,20), encode ␥-glutamylcysteine synthetase (Gsh1) and GR, respectively (43,44). The Gsh1 catalyzes the first and rate-limiting step of glutathione synthesis (45). Therefore, it is conceivable that the levels of intracellular glutathione and GR activity reflect the activity of Yap1. Table I shows the levels of total glutathione and GR activity in various mutants. As reported previously, the yap1⌬ mutant showed lower levels of total glutathione and GR activity than the wild type (19,20). Total glutathione and GR activity in the wild type were increased by the treatment with 0.2 mM H 2 O 2 , but not in the yap1⌬ mutant (46). The levels of glutathione and GR activity in trx1⌬/trx2⌬ were constitutively high under nonstressed conditions, and the values were higher than those in the wild type which was treated with H 2 O 2 . This indicates that the loss of both Trx1 and Trx2 affects intracellular glutathione metabolism. GRX deficiency (grx1⌬/grx2⌬) and single-gene mutation of TRX or GRX genes did not affect total glutathione levels and GR activity (data for single-gene mutants are not shown), and our results are consistent with the previous reports (34,47). The results in Table I imply that Yap1 is constitutively activated in trx1⌬/trx2⌬ even though the cells are not exposed to oxidative stress.
To confirm the constitutive activation of Yap1 in the TRXdeficient mutant, we next assessed the expression of the lacZ reporter gene under the control of the GSH1 promoter. A fusion gene was constructed containing the lacZ coding region fused to the GSH1 promoter sequence which contains the Yap1 binding site (5Ј-TTAGTCA-3Ј) (19). As shown in Fig. 1A, ␤-galactosidase activity was increased with H 2 O 2 treatment in the wild type, but not in the yap1⌬, as reported previously (48). The ␤-galactosidase activities in single mutants (trx1⌬ or trx2⌬) under non-stressed conditions were the same to that in the wild type. On the contrary, the basal level of ␤-galactosidase activity in trx1⌬/trx2⌬ was significantly higher than that of the wild type, and the value was almost the same as that of the wild type treated with 0. 2  We reconfirmed the increased expression of the GSH1 gene in trx1⌬/trx2⌬ by Northern blotting analysis (Fig. 1B). Basal level of the GSH1 gene expression in trx1⌬/trx2⌬ was almost the same as that of the H 2 O 2 -treated wild type. Expression of the GSH1 gene in trx1⌬/trx2⌬ did not increase with 0.2 mM H 2 O 2 stress. These results were in good agreement with those obtained by using the GSH1-lacZ reporter gene assay. Expression of the TRR1 gene, another target gene for Yap1 (22), was also constitutively high in trx1⌬/trx2⌬, but not in grx1⌬/grx2⌬ (Fig. 1C).
In yeast and other eukaryotic cells, expression of most genes is controlled by multiple transcription factors acting co-operatively (49). To see whether other factor(s) that may bind to the GSH1 promoter region are involved in increased expression of the GSH1-lacZ reporter gene in the trx1⌬/trx2⌬ mutant, we also examined the expression of a lacZ reporter gene driven by three SV40 AP-1 sites and the TATA element of the CYC1 promoter (21,26,29). This reporter gene was proven to be solely regulated by Yap1 (21, 26, 29). Expression of the Yap1dependent lacZ reporter gene was also constitutively induced  1. Effect of the TRX deficiency on the expression of Yap1 target genes. Cells at exponential phase were harvested, resuspended in fresh SD medium, and treated with or without 0.2 mM H 2 O 2 for 60 min. Cell extracts were prepared and ␤-galactosidase activity was assayed (A and D). Data are means Ϯ S.D. from three independent experiments. A, expression level of the GSH1-lacZ reporter gene. B and C, Cells were treated with or without 0.2 mM H 2 O 2 for 30 min. Total cellular RNA was prepared, and 25 g of RNA was applied to each slot. D, expression level of the lacZ reporter gene containing AP-1-binding site followed by TATA box of the CYC1 gene.
in trx1⌬/trx2⌬ (Fig. 1D), as well as the GSH1-lacZ fusion. This suggests that Yap1 is the main factor for the increased expression of Yap1 target genes in the trx1⌬/trx2⌬ mutant. These experiments showed that the deficiency of TRX (trx1⌬/trx2⌬) resulted in constitutive acceleration of the expression of Yap1 target genes even under non-stressed conditions. Yap1 Is Constitutively Localized in the Nucleus in trx1⌬/ trx2⌬-To determine whether constitutive high expression of Yap1 target genes is caused by the increase of Yap1 protein or not, we compared the amounts of Yap1 in the wild type and trx1⌬/trx2⌬ under non-stressed conditions by Western blotting analysis. The Yap1-specific bands were detected at similar intensities both in the wild type and the trx1⌬/trx2⌬ mutant (Fig. 2). With the consideration of previous reports that the levels of Yap1 do not change during oxidative stress (27,28), the constitutive high activity of Yap1 in trx1⌬/trx2⌬ under non-stressed conditions can be attributed to a post-translational modification of Yap1, where TRX is likely to be involved.
It is reported that the Yap1 activity is regulated at the level of its localization. Yap1 exists both in the cytoplasm and in the nucleus under non-stressed conditions, while it is concentrated in the nucleus by oxidative stress (26). We reconfirmed this by using the GFP-Yap1 fusion in the wild type (Fig. 3). In a single mutant of trx1⌬ or trx2⌬, Yap1 behaved similarly to that in the wild type (data not shown). On the other hand, Yap1 was constitutively localized in the nucleus in trx1⌬/trx2⌬ not only under oxidative-stressed conditions but also under nonstressed conditions. Highly constitutive expression of the Yap1 target genes in trx1⌬/trx2⌬ was most likely caused by the constitutive localization of Yap1 in the nucleus.
High Activity of Yap1 in trx1⌬/trx2⌬ Was Repressed under Anaerobic Conditions-It has been reported that cysteine residues are important in the redox-regulation of AP-1, NF-B, and OxyR (2, 3). Yap1 also contains a CRD at the C terminus, and it was reported that cysteine residue(s) within the CRD is necessary to mediate the response to oxidative stress (26 -28, 30). Thus, Yap1 also might be regulated via the redox modification of cysteine residue(s) and TRX may participate in this redox-regulation. To assess this, we investigated the activity of Yap1 in the trx1⌬/trx2⌬ mutant under anaerobic conditions. ␤-Galactosidase activity derived from GSH1-lacZ in the wild type was virtually the same under both anaerobic and aerobic conditions without oxidative stress. On the other hand, ␤-galactosidase activity in trx1⌬/trx2⌬ was markedly repressed under anaerobic conditions to the level similar to the wild type (Fig. 4A). Furthermore, this repression was restored by the transfer of the cells to the aerobic conditions. Similar results were observed using the Yap1-dependent lacZ reporter gene (Fig. 4B). These results strongly suggest that redox states in the cells reflect the activity of Yap1 in trx1⌬/trx2⌬.
Additionally, we measured the levels of intracellular oxidation in trx1⌬/trx2⌬ using the oxidant-sensitive probe DCFH-DA. The level of intracellular oxidation in trx1⌬/trx2⌬ was more than 2-fold higher than that in the wild type, while that in grx1⌬/grx2⌬ was only 1.1-fold higher (Fig. 4C). These results indicate that the intracellular environments of trx1⌬/ trx2⌬ were in more oxidized states than those of the wild type. Redox states of the intracellular environments and the redox regulation of transcription factors are suggested to be well linked (3,13). Thus, the oxidative environments in trx1⌬/trx2⌬ may also account for the constitutive activation of Yap1.
GRX Does Not Substitute for TRX-The constitutive activation of Yap1 was observed only in the TRX-deficient mutant (trx1⌬/trx2⌬), but not in the GRX-deficient mutant (grx1⌬/ grx2⌬). However, in E. coli, the thioredoxin system and the glutaredoxin system can partially substitute for each other in vivo (11,12). Overexpression of the grxA gene in a trxA mutant can substitute for thioredoxin 1 in E. coli (50). We examined whether or not the overexpression of GRX gene in trx1⌬/trx2⌬ can repress the constitutive activation of Yap1. The Yap1 activity in trx1⌬/trx2⌬ was reduced to the basal level of the wild type by overexpression of either the TRX1 or TRX2 gene alone, i.e. the levels of total glutathione, GR activity, and the expression of GSH1-lacZ were reduced to those of the wild type ( Fig.  5; data of total glutathione and GR activity are not shown). On the other hand, overexpression of the GRX1 or GRX2 gene did not affect the Yap1 activity in trx1⌬/trx2⌬. Therefore, GRX was thought to be unable to substitute for TRX in the regulation of Yap1 activity in vivo. Overexpression of the TRX3 gene encoding mitochondrial TRX also did not affect the Yap1 activity in trx1⌬/trx2⌬, suggesting the difference of functions between cytosolic TRX and mitochondrial TRX.
Interaction between TRX and Yap1-The constitutive activation of Yap1 in trx1⌬/trx2⌬ suggests that TRX may regulate the Yap1 activity negatively. To see whether TRX associates with Yap1 directly or not, we first investigated the intracellular localization of TRX by immunofluorescence microscopy using anti-TRX antibody. If TRX interacts with Yap1 directly to repress the Yap1 activity, the intracellular localization of TRX is expected to be similar to that of Yap1. Yeast TRX would be able to enter or leave the nucleus by diffusion through the nuclear pore complex on account of its molecular size (molecular masses of both Trx1 and Trx2 are approximately 11 kDa) (51). TRX was observed predominantly in the cytoplasm under non-stressed conditions (Fig. 6). After the treatment of the wild type cells with a mild oxidative stress (0.2 mM H 2 O 2 ), TRX was observed both in the cytoplasm and in the nucleus, suggesting the possibility that TRX can associate with Yap1 both in the nucleus and in the cytoplasm. The increase of the levels of TRX protein by H 2 O 2 (0.2 mM) treatment was also confirmed by Western blotting analysis (data not shown). In the case of HeLa cells, TRX is translocated from the cytoplasm to the nucleus after the stress treatment and concentrated in the nucleus for a long time (10). In S. cerevisiae, however, the concentration of TRX in the nucleus was not observed under oxidative conditions.
Next, we performed two-hybrid assay experiments to examine the direct association of Yap1 with Trx1 or Trx2 (Matchmaker Two-Hybrid System 2, CLONTECH). However, we could not detect the direct association between Yap1 and Trx1 or Trx2. This might be because their association is too weak to detect by this assay system, or because TRX associates with Yap1 via other factors, such as Ref-1 in the case of AP-1 regulation (10).
Synthetic Lethality of yap1 Null Allele and trx1⌬/trx2⌬-Because the expression of the Yap1 target genes are enhanced in the trx1⌬/trx2⌬ mutant, which resulted in the increase of the activity of antioxidant enzymes, we suspected that the trx1⌬/trx2⌬ mutant might acquire the tolerance to oxidative FIG. 2. The levels of Yap1 protein are unaffected by the TRX deficiency. Cells were cultured until exponential phase without stress treatment, and cell extracts were prepared. An equal amount of protein (170 g) from the wild type, trx1⌬/trx2⌬, and yap1⌬ was subjected to SDS-PAGE (10% polyacrylamide gel). Western blotting analysis was carried out using anti-Yap1 antiserum. Lane 1, wild type; lane 2, trx1⌬/trx2⌬; lane 3, yap1⌬. stress. We investigated the susceptibility to H 2 O 2 (0.2-5 mM) and the induction of adaptation to H 2 O 2 stress in trx1⌬/trx2⌬. The trx1⌬/trx2⌬ mutant showed great susceptibilities to H 2 O 2 despite the increase of glutathione and antioxidant enzymes (Fig. 7A). Furthermore, the trx1⌬/trx2⌬ mutant was unable to induce the adaptation to H 2 O 2 despite of the constitutive activation of Yap1 (Fig. 7B). The wild type induced a large increase in tolerance to 2 mM H 2 O 2 by pretreatment with a sublethal dose of H 2 O 2 (0.2 mM, 60 min), while the trx1⌬/trx2⌬ mutant did not. These results indicate that the constitutive activation of Yap1 may be essential for the trx1⌬/trx2⌬ mutant to survive under aerobic conditions rather than to acquire the resistance against oxidative stress.
In this study, we tried to construct a yap1⌬/trx1⌬/trx2⌬ triple mutant to see the genetic interaction of these genes. We tried several times to disrupt the TRX1 gene in YT-2a ( MATa   FIG. 3. Yap1 is constitutively localized in the nucleus in the trx1⌬/trx2⌬ mutant. GFP fluorescence was visualized in living yeast cells carrying pRS cp-GFP-YAP1 (25). Cells at exponential phase were harvested, resuspended in fresh SD medium, and treated with or without 0.2 mM H 2 O 2 . Cells were stained with 1 g/ml 4Ј,6Ј-diamidino-2-phenylindole dihydrochloride (DAPI) to visualize DNA. trx1⌬/trx2⌬. A and B, wild type and trx1⌬/trx2⌬ cells were cultured until exponential phase without shaking under anaerobic conditions. Cells were then harvested and resuspended in fresh SD medium and incubated with shaking under aerobic conditions for another 3 or 6 h at 30°C. Cell extracts were prepared, and ␤-galactosidase activity was assayed as described under "Experimental Procedures." For control experiments, cells were grown with shaking under aerobic conditions. A, expression of the GSH1-lacZ reporter gene. B, expression level of the lacZ reporter gene containing AP-1-binding site followed by TATA box of the CYC1 gene. Yeast strains were as follows: empty bar, wild type; filled bar, trx1⌬/trx2⌬. C, cells at exponential phase were harvested and resuspended in fresh SD medium containing 0.1 mM DCFH-DA for 20 min at 30°C, and then cell extracts were prepared. yap1⌬::HIS3 TRX1 trx2⌬::LEU2), but no viable transformant was obtained. In contrast, it was possible to construct the yap1⌬/grx1⌬/grx2⌬ mutant. To confirm the lethality of this triple mutant (yap1⌬/trx1⌬/trx2⌬), we constructed a diploid strain YT-21 (MATa/MAT␣ yap1⌬::HIS3/yap1⌬::HIS3 TRX1/trx1⌬::URA3 trx2⌬::LEU2/TRX2) for tetrad analysis, by crossing YT-1␣ (MAT␣ yap1⌬::HIS3 trx1⌬::URA3 TRX2) and YT-2a. Thirty-seven asci were subjected to tetrad analysis and all spores (37 ϫ 4 ϭ 148) were incubated on YPD plates at 30°C for 2 days. Thirty-eight spores among 148 spores did not germinate. The genotype of each germinated strain (148 Ϫ 38 ϭ 110 strains) was analyzed to determine the genotypes of the 38 spores that did not germinate. The TRX1-TRX2 ascus type showed random assortment (parental ditype:nonparental ditype:tetratype was 6:7:24, or approximately 1:1:4). This result seems reasonable because the TRX1 and TRX2 genes locate on the different chromosomes (TRX1 on chromosome XII and TRX2 on chromosome VII). The genotypes of all 38 spores were inferred to be triple mutant (yap1⌬/trx1⌬/trx2⌬). Therefore, we concluded that the combination of the yap1⌬ mutation with the trx1⌬/trx2⌬ mutation was lethal at least under aerobic conditions. This result also strongly indicates that the constitutive activation of Yap1 is essential for the trx1⌬/trx2⌬ mutant to survive under aerobic conditions. DISCUSSION TRX in the Regulation of Yap1-In this study, we demonstrated that Yap1 is constitutively activated and concentrated in the nucleus in trx1⌬/trx2⌬ under non-stressed conditions.

FIG. 4. The activity of Yap1 is repressed under anaerobic conditions in
No difference was observed in the levels of Yap1 protein between the wild type and trx1⌬/trx2⌬, indicating that the constitutive activation of Yap1 was due to its post-translational modification. Additionally, we also indicated the possibility that TRX negatively controls the activity of Yap1 via redox states because the constitutive activation was observed only under aerobic conditions. It was clarified that the localization of Yap1 (and the activity of Yap1) is regulated by the nuclear export of Yap1 mediated by the interaction between Crm1 and the NES within the CRD of Yap1, and that the cysteine residue(s) within the CRD serve as redox sensors which regulate this nuclear export (29,30). It is conceivable that TRX affects the redox states of cysteine residue(s) within the CRD, because TRX is a reductant of disulfide bonds. TRX may reduce the cysteine residue(s) to keep the CRD in a form which is able to interact with Crm1. Therefore, Yap1 was constitutively concentrated in the nucleus in trx1⌬/trx2⌬, but not concentrated in the nucleus under normal conditions in the cells that contain TRX (Fig. 3).
It is still controversial whether TRX directly associates with cysteine residue(s) of the CRD or not. We confirmed that TRX exists not only in the cytoplasm but also in the nucleus, especially after the treatment with a mild oxidative stress (Fig. 6). Therefore, TRX would be possible to associate with Yap1 both in the cytoplasm and in the nucleus. However, we were not able to detect the direct interaction of Yap1 with Trx1 or Trx2 by the two-hybrid assay experiments. This might be because their association is too weak to detect by this assay system, or because TRX cannot associate directly with Yap1. If TRX has no direct association with Yap1, TRX may affect the redox states of the CRD via other factors, such as Ref-1 in the case of AP-1 regulation (10). Yan et al. (30) reported that a single cysteine residue within the CRD is necessary and sufficient for the response to oxidative stress, and speculated that the cysteine residue(s) within the CRD form linkages with a third protein, as yet unidentified, which serves to mask the NES. TRX may regulate the formation of these linkages. Interestingly, the TRX2 gene is one of the target genes for Yap1. Therefore, Yap1 may be autoregulated; active form of Yap1 enhances transcription of the TRX2 gene, and Trx2 may promote the interaction between Crm1 and Yap1 by unmasking the NES to export Yap1, as one of the down-regulations. This idea may be supported by the result that TRX exists in the nucleus after the treatment with a mild oxidative stress (Fig. 6).
Another possible explanation is that TRX may take part in the regulation of Yap1 via the redox states of the whole intracellular environments. TRX not only modifies proteins such as transcription factors, it also maintains intracellular environments in reduced states (11,12). The deficiency of the thioredoxin system causes the oxidation of the cytoplasm in E. coli (12). The TRX deficiency in yeast may change the intracellular redox states, and consequently this change would cause the redox modification of cysteine residues of the CRD and the activation of Yap1. Indeed, we showed that the level of intracellular oxidation in trx1⌬/trx2⌬ was higher than that of the wild type (Fig. 4C), and it has been reported that the trx1⌬/ trx2⌬ mutant accumulates high levels of oxidized glutathione (47). These results support the possibility of the change in intracellular redox states caused by TRX deficiency in S. cerevisiae.
Taking these results together with the results in this study, we present a tentative model for the roles of TRX in the regulation of Yap1. Under normal conditions, TRX reduces intracellular environments and may maintain the CRD in a reduced form, which can interact with Crm1. After the stimulation with a mild oxidative stress, the synthesis of TRX is induced by the activated Yap1, and then TRX may function as a deactivator for Yap1 in the nucleus besides as an antioxidant enzyme. TRX may unmask the NES and facilitate the interaction between the CRD and Crm1. TRX in the regulation of Yap1 in yeast may play a similar role of GRX in the deactivation of OxyR in E. coli (3). The details of the interaction between Yap1 and TRX are currently under investigation.
The Difference in Functions between TRX and GRX-In addition to TRX, GRX is also a reductant of disulfide bonds, and known as a redox regulator of OxyR in E. coli (3). It has been reported that both TRX and GRX can partially substitute for each other in vivo, and contribute to maintaining the cytoplasm in a reduced state in E. coli (12,50). Both of these thioldisulfide oxidoreductases are required for protection against oxidative stress in S. cerevisiae (21,34). However, GRX seems to be not involved in the regulation of Yap1 because GRX deficiency did not affect the activity of Yap1 (Table I and Fig.  1C) and the constitutive activation of Yap1 in trx1⌬/trx2⌬ was not repressed by the overexpression of the GRX1 or GRX2 gene (Fig. 5). Additionally, the yap1⌬/grx1⌬/grx2⌬ triple mutant was viable but the yap1⌬/trx1⌬/trx2⌬ triple mutant showed synthetic lethality. There may be a distinct difference between TRX and GRX in their functions regarding the regulation of Yap1 activity. Differences were also observed in other cellular metabolism, i.e. the level of intracellular oxidation of the trx1⌬/trx2⌬ mutant was higher than that of the grx1⌬/grx2⌬ mutant (Fig. 4C), and grx⌬/grx2⌬ mutant did not show the methionine auxotrophy and the increase of cell size. In E. coli, there is difference in the substrate specificity between TRX and GRX (50). TRX may be playing irreplaceable role(s) and be more important than GRX as an antioxidant enzyme in S. cerevisiae. The difference in redox potential may be one of causes of the difference in functions between TRX and GRX (52), but it is still obscure what the cause of the difference in functions is.
Synthetic Lethality of yap1⌬ and trx1⌬/trx2⌬-Another important phenotype to be noted in the relationship between Yap1 and TRX is the synthetic lethality of yap1 null mutation and TRX deficiency (trx1⌬/trx2⌬). The synthetic lethality of the yap1⌬/trx1⌬/trx2⌬ triple mutation strongly indicates that Yap1 is essential in trx1⌬/trx2⌬ for the growth under aerobic conditions. As we demonstrated here, the levels of transcription of the genes encoding antioxidant enzymes, such as GSH1, GLR1, and TRR1 (whose expressions are under the control of Yap1), were increased in trx1⌬/trx2⌬ ( Fig. 1 and Table I). Such an increase of antioxidant enzymes must have some physiological effect(s) that allow the trx1⌬/trx2⌬ mutant to survive under aerobic conditions. The simplest explanation would be that the constitutive activation of Yap1 is essential for trx1⌬/ trx2⌬ to compensate for TRX deficiency. The trx1⌬/trx2⌬ mutant may barely survive by increasing other antioxidant enzymes in the Yap1-dependent manner to keep the level of intracellular oxidation as low as possible. Thus, despite the constitutive activation of Yap1, the trx1⌬/trx2⌬ mutant was hypersensitive to H 2 O 2 (Fig. 7) and its intracellular oxidation level was still higher than that of the wild type (Fig. 4C). The synthetic lethality of yap1⌬/trx1⌬/trx2⌬ presumably reflects the more oxidized states of intracellular environments. The intracellular oxidation level of grx1⌬/grx2⌬ was almost the same with that of the wild type (Fig. 4C), and the yap1⌬/grx1⌬/ grx2⌬ triple mutant was viable. The disruption of the GLR1 gene was also lethal in the trx1⌬/trx2⌬ background under aerobic conditions (47). These observations may support our idea.