The Protein Interaction of Saccharomyces cerevisiae Cytoplasmic Thiol Peroxidase II with SFH2p and Its in Vivo Function*

Previously, we reported that the yeast cytoplasmic thiol peroxidase type II isoform (cTPx II), a member of the TSA/AhpC family, showed a very low peroxidase ac- tivity when compared with other cytoplasmic yeast isoforms, and that cTPx II mutant (cTPx II (cid:1) ) showed a severe growth retardation compared with that of the wild-type cells. To reveal the physiological function of cTPx II in yeast cell growth, we searched for proteins which react with cTPx II. In this study, we identified a novel interaction between cTPx II and CSR1p using the yeast two-hybrid system. CSR1p (SFH2p) has been known to be one member of Sec14 homologous (SFH2) proteins. SFH2p exhibits phosphatidylinositol transfer protein activity. Interestingly, we found that cTPx II selectively bound to SFH2p among the five types of SFH proteins and Sec14p. The interaction required the dimerization of cTPx II. In addition, SFH2p also specifically bound to cTPx II among the yeast thiol peroxidase isoforms. The selective interaction of the dimer form of cTPx II (the oxidized form) with SFH2p was also con- firmed by glutathione S -transferase pull-down and immunoprecipitation assays. The growth retardation, clearly reflected by the length of the lag phase, of cTPx a novel protein-protein CSR1p

Aerobically growing cells are continuously challenged by reactive oxygen species. Reactive oxygen species are potent oxidants capable of damaging all cellular components including DNA, protein, and membrane lipid. To protect against the toxicity of reactive oxygen species, aerobic organisms are equipped with an array of defense mechanisms (1). Among these, a new type of peroxidase, named thiol peroxidase(TPx), 1 thioredoxin peroxidase protector protein, thiol-specific antioxidant protein (TSA), or peroxiredoxin, has been known to eliminate H 2 O 2 and alkyl hydroperoxides using a thiol-reducing equivalent (2)(3)(4)(5)(6)(7). The new type of peroxidase with cysteine as the primary site of catalysis has been discovered from prokaryotes to eukaryotes . The TPx family, also referred to as the TSA/alkyl hydroperoxide reductase family or peroxiredoxin family, is a large family of a new type of antioxidant. In mammalian tissue, at least six types of TPx isoenzymes have been identified. Recently, in addition to two types of TPx isoenzymes (TSA I, YML028W (Ref. 2); TSA II/alkyl hydroperoxide reductase 1, YLR109W (Refs. 15 and 16)) described previously as yeast members of the TSA/alkyl hydroperoxide reductase family, we have characterized three TPx homologues (YDR453C, YBL064C, and YIL010W) as a new member of the yeast TPx family (22). Evidence from our recent work indicates that different TPx isoenzymes are localized in distinct cellular organelles, where they are likely to serve diverse functions in yeast cells (22). TSA I and TSA II (Ahp1) were described as a general hydroperoxide peroxidase to remove H 2 O 2 and alkyl hydroperoxide (2,15,16,22). Three novel isoforms showed a thiol peroxidase activity supported by thioredoxin and appeared to be distinctively localized in the cytoplasm, mitochondria, and nucleus. Each isoform was named after its subcellular localization such as cytoplasmic TPx I (cTPx I or TSA I), cTPx II (YDR453C), cTPx III (TSA II/alkyl hydroperoxide reductase 1), mitochondrial TPx (YBL064C), and nuclear TPx (YIL010W) (22).
Recently, we have reported that, unlike other TPx null mutants, cTPx II null mutant showed a slow growth phenotype and accumulation of G 1 -phased cells during the log phase (23). The growth defect appeared to be caused by the accumulation of G 1 -phased cells, even in the exponentially growing condition (22). We have demonstrated that Msn2p/4p-mediated transcription of the cTPx II gene under negative control of Ras-TOR signaling pathway is turned on at diauxic shift (23). Thus, on the basis of previous observations, we proposed that cTPx II might be one of the candidates for signaling mediators to maintain the aerobic life of stationary-phased yeast, although the function is not clearly understood (22,23).
Proteins are frequently engaged in multiple interactions, and that governance of protein interaction specificity is a primary means of regulating biological systems. One member of mammalian TPx isoenzymes (i.e. PAG, for proliferation-associated gene product), which is thought to be a mammalian counter part of Saccharomyces cerevisiae cTPx II, is an Abl SH3binding protein and a physiological inhibitor of c-Abl tyrosine kinase activity (24). To gain more insight into the physiological function of cTPx II in maintaining the aerobic life of stationaryphased yeast, we explored a possible protein interaction with cTPx II using the yeast two-hybrid system. We first identified a novel protein-protein interaction between cTPx II and CSR1p (SFH2p). SFH2p has been known to be one member of Sec14 homologous (SFH) proteins (25). SFH2p exhibits phosphatidylinositol transfer protein (PITP) activity, which is essential for Golgi function and cell viability (26,27). In this study, we find the oxidative stress-mediated protein-protein interaction between the dimeric form (oxidized form) of cTPx II and SFH2p and the rescue of the growth defect in cTPx II⌬ strain by the deletion of SFH2p. These results suggest that cTPx II implicated in the regulation of yeast cell proliferation driven by oxidative stress. This work may help to elucidate a reason for the slow growth phenotype and furthermore, a physiological function of cTPx II during post-diauxic growth in S. cerevisiae.

EXPERIMENTAL PROCEDURES
Construction of Bait and Prey Plasmid for the Two-hybrid System-pEG202, which was used as a vector to express the LexA-cTPx II fusion protein, contains the his3 selectable marker, yeast 2-m origin, Escherichia coli pBR origin, and LexA DNA-binding domain. The cTPx II DNA fragment encoding cTPx II (192 amino acids) was amplified by PCR and cloned into EcoRI-XhoI sites of pEG202. The resulting plasmid, named pLexA-cTPx II, was used as the bait. The yeast strain EGY48 (MATa, his3, trp1, ura3-52, leu; pLeu-LexAop/pSH18 -34 (Lex-AoplacZ reporter)) was transformed with the plasmids. The expression of the fusion protein was identified by immunoblot using cTPx II antibody.
Two-hybrid Screening of Yeast Genomic DNA Library-The complete coding region of cTPx II was inserted into pLexA (Clontech) and used as bait to screen a S. cerevisiae strain S288C (ATCC 26108) genomic library (OriGene) in the transcription activator B42 fusion vector pJG4 in a yeast strain EGY48 possessing the p8op-Laz reporter plasmid (Clontech). Plasmid pJG4-5 contains the TRP1-selectable marker, yeast 2-m origin, and E. coli pUC origin. Expression of the fusion protein in this plasmid is under the control of GAL1, a galactose-inducible promoter. Approximately 2 ϫ 10 6 yeast transformants were screened according to the instruction from the manufacturer. The yeast strain EGY48/pLexA-cTPx II was transformed with the yeast genomic DNA library by the lithium acetate method. Transformants were selected for tryptophan prototrophy on synthetic agar medium (Ura Ϫ , His Ϫ , Trp Ϫ ) that contained 2% glucose. All of the transformants were pooled and re-spread on a synthetic medium (Ura Ϫ , His Ϫ , Trp Ϫ , Leu Ϫ ) that contained 2% galactose to induce the introduced DNA. To confirm the dependence of their growth in the presence of galactose, cells grown on the selection medium were re-plated on the synthetic medium (Ura Ϫ , His Ϫ , Trp Ϫ , Leu Ϫ ) containing 2% galactose and 2% glucose. Cells grown only on the galactose media were subjected to further characterization. To test for ␤-galactosidase activity, the selected cells were also streaked on a synthetic medium (Ura Ϫ , His Ϫ , Trp Ϫ ) consisting of 2% galactose (or 2% glucose as a control) and 5-bromo-4-chloro-3-indolyl-␤-D-galactoside. The cells showing the galactose-dependent expression of both reporter genes were finally selected for the purpose of isolation of the plasmid. The isolated plasmids were transformed in the E. coli strain XL-1 Blue (SupE44, hsdR17, recA1, endA1, gyrA46, thi Ϫ , relA1, lac Ϫ ). The resulting transformants were selected on the basis of their growth ability on a LB minimal medium supplemented with ampicillin. The plasmids, obtained from the Amp ϩ E. coli transformants, were used as the prey for an additional confirming test. The inserted genomic DNAs were sequenced using the B42 primer (5Ј-CC AGC CTC TTG CTG AGT GGA GATG), and identified on a data base for S. cerevisiae (YGD).
Specificity Test for Interaction between cTPx II (Bait) and SFH Proteins Including SFH2p and Sec14p (Prey)-The genes for Sec14, and five types of SFH proteins including SFH2p, were PCR-amplified and cloned into pJG4-5. These plasmids were co-transformed with pLexA-cTPx II. Transformants were selected on the synthetic medium (Ura Ϫ , His Ϫ , Trp Ϫ ) containing 2% glucose. After 3-day incubation at 30°C, each transformant was picked up and re-spread on the synthetic medium (Ura Ϫ , His Ϫ , Trp Ϫ ) consisting of 2% galactose (or 2% glucose for a control test) and 5-bromo-4-chloro-3-indolyl-␤-D-galactoside to identify the protein-protein interaction, which resulted in the expression of lacZ gene. They were retested on the synthetic medium (Ura Ϫ , His Ϫ , Trp Ϫ , Leu Ϫ ) supplemented with 2% galactose (or 2% glucose) to confirm the galactose-dependent growth.
Specificity Test for Interaction between Yeast TPx Isoforms (Bait) and SFH2p (Prey)-The full genes encoding five types of yeast TPx proteins (cTPx I, cTPx II, cTPx III, mTPx, and nTPx) were amplified by PCR using appropriate primers, and cloned into pLexA. These plasmids were co-transformed with pJG4-5-SFH2.
␤-Galactosidase Assay-The relative strength of the protein interaction between bait and prey proteins was determined by measuring the expression level of lacZ reporter gene. Cells were suspended in Z buffer (60 mM Na 2 HPO 4 , 40 mM Na 2 HPO 4 , 10 mM KCl, 1 mM MgSO 4 , pH 7.0) containing 2-mercaptoethanol and disrupted by vortexing with glass beads, and the ␤-galactosidase activity was assayed using o-nitrophenyl-␤-D-galactoside (ONPG) as a artificial substrate according to the method described previously (23,29). The ␤-galactosidase activity was expressed as unit (increase of OD at 412 nm that resulted from ONPG hydrolyzed by ␤-galactosidase per 10 min per mg of protein). Protein concentration was determined using Bradford protein assay kit (Bio-Rad). Yeast transformation, DNA, protein extraction from yeast, and other methods not mentioned were carried out according to supplier manual or a standard protocol described elsewhere (28).
In Vitro Protein Binding Assay Using Glutathione S-Transferase (GST) Pull-down and Immunoprecipitation Methods-The complete coding region of SFH2 was inserted into pGEX-4T-1 (Amersham Biosciences). The resulting plasmid GST-SFH2 was transformed into E. coli BL21 (DE3). GST-Sfh2p fusion was induced by addition of 0.5 mM IPTG at 20°C for 8 h and purified on glutathione (GSH)-Sepharose 4B beads (Amersham Biosciences). For the pull-down assay, 20 g of cTPx II, 20 g of GST-SFH2p, and 100 l of buffer (50 mM Tris-Cl, pH 8.0, 150 mM NaCl) were mixed and incubated in the presence or absence of 1 mM DTT for 30 min at 4°C with end-on-end shaking. After addition of GSH beads, the mixture was incubated at 4°C for an additional 30 min, and the GSH beads were separated by a brief centrifugation, washed five times with phosphate-buffered saline, and suspended in 25 l of SDS sample loading buffer (50 mM Tris-Cl, pH 6.8, 100 mM DTT, 2% bromphenol blue, 10% glycerol). The bound protein was analyzed by Western blot using GST and cTPx II antibodies. For the immunoprecipitation assay, 20 g of cTPx II and 20 g of GST-SFH2p were mixed and incubated in the presence or absence of 1 mM DTT for 30 min at 4°C with end-on-end shaking. After addition of cTPx II antibody, the mixture was incubated at 4°C for another 30 min, and protein Aagarose beads (Amersham Biosciences) were added to the mixture. After a 30-min incubation at 4°C, the beads were separated by a brief centrifugation and washed five times with the buffer. The beads were washed extensively with phosphate-buffered saline and then boiled in SDS-loading buffer for Western blot.
Gene Disruption by Fusion PCR-Null mutant of cTPx II was a laboratory stock used in previous work (22). The SFH2 gene was disrupted by a short flanking homology method. The 5Ј and 3Ј ends of SFH2 gene were amplified with a pair of primers, which were designed for amplification of the 0.41-kb DNA fragments from the ends. The TRP gene was used as select markers for the SFH2⌬ strain. TRP gene was amplified pJG4-5 vector with primers (5Ј-ATG TCT GTT ATT AAT TTC ACA GGTAT, forward; and 5Ј-CTA TTT CTT AGC A TTT TTG ACG AAAT, reverse). This marker was used for producing a DNA fragment where the marker is flanked by the two 0.41-kb end regions of SFH2 gene by a fusion PCR. The first fusion PCR was performed by melting the 5Ј end fragment of SFH2 gene, which was amplified with forward primer (5Ј-ATG TCT GTT ATT AAT TTC ACA GGT AT) and reverse primer (5Ј-TRPCF-CTA CCT GTG AAA TTA ATA ACA GAC ATG CCA GTA CAA ATC GTG ACC CAA). The additional bases complementary to the TRP forward primer are represented as TRPCF. The PCR product was gel-purified, and the second fusion PCR was set up by melting the product of the first fusion PCR product and the 3Ј end fragment of SFH2 gene, which was amplified with forward primer (5Ј-TRPCR-ATT TCG TCA AAA ATG CTA AGA AAT AGG GAA CAT CAT CAA GAA TTG GCT) and reverse primer (5Ј-CTA TTT CTT AGC ATT TTT GAC GAA AT). The additional bases complementary to the TRP reverse primer are represented as TRPCR. The final PCR product was gelpurified and used for the disruption of Sfh2 gene. The cTPx II⌬SFH2⌬ strain was derived from cTPx II⌬ strain, and the SFH2⌬ strain was derived from cTPx II⌬ wild strain, W303-1A. Correct replacement of the target open reading frames was confirmed by PCR directly on yeast cells.
Determination of Thioredoxin-linked Peroxidase Activity of cTPx II-Peroxidase reaction was performed in 350 l of reaction mixture containing 50 mM Hepes-NaOH, pH 7.0, 0.8 M thioredoxin, 0.3 M thioredoxin reductase, 0.26 mM NADPH, 2.5 M cTPx II, varying concentrations of SFH2p, and 300 M t-butyl hydroperoxide at room temperature. The peroxidase activity of TPx linked to NADPH oxidation was traced for 3 min as the decrease of A 340 . cTPx II and SFH2p were purified from BL21 strain carrying corresponding expression vector. For SFH2p, the E. coli stain was cultured for 12 h at 20°C after addition of IPTG.

Screening of cTPx II-interacting Proteins in the Yeast Two-
hybrid System-The protein interaction with cTPx II was studied using a yeast two-hybrid system. Approximately 2 ϫ 10 6 independent transformants were pooled and re-suspended on the selection plates (Ura Ϫ , His Ϫ , Trp Ϫ , Leu Ϫ ) containing galactose. Among approximately 2500 colonies selected on the plates, a total of 56 colonies showed galactose-dependent ␤-galactosidase activity. For determining the relative expression level of the lacZ reporter gene caused by the protein-protein interaction, 56 colonies were cultured in the presence of galactose, and the expressed ␤-galactosidase activities were measured. Among the 56 colonies tested, 2 colonies showed relatively stronger ␤-galactosidase activities (data not shown). The two colonies were finally selected for sequencing. Sizes of the DNA inserts appeared to be same (2.4 kbp) on agarose gel. Sequence analysis revealed that the clones encode the C-terminal fragment of SFH2p. The DNA fragment encoded SFH2p fragment (the region from Ala-109 to the C-terminal end, Val-408). To investigate the interaction of cTPx II and complete SFH2p, the full SFH2 gene was amplified by PCR, and the PCR fragments were inserted into pJG4-5. The plasmids were cotransformed with pLexA-cTPx II. The expression level of the lacZ reporter gene caused by the protein-protein interaction between cTPx II and SFH2p was ϳ20% higher than the SFH2p fragment-mediated expression level (data not shown), which indicates the protein-protein interaction between SFH2p and cTPx II.
Cytoplasmic TPx II Protein Specifically Binds to SFH2p-Yeast expresses six members of a family of PITPs including Sec14p dedicated to divergent sets of cellular functions. The sequences of the family share a high degree of sequence similarity with Sec14p (from 43 to 79%) (25). Four of these proteins (SFH2p, SFH3p, SFH4p, and SFH5p) exhibit phosphatidylinositol transfer activity (30). Thus, to investigate a possible protein-protein interaction between cTPx II and other members of PITP family, the full genes of PITP family were fused in frame to the B42-activating domain of the pJG4-5 vector and used as prey for cTPx II. The plasmids were co-transformed with pLexA-cTPx II to analyze in vivo protein-protein interaction. The relative ␤-galactosidase activities (Fig. 1) showed that any significant ␤-galactosidase activities were not expressed in the various transformants except for the yeast cells containing SFH2p as prey, which suggests that cTPx II protein specifically binds only to Sfh2p.
The Dimeric Form of cTPx II Protein Selectively Binds to SFH2p-Among the yeast members, four TPx isoforms (cTPx 1, cTPx II, cTPx III, and mTPx) (but not nTPx) were dimerized upon oxidation via intermolecular disulfide bond (22). The alignment of amino acid sequences revealed one highly conserved cysteine within the TPx homologous proteins. The conserved cysteine (Cys-48 for cTPx II) was previously reported to serve as one part of a sulfhydryl group forming the intermolecular disulfide linkage (22). For cTPx II, Cys-171 acts as a counterpart of a sulfhydryl group to form intermolecular sulfide bond. Thus, it is interesting to investigate which form among the monomeric and dimeric forms is more favorable for protein interaction with SFH2p. To answer the question, first, for the purpose of confirming the dimerization of cTPx II in vivo using yeast two-hybrid system, we constructed pLexA-cTPx II mutant, in which conserved cysteine(s) (C71S, C171S, and C48S/C171S) is replaced with serine as a bait, and pJG4-5-cTPx II for prey. To analyze in vivo dimerization of cTPx II, the bait plasmid (pLexA-cTPx II mutant) was co-transformed with the prey plasmid (pJG4-5-cTPx II). The relative ␤-galactosidase activities and the growth capability on the selection media show that the wild-type cTPx II (bait) binds to the cTPx II expressed from the prey plasmid, whereas any other cTPx II mutant proteins as a bait could not dimerize with the wild-type prey protein (Fig. 2), confirming the in vivo dimerization via the intermolecular disulfide linkage.
To examine the effect of the dimerization of cTPx II on the protein-protein interaction with SFH2p, pLexA-cTPx II mutants (C48S, C171S, and C48S/C171S) were introduced to the yeast cell containing pJG4-5-SFH2. To monitor the proteinprotein interaction, the expression levels of lacZ gene were determined in the various yeast transformants (Fig. 3). Taken together with the capability for the dimerization between the wild-type cTPx II proteins shown in Fig. 2, the unique expression of ␤-galactosidase caused by the protein-protein interaction between the wild-type cTPx II and SFH2p suggests that the SFH2p selectively binds to the dimeric form of cTPx II protein.
SFH2p has two cysteine residues (Cys-246 and Cys-383) within the C-terminal region of the sequence consisting of 407 amino acids. To rule out the possibility of an intermolecular disulfide linkage between cTPx II and SFH2p, which might result in the expression of ␤-galactosidase activity, pJG4-5-SFH2 mutant lacking the DNA fragment encoding the protein fragment (from Val-240 to Val-407) containing two cysteine was constructed and co-transformed with pLexA-cTPx II vec- vector and used as prey for cTPx II. Each vector was co-transformed with pLexA-cTPx II. The induced ␤-galactosidase activity of the transformed yeast cells by galactose was measured. As a control test, the activity of the yeast cells grown on medium containing glucose as a carbon source was also measured. The ␤-galactosidase activity was assayed. The ␤-galactosidase activity is expressed as increase of OD at 412 nm that resulted from ONPG hydrolyzed by ␤-galactosidase (OD increase per mg of protein per 10 min).
tor. The ␤-galactosidase activity, which was induced by the protein-protein interaction between cTPx II and the SFH2p mutant, was ϳ30% lower than that by the interaction between the corresponding wild-type proteins (data not shown). These data indicate that there is no intermolecular disulfide bond between cTPx II and SFH2p.
SFH2p Selectively Interacts to cTPx II among Yeast TPx Isoforms-Recently, we have characterized five TPx homologues as new members of yeast TSA/AhpC family (cTPx 1, cTPx II, cTPx III, mTPx, and nTPx) (22). Thus, to test whether other yeast TSA/AhpC members could act as interacting partners of SFH2p, the complete encoding DNA legions of TPx isoforms were fused in frame to the LexA-DNA binding domain of the pLexA vector and used as bait for SFH2p. To evaluate the protein-protein interactions, the expressed ␤-galactosidase activities were measured. Fig. 4 shows that except for cTPx I, the amino acid sequence of which shares a high degree of sequence identity with cTPx II (86% identity, 96% positives), the other TPx isoforms (cTPx III, mTPx, and nTPx) did not significantly interact with SFH2p when compared with cTPx II. To see the dimerization effect on the protein-protein inter-action, the same experiment as that of cTPx II was performed using cTPx I mutant, in which the corresponding cysteine was replaced with serine. In contrast to the case of cTPx II, the ␤-galactosidase activities induced by the protein-protein interaction were regardless of presence or absence of the conserved cysteine (data not shown). This result suggests that the protein-protein interaction between cTPx I and SFH2p is different from that between cTPx II and SFH2p. Taken together, these results suggest that the protein-protein interaction between cTPx II and SFH2p is very specific, and that SFH2p selectively binds to the dimeric form of cTPx II Biochemical Verification of Two-hybrid Interaction of Dimeric Form of cTPx II with SFH2p-Interaction between the dimeric form of cTPx II and SFH2p was verified by both methods of immunoprecipitation and GST pull-down. For immunoprecipitation assay, after purified GST-SFH2 fusion protein and native cTPx II protein were incubated in the presence or absence of DTT, cTPx II antibody was employed to pull down GST-SFH2 fusion protein in vitro. GST protein was used as a control to eliminate a possible binding of GST itself to cTPx II protein. Immunoblot analysis of the protein complex with GST antibody demonstrates that GST-SFH2 fusion protein interacts to the dimeric (oxidized) form of cTPx II, but not to the monomeric (reduced by DTT) form of cTPx II (Fig. 5A). GST protein itself did not bound to cTPx II protein (data not shown).
For GST pull-down assay, after purified GST-SFH2 fusion protein and native cTPx II protein were incubated in the presence or absence of DTT, SFH2-GST fusion protein was employed to pull down cTPx II protein in vitro. GST protein was used as a control to test the specificity of the binding. Immunoblot analysis of the bound proteins with GST and cTPx II antibodies was performed. Fig. 5B shows that the dimeric form of cTPx II protein, which was incubated without DTT, binds to immobilized GST-SFH2 protein, whereas the monomeric form of cTPx II, which was incubated in the presence of DTT, does not bind to the GST-SFH2 fusion. Cytoplasmic TPx II did not bind to immobilized GST protein alone (data not shown).
The Growth Phenotype of cTPx II⌬SFH2⌬ Strain-Recently, we have reported that, unlike other TPx null mutants, cTPx II null mutant showed a slow growth phenotype and cells were arrested at G 1 -phase during the log phase. The growth defect was not recovered by prolong the cell culture to the stationary phase (22). The transcription of cTPx II was activated by the diauxic shift, which is strictly negative-regulated by Ras-TOR signaling pathway (23). The previous data suggest that cTPx II FIG. 2. Verification of the dimerization of cTPx II using yeast two-hybrid system. Each point-mutated cTPx II gene (wild (W), C48S (C1S), C171S (C2S), and C48S/C171S (C1/2S)) was fused in frame to the LexA-DNA binding domain of the pLexA vector and used as bait for the wild type of cTPx II. Each vector was co-transformed with pJG4-5-cTPx II (plasmid for prey). The expressed ␤-galactosidase activity in the resulting yeast cells in the presence of galactose was measured. The galactose-dependent growth on plate deleting Ura, His, Trp, and Leu was shown in inset. The ␤-galactosidase activity is expressed as increase of OD at 412 nm that resulted from ONPG hydrolyzed by ␤-galactosidase (OD increase per mg of protein per 10 min).

FIG. 3. CTPx II dimeric form-specific interaction of SFH2p.
Each point-mutated cTPx II gene (wild (W), C48S (C1S), C171S (C2S), and C48S/C171S (C1/2S)) was fused in frame to the LexA-DNA binding domain of the pLexA vector and used as bait for Sfh2p. Each vector was co-transformed with pJG4-5-SFH2 (the plasmid for prey). The expressed ␤-galactosidase activity in the resulting yeast cells in the presence of galactose was measured. The ␤-galactosidase activity is expressed as increase of OD at 412 nm that resulted from ONPG hydrolyzed by ␤-galactosidase (OD increase per mg of protein per 10 min). has a physiological role in stationary-phased growth. Therefore, it is worth investigating how cTPx II functions in the stationary-phased growth. In an attempt to reveal the physiological role of cTPx II interaction with SFH2p in stationary phase, we constructed the SFH2⌬ strain from cTPx II⌬ and its parent strains (W303-1a) and monitored their effects on cell growth. Fig. 6 shows that the colony-forming units are significantly higher for the cTPx II⌬ cells relative to those for the wild type, SFH2⌬, and double mutant cells, with the latter three cases forming large colonies. SFH2⌬ strain did not give any a slow growth phenotype compared with its wild-type strain. However, the deletion of SFH2 in cTPx II⌬ strain resulted in recovering a slow growth phenotype shown by cTPx II⌬ strain, which suggests the inhibitory action of a free form of SFH2p on a cTPx II-dependent growth. Therefore, the growth retardation of cTPx II⌬ strain could be interpreted as the result of an inhibitory action of a free form of SFH2p in the absence of cTPx II protein. Taken together, these results could provide a new function of the cTPx II-SFH2p complex. The growth recovery of the cTPx II⌬ strain by deletion of SFH2p supports the in vivo protein-protein interaction between cTPx II and SFH2p.
To investigate further the physiology of the double mutant, the growth was monitored throughout the yeast growth phases. Fig. 7A shows that the growth of the double mutant exhibits a shorter lag phase relative to that of the cTPx II⌬ strain, which is visualized in Fig. 6, but the growth is decreased to a lower level than that of cTPx II⌬ strains during stationary-phased growth. The growth of SFH2p⌬ strain was nearly the same as that of its parent strain, which is consistent with the result shown by Fig. 6. To test the possibility that the severe growth retardation of the double mutant during the stationary phase might be caused by the high susceptibility of the double mutant to oxidative stress, oxidative stress was subjected to the mutant (Fig. 7B). Exponential-and stationary growth-phased cells were plated on the YPD plate containing 0.1 mM t-butyl hydroperoxide, and the cell viability was measured in terms of the number of the survival colonies. For the viability test using exponential-phased cells, each single mutant did not show any significant change in cell viability against the oxidative stress. However, the cell viability of the double mutant was significantly lower than either of the single mutants and its wild-type strain. In case of the stationary-phased cells, except for the wild-type strain, the viability of all mutants (cTPx II⌬, SFH⌬, FIG. 5. Biochemical verification of two-hybrid interaction of dimeric form of cTPx II with SFH2p. In vitro protein-protein interaction between dimeric form of cTPx II and SFH2p was verified by both methods of immunoprecipitation (A) and GST pull-down (B). A, for immunoprecipitation assay, the mixture of GST-SFH2p fusion protein (20 g) and cTPx II (20 g) was incubated in the presence (lane A-3) or absence of DTT (lane A-2); cTPx II antibody was employed to pull down GST-SFH2p fusion protein. GST protein was used as a control to test the specificity of the binding (data not shown). The immune complex was precipitated with protein A-Sepharose beads and resolved by 10% SDS-PAGE following repeated washings of the beads, transferred to nitrocellulose paper, and detected with GST antibody. B, for GST pulldown assay, cTPx II (20 g) was incubated with GST-SFH2p fusion protein (20 g) in the presence (lanes B-3 and B-6) or absence of DTT (lanes B-2 and B-5), followed by the addition of GSH-Sepharose 4B beads. After three washes of the beads, the equal volume of the beads was loaded to 10% SDS-PAGE for analyzing GST-SFH2p and cTPx II using GST antibody (lanes B-1 to B-3) and cTPx II antibody (lanes B-4 to B-6), respectively. GST protein only was used as a control to test the specificity of the binding (data not shown). and cTPx II⌬SFH⌬) is significantly lower when compared with the viability derived from the corresponding exponential cells. For the wild-type strain (W303-1a), stationary-phased cells survived more than those plated with cells from the exponential growth phase, which is consistent with the fact that stationary-phased cells are more resistant to oxidative stress. Analysis of viability data indicated that deletion of SFH2p is more harmful for the stationary-phased yeast cells in the presence of oxidative stress. Collectively, these results demonstrate that the growth retardation of the aged double mutant is caused by the high susceptibility to oxidative stress, which is acquired by deletion of SFH2p rather than by cTPx II. Therefore, the inhibitory action of Sfh2p on yeast growth (cTPx II-dependent growth) before the diauxic shift (i.e. before the induction cTPx II) (23) is necessary for maintaining yeast life against oxidative stress.
Peroxidase Activity of the Complex of cTPx II and SFH2p-Previous data suggest that the protein-protein interaction between cTPx II and SFH2p could act as a signal event involved in yeast growth. The biochemical characterization of complex is the obvious key to understanding how the complex-forming process is involved in yeast growth. The physiology of the double mutant described previously suggests that the complex itself are necessary for maintaining yeast aerobic life. Now, it is interesting to examine the changes of antioxidant activity as forming the complex. The change of thioredoxin-linked peroxidase given by cTPx II was monitored as the concentration of SFH2p was increased to one equivalent to the concentration of cTPx II. After a 30-min incubation of the mixture consisting of cTPx II and SFH2p for forming the complex, the peroxidase activity in terms of NADPH decrease was spectrophotometrically traced at 340 nm (Fig. 8). The peroxidase activity was not significantly changed throughout the experiments, which shows that the activity of the free form of cTPx II is nearly same as that of the complex form. Therefore, we ruled out the possibility that some changes of peroxidase activity of cTPx II by forming the complex could directly participate in changing the phenotype of the double mutant (i.e. susceptibility against oxidative stress). DISCUSSION ROS do not have exclusively toxic effects. Low levels of ROS can act as signaling molecule under physiological condition (31). In mammals, ROS such as H 2 O 2 produced in physiological condition can activate transcription factor, such as NFB and AP-1 (32), and can function as signals in apoptosis that is induced by tumor necrosis factor-␣ (33). It is known that at least six types of TPx isoenzymes exist in mammalian. Evidence from our recent work indicates that different TPx isoenzymes are localized in distinct cellular organelles, where they are likely to serve diverse functions in yeast cells (22). We reported that five types of TPx isoforms exist in yeast. The amino acid sequence similarity among yeast TPx homologous proteins is 11.6%. Two cytoplasmic isoforms (cTPx I and cTPx II) share the highest degree of amino acid sequence similarity (86% identity, 96% positives) among yeast TPx isoforms, but thioredoxin-dependent thiol peroxidase activity of cTPx II is much lower than that of cTPx I (22). In addition, we also found that cTPx II null mutant showed a slow growth phenotype when compared with its parent and the other TPx mutants (22). Furthermore, the protein expression of cTPx II, which is very low compared with other TPx isoforms, is highly sensitive to intracellular redox state (22,23). In the present study, we have provided a line of evidence that cTPx II may act as a candidate for signaling mediator upon oxidative stress In our search for interacting molecules with cTPx II, we have isolated SFH2p using the yeast two-hybrid system, and demonstrated that cTPx II protein specifically binds the oxidative form (dimeric form) of SFH2p. Taken together with our observations, our results suggest the possibility that cTPx II might be a potential candidate for the oxidative-stress-mediated signaling mediator. Similarly, a putative mammalian counterpart for yeast cTPx II, PAG (for proliferation-associated gene product), gene expression of which occurs in two cellular events (oxidative stress and proliferation), is an Abl SH3-binding protein and a physiological inhibitor of c-Abl tyrosine kinase activity (24).
Previously we reported that transcription of cTPx II is turned on at diauxic shift, which is under negative control of Ras/cAMP-TOR signaling pathway (23). This result implies that cTPx II works at the stationary phase of yeast life cycle. In addition to the selective interaction of SFH2p with the dimeric (oxidative) form of cTPx II protein, we have found a line of evidence supporting the in vivo physiological function of the CTPx II-SFH2p complex. The rescue of the growth defect in cTPx II⌬ strain by deletion of SFH2p supports a role for cTPx II in triggering a cTPx II-dependent growth, especially stationary-phased growth through the action of the cTPx II and SFH2p complex. We suggest that the process for formation of the complex is necessary for yeast cell viability against oxidative stress, thus, especially, for maintaining the aerobic life of stationary-phased yeast cells. It is generally believed that some of the TSA/AhpC family, including cTPx II, undergo a reversible conversion between the monomeric and dimeric forms upon redox state of cells. Thus, the protein-protein interaction between cTPx II and SFH2p regulated by intracellular redox state could act as a signal event to trigger oxidative stress-dependent (stationary-phased) yeast cell proliferation. Based on the present results, we suggest a working model that the oxidative stress induces the protein-protein interaction between SFH2p and cTPx II proteins, which in part turns on stationaryphased growth of yeast cell (Fig. 9).
We demonstrate that the changes of the antioxidant activity given by cTPx II-SFH2p complex do not directly participate in the function of the complex. It remains to be solved how SFH2p inhibits cTPx II-dependent growth (i.e. stationary-phased growth) and how the cTPx II-SFH2p complex works. Although more characterization of the complex consisting of cTPx II protein and SFH2p is the obvious key to understanding of the function of cTPx II in oxidative stress-mediated cell signaling, overall, the results reported here suggest that the intracellular redox state-dependent protein interaction between cTPx II and SFH2p is a physiological process to mediate growth signaling involved in cTPx II-dependent growth, probably stationaryphased growth.