Tumor and Stem Cell Biology Cancer Research The ... · Tumor and Stem Cell Biology The Deubiquitinating Enzyme USP17 Is Highly Expressed in ... F. Burrows1,2, and James A. Johnston1

Published OnlineFirst April 13, 2010; DOI: 10.1158/0008-5472.CAN-09-4152

Tumor and Stem Cell Biology
Cancer
Research

The Deubiquitinating Enzyme USP17 Is Highly Expressed inTumor Biopsies, Is Cell Cycle Regulated, and Is Required forG1-S Progression

Cheryl McFarlane1, Alyson A. Kelvin1, Michelle de la Vega1, Ureshnie Govender1, Christopher J. Scott2,James F. Burrows1,2, and James A. Johnston1

Abstract

Authors' AMedicine, DSchool ofQueen's Un

Note: SupResearch O

CorresponImmunity, SMedical BuPhone: 44-qub.ac.uk.

doi: 10.115

©2010 Am

www.aacr

Do

Ubiquitination is a reversible posttranslational modification that is essential for cell cycle control, and it isbecoming increasingly clear that the removal of ubiquitin from proteins by deubiquitinating enzymes (DUB) isequally important. In this study, we have identified high levels of the DUB USP17 in several tumor-derived celllines and primary lung, colon, esophagus, and cervix tumor biopsies. We also report that USP17 is tightlyregulated during the cell cycle in all the cells examined, being abundantly evident in G1 and absent in S phase.Moreover, regulated USP17 expression was necessary for cell cycle progression because its depletion signifi-cantly impaired G1-S transition and blocked cell proliferation. Previously, we have shown that USP17 regulatesthe intracellular translocation and activation of the GTPase Ras by controlling Ras-converting enzyme 1(RCE1) activation. RCE1 also regulates the processing of other proteins with a CAAX motif, including Rhofamily GTPases. We now show that USP17 depletion blocks Ras and RhoA localization and activation. More-over, our results confirm that USP17-depleted cells have constitutively elevated levels of the cyclin-dependentkinase inhibitors p21cip1 and p27kip1, known downstream targets of Ras and RhoA signaling. These observa-tions clearly show that USP17 is tightly regulated during cell division and that its expression is necessary tocoordinate cell cycle progression, and thus, it may be considered a promising novel cancer therapeutic target.Cancer Res; 70(8); 3329–39. ©2010 AACR.

Introduction

Ubiquitination is an important mechanism for regulatingcell cycle progression. Numerous components of the cell cyclemachinery, including various cyclins and cyclin-dependent ki-nase inhibitors (CDKI), are modified by ubiquitination (1).Ubiquitination is relatively well characterized and is me-diated by either the E3 ligases SKP1-CUL-F-box complexor the anaphase-promoting complex/cyclosome at distinctstages of the cell cycle (2). Similar to other covalent modifica-tions, such as phosphorylation or methylation, ubiquitination isa reversible posttranslational modification. Deubiquitinatingenzymes (DUBs) catalyze the removal of ubiquitin by hydrolyz-ing the isopeptide bond linking ubiquitin to its substrate.Accumulating evidence suggests that DUBs also play an impor-

ffiliations: 1Centre for Infection and Immunity, School ofentistry and Biomedical Sciences; 2Molecular Therapeutics,Pharmacy, Faculty of Medicine, Health and Life Sciences,iversity Belfast, Belfast, United Kingdom

plementary data for this article are available at Cancernline (http://cancerres.aacrjournals.org/).

ding Author: James A. Johnston, Centre for Infection andchool of Medicine, Dentistry and Biomedical Sciences, Whitlailding, 97 Lisburn Road, Belfast BT9 7BL, United Kingdom.289-097-2260; Fax: 44-289-032-5839; E-mail: jim.johnston@

8/0008-5472.CAN-09-4152

erican Association for Cancer Research.

journals.org

Researcon March 1cancerres.aacrjournals.org wnloaded from

tant role in the cell cycle as they help maintain the dynamicbalance between ubiquitination and deubiquitination (3, 4).Recently, several DUBs have been identified that regulate

the cell cycle and control mitotic and spindle assemblycheckpoints. For instance, anaphase initiation is facilitatedby USP44, which regulates mitosis by deubiquitinatingcdc20 (5). In addition, mitotic entry is facilitated by the generesponsible for cylindromatosis (CYLD), which modulatespolo-like kinase 1 (Plk1) activity (6). Moreover, chromosomalalignment and segregation during mitosis is also regulated bythe deubiquitination of survivin by USP9 (7).DNA damage checkpoints are regulated by USP1 and USP7.

USP1 catalyzes the deubiquitination of two important compo-nents of DNA damage repair pathways: FANCD2 and prolifer-ating cell nuclear antigen (PCNA; refs. 8, 9). USP7 (HAUSP)deubiquitinates Mdm2, and Mdmx, which alters the stabilityand activity of p53 (10). DUBs can also modulate chromatinstructure, thus altering gene transcription in a cell cycle–specific manner. Histone H2A is deubiquitinated by USP16during mitosis and enables histone phosphorylation and chro-mosomal segregation (11). Additionally, USP3 catalyzes the re-moval of ubiquitin from histones in a cell cycle–dependentmanner and is essential to promote S-phase entry (12).The importance of DUBs for cell cycle control has been

underscored by the strong association with tumorigenesis.Mutations in CYLD have been associated with cylindromato-sis, Brooke-Spiegler syndrome, and familial trichoepithelioma,

3329

h. 4, 2020. © 2010 American Association for Cancer

http://cancerres.aacrjournals.org/

McFarlane et al.

3330


all of which are predisposed to multiple skin tumors of thehead and neck (13). Additionally, downregulation of USP7has been linked with non–small cell lung carcinoma, and inconjunction with p53 status, USP7 was shown to be a prog-nostic marker for adenocarcinoma patients (14).The DUB/USP17 family of cytokine-inducible DUBs was

initially identified in murine hematopoietic cells. There areseveral murine family members (DUB-1, DUB-1A, DUB-2,and DUB-2A) and a human family member (DUB-3/USP17,subsequently referred to as USP17). USP17 family membersare immediate-early genes, which are rapidly induced follow-ing cytokine stimulation (15–18). Several lines of evidencehave implicated these enzymes in the regulation of cellgrowth and survival. Constitutive expression of murineDUB-1 has been shown to induce G1 cell cycle arrest, whereasDUB-2 expression can markedly inhibit apoptosis followingcytokine withdrawal (15, 17, 19, 20). In addition, we have con-firmed that constitutive overexpression of human USP17blocks growth factor–dependent cell proliferation (20). Morerecently, we have also shown that USP17 regulates the activityof the GTPase CAAX-box processing enzyme Ras-convertingenzyme 1 (RCE1) to alter Ras processing and membrane local-ization, and confirmed that the ability of USP17 to block cellproliferation was RCE1 dependent (21).In this study, we report high levels of USP17 expression in

many primary tumor biopsies and tumor-derived cell lines.In addition, we find that USP17 expression is tightly regulat-ed during the cell cycle and plays a fundamental role in me-diating cell cycle progression. USP17 depletion markedlyinhibits cell proliferation and prevents the passage of cellsfrom G1 into S phase. We report that USP17 depletion blocksthe translocation and proper activation of Ras and RhoA inG1, resulting in the accumulation of the CDKIs p21cip1 andp27kip1, which we suggest is the underlying cause of the G1

block. Our findings indicate that tight regulation of USP17expression is essential for G1-S transition and that targetingUSP17 could be a novel anticancer approach.

Materials and Methods

Immunohistochemistry. Immunohistochemical stainingof tissue microarray (A301, Stretton Scientific, http://www.strettonscientific.co.uk) was done using 2 μg/mL of USP17monoclonal antibody.Cell culture, transfections, and synchronization. HeLa

and MCF-7 cells were grown in 10% FCS, 1% penicillin/strep-tomycin, and 1% L-glutamine (DMEM) and maintained in a37°C, 5% CO2 incubator. Cells were transfected with 3 μgDNA using FuGENE6 (Roche). The pSUPER-USP17shRNAconstructs were a kind gift from Dr. Rene Medema (Univer-sity Medical Centre, Utrecht, the Netherlands; shRNA#1,GCAGGAAGATGCCCATGAA). The pRS-USP17shRNA (shRNA#2,GATGATTTGGCTCCTGTGGCAAGACAGCT) was from Ori-gene Technologies. Cells were synchronized at G1-S by doublethymidine block as described previously (22).Propidium iodide staining and analysis by fluorescence-

activated cell sorting. Cells were trypsinized, fixed in etha-nol for 1 h at 4°C, stained with propidium iodide (PI) solution

Cancer Res; 70(8) April 15, 2010

Researcon March 1cancerres.aacrjournals.org Downloaded from

(10 μg/mL) with RNase A (250 μg/mL), incubated at 37°C for30 min, and analyzed by FACSCanto II (BD Biosciences) andFlowJo.Immunoblotting and immunoprecipitation. Cells were

lysed in radioimmunoprecipitation assay buffer, and whole-cell lysates or immunoprecipitates were prepared as describedpreviously (21). Antibodies were as follows: anti-USP17 (FusionAntibodies); anti–pan-Ras, anti–extracellular signal-regulatedkinase (ERK) 1/2, anti–phospho-ERK1/2, anti–mitogen-activated protein kinase (MAPK)/ERK kinase (MEK) 1/2, andanti–phospho-MEK1/2 (Cell Signaling); anti-RhoA, anti–cyclinD1, cyclin A, and cyclin E (Calbiochem); anti–cyclin B1 (SantaCruz Biotechnology); anti-p21cip1 (Upstate Antibodies); andanti-p27kip1 (Calbiochem). Densitometry was carried out us-ing ImageJ software (NIH).Confocal microscopy. Cells were fixed with 10% trichlor-

oacetic acid (TCA) as described previously (23). Ras andRhoA were stained according to antibody manufacturer's in-structions and detected with tetramethylrhodamine isothio-cyanate (TRITC)–conjugated donkey anti-mouse (JacksonImmunoResearch). Nuclei were counterstained with 300nmol/L 4′,6-diamidino-2-phenylindole (DAPI; MolecularProbes). Slides were mounted using Vectashield (Vector Lab-oratories) and then viewed with LSM 510 META NLO confo-cal microscope (Zeiss).GTPase activation assays. Ras and RhoA activation was

determined by pull-downs using glutathione S-transferase(GST)–Raf–RBD and GST-Rhotekin, respectively, as de-scribed previously (21, 24).RNA extraction and reverse transcription-PCR. RNA was

extracted using STAT-60 according to the manufacturer's in-structions (Biogenesis). Reverse transcription-PCR (RT-PCR)was performed on 1 μg of total RNA using OneStep RT-PCR kit (Qiagen) as described previously (25). The followingprimers were used: USP17, 5′-CAGTGAATTCGTGGGAATG-GAGGACGACTCACTCTAC-3′ (forward) and 5′-AGTCATC-GATCTGGCACACAAGCATAGCCCTC-3′ (reverse); β-actin,5 ′ -GGACTTCGAGCAAGAGATGG-3 ′ ( forward) and5′-AGCACTGTGTTGGCGTACAG-3′ (reverse).

Results

USP17 is highly expressed in tumor cells. Previous re-ports have linked USP17 with proliferation and survival,where constitutive USP17 overexpression in NIH3T3, mouseembryonic fibroblast, and bone marrow–derived pro–B-celllines inhibited proliferation (20, 21). However, USP17 is alsoexpressed at a high level in various hematopoietic tumor-derived cells (20). To further examine these apparentlycontradictory findings, USP17 expression was assessed innumerous cell lines derived from various solid tumors,including prostate (PC3 and DU145), breast (MDA-MB-231,MDA-MB-157, BT474, and T47D), and lung (A549, H1299, andH460; Fig. 1A and B; data not shown). USP17 mRNA and pro-tein were clearly evident in all tumor-derived cell lines exam-ined. The highest USP17 expression was observed in the PC3prostate adenocarcinoma cell line, although high USP17 ex-pression was also evident in the breast cancer cell lines

Cancer Research



USP17 Mediates G1-S Cell Cycle Progression


derived from medullary (MDA-MB-231 and MDA-MB-157)and ductal (BT474 and T47D) carcinomas. To determinewhether the expression of USP17 in cell lines was representa-tive of USP17 levels in clinical samples, USP17 was alsoexamined in primary tumor tissue. Consistent with our obser-vations of the tumor-derived cell lines, examination of the in-tensity and distribution of USP17 revealed positive staining incolorectal adenocarcinoma and squamous cell carcinomas ofthe lung, cervix, and esophagus (Fig. 1C). Comparativelynegligible USP17 levels were observed in the matched nonma-lignant tissues (Fig. 1C). USP17 was largely distributedthroughout the cytosol of squamous carcinoma cells andwas not detected in the surrounding stromal tissue. In addi-tion, themoderately differentiated colon adenocarcinoma alsoexhibited strong USP17 staining of the nuclei. Collectively,

www.aacrjournals.org


these observations clearly confirmed that tumor cells ex-pressed elevated levels of USP17, suggesting that it mayplay a role in tumor progression. Clearly, the initial observa-tions suggested that USP17 was expressed in multiple tumorsand was particularly evident in rapidly dividing tumor-derivedcell lines. By contrast, however, it has previously been shownthat constitutive overexpression of USP17 blocked prolifera-tion (20). A possible explanation for these apparently contra-dictory observations could be that USP17 expression may bemodulated during cell division in a similar manner to corecomponents of the cell cycle machinery, including cyclinsand CDKI (26).USP17 expression is cell cycle regulated. To examine

USP17 expression during each cell cycle phase, HeLa cellswere synchronized in G1 by double thymidine block and

Figure 1. USP17 is highly expressed in tumor cells. A, USP17 mRNA expression assessed by RT-PCR in various tumor-derived cell lines (PC3,DU145, MDA-MB-231, MDA-MB-157, BT474, and T47D). B, USP17 protein expression was determined by immunoprecipitation (IP) followed byimmunoblotting (WB) for USP17. β-Actin was used as a loading control. C, immunohistochemical staining of USP17 in colon adenocarcinoma andsquamous cell carcinomas of the lung, cervix, and esophagus in comparison with corresponding normal tissue. The brown color represents USP17protein expression and the blue (hematoxylin) counterstain shows nuclei. Magnifications, ×20 and ×40 (as specified).

Cancer Res; 70(8) April 15, 2010 3331



McFarlane et al.

3332


the synchronous cell cycle progression was verified by PIstaining (Fig. 2A, left). Following release from the double thy-midine block, >90% of the cells had consistently entered Sphase within 4 hours, successfully undergone mitosis by 10hours, and completed the cell cycle within 12 hours. Consis-



tent with the fluorescence-activated cell sorting (FACS) anal-ysis, the elevated expression of cyclins E and A within 4 hoursand increased cyclin B1 levels after 6 hours verified the syn-chronicity of cells in G1-S and G2-M, respectively (Fig. 2,right). USP17 expression was examined at each cell cycle

Figure 2. USP17 expression is cell cycle regulated. A, left, HeLa cells were synchronized in G1 by double thymidine block and cell cycle was assessed byPI staining and FACS analysis; right, cyclins E, A, and B1 were immunoblotted at indicated times following thymidine release. B, USP17 mRNA expressionwas assessed by RT-PCR (left) and USP17 protein was determined by immunoprecipitation followed by immunoblotting at indicated times followingthymidine release (right). C, USP17 was stained using a TRITC-conjugated antibody (red staining and white arrows) and the nuclei were counterstained withDAPI (blue) at indicated times following thymidine release. D, MCF-7 cells were synchronized by double thymidine block and cell cycle was assessedby PI staining and FACS analysis (left) and USP17 mRNA expression was determined by RT-PCR (right) at indicated times following thymidine release.

Cancer Research





phase and found to be exquisitely regulated (Fig. 2B). USP17mRNA was not expressed in G1 (0 hour) but was abundant2 hours after thymidine release as the cells entered the DNAsynthesis phase. However, we consistently found negligibleUSP17 mRNA following 4-hour release from thymidine when>90% of the cells were in S phase (Fig. 2B). USP17 mRNA wasalso expressed on entry into G2-M and mitosis. Immunopre-cipitation and Western blotting confirmed that USP17 pro-tein levels reflected the tightly regulated expression ofUSP17 mRNA levels during cell cycle progression (Fig. 2B,right). This trend was also verified by confocal microscopy.Negligible levels of USP17 were consistently observed in G1

(Fig. 2C, left) and S phase (Fig. 2C, right), but USP17 washighly expressed during the G1-S transition (Fig. 2C, middle).To ensure that these findings were not cell type specific,MCF-7 cells were similarly examined (Fig. 2D, left). Consis-tent with our observation in HeLa cells, USP17 mRNA sub-stantially increased within 2 hours of thymidine releaseand was undetectable by 4 hours when cells had entered Sphase (Fig. 2D, right). Collectively, these findings clearly



showed that USP17 expression was tightly regulated in a cellcycle–specific manner, and this may account for the elevatedUSP17 levels observed in rapidly dividing tumor cells (Fig. 1).Moreover, these findings indicated that cell cycle phase–specific expression of USP17 could underpin an importantrole in cell cycle control.USP17 silencing inhibits proliferation and impairs G1-S

transition. Cell cycle progression is driven by the oscillatingactivation of CDKs alongside precisely timed fluctuations inthe synthesis and degradation of cyclins (26, 27). As USP17displayed a similar cyclical expression pattern during cell di-vision, we hypothesized that it also played a part in regulat-ing cell cycle progression. To understand the role of USP17during cell division and proliferation, we validated short hair-pin RNA (shRNA) to silence endogenous USP17. Two inde-pendent USP17 shRNAs effectively depleted endogenousUSP17 in a range of tumor-derived cell lines (Fig. 3A; datanot shown). We next examined the effect of transient USP17depletion on cell proliferation by MTT (Fig. 3B) and try-pan blue exclusion assays (Fig. 3C). Silencing of USP17 had a

Figure 3. USP17 silencing blockscell proliferation. A, HeLa cellswere transfected with scrambledshRNA (SCR shRNA) or USP17shRNA#1 or shRNA#2. USP17mRNA was determined byRT-PCR. B, HeLa cells weretransfected with scrambledshRNA or USP17 shRNA andproliferation was assessed by MTTassay over 96 h. C, proliferationof scrambled shRNA–transfectedor USP17 shRNA–transfectedcells was assessed by trypanblue exclusion assay over 96 h.D, cell cycle profiles of scrambledshRNA–transfected or USP17shRNA–transfected cells 72 hfollowing transfection.

Cancer Res; 70(8) April 15, 2010 3333



McFarlane et al.

3334


profound effect on cell growth in comparison with scrambledshRNA–transfected or mock-transfected cells. USP17 deple-tion blocked proliferation for up to 72 hours, after which pro-liferation gradually increased, probably due to transienttransfection (Fig. 3B). Similar effects where determined withUSP17 shRNA#2 (Supplementary Fig. S1A). The effect ofUSP17 depletion on cell growth was not due to increased celldeath, as a sub-G0 peak was not detected in USP17 shRNAcells. In addition, we did not observe any profound S orG2-M cell cycle block in USP17-silenced cells. However, an in-creased number of USP17-depleted cells were consistently de-tected in G1 (62.6%), in comparison with scrambled shRNAcells (53.1%), which suggested that USP17 silencing may havecaused cells to accumulate in G1 (Fig. 3D). These findings in-dicated that USP17 depletion had a significant cytostatic ef-fect on tumor-derived cell lines, as USP17 shRNA markedlyinhibited cell proliferation but had a negligible effect on celldeath. In addition, these observations also indicated thatUSP17 had a role in cell cycle regulation, as USP17 silencingcaused cells to accumulate in G1.To further examine this effect, shRNA-transfected cells

were synchronized in G1 and the effect on the cell cycle



was monitored following thymidine release. We consistentlyfound that USP17 silencing significantly impaired the transi-tion from G1-S (Fig. 4A). Following release from the thymi-dine block, the vast majority (84%) of scrambled shRNAcells were synchronized in S within 4 hours. In contrast, only50% of USP17-depleted cells had progressed into S and theremaining 47% had failed to exit G1 (Fig. 4A). The proportionof cells entering S phase were also examined by monitoringbromodeoxyuridine (BrdUrd) incorporation in synchronizedcells following thymidine release, and the results of four inde-pendent experiments are summarized in Fig. 4B. Consistentwith our previous observations (Fig. 2A), the vast majority(50%) of scrambled shRNA cells had incorporated BrdUrd fol-lowing 4 hours of thymidine release, whereas USP17 deple-tion significantly delayed the entry into S phase with only20% of the USP17-silenced cells incorporating BrdUrd overthe same time period (Fig. 4B). Impaired G1-S–phase progres-sion was also observed using USP17 shRNA#2 (SupplementaryFig. S1B). These observations clearly showed that USP17 silenc-ing abrogated the passage of cells through G1 into S phase. Ourprevious observations confirmed negligible USP17 expressionwhen cells exclusively resided in G1 (0 hour) or S (4 hours);

h. 4, 2020. © 2010 Ame

Figure 4. USP17 silencing inhibits G1-Stransition. A, scrambled or USP17 shRNAHeLa cells were synchronized in G1 bydouble thymidine block. Cell cycle wasassessed by PI staining and FACS analysis.B, scrambled or USP17 shRNA HeLa cellssynchronized in G1 were pulsed labeledfor 30 min with 33 μmol/L BrdUrd andS-phase entry was monitored by BrdUrdincorporation following thymidine releaseusing a FITC-conjugated BrdUrd antibody.

Cancer Research

rican Association for Cancer




however, USP17 was highly expressed during the G1-S transi-tion (2 hours; Fig. 2C). Therefore, collectively, these results sug-gest that it is necessary to induce USP17 expression to enablecells to exit G1 and enter S phase.shRNA knockdown of USP17 impairs GTPase activation

during the cell cycle. We have previously documented thatUSP17 regulates the activity and intracellular localization ofthe GTPase Ras (21). Ras and Rho family GTPases play a ma-jor role in promoting G1-S progression through their modu-lation of cyclin and CDKI levels (28, 29). Because USP17expression was also important for G1-S transition, we inves-tigated whether USP17 silencing was causing G1 arrest by at-tenuating GTPase function. Ras and RhoA localization inscrambled or USP17 shRNA cells was examined by confocalmicroscopy in G1 (Fig. 5). Before thymidine release, RhoA(Fig. 5A) and Ras (Fig. 5B) were distributed diffusely through-out the cytoplasm in both the scrambled shRNA and theUSP17 shRNA cells. Consistent with previous studies thathave reported a redistribution of GTPases to the plasmamembranes during G1, a rapid translocation of Ras and RhoAto the plasma membrane was observed in scrambled shRNAcells 5 minutes following thymidine release (Fig. 5A and B,arrows; ref. 30). The distribution of GTPases at the plasmamembrane dissipated after 15 minutes, and within 30 min-utes of release, Ras and RhoA resided predominantly in thecytoplasm. A corresponding translocation was not observedin the USP17-depleted cells; Ras and RhoA failed to relocalizeto the plasma membrane following the thymidine release andwere dispersed evenly throughout the cytoplasm. Membranelocalization of GTPases is frequently used as a surrogatemarker of activation; therefore, the lack of RhoA and Rasmembrane association suggested attenuated GTPase activa-tion in USP17-silenced cells. RhoA and Ras activity in G1 werethen directly examined using GST-Rhotekin and GST-Rafpull-down assays, respectively (Fig. 5C and D). As expected,a significant induction of RhoA (Fig. 5C) and Ras (Fig. 5D)activation was observed in the scrambled shRNA cells imme-diately (5 minutes) following thymidine release. Consistentwith the localization data, RhoA and Ras activation persistedfor 15 minutes following thymidine release and diminishedwithin 30 minutes. In comparison, USP17-silenenced cellswere unable to regulate the activation of RhoA and Ras.Con-stitutive activation of RhoA and Ras was consistently observedin USP17-depleted cells, and yet, these GTPases failed to trans-locate to the plasma membrane (Fig. 5C and D). In agreementwith our previous reports, these data show that USP17 is es-sential for coordinating the localization and activation ofGTPases. Compellingly, these findings showed that USP17 si-lencing abrogates the rapid activation and plasma membranelocalization of RhoA and Ras, following thymidine release,which are essential for driving G1-S progression and are likelyto be the underlying cause of impaired cell cycle progression.Depletion of USP17 stabilizes the levels of the CDKIs

p21cip1 and p27kip1. It is well established that GTPase acti-vation controls G1-S progression by regulating transcription,translation, and degradation of numerous proteins (28, 29).Our observations suggested that USP17 regulated GTPase lo-calization and activation in a cell cycle–dependent manner



(Fig. 5). Therefore, the effects of USP17 silencing were as-sessed by examining the effect on downstream effectors ofGTPases in G1. Ras and RhoA activation stimulates the sus-tained ERK activation, which is necessary to drive cyclin D1transcription in G1 (31, 32). We observed a corresponding ac-tivation of MEK and ERK in scrambled shRNA cells 1 hourfollowing release from thymidine block, and this activationwas sustained for 4 hours to drive entry into S. However,both MEK and ERK activation were markedly blunted inUSP17-depleted cells (Fig. 6, top). This further supports ourprevious findings that USP17 depletion is associated withmislocalization and improper activation of GTPases, and thatthis in turn inhibits the downstream activation of MEK andERK. The other major targets of GTPase signaling in G1 arethe CDKIs p21cip1 and p27kip1, which are downregulated inresponse to GTPase activation (31, 33–35). CDKIs functionas negative regulators of cell cycle progression by modulatingthe activities of cyclin-CDK complexes. RhoA stimulatesthe degradation of p27kip1 through the activation of cyclinE/CDK2, whereas Ras promotes depletion of p27kip1 throughthe ERK/MAPK pathway (36, 37). The levels of p21cip1 arenegatively affected by Ras and RhoA, which culminates inthe repression of p21cip1 transcription and increased proteinturnover (31, 33, 35). These CDKIs have a potent inhibitoryeffect on the CDKs CDK2 and CDK4, which drive G1-S cellcycle progression. The levels of both p21cip1 and p27kip1 aredownregulated in scrambled shRNA cells immediately fol-lowing thymidine release to relieve the inhibition on the cy-clin/CDK complexes and thus facilitate G1-S passage (Fig. 6).However, constitutively elevated levels of p21cip1 and p27kip1

are consistently observed in USP17-depleted cells (Fig. 6). Itis widely accepted that elevated levels of p21cip1 and p27kip1

induce G1 arrest; therefore, it is highly likely that the im-paired G1-S transition noted in USP17-depleted cells occursas a result of deregulation of p21cip1 and p27kip1 levels (31,38). Collectively, these findings show that USP17 silencing in-hibits signaling pathways that are integral to G1-S progres-sion, resulting in constitutively high levels of p21cip1 andp27kip1 that block cell cycle progression.

Discussion

It is becoming increasingly apparent that cell cycle pro-gression and proliferation are regulated by the dynamic bal-ance between ubiquitination and deubiquitination (1, 2, 4).Much attention has been paid to ubiquitin-mediated cell cy-cle control and the role of DUBs (1, 4). In this study, we revealthat USP17 expression is tightly regulated during cell cycleprogression and that temporal expression of this DUB is es-sential for cell division. We show that induction of USP17 ex-pression in G1 is necessary for progression into S phase andthat the G1 block observed in USP17-silenced cells coincideswith deregulation of Ras and RhoA relocalization and activa-tion. Moreover, the impaired activation of GTPases in USP17-depleted cells is associated with constitutively elevated levelsof p21cip1 and p27kip1, causing cells to accumulate in G1.The exquisite cell cycle regulation of USP17 seems to ex-

plain the apparent paradox of high expression in tumors and

Cancer Res; 70(8) April 15, 2010 3335



McFarlane et al.

3336


tumor-derived cell lines with the fact that constitutive over-expression blocks proliferation (20). Cell cycle progression ischaracterized by oscillating levels of numerous proteins in-volved in regulating cell cycle progression, such as cyclins,



CDKIs, Plk1, securin, and Aurora A/B; as their levels are in-timately regulated by ubiquitination, we wondered whetherUSP17 played a role in cell cycle control (2, 39–41). Consis-tent with this hypothesis, USP17 transcription and protein

h. 4, 2020. © 2010 American As

Figure 5. USP17 alters thelocalization and activation ofRas and Rho in G1. A and B,synchronous scrambled or USP17shRNA HeLa cells were fixedusing 10% TCA at specified timesfollowing thymidine release andstained for endogenous RhoA andRas (red), respectively. The nuclei(blue) were counterstained withDAPI. C and D, RhoA and Raspull-downs were performed usingGST-tagged Rhotekin-RBD andGST-tagged Raf-RBD fusionproteins, respectively. Pull-downsand whole-cell lysates wereimmunoblotted with either RhoAor pan-Ras antibodies. Proteinloaded was confirmed byimmunoblotting for γ-tubulin.Right, densitometry of Ras andRhoA activity.

Cancer Research

sociation for Cancer




expression fluctuated significantly between each phase of thecell cycle, similarly to other cell cycle regulators. The tightlyregulated expression of USP17 in a cell cycle–dependentmanner is similar to other DUBs that have previously beenimplicated in cell cycle control. USP1 expression is inducedin S phase and enables the deubiquitination of two importantcomponents of DNA damage repair pathways, FANCD2 andPCNA (8, 9), whereas DUBs with recognized roles in mitoticspindle assembly checkpoints, such as CYLD, USP9, andUSP44, are also temporally expressed during cell divisionand levels are found to peak during mitosis (5–8).Although the DUB/USP17 family has previously been im-

plicated in proliferation, the mechanism underpinningUSP17 control of cell cycle progression was unclear. Recentstudies identified RCE1 as a substrate for USP17 and con-firmed that RCE1 activity was regulated by deubiquitination,and more significantly, the ability of USP17 to block prolifer-ation was completely dependent on RCE1 (21). In addition toRas, RCE1 is also responsible for the posttranslational mod-ification of numerous proteins that contain a CAAX motif,including the Rho family of GTPases. The role of Ras andRho family GTPases in G1 progression has been comprehen-sively investigated and shown to modulate the levels of cy-clins A, E, and D1 and the CDKIs p21cip1 and p27kip1 bydistinct mechanisms (28, 29, 33, 34). Signals from Ras andRhoA interact to regulate the levels of p21cip1 and p27kip1



through the modulation of their transcription, translation,and degradation (34, 42). Moderate levels of Ras andRaf-MEK-ERK signaling increase transcription and stabiliza-tion p21cip1 protein levels to promote cell cycle progressionthrough its interaction with cyclin D1 (31, 35, 43). Cross talkbetween Ras and RhoA leads to downregulation of p21cip1

transcription, which is necessary for G1-S progression (24, 31).Ras and RhoA activation also promotes cell cycle progressionby downregulating p27kip1 levels through enhanced degrada-tion and decreased protein synthesis (34, 42). Our findingssuggest that the improper localization and activation ofRas and RhoA in the USP17 silenced cells could be responsi-ble for the constitutively high levels of the CDKIs p21cip1 andp27kip1 observed.The subcellular distribution of GTPases determines their

interaction with corresponding guanine nucleotide exchangefactors (GEF) and GTPase-activating proteins (GAP) and iscritical in determining appropriate GTPase activation. There-fore, it was not surprising that improper GTPase activationwas also observed in USP17-depleted cells, which had consti-tutively high levels of GTP-bound RhoA and Ras. Further-more, recent advances in the field have identified pools ofactive Ras that signal from intracellular compartments, in-cluding the Golgi and endoplasmic reticulum (44). Indeed, ithas been reported that the effectors downstream of ERK1/2activation are dependent on the subcellular compartment

Figure 6. Depletion of USP17 stabilizes the levels of the CDKIs p21cip1 and p27kip1. HeLa cells transfected with scrambled or USP17 shRNA weresynchronized by double thymidine block and lysates were collected at specified times following thymidine release. Cell lysates were immunoblottedwith antibodies for phospho-ERK (pERK), ERK, phospho-MEK (pMEK), MEK, p21cip1, and p27kip1. Protein loaded was confirmed by immunoblotting forγ-tubulin. Right, densitometry of MEK and ERK activity and the relative levels of p21cip and p27kip standardized to γ-tubulin.

Cancer Res; 70(8) April 15, 2010 3337



McFarlane et al.

3338


from which the Ras signal originates (45). Therefore, it is plau-sible that constitutively active Ras is signaling from an intra-cellular compartment in USP17 depleted cells and driving analternative signaling cascade that impairs cell cycle progres-sion. In support of our hypothesis that improper localizationand activation of GTPases causes USP17-depleted cells to ac-cumulate in G1, it has recently been reported that constitutiveactivation of RhoA delays the entry of cells into S phase. Thiscell cycle defect was thought to have occurred as a resultof suppressed ERK activation that was necessary to drivecyclin D1 expression (46). Consistent with these findings, wealso observe attenuated ERK activation in association withconstitutive GTPase activation in USP17-silenced cells. There-fore, we propose that USP17 must be required to facilitateproper localization and activation of GTPases in a cell cycle–dependent manner.To date, overwhelming evidence has attributed the con-

trolled timing of GTPase activation during the cell cycle toa range of complementary GEF (such as Sos1/2, Dbl, Vav,and ECT2) and GAPs (such as p120 RasGAP and Cyk4). How-ever, it has become increasingly apparent that GTPase pro-cessing enzymes are also implicitly involved in determiningGTPase activation and consequent function. Indeed, en-zymes involved in posttranslational modification of GTPaseshave been widely investigated as novel therapeutic targets toinhibit Ras and RhoA signaling and block tumor growth andprogression. Farnesyltransferase (FT) and geranylgeranyl-transferase I (GGT) are responsible for the prenylation ofGTPases before CAAX-box cleavage by RCE1. Inhibitors ofFT and GGT (FTI and GGTI) have been developed that blockprenylation and, therefore, GTPase activation. Attenuation ofGTPase processing using these inhibitors causes cells to



arrest in G1 (47, 48). In agreement with our observations inUSP17-depleted cells, this was thought to occur as a result ofimproper GTPase activation that leads to elevated p21cip1

and p27kip1 levels (49, 50).In summary, our findings clearly show that, like other cell

cycle–regulated proteins, the expression of USP17 is tightlycontrolled during cell cycle progression. Additionally, weshow that this rapid and transient expression of USP17 isnecessary for successful cell cycle progression, as silencingof USP17 expression impairs G1-S transition and blocks pro-liferation. This would seem to occur through the deregula-tion of GTPase signaling and stabilization of CDKIs p21cip1

and p27kip1. These findings, coupled with the elevated levelsof USP17 in primary tumor biopsies, suggest that this DUBmay be an interesting target for cancer therapy.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Dr. Rene Medema for constructs and Drs. Shauna Hegarty,Suzanne McFarlane, and Perry Maxwell for their assistance.

Grant Support

Action Cancer and Biotechnology and Biological Sciences Research Council(BB/FO134647).

The costs of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 11/11/2009; revised 01/28/2010; accepted 02/15/2010; publishedOnlineFirst 04/06/2010.

References
1. Wickliffe K, Williamson A, Jin L, Rape M. The multiple layers of ubi-
quitin-dependent cell cycle control. Chem Rev 2009;109:1537–48.2. Nakayama KI, Nakayama K. Ubiquitin ligases: cell-cycle control and

cancer. Nat Rev Cancer 2006;6:369–81.3. Hussain S, Zhang Y, Galardy PJ. DUBs and cancer: the role of deu-

biquitinating enzymes as oncogenes, non-oncogenes and tumorsuppressors. Cell Cycle 2009;8:1688–97.

4. Song L, Rape M. Reverse the curse—the role of deubiquitination incell cycle control. Curr Opin Cell Biol 2008;20:156–63.

5. Stegmeier F, Rape M, Draviam VM, et al. Anaphase initiation is reg-ulated by antagonistic ubiquitination and deubiquitination activities.Nature 2007;446:876–81.

6. Stegmeier F, Sowa ME, Nalepa G, Gygi SP, Harper JW, Elledge SJ.The tumor suppressor CYLD regulates entry into mitosis. Proc NatlAcad Sci U S A 2007;104:8869–74.

7. Vong QP, Cao K, Li HY, Iglesias PA, Zheng Y. Chromosome align-ment and segregation regulated by ubiquitination of survivin. Science2005;310:1499–504.

8. Nijman SM, Huang TT, Dirac AM, et al. The deubiquitinating enzymeUSP1 regulates the Fanconi anemia pathway. Mol Cell 2005;17:331–9.

9. Huang TT, Nijman SM, Mirchandani KD, et al. Regulation of monoubi-quitinated PCNA by DUB autocleavage. Nat Cell Biol 2006;8:339–47.

10. Brooks CL, Li M, HuM, Shi Y, GuW. The p53-Mdm2-HAUSP complexis involved in p53 stabilization by HAUSP. Oncogene 2007;26:7262–6.

11. Joo HY, Zhai L, Yang C, et al. Regulation of cell cycle progressionand gene expression by H2A deubiquitination. Nature 2007;449:1068–72.

12. Nicassio F, Corrado N, Vissers JH, et al. Human USP3 is a chromatinmodifier required for S phase progression and genome stability. CurrBiol 2007;17:1972–7.

13. Saggar S, Chernoff KA, Lodha S, et al. CYLD mutations in familialskin appendage tumours. J Med Genet 2008;45:298–302.

14. Masuya D, Huang C, Liu D, et al. The HAUSP gene plays an im-portant role in non-small cell lung carcinogenesis through p53-dependent pathways. J Pathol 2006;208:724–32.

15. Zhu Y, Pless M, Inhorn R, Mathey-Prevot B, D'Andrea AD. The murineDUB-1 gene is specifically induced by the βc subunit of interleukin-3receptor. Mol Cell Biol 1996;16:4808–17.

16. Baek KH, Kim MS, Kim YS, Shin JM, Choi HK. DUB-1A, a novel deu-biquitinating enzyme subfamily member, is polyubiquitinated andcytokine-inducible in B-lymphocytes. J Biol Chem 2004;279:2368–76.

17. Migone TS, Humbert M, Rascle A, Sanden D, D'Andrea A, Johnston JA.The deubiquitinating enzyme DUB-2 prolongs cytokine-induced signaltransducers and activators of transcription activation and suppressesapoptosis following cytokine withdrawal. Blood 2001;98:1935–41.

18. Zhu Y, Lambert K, Corless C, et al. DUB-2 is a member of a novelfamily of cytokine-inducible deubiquitinating enzymes. J Biol Chem1997;272:51–7.

19. Zhu Y, Carroll M, Papa FR, Hochstrasser M, D'Andrea AD. DUB-1, adeubiquitinating enzyme with growth-suppressing activity. Proc NatlAcad Sci U S A 1996;93:3275–9.

20. Burrows JF, McGrattan MJ, Rascle A, Humbert M, Baek KH, JohnstonJA. DUB-3, a cytokine-inducible deubiquitinating enzyme that blocksproliferation. J Biol Chem 2004;279:13993–4000.

Cancer Research





21. Burrows JF, Kelvin AA, McFarlane C, et al. USP17 regulates ras ac-tivation and cell proliferation by blocking RCE1 activity. J Biol Chem2009;284:9587–95.

22. Whitfield ML, Sherlock G, Saldanha AJ, et al. Identification of genesperiodically expressed in the human cell cycle and their expression intumors. Mol Biol Cell 2002;13:1977–2000.

23. Yonemura S, Hirao-Minakuchi K, Nishimura Y. Rho localization incells and tissues. Exp Cell Res 2004;295:300–14.

24. Sahai E, Olson MF, Marshall CJ. Cross-talk between ras and rho sig-nalling pathways in transformation favours proliferation and in-creased motility. EMBO J 2001;20:755–66.

25. Burrows JF, McGrattan MJ, Johnston JA. The DUB/USP17 deubiqui-tinating enzymes, a multigene family within a tandemly repeated se-quence. Genomics 2005;85:524–9.

26. Sherr CJ. The Pezcoller lecture: cancer cell cycles revisited. CancerRes 2000;60:3689–95.

27. Satyanarayana A, Kaldis P. Mammalian cell-cycle regulation: severalcdks, numerous cyclins and diverse compensatory mechanisms.Oncogene 2009;28:2925–39.

28. Coleman ML, Marshall CJ, Olson MF. RAS and RHO GTPases inG1-phase cell-cycle regulation. Nat Rev Mol Cell Biol 2004;5:355–66.

29. Villalonga P, Ridley AJ. Rho GTPases and cell cycle control. GrowthFactors 2006;24:159–64.

30. Noguchi Y, Nakamura S, Yasuda T, et al. Newly synthesized rho A,not ras, is isoprenylated and translocated to membranes coincidentwith progression of the G1 to S phase of growth-stimulated ratFRTL-5 cells. J Biol Chem 1998;273:3649–53.

31. Olson MF, Paterson HF, Marshall CJ. Signals from ras and rhoGTPases interact to regulate expression of p21Waf1/Cip1. Nature1998;394:295–9.

32. Swant JD, Rendon BE, Symons M, Mitchell RA. Rho GTPase-dependent signaling is required for macrophage migration inhibitoryfactor-mediated expression of cyclin D1. J Biol Chem 2005;280:23066–72.

33. Coleman ML, Densham RM, Croft DR, Olson MF. Stability ofp21Waf1/Cip1 CDK inhibitor protein is responsive to RhoA-mediatedregulation of the actin cytoskeleton. Oncogene 2006;25:2708–16.

34. Croft DR, Olson MF. The rho GTPase effector ROCK regulates cyclinA, cyclin D1, and p27Kip1 levels by distinct mechanisms. Mol CellBiol 2006;26:4612–27.

35. Coleman ML, Marshall CJ, Olson MF. Ras promotes p21(Waf1/Cip1)protein stability via a cyclin D1-imposed block in proteasome-mediated degradation. EMBO J 2003;22:2036–46.

36. Hu W, Bellone CJ, Baldassare JJ. RhoA stimulates p27(kip) degrada-tion through its regulation of cyclin E/CDK2 activity. J Biol Chem1999;274:3396–401.

37. Delmas C, Manenti S, Boudjelal A, Peyssonnaux C, Eychene A,Darbon JM. The p42/p44 mitogen-activated protein kinase activation



triggers p27Kip1 degradation independently of CDK2/cyclin E in NIH3T3 cells. J Biol Chem 2001;276:34958–65.

38. Woods D, Parry D, Cherwinski H, Bosch E, Lees E, McMahon M.Raf-induced proliferation or cell cycle arrest is determined by thelevel of raf activity with arrest mediated by p21Cip1. Mol Cell Biol1997;17:5598–611.

39. van Leuken R, Clijsters L, van Zon W, et al. Polo-like kinase-1 con-trols aurora A destruction by activating APC/C-Cdh1. PLoS One2009;4:e5282.

40. Floyd S, Pines J, Lindon C. APC/C Cdh1 targets aurora kinase tocontrol reorganization of the mitotic spindle at anaphase. Curr Biol2008;18:1649–58.

41. Marangos P, Carroll J. Securin regulates entry into M-phase by mod-ulating the stability of cyclin B. Nat Cell Biol 2008;10:445–51.

42. Mammoto A, Huang S, Moore K, Oh P, Ingber DE. Role of RhoA,mDia, and ROCK in cell shape-dependent control of the Skp2-p27kip1 pathway and the G1/S transition. J Biol Chem 2004;279:26323–30.

43. Bottazzi ME, Zhu X, Bohmer RM, Assoian RK. Regulation of p21(cip1) expression by growth factors and the extracellular matrix re-veals a role for transient ERK activity in G1 phase. J Cell Biol 1999;146:1255–64.

44. Rocks O, Peyker A, Kahms M, et al. An acylation cycle regulates lo-calization and activity of palmitoylated ras isoforms. Science 2005;307:1746–52.

45. Casar B, Arozarena I, Sanz-Moreno V, et al. Ras subcellular localiza-tion defines extracellular signal-regulated kinase 1 and 2 substratespecificity through distinct utilization of scaffold proteins. Mol CellBiol 2009;29:1338–53.

46. Morin P, Flors C, Olson MF. Constitutively active RhoA inhibits pro-liferation by retarding G(1) to S phase cell cycle progression and im-pairing cytokinesis. Eur J Cell Biol 2009;88:495–507.

47. Adnane J, Bizouarn FA, Qian Y, Hamilton AD, Sebti SM. p21(WAF1/CIP1) is upregulated by the geranylgeranyltransferase I inhibitor GGTI-298 through a transforming growth factor β- and Sp1-responsiveelement: involvement of the small GTPase rhoA. Mol Cell Biol 1998;18:6962–70.

48. Sepp-Lorenzino L, Rosen N. A farnesyl-protein transferase inhibitorinduces p21 expression and G1 block in p53 wild type tumor cells.J Biol Chem 1998;273:20243–51.

49. Vogt A, Sun J, Qian Y, Hamilton AD, Sebti SM. The geranylgeranyl-transferase-I inhibitor GGTI-298 arrests human tumor cells in G0/G1

and induces p21(WAF1/CIP1/SDI1) in a p53-independent manner.J Biol Chem 1997;272:27224–9.

50. Kazi A, Carie A, Blaskovich MA, et al. Blockade of protein geranyl-geranylation inhibits Cdk2-dependent p27Kip1 phosphorylation onThr187 and accumulates p27Kip1 in the nucleus: implications forbreast cancer therapy. Mol Cell Biol 2009;29:2254–63.

Cancer Res; 70(8) April 15, 2010 3339



2010;70:3329-3339. Published OnlineFirst April 13, 2010.Cancer Res Cheryl McFarlane, Alyson A. Kelvin, Michelle de la Vega, et al. -S Progression1

Tumor Biopsies, Is Cell Cycle Regulated, and Is Required for GThe Deubiquitinating Enzyme USP17 Is Highly Expressed in

Updated version

10.1158/0008-5472.CAN-09-4152doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/04/05/0008-5472.CAN-09-4152.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/8/3329.full#ref-list-1

This article cites 50 articles, 28 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/70/8/3329.full#related-urls

This article has been cited by 11 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/70/8/3329To request permission to re-use all or part of this article, use this link

Research. on March 14, 2020. © 2010 American Association for Cancercancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-09-4152

http://cancerres.aacrjournals.org/content/suppl/2010/04/05/0008-5472.CAN-09-4152.DC1

http://cancerres.aacrjournals.org/content/70/8/3329.full#ref-list-1

http://cancerres.aacrjournals.org/content/70/8/3329.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/70/8/3329


Documents

Tumor and Stem Cell Biology Cancer Research The ... · Tumor and Stem Cell Biology The Deubiquitinating Enzyme USP17 Is Highly Expressed in ... F. Burrows1,2, and James A. Johnston1