INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Subfamilies of heat shock proteins that are constitutively expressed in cytosol, such as HSP70, function as molecular chaperones under both normal and stressful conditions in mediating the folding, oligomeric assembly, and intracellular transport of proteins (14). The 90-kDa heat shock protein, HSP90, is the most abundant protein in cytosols of eucaryotic cells. HSP90 is a major molecular chaperone which is distributed ubiquitously among all living organisms and which is indispensable for yeast cells to survive even under nonstressful conditions (2). In vitro, HSP90 prevents the aggregation of chemically denatured or heat-denatured proteins (16, 47, 50) and inherently unstable proteins such as casein kinase II (30). Nonnative proteins associated with HSP90 were released from the complexes and then refolded with the aid of reticulocyte lysates (50) or a combination of HSP70 and HSP40 (10, 29). Consistent with the results of these in vitro studies, overexpression of HSP90 confers a tolerance to stress in mammalian and yeast cells (2, 49). On the other hand, in vivo studies have revealed that HSP90 functions as a molecular chaperone for specific substrate proteins, most of which are involved in signal transduction (33, 38).

The budding yeast Saccharomyces cerevisiae has two genes, HSP82 and HSC82, that encode homologues of HSP90 (2). In cells grown under normal conditions, Hsc82 is constitutively expressed, while the expression of Hsp82 is very low. When cells are exposed to stress, the expression of Hsp82 is markedly enhanced. Since the induction and accumulation of Hsp82 in stressed cells take 1 h or longer, HSC82 rather than HSP82 must be biologically active in the protection of normally growing yeast cells from acute stress. In the course of carrying out this study, we made the unexpected finding that cells overexpressing Hsc82 are hypersensitive to various stresses including high concentrations of NaCl. The phenotypes associated with the sensitivity to salt stress of Hsc82-overexpressing cells are similar to those of calcineurin-defective mutants. This finding led us to investigate the genetic and physical interactions between Hsp82 and Hsc82 and calcineurin.

Calcineurin is a serine/threonine protein phosphatase activated by calcium and the Ca²⁺-calmodulin complex which is involved in a variety of signal transduction processes (21, 45). Purified calcineurin from tissues is a heterodimer consisting of a catalytic subunit and a Ca²⁺-binding regulatory subunit. In S. cerevisiae, genes CNA1 and CNA2 encode the catalytic subunit, while CNB1 encodes the regulatory subunit (8, 9, 22, 24). While calcineurin is dispensable in yeast cells for growth under normal conditions, it is essential for growth under salt stress conditions (27, 32). Recently a substrate phosphoprotein for calcineurin, Tcn1/Crz1, was identified in yeast (44). Calcineurin activity is required for the transcriptional activation by transcriptional factor Tcn1/Crz1 (26, 43, 44) of several genes including PMR2, PMC1, and FKS2 (7, 11).

Ca²⁺-calmodulin activates calcineurin by binding to the catalytic subunit (13). Calcineurin binds to anchoring proteins such as FKBP12 (4, 5, 17, 46) and AKAP79 (20). Since the calcineurin catalytic subunit has an unstable conformation when present alone (19), it is not surprising that calcineurin interacts with a variety of proteins. Calmodulin is the only substance to bind to the catalytic subunit in vitro and protect it from proteolysis (25). We report that Hsc82 protects Cna2 from degradation in vivo when expressed to an appropriate extent but interferes with its activation when overexpressed.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Microbiological techniques. Yeast transformations were performed according to the method of Ito et al. (15). Rich medium containing glucose (YPD) and synthetic minimal medium (SD), as described elsewhere, were used (40). SC is SD supplemented with 0.5% Casamino Acids (Difco) and 100 mg of uracil, adenine sulfate, and tryptophan/liter. YPGal, SDGal, and SCGal are YPD, SD, and SC, respectively, in which 2% glucose is replaced with 2% galactose and 0.2% sucrose. SC-U and SCGal-U are SC and SCGal, respectively, lacking uracil. SC-U,W is SC lacking tryptophan and uracil. Medium was buffered to pH 5.5 via the addition of 5 mM succinic acid and supplemented with CaCl₂, NaCl, or LiCl. HGal-L,H,W is SDGal supplemented with 20 mg each of uracil, arginine, and methionine, 30 mg each of tyrosine, isoleucine, and lysine, 150 mg of valine, 60 mg of phenylalanine, and 50 mg of adenine/liter and 10 mM KPB buffer (pH 7.5). HGal-L,H,W,U is HGal-L,H,W lacking uracil. A 2% mixture of galactose and glucose with 0.2% sucrose was used in medium as a carbon source. HS-U is SD supplemented with 20 mg each of argentine, tryptophan, and methionine, 30 mg each of tyrosine, leucine, histidine, isoleucine, and lysine, 150 mg of valine, 60 mg of phenylalanine, and 50 mg of adenine/liter, and was buffered to pH 5.0 via the addition of 5 mM succinic acid. HS-U,W is HS-U lacking tryptophan. HS-L,H,W,U is HS-U lacking tryptophan, leucine, and histidine. Medium was sometimes supplemented with CaCl₂. Yeast cells were cultured at 30°C, unless otherwise indicated.

Antibodies and reagents. A mouse monoclonal antihemagglutinin (anti-HA) antibody (16B12) was obtained from Berkeley Antibody Company. Rabbit polyclonal anti-calcineurin B antibodies (PA3-025) were obtained from Affinity Bioreagents, Inc. Canavanine was obtained from Sigma, and geldanamycin was obtained from Gibco BRL. FK506 was obtained from Fujisawa Co. Ltd. Bovine calmodulin was obtained from BIOMOL Research Laboratories, Inc., as was the BIOMOL GREEN calcineurin assay kit. Protein concentration was determined using BCA protein assay reagent (Pierce Chemical Co.) with bovine serum albumin (Sigma) as the standard.

Gene replacements. A 1.1-kbp SmaI-PvuII fragment from pJJ281 containing TRP1 was inserted into the EcoRV site of pRS90 (18) to produce KS-HSP82. KS-HSP82 was digested by SalI and SacI to linearize the hsp82::TRP1 fragment and transformed into YPH500 using one-step gene replacement, resulting in JH82D. Gene disruption was confirmed by Western blotting.

Strains and plasmids. The yeast strains used are listed in Table 1. A hsc82 mutant (Y499-1H) was crossed with a hsp82 mutant (YJH82D/pRS315-1GAL10-HSC82) and then sporulated and dissected to produce a hsc82 hsp82/pGAL1:HSC82 double mutant (YJH82DD). Plasmids pYO324-HSC82 and pYO326-HSC82 carry the 4.2-kbp SpeI-SpeI fragment of HSC82 in an XbaI site of 2 µm plasmids pYO324 and pYO326 (34), respectively. Plasmid pYO326-CNA2 was constructed in the following manner. The complete CNA2 coding region was amplified from the yeast genome by PCR using convergent primers 5'-GCGGGATCCGATGTCTTCAGACGCTAT and 5'-GCGGGATCCCTATTTCGTATCATTCTTT. The amplified fragments were inserted into the KpnI and SacI sites of pRS304 after blunting to construct pRS304-CNA2. pRS304-CNA2 was digested with SalI and integrated at the CNA2 locus of the yeast genome to create YJC2-W. The genome DNA from YJC2-W was digested with BglII and self-ligated to produce pRS304-CNA2G. Next, 1.5- and 3-kbp BglII-SalI fragments from pRS304-CNA2G were inserted into the BamHI-SalI sites of pYO326 and pUC119, respectively, to produce pYO326-CNA2N and pUC119-CNA2C. A 3-kbp SalI-KpnI fragment from pUC119-CNA2C was inserted into the SalI-KpnI site of pYO326-CNA2N to produce pYO326-CNA2. Plasmid pSM102-CNA2 was constructed in the following manner and used for expression of triple-HA epitope-tagged product of CNA2. The SphI site of pSM102 was changed into a BamHI site via an insertion of a BamHI linker (Takara Syuzo Co.) after blunting to produce pSM102BamHI. The complete CNA2 coding region was amplified from pYO326-Cna2 by PCR using convergent primers 5'-GCGGGATCCGATGTCTTCAGACGCTAT and 5'-GCGGGATCCCTATTTCGTATCATTCTTT. The amplified fragments were then inserted into the BamHI site of pSM102BamHI to construct pSM102-CNA2. A 2-kbp BglII (just before ATG of the triple-HA coding sequence)-SacI fragment of pSM102-CNA2 was inserted in frame into the BamHI-SacI site of pUCGPD to produce pRS316-pGPD-ha-CNA2. A 0.5-kbp BamHI-BamHI fragment from pMC1871 was inserted into the BamHI site of p2UGPD to produce pYO326-pGPD-lacZ. A 0.5-kbp XhoI-SacI fragment from pRS316-1GAL10 was inserted into the XhoI-SacI site of pRS315 to produce pRS315-1GAL10. A 2-kbp ScfI fragment of pYO326-HSC82 was inserted into the BamHI site of pRS315-1GAL10 after blunting to produce pRS315-1GAL10-HSC82. The complete CNB1 coding region was amplified from the yeast genome by PCR using convergent primers 5'-GCGAGATCTGATGGTGCTGCTCCTTC and 5'-GCGAGATCTTTACACATCGTATTGCAA. The amplified fragments were then inserted into the BamHI site of pUCGPD to construct pRS316-pGPD-CNB1. Plasmid pYO326-cna2^H105Q was constructed in the following manner. A 600-bp fragment of the N terminus-encoding region of CNA2 was amplified from pYO326-CNA2 by PCR using convergent primers 5'-GCGGGATCCGATGTCTTCAGACGCTAT and 5'-CGCCCGCGGAGTAGCCAGAAATGGTC. The amplified fragments were then inserted into the EcoRV site of pT7Blue to produce pT7-Cna2DPP. A 1.2-kbp fragment of the C terminus-encoding region of CNA2 was amplified from pYO326-CNA2 by PCR using convergent primers 5'-TTCTGGCTACTCCGCGGTAACCAAGAATGTAAGCAT and 5'-GCGGGATCCCTATTTCGTATCATTCTTT. The amplified fragments were then inserted into the EcoRV site of pT7Blue to produce pT7-Cna2DPPC. A 600-bp SacI-SacII fragment of pT7-Cna2DPPN was inserted into the SacI-SacII site of pT7-Cna2DPPC to produce pT7-Cna2DPP. A 0.2-kbp SalI-XbaI fragment of pT7-Cna2DPP was inserted into the SalI-XbaI site of pYO326-CNA2 to produce pYO326-Cna2^H105Q. For construction of pRS314-pCNA2-ha-CNA2 and pYO324-pCNA2-ha-CNA2, the complete CNA2 promoter region was amplified from pYO326-Cna2 by PCR using convergent primers 5'-CGCGAGCTCAAAACGTGCC and 5'-CGCAGATCTTGGCGTTGAGAGTGT. The amplified fragments were then inserted into the EcoRV-SacI site of pRS314 to construct pRS314-CNA2p. A 2-kbp BglII-XhoI fragment of pSM102-CNA2 was inserted into the BglII-SalI site of pUCGPD to produce pRS314-pCNA2-ha-CNA2. A 2-kbp SacI-SacI fragment of pRS314-pCNA2-ha-CNA2 was inserted into the SacI site of pYO324 to produce pYO324-pCNA2-ha-CNA2. Plasmid pSM102 was a gift from S. Matsumoto (Tokyo Metropolitan Institute of Medical Science). Plasmids p2UGPD (GPD, glyceraldehyde-3-phosphate dehydrogenase) (31), pUCGPD, and pRS316-1GAL10 were gifts from Y. Kimura (Tokyo Metropolitan Institute of Medical Science). Plasmids pAMS363 and pAMS366 were gifts from M. S. Cyert (Stanford University).

PAGE. To resolve isoforms Hsp82 and Hsc82, sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was carried out using a 4 to 20% polyacrylamide gradient gel (Daiichi Pure Chemicals). In this system, no SDS was added in either the upper or lower electrode buffers.

Immunoblotting. Yeast cells in the exponential-growth phase were harvested and then disrupted with glass beads in the presence of 1% SDS containing 1 mM phenylmethylsulfonyl fluoride and (PMSF), 1 mg of leupeptin, 1 mg of antipain, 1 mg of aprotinin, and 1 mg of pepstatin A/ml. Protease inhibitors were purchased from Sigma. After unbroken cells and glass beads were removed by brief centrifugation at 800 × g, the supernatant was used as the lysate. Lysate was resolved by SDS-PAGE, transferred to a membrane, and subjected to Western blotting with various antibodies.

Immunoprecipitation. Yeast cells in the exponential-growth phase were harvested and then disrupted with glass beads in 200 µl of radioimmunoprecipitation assay (RIPA)-Mo buffer (150 mM NaCl, 1% NP-40, 0.5% deoxycholate, 0.1% SDS, 50 mM Tris, [pH 8.0] 10 mM sodium molybdate, 10 mM MgCl₂) containing 1 mM PMSF and 1 mg of leupeptin, antipain, aprotinin, and pepstatin A/ml. All protease inhibitors were purchased from Sigma. After unbroken cells and glass beads were removed by brief centrifugation at 800 × g, the lysates were clarified twice by centrifugation at 15,000 × g at 4°C for 10 min. The protein concentration of each lysate was determined using BCA protein assay reagent, and the lysates were diluted to the same concentration with RIPA-Mo buffer with protease inhibitors. Protein G beads (Sigma) (50% [vol/vol]) were added to aliquots of lysates containing equivalent amounts of the cell lysates, after which the mixtures were kept at 4°C for 1 h with gentle rotation. The beads were removed by centrifugation. Antibodies were added to the preabsorbed lysates and the mixtures were incubated at 4°C for 2 h with gentle rotation. Subsequently, protein G beads (50% [vol/vol]) were added to the mixtures, which were then incubated at 4°C for 2 h with gentle rotation. The beads were then collected by centrifugation and washed four times with RIPA-Mo buffer. Proteins were eluted by boiling the beads in sample buffer (50 mM Tris [pH 6.8], 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol). They were then resolved by SDS-PAGE, transferred to a membrane, and subjected to Western blotting with various antibodies.

-Galactosidase activity. Cells were grown overnight in an appropriate medium to mid-log phase and incubated for an additional 6 h in the presence of 0.2 mM CaCl₂. FK506 or geldanamycin was added to the incubation medium. -galactosidase activity was determined at room temperature using chloroform-SDS-permeabilized cells as described previously and normalized by cell numbers (12).

Calcineurin phosphatase assay. Phosphatase activities of immune complexes were determined using the BIOMOL GREEN calcineurin assay kit (BIOMOL Research Laboratories, Inc.). Immune complex-bound beads were prepared by immunoprecipitation. The immune complex-bound beads (10 µl) were added to 50 µl of reaction cocktail containing 50 mM Tris, pH 8.0, 100 mM NaCl, 6 mM MgCl₂, 0.5 mM dithiothreitol, 0.1 mg of bovine serum albumin/ml, 0.1 mM CaCl₂, and 0.3 mM RII phosphopeptide substrate. The mixtures were incubated at 30°C for 10 min. The supernatants of the mixtures were collected using a spin filter. Amounts of the released free phosphate in the supernatants were determined by the classic malachite green assay according to the supplier's protocol, after which calcineurin activity was calculated by subtracting background phosphate.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Hypersensitivity of Hsc82-overexpressing cells to various stresses. To investigate the role of Hsp82 and Hsc82 in stress tolerance using S. cerevisiae, a wild-type strain was transfected with a multicopy plasmid harboring HSC82 including its own promoter, creating a strain overexpressing Hsc82. Western blot analysis with the anti-Hsp82/Hsc82 antibody revealed that the strain overexpressed Hsc82 approximately 20-fold under normal conditions (Fig. 1A). Hsc82-overexpressing cells were examined for resistance to various stresses including heat, canavanine, and salt. The expression levels of Hsp82 and Hsc82 in Hsc82-overexpressing cells did not change significantly when these cells were continuously exposed to stresses over a period of a few days (Fig. 1A). Unexpectedly, Hsc82-overexpressing cells were found more sensitive to stresses than cells expressing Hsp82 and Hsc82 at normal levels (Fig. 1B). For instance, growth of Hsc82-overexpressing cells was severely restricted on plates containing 100 µg of canavanine/ml, 1 M NaCl, or 0.2 M LiCl. Experiments using liquid cultures showed that these cells were also hypersensitive to heat, 15% (vol/vol) ethanol, and 2.5 M NaCl (Fig. 2). When Hsc82-overexpressing cells were incubated at a nonlethal high temperature, the cells acquired the same level of heat resistance as control cells (HSC82 HSP82), but the preincubation did not affect the hypersensitivity to 2.5 M NaCl (Fig. 2). The hypersensitivity to 1 M NaCl was ascribed to a high concentration of Na⁺ and not to the concentration of Cl or to a high osmolarity given that Hsc82-overexpressing cells are hypersensitive to 1 M sodium acetate (data not shown) but not to 1 M KCl (Fig. 1B), 0.5 M MgCl₂ (data not shown), or 2 M sorbitol (Fig. 1B). The deficient growth in media containing 1 M NaCl or 0.2 M LiCl was suppressed via the addition of 0.2 M KCl (Fig. 1B). Hsc82-overexpressing cells are also sensitive to 6.0 mM MnCl₂ (data not shown). On the other hand, Hsc82-overexpressing cells are more resistant to 0.6 M CaCl₂ than control cells (Fig. 1B). The salt sensitivities of a cnb1 strain of MCY3-1C cells were not affected by Hsc82 overexpression (data not shown). Taken together, the above findings indicate that the phenotypes of Hsc82-overexpressing cells as to sensitivity to high-salt media and concomitant calcium tolerance are similar to those of calcineurin-defective mutants (27, 32).

View larger version (78K): [in this window] [in a new window]   FIG. 1.   Effects of overexpression of Hsc82 on stress sensitivity. (A) Quantification of Hsp82 and Hsc82 expression in control (YPH500/pYO326) and Hsc82-overexpressing (YPH500/pYO326-HSC82) cells. Lanes 1 and 2, expression levels of Hsp82 and Hsc82 in control cells incubated with or without canavanine (100 µg/ml), respectively; lanes 3 and 4, expression levels of Hsp82 and Hsc82 in Hsc82-overexpressing cells incubated with or without canavanine (100 µg/ml), respectively. Lysates were prepared from control and Hsc82-overexpressing cells that had been cultured in SC-U medium with or without canavanine for 3 days at 22°C. Samples equivalent to proteins from 107 cells were resolved by SDS-PAGE and Western blotted with an anti-Hsp82/Hsc82 antibody. (B) Hypersensitivity of Hsc82-overexpressing cells to various stresses. Control and Hsc82-overexpressing cells were grown overnight in SC-U medium, appropriately diluted, and spotted on SC-U plates with or without the indicated reagents. The plates were incubated at 18°C for 4 days (control and 100 µg canavanine/ml), at 30°C for 2 days after being exposed to 50°C for 30 min (heat shock) and at 36°C for 2 days (others).

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Hypersensitivity of Hsc82-overexpressing cells to various stresses determined in liquid cultures. Control (YPH500/pYO326; solid squares) and Hsc82-overexpressing (YPH500/pYO326-HSC82; solid circles) cells were grown overnight in SC-U medium, appropriately diluted, and then incubated at 50°C in SC-U (A), in SC-U containing 15% (vol/vol) ethanol (Et-OH) at 30°C (B), or in SC-U containing 2.5 M NaCl at 30°C (C) for the indicated periods. Both control cells (open squares) and Hsc82-overexpressing cells (open circles) were pretreated at 37°C for 1 h and exposed to stress for the indicated periods. CFUs were then determined by plating on YPD plates at 30°C.

Suppression of hypersensitivity of Hsc82-overexpressing cells to salt stress by co-overexpression of the catalytic subunit of calcineurin, Cna2. The phenotypic similarity between Hsc82-overexpressing cells and calcineurin-defective cells prompted us to examine both the genetic and physical interactions between Hsc82 and calcineurin. Given that the expression of CNA2, encoding the catalytic subunit of calcineurin, suppresses phenotypes associated with calcineurin-defective strains, we examined whether or not an overexpression of Cna2 affects the phenotypes of Hsc82-overexpressing cells. We observed that the overexpression of Cna2 under the control of its own promoter suppressed the hypersensitivity of Hsc82-overexpressing cells to 1.0 M NaCl (data not shown) or 0.2 M LiCl (Fig. 3). However, the overexpression of Cna2 neither lowered the expression level of Hsc82 (data not shown) nor compensated for the hypersensitivity to canavanine or heat (Fig. 3). When cna2^H105Q, a mutant encoding an amino acid substitution near the metal-binding site (28), was used instead of CNA2, no suppression of hypersensitivity was observed (Fig. 3). Overexpression of the regulatory subunit of calcineurin, Cnb1, did not suppress the hypersensitivity (data not shown).

View larger version (65K): [in this window] [in a new window]   FIG. 3.   Suppression of the hypersensitivity of Hsc82-overexpressing cells to LiCl by co-overexpression of the catalytic subunit of calcineurin. Shown are the effects of co-overexpression of Cna2. Four strains, YPH500/pYO324 pYO326 (two-vector plasmids), YPH500/pYO324-HSC82 (Hsc82 overexpression plasmid) pYO326, YPH500/pYO324-HSC82 pYO326-CNA2 (Cna2 overexpression plasmid), and YPH500/pYO324-HSC82 pYO326-cna2H105Q (a Cna2 mutant overexpression plasmid), were used. Cells of these strains were grown overnight in SC-U,W medium, appropriately diluted, and spotted on SC-U,W plates with or without the indicated reagents (Can., canavanine) and incubated at 36°C for 2 days (control and 0.2 M LiCl) or 18°C for 4 days (100 µg of canavanine/ml).

Reduction in calcineurin-dependent gene expression by overexpression of Hsc82. The in vivo activity of calcineurin can be assessed by determining calcineurin-dependent gene expression. Repeated CDRE (calcineurin-dependent response element) of FKS2, a calcineurin-dependent gene, was fused to the 5' end of LacZ and expressed in the cells to be examined. Plasmids, pAMS363 (two copies of CDRE) (43) and pAMS366 (four copies of CDRE) (43) were separately transfected into wild-type and Hsc82-overexpressing cells, and the transfected cells were exposed to 0.2 M CaCl₂. Calcineurin-dependent gene expression, as indicated by the activity of -galactosidase, was significantly reduced by the overexpression of Hsc82, whereas the expression of pGPD-LacZ was not affected (Fig. 4). No -galactosidase activity was detected when the plasmids (pAMS363 and pAMS366) were not transfected (data not shown). CDRE-dependent gene expression was abolished in the presence of FK506, a calcineurin inhibitor (Fig. 4). These results suggest that the catalytic subunit of calcineurin, Cna2, might be bound to and sequestered by an excess of Hsc82.

View larger version (21K): [in this window] [in a new window]   FIG. 4.   Effects of Hsc82 overexpression on calcineurin-dependent gene (two and four copies of CDRE) expression. Control (YPH500/pYO324) and Hsc82-overexpressing (YPH500/pYO324-HSC82) cells were transformed with pAMS363 (two copies of CDRE) or pAMS366 (four copies of CDRE) and grown to mid-log phase in H-U,W medium. The cells were appropriately diluted and incubated at 30°C for 6 h in HS-U,W (pH 5.0) medium with or without CaCl2 (0.2 M) or FK506 (1 µg/ml). -Galactosidase was also expressed under the control of the GPD promoter in wild-type and Hsc82-overexpressing cells. The cells were exposed to CaCl2 and FK506 (1 µg/ml), and the -galactosidase activity was determined as described above.

Association of Cna2 with Hsp82 and Hsc82. To examine the physical association between Cna2 and Hsp82 and Hsc82 in vivo, HA-CNA2 was constructed and expressed in the cells to be examined. Cells arrested in a high-salt medium regained growth by simultaneous overexpression of Cna2 or HA-Cna2 and Cnb1, indicating that HA-Cna2 is as functional in cells as Cna2 (data not shown). HA-CNA2 was transfected into control and Hsc82-overexpressing cells. The overexpression of Hsc82 did not reduce the expression of HA-Cna2 (data not shown). Cells of the transfectants were cultured and then exposed to 1 M NaCl stress, after which cell extracts were prepared. Cna2 was immunoprecipitated from cell extracts with an anti-HA antibody. Immunoprecipitates were then resolved by SDS-PAGE and blotted with anti-HA and anti-Hsp82 antibodies. The results showed that Hsp82 and Hsc82 were coimmunoprecipitated with Cna2 from cell extracts of normally growing control cells transfected with HA-CNA2, whereas coimmunoprecipitation from extracts of cells incubated in 0.75 M NaCl was decreased (Fig. 5A). No coimmunoprecipitation of Hsp82 and Hsc82 was seen by using the anti-HA antibody from cell extracts of normally growing control cells transfected with the control plasmid, which did not contain HA-Cna2 (data not shown). In contrast, Hsc82 was coimmunoprecipitated with Cna2 from extracts of both NaCl-stressed and nonstressed Hsc82-overexpressing cells that had been transfected with HA-CNA2 (Fig. 5A). Essentially the same results were obtained when immunoprecipitation was performed with the anti-Hsp82/Hsc82 antibody instead of the anti-HA antibody (Fig. 5B). The coimmunoprecipitation of Cna2 and Hsp82 or Hsc82 was not detected in extracts of either control or Hsc82-overexpressing cells when these cells were treated with geldanamycin (Fig. 5B). Treatment of cells with geldanamycin reduced calcineurin-dependent gene expression, however (Fig. 5C), indicating that dissociation of Cna2 from Hsc82 did not always activate Cna2. We found that the amount of Cna2 was decreased upon treatment of cells with geldanamycin (Fig. 5D). Although calcineurin-dependent gene expression was inhibited by FK506 (Fig. 4B), treatment of cells with FK506 did not affect the results of coimmunoprecipitation using any strain (data not shown). The above results support the notion that Cna2 forms a complex with Hsp82 and Hsc82 in normally growing control cells but that the association is disrupted when these cells are exposed to NaCl stress. For Hsc82-overexpressing cells, the association between Cna2 and Hsc82 is not disrupted even when these cells are exposed to stress, probably as a result of the excess of Hsc82 present in cells. Treatment of cell extracts from either control or Hsc82-overexpressing cells with an excess amount of calmodulin in the presence of 2 mM CaCl₂ decreased the coimmunoprecipitation of Cna2 and Hsc82 (Fig. 5E).

View larger version (17K): [in this window] [in a new window]   FIG. 5.   Coimmunoprecipitation of Hsp82 and Hsc82 with HA-Cna2. (A) Coimmunoprecipitation disappeared in extracts from control cells exposed to NaCl stress, whereas it was still observed in extracts from Hsc82-overexpressing cells exposed to stress. Control (YPH500/pYO324 pRS316-pGPD-HA-CNA2) and Hsc82-overexpressing (YPH500/pYO324-HSC82 pRS316-pGPD-HA-CNA2) cells were grown to mid-log phase in SC-U,W medium and appropriately diluted in SC-U,W medium with or without NaCl (0.75 M) at 30°C for 4 h. Lysates were prepared from these cells and subjected to immunoprecipitation (IP) with an anti-HA antibody. Immunoprecipitates were resolved by SDS-PAGE and blotted with anti-HA and anti-Hsp82/Hsc82 antibodies. (B) Disappearance of the coimmunoprecipitation by treatment of cells with geldanamycin. Control cells (see panel A) were grown to mid-log phase in SC-U,W medium and treated with various concentrations of geldanamycin (GA) in SC-U,W medium for 1 h at 30°C. Lysates were prepared from these cells and subjected to immunoprecipitation with the anti-Hsp82/Hsc82 antibody. Immunoprecipitates were resolved by SDS-PAGE and blotted with anti-HA and anti-Hsp82/Hsc82 antibodies. Only data for control cells are shown, but the results were essentially the same for Hsc82-overexpressing cells. (C) Effects of geldanamycin on calcineurin-dependent gene expression. Wild-type cells (YPH499/pAMS363 or YPH499/pAMS366) were grown in HS-U medium overnight at 30°C, appropriately diluted, and incubated for 6 h at 30°C in HS-U (pH 5.0) medium containing 0.2 M CaCl2 in the presence or absence of geldanamycin (18 µM). The -galactosidase activity of these cells was determined as described in Materials and Methods. (D) Decrease in the quantity of HA-Cna2 by treatment of cells with geldanamycin. The cell lysates, equivalent to proteins in panel B, were resolved by SDS-PAGE and blotted with the anti-HA antibody. (E) Abolishment of the coimmunoprecipitation by treating cell lysate with bovine calmodulin (CaM). Control cells (see panel A) were grown to mid-log phase in SC-U,W medium at 30°C, after which lysates were prepared from these cells and subjected to immunoprecipitation with the anti-HA antibody. CaCl2, to a final concentration of 2 mM, and, when indicated, calmodulin were added to the preabsorbed lysates 1 h after the addition of anti-HA antibody. The mixtures were incubated at 4°C for 1 h, after which immune complexes were collected. Concentrations of calmodulin were 0 (lane 2), 0.4 (lane 3), and 4 µg/ml (lane 4). Immunoprecipitates were resolved by SDS-PAGE and blotted with anti-HA and anti-Hsp82/Hsc82 antibodies.

View larger version (20K): [in this window] [in a new window]   FIG. 6.   (A) Reduction of calcineurin activity of cells expressing HA-Cna2 by overexpression of Hsc82 and its suppression by further co-overexpression of Cna2. HA-Cna2 (YPH499/pRS314-pCNA2-ha-CNA2) cells, HA-Cna2 cells overexpressing Hsc82 (YPH499/pRS314-pCNA2-ha-CNA2 pYO326-HSC82), and cells co-overexpressing HA-Cna2 and Hsc82 (YPH499/pYO324-pCNA2-ha-CNA2 pYO326-HSC82) were grown to mid-log phase in SC-U,W medium. The cells were appropriately diluted and incubated at 30°C for 4 h in SC-U,W (pH 5.5) medium with 0.75 M NaCl. From these cells, cell lysates were prepared. Complexes containing HA-Cna2 were immunoprecipitated by anti-HA antibodies from the cell lysates, after which phosphatase activities associated with the immune complexes were determined as described in Materials and Methods with or without EGTA (5 mM) and presented as averages of three independent determinations (± standard deviations). The amount of HA-Cna2 was determined by Western blotting. (B) Suppression by Ca2+-calmodulin of the inhibitory effect of excess Hsc82 on calcineurin activity. Cell lysates were prepared from control and Hsc82-overexpressing cells as for panel A, in which HA and calcineurin were coexpressed, and incubated for 1 h with 4 µg of calmodulin (CaM)/ml. Immunoprecipitation was performed as described for panel A. Calcineurin activities associated with the immune complexes were determined as described in Materials and Methods.

Requirement of Hsp82 and Hsc82 for activation of calcineurin. Although steroid hormone receptors are inactive as transcription factors in complexes with HSP90, they have high-affinity ligand-binding sites only in the complexes, and thus HSP90 is necessary for steroid hormone receptor functions (36). The above results strongly suggest that calcineurin in Hsp82 complexes is as inactive as steroid hormone receptors. We next examined whether Hsp82 and Hsc82 are required for calcineurin functions in S. cerevisiae. A correlation between the activity of calcineurin and the expression levels of Hsp82 and Hsc82 was determined using the following experimental system. We constructed a strain in which the expression of Hsc82 is controlled by the Gal1 promoter and both endogenous HSP82 and HSC82 were disrupted (JH82DD/pRS315-1GAL10-HSC82). Then, 2% (wt/vol) mixtures of galactose and glucose were used as carbon sources in culture media. In these cells, the GAL1 promoter is activated as a function of galactose concentrations, and, therefore, the intracellular concentration of Hsc82 decreases as the concentration of galactose in media is decreased. The expression of Hsc82 was greater than 60% of the normal expression level in HSP82 HSC82 cells growing in SC medium when 1.8% galactose plus 0.2% glucose were used as a carbon source mixture; it was reduced as the fraction of galactose was further decreased (Fig. 7A). The cells expressed approximately 1/17 of the Hsc82 expressed by wild-type cells and grew very slowly in medium containing 1% galactose and 1% glucose (Fig. 7B). Cells of a wild-type strain grown in media containing different galactose/glucose ratios did not show different sensitivities to LiCl, suggesting that the sensitivity was not affected by catabolite repression (Fig. 7B). These results clearly demonstrated that cells expressing lower levels of Hsc82 than wild-type cells are more sensitive to 0.2 M LiCl (Fig. 7B) or 1 M NaCl (data not shown).

View larger version (45K): [in this window] [in a new window]   FIG. 7.   Requirement of Hsp82 and Hsc82 for resistance against LiCl stress and calcineurin-dependent gene expression. (A) Galactose-dependent expression of Hsc82 in hsp82 hsc82/pGAL1:HSC82 cells (YJH82DD/pRS315-1GAL10-HSC82). YJH82DD cells were grown overnight in H-L,H,W medium containing various ratios of galactose and glucose, appropriately diluted, and cultured in the same medium for 6 h. Lysates equivalent to proteins from these cells were resolved by SDS-PAGE and blotted with the anti-Hsp82/Hsc82 antibody. The amount of Hsc82 expressed in these cells is expressed as a percentage of Hsp82 and Hsc82 expressed in wild-type cells grown in SC medium. (B) Hypersensitivity to LiCl stress of cells expressing Hsc82 at low levels. YPH499 (HSP82 HSC82/pRS313 pRS314 pRS315-1GAL10) and YJH82DD (hsp82 hsc82/pRS315-1GAL10-HSC82) cells were separately cultured overnight in HGal-L,H,W medium, diluted, and spotted on YP plates containing various ratios of galactose and glucose with or without LiCl (0.2 M). The plates were incubated at 30°C for 2 days. (C) Requirement of Hsc82 for calcineurin-dependent gene expression. YJH82DD cells were transformed with pAMS363 (two copies of CDRE; open squares), pAMS366 (four copies of CDRE; solid squares) and pYO326-pGPD-lacZ (pGPD-lacZ; open circles). The transformants were grown overnight in HS-L,H,W,U medium containing various ratios of galactose and glucose and diluted in HS-L,H,W,U medium (pH 5.0) containing various ratios of galactose and glucose with 0.2 M CaCl2 for 6 h at 30°C. Then the -galactosidase activity was determined as described in Materials and Methods. The relative amount of -galactosidase activity (percentage of that for control cells in the same medium) was plotted against the amount of Hsc82 (percentage of Hsp82 plus Hsc82 in wild-type cells).

hsc82

HSP82 hsc82

View larger version (25K): [in this window] [in a new window]   FIG. 8.   Effects of ts mutations of Hsp82 on salt sensitivity. (A) Hypersensitivity to 0.2 N LiCl and tolerance of 0.5 M CaCl2 of ts hsp82 mutants. ts hsp82G170D hsc82 (YOK5H) and hsp82S485Y hsc82 (YOK9H) mutants were examined. Cells of the ts mutants were grown in YPD medium overnight at 25°C, diluted, spotted on YPD plates with or without LiCl (0.2 M) or CaCl2 (0.5 M), and incubated at 25°C for 3 days. (B) Reduction in calcineurin-dependent gene expression in a ts hsp82 mutant. Wild-type (YPH500; bars 1), HSP82 hsc82 (Y500-1H; bars 2), hsp82G170D hsc82 (YOK5H; bars 3), and hsp82S485Y hsc82 (YOK9H; bars 4) cells were transformed with plasmid pAMS363 (two copies of CDRE) or pAMS366 (four copies of CDRE). Cells of each transformant were grown in HS-U medium overnight at 25°C, appropriately diluted, and incubated for 6 h at 25°C in HS-U (pH 5.0) medium with or without CaCl2 (0.2 M). The -galactosidase activity of these cells was determined as described in Materials and Methods.

Instability of the calcineurin catalytic subunit, Cna2, in cells expressing reduced levels of Hsc82. As shown above, the activity of calcineurin, as determined by calcineurin-dependent gene expression, was greatly reduced in cells expressing low levels of Hsc82. This may be the consequence of a reduced quantity of Cna2 in these cells. Alternatively, Cna2 may be rendered inactive by forming complexes with proteins other than Hsc82, and furthermore the association may not be disrupted upon receipt of activation signals. To examine these possibilities, the amount of Cna2 in cells expressing a reduced level of Hsc82 was determined and compared with that in cells expressing a normal level of Hsc82. HA-Cna2 was expressed under the control of GPD promoter in a hsp82 hsc82/pGAL1:HSC82 strain (YJH82DD/pRS315-1GAL10-HSC82). Cells of this strain were separately cultured in media containing 2% mixtures of galactose and glucose. Extracts were prepared from these cells, resolved by SDS-PAGE, and blotted with the anti-HA antibody (Fig. 9A). A significant amount of Cna2 was detected in extracts from cells cultured in 2% galactose medium, whereas only a trace of Cna2 was detected in extracts from those cultured in 1.6% galactose and 0.4% glucose medium. Exposure of these cells to 1 M NaCl stress did not alter the results.

View larger version (20K): [in this window] [in a new window]   FIG. 9.   Effects of Hsp82 on stability of Cna2 and Cnb1. (A) Destabilization of Cna2 in cells expressing reduced levels of Hsc82. hsp82 hsc82/pGAL1:HSC82 cells (YJH82DD/pRS315-p1GAL10-HSC82 pRS316-pGPD-ha-CNA2) were grown overnight in H-L,H,W,U medium containing various ratios of galactose and glucose, appropriately diluted, and cultured in the same medium for 6 h at 30°C. A cell lysate equivalent to proteins from each culture was resolved by SDS-PAGE and blotted with anti-HA and anti-Hsp82/Hsc82 antibodies for determination of the amounts of Cna2 and Hsc82, respectively. (B) Destabilization of Cna2 in ts hsp82 cells. Lysates equivalent to proteins of hsp82 (Y500-1H/pRS316-pGPD-ha-CNA2; lane 1) and ts (YOK5H/pRS316-pGPD-ha-CNA2; lane 2) cells grown to mid-log phase in SC-U medium at 25°C were resolved by SDS-PAGE and blotted with anti-Hsp82/Hsc82 and anti-HA antibodies, respectively. (C) No effect of ts hsp82 on Cnb1 was observed. Lysates equivalent to proteins of hsp82 (Y500-1H/316-pGPD-CNB1; lane 1) and ts (YOK5H/316-pGPD-CNB1; lane 2) cells grown to mid-log phase in SC-U medium at 25°C were resolved by SDS-PAGE and blotted with the anti-Cnb1 antibody.

hsc82

hsc82

hsc82

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

HSP90 is a major cytosolic molecular chaperone that has been shown to protect denatured proteins from irreversible aggregation in vitro (16, 47). Higher concentrations of HSP90 (Hsp82 and Hsc82) are required for growth of yeast cells at higher temperatures (2). A CHO variant which expresses HSP90 at a level severalfold higher than normal is relatively heat resistant compared to the parental strain (49). It is plausible that HSP90 functions like HSP70 in vivo by protecting cells from damage caused by various stresses including elevated temperatures. For this reason, we expected that overexpression of either Hsp82 or Hsc82 in S. cerevisiae might confer stress tolerance on this organism. On the contrary, as was shown above, a strain overexpressing Hsc82 approximately 20-fold was hypersensitive to various stresses including canavanine, heat shock, and high salt. These results are consistent in part with the report by Cheng et al. (6), wherein overexpression of Hsp82 in yeast caused strain-dependent reductions in growth at 37.5°C and in thermotolerance. We observed that preconditioned Hsc82-overexpressing cells acquired thermotolerance in the same manner as wild-type cells did, indicating that the induced stress tolerance mechanism suppressed the hypersensitivity. These preheated, Hsc82-overexpressing cells were still hypersensitive to high salt stress, however, suggesting that the molecular mechanism operating in salt stress differs at least in part from that operating in thermotolerance.

Hsc82-overexpressing cells are also hypersensitive to 1 M NaCl and 0.2 M LiCl, but not to 1 M KCl or 1 M NaCl plus 0.2 M KCl. These results and those for the sensitivity to 2 M sorbitol and 0.6 M CaCl₂ suggest that Hsc82-overexpressing cells are similar in their phenotypes to a calcineurin-defective mutant (27, 32). In fact, we observed that calcineurin-dependent gene expression was reduced. We found that co-overexpression of the catalytic subunit of calcineurin, Cna2, in Hsc82-overexpressing cells at least partially suppresses the hypersensitivity. This raised the possibility that Cna2 not only genetically but also physically interacts with Hsc82, which turned out to be the case. Coimmunoprecipitation experiments clearly demonstrated that, when cells are exposed to salt stress, Cna2 is dissociated from complexes of Cna2 and Hsc82 in wild-type cells but that Cna2 still exists as a complex with Hsc82 in Hsc82-overexpressing cells. As the complexes of steroid receptors and a complex of pp60^v-src with HSP90 are inactive as transcription factors and as a tyrosine protein kinase, respectively (23), Cna2 in a complex with Hsc82 was shown in vitro to be inactive as a protein phosphatase. Thus, the overexpression of Hsc82 caused hypersensitivity to stresses in yeast at least in part by sequestering Cna2 from its activation process. This interpretation was supported by the fact that the hypersensitivity of cnb1 cells to salt stresses was not affected by overexpression of Hsc82. Co-overexpression of Cna2 in Hsc82-overexpressing cells suppressed the hypersensitivity to salt stress but not to canavanine or heat shock. This suggests that calcineurin is involved in the protection of cells from salt stress, but not from canavanine or heat shock, and that other Hsc82 substrates, whose functions are necessary for survival under other types of stress, are also sequestered by an excess of Hsc82. Consistent with these results, cna1 cna2 cells are not hypersensitive to heat shock or canavanine (unpublished results). Although yeast cells treated with the immunosuppressing drug FK506 have salt-sensitive phenotypes similar to those of calcineurin-deficient strains (32), our results showed that FK506 did not affect the complex formation of Cna2 and Hsc82 but reduced calcineurin-dependent gene expression. Therefore, the drug is likely to affect calcineurin after it is dissociated from Hsp82 and Hsc82.

HSP90 was initially thought to simply suppress the function of steroid receptors (23, 37) and pp60^v-src (3, 23). Genetic analysis of yeast (1, 35, 48) and reconstitution experiments in vitro (39) have demonstrated that HSP90 is also a positive regulator of steroid receptors and pp60^v-src. Thus, we examined whether Hsc82 is simply a negative regulator of calcineurin or, alternatively, whether the complex formation of Cna2 with Hsc82 is necessary for activation of Cna2 if Hsc82 is appropriately expressed. We demonstrated that underexpression of Hsc82 increased susceptibility to stress and reduced NaCl-induced, calcineurin-dependent gene expression. The present results also suggested that Cna2 was unstable and susceptible to proteolytic degradation in Hsc82-deficient cells. Similarly, when pp60^v-src was expressed in yeast cells, the expression of the protein was significantly reduced by lowering the levels of Hsp82 and Hsc82 expression (48). Furthermore, ts hsp82 mutant strains were hypersensitive to stress and Cna2 was decreased in these mutants even under nonstressful conditions. NaCl-induced, calcineurin-dependent gene expression was also reduced in a ts hsp82 mutant even at the permissive temperature. This effect of the hsp82 mutation appeared to be more severe on calcineurin than on glucocorticoid receptors and pp60^v-src, given that the effects on the last were not detected at the permissive temperature (18, 33). Geldanamycin, an HSP90-specific inhibitor, dissociates Cna2 from Hsc82 and Hsp82 but does not activate it because Cna2 is destabilized as a consequence of the dysfunction of Hsc82 and Hsp82. Taken as a whole, the above findings suggested that, like steroid receptors and pp60^v-src, newly translated Cna2 binds to and forms a complex with Hsp82 and Hsc82 likely because it has a partially unstable structure when present alone (19). As revealed by coimmunoprecipitation experiments, Hsp82 and Hsc82 form complexes with and stabilize Cna2 until cells are exposed to stress. The results also demonstrated that dissociation of the complexes leading to the activation of calcineurin involves mobilization of Ca²⁺ to activate calmodulin. An excess of Hsc82 probably competitively interferes with the binding of Ca²⁺-calmodulin to Cna2, thus inhibiting the activation of calcineurin. In addition, stress must mobilize preexisting Hsp82 and Hsc82, which prevents harmful aggregation of denatured proteins, causing a shift in the equilibrium that dissociates the complexes. However, this equilibrium shift is not sufficient for the activation of calcineurin because neither canavanine nor heat shock activated calcineurin (unpublished observations). The present finding is consistent in part with the report by Someren et al. (42), in that Hsp90 interacts with the catalytic subunit of calcineurin in vivo. However, as described above, we observed that HSP90 stabilizes, but does not activate, calcineurin; this is inconsistent with the notion that HSP90 increases calcineurin activity in a dose-dependent manner (42). The discrepancy remains to be elucidated.