Molecular Chaperone Hsp90 Regulates REV1-Mediated Mutagenesis

2 downloads 0 Views 4MB Size Report
Jan 26, 2011 - Kim, S. H., and W. M. Michael. 2008. Regulated proteolysis .... 282:20256–20263. 62. Xie, K., J. Doles, M. T. Hemann, and G. C. Walker. 2010.
MOLECULAR AND CELLULAR BIOLOGY, Aug. 2011, p. 3396–3409 0270-7306/11/$12.00 doi:10.1128/MCB.05117-11 Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Vol. 31, No. 16

Molecular Chaperone Hsp90 Regulates REV1-Mediated Mutagenesis䌤 Franklin Mayca Pozo,1† Tsukasa Oda,1† Takayuki Sekimoto,1 Yoshiki Murakumo,2 Chikahide Masutani,3‡ Fumio Hanaoka,3§ and Takayuki Yamashita1* Laboratory of Molecular Genetics, The Institute for Molecular and Cellular Regulation, Gunma University, Maebashi, Gunma 371-8512, Japan1; Department of Pathology, Nagoya University Graduate School of Medicine, Showa-ku, Nagoya 466-8550, Japan2; and Cellular Biology Laboratory, Graduate School of Frontier Biosciences, Osaka University, and SORST, Japan Science and Technology Agency, Suita, Osaka 565-0871, Japan3 Received 26 January 2011/Returned for modification 16 February 2011/Accepted 13 June 2011

REV1 is a Y-family polymerase that plays a central role in mutagenic translesion DNA synthesis (TLS), contributing to tumor initiation and progression. In a current model, a monoubiquitinated form of the replication accessory protein, proliferating cell nuclear antigen (PCNA), serves as a platform to recruit REV1 to damaged sites on the DNA template. Emerging evidence indicates that posttranslational mechanisms regulate REV1 in yeast; however, the regulation of REV1 in higher eukaryotes is poorly understood. Here we show that the molecular chaperone Hsp90 is a critical regulator of REV1 in human cells. Hsp90 specifically binds REV1 in vivo and in vitro. Treatment with a specific inhibitor of Hsp90 reduces REV1 protein levels in several cell types through proteasomal degradation. This is associated with suppression of UV-induced mutagenesis. Furthermore, Hsp90 inhibition disrupts the interaction between REV1 and monoubiquitinated PCNA and suppresses UV-induced focus formation. These results indicate that Hsp90 promotes folding of REV1 into a stable and/or functional form(s) to bind to monoubiquitinated PCNA. The present findings reveal a novel role of Hsp90 in the regulation of TLS-mediated mutagenesis. mals (9, 29, 62). REV1 has a unique deoxycytidyl transferase activity in vitro, incorporating dCMP opposite several types of DNA lesions, such as abasic sites and damaged guanines (17, 18, 30, 38, 65); however, this enzymatic activity is dispensable for REV1-mediated TLS (37). Instead, the function of REV1 depends on protein-protein interactions via three key domains: the N-terminal BRCT domain (named after the C-terminal domain of a breast cancer susceptibility protein) (15, 22, 23), ubiquitin (Ub)-binding motifs (UBMs) in the central region (16, 61), and the C-terminal polymerase interaction domain (PID) (13, 27, 35, 42). The BRCT domain, which is not contained in the other Y-Pols, mediates the physical interaction between REV1 and PCNA (15). Cooperating with this domain, two UBMs in the central region of REV1 participate in the interaction between REV1 and Ub-PCNA (16, 61). The C-terminal domain of REV1 mediates interaction with other Y-Pols, Pol ␩, Pol ␫, and Pol ␬ (13, 27, 42). Pol ␨ also interacts with REV1 via its noncatalytic REV7 subunit (13, 35). Based on these findings, it was proposed that REV1 acts as a scaffold protein to coordinate TLS polymerase switching (10, 14, 28, 43, 58). To support this notion, recent studies have indicated that the REV1/Pol ␬ interaction is required for Pol ␬ function (41) and that the REV1/Pol ␩ interaction promotes targeting of REV1 to UV-damaged DNA (1). Among the Y-Pols, REV1 and Pol ␩ are conserved throughout the eukaryotes and have been extensively studied. Substantial evidence indicates that posttranslational control of Pol ␩ plays an important role in the regulation of its function. In Saccharomyces cerevisiae, Pol ␩ protein levels are regulated by proteasomal degradation (52). In addition, various types of covalent modifications, including monoubiquitination, phosphorylation, and sumoylation, are reported to regulate protein interactions and stability of Pol ␩ in higher eukaryotes (3, 4, 6, 26). In contrast, understanding of REV1 regulation is limited.

Genomic DNA is constantly exposed to both extrinsic and intrinsic genotoxic agents, such as UV light and oxidative stress. Although most DNA lesions are removed by multiple DNA repair pathways, some escape these mechanisms and persist in the genome. Such unrepaired DNA lesions usually block the progression of replication forks catalyzed by highfidelity DNA polymerases (Pols), which may lead to downstream reinitiation of DNA synthesis, leaving single-stranded DNA gaps. Translesion synthesis (TLS) is an essential mechanism for bypassing such replication blocks and postreplicative gaps by employing specialized Pols, including Pol ␩, Pol ␬, Pol ␫, and REV1, members of the Y-family Pols (Y-Pols), and Pol ␨, a member of the B family (10, 14, 28, 43, 58). These TLS Pols display low stringency for the active site and a lack of proofreading, and thus they contribute to mutagenesis and to DNA damage tolerance. Recruitment of the TLS Pols at sites of DNA damage depends on Rad6/Rad18- and CRL4Cdt2-mediated production of monoubiquitinated proliferating cell nuclear antigen (PCNA) (Ub-PCNA), to which Y-Pols bind with high affinity (3, 21, 54, 57). REV1 plays a central role in promoting mutagenesis in lower and higher eukaryotes. Of note, the mutagenic activity of REV1 contributes to tumor initiation and progression in mam-

* Corresponding author. Mailing address: Laboratory of Molecular Genetics, The Institute for Molecular and Cellular Regulation, Gunma University, 3-39-15 Showa-machi, Maebashi, Gunma 3718512, Japan. Phone: 81-27-220-8830. Fax: 81-27-220-8834. E-mail: [email protected]. † These authors contributed equally to this work. ‡ Present address: Research Institute of Environmental Medicine, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan. § Present address: Department of Chemistry, Faculty of Science, Gakushuin University, Toshima-ku, Tokyo 171-8588, Japan. 䌤 Published ahead of print on 20 June 2011. 3396

VOL. 31, 2011

Hsp90 REGULATES REV1

In S. cerevisiae, Mec1 protein kinase, the ortholog of human ATR, phosphorylates Rev1, promoting efficiency of mutagenic TLS in a Pol ␨-dependent manner (20, 44, 49). Yeast Rev1 protein levels increase to be ⬃50-fold higher during the G2/M phase compared to those during the G1 and S phases, whereas its transcript levels vary only ⬃3-fold over the same phases, suggesting posttranscriptional control of Rev1 levels (59). Indeed, a recent study revealed that proteasomal degradation is involved in cell cycle-linked Rev1 regulation (60). However, the amino acid residue phosphorylated by Mec1 is not conserved, and REV1 protein levels are not affected by cell cycle progression in mammalian cells (1). Thus, the REV1 regulatory mechanism is poorly understood in higher eukaryotes. Hsp90 is an ATP-dependent molecular chaperone that regulates conformation, stability, protein interactions, and trafficking of “client” proteins in cooperation with Hsp70 and cochaperones, such as p23 (46, 53, 56). Most Hsp90 clients are proteins essential for cell survival/growth, including protein kinases, steroid receptors, and transcription factors. In addition, Hsp90 promotes stability and activity of mutant or overexpressed oncoproteins. Hence, Hsp90 is a promising target for cancer chemotherapy. Indeed, specific inhibitors of Hsp90, including geldanamycin analogues, such as 17-allylamino-17demethoxygeldanamycin (17-AAG), are currently in preclinical and clinical studies (46, 56). Of note, an increasing number of proteins in the DNA damage response network, such as Chk1 and FANCA, were shown to be regulated by Hsp90 (2, 39, 64). We recently reported that Hsp90 binds Pol ␩ and regulates its conformation, promoting its stability and binding to Ub-PCNA (50). These findings led us to further study the role of Hsp90 in the regulation of TLS. In the present work, we provide evidence that REV1 is also a client of Hsp90. Our findings reveal a new role of Hsp90 in the regulation of TLS. MATERIALS AND METHODS Plasmids, siRNAs, and PCR primers. cDNAs encoding N-terminally green fluorescent protein (GFP)-tagged full-length human REV1 (1 to 1251), a REV1 fragment (1 to 1178) lacking PID (delPID-REV1), a REV1 fragment (153 to 1251) lacking BRCT (delBRCT-REV1), and a REV1 mutant carrying four missense mutations (L946A, P947A, L1024A, and P1025A) in UBMs (UBM mt-REV1) (16) were inserted into a lentiviral vector, CSII-CMV-MCS-IRES2Bsd (kindly provided by H. Miyoshi, RIKEN). The mutations of UBMs were generated with the QuikChange mutagenesis kit (Stratagene). cDNAs encoding full-length and various truncated REV1 proteins were subcloned into pEGFP-C3 (34). Expression vectors for glutathione S-transferase (GST) fusion proteins were constructed as follows. Human PCNA and ubiquitin cDNAs were subcloned into pGEX6P-1 and pGEX4T-1 (GE Healthcare), respectively (50). cDNA encoding Pol ␩ fragment (positions 350 to 713) was subcloned into pGEX4T-1. cDNA encoding full-length REV7 was subcloned into pGEX6P-1 (36). The target sequences for small interfering RNAs (siRNAs) were as follows: LacZ, 5⬘-CUCGGCGUUUCAUCUGUGG-3⬘; REV1, 5⬘-CCGCUGAGGAAU UGAGAAA-3⬘. The target sequences of Hsp90 small interfering RNA (siRNA) were previously described (50). The sequences of the PCR primers were as follows: glyceraldehyde-3-phosphate dehydrogenase (GAPDH), forward primer, 5⬘-ACCACAGTCCATGCCATCAC-3⬘; reverse primer, 5⬘-TCCACCACCCTG TTGCTGTA-3⬘; REV1, forward primer, 5⬘-CCCAGGAGGAGGATAAGGCT G-3⬘; reverse primer, 5⬘-GTCTTTGTAGGGTATTGACAAACTCAGTC-3⬘. Oligonucleotides and siRNAs were purchased from Invitrogen and Sigma, respectively. Cells. HEK293T (human embryonic kidney cell line), OUMS-36T-3F (an hTERT-immortalized human fibroblast cell line), XP2SASV3 (Pol ␩-deficient simian virus 40 [SV40]-transformed human fibroblast cell line), U2OS, SaOS2 (human osteosarcoma cell lines), DU-145, PC-3 (human prostatic carcinoma cell lines), U87 (human glioblastoma cell line), and COS7 (monkey kidney cell line) cells were maintained in Dulbecco’s modified Eagle medium containing 10%

3397

fetal bovine serum at 37°C under 5% CO2. HEK293T, OUMS-36T-3F, and XP2SASV3 cells stably expressing GFP-REV1, GFP-delPID-REV1, GFP-delBRCT-REV1, or GFP-UBM mt-REV1 were obtained using recombinant lentiviral vectors, CSII-CMV-MCS-IRES2-Bsd/GFP-REV1, CSII-CMV-MCSIRES2-Bsd/GFP-delPID-REV1, CSII-CMV-MCS-IRES2-Bsd/GFP-delBRCT-REV1, or CSII-CMV-MCS-IRES2-Bsd/GFP-UBM mt-REV1, respectively. Antibodies. Guinea pig polyclonal anti-REV1 used for direct Western blotting was previously described (1). Other antibodies used in this study were as follows: mouse monoclonal anti-Hsp90, which recognizes both Hsp90␣ and Hsp90␤ (SPA-830; StressGen), mouse monoclonal anti-Hsp90␣ (2AHSP116A; Institute of Immunology), mouse monoclonal anti-Hsp90␤ (sc-13119; Santa Cruz Biotechnology), rabbit polyclonal anti-REV1 (sc-48806; Santa Cruz Biotechnology), mouse monoclonal anti-PCNA (sc-56; Santa Cruz Biotechnology), rabbit monoclonal anti-GAPDH (14C10; Cell signaling technology), rabbit polyclonal antiGFP (594; MBL), mouse monoclonal anti-Myc (05-724; Upstate), rabbit polyclonal anti-HDAC6 (sc-11420; Santa Cruz Biotechnology), mouse monoclonal anti-PARP-1 (sc-8007; Santa Cruz Biotechnology), and rabbit polyclonal antiubiquitin (SPA-200; StressGen). Cell extracts. Whole-cell lysates (WCL) for immunoblotting were prepared by directly lysing cells in Laemmli buffer. Cell lysates for immunoprecipitation were prepared by incubating cells in HEPES lysis buffer (10 mM HEPES [pH 7.6], 150 mM KCl, 10 mM MgCl2, 0.1% NP-40, 5 mM NaF, 2 mM Na3VO4, 20 mM Na2MoO4, and protease inhibitor cocktail [Roche]) for 30 min at 4°C, as described previously (39). The lysates were centrifuged at 15,000 ⫻ g for 10 min at 4°C, and supernatants were incubated with the indicated antibodies. To study the interaction between REV1 and PCNA, cell extracts were prepared as described previously (25). First, soluble fractions of cells were removed by extraction in CSK buffer [10 mM piperazine-N,N⬘-bis(2-ethanesulfonic acid) (pH 6.8), 100 mM NaCl, 3 mM MgCl2, 1 mM EGTA, 0.5% Triton X-100, and 300 mM sucrose] on ice for 5 min. The CSK-insoluble fractions of cells were fixed with 3.7% formaldehyde in phosphate-buffered saline (PBS) for 10 min on ice. After washing with PBS, the CSK-insoluble fractions were incubated with 0.1 M glycine in PBS for 10 min and then harvested with a cell scraper and suspended in 0.1 M glycine in PBS and centrifuged at 15,000 ⫻ g for 10 min at 4°C. Pellets were dissolved in lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.1% sodium dodecyl sulfate) for 10 min. After sonication with a Bioruptor instrument (Sonifier 150; Branson) with 3 cycles of 5 s on, 30 s off, the lysates were centrifuged at 15,000 ⫻ g for 10 min at 4°C and supernatants were used for immunoprecipitation. To detect polyubiquitinated REV1, cells were lysed with lysis buffer (20 mM HEPES [pH 7.4], 150 mM NaCl, 5 mM EDTA, 10% glycerol, 0.5% Triton X-100, 0.5 mM N-ethylmaleimide, 0.5 mM iodoacetamide, 5 nM ubiquitin aldehyde, and protease inhibitor cocktail) as described previously (40). Immunoprecipitation, immunoblotting, and GST pulldown assay. Immoprecipitation and immunoblotting were performed by standard procedures. Protein signals were densitometrically quantified using National Institutes of Health imaging software (ImageJ). GST fusion proteins were expressed in Escherichia coli (DH5␣) and bound to 20 ␮l of glutathione-Sepharose 4B beads (GE Healthcare) at 4°C for 2 h, followed by washing with PBS, as described previously (50). Cell lysates for GST pulldown assay were extracted with the HEPES lysis buffer and incubated with 5 ␮g of GST fusion proteins bound to Sepharose beads. After 6 h of incubation at 4°C, beads were washed four times with the cell lysis buffer and eluted in Laemmli buffer. Samples were subjected to immunoblotting analysis. In vitro protein synthesis. Myc-REV1 was expressed in rabbit reticulocyte lysates (RRL) using the TNT T7 Quick Coupled transcription/translation system (Promega) according to the supplier’s instructions. In vitro transcription/translation reactions were carried out in a total volume of 50 ␮l for 90 min at 30°C. The reaction mixture was diluted with 500 ␮l of the HEPES lysis buffer and subjected to further analysis. Fluorescence microscopy. Fluorescence microscopy was performed as described previously (39). After UV irradiation, the soluble components of cells were removed by extraction in CSK buffer on ice for 5 min, and the CSKinsoluble fractions were fixed in cold methanol for 10 min at ⫺30°C, as described previously (55). Samples were subsequently washed twice with PBS and incubated for 1 h at room temperature with blocking buffer (PBS containing 0.1% NP-40, 3% bovine serum albumin, and 10% normal goat serum). Samples were incubated with anti-PCNA antibody for 2 h and subsequently incubated with Cy3-conjugated secondary antibody for 1 h (Sigma). Semiquantitative RT-PCR. Total RNA was isolated using RNAiso Plus reagent (TaKaRa). Reverse transcription (RT) reactions were performed using ReverTra Ace (Toyobo) according to the manufacturer’s instructions. Briefly, 1 ␮g of total RNA was added to 20 ␮l of reaction mixture and incubated at 37°C for 30 min, followed by an additional 5 min at 95°C. After an RT reaction, the

3398

MAYCA POZO ET AL.

MOL. CELL. BIOL.

FIG. 1. Physical interaction between Hsp90 and REV1. (A and B) 293T/GFP-REV1 cells were treated with vehicle or 1 ␮M 17-AAG for 6 h. Cell lysates were immunoprecipitated (IP) followed by immunoblotting or were directly immunoblotted, as indicated. cIg, control IgG; IgH, IgG heavy chain; IgL, IgG light chain. (C) HEK293T, OUMS-36T-3F, and XP2SASV3 cells were incubated with vehicle or 1 ␮M 17-AAG for 6 h. Cell lysates were immunoprecipitated and immunoblotted as indicated. (D) Myc-REV1 was synthesized in RRL with vehicle or 10 ␮M 17-AAG. Reaction mixtures were immunoprecipitated and immunoblotted as indicated. As a negative control, a reaction mixture without Myc-REV1 template was similarly analyzed (Mock).

reaction mixture was diluted in 170 ␮l Tris-EDTA (TE) buffer, and 1 ␮l of the cDNA solution was used for semiquantitative PCR with ExTaq polymerase, Hot Start version (TaKaRa). The conditions of PCR were as follows: 28 cycles of denaturation at 94°C for 30 s, annealing at 60°C for 30 s, and elongation at 72°C for 30 s. The relationship between the band intensity and the number of PCR cycles was linear. supF mutagenesis assay. The pSP189 plasmid DNA (50 ␮g) dissolved in 1 ml of TE buffer (10 mM Tris-HCl [pH 7.5], 1 mM EDTA) was irradiated with 100 J/m2 UV in a sterile plastic 60-mm tissue culture dish. The plasmid DNA was then precipitated with ice-cold ethanol, redissolved in TE buffer, and transfected into cells using FuGENE6 reagent (Roche). After 54 h, the plasmid DNA was isolated and transformed into MBM7070. The mutation frequency of supF was determined as described previously (7). Sequencing analysis. The plasmids were isolated from mutant (white) colonies in the supF mutagenesis assay and then sequenced using a Genetic Analyzer 3130 instrument (Applied Biosystems).

RESULTS Physical interaction between Hsp90 and REV1. To address the hypothesis that Hsp90 regulates REV1, we first studied their physical association. We used HEK293T cells and TERTimmortalized human fibroblast OUMS-36T-3F cells that stably express GFP-REV1 (293T/GFP-REV1 and OUMS/GFPREV1 cells, respectively). GFP-REV1 expression levels in

these cell lines were approximately 10-fold and 40-fold greater, respectively, than that of endogenous REV1 (data not shown). In 293T/GFP-REV1 cells, Hsp90 coimmunoprecipitated with GFP-REV1 (Fig. 1A). This interaction was blocked in the presence of 17-AAG, a specific inhibitor of Hsp90. Both Hsp90␣ and Hsp90␤, two major Hsp90 isoforms, associated with REV1. The interaction between Hsp90 and REV1 was confirmed by reciprocal immunoprecipitation using an antiHsp90 antibody (Fig. 1B). Similar results were obtained in OUMS/GFP-REV1 cells (data not shown). Furthermore, endogenous REV1 specifically associated with Hsp90 in HEK293T, OUMS-36T-3F, and XP2SASV3 cells (Fig. 1C). To further test whether REV1 directly binds Hsp90, we synthesized Myc-tagged REV1 in RRL, which are a rich source of Hsp90. In this mixture, REV1 was associated with Hsp90 in a 17-AAG-sensitive manner (Fig. 1D). To determine the region of REV1 that was responsible for the interaction with Hsp90, we next expressed a series of REV1 deletion mutants and examined their interaction with Hsp90 (Fig. 2). The full-length protein, as well as fragments including the central portion (387 to 825), bound Hsp90,

VOL. 31, 2011

Hsp90 REGULATES REV1

3399

FIG. 2. Hsp90 binds to the central region of REV1. The REV1 full-length protein (1–1251) and various truncated mutants with an N-terminal GFP tag were transiently expressed in HEK293T cells using FuGENE6 (Roche). Cell lysates were immunoprecipitated followed by immunoblotting, as indicated. Arrows indicate GFP or GFP-fused REV1 proteins. Structures of the truncated mutants are schematically shown at the bottom. Hsp90 interaction, ranging from negative (⫺) to strongly positive (⫹⫹), is summarized on the right for each mutant.

whereas the N-terminal (1 to 152) and C-terminal (826 to 1251) fragments failed to bind Hsp90. 17-AAG induces proteasomal degradation of REV1 in a cell type-dependent manner. To understand the functional significance of the association between REV1 and Hsp90, we examined effects of 17-AAG on REV1 protein levels in multiple cell lines. In some of these cell lines, such as OUMS-36T-3F, DU145, U2OS, and COS7, 17-AAG induced a significant reduction of REV1 levels (Fig. 3A and data not shown). Of note, the drug induced REV1 reduction in a similar time course in Pol ␩-deficient XP2SASV3 cells (Fig. 3A). On the other hand, such an effect was negligible in HEK293T, PC-3, U87, and SaOS2 cell lines (Fig. 3A and data not shown). Basal protein levels of REV1 in these cell lines and XP2SASV3 cells complemented with wild-type Pol ␩ were compared (Fig. 3B). Except for HEK293T cells, REV1 was expressed at similar levels. 17-AAG typically induces proteasomal degradation of Hsp90 clients; therefore, we next examined effects of proteasome inhibitors. As shown in Fig. 4A, lactacystin and MG-132 abolished the 17-AAG-induced reduction of REV1 in OUMS36T-3F and DU-145 cells, respectively. Cycloheximide chase analysis confirmed that 17-AAG significantly reduced the halflife of REV1 in OUMS-36T-3F cells (Fig. 4B). In addition, 17-AAG had little effect on REV1 mRNA levels in OUMS36T-3F, DU-145, and U2OS cells (Fig. 4C). Taken together, these results indicate that the 17-AAG-induced reduction of endogenous REV1 is mostly explained by acceleration of proteasomal degradation. Proteasomal degradation of various

proteins is usually preceded by polyubiquitination catalyzed by specific E3 ubiquitin ligases. To examine effects of 17-AAG on polyubiquitination of REV1, we used OUMS/GFP-REV1 cells. In these cells, lactacystin by itself markedly increased steady-state levels of GFP-REV1, indicating that its basal turnover rate is rapid, probably due to overexpression (Fig. 4D, lower panel). Under this condition, 17-AAG still significantly increased polyubiquitination of REV1 (Fig. 4D, upper panel, compare lanes 3 and 4), suggesting that an E3 ubiquitin ligase is involved in targeting REV1 to the proteasome. 17-AAG suppresses UV-induced nuclear focus formation of REV1. Although Hsp90 inhibition had little effect on REV1 protein levels in several cell lines, it is still possible that Hsp90 regulates the function of REV1 by regulating its conformation. Recruitment of REV1 to DNA damage-induced stalled forks and postreplication gaps is a key regulatory step for its functions. To study the role of Hsp90 in this step, we first monitored effects of 17-AAG on UV-induced nuclear focus formation of REV1 in 293T/GFP-REV1 cells. In these cells at basal states, GFP-REV1 was diffusely distributed in the nucleoplasm, while 5 to 20 h after UV irradiation, GFP-REV1 accumulated in nuclear foci that colocalized with PCNA (Fig. 5A), consistent with previous observations (15, 16, 34). In the UVirradiated cells, 17-AAG suppressed REV1 focus formation but had little effect on PCNA focus formation (Fig. 5A), raising the possibility that Hsp90 regulates relocalization of REV1 to sites of DNA damage. Consistent with previous reports (15, 16), GFP-delBRCT-REV1 and GFP-UBM mt-REV1 were defective for UV-induced accumulation in replication foci (Fig.

3400

MAYCA POZO ET AL.

FIG. 3. 17-AAG induces reduction of REV1 protein levels in a cell type-dependent manner. (A) OUMS-36T-3F, DU-145, XP2SASV3, U2OS, and HEK293T cells were treated with 1 ␮M 17-AAG for the indicated times, and WCL were immunoblotted as indicated. REV1 signals were quantified and normalized against GAPDH signals. Data represent means ⫾ SD from three independent experiments (lower graphs). (B) REV1 protein levels in various cell lines. WCL were immunoblotted as indicated.

5B). Of note, a significant portion of GFP-delBRCT-REV1 was distributed in the cytoplasm after Triton X-100 extraction, whereas almost no signals of GFP-UMB mt-REV1 were detected after the extraction. A recent study by Akagi et al. showed that interaction between Pol ␩ and REV1 promotes targeting of endogenous REV1 to sites of UV-induced DNA damage (1). Since 17AAG suppresses Pol ␩ accumulation in replication foci (50), 17-AAG-induced inhibition of REV1 focus formation may depend on its interaction with Pol ␩. To test this possibility, we stably expressed GFP-delPID-REV1 in HEK293T cells and GFP-REV1 in Pol ␩-deficient cells. In both cell lines, UV irradiation induced REV1 accumulation in nuclear foci, which

MOL. CELL. BIOL.

colocalized with PCNA (Fig. 5B and C). These results are consistent with previous observations (34, 55) and suggest that REV1 is targeted to sites of UV-induced DNA damage in a Pol ␩-independent manner, at least when expressed at high levels. In these cell lines, 17-AAG also inhibited REV1 focus formation, indicating that the major effect of the drug is independent of the interaction between REV1 and Pol ␩. Effects of siRNA-mediated depletion of Hsp90 on REV1 levels and focus formation. To further verify the role of Hsp90 in the regulation of REV1, we first examined effects of siRNAmediated Hsp90 reduction on REV1 protein levels. Transfection of siRNAs for Hsp90␣ and Hsp90␤, major isoforms of Hsp90, into OUMS-36T-3F cells induced a marked reduction in REV1 protein levels, which was partially restored by treatment with lactacystin (Fig. 6A). On the other hand, siRNAmediated knockdown of Hsp90 had little effect on REV1 protein levels in HEK293T cells (Fig. 6A). Furthermore, Hsp90 knockdown in 293T/GFP-REV1 cells inhibited UV-induced nuclear focus formation of GFP-REV1 without significant effects on its protein levels (Fig. 6B). These results are consistent with the effects of 17-AAG on REV1 protein levels and localization described above and confirm the Hsp90-mediated regulation of REV1. Hsp90 regulates the interaction between REV1 and UbPCNA. We next studied the biochemical basis of the 17-AAGinduced suppression of REV1 nuclear focus formation in 293T/GFP-REV1 cells. For this purpose, we prepared WCL from 293T/GFP-REV1 cells treated with UV irradiation and 17-AAG under conditions comparable to those described for Fig. 5A. Immunoblot analyses of these samples revealed that Ub-PCNA levels increased in a time course similar to that of nuclear focus formation, whereas effects of 17-AAG on GFPREV1 and Ub-PCNA levels were mild and negligible, respectively (Fig. 7A). Based on the above results, we reasoned that Hsp90 might regulate the recruitment of REV1 to chromatin at sites of DNA damage by affecting the interaction between REV1 and Ub-PCNA. To test this idea, we analyzed the CSK-insoluble fraction. The results showed that 17-AAG suppressed a UVinduced increase in chromatin-bound GFP-REV1, with little change in Ub-PCNA levels (Fig. 7B, lower panels). Furthermore, 17-AAG significantly inhibited the UV-induced interaction between GFP-REV1 and Ub-PCNA (Fig. 7B, upper panels). Of note, 17-AAG suppressed the interaction between endogenous REV1 and Ub-PCNA (Fig. 7C), confirming the physiological role of Hsp90 in the regulation of this protein interaction. To characterize the interaction between REV1 and UbPCNA, we analyzed REV1 mutant proteins stably expressed in HEK293T cells. GFP-delPID-REV1 and GFP-UBM mtREV1 levels in WCL were similar to GFP-REV1 levels, whereas GFP-delBRCT-REV1 levels were much higher (Fig. 7D). Similarly to GFP-REV1, GFP-delPID-REV1 interacted with Ub-PCNA after UV irradiation in a 17-AAG-sensitive manner (Fig. 7E). GFP-delBRCT-REV1 was relatively abundant in the CSK-insoluble fraction, probably reflecting its localization in the Triton X-100-insoluble cytoplasmic fraction. Indeed, the interaction between GFP-delBRCT-REV1 and Ub-PCNA was undetectable (Fig. 7E). GFP-UBM mt-REV1 was not detectable in the CSK-insoluble fraction, nor in anti-

VOL. 31, 2011

Hsp90 REGULATES REV1

3401

FIG. 4. 17-AAG induces proteasomal degradation of REV1. (A) OUMS-36T-3F cells were treated with 1 ␮M 17-AAG with vehicle or 10 ␮M lactacystin for 20 h. DU-145 cells were treated with 1 ␮M 17-AAG for 12 h and with 10 ␮M MG-132 for last 4 h of culture. WCL were immunoblotted as indicated. (B) OUMS-36T-3F cells were treated with 100 ␮g/ml cycloheximide (CHX) with or without 1 ␮M 17-AAG for the indicated times, and WCL were immunoblotted as indicated. REV1 signals were quantified and normalized against GAPDH signals. Data represent means ⫾ SD from three independent experiments (lower graph). (C) OUMS-36T-3F, DU-145, and U2OS cells were treated with 1 ␮M 17-AAG for the indicated times. REV1 mRNA levels were estimated by semiquantitative RT-PCR. GAPDH mRNA levels were measured as an internal control. (D) OUMS/GFP-REV1 cells were treated with 1 ␮M 17-AAG and/or 10 ␮M lactacystin for 12 h. Cell lysates were immunoprecipitated (IP) followed by immunoblotting or directly immunoblotted, as indicated. An arrow indicates monoubiquitinated GFP-REV1 (16).

PCNA immunoprecipitates (Fig. 7E). Overall, these biochemical data are consistent with the results of the immunofluorescence studies shown in Fig. 5. To further analyze the molecular mechanism by which Hsp90 regulates the interaction between REV1 and UbPCNA, we performed GST pulldown experiments. Our results showed that GFP-REV1 extracted from 293T/GFP-REV1 cells specifically bound GST-PCNA and GST-Ub (Fig. 8A). Of note, GFP-REV1 extracted from 17-AAG-treated cells exhibited significantly lower binding to GST-PCNA, whereas GFPREV1 binding to GST-Ub was not affected by Hsp90 inhibition (Fig. 8A). Consistent with the results of nuclear focus formation and the in vivo interaction, GST-PCNA pulldown of GFP-delPID-REV1 was inhibited similarly to that of GFPREV1 (Fig. 8B). To complement these results, binding of GFP-REV1 to GST-Pol ␩ (350–713) or GST-REV7 was not affected by Hsp90 inhibition (Fig. 8A). Collectively, these results suggest that 17-AAG-induced disruption of the in vivo

interaction between REV1 and Ub-PCNA is, at least in part, attributable to impaired REV1-binding to PCNA. In these experiments, PARP-1 and HDAC6 were employed as internal controls of GST-PCNA and GST-Ub pulldowns, respectively. PARP-1 binds directly to PCNA, and HDAC6 binds directly to Ub (5, 11). As described previously (50), 17-AAG had no effect on GST-PCNA pulldown of PARP-1 or GST-Ub pulldown of HDAC6. To ensure that the effect of 17-AAG depends on Hsp90, GFP-REV1 (1–152), which contains the BRCT domain but fails to bind Hsp90 (Fig. 2), was transiently expressed in HEK293T cells and subjected to GST pulldown experiments. The results showed that 17-AAG had little effect on binding of this fragment to GST-PCNA (Fig. 8C), suggesting that the Hsp90-mediated folding of REV1 regulates the accessibility of the BRCT domain. Effects of Hsp90 inhibition on REV1-mediated mutagenesis. We reasoned that Hsp90 may have paradoxical effects on UVinduced mutagenesis. Hsp90 suppresses UV-induced mutagen-

3402

MAYCA POZO ET AL.

MOL. CELL. BIOL.

FIG. 5. Effects of Hsp90 inhibition on UV-induced nuclear focus formation of REV1. (A) 293T/GFP-REV1 cells were irradiated with 6 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG. GFP-REV1 focus-positive cells and PCNA focus-positive cells were scored at the indicated time points (lower graphs). Data represent means ⫾ SD for three independent experiments. Representative images were obtained from cells 18 h after treatment. (B) HEK293T cells stably expressing GFP-delPID-REV1, GFP-delBRCT-REV1, and GFP-UBM mt-REV1 were irradiated with 6 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG for 18 h. Then, cells were examined as described above. Data represent means ⫾ SD for three independent experiments. (C) XP2SASV3 cells stably expressing GFP-REV1 were irradiated with 4.5 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG for 18 h. Then, cells were examined as described above. Data represent means ⫾ SD for three independent experiments.

VOL. 31, 2011

Hsp90 REGULATES REV1

3403

FIG. 6. Effects of siRNA-mediated Hsp90 depletion on REV1 protein expression and focus formation. (A) Hsp90␣ and Hsp90␤ siRNAs were transfected into OUMS-36T-3F and HEK293T cells using Lipofectamine 2000 (Invitrogen). LacZ siRNA was used as a control. After 72 h, WCL were immunoblotted as indicated. Cells were treated with vehicle or 10 ␮M lactacystin for the last 20 h of culture. The fold increase is shown under the panels. (B) 293T/GFP-REV1 cells were transfected with Hsp90␣ and Hsp90␤ siRNAs. After 48 h, cells were irradiated with 6 J/m2 UV and subsequently incubated for 18 h. Then, GFP-REV1 focus-positive cells were scored. Data represent means ⫾ SD for three independent experiments (upper graph). WCL were immunoblotted as indicated (lower panels).

esis by promoting accurate Pol ␩-mediated TLS, as reported in our recent study (50). On the other hand, Hsp90 may promote REV1-mediated mutagenic TLS across UV-induced DNA damage. Therefore, we used a Pol ␩-deficient cell line, XP2SASV3, to study the role of Hsp90 in the regulation of REV1-mediated TLS. UV-irradiated pSP189, a supF shuttle vector (45), was transfected into these XP2SASV3 cells. In a recent study by Akagi et al. (1), the interaction of REV1 with Pol ␩ was described as promoting REV1 targeting to replication foci, raising a possibility that REV1 might not function properly in Pol ␩-deficient cells. However, analyses of Pol ␩ mutants in this report indicated that disruption of the interaction between Pol ␩ and REV1 does not affect UV-induced mutagenesis, suggesting that REV1-mediated mutagenesis occurs in a Pol ␩-independent manner. Indeed, siRNA-mediated reduction of REV1 in XP2SASV3 cells significantly decreased mutation rates of the supF gene, indicating that REV1-mediated TLS contributes to mutagenesis in this assay system (Fig. 9A and B). In XP2SASV3 cells, 17-AAG reduced REV1 protein levels to an extent similar to that of siRNA-mediated reduction (Fig. 9C); therefore, we predicted that the drug would suppress UV-induced mutagenesis. In fact, in LacZ siRNA-transfected cells, 17-AAG decreased supF mutation rates to a level comparable to that observed in REV1 siRNAtransfected cells. Combined transfection of REV1 siRNA and 17-AAG treatment reduced REV1 protein levels to a greater extent; however, no additional decrease in mutation rates was observed. These results indicate that 17-AAG affects supF mutagenesis mainly through reduction of REV1. Sequencing mutant supF clones from each culture revealed that neither 17AAG nor siRNA-mediated reduction of REV1 caused significant changes in the UV-induced mutation spectrum of the gene (Fig. 10). In an attempt to perform another functional assay of REV1mediated TLS, we first tested effects of siRNA-mediated de-

pletion of REV1 on UV-induced cytotoxicity in XP2SASV3 cells. These cells were highly sensitive to UV, and cell survival curves were obtained in the range of 0 to 5 J/m2. However, no significant difference was observed between REV1-depleted cells and control cells (data not shown), indicating that it is difficult to assess effects of Hsp90 inhibition on REV1 function using this assay. These results are consistent with previous observations, which showed that antisense or ribozyme-mediated suppression of REV1 levels significantly decreased UVinduced mutagenesis but had little effects on UV-induced cytotoxicity (8, 12). Thus, the mutagenesis assay seems more sensitive to reduced REV1 protein levels compared with the cell survival assay. DISCUSSION REV1 is a member of the Y-Pol family and plays a central role in the introduction of mutations through promoting errorprone TLS (10, 14, 28, 43, 58). While emerging evidence indicates that posttranslational control of REV1 plays an important regulatory role in yeast (20, 44, 49, 60), little is known concerning the regulatory mechanisms of REV1 in mammalian cells. The present study provides several lines of evidence that Hsp90 is a critical regulator of REV1. Hsp90 specifically associated with REV1 in vivo and in vitro. In addition, a specific inhibitor of Hsp90, 17-AAG, reduced REV1 protein levels in several cell types through proteasomal degradation. This reduction was associated with decreased UV-induced mutagenesis. Furthermore, 17-AAG suppressed the recruitment of REV1 to sites of DNA damage by disrupting the interaction between REV1 and Ub-PCNA. From these observations, we conclude that Hsp90 is required for REV1 to be properly folded into a stable and functional conformation. REV1 interacts with Pol ␩ (1, 13, 27, 42), and as we recently reported (50), Hsp90 recognizes Pol ␩ as a client; therefore,

FIG. 7. 17-AAG inhibits the in vivo interaction between REV1 and Ub-PCNA. (A) 293T/GFP-REV1 cells were irradiated with 6 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG for the indicated times. WCL were immunoblotted as indicated. GFP-REV1 and Ub-PCNA signals were quantified and normalized against GAPDH signals (lower graphs). Data represent means ⫾ SD from three independent experiments. (B) 293T/GFP-REV1 cells were irradiated with 6 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG for 18 h. Cell extracts from the CSK buffer-insoluble fraction were prepared as described in “Materials and Methods” and were immunoprecipitated (IP) followed by immunoblotting or directly immunoblotted, as indicated. (C) HEK293T cells were irradiated with 6 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG for 24 h. The CSK buffer-insoluble fraction was analyzed as described above. (D) WCL were prepared from 293T/GFP-REV1 cells and HEK293T cells stably expressing GFP-delPID-REV1, GFP-delBRCT-REV1, or GFP-UBM mt-REV1 and were immunoblotted as indicated. (E) HEK293T cells stably expressing GFP-delPID-REV1, GFP-delBRCT-REV1, or GFP-UBM mt-REV1 and 293T/GFP-REV1 cells were irradiated with 6 J/m2 UV and subsequently incubated with vehicle or 1 ␮M 17-AAG for 18 h. The CSK bufferinsoluble fraction was analyzed as described above. 3404

VOL. 31, 2011

Hsp90 REGULATES REV1

3405

FIG. 8. 17-AAG inhibits the in vitro interaction between REV1 and PCNA. 293T/GFP-REV1 (A) or 293T/GFP-delPID-REV1 (B) cells were treated with vehicle or 1 ␮M 17-AAG for 18 h. Cell lysates were incubated with 5 ␮g of GST fusion proteins. Cell lysates and pulldown fractions were immunoblotted as indicated. Relative amounts of GFP-REV1 proteins were determined (bottom graphs). Data represent means ⫾ SD from three independent experiments. Statistical analyses were performed using a two-tailed Student t test. ⴱ, P ⬍ 0.05; NS, not significant. PARP-1 and HDAC6 serve as an internal control of PCNA- and Ub-binding proteins, respectively. (C) GFP-REV1 (1–152) was transiently expressed in HEK293T cells, and its binding to GST-PCNA was analyzed as described above.

Hsp90 may indirectly bind REV1. However, the interaction between REV1 and Hsp90 was also observed in Pol ␩-deficient cells, indicating that Pol ␩ is not required for the interaction. In addition, analysis of REV1 deletion mutants revealed that the PID that mediates the interaction with Y-Pols, including Pols ␩, ␫, and ␬ (13, 27, 42), was not required for the interaction with Hsp90. Furthermore, the BRCT domain and the UBMcontaining region, which are critical for protein-protein interactions of REV1, were unnecessary for the interaction with Hsp90. These results indicate that Hsp90 directly recognizes REV1. This notion is further supported by the observation that in vitro-synthesized REV1 protein was associated with Hsp90. REV1 protein levels affect mutation rates (9, 29, 62); however, little is known about the mechanisms regulating REV1 protein levels in mammalian cells. Our results indicate that Hsp90 regulates REV1 protein levels in several cell lines. 17AAG significantly decreased REV1 protein levels by accelerating proteasomal degradation. Although a precise reason for the cell type-dependent variability of 17-AAG effects is unknown, it might be simply explained by different expression levels of a REV1-targeting E3 ubiquitin ligase. A chaperoneassociated ubiquitin ligase, CHIP, which catalyzes polyubiquitination of some Hsp90-interacting clients, is an attractive candidate (31, 63). However, in our preliminary results, CHIP expression levels did not correlate with 17-AAG-induced reduction of REV1, and short hairpin RNA (shRNA)-mediated

depletion of CHIP had no effects on 17-AAG-induced reduction of REV1 levels (data not shown). Of note, the 17-AAGinduced reduction of REV1 levels suppressed UV-induced mutagenesis. Hence, clarification of the molecular mechanism for Hsp90-mediated regulation of REV1 levels, including identification of a REV1-targeting E3 ligase, will be an important challenge for future research. Mild reduction of REV1 levels in HEK293T cells allowed us to study the regulatory role of Hsp90 in REV1 protein-protein interactions. An initial observation was that Hsp90 inhibition suppressed UV-induced relocalization of GFP-REV1 to nuclear foci. This may reflect inhibition of the recruitment of REV1 to sites of DNA damage via binding to Ub-PCNA. Indeed, Hsp90 inhibition disrupts the interaction between GFP-REV1 and Ub-PCNA in vivo. Similar results were obtained with endogenous REV1, supporting the physiological significance of this regulatory mechanism. These results suggest that Hsp90 promotes folding of REV1 into a functional form to bind Ub-PCNA. To strengthen this idea, our GST pulldown experiments showed that 17-AAG suppressed in vitro binding of REV1 to PCNA. Since the BRCT domain is critical for PCNA binding in this assay (15), these results suggest that Hsp90-mediated folding of REV1 promotes the accessibility of the BRCT domain for efficient interaction with Ub-PCNA. To support this notion, we confirmed a previous observation that UV-induced nuclear focus formation of the REV1 mutant

3406

MAYCA POZO ET AL.

MOL. CELL. BIOL.

FIG. 9. Effects of 17-AAG on UV-induced mutation rates. LacZ or REV1 siRNA was transfected into XP2SASV3 cells using Lipofectamine 2000 (Invitrogen). After 24 h of the siRNA transfection, UV-irradiated pSP189 was transfected into the cells using FuGENE6 and incubated for 54 h. 17-AAG (1 ␮M) or vehicle was added for the last 42 h of culture. Then, replicated pSP189 was recovered and WCL were prepared from each culture. (A) The supF mutant frequencies were determined in three independent experiments (experiments 1, 2, and 3), as described in Materials and Methods. *, P ⬍ 0.01; **, P ⬍ 0.05 (as determined by chi-square test). (B) Relative mean fold changes in mutation frequency from the three experiments are shown with SD. Statistical analyses were performed using a two-tailed Student t test. *, P ⬍ 0.01; **, P ⬍ 0.05. (C) WCL were immunoblotted with the indicated antibodies.

protein lacking the BRCT domain was significantly attenuated compared with that of full-length REV1. We also confirmed previous findings that UBMs are critical for UV-induced focus formation of REV1 (16). Although we could not detect any significant effect of Hsp90 inhibition on GST-Ub pulldown of REV1, it is still possible that Hsp90 regulates UBM-mediated protein interactions in vivo. One interesting possibility is that DNA damage-induced modification of REV1 and/or expression of cofactors might be required for the Hsp90-mediated regulation of REV1-binding to Ub. An alternative, but not mutually exclusive, possibility is that Hsp90 indirectly regulates REV1 focus formation. For instance, the Fanconi anemia (FA) “core” complex, containing eight FA proteins, such as FANCA and FA-associated proteins, is required for efficient REV1 focus formation and REV1-mediated mutagenesis (19, 32). Since Hsp90 inhibition induces proteasomal degradation and cytoplasmic retention of FANCA and consequent disruption of the FA core complex (39), it is plausible that the effects of 17-AAG on REV1 focus formation and REV1-mediated mutagenesis are, at least in part, mediated by the effect of the Hsp90 inhibitor on the FA core complex. What is the overall regulatory role of Hsp90 for TLS? Our

knowledge is too premature to answer this question. For instance, a noncatalytic regulatory function of REV1 in TLS is poorly defined, and Hsp90 may have other targets in the TLS machinery. In a current model, TLS consists of two major processes, insertion and extension: first, a TLS polymerase incorporates nucleotides opposite DNA lesions, and then a second TLS polymerase extends a few additional nucleotides before restart of replicative DNA synthesis (10, 14, 28, 43, 58). This two-polymerase model was previously proposed based on in vitro studies and was further supported by a recent study using a plasmid-based assay in human cells (51, 66). This study indicated that UV-induced cyclobutane pyrimidine dimers are bypassed by Pol ␩ by itself in an error-free manner, whereas a combination of Pol ␫ or Pol ␬-mediated error-prone insertion and a Pol ␨-mediated extension has a major role in mutagenic TLS across the DNA lesion in Pol ␩-deficient cells. It is proposed that REV1 functions in the latter mutagenic TLS pathway through its interaction with Pol ␫ and Pol ␬ as well as Pol ␨, possibly facilitating a switch of Pols (10, 14, 28, 43, 58). This model explains the apparently paradoxical effects of Hsp90 inhibition on UV-induced mutagenesis in Pol ␩-expressing and Pol ␩-deficient cells in our previous and present studies (50).

VOL. 31, 2011

Hsp90 REGULATES REV1

3407

FIG. 10. Effects of 17-AAG on the mutation spectrum of supF. At least 50 pSP189 mutant clones were obtained from each culture of XP2SASV3 cells treated as described in the legend to Fig. 9, and these were sequenced. Mutations within the structural supF gene are shown (upper sequence). Tandem base substitutions are underlined. Deletions are indicated with lines. Types of mutations in the supF gene are shown in the lower graph.

Thus, the impact of Hsp90 inhibition probably depends on what TLS pathways and DNA lesions predominate. The present findings have several important implications for cancer prevention and therapeutics. Carcinogenic activities of genotoxic agents, such as UV and benzo[a]pyrene, are mediated mainly by introduction of point mutations through errorprone TLS. Indeed, recent sequencing of the genomes of cell lines derived from a melanoma and a lung cancer indicated that the majority of mutations were derived from mutagenic TLS across DNA lesions induced by UV and carcinogenic chemicals contained in tobacco smoke, respectively (47, 48). Since REV1 plays a critical role in TLS-mediated mutagenesis

(12, 33), it is plausible that REV1 contributes to DNA damageinduced carcinogenesis. This notion is supported by a recent report showing that reduced expression of REV1 suppresses benzo[a]pyrene-induced development of lung cancer in mice (9). Based on this and our present findings, it is reasonable to speculate that Hsp90 inhibitors might be useful for suppressing DNA damage-induced carcinogenesis in vivo. In addition, REV1-mediated mutagenesis may contribute to tumor progression. To support this notion, increased expression of REV1 elevates the rate at which cultured ovarian cancer cells acquire cisplatin resistance (29). Furthermore, a recent study documented that decreased expression of REV1 prevents cy-

3408

MAYCA POZO ET AL.

clophosphamide-induced mutagenesis and acquisition of chemoresistance to the drug in B-cell lymphoma (62). Given that Hsp90 activity is increased in tumor cells (24), the impact of Hsp90-targeting drugs on REV1-mediated mutagenesis may be more significant in prevention of chemoresistance acquisition. In conclusion, our present work provides new insights into the pharmacological actions of Hsp90 inhibitors. ACKNOWLEDGMENTS We thank M. Seidman (National Institutes of Health) and K. Tanaka (Osaka University) for kindly providing pSP189 and XP2SASV3 cells, respectively. We thank K. Tomizawa for technical help. This work was supported by a Grant-in-Aid for Scientific Research, a grant from the Global COE Program from the Ministry of Education, Science, Technology, Sports and Culture of Japan, and grants from the Ministry of Health, Labor and Welfare of Japan. REFERENCES 1. Akagi, J., et al. 2009. Interaction with DNA polymerase ␩ is required for nuclear accumulation of REV1 and suppression of spontaneous mutations in human cells. DNA Repair (Amst.) 8:585–599. 2. Arlander, S. J., et al. 2003. Hsp90 inhibition depletes Chk1 and sensitizes tumor cells to replication stress. J. Biol. Chem. 278:52572–52577. 3. Bienko, M., et al. 2005. Ubiquitin-binding domains in Y-family polymerases regulate translesion synthesis. Science 310:1821–1824. 4. Bienko, M., et al. 2010. Regulation of translesion synthesis DNA polymerase ␩ by monoubiquitination. Mol. Cell 37:396–407. 5. Boyault, C., K. Sadoul, M. Pabion, and S. Khochbin. 2007. HDAC6, at the crossroads between cytoskeleton and cell signaling by acetylation and ubiquitination. Oncogene 26:5468–5476. 6. Chen, Y. W., et al. 2008. Human DNA polymerase ␩ activity and translocation is regulated by phosphorylation. Proc. Natl. Acad. Sci. U. S. A. 105: 16578–16583. 7. Choi, J. H., and G. P. Pfeifer. 2005. The role of DNA polymerase ␩ in UV mutational spectra. DNA Repair (Amst.) 4:211–220. 8. Clark, D. R., W. Zacharias, L. Panaitescu, and W. G. McGregor. 2003. Ribozyme-mediated REV1 inhibition reduces the frequency of UV-induced mutations in the human HPRT gene. Nucleic Acids Res. 31:4981–4988. 9. Dumstorf, C. A., S. Mukhopadhyay, E. Krishnan, B. Haribabu, and W. G. McGregor. 2009. REV1 is implicated in the development of carcinogeninduced lung cancer. Mol. Cancer Res. 7:247–254. 10. Friedberg, E. C., A. R. Lehmann, and R. P. Fuchs. 2005. Trading places: how do DNA polymerases switch during translesion DNA synthesis? Mol. Cell 18:499–505. 11. Frouin, I., et al. 2003. Human proliferating cell nuclear antigen, poly(ADPribose) polymerase-1, and p21waf1/cip1. A dynamic exchange of partners. J. Biol. Chem. 278:39265–39268. 12. Gibbs, P. E., et al. 2000. The function of the human homolog of Saccharomyces cerevisiae REV1 is required for mutagenesis induced by UV light. Proc. Natl. Acad. Sci. U. S. A. 97:4186–4191. 13. Guo, C., et al. 2003. Mouse Rev1 protein interacts with multiple DNA polymerases involved in translesion DNA synthesis. EMBO J. 22:6621–6630. 14. Guo, C., J. N. Kosarek-Stancel, T. S. Tang, and E. C. Friedberg. 2009. Y-family DNA polymerases in mammalian cells. Cell Mol. Life Sci. 66:2363– 2381. 15. Guo, C., et al. 2006. REV1 protein interacts with PCNA: significance of the REV1 BRCT domain in vitro and in vivo. Mol. Cell 23:265–271. 16. Guo, C., et al. 2006. Ubiquitin-binding motifs in REV1 protein are required for its role in the tolerance of DNA damage. Mol. Cell. Biol. 26:8892–8900. 17. Haracska, L., S. Prakash, and L. Prakash. 2002. Yeast Rev1 protein is a G template-specific DNA polymerase. J. Biol. Chem. 277:15546–15551. 18. Haracska, L., et al. 2001. Roles of yeast DNA polymerases ␦ and ␨ and of Rev1 in the bypass of abasic sites. Genes Dev. 15:945–954. 19. Hicks, J. K., et al. 2010. Differential roles for DNA polymerases eta, zeta, and REV1 in lesion bypass of intrastrand versus interstrand DNA crosslinks. Mol. Cell. Biol. 30:1217–1230. 20. Hirano, Y., and K. Sugimoto. 2006. ATR homolog Mec1 controls association of DNA polymerase ␨-Rev1 complex with regions near a double-strand break. Curr. Biol. 16:586–590. 21. Hoege, C., B. Pfander, G. L. Moldovan, G. Pyrowolakis, and S. Jentsch. 2002. RAD6-dependent DNA repair is linked to modification of PCNA by ubiquitin and SUMO. Nature 419:135–141. 22. Jansen, J. G., et al. 2009. Separate domains of Rev1 mediate two modes of DNA damage bypass in mammalian cells. Mol. Cell. Biol. 29:3113–3123. 23. Jansen, J. G., et al. 2005. The BRCT domain of mammalian Rev1 is involved in regulating DNA translesion synthesis. Nucleic Acids Res. 33:356–365.

MOL. CELL. BIOL. 24. Kamal, A., et al. 2003. A high-affinity conformation of Hsp90 confers tumour selectivity on Hsp90 inhibitors. Nature 425:407–410. 25. Kannouche, P. L., J. Wing, and A. R. Lehmann. 2004. Interaction of human DNA polymerase ␩ with monoubiquitinated PCNA: a possible mechanism for the polymerase switch in response to DNA damage. Mol. Cell 14:491– 500. 26. Kim, S. H., and W. M. Michael. 2008. Regulated proteolysis of DNA polymerase ␩ during the DNA-damage response in C. elegans. Mol. Cell 32:757– 766. 27. Kosarek, J. N., et al. 2008. Comparative analysis of in vivo interactions between Rev1 protein and other Y-family DNA polymerases in animals and yeasts. DNA Repair (Amst.) 7:439–451. 28. Lehmann, A. R., et al. 2007. Translesion synthesis: Y-family polymerases and the polymerase switch. DNA Repair (Amst.) 6:891–899. 29. Lin, X., T. Okuda, J. Trang, and S. B. Howell. 2006. Human REV1 modulates the cytotoxicity and mutagenicity of cisplatin in human ovarian carcinoma cells. Mol. Pharmacol. 69:1748–1754. 30. Masuda, Y., M. Takahashi, S. Fukuda, M. Sumii, and K. Kamiya. 2002. Mechanisms of dCMP transferase reactions catalyzed by mouse Rev1 protein. J. Biol. Chem. 277:3040–3046. 31. McDonough, H., and C. Patterson. 2003. CHIP: a link between the chaperone and proteasome systems. Cell Stress Chaperones 8:303–308. 32. Mirchandani, K. D., R. M. McCaffrey, and A. D. D’Andrea. 2008. The Fanconi anemia core complex is required for efficient point mutagenesis and Rev1 foci assembly. DNA Repair (Amst.) 7:902–911. 33. Mukhopadhyay, S., D. R. Clark, N. B. Watson, W. Zacharias, and W. G. McGregor. 2004. REV1 accumulates in DNA damage-induced nuclear foci in human cells and is implicated in mutagenesis by benzo[a]pyrenediolepoxide. Nucleic Acids Res. 32:5820–5826. 34. Murakumo, Y., S. Mizutani, M. Yamaguchi, M. Ichihara, and M. Takahashi. 2006. Analyses of ultraviolet-induced focus formation of hREV1 protein. Genes Cells 11:193–205. 35. Murakumo, Y., et al. 2001. Interactions in the error-prone postreplication repair proteins hREV1, hREV3, and hREV7. J. Biol. Chem. 276:35644– 35651. 36. Murakumo, Y., et al. 2000. A human REV7 homolog that interacts with the polymerase ␨ catalytic subunit hREV3 and the spindle assembly checkpoint protein hMAD2. J. Biol. Chem. 275:4391–4397. 37. Nelson, J. R., P. E. Gibbs, A. M. Nowicka, D. C. Hinkle, and C. W. Lawrence. 2000. Evidence for a second function for Saccharomyces cerevisiae Rev1p. Mol. Microbiol. 37:549–554. 38. Nelson, J. R., C. W. Lawrence, and D. C. Hinkle. 1996. Deoxycytidyl transferase activity of yeast REV1 protein. Nature 382:729–731. 39. Oda, T., T. Hayano, H. Miyaso, N. Takahashi, and T. Yamashita. 2007. Hsp90 regulates the Fanconi anemia DNA damage response pathway. Blood 109:5016–5026. 40. Ogawa, M., et al. 2008. Rines/RNF180, a novel RING finger gene-encoded product, is a membrane-bound ubiquitin ligase. Genes Cells 13:397–409. 41. Ohashi, E., et al. 2009. Identification of a novel REV1-interacting motif necessary for DNA polymerase ␬ function. Genes Cells 14:101–111. 42. Ohashi, E., et al. 2004. Interaction of hREV1 with three human Y-family DNA polymerases. Genes Cells 9:523–531. 43. Ohmori, H., T. Hanafusa, E. Ohashi, and C. Vaziri. 2009. Separate roles of structured and unstructured regions of Y-family DNA polymerases. Adv. Protein Chem. Struct. Biol. 78:99–146. 44. Pages, V., S. R. Santa Maria, L. Prakash, and S. Prakash. 2009. Role of DNA damage-induced replication checkpoint in promoting lesion bypass by translesion synthesis in yeast. Genes Dev. 23:1438–1449. 45. Parris, C. N., and M. M. Seidman. 1992. A signature element distinguishes sibling and independent mutations in a shuttle vector plasmid. Gene 117:1–5. 46. Pearl, L. H., C. Prodromou, and P. Workman. 2008. The Hsp90 molecular chaperone: an open and shut case for treatment. Biochem. J. 410:439–453. 47. Pleasance, E. D., et al. 2010. A comprehensive catalogue of somatic mutations from a human cancer genome. Nature 463:191–196. 48. Pleasance, E. D., et al. 2010. A small-cell lung cancer genome with complex signatures of tobacco exposure. Nature 463:184–190. 49. Sabbioneda, S., I. Bortolomai, M. Giannattasio, P. Plevani, and M. MuziFalconi. 2007. Yeast Rev1 is cell cycle regulated, phosphorylated in response to DNA damage and its binding to chromosomes is dependent upon MEC1. DNA Repair (Amst.) 6:121–127. 50. Sekimoto, T., et al. 2010. The molecular chaperone Hsp90 regulates accumulation of DNA polymerase ␩ at replication stalling sites in UV-irradiated cells. Mol. Cell 37:79–89. 51. Shachar, S., et al. 2009. Two-polymerase mechanisms dictate error-free and error-prone translesion DNA synthesis in mammals. EMBO J. 28:383–393. 52. Skoneczna, A., J. McIntyre, M. Skoneczny, Z. Policinska, and E. Sledziewska-Gojska. 2007. Polymerase eta is a short-lived, proteasomally degraded protein that is temporarily stabilized following UV irradiation in Saccharomyces cerevisiae. J. Mol. Biol. 366:1074–1086. 53. Taipale, M., D. F. Jarosz, and S. Lindquist. 2010. HSP90 at the hub of protein homeostasis: emerging mechanistic insights. Nat. Rev. Mol. Cell Biol. 11:515–528.

VOL. 31, 2011 54. Terai, K., T. Abbas, A. A. Jazaeri, and A. Dutta. 2010. CRL4Cdt2 E3 ubiquitin ligase monoubiquitinates PCNA to promote translesion DNA synthesis. Mol. Cell 37:143–149. 55. Tissier, A., et al. 2004. Co-localization in replication foci and interaction of human Y-family members, DNA polymerase pol ␩ and REVl protein. DNA Repair (Amst.) 3:1503–1514. 56. Trepel, J., M. Mollapour, G. Giaccone, and L. Neckers. 2010. Targeting the dynamic HSP90 complex in cancer. Nat. Rev. Cancer 10:537–549. 57. Watanabe, K., et al. 2004. Rad18 guides pol␩ to replication stalling sites through physical interaction and PCNA monoubiquitination. EMBO J. 23: 3886–3896. 58. Waters, L. S., et al. 2009. Eukaryotic translesion polymerases and their roles and regulation in DNA damage tolerance. Microbiol. Mol. Biol. Rev. 73: 134–154. 59. Waters, L. S., and G. C. Walker. 2006. The critical mutagenic translesion DNA polymerase Rev1 is highly expressed during G(2)/M phase rather than S phase. Proc. Natl. Acad. Sci. U. S. A. 103:8971–8976. 60. Wiltrout, M. E., and G. C. Walker. 2011. Proteasomal regulation of the mutagenic translesion DNA polymerase, Saccharomyces cerevisiae Rev1. DNA Repair (Amst.) 10:169–175.

Hsp90 REGULATES REV1

3409

61. Wood, A., P. Garg, and P. M. Burgers. 2007. A ubiquitin-binding motif in the translesion DNA polymerase Rev1 mediates its essential functional interaction with ubiquitinated proliferating cell nuclear antigen in response to DNA damage. J. Biol. Chem. 282:20256–20263. 62. Xie, K., J. Doles, M. T. Hemann, and G. C. Walker. 2010. Error-prone translesion synthesis mediates acquired chemoresistance. Proc. Natl. Acad. Sci. U. S. A. 107:20792–20797. 63. Xu, W., et al. 2002. Chaperone-dependent E3 ubiquitin ligase CHIP mediates a degradative pathway for c-ErbB2/Neu. Proc. Natl. Acad. Sci. U. S. A. 99:12847–12852. 64. Yamashita, T., T. Oda, and T. Sekimoto. 2007. Hsp90 and the Fanconi anemia pathway: a molecular link between protein quality control and the DNA damage response. Cell Cycle 6:2232–2235. 65. Zhang, Y., et al. 2002. Response of human REV1 to different DNA damage: preferential dCMP insertion opposite the lesion. Nucleic Acids Res. 30: 1630–1638. 66. Ziv, O., N. Geacintov, S. Nakajima, A. Yasui, and Z. Livneh. 2009. DNA polymerase ␨ cooperates with polymerases ␬ and ␫ in translesion DNA synthesis across pyrimidine photodimers in cells from XPV patients. Proc. Natl. Acad. Sci. U. S. A. 106:11552–11557.