The Peptidyl-prolyl Isomerase Domain of the CyP-40 Cyclophilin Homolog Cpr7 Is Not Required to Support Growth or Glucocorticoid Receptor Activity in Saccharomyces cerevisiae*

CyP-40 cyclophilins are found in association with molecular chaperone Hsp90·steroid receptor complexes. The amino-terminal portion of these cyclophilins harbors the characteristic peptidyl-prolyl isomerase (PPIase) domain, whereas three copies of the tetratricopeptide (TPR) motif, a structure shown to be involved in protein-protein interactions, and a putative calmodulin-binding domain are located in the carboxyl-terminal half of the protein. The TPR domains mediate binding to Hsp90, but a requirement for the PPIase domain has not been established. To address this, we have investigated the effects of mutations that alter the PPIase domain of the Saccharomyces cerevisiae CyP-40 homolog, Cpr7. Because Cpr7 is required for rapid growth and full Hsp90 activity, a functional assessment of the PPIase domain could be performed in vivo. A mutation in the catalytic domain altering a conserved site predicted to be essential for isomerase activity did not compromise Cpr7 function. Furthermore, deletion of the entire PPIase domain did not significantly affect growth or Hsp90-mediated steroid receptor activity. These results indicate that the TPR-containing carboxyl terminus of Cpr7 is sufficient for fundamental Cpr7-dependent activity.

The molecular chaperone Hsp90 can serve as a general chaperone in vitro when present at high concentration by prevent-ing aggregation of a variety of substrates (1)(2)(3)(4)(5)(6)(7)(8), but its role in vivo appears to be restricted to the maturation of a subset of proteins involved in signal transduction pathways (9 -12). The best characterized Hsp90-dependent substrates are members of the steroid receptor family. Maturation of steroid receptor complexes involves a dynamic assembly pathway in which the receptor first enters into an intermediate complex containing Hsp90, Hsp70, p60 (Hop), and p48 (Hip) (13)(14)(15) and is later found in a mature complex containing Hsp90, low levels of Hsp70, p23, and one of three immunophilins (15). The mature complex is required for the steroid receptor to achieve a conformation that enables it to bind and respond to hormone.
Three immunophilins, FKBP51, FKBP52, and Cyp-40, have been identified as components of distinct steroid hormone receptor complexes. FKBP51 and FKBP52 belong to the FK506binding protein family, whereas Cyp-40 is a cyclophilin (15,16). In addition to binding to their respective immunosuppressive drugs (FK506 or cyclosporin A), most immunophilins tested thus far exhibit peptidyl-prolyl isomerase (PPIase) 1 activity in vitro (17). Because these proteins are found in Hsp90-steroid receptor complexes in organisms as diverse as mammals and yeast, it would seem likely that prolyl isomerization plays a role in governing the activity of the complex. In vitro experiments have shown that CyP-40 can act as a molecular chaperone by maintaining a substrate in a non-native folding-competent intermediate (18). However, isomerization does not appear to be required for this function because treatment with cyclosporin A, which inhibits isomerase activity of cyclophilins in vitro, did not have an effect in this assay (18). Furthermore, biochemical experiments have suggested that the isomerase activity of immunophilins is not required for reconstitution of the glucocorticoid receptor in mature heterocomplexes (19). Thus, the in vivo role of the PPIase activity of immunophilins in Hsp90-dependent events, if any, remains unknown.
In addition to the PPIase domain, Hsp90-associated immunophilins harbor three units of the tetratricopeptide repeat (TPR), a degenerate 34-amino acid motif involved in proteinprotein interactions (20). FKBP52 and CyP-40 interact directly with Hsp90 through this domain (21,22) by binding to what appears to be a single TPR-accepting pocket on Hsp90 (19,23). Two additional components of steroid receptor-Hsp90 complexes, PP5 (a protein phosphatase) and p60, also contain TPR motifs and bind to the same site on Hsp90 as do the immunophilins (24,25). Interestingly, overexpression of the TPR domain of PP5 results in a dominant negative effect that inhibits glucocorticoid receptor-mediated transactivation (26), supporting a function for TPR-containing proteins in steroid receptor signaling. Some of the TPR proteins that interact with Hsp90 have been proposed to target Hsp90 substrates to their cellular destination (24).
Two Saccharomyces cerevisiae CyP-40 homologs, Cpr6 and Cpr7, have been identified (27,28) and shown to interact directly with Hsp90 via their TPR-containing carboxyl termini (29). Disruption of CPR7 confers a modest decrease in the rate of cell growth (28,30), but this effect is severe when combined with mutations that either decrease the levels of Hsp90 or disrupt STI1, the gene encoding the yeast homolog of p60 (29). CPR7 is also required for full activity of two heterologous Hsp90 substrates, the glucocorticoid receptor (GR) and the oncogenic tyrosine kinase p60 v-src (29). In this report we test the requirement for the PPIase domain in Cpr7. We conclude that fundamental functions associated with Cpr7 do not depend on its role as a prolyl isomerase but instead require the TPR-containing carboxyl terminus.

EXPERIMENTAL PROCEDURES
Media-The standard genetic techniques and growth media used were those described in Sherman et al. (31).
Plasmid Construction-Plasmid pJM99, a centromeric plasmid expressing Cpr7 from the native CPR7 promoter, was constructed as follows. Oligonucleotides Cpr7gn1 (5Ј-AAATCTAGAGTAGGTGGATC-TAGTAACTTC-3Ј) and Cpr7gn4 (5Ј-AGTCTCGAGGACAGCGCAAAT-GCTACCTG-3Ј) were synthesized and used to amplify a region of the S. cerevisiae genome encompassing ϳ1 kb upstream and ϳ900 bp downstream of the CPR7 open reading frame (ORF) by polymerase chain reaction (PCR). The amplified product was treated with the restriction endonucleases XbaI and XhoI and subcloned into similarly treated vector pRS316 (32).
The Cpr7 R64A protein was expressed from plasmid pAADB163, which was constructed as follows. PCR primers Cpr7gn1 and Cpr7gn2 (5Ј-A-TCCCCGGGAACGACTGCATGAAAACCATTTCCCTTATAGCTCAGG-3Ј) were used to amplify a region of the S. cerevisiae genome that encompasses a region approximately 1 kb upstream from the start codon of the CPR7 ORF to ϳ200 bp into the CPR7 ORF. The amplified product was treated with XbaI and SmaI and subcloned into pRS316 digested with the same enzymes. A second DNA segment corresponding to a region from ϳ200 bp into the CPR7 ORF to ϳ900 bp downstream from the stop codon of the CPR7 ORF was amplified by PCR with primers Cpr7gn3 (5Ј-TGTCCCGGGTTTATGATCCAGGCAGGTGAC-3Ј) and Cpr7gn4. This PCR product was then digested with SmaI and XhoI and inserted into the same sites of pRS316, which already contained the XbaI/SmaI insert. The resulting plasmid expresses a mutant Cpr7 protein from the native CPR7 promoter with the following amino acid substitutions: R64A, K67P,and N68G.
Plasmid pAADB158 expresses the Cpr7 ⌬PPIase mutant protein from the CPR7 native promoter. The first step in its construction was to amplify the promoter region of CPR7 by PCR using primers Cpr7gn1 and Cpr7gn5 (5Ј-ATTGGATCCGCGGTTGGATGTAGGTTGTTG-3Ј) and to subclone the PCR product into pRS316 using XbaI and BamHI. Next, primers BamHIT7#1 (5Ј-CCCCCCGGATCCGACTGTGGAGTGT-GGGAAAAA-3Ј) and Cpr7gn4 were used to PCR amplify a DNA fragment encoding the carboxyl-terminal region of Cpr7 and containing ϳ900 bp of the 3Ј-untranslated region of CPR7. This PCR product was then treated with BamHI and XhoI and subcloned into pRS316, which already contained the XbaI/BamHI fragment. Plasmid pAADB158 expresses a mutant form of Cpr7, Cpr7 ⌬PPIase , that contains a deletion of amino acids 2-200.
Plasmid pAADB131 expresses the Cpr7 ⌬PPIase mutant protein from the GAL1 promoter and was created by inserting a BamHI/EcoRI fragment from pJM12 (29) into the yeast expression plasmid pYES2. The protein expressed from this plasmid encompasses the same region of Cpr7 as the protein encoded by pAADB158.
The Cpr7 ⌬PPIase -overexpressing plasmid used in the GR experiments (pAADB175) was constructed as follows. pAADB158 was treated with XbaI, and the 5Ј overhangs were filled in using the Klenow fragment of DNA polymerase I from Escherichia coli (Amersham Pharmacia Biotech). Next, the plasmid was treated with XhoI, releasing a ϳ2.5-kb fragment. This fragment was then subcloned into plasmid pRS426 (a multi-copy version of pRS316 (32)), which was previously digested with SmaI and XhoI.
The Cpr6-overexpressing plasmid (pJM90) used in the GR experiments was constructed as follows. PCR primers Cpr6xgst (5Ј-CCCCG-GATCCATGACTAGACCTAAAACTTTT-3Ј) and Cpr6-3Јxgpd (5Ј-AAA-GAGCTCTTGGGCAGGACGACATCATG-3Ј) were used to amplify a region of the S. cerevisiae genome beginning at the translational start site and ending approximately 1 kb downstream of the CPR6 ORF. The amplified fragment was treated with BamHI and SacI and subcloned into vector p2UGpd (a version of pRS426 that contains the promoter from the GPD gene). Overexpression of Cpr6 from this plasmid was confirmed by Western blot analysis.
Western Blot Analysis of Proteins-Plasmids expressing the wild-type or mutant forms of Cpr7 were introduced into S. cerevisiae cells using the lithium acetate method (33). For detection of Cpr7 proteins, cells were grown in YNB liquid medium (0.16% yeast nitrogen base, 0.5% ammonium sulfate, 0.76% KCl) containing all amino acids but lacking uracil and containing either 2% glucose or 2% galactose, as indicated. Cells were lysed as described (27), and an approximately equal amount of total cell lysate was subjected to SDS-polyacrylamide gel electrophoresis through 10% acrylamide. Equal sample loading was confirmed by Coomassie Blue staining of the gels. Separated proteins were then transferred to nitrocellulose membrane (Schleicher & Schuell). The Cpr7 proteins were probed on immunoblots using rabbit polyclonal antibodies raised against a region of Cpr7 from amino acids 193 to 393. Horseradish peroxidase-conjugated anti-rabbit antibodies (Promega) were then allowed to bind the anti-Cpr7 antibodies. Antibody detection was performed using the ECL Western blotting detection system (Amersham Pharmacia Biotech) following the manufacturer's guidelines.
Glucocorticoid Receptor Activity Assay-Cells deleted for CPR7 were transformed with a GR expression plasmid, p2A/GRGZ, which also harbors the lacZ gene under the control of GR elements (10). Plasmids expressing the wild-type or mutant cyclophilins were introduced into these cells, and the transformants were assayed for ␤-galactosidase activity as follows. Mid-logarithmic growing cells were grown to a density of ϳ1.5 ϫ 10 7 cells/ml in plasmid-selective medium containing 2% glucose were treated with saturating amounts (20 M) of deoxycortisone for 1 h at 25°C. Cells were immediately harvested and resuspended in lysis buffer (10 mM Tris, pH 7.3, 50 mM NaCl, 50 mM KCl, 10 mM MgCl 2 , 20% (w/v) glycerol, 1 mM dithiothreitol, 0.4 g/ml aprotonin (Sigma), 0.4 g/ml leupeptin (Sigma), and 2 mM 4-aminoethyl-benzensulfonyl fluoride (Sigma)). Cells were then lysed using acid-washed glass beads (B. Braun Biotech International) by vortexing at full speed for 30 s and placed on ice for 1 min (repeated for six cycles). ␤-Galactosidase activity was measured and normalized to total cellular protein essentially as described previously (34).

The PPIase Domain of Cpr7 Is Not Required for Normal
Growth-To begin our assessment of the role PPIase activity might play in the function of CyP-40 cyclophilins, we tested whether Cpr7 exhibits isomerase activity in vitro. Using a standard assay (35), isomerase activity was detected with the related cyclophilin Cpr6 2 (30, 36) but not with Cpr7 (30). 2 Similar observations have been reported for the bovine cyclophilin RanBP2, which showed only marginal isomerase activity in vitro (37). It is possible that the isomerase activity of Cpr7 was missed because the peptide substrate used in the assay does not sufficiently mimic natural Cpr7 substrates. Alternatively, Cpr7 may have evolved to carry out functions that are independent of peptidyl-prolyl isomerization.
To test whether PPIase activity is required for Cpr7 function in vivo, we first mutated the highly conserved arginine residue in the catalytic region (Arg 64 in Cpr7) to alanine (Fig. 1) and tested for the ability of the mutant protein (Cpr7 R64A ) to com-   1. Cpr7 proteins used in this study. Plasmids expressing the Cpr7 proteins were constructed as described under "Experimental Procedures." The relevant amino acid substitution within the PPIase domain of Cpr7 R64A is indicated with an asterisk. Cpr7 R64A also harbors two amino acid substitutions (K67P and N68G) at positions that have not been implicated in PPIase activity. The Cpr7 ⌬PPIase protein lacks the entire PPIase domain. In the experiments outlined in this work, Cpr7 ⌬PPIase is expressed from either the native CPR7 promoter or from the galactose inducible GAL1 promoter. Gray rectangles indicate the PPIase domain, hatched boxes represent TPR units, and black boxes represent the putative calmodulin-binding domain (28,43). Numbers below the diagrams indicate amino acid positions. plement a cpr7⌬ recipient. Enzymatic studies have shown that the analogous arginine in the human CyP-18 protein (Arg 55 ) is essential for PPIase activity in vitro (38), and the equivalent substitution in the retina-specific bovine cyclophilin RanBP2 (R143A) resulted in a dramatic in vivo defect in the processing of opsin in cone photoreceptor cells (39). Expression of Cpr7 R64A from the native CPR7 promoter produced a mutant protein with the same size and abundance as wild-type Cpr7 (Fig. 2).
We examined the ability of Cpr7 R64A to rescue the growth defect of Cpr7-deficient cells. The modest slow growth phenotype of cpr7⌬ cells (28) is greatly enhanced when combined with deletion of HSC82 (29), the major source of Hsp90 protein at normal temperatures (40). 3 Expression of Cpr7 R64A completely rescued the growth defects of both cpr7⌬ cells 2 and cpr7⌬ hsc82⌬ cells (Fig. 3). This result shows that the highly conserved arginine within the catalytic site of Cpr7 is not required for normal growth and suggested that PPIase activity is not involved in this Cpr7-dependent function.
To rigorously test this possibility, a plasmid that expresses a mutant form of Cpr7 in which the entire PPIase domain is deleted was constructed and tested for Cpr7 functions (Fig. 1). Western blot analysis confirmed that constructs with this deletion produced a protein of the anticipated size (Fig. 2). Expression of CPR7 ⌬PPIase from its native promoter resulted in partial rescue of the slow growth phenotype of cpr7⌬ hsc82⌬ cells (Fig. 3). Because Cpr7 ⌬PPIase is not expressed at the same levels as wild-type Cpr7 with this construct (Fig. 2), we tested whether increasing expression of the truncated protein would confer more complete rescue of the growth defect of cpr7⌬ hsc82⌬ cells. Expression of Cpr7 ⌬PPIase from the galactoseinducible GAL1 promoter resulted in abundant production of the truncated protein (Fig. 2) and conferred growth rates in cpr7⌬ hsc82⌬ cells that are indistinguishable from those of wild-type cells (Fig. 3). These results demonstrate that the TPR-containing carboxyl terminus of Cpr7 is sufficient to mediate normal Cpr7-dependent growth.

Expression of the TPR-containing Carboxyl Terminus of Cpr7 Increases Activity of the Glucocorticoid Receptor in cpr7⌬ Cells-
The molecular components of the chaperone machinery that associate with the GR are conserved between mammalian and yeast cells (27). Furthermore, when GR is expressed in S. cerevisiae it exhibits hormone-dependent signal transduction activity, and this activity depends on Hsp90 (41). We have previously shown that deletion of CPR7 results in 5-fold reduction in the activity of GR (29). To determine whether expression of the TPR-containing carboxyl terminus of Cpr7 is sufficient to mediate Cpr7-dependent GR function, we tested its ability to restore GR activity in cpr7⌬ cells. Expression of a wild-type copy of CPR7 from plasmid pJM99 in cpr7⌬ cells produced slightly less GR activity compared with cells harboring a chromosomal copy of CPR7. 2 Because Western blot analysis showed that the expression level of Cpr7 from this plasmid was essentially normal, 2 the modest decrease in GR activity is likely due to plasmid loss from a small percentage of the cells carrying the Cpr7-expressing plasmid during the growth of the culture. Expression of Cpr7 R64A FIG. 4. Expression of Cpr7 R64A or Cpr7 ⌬PPIase increases GR activity in cpr7⌬ cells. ␤-Galactosidase activity from a GR-dependent reporter gene in cpr7⌬ cells harboring plasmids expressing the indicated protein. Cells were grown to mid-logarithmic stage in glucose medium, exposed to hormone, and assayed as described under "Experimental Procedures." The plasmid used to overexpress Cpr7 ⌬PPIase in this assay (pAADB175) is different from the one used in the previous experiments (to confer overexpression of the protein in glucose media) and produces the truncated protein at levels comparable with the galactose-inducible plasmid. 2 Cpr6 is overexpressed from plasmid pJM90 (see "Experimental Procedures"). in cpr7⌬ cells resulted in GR activity comparable with that of cells expressing wild-type Cpr7 from a plasmid (Fig. 4). Furthermore, Cpr7 ⌬PPIase restored some GR activity (55%) when expressed from the native CPR7 promoter and produced even greater GR activity (81%) when it was overexpressed (Fig. 4). In contrast, overexpression of Cpr6, a highly related cyclophilin that also contains three TPR motifs, did not increase GR activity (Fig.  4), showing that the role played by the carboxyl half of Cpr7 in the regulation of GR is specific. DISCUSSION Two homologs of the mammalian cyclophilin CyP-40 occur in S. cerevisiae, but only one, Cpr7, has been shown to have a functional role in Hsp90-associated activity (29). Thus far, no phenotype has been associated with the deletion of CPR6 (28 -30, 36). Interestingly, in vitro assays have detected isomerase activity for Cpr6 2 (30,36) but not for Cpr7 2 (30). The results described here strongly suggest that the in vivo functions thus far associated with the yeast Cyp-40 homolog Cpr7, namely the maintenance of rapid growth rates and GR activity, are not dependent on PPIase activity but are mediated by the TPRcontaining carboxyl terminus. This is consistent with the observation that pharmacological inhibition of isomerase activity does not prevent reconstitution of salt-dissociated GR heterocomplexes in vitro (19). The TPR-containing domain is capable of mediating both physical (29) and functional (data herein) interactions with Hsp90.
What, then, is the functional role of the PPIase domain of Cpr7? The steady-state level of Cpr7 ⌬PPIase was less than that of wild-type Cpr7 using the same promoters and plasmids. Thus, it is possible that the PPIase domain was adventitiously acquired in evolution as a mechanism for stabilizing the TPRcontaining domain. If so, it seems unlikely that this is the only purpose of the PPIase domain because it has been highly conserved. The Cyp-40 member of the cyclophilin family is found in steroid receptor complexes from yeast to mammals. Moreover, a critical amino acid (Arg 64 ), which is required for PPIase activity in other cyclophilins, is not required to maintain normal expression levels of Cyp-40 in yeast. Yet it too has been conserved from yeast to mammals. If stabilization of the TPR domain was the sole function of the PPIase domain, greater sequence divergence would be expected over one billion years of evolution. We favor the possibility that the PPIase domain plays a role in Hsp90-mediated events, but a subtle one. Effects that are too subtle to detect in the types of assays we have employed here can nevertheless provide sufficient advantage for selection in evolution. For example, the PPIase domain might affect the rate at which other co-chaperones join or leave the Hsp90 complex, enhance the proper localization of the complex in the cell, or act on an unidentified substrate whose function is not relevant in our assays.
Whether such functions might involve catalyzing the isomerization of peptidyl-prolyl bonds is unclear. Although PPIase activity was not detected with Cpr7 in vitro, its activities might be restricted to particular substrates and therefore not be apparent in a standard in vitro peptide assay. Examples consistent with this scenario are presented by the bovine cyclophilin RanBP2 and the well characterized Drosophila cyclophilin homolog NinaA. RanBP2 exhibits only modest PPIase activity in vitro, but mutation of the site analogous to the R64A mutation in Cpr7 significantly decreased its function in vivo (39). Furthermore, although to our knowledge NinaA has not been reported to possess PPIase activity in vitro, mutations in NinaA that map to amino acid residues known to be crucial for PPIase activity in mammalian cyclophilins conferred strong ninaA mutant phenotypes in vivo (42).
In any case, with respect to growth control and GR activity in yeast, it is clear that the TPR-containing domain of Cpr7 plays a much stronger role than the PPIase domain. TPR domains have been identified in a number of proteins involved in varied cellular functions, including cell cycle control, transcriptional regulation, and steroid receptor regulation (15,20). In addition to a well established role in protein-protein interactions, recent reports have hypothesized that the TPR domains present in Hsp90-associated immunophilins are involved in subcellular targeting (24). How might the carboxyl terminus of Cpr7 function in Hsp90-dependent events? As suggested by others (24), the TPR domain of Cpr7 may indeed be involved in subcellular targeting of the Hsp90⅐steroid receptor complex. Alternatively, but not exclusively, we speculate that the TPR region of Cpr7 might cause a conformational change in Hsp90, enhancing its effectiveness in chaperoning steroid receptors and growth regulators into active or at least signal-responsive conformations.