The Chaperone-Associated Ubiquitin Ligase CHIP Is Able to Target P53 for Proteasomal Degradation*

Erschienen in: Journal of Biological Chemistry ; 280 (2005), 29. - S. 27443-27448 https://dx.doi.org/10.1074/jbc.M501574200

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 29, Issue of July 22, pp. 27443–27448, 2005 © 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. The Chaperone-associated Ubiquitin Ligase CHIP Is Able to Target p53 for Proteasomal Degradation*

Received for publication, February 10, 2005, and in revised form, May 13, 2005 Published, JBC Papers in Press, May 23, 2005, DOI 10.1074/jbc.M501574200

Claudia Esser‡, Martin Scheffner§, and Jo¨rgHo¨ hfeld‡¶ From the ‡Institute for Cell Biology and Bonner Forum Biomedizin, Rheinische Friedrich-Wilhelms-University Bonn, Ulrich-Haberland-Str. 61a, D-53121 Bonn, Germany and the §Department of Biology, Laboratory for Biochemistry, University of Konstanz , 78457 Konstanz, Germany

The cellular level of the tumor suppressor p53 is (12). Mdm2-mediated ubiquitylation targets p53 for degrada- tightly regulated through induced degradation via the tion by the 26 S proteasome and is of central importance for ubiquitin/proteasome system. The ubiquitin ligase establishing low p53 levels in normal cells (13, 14). The func- Mdm2 plays a pivotal role in stimulating p53 turnover. tional interplay between Mdm2 and p53 was elegantly demon- However, recently additional ubiquitin ligases have strated in gene knock-out studies, in which the embryonic been identified that participate in the degradation of lethality of mdm2 null mice was rescued by simultaneous de- the tumor suppressor. Apparently, multiple degradation letion of the p53 gene (15, 16). Stress-induced phosphorylation pathways are employed to ensure proper destruction of of p53 attenuates the interaction with Mdm2, leading to stabi- p53. Here we show that the chaperone-associated ubiq- lization and activation of the transcription factor (17). Intrigu- uitin ligase CHIP is able to induce the proteasomal deg- ingly, Mdm2 is itself a transcriptional target of p53, which radation of p53. CHIP-induced degradation was ob- establishes a negative feedback loop to terminate p53-mediated served for mutant p53, which was previously shown to stress responses (18). Additional mechanisms that regulate the associate with the chaperones Hsc70 and Hsp90, and for the wild-type form of the tumor suppressor. Our data Mdm2-p53 interplay include autoubiquitylation of Mdm2, the reveal that mutant and wild-type p53 transiently asso- association of Mdm2 with diverse binding partners, and alter- ciate with molecular chaperones and can be diverted ations of the intracellular localization of Mdm2 and p53 (8, 10, onto a degradation pathway through this association. 19, 20). Furthermore, Mdm2 not only mediates ubiquitylation of p53 but can also conjugate Nedd8 to the tumor suppressor to inactivate its transcriptional activity (7). The p53 tumor suppressor has been termed “the guardian of Despite the central role of Mdm2 in proteasomal targeting of the genome” (1). In normal cells p53 is present at low concen- p53, additional ubiquitin ligases were recently shown to par- tration. DNA damage and other stresses such as hypoxia cause ticipate in the degradation of the tumor suppressor in normal an accumulation of p53, which leads to cell cycle arrest or cells, including p300, Pirh2, and COP1 (21–23). Although p300 apoptosis (2, 3). p53 acts as a transcription factor to activate seems to cooperate with Mdm2 during ubiquitylation, Pirh2 target genes that are involved in these responses to prevent and COP1 trigger the destruction of p53 independent of Mdm2. damaged cells from proliferating and passing mutations on to Multiple degradation pathways apparently exist to maintain the next generation (4). Cells that lack functional p53 are low levels of p53 in normal cells. It is unclear whether these unable to respond appropriately to stress and are prone to degradation pathways are truly redundant or whether they are oncogenic transformation. In fact, missense mutations that selectively engaged in p53 destruction dependent on cell line- inactivate p53 are found in ϳ50% of all human tumors making age, developmental stage, or physiological situation. In any them the most frequent genetic alterations in cancer (5, 6). case, the complexity of p53 degradation may allow to integrate p53 is regulated through a variety of posttranslational mod- diverse signaling events through which p53 can be regulated. ifications, including phosphorylation, acetylation, and attach- We have recently identified a pathway for protein degrada- ment of ubiquitin, the small ubiquitin-like modifier SUMO and tion in the mammalian cytoplasm and nucleus that involves a the ubiquitin-like protein Nedd8 (4, 7, 8). Central to the regu- close cooperation of the molecular chaperones Hsc70 and lation of p53 is the ubiquitin ligase Mdm2 (8–10). Ubiquitin Hsp90 with the ubiquitin-proteasome system (24). Of central ligases (E3(s))1 provide specificity to ubiquitin conjugation as importance on this degradation pathway is the chaperone- they mediate the final step in the conjugation process, follow- associated ubiquitin ligase CHIP (25). Through binding to the ing the activation of ubiquitin by the E1 enzyme and its trans- carboxyl termini of Hsc70 and Hsp90, CHIP mediates the ubiq- fer onto a ubiquitin-conjugating (E2) enzyme (11). Mdm2 be- uitylation of chaperone-bound client proteins in conjunction longs to the RING finger E3s, which facilitate ubiquitylation by with E2 enzymes of the Ubc4/5 family and induces client deg- tethering the E2-ubiquitin complex to the substrate protein radation by the 26 S proteasome (26–28). Affected chaperone clients can be broadly divided into two subclasses: (i) Hsc70- and Hsp90-associated signaling proteins, for example the * The costs of publication of this article were defrayed in part by the glucocorticoid hormone receptor (26, 29) and the oncogenic payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to receptor tyrosine kinase ErbB2 (30) and (ii) aggregation- indicate this fact. prone proteins that are subjected to chaperone-assisted qual- ¶ To whom correspondence should be addressed. Tel.: 49-228-735308; ity control, such as misfolded cystic fibrosis transmembrane Fax: 49-228-735302; E-mail: [email protected]. conductance regulator (31, 32) and hyperphosphorylated tau 1 The abbreviations used are: E3, ubiquitin-protein isopeptide ligase; (33, 34). However, the full range of cellular substrates of E2, ubiquitin carrier protein; E1, ubiquitin-activating enzyme; MOPS, 3-(N-morpholino)propanesulfonic acid; siRNA, small interfering RNA; CHIP remains to be explored. Remarkably, mice that lack GA, geldanamycin. CHIP develop apoptosis in multiple organs after environmen-

This paper is available on line at http://www.jbc.org 27443 Konstanzer Online-Publikations-System (KOPS) URL: http://nbn-resolving.de/urn:nbn:de:bsz:352-167692 27444 CHIP Induces p53 Degradation

tal challenge (35). This seems to reflect the role of CHIP in the conformational regulation of the heat shock transcription factor but may also mirror altered associations between the chaperone machinery and diverse apoptosis regulators in the

absence of CHIP (36). FIG.1. p53R175H forms stable complexes with Hsc/Hsp70 in Many transforming mutants of p53 display structural de- contrast to wild-type p53. H1299 cells were transiently transfected fects and were previously found associated with the molecular with pRC/CMV-p53wt (p53) and pcDNA3.1-p53R175H (p53R175H)as indicated. Control cells were untransfected. For immunoprecipitation a chaperones Hsc70 and Hsp90 in tissue culture cells and in ␣ 175 specific anti-p53 antibody ( -p53 IP) was used. After SDS-PAGE and human tumor specimens (37–39). For example, Arg muta- immunoblotting p53-associated Hsc/Hsp70 was detected using a poly- tions destabilize loop regions within the DNA binding domain, clonal anti-Hsc/Hsp70 antibody. To monitor expression levels, 25 ␮gof leading to partial unfolding and association with Hsc70 (40– protein extracts (ex.) were loaded. 44). In contrast to the findings for mutant p53, an interaction of ␮ ␮ wild-type p53 with molecular chaperones could not be detected ubiquitin, 1 g/ l ubiquitin-aldehyde, 10 mM ATP, 10 mM MgCl2,10mM by coimmunoprecipitation (39, 45). However, this does not ex- dithiothreitol, 10 mM phosphocreatine, and 10 mM creatine kinase. The total clude that the extended chaperone interactions observed for volume of the samples was adjusted to 20 ␮l with 25 mM MOPS, pH 7.2, 100 mutant p53 actually represent a pathological exaggeration of mM KCl, and 1 mM phenylmethylsulfonyl fluoride. Samples were incubated for2hat30°Candthen analyzed by SDS-PAGE and phosphorimaging. transient interactions between wild-type p53 and the cellular Reporter Gene Assays—H1299 cells were seeded in six-well plates and chaperone machinery, similar to observations made for other were transfected with 0.6 ␮g of a firefly luciferase reporter gene plasmid signaling proteins (46, 47). Wild-type p53 possesses an intrinsic (pp53-TA-Luc; BD Biosciences Clontech), 0.2 ␮g pRC/CMV-p53wt, and 1.2 conformational lability (48, 49), and may therefore undergo ␮g pcDNA3.1-CHIP as indicated. The total amount of added DNA was dynamic associations with molecular chaperones at a post- kept constant at 2.5 ␮g/well by addition of pcDNA3.1. U2OS cells were ␮ translational stage. Such associations may modulate the pres- seeded in six-well plates and were transfected with 0.2 g of ppluc-p53 and increasing amounts of pcDNA3.1-CHIP from 0 to 0.5 ␮g as indicated. entation of p53 for degradation and could serve as sensors of The total amount of DNA was kept constant at 0.8 ␮g/well by the addition 175 cell stress. Here we show that wild-type p53 and an Arg of pcDNA3.1. Cells were washed once with phosphate-buffered saline and mutant form of the tumor suppressor (p53R175H) can be tar- lysed in 100 ␮l of lysis buffer (50 mM MOPS, pH 7.2, 100 mM KCl, 0.5% geted for proteasomal degradation by the CHIP ubiquitin li- Tween 20) containing Complete protease inhibitor. The lysate was cen- gase. Our data thus reveal a novel degradation pathway for the trifuged at 20,000 ϫ g for 30 min at 4 °C, and the supernatant was used ␮ tumor suppressor that is entered through a transient associa- as a soluble extract. 5 l of cell extract were analyzed with luciferase assay reagent (Promega). Levels of p53 and CHIP were determined after tion of wild-type and mutant p53 with molecular chaperones. SDS-PAGE and immunoblotting. RNA Interference—Endogenous CHIP was depleted in U2OS cells MATERIALS AND METHODS using siRNA oligonucleotides (Dharmacon, Lafayette, CO). The Purified Proteins and Antibodies—The following proteins were ex- oligonucleotide CHIP1 is directed against the sequence GAA- pressed recombinantly and purified as described previously: rat Hsc70, GAAGCGCTGGAACAGC, and CHIP2 targets the sequence ACCAC- human Hsp40, human UbcH5b, human CHIP, and wheat E1 (26, 50, GAGGGTGATGAGGA of the human CHIP gene. Green fluorescent 51). Purified bovine ubiquitin was purchased from Sigma. For immu- protein siRNA (GGCTACGTCCAGGAGCGCACC) served as the con- noblotting anti-p53 (DO-1; Oncogene Research Products, San Diego, trol. U2OS cells were seeded in six-well plates and were transfected CA), polyclonal anti-Hsc/Hsp70 (F. U. Hartl, MPI for Biochemistry), twice at a 72-h interval with siRNA oligonucleotides using Lipo- monoclonal anti-Hsc/Hsp70 (SPA-820; StressGen Biotechnologies, San fectamine 2000 (Invitrogen) according to the manufacturer’s recom- Diego, CA), and anti-CHIP antibodies (25) were used. mendations. 72 h after the second transfection cells were washed once Cell Culture and Transfection—H1299 cells were grown in RPMI with phosphate-buffered saline and lysed in 100 ␮l of radioimmune media (Sigma) supplemented with 10% fetal calf serum, penicillin, and precipitation assay buffer supplemented with Complete protease in- streptomycin. Transfection of H1299 cells was performed using DOTAP hibitor. The lysate was centrifuged at 20,000 ϫ g for 30 min at 4 °C, liposomal transfection reagent according to the protocol of the manu- and the supernatant was used as a soluble extract. Equal amounts of facturer (Roche Diagnostics). U2OS cells were grown in Dulbecco’s protein were separated by SDS-PAGE. Levels of CHIP and p53 were modified Eagle’s medium (Invitrogen) supplemented with 10% fetal calf determined by immunoblotting. serum, penicillin, and streptomycin. For transfection of U2OS cells Effectene transfection reagent (Qiagen, Valencia, CA) was used. Cell RESULTS extracts were prepared 24 h post-transfection. p53R175H but Not Wild-type p53 Forms Stable Complexes Degradation Assays—To analyze the degradation kinetics of p53, H1299 cells were seeded in six-well plates and were transfected with 0.1 with Hsc70 in H1299 Cells—To investigate a potential influ- ␮g of pRC/CMV-p53wt or 0.1 ␮g of pcDNA3.1-p53R175H and 1.2 ␮gof ence of the chaperone-associated ubiquitin ligase CHIP on the pcDNA3.1-CHIP as indicated. The total amount of added DNA was kept turnover of p53, the human lung cancer cell line H1299 was constant at 2.5 ␮g/well by the addition of pcDNA3.1. Protein lysates used. In an initial experiment we analyzed the association of were prepared at indicated time points after addition of cycloheximide wild-type p53 and the conformational mutant p53R175H with ␮ (60 g/ml). Cells were washed once with phosphate-buffered saline and Hsc70 following transient transfection of the p53-deficient cell lysed in radioimmune precipitation assay buffer (25 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% SDS, 0.5% sodium deoxycholate, 1% Nonidet line with corresponding expression plasmids. Although wild- P-40, 10% glycerol, 2 mM EDTA) supplemented with Complete protease type p53 was not found in association with Hsc70, complexes inhibitor (Roche Diagnostics). The lysate was centrifuged at 20,000 ϫ g between p53R175H and Hsc70 were readily detectable after for 30 min at 4 °C, and the supernatant was used as a soluble extract. immunoprecipitation (Fig. 1). Transient transfection of H1299 Equal amounts of protein were separated by SDS-PAGE. Levels of p53 cells thus recapitulates the findings obtained in p53-expressing and p53R175H were determined by immunoblotting and quantified at cell lines and tumor specimens that conformational mutants indicated time points. To analyze the effect of geldanamycin on the degradation of p53, H1299 cells were treated with geldanamycin (1 ␮M) but not wild-type p53 form stable complexes with molecular and dimethyl sulfoxide, respectively, 12 h after transfection. chaperones (37–39). In Vitro Ubiquitylation Assay—In vitro transcription of p53 and p53R175H and Wild-type p53 Can Be Targeted for Degrada- p53R175H was performed with pRC/CMV-p53wt and pcDNA3.1-p53R175H tion by the CHIP Ubiquitin Ligase—As the p53R175H mutant of using the T7 RiboMax system according to the manufacturer’s instructions p53 was found associated with Hsc70, we investigated whether (Promega). The obtained RNA was used for in vitro translation of radiola- the turnover of the mutant form is affected by CHIP. Upon beled p53 and p53R175H with nuclease-treated rabbit reticulocyte lysate (Promega). For ubiquitylation of p53 and p53R175H, 8 ␮l of the translation elevation of the cellular levels of CHIP in H1299 cells a signifi- reactions was incubated with 0.1 ␮M E1, 4 ␮M UbcH5b, 6 ␮M CHIP, 6 ␮M cant decline in the levels of coexpressed p53R175H was observed Hsc70, and 0.6 ␮M Hsp40 as indicated. Each sample received 2.5 ␮g/␮l (Fig. 2A). This decline could be attributed to an accelerated deg- CHIP Induces p53 Degradation 27445

FIG.2.CHIP targets p53R175H for proteasomal degradation. H1299 cells were transiently transfected with pcDNA3.1-p53R175H FIG.3. CHIP targets wild-type p53 for proteasomal degrada- (p53R175H) and pcDNA3.1-CHIP (CHIP) as indicated. A, to analyze the tion. H1299 cells were transiently transfected with pRC/CMV-p53wt effect of CHIP on steady state levels of p53R175H, protein lysates were (p53) and pcDNA3.1-CHIP (CHIP) as indicated. A, to analyze the effect prepared. After SDS-PAGE and immunoblotting p53R175H and CHIP of CHIP on steady state levels of p53, protein lysates were prepared. were detected by specific antibodies. Immunodetection of Hsc/Hsp70 After SDS-PAGE and immunoblotting p53 and CHIP were detected by served as loading control. B, to determine the effect of CHIP on the specific antibodies. Immunodetection of Hsc/Hsp70 served as loading half-life of p53R175H, protein lysates were prepared at the indicated control. B, to determine the effect of CHIP on the half-life of p53, protein time points after cycloheximide (60 ␮g/ml) treatment. The amount of lysates were prepared at the indicated time points after cycloheximide ␮ p53R175H was analyzed by immunoblotting using a specific anti-p53 (60 g/ml) treatment. The amount of p53 was analyzed by immunoblot- ␮ antibody. 25 ␮g of total protein were loaded for each time point. C, the ting using a specific anti-p53 antibody. 25 g of total protein were amount of p53R175H in presence of CHIP (open triangles) and absence loaded for each time point. C, the amount of p53 in presence of CHIP of CHIP (closed triangles) was quantified. Presented data were aver- (open squares) and absence of CHIP (closed squares) was quantified. aged from six independent experiments. Error bars represent the S.D. Presented data were averaged from six independent experiments. Error bars represent the S.D.

radation of p53R175H induced by the CHIP ubiquitin ligase (Fig. 2, B and C). Apparently, the oncogenic mutant protein can be turned over by a chaperone-assisted degradation pathway. Despite the fact that complexes between wild-type p53 and Hsc70 were not detectable by immunoprecipitation, it was previously speculated that wild-type p53 may undergo highly transient interactions with molecular chaperones (39). We reasoned that such transient interactions might become detectable upon CHIP overexpression, when even those proteins that associate FIG.4. Geldanamycin stimulates the degradation of p53 and with Hsc70 in a highly transient manner would be irreversibly p53R175H. The effect of geldanamycin on the degradation of p53 and diverted onto a degradation pathway (29, 31, 32). In fact, CHIP p53R175H was investigated in H1299 cells coexpressing CHIP when indicated. Cells were cultured overnight with GA and Me2SO (DMSO) was able to induce the degradation of wild-type p53, albeit with as indicated. Data are shown as the ratio of degradation rates deter- a slightly reduced efficiency when compared with the findings mined in control cells versus treated cells. Error bars represent the S.D. for p53R175H (Fig. 3). Our data reveal that wild-type p53 of three independent experiments. associates transiently with molecular chaperones and can be diverted onto a degradation pathway through this association. Geldanamycin Induces the Degradation of Wild-type and Mutant p53 in H1299 Cells—We investigated how the ansamy- cin antibiotic geldanamycin (GA) affects p53 degradation. GA specifically inhibits Hsp90, which usually results in the proteasomal degradation of client proteins that depend on the activity of the chaperone (52). The turnover of wild-type and mutant p53 was stimulated upon GA treatment, consistent FIG.5. The ubiquitin ligase CHIP induces ubiquitylation of with an association of both forms with Hsp90 (Fig. 4). However, p53 in cooperation with UbcH5b and Hsc70. A, in vitro translated only in the case of p53R175H was a synergistic effect of GA p53 and p53R175H were incubated with purified Hsc70, Hsp40, treatment and CHIP elevation observed. The prolonged and UbcH5b, and CHIP as indicated. All reactions received the ubiquitin- more stable association of mutant p53 with Hsp90 and Hsc70 activating enzyme E1. Ubiquitylation was assessed by autoradiogra- phy. B, quantification of data obtained under A. The amount of ubiq- may provide the molecular basis for this synergism. uitylated p53 detected in the presence of Hsc70, Hsp40, and UbcH5b CHIP Mediates Ubiquitylation of p53 and p53R175H in Co- was set to 1. operation with UbcH5b and Hsc70—To verify that CHIP acts as an E3 ubiquitin ligase during p53 degradation, p53 and CHIP Affects p53-mediated Transcription—We analyzed p53R175H were in vitro translated in rabbit reticulocyte lysate. how CHIP influences p53-mediated transcription using a re- Upon addition of purified CHIP to translation reactions, ubiq- porter construct that contains firefly luciferase under the con- uitylated forms of wild-type and mutant p53 accumulated (Fig. trol of a p53-response element. Overexpression of CHIP led to 5). An increase in the amount of ubiquitylated p53 was also a significant reduction of luciferase activity in corresponding observed when UbcH5b and Hsc70 were added, suggesting a cell extracts (Fig. 6A). Conceivably, such a reduction may be close cooperation of CHIP, UbcH5b, and Hsc70 during the caused by a CHIP-induced degradation of luciferase itself. ubiquitylation of p53, similar to recent observations for other However, it was previously shown that CHIP does not target CHIP substrates (26, 32). luciferase for degradation but stimulates the folding of lucifer- 27446 CHIP Induces p53 Degradation

FIG.7.CHIP ablation by siRNA stabilizes p53. U2OS cells were transfected with siRNA oligonucleotides targeting two different se- quences of CHIP (CHIP1 and CHIP2) and green fluorescent protein (GFP) siRNA as indicated. Levels of CHIP and p53 were analyzed after the separation of 30 ␮g of protein extracts on a SDS-PAGE and immunoblotting using specific antibodies. Actin served as loading control.

(39, 54). Therefore, oncogenic mutants are amenable to a regulation through chaperone-assisted degradation. A similar association with molecular chaperones was not yet unequivocally demonstrated for wild-type p53, as complexes between the tumor suppressor and Hsc70 or Hsp90 could not be isolated (39). It was speculated, however, that such complexes may escape detection because of their highly transient nature (39). Such transient associations with Hsc70 or Hsp90 may become FIG.6.CHIP influences p53-mediated transcription. A, H1299 detectable when the chaperone client is irreversibly diverted cells were transiently transfected with a reporter plasmid expressing onto a degradation pathway through the action of the CHIP firefly luciferase under control of a p53 response element. Cotransfec- ubiquitin ligase. In fact, we observed that wild-type p53 is tions were carried out using p53 and CHIP expression vectors as indicated. Protein lysates were prepared, and luciferase activity was meas- sensitive to an elevation of the cellular levels of the chaperone- ured. Activities are expressed relative to uninduced promoter activity, associated ubiquitin ligase. Furthermore, depletion of endoge- which was set to 1. The error bars represent the S.D. of three independ- nous CHIP stabilized wild-type p53 in U2OS cells. Taken to- ent experiments. B, levels of p53 and CHIP were detected by immuno- gether, our data establish a role of molecular chaperones in the blotting using specific antibodies. 40 ␮g of total protein were loaded. Immunodetection of Hsc/Hsp70 served as loading control. C, U2OS cells regulation of wild-type p53. were transiently transfected with a reporter plasmid expressing firefly Evidence suggests that chaperones associate with p53 at luciferase under control of a p53 response element. Increasing amounts multiple stages and influence the oligomeric state, the nucleo- of pcDNA-CHIP were cotransfected ranging from 0 to 0.5 ␮g DNA. cytoplasmic transport and the transcriptional activity of the Activities are expressed relative to p53-induced promoter activity, which was set to 100%. tumor suppressor (reviewed in Refs. 55–59). For example, Hsc70 was shown to sequester a mutant form of p53 in the ase in tissue culture cells (53). In this regard the observed cytoplasm by masking a nuclear localization signal present at reduction of luciferase activity rather appears to be an under- the carboxyl terminus (56). p53 displays multiple binding sites estimate of the effect of CHIP on p53-mediated transcription. for Hsc70, located in the amino-terminal transactivation do- Notably, the observed reduction in luciferase activity was com- main, the DNA-binding core domain, and the carboxyl termi- parable with the reduction of p53 levels (Fig. 6B). CHIP-in- nus (60–63). Binding sites within the core domain are recog- duced attenuation of p53-mediated transcription was also ob- nized with particularly high affinity by Hsc70, but they appear served in U2OS cells that endogenously express p53 (Fig. 6C). to be buried within the native protein. On the other hand, the It seems that CHIP alters p53-dependent transcriptional re- amino-terminal region of p53 seems to be natively unfolded, sponses through an induced degradation of p53. and the carboxyl terminus is relatively unstructured (64–66). Depletion of Endogenous CHIP Stabilizes p53—Endogenous Binding sites within these regions may remain accessible for levels of CHIP were depleted in U2OS cells following transfec- Hsc70 when the protein is largely in a native conformation. tion with siRNAs (Fig. 7). Intriguingly, depletion of the chap- Notably, studies using conformational antibodies indicate that erone-associated ubiquitin ligase caused a significant increase even the core domain is a metastable structure that is easily of p53 levels. The findings emphasize the role of chaperone- perturbed upon treatment with chelating or oxidizing agents or assisted degradation in maintaining low concentrations of p53 by raising the temperature (49, 67). The core domain is also under physiological conditions. recognized by Hsp90 (57, 68), and Walerych et al. (58) recently demonstrated that Hsp90 is required to maintain the DNA DISCUSSION binding activity of the core domain under physiological condi- Here we identify a novel degradation pathway for the tumor tions. Apparently, p53 is a protein of large conformational suppressor p53. This pathway is entered through a transient flexibility and seems to be in a conformational equilibrium association of wild-type and mutant p53 with molecular chap- between native and less structured states. This flexibility may erones and involves the chaperone-associated ubiquitin ligase provide the means for complex intra- and intermolecular inter- CHIP. CHIP cooperates with E2 enzymes of the Ubc4/5 family actions and for the association of p53 with a multitude of to mediate the attachment of an ubiquitin-derived degradation regulatory proteins. The accompanying conformational signal to chaperone-bound p53, which leads to the proteasomal changes are apparently assisted by Hsc70 and Hsp90. In this destruction of the tumor suppressor. CHIP-mediated degrada- regard the extended chaperone interactions observed for mu- tion critically depends on the interaction of the ubiquitin ligase tant p53 seem to represent a pathologic exaggeration of phys- with the chaperones Hsc70 and Hsp90, which were shown to iologic interactions of wild-type p53 with the chaperone present chaperone clients to the ubiquitin ligase (26, 29, 31, machinery. 32). It has long been appreciated that oncogenic mutant forms Conceivably, the CHIP-mediated degradation pathway of p53 associate with Hsc70 and Hsp90 in tissue culture cells might be entered through an association of a chaperone client and in tumor specimens (37, 39). Because of conformational with either Hsc70 or Hsp90, as CHIP is able to bind both alterations mutant p53s are retained in complexes with the two chaperones. Treatment of tissue culture cells with small mo- chaperones and with several of their regulatory cochaperones lecular inhibitors of Hsp90, such as GA, has helped to verify CHIP Induces p53 Degradation 27447 this hypothesis. Geldanamycin blocks the interaction of Hsp90 16. Montes de Oca Luna, R., Wagner, D. S., and Lozano, G. (1995) Nature 378, 203–206 with chaperone clients. Remarkably, the Hsp90 inhibitor stim- 17. Prives, C., and Hall, P. A. (1999) J. Pathol. 187, 112–126 ulates the CHIP-mediated degradation of signaling proteins 18. Barak, Y., Juven, T., Haffner, R., and Oren, M. (1993) EMBO J. 12, 461–468 that rely on an association with Hsc70 and Hsp90, i.e. the 19. Linares, L. K., Hengstermann, A., Ciechanover, A., Mu¨ller, S., and Scheffner, M. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 12009–12014 oncogenic receptor tyrosine kinase ErbB2 (30). GA treatment 20. Sui, G., Affar el, B., Shi, Y., Brignone, C., Wall, N. R., Yin, P., Donohoe, M., dissociates Hsp90 while increasing the association of Hsc70 Luke, M. P., Calvo, D., Grossman, S. R., and Shi, Y. (2004) Cell 117, with ErbB2. Redistribution into Hsc70 complexes followed by 859–872 21. Grossman, S. R., Deato, M. E., Brignone, C., Chan, H. M., Kung, A. L., Tagami, proteasomal degradation was also observed for oncogenic mu- H., Nakatani, Y., and Livingston, D. M. (2003) Science 300, 342–344 tants of p53 upon treatment of tumor cell lines with GA (39). 22. Leng, R. P., Lin, Y., Ma, W., Wu, H., Lemmers, B., Chung, S., Parant, J. M., Here we show that GA-mediated inhibition of Hsp90 induces Lozano, G., Hakern, R., and Benchimol, S. (2003) Cell 112, 779–791 23. Dornan, D., Wertz, I., Schimizu, H., Arnott, D., Frantz, G. D., Dowd, P., the degradation of mutant and wild-type p53 in H1299 cells. O’Rourke, K., Koeppen, H., and Dixit, V. M. (2004) Nature 429, 86–92 This provides additional evidence for an association of both 24. Esser, C., Alberti, S., and Ho¨hfeld, J. (2004) Biochim. Biophys. Acta 1695, forms with molecular chaperones. GA-induced degradation of 171–188 25. Ballinger, C. A., Connell, P., Wu, Y., Hu, Z., Thompson, L. J., Yin, L.-Y., and mutant but not wild-type p53 was further stimulated by the Patterson, C. (1999) Mol. Cell. Biol. 19, 4535–4545 overexpression of CHIP. The highly transient association of 26. Demand, J., Alberti, S., Patterson, C., and Ho¨hfeld, J. (2001) Curr. Biol. 11, wild-type p53 with Hsc70 and Hsp90 may limit the synergistic 1569–1577 27. Murata, S., Minami, Y., Minami, M., Chiba, T., and Tanaka, K. (2001) EMBO effects of GA treatment and CHIP overexpression. In any case, Rep. 2, 1133–1138 the diversion of chaperone clients, including p53, onto a pro- 28. Jiang, J., Ballinger, C. A., Wu, Y., Dai, Q., Cyr, D. M., Ho¨hfeld, J., and teasomal degradation pathway seems to be preferentially me- Patterson, C. (2001) J. Biol. Chem. 276, 42938–42944 29. Connell, P., Ballinger, C. A., Jiang, J., Wu, Y., Thompson, L. J., Ho¨hfeld, J., diated by an Hsc70/CHIP chaperone machinery. and Patterson, C. (2001) Nat. Cell Biol. 3, 93–96 The observation that elevated CHIP levels increase the sen- 30. Xu, W., Marcu, M., Yuan, X., Mimnaugh, E., Patterson, C., and Neckers, L. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 12847–12852 sitivity of oncogenic mutant forms of p53 against Hsp90 inhib- 31. Meacham, G. C., Patterson, C., Zhang, W., Younger, J. M., and Cyr, D. M. itors might be of profound relevance for cancer treatment. (2001) Nat. Cell Biol. 3, 100–105 Several Hsp90-inhibitors are currently tested in clinical trials 32. Alberti, S., Bo¨hse, K., Arndt, V., Schmitz, A., and Ho¨hfeld, J. (2004) Mol. Biol. Cell 15, 4003–4010 as anti-tumor agents (52). The expression level of CHIP might 33. Shimura, H., Schwartz, D., Gygi, S. P., and Kosik, K. S. (2004) J. Biol. Chem. be an important determinant of the therapeutic success of such 279, 4869–4876 pharmacological interventions. 34. Petrucelli, L., Dickson, D., Kehoe, K., Taylor, J., Snyder, H., Grover, A., De Lucia, M., McGowan, E., Lewis, J., Prihar, G., Kim, J., Dillmann, W. H., Notably, siRNA-mediated depletion of CHIP significantly Browne, S. E., Hall, A., Voellmy, R., Tsuboi, Y., Dawson, T. M., Wolozin, B., stabilized p53. This suggests a critical role of chaperone-as- Hardy, J., and Hutton, M. (2004) Hum. Mol. Genet. 13, 703–714 sisted degradation in maintaining low concentrations of the 35. Dai, Q., Zhang, C., Wu, Y., McDonough, H., Whaley, R. A., Godfrey, V., Li, H. H., Madamanchi, N., Xu, W., Neckers, L., Cyr, D., and Patterson, C. tumor suppressor under physiological conditions and points to (2003) EMBO J. 22, 5446–5458 a novel link between the cellular chaperone machinery and 36. Mosser, D. D., and Morimoto, R. I. (2004) Oncogene 23, 2907–2918 apoptosis regulation. Intriguingly, recent evidence also sug- 37. Davidoff, A. M., Iglehart, J. D., and Marks, J. R. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 3439–3442 gests a cross-talk between CHIP and Mdm2 in p53 regulation 38. Blagosklonny, M. V., Toretsky, J., and Neckers, L. (1995) Oncogene 11, (59). The molecular basis for this cross-talk is provided by the 933–939 formation of a heterocomplex comprising Mdm2, p53, and 39. Whitesell, L., Sutphin, P. D., Pulcini, E. J., Martinez, J. D., and Cook, P. H. (1998) Mol. Cell. Biol. 18, 1517–1524 Hsp90 (59, 69). In the heterocomplex the activity of Mdm2 to 40. Hinds, P. W., Finlay, C. A., Quartin, R. S., Baker, S. J., Fearon, E. R., ubiquitylate p53 is inhibited by Hsp90 (64). At the same time Vogelstein, B., and Levine, A. J. (1990) Cell Growth & Differ. 1, 571–580 Mdm2 and Hsp90 cooperate to stimulate the unfolding of na- 41. Gannon, J. V., Greaves, R. Iggo, R., and Lane, D. P. (1990) EMBO J. 9, 1595–1602 tive p53 tetramers (59). Because of its ability to bind Hsp90, 42. Stephen, C. W., and Lane, D. P. (1992) J. Mol. Biol. 225, 577–583 CHIP can enter the heterocomplex and then further promote 43. Bargonetti, J., Manfredi, J. J., Chen, X., Marshak, D. R., and Prives, C. (1993) p53 unfolding (59). This may represent an initial step in divert- Genes Dev. 7, 2565–2574 44. Cho, Y., Gorina, S., Jeffrey, P. D., and Pavletich, N. P. (1994) Science 265, ing p53 from Mdm2-mediated regulation onto a CHIP-induced 346–355 degradation pathway. In line with a role of CHIP in apoptosis 45. Finlay, C. A., Hinds, P. W., Tan, T. H., Eliyahu, D., Oren, M., and Levine, A. J. regulation, an anti-apoptotic activity of CHIP was recently (1988) Mol. Cell. Biol. 8, 531–539 46. Xu, Y., Singer, M. A., and Lindquist, S. (1999) Proc. Natl. Acad. Sci. U. S. A. demonstrated based on the analysis of CHIP knock-out mice 96, 109–114 (35). Although the anti-apoptotic activity was largely attrib- 47. Cheung, J., and Smith, D. F. (2000) Mol. Endocrinol. 14, 939–946 uted to the role of CHIP in the conformational regulation of the 48. Milner, J., and Watson, J. V. (1990) Oncogene 5, 1683–1690 49. Hainaut, P., Butcher, S., and Milner, J. (1995) Br. J. Cancer 71, 227–231 heat shock transcription factor, the ability of the chaperone- 50. Ho¨hfeld, J., and Jentsch, S. (1997) EMBO J. 16, 6209–6216 associated ubiquitin ligase to induce p53 degradation may con- 51. Alberti, S., Demand, J., Esser, C., Emmerich, N., Schild, H., and Ho¨hfeld, J. (2002) J. Biol. Chem. 277, 45920–45927 tribute to this activity. 52. Neckers, L. (2002) Trends Mol. Med. 8, S55–61 REFERENCES 53. Kampinga, H. H., Kanon, B., Salomons, F. A., Kabakov, A. E., and Patterson, C. (2003) Mol. Cell. Biol. 23, 4948–4958 1. Lane, D. P. (1992) Nature 358, 15–16 54. King, F. W., Wawrzynow, A., Ho¨hfeld, J., and Zylicz, M. (2001) EMBO J. 20, 2. Levine, A. J. (1997) Cell 88, 323–331 6297–6305 3. Vousden, K. H., and Lu, X. (2002) Nat. Rev. Cancer 2, 594–604 55. Zylicz, M., King, F. W., and Wawrzynow, A. (2001) EMBO J. 20, 4634–4638 4. Vogelstein, B., Lane, D., and Levine, A. J. (2000) Nature 408, 307–310 56. Akakura, S., Yoshida, M., Yoneda, Y., and Horinouchi, S. (2001) J. Biol. Chem. 5. Hainaut, P., and Hollstein, M. (2000) Adv. Cancer Res. 77, 81–137 276, 14649–14657 6. Woods, Y. L., and Lane, D. P. (2003) Hematol. J. 4, 233–247 57. Mu¨ller, L., Schaupp, A., Walerych, D., Wegele, H., and Buchner, J. (2004) 7. Xirodimas, D. P., Saville, M. K., Bourdon, J. C., Hay, R. T., and Lane, D. P. J. Biol. Chem. 279, 48846–48854 (2004) Cell 118, 83–97 58. Walerych, D., Kudla, G., Gutkowska, M., Wawrzynow, B., Mu¨ller, L., King, 8. Harper, J. W. (2004) Cell 118, 2–4 9. Freedman, D. A., Wu, L., and Levine, A. J. (1999) Cell. Mol. Life Sci. 55, F. W., Helwak, A., Boros, J., Zylicz, A., and Zylicz, M. (2004) J. Biol. Chem. 96–107 279, 48836–48845 10. Michael, D., and Oren, M. (2003) Semin. Cancer Biol. 13, 49–58 59. Burch, L., Shimizu, H., Smith, A., Patterson, C., and Hupp, T. R. (2004) J. Mol. 11. Hershko, A., Ciechanover, A., and Varshavsky, A. (2000) Nat. Med. 10, Biol. 337, 129–145 1073–1081 60. Stu¨rzbecher, H. W., Addison, C., and Jenkins, J. R. (1988) Mol. Cell. Biol. 8, 12. Fang, S., Jensen, J. P., Ludwig, R. L., Vousden, K. H., and Weissman, A. M. 3740–3747 (2000) J. Biol. Chem. 275, 8945–8951 61. Lam, K. T., and Calderwood, S. K. (1992) Biochem. Biophys. Res. Commun. 13. Kubbutat, M. H., Jones, S. N., and Vousden, K. H. (1997) Nature 387, 299–303 184, 167–174 14. Haupt, Y., Maya, R., Kazaz, A., and Oren, M. (1997) Nature 387, 296–299 62. Fourie, A. M., Sambrook, J. F., and Gething, M. J. (1994) J. Biol. Chem. 269, 15. Jones, S. N., Roe, A. E., Donehower, L. A., and Bradley, A. (1995) Nature 378, 30470–30478 206–208 63. Fourie, A. M., Hupp, T. R., Lane, D. P., Sang, B. C., Barbosa, M. S., Sambrook, 27448 CHIP Induces p53 Degradation

J. F., and Gething, M. J. (1997) J. Biol. Chem. 272, 19471–19479 (2003) J. Mol. Biol. 332, 1131–1141 64. Pavletich, N. P., Chambers, K. A., and Pabo, C. O. (1993) Genes Dev. 7, 67. Hainaut, P., and Milner, J. (1993) Cancer Res. 53, 4469–4473 2556–2564 68. Ru¨diger, S., Freund, S. M., Veprintsev, D. B., and Fersht, A. R. (2002) Proc. 65. Bell, S., Klein, C., Mu¨ller, L., Hansen, S., and Buchner, J. (2002) J. Mol. Biol. Natl. Acad. Sci. U. S. A. 99, 11085–11090 322, 917–927 69. Peng, Y., Chen, L., Li, C., Lu, W., and Chen, J. (2001) J. Biol. Chem. 276, 66. Dawson, R., Mu¨ller, L., Dehner, A., Klein, C., Kessler, H., and Buchner, J. 40583–40590