Revised version received23 April 2009Accepted 23 April 2009Available online 5 May 2009
Previously, we found that in t(X;1)(p11;q21)-positive renal cell carcinomas the bHLH-LZtranscription factor TFE3 is fused to a novel protein designated PRCC. In addition, we found that
the PRCCTFE3 fusion protein, which has retained all known functional domains of TFE3, acts as amore potent transcriptional activator than wild type TFE3. We also found that PRCCTFE3expression confers in vitro and in vivo transformation onto various cell types, including those ofthe kidney. Here we show that de novo expression of the PRCCTFE3 fusion protein provokes cellcycle delay. This delay, which is mediated by induction of the cyclin-dependent kinase inhibitorp21
WAF1/CIP1, affects both the G1/S and the G2/M phases of the cell cycle and prevents the cells from
undergoing polyploidization. We also show that the PRCCTFE3 fusion protein binds directly to thep21
WAF1/CIP1promoter and that the PRCCTFE3-induced up-regulation of p21
WAF1/CIP1leads to activation
of the pRB pathway. Finally, we show that in t(X;1)(p11;q21)-positive renal tumor cells severalprocesses that link PRCCTFE3 expression to p21
WAF1/CIP1-mediated cell cycle delay are abrogated.
Our data suggest a scenario in which, during the course of renal cell carcinoma development, an
initial PRCCTFE3-induced cell cycle delay must be numbed, thus permitting continuedproliferation and progression towards full-blown malignancy.
Oncogenic stimulation can induce cell cycle delay and senescenceto ensure that damaged cells stop to proliferate [1–5]. Recently, itwas shown that such a delay may play a critical role in the in vivoprotection from tumor development [6–10], and may be mediatedby cyclin-dependent kinase inhibitors such as p16INK4a andp21
WAF1/CIP1[11,12]. Subsequent elimination of this protective
mechanism by suppression of the p53 and/or p16INK4a pathwayshas been shown to permit continued proliferation of primary cells in
Genetics 855, Radboud Uni
cn.nl (A. Geurts van Kessel
r Inc. All rights reserved.
the presence of the oncogenic event, thereby leading to tumori-genicity [13]. In full accordancewith these findings, Voorhoeve et al.[14] identified twomicroRNAswhose expression could substitute forthe loss ofwild typep53 and, by doing so, could overcomeoncogenicH-RASV12-mediated cell cycle delay in primary fibroblasts.
A well-defined subset of human renal cell carcinomas ischaracterized by the presence of recurrent chromosome transloca-tions, each affecting members of the MiT family of transcriptionfactors [15–17]. Previously, we and others found that one of thesetranslocations, i.e., t(X;1)(p11;q21), leads to an in frame fusion of
versity Nijmegen Medical Centre, PO Box 9101, 6500 HB Nijmegen, The
2400 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
MiT protein TFE3 to a novel protein, designated PRCC [18,19].TFE3 contains several archetypal transcription factor-associatedfunctional domains, among which are two transactivationdomains, a basic domain, a helix–loop–helix domain, and aleucine-zipper domain [20]. PRCC neither contains domainsindicative for its function nor exhibits homology to any knownprotein [18,19]. It has been shown, however, that PRCC canassociate with pre-mRNA splicing factors such as SC35, PRL1 andCDC5 and that, after fusion with TFE3, this association isweakened [21]. By employing transactivation assays, we andothers found that the PRCCTFE3 fusion protein, which hasretained all the functional domains of TFE3, acts as a significantlymore potent transcriptional activator than wild type TFE3 [21,22].Moreover, we found that PRCCTFE3 is able to confer malignanttransformation onto various cell types, including those of thekidney, in both in vitro and in vivo assays [23]. In addition, it hasrecently been shown that the MET tyrosine kinase receptor genecan act as a downstream transcriptional target of several TFE3fusion proteins, including PRCCTFE3 [24–26] [own unpublisheddata], thereby providing a functional rationale for the oncogenicnature of these fusion proteins.
Here, we report that de novo expression of the oncogenicPRCCTFE3 fusion protein provokes cell cycle delay. This delay ismediated through induction of the cyclin-dependent kinaseinhibitor p21
WAF1/CIP1and a concomitant activation of the pRB
pathway. In addition, we report that in t(X;1)(p11;q21)-positiverenal tumor cells this PRCCTFE3-induced cell cycle delay isnumbed, thus allowing the development of a full-blownmalignancy.
Materials and methods
Cell culture
Hela, U2OS, Rat-1 and Phoenix cells were cultured in DMEM,supplemented with 10% FCS, penicillin (100 U/ml) and strepto-mycin (100 μg/ml). The renal cell carcinoma-derived cell line Cl89-12117 [27] was cultured in RPMI1640, supplemented with 10% FCS,penicillin (100 U/ml) and streptomycin (100 μg/ml). HEK293/T-REx/PRCCTFE3 [27] and similarly obtained HEK293/T-REx/PRCCcells were cultured in DMEM, supplemented with 10% FCS, zeocin(500 μg/ml), blasticidin (5 μg/ml), penicillin (100 U/ml) andstreptomycin (100 μg/ml). For induction of PRCCTFE3 and/or PRCCexpression, tetracyclin was added to the culture medium (24 h;1 μg/ml, unless stated otherwise). HEK293/PRCCTFE3-TY1 wascultured in DMEM, supplemented with 10% FCS, penicillin(100 U/ml), streptomycin (100 μg/ml), G418 (Gibco; 0.5 μg/ml)and tetracycline (1 μg/ml). For induction of TY1-tagged PRCCTFE3,tetracyclin was washed away and cells were grown on completemedium without tetracycline. For an arrest at the metaphase/anaphase transition of the cell cycle, cells were treated withnocodazole (Sigma Aldrich; 16 h, 250 ng/ml). Cells weresubsequently released from the nocodazole-induced arrest inpre-warmed complete medium after two washes in pre-warmedPBS. To inhibit translation of proteins, cells were treated withcycloheximide (CHX; Sigma Aldrich, 16 h, 10 μg/ml). For theestablishment of growth curves cells were trypsinized, resus-pended in complete medium and counted using a Bürker-Türkchamber (W. Schreck). Counts from duplicate cultures were
averaged for each time point and plotted on a log scale. Cellpopulation doubling times were calculated from the linear phaseof the growth curve. For retroviral infection, supernatants wereproduced through transfection of Phoenix packaging cells with ashRNA construct, pRS-p21 [28], using Lipofectamine 2000 (Invi-trogen). 48 h post transfection, the tissue culture medium wasfiltered through a 0.45 μm millipore filter and the resulting viralsupernatant was used for a 6 h infection of HEK293/T-REx/PRCCTFE3 cells through the addition of 4 μg/ml polybrene (SigmaAldrich). Subsequently, the cells were allowed to recover for 24 hin fresh medium after which infected cells were selected using2 μg/ml puromycin (Sigma Aldrich).
Western blot analysis
For western blot analysis, proteins were extracted using a lysisbuffer (50 mM Tris–HCl pH 8.0, 150 mM NaCl, 1% NP40, 0.1% SDS,0.5% DOC). Subsequently, cell debris was removed by centrifuga-tion and the cleared samples were incubated for 5 min in samplebuffer before loading (20 μg per lane) onto 4–12% NuPage bis–trisgels (Invitrogen). After electrophoresis, the gels were blotted ontonitrocellulose membranes (Protran; Schleicher and Schuell). Theresulting blots were blocked in 5% non-fat dry milk/Tris bufferedsaline containing 0.05% Tween-20 (TBST), and incubated over-night with primary antibody and, subsequently, with a peroxidaseconjugated secondary antibody (Zymed) or with a fluorescentconjugated secondary antibody (Molecular Probes). Signals werevisualized using autoradiographic exposure to Kodak X-Omat filmsor scanned and analyzed using the fluorescence-scanning Odysseysystem and associated software (Li-Cor). For quantification of thesignals the ImageJ software tool was used (http://rsb.info.nih.gov/ij/). The area under the curve (AUC) of a specific signal wascorrected by the AUC of the loading control. The measurements ofthe untransfected or uninduced samples were arbitrarily set at 1(or 100%) and the other conditions were recalculated correspond-ingly, thereby allowing ratio comparisons.
Antibodies
A polyclonal anti-PRCC antibodywas used as described before [27].Anti-p21
Ran (BD Transduction Laboratories), anti-γ-tubulin (Santa Cruz),anti-CDK2 (Abcam) and anti-pRB (phospho S807 and S811;Abcam) antibodies were used according to the instructions ofthe manufacturers.
Cell cycle analysis using flow cytometry
Replicative DNA synthesis and DNA content were analyzed bybivariate flow cytometry using a fluorescence activated cell sorter(FACS) as described before [27]. For analysis of the progression ofcells through the S phase, and the subsequent G2/M and G0/G1phases of the cell cycle, cells were pulse labeled for 10 min at 37 °Cwith 10 μM BrdU and harvested at different time intervals bytrypsinization followed by overnight fixation in 70% ethanol at4 °C. After adding 10 ml PBS, cells were spun down and treatedwith 0.1 M HCl containing 0.4 mg/ml pepsin for 20 min at roomtemperature. Subsequently, cells were permeabilized (0.5%Tween-20, 0.1% BSA in PBS) and incubated for 12 min at 37 °C in2 M HCl, followed by the addition of 0.05 M borate buffer (pH 8.5).
2401E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
Cells were washed with 0.5% Tween-20, 0.1% BSA in PBS andincubated with anti-BrdU-FITC antibody (Becton Dickinson; 1 h,4 °C). Finally, cells were counterstained with 40 μg/ml propidiumiodide supplemented with 250 μg/ml RNase A and analyzed byFACS. Apoptosis assays were performed using an Annexin V-FITCapoptosis detection kit I (BD Biosciences), according to theinstructions of the manufacturer.
Chromatin immunoprecipitation analysis
Chromatin immunoprecipitations were performed essentially asdescribed before [29]. Protein–DNA complexes were immunopre-cipitated using an antibody directed against the TY1-tag (BB2; agift from H. Stunnenberg) and a mouse pre-immune serum. Therecovered genomic DNAs were analyzed by PCR using primer pairsspanning the p21
WAF1/CIP1promoter [30].
Results
PRCCTFE3 expression results in cell cycle delay
In order to assess the effect of de novo PRCCTFE3 expression on cellcycle progression, we performed BrdU pulse–chase experimentson both PRCCTFE3 expressing and non-expressing, i.e., tetracyclin-induced and non-induced, HEK293/T-REx/PRCCTFE3 cells (seeMaterials andmethods and Fig. S1). After pulse labeling, cells wereharvested at various time points, followed by cell cycle profileanalyses using FACS. Through such analyses, BrdU-positive cellscan be followed as they pass through the S, G2/M, and G0/G1phases of the cell cycle, respectively.
Immediately after pulse labeling (t=0) all BrdU-positive cellswere in the S phase, as expected (Fig. 1). Of the non-induced(−tet) cells 37.1% were in the S phase (BrdU-positive), 41.5% in theG0/G1 (2 N) phase and 16.4% in the G2/M (4 N) phase of the cellcycle (BrdU-negative). Similarly, of the induced, PRCCTFE3expressing, cells (+tet) 19.4% were in the S phase, 55.9% in theG0/G1 phase and 19.4% in the G2/M phase of the cell cycle,respectively. Notably, the percentage of BrdU-positive (S phase)cells was significantly lower (19.4% versus 37.1%) in the PRCCTFE3expressing cells than in the non-expressing cells (+tet and −tet,respectively). In reverse, the percentage of G0/G1 (BrdU-negative)cells was significantly higher (55.9% versus 41.5%) in the PRCCTFE3expressing cells than in the non-expressing cells. These resultssuggest that during pulse labeling less PRCCTFE3 expressing cellshad entered the S phase and, thus, that these cells were delayed inthe G1 to S transition of the cell cycle.
A less significant difference was observed for the respectiveinduced and non-induced G2/M phase cells (19.4% versus 16.4%).After a 4 h chase (Fig. 1), 69.5% of the non-induced BrdU-positivecells had entered the G2/M phase, whereas 15.8% of these cells hadentered the G0/G1 phase of the cell cycle. In the simultaneouslyinduced, PRCCTFE3 expressing, BrdU-positive cells 74.6% hadentered the G2/M phase, whereas only 6.8% of these cells hadentered the G0/G1 phase. After a 6 h chase, 37.5% of the non-induced BrdU-positive cells had progressed through the M phaseinto the G0/G1 phase. In contrast, only 14.6% of the induced,PRCCTFE3 expressing, BrdU-positive cells had entered the G0/G1phase, whereas 71.8% of these latter cells were still in the G2/Mphase of the cell cycle. After a 12 h chase 69.6% of the non-induced
BrdU-positive cells had re-entered the G0/G1 phase, whereas theinduced, PRCCTFE3 expressing, BrdU-positive cells were predomi-nantly present in the G2/M phase of the cell cycle (53.6%) and,thus, were delayed.
The cell cycle data also revealed a relatively slow passage of thePRCCTFE3 expressing cells through the S phase (Fig. 1; +tet). Fromthese results we conclude that exogenous PRCCTFE3 expression inHEK293 cells results in a delay of the G1/S, G2/M, and S phases ofthe cell cycle.
In order to independently confirm these data, we establishedgrowth curves of PRCCTFE3 expressing and non-expressingHEK293/T-REx cells (Fig. S2). Cells from duplicate cultures werecounted at days 0, 2, 5, 7, 9 and 12, and averaged for each timepoint. Cell population doubling times were calculated from thelinear phases of the growth curves. By doing so, we found thatPRCCTFE3 non-expressing cells (−tet) exhibited a doubling timeof ∼24 h. Based on this doubling time, subsequent FACS analysisindicated that the cells remained 14.5 h in the G0/G1 phase, 9 h inthe S phase and 30 min in the G2/M phase of the cell cycle (notshown). In contrast, we found that PRCCTFE3 expressing cells(+tet) only doubled once during the first 5 days and stoppedgrowing from that point on. Subsequent FACS analysis (day 1–5)indicated that the cells were almost evenly distributed over G1/G0and G2/M phases of the cell cycle, with few cells in S phase (notshown).
Our observation that PRCCTFE3 expression results in a G2/M phasedelay, prompted us to assess the effect of the microtubuledestabilizing drug nocodazole, which normally arrests cells atthe metaphase/anaphase transition of the cell cycle [31], on thisdelay. In order to do so, we again pulse labeled both non-inducedand induced HEK293/T-REx/PRCCTFE3 cells with BrdU, followedby a 16 h treatment of these cells with nocodazole and asubsequent release from this treatment during various timeintervals.
Of the non-induced cells 35.3% were in the S phase (BrdU-positive), 37.8% in the G0/G1 phase, and 21.5% in the G2/M phaseof the cell cycle (Fig. S3). After PRCCTFE3 induction 49.9% of thecells were in the G0/G1 phase, only 22.8% in the S phase, and 19.8%in the G2/M phase of the cell cycle (Fig. S3), which is in fullagreement with the above notion that PRCCTFE3 expressing cellsare delayed in the G1 to S phase transition. Subsequently, the BrdU-positive cells were chased for different time intervals after a 16 hnocodazole treatment and subjected to FACS analysis (Fig. 2).
Immediately after nocodazole treatment (0 h)most cells (78.4–81.2%) were arrested in the G2/M phase of the cell cycle, asexpected. After a 2 h release from nocodazole 25.1% of the non-induced cells had exited the M phase and re-entered the G0/G1phase, whereas of the PRCCTFE3 expressing cells only 14.6% had re-entered the G0/G1 phase of the cell cycle. After an 8 h release fromnocodazole 72.7% of the PRCCTFE3 expressing cells were stillpresent in the G2/M phase, whereas 29.2% of the non-inducedcells had re-entered the G0/G1 phase.
Notably, 16.5% of the BrdU-positive non-induced cells failed tore-enter the G0/G1 phase of the cell cycle but, instead, becamepolyploid. Subsequent microscopic analysis revealed that thesepolyploid cells are mononuclear in nature, thus indicating that
Fig. 1 – Cell cycle delay after exogenous PRCCTFE3 expression. HEK293/T-REx/PRCCTFE3 cells, incubated with or without tetracyclin(+ or −tet), were pulse labeled with BrdU. Subsequently, cell cycle profiles were obtained by FACS analysis. In the upper boxes dotplots representing BrdU fluorescence (y-axis) versus DNA content (x-axis) are shown. In the lower boxes DNA histograms of all cellsin the above dot plots are shown. At time point 0 h, cells in the G0/G1 (2 N), S and G2/M (4 N) phases of the cell cycle, and polyploidcells (>4 N) are marked by boxes 1, 2, 3 and 4, respectively. The percentages of cells in each of these cell cycle phases are listed (1–4).At later time points (4–12 h), cell cycle profiles are calculated based on BrdU-labeled cells, which are represented in the upper squareof each dot plot. Percentages of cells in the G0/G1 (2 N), S and G2/M (4 N) phases of the cell cycle, and polyploid cells (>4 N) are listedbelow the dot plots. Non-induced and induced, PRCCTFE3 expressing, cells are indicated by – or +tet, respectively.
2402 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
they result from endoreduplication. By combining these data, weconclude that PRCCTFE3 expression leads to an abrogation of cellcycle recovery after nocodazole treatment.
In order to independently confirm these data, we calculatedthe cell doubling times in PRCCTFE3 expressing and non-expressing HEK293/T-REx cells after nocodazole treatment(Fig. S4). By doing so, we found that the non-expressing cellsexhibited a doubling time of ∼46 h during the first 4 days afterrelease from nocodazole. After this time point, a doubling time of∼24 h was measured, which compares to that of non-treatedcells (see above). Conversely, the number of PRCCTFE3 expres-sing cells only increased 6% during the first day and stoppedgrowing after this time point, thus confirming an abrogation ofcell cycle recovery after nocodazole treatment upon PRCCTFE3expression.
PRCCTFE3 expression prevents polyploidy in the presence ofnocodazole
Since we observed polyploidy in nocodazole treated non-PRCCTFE3 expressing cells, but not in similarly treated PRCCTFE3expressing cells, we decided to assess the role of the PRCCTFE3fusion protein in polyploidization in further detail. Therefore, weagain BrdU-pulse labeled non-induced and induced HEK293/T-REx/PRCCTFE3 cells, and assayed by FACS the ploidy status of theBrdU-positive and BrdU-negative cells after incubation duringseveral time intervals in medium containing nocodazole.
After a 16 h incubation in the presence of nocodazole, mostnon-induced cells were arrested at the G2/M phase of the cellcycle, as expected (Fig. 3; −tet, 4 N). Completely in line with theabove results, again a significant proportion of the induced,
Fig. 2 –Delayed recovery of PRCCTFE3 expressing cells fromnocodazole treatment. HEK293/T-REx/PRCCTFE3 cells, incubatedwith orwithout tetracyclin (+ or −tet), were pulse labeled with BrdU and, subsequently, subjected to cell cycle profiling as indicated inFig.1. In the upper boxes dot plots representing BrdU fluorescence (y-axis) versus DNA content (x-axis) are shown. In the lower boxesDNA histograms of the BrdU-positive cells, which are represented in the upper square of each dot plot, are shown. After pulselabeling, cells were grown overnight in the presence of nocodazole, and harvested at different time points (0–8 h) after nocodazoleremoval. The percentages of cells in each of the cell cycle phases are listed.
2403E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
PRCCTFE3-expressing, BrdU-negative cells were delayed in the G0/G1 phase of the cell cycle (Fig. 3; +tet, 2 N). After a 24 h incubationin nocodazole 19.0% of the non-induced BrdU-positive cells werepolyploid (>4 N), whereas the induced, PRCCTFE3 expressing,BrdU-positive cells were predominantly residing in the G2/Mphase of the cell cycle (76.3%), i.e., they were diploid (4 N). After a48 h incubation in the presence of nocodazole 60.4% of the non-induced cells were polyploid. Further extension of the incubationin nocodazole (72 and 96 h, respectively) resulted in a further
increase in the percentage of polyploid non-induced cells (up to66.8%). In reverse, however, even after a 96 h exposure tonocodazole of the induced, PRCCTFE3 expressing, cells thepercentage of polyploid cells remained low (6.2%).
In order to assess the presence of putative apoptotic effectselicited by PRCCTFE3 expression induction in the cell system used,we performed apoptosis assays on nocodazole treated cells, withor without PRCCTFE3 expression induction. By doing so, weobserved very similar apoptotic levels under the various tissue
2404 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
culture conditions tested (Fig. S5) indicating that, next tonocodazole, PRCCTFE3 expression induction does not have anyeffect.
From our results we conclude that exogenous PRCCTFE3expression prevents polyploidization in the presence of themicrotubule destabilizing drug nocodazole. In contrast, polyploi-dizationwas readily observed in similarly treated t(X;1)(p11;q21)-positive Cl89-12117 and Cl89-17872 renal cell carcinoma-derivedcells which endogenously express PRCCTFE3 (19.4% and 46.0%polyploid cells after 96 h in nocodazole versus 6.1% and 6.7% innon-treated cells, respectively). This latter result indicates that inthese tumor cells the PRCCTFE3-mediated mechanism that pre-vents polyploidization is abrogated.
Since degradation of cyclin B1 is a prerequisite for progressionof cells through mitosis, inhibition of this process is expected toresult in a concomitant inhibition of cyclin B1 degradation [32].Indeed, after PRCCTFE3 expression induction, the levels of cyclinB1 remained constant even after release of the cells from anocodazole-induced mitotic arrest (not shown).
PRCCTFE3 induces p21WAF1/CIP1
expression and directly binds toits promoter
Since G1/S and G2/M cell cycle arrest has been shown to bemediated by cyclin-dependent kinase inhibitors such as p21
WAF1/CIP1
Fig. 3 – PRCCTFE3 expression prevents polyploidy in the presence owithout tetracyclin (+ or −tet) were pulse labeled with BrdU andintervals (16–96 h). Subsequently, cell cycle profiles were generatedfluorescence (y-axis) versus DNA content (x-axis), whereas the lowthe dot plots.
[12,33,34]we decided to assess the expression of this protein in ourPRCCTFE3 induction system. Therefore, HEK293/T-REx/PRCCTFE3cells were incubated for 24 h with tetracyclin and subsequentlyassayed by western blot analysis for expression of the PRCCTFE3and p21
WAF1/CIP1proteins (Fig. 4A). As an internal loading control the
small GTPase Ran was used. By doing so, we found that thePRCCTFE3 protein was abundantly expressed after tetracyclininduction, as expected. Concomitantly, we found that thep21
WAF1/CIP1protein exhibited a clear up-regulation in expression.
No overt changes in expression were observed for theendogenous PRCC (not shown), p53 and (control) Ran proteins.
In order to correlate PRCCTFE3 expression induction with theobserved p21
were incubated in different concentrations of tetracyclin and,again, analyzed by western blotting (Fig. S1A). Whereas at lowconcentrations (0–1 ng/ml) only a minor induction of PRCCTFE3expression was observed, this induction was apparent at concen-trations of 10 ng/ml and up. Simultaneously, up-regulation ofp21
WAF1/CIP1was observed. This up-regulationwas not observed after
induction of control empty vector HEK293/T-REx cells withtetracyclin (Fig. S1A; vector). In order to exclude PRCCTFE3induction unrelated effects, non-induced and induced HEK293/T-REx/PRCCTFE3 cells were analyzed for p21
WAF1/CIP1protein levels at
different time intervals. By doing so, we found that as early as 2–4 h after PRCCTFE3 expression induction the p21
WAF1/CIP1proteinwas
f nocodazole. HEK293/T-REx/PRCCTFE3 cells incubated with orgrown in the presence of nocodazole during different timeas indicated in Fig.1. The upper boxes represent dot plots of BrdUer boxes represent DNA histograms of the BrdU-positive cells in
2405E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
up-regulated (Fig. S1B). This up-regulation was further increasedafter 8 h and matched closely with the concomitant up-regulationof the PRCCTFE3 protein. Taken together, we conclude that de novoPRCCTFE3 expression up-regulates p21
WAF1/CIP1expression.
Fig. 4 – PRCCTFE3 induces p21WAF1/CIP1
expression. (A) Proteinswere extracted from HEK293/T-REx/PRCCTFE3 cells incubated24 h with or without tetracyclin (+ or−tet) and, subsequently,analyzed by western blotting using anti-PRCC, anti-p53, anti-p21
T-REx/PRCC and HEK293/T-REx/PRCCTFE3 cells were incubatedwith or without tetracyclin (+ or −tet) and/or cycloheximide(+ or −CHX), and analyzed by western blotting using anti-p53and anti-p21
WAF1/CIP1antibodies. (C) Protein–DNA complexes
were immunoprecipitated from HEK293/PRCCTFE3-TY1 cellsusing an anti-tag antibody (αTY1) and a mouse pre-immuneserum (IgG). The recovered genomic DNAs, as well as the inputchromatin, were analyzed by PCR using primer pairs spanningthe p21
WAF1/CIP1gene promoter. (D) Hela, U2OS and Rat-1 cells
were either not transfected (—) or transiently transfected withan empty vector or a PRCCTFE3 expression construct. Cells werelysed 24 h after transfection and analyzed by western blottingusing anti-PRCC, anti-p21
WAF1/CIP1and anti-Ran (control)
antibodies. Protein bands were derived from single membranesand band intensities of at least 3 independent experimentswere quantified and averaged using the ImageJ software tool.The amount of proteinwas corrected for by the loading control.The non-transfected condition (—) was arbitrarily set at 100%.
In order to assess whether the observed PRCCTFE3-mediatedup-regulation of p21
WAF1/CIP1results from de novo protein synthesis,
we cultured both tetracyclin-induced and -uninduced HEK293/T-REx/PRCCTFE3 cells and control HEK293/T-REx/PRCC cells in thepresence or absence of the mRNA translation inhibitor cyclohex-imide (CHX; Fig. 4B). As expected, no p21
WAF1/CIP1expression was
observed under the various culture conditions applied to thecontrol HEK293/T-REx/PRCC cells. In the HEK293/T-REx/PRCCTFE3 cells, however, p21
WAF1/CIP1protein up-regulation was
readily observed after PRCCTFE3 induction in the absence ofcycloheximide but not in its presence, thus indicating that theabove-observed PRCCTFE3-induced up-regulation of p21
WAF1/CIP1
indeed reflects de novo protein synthesis. As in Fig. 4A, no overtchanges were observed in p53 expression.
Since the PRCCTFE3 fusion protein can act as a transcriptionalactivator [22], we set out to assess whether the observed p21
WAF1/CIP1
up-regulation results fromdirect bindingof thePRCCTFE3protein toits gene promoter. To this end, we performed chromatin immuno-precipitation (ChIP) experiments in HEK293 cells expressing a TY-1tagged (see Materials and methods) full-length PRCCTFE3 protein.The recovered ChIP DNAs were analyzed for the enrichment ofp21
WAF1/CIP1promoter sequences. By doing so, we indeed found that
the anti-TY1ChIPDNAswere enriched for these sequences (Fig. 4C).From these results we conclude that the PRCCTFE3 protein activatesthe p21
WAF1/CIP1promoter through direct binding.
In order to exclude HEK293-specific effects three independentcell lines, Hela, U2OS and Rat-1, were transiently transfected witheither empty vectors or PRCCTFE3 expression constructs andanalyzed for p21
WAF1/CIP1expression by western blotting (Fig. 4D). As
in HEK293 cells, also in Hela and U2OS cells de novo expression ofPRCCTFE3 elicited an increase in p21
WAF1/CIP1protein expression
levels. As expected, the p21WAF1/CIP1
-deficient Rat-1 cells [35–37]failed to show such an induction.
Exogenous PRCCTFE3 expression does not activate the pRBpathway in t(X;1)(p11;q21)-positive tumor cells
Since p21WAF1/CIP1
is known to act within the pRB cell cycle controlpathway [12], we set out to investigate whether the p21
WAF1/CIP1
downstream effectors CDK2 and/or pRB are affected by de novoPRCCTFE3 expression. To this end, HEK293/T-REx/PRCCTFE3 cellswere again incubated with tetracyclin and, subsequently, assayedby western blot analysis for CDK2 protein expression and pRBphosphorylation levels (Fig. 5). After de novo PRCCTFE3 expressioninduction, we observed a clear decrease in the CDK2 protein level(Fig. 5A). Since CDK2 is known to catalyze phosphorylation of pRBat serine 807, which in turn disrupts its binding to E2F and enablescell cycle progression [38–41], we used an anti-pRB phospho S807antibody (see Materials and methods) to assess the pRB phos-phorylation status. By doing so, we observed a clear decrease in thelevel of phosphorylated pRB protein after de novo PRCCTFE3expression induction (Fig. 5B). Together, these observationssupport a scenario in which the PRCCTFE3-induced cell cycledelay is brought about through a p21
WAF1/CIP1-mediated activation of
the pRB cell cycle control pathway.In order to further substantiate this notion, we stably knocked
down p21WAF1/CIP1
expression in our HEK293/T-REx/PRCCTFE3inducible cell line using a retroviral shRNA construct (Fig. S6A).Next, we determined the doubling times of the shRNA engineeredPRCCTFE3 expressing and non-expressing cells. By doing so, no
Fig. 5 – PRCCTFE3 expression activates the pRB signaling pathway. Proteins were extracted from HEK293/T-REx/PRCCTFE3 cellsincubated 0 and 48 h with tetracycline and, subsequently, analyzed by western blotting using anti-CDK2 (A), anti-phospho-pRB (B)and anti-Ran (control) antibodies. Protein bands were derived from single membranes and quantified using the ImageJ softwaretool. The amount of proteinwas corrected for by using the loading control. The non-induced condition (t=0)was arbitrarily set at 1.
2406 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
overt differences in cell doubling times were observed in therespective p21
WAF1/CIP1knockdown cells (both ∼24 h; Fig. S6B),
which fully supports our notion that the PRCCTFE3-induced cellcycle delay is mediated by p21
WAF1/CIP1.
In order to subsequently assess whether exogenous PRCCTFE3expression also affects p21
WAF1/CIP1protein levels in t(X;1)(p11;q21)-
positive renal tumor cells, we transfected the Cl89-12117 cell line(see above) with a PRCCTFE3 expression construct, resulting in amarked increase in the PRCCTFE3 protein level (Fig. 6A). Incontrast to the above obtained results, however, we failed toobserve any concurrent effect on the p21
WAF1/CIP1expression level
(Fig. 6A). In order to assess whether other components of the pRBpathway, i.e., CDK2 and/or pRB, might still be affected by theexogenous PRCCTFE3 expression in renal tumor cells, we againperformed western blot analyses on non-transfected and trans-fected Cl89-12117 cells using anti-CDK2 and anti-pRB phosphoS807 antibodies. By doing so, we failed to detect any changes inCDK2 protein expression (Fig. 6B) and/or pRB phosphorylation(Fig. 6C) levels. These data are completely in line with a model inwhich in t(X;1)(p11;q21)-positive renal cell carcinomas an initialPRCCTFE3-induced cell cycle delay is numbed, thus allowing aprogression towards full-blown malignancy.
Discussion
We found that de novo expression of the renal cell carcinoma-associated oncogenic fusion protein PRCCTFE3 provokes cell cycledelay. This delay was mediated by direct induction of the cyclin-dependent kinase inhibitor p21
WAF1/CIP1and affected the G1/S, G2/M
and S phases of the cell cycle. In addition, we found that thePRCCTFE3-induced p21
WAF1/CIP1up-regulation was associated with
an activation of the pRB pathway, i.e., with CDK2 protein and pRBphosphorylation down-regulation. The PRCCTFE3-induced cellcycle delay could be rescued by knocking down p21
WAF1/CIP1
expression. No PRCCTFE3-mediated induction of the cyclin-dependent kinase inhibitor p16INK4a was observed (not shown).Our results are in accordance with literature data [12,33,34],
including the p21WAF1/CIP1
-mediated S phase delay in PRCCTFE3expressing cells [42,43].
We also found that, after exposure of cells to the microtubuledestabilizing drug nocodazole, PRCCTFE3 expression preventspolyploidization. This finding is in line with, and substantiates,the observed PRCCTFE3-induced G2/M and G1/S cell cycle delay,thus preventing mitotically compromised cells from undergoingendoreduplication. The observed effects were not due to putativeapoptotic effects elicited by PRCCTFE3 expression induction.Others have found that induction of p21
WAF1/CIP1is required for a
G1/S cell cycle arrest after activation of the mitotic spindlecheckpoint, and that this arrest prevents endoreduplication[44–47]. Subsequently, Vogel et al. [48] found that in normal cellcycle control the prevention of polyploidization is based on cross-talk between the mitotic spindle assembly checkpoint and theG2/M and G1/S checkpoints. Our observation that PRCCTFE3 iscapable of inducing p21
WAF1/CIP1expression and, concomitantly,
G2/M and G1/S cell cycle delay is in agreement with these data.More interestingly we found that, in contrast to HEK293, Hela,
and U2OS cells, exogenous expression of PRCCTFE3 in t(X;1)(p11;q21)-positive renal tumor cells failed to affect p21
WAF1/CIP1expres-
sion levels and/or pRB pathway activation. This latter resultsuggests that the processes that link PRCCTFE3 expression top21
WAF1/CIP1-mediated cell cycle delay are uncoupled in these tumor
cells. Such an uncoupling is completely in line with our observa-tion that in t(X;1)(p11;q21)-positive renal tumor cells theprevention of nocodazole-induced polyploidization is abrogated.
Our observations are also in accordance with the notion [49]that an oncogenic chromosome translocation by itself may not besufficient to induce malignant transformation, and that comple-mentary alterations are required to provoke transformation. Such astep-wise mechanism is compatible with recent observationsindicating that oncogenic stimulation may induce cell cycle arrestand premature senescence [1–9], and that a subsequent elimina-tion of this protective mechanism may permit continued pro-liferation in the presence of the initial oncogenic event, therebyleading to a full-blown malignancy [13]. As we observed p21
WAF1/CIP1
expression in t(X;1)(p11;q21)-positive renal tumor cells, we
Fig. 6 – PRCCTFE3 expression does not activate the pRB pathwayin t(X;1)(p11;q21)-positive tumor cells. Proteins were extractedfrom non-transfected Cl89-12117 cells and Cl89-12117 cells 24 hafter transfection with a PRCCTFE3 expression construct.Subsequently, these proteins were subjected to western blotanalysis using (A) anti-PRCC and anti-p21
WAF1/CIP1, (B) anti-CDK2
and (C), anti-phospho-pRB antibodies. As a loading control ananti-Ran antibody was used. Protein bands were derived fromsingle membranes.
2407E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
hypothesize that in these PRCCTFE3 expressing cells [18] the cellcycle delay is numbed. Since PRCCTFE3 directly binds to thep21
WAF1/CIP1promoter, we propose that this numbing must be
brought about by an alternative, as yet unknown, mechanismpossibly involving the recruitment of one or more repressorproteins and/or the up-regulation of genetic factors such asuncovered by Voorhoeve et al. [14]. By using a functional screen,these latter authors identified two microRNAs, miRNA-372 andmiRNA-373, whose expression could substitute for the loss of wildtype p53 and, by doing so, could overcome oncogenic H-RASV12-mediated cell cycle delay. These observations nicely correlate witha previous finding indicating that the Ewing's sarcoma-associatedoncogenic fusion protein EWSFLI induces a p53-dependent growtharrest in primary human fibroblasts [50], and that this growtharrest can be rescued by inhibition of p53.
More recently, it was found that also the human synovialsarcoma-associated oncogenic fusion protein SYTSSX1 inducesp21
WAF1/CIP1expression [3]. Introduction of mutations in the Sp1/Sp3
binding sites of the p21WAF1/CIP1
gene promoter abolished SYTSSX1-induced transcription activity, thus indicating that in synovialsarcoma cells the observed effect may be Sp1/Sp3-dependent.Sequencing of the p21
WAF1/CIP1promoter in the t(X;1)(p11;q21)-
positive renal cell carcinoma cell line Cl89-121178 failed to reveal
any mutations (not shown), which is in line with the above notionthat in these cells the escape from cell cycle delay may be broughtabout through an alternative mechanism(s).
In conclusion, our data indicate that exogenous expression ofthe renal cell carcinoma-associated oncogenic fusion proteinPRCCTFE3 provokes a p21
WAF1/CIP1-mediated cell cycle delay. Since
p21WAF1/CIP1
is also expressed in primary PRCCTFE3-positive renaltumor cells, and since we found that in these tumor cells p21
WAF1/CIP1
levels are no longer affected by exogenous PRCCTFE3, we suggestthat during the course of tumor development the initial p21
WAF1/CIP1-
induced cell cycle delay must be numbed, thus allowing a step-wise progression towards full-blown malignancy. This step-wiseprogression may include a PRCCTFE3 mediated acquisition ofresistance to anti-proliferative signals, as was recently observedafter TFE3 over-expression in p21
WAF1/CIP1-deficient Rat-1 cells [51].
Acknowledgments
The authors thank Marion Lohrum, Reuven Agami, Eric Schoen-makers, Diederik de Bruijn, Piet van Erp and Frank van Leeuwen foradvice and support and Henk Stunnenberg for providing the BB2antibody. This work was supported by the Dutch Cancer Society.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.yexcr.2009.04.022.
R E F E R E N C E S
[1] M. Serrano, A.W. Lin, M.E. McCurrach, D. Beach, S.W. Lowe,Oncogenic ras provokes premature cell senescence associatedwith accumulation of p53 and p16INK4a, Cell 88 (1997) 593–602.
[2] M. Braig, C.A. Schmitt, Oncogene-induced senescence: putting thebrakes on tumor development, Cancer Res. 66 (2006) 2881–2884.
[3] M. Tsuda, T. Watanabe, T. Seki, T. Kimura, H. Sawa, A. Minami,T. Akagi, K. Isobe, K. Nagashima, S. Tanaka, Induction ofp21(WAF1/CIP1) by human synovial sarcoma-associated chimericoncoprotein SYT-SSX1, Oncogene 24 (2005) 7984–7990.
[4] M. Collado, M. Serrano, The power and the promise ofoncogene-induced senescence markers, Nat. Rev. Cancer 6 (2006)472–476.
[5] W.J. Mooi, D.S. Peeper, Oncogene-induced cellsenescence—halting on the road to cancer, N. Engl. J. Med. 355(2006) 1037–1046.
[6] M. Braig, S. Lee, C. Loddenkemper, C. Rudolph, A.H. Peters,B. Schlegelberger, H. Stein, B. Dorken, T. Jenuwein, C.A. Schmitt,Oncogene-induced senescence as an initial barrier in lymphomadevelopment, Nature 436 (2005) 660–665.
[7] Z. Chen, L.C. Trotman, D. Shaffer, H.K. Lin, Z.A. Dotan, M. Niki,J.A. Koutcher, H.I. Scher, T. Ludwig, W. Gerald, C. Cordon-Cardo,P.P. Pandolfi, Crucial role of p53-dependent cellular senescence insuppression of Pten-deficient tumorigenesis, Nature 436 (2005)725–730.
[8] M. Collado, J. Gil, A. Efeyan, C. Guerra, A.J. Schuhmacher,M. Barradas, A. Benguria, A. Zaballos, J.M. Flores, M. Barbacid,D. Beach, M. Serrano, Tumour biology: senescence inpremalignant tumours, Nature 436 (2005) 642.
[9] C. Michaloglou, L.C. Vredeveld, M.S. Soengas, C. Denoyelle,T. Kuilman, C.M. van der Horst, D.M. Majoor, J.W. Shay, W.J. Mooi,
2408 E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
D.S. Peeper, BRAFE600-associated senescence-like cell cyclearrest of human naevi, Nature 436 (2005)720–724.
[10] A.L. Fridman, M.A. Tainsky, Critical pathways in cellularsenescence and immortalization revealed by gene expressionprofiling, Oncogene 27 (2008) 5975–5987.
[11] E. Hara, R. Smith, D. Parry, H. Tahara, S. Stone, G. Peters, Regulationof p16CDKN2 expression and its implications for cellimmortalization and senescence, Mol. Cell Biol. 16 (1996)859–867.
[12] F. Bunz, A. Dutriaux, C. Lengauer, T. Waldman, S. Zhou, J.P. Brown,J.M. Sedivy, K.W. Kinzler, B. Vogelstein, Requirement for p53 andp21 to sustain G2 arrest after DNA damage, Science 282 (1998)1497–1501.
[13] P.M. Voorhoeve, R. Agami, Unraveling human tumor suppressorpathways: a tale of the INK4A locus, Cell Cycle 3 (2004)616–620.
[14] P.M. Voorhoeve, C.le Sage, M. Schrier, A.J. Gillis, H. Stoop, R. Nagel,Y.P. Liu, J.van Duijse, J. Drost, A. Griekspoor, E. Zlotorynski,N. Yabuta, G. De Vita, H. Nojima, L.H. Looijenga, R. Agami,A genetic screen implicates miRNA-372 and miRNA-373 asoncogenes in testicular germ cell tumors, Cell 124 (2006)1169–1181.
[15] J.N. Eble, G. Sauter, J.I. Epstein, I.A. Sesterhenn, Tumors of thekidney, World Health Organization Classification of Tumors.Pathology and Genetics of Tumours of the Urinary System andMale Genital Organs, IARC Press, Lyon, France, 2004, pp. 37–38.
[16] K. Medendorp, J.J.M. van Groningen, M. Schepens, L. Vreede,J. Thijssen, E.F. Schoenmakers, W.H. van den Hurk, A. Geurts vanKessel, R.P. Kuiper, Molecular mechanisms underlying the MiTtranslocation subgroup of renal cell carcinomas, Cytogenet.Genome Res. 118 (2007) 157–165.
[17] R. Ramphal, A. Pappo, M. Zielenska, R. Grant, B.Y. Ngan, Pediatricrenal cell carcinoma: clinical, pathologic, and molecularabnormalities associated with the members of the MiTtranscription factor family, Am. J. Clin. Pathol. 126 (2006)349–364.
[18] M.A. Weterman, M. Wilbrink, A. Geurts van Kessel, Fusion of thetranscription factor TFE3 gene to a novel gene, PRCC, in t(X;1)(p11;q21)-positive papillary renal cell carcinomas, Proc. Natl.Acad. Sci. U. S. A. 93 (1996) 15294–15298.
[19] S.K. Sidhar, J. Clark, S. Gill, R. Hamoudi, A.J. Crew, R. Gwilliam,M. Ross, W.M. Linehan, S. Birdsall, J. Shipley, C.S. Cooper,The t(X;1)(p11.2;q21.2) translocation in papillary renal cellcarcinoma fuses a novel gene PRCC to the TFE3 transcriptionfactor gene, Hum. Mol. Genet. 5 (1996) 1333–1338.
[20] H. Beckmann, L.K. Su, T. Kadesch, TFE3: a helix–loop–helix proteinthat activates transcription through the immunoglobulinenhancer muE3 motif, Genes Dev. 4 (1990) 167–179.
[21] Y.M. Skalsky, P.M. Ajuh, C. Parker, A.I. Lamond, G. Goodwin,C.S. Cooper, PRCC, the commonest TFE3 fusion partner in papillaryrenal carcinoma is associated with pre-mRNA splicing factors,Oncogene 20 (2001) 178–187.
[22] M.A. Weterman, J.J.M. van Groningen, A. Jansen, A. Geurts vanKessel, Nuclear localization and transactivating capacities of thepapillary renal cell carcinoma-associated TFE3 and PRCC (fusion)proteins, Oncogene 19 (2000) 69–74.
[23] M.A.Weterman, J.J.M. van Groningen, A. den Hartog, A. Geurts vanKessel, Transformation capacities of the papillary renal cellcarcinoma-associated PRCCTFE3 and TFE3PRCC fusion genes,Oncogene 20 (2001) 1414–1424.
[24] Y. Sagara, Y. Miyata, K. Nomata, K. Abe, J. Eguchi, T. Hayashi,H. Sakai, S. Kanda, H. Kanetake, TFE3-renal carcinoma in an adultpatient: a case with strong expression of phosphorylatedhepatocyte growth factor (HGFR)/Met, Pathol. Res. Pract. 205(2009) 57–61.
[25] M. Tsuda, I.J. Davis, P. Argani, N. Shukla, G.G. McGill, M. Nagai,T. Saito, M. Lae, D.E. Fisher, M. Ladanyi, TFE3 fusions activate METsignaling by transcriptional up-regulation, defining another class
of tumors as candidates for therapeutic MET inhibition, CancerRes. 67 (2007) 919–929.
[26] P. Argani, M. Tsuda, M.-S. Chen, M. Edgar, M. Landanyi,Overexpression of the MET receptor tyrosine kinase in tumorswith TFE3 gene fusions, Mod. Pathol. 19 (suppl. 1) (2006) 297ARef Type: Abstract.
[27] M.A. Weterman, J.J.M. van Groningen, L. Tertoolen, A. Geurts vanKessel, Impairment of MAD2B–PRCC interaction in mitoticcheckpoint defective t(X;1)-positive renal cell carcinomas,Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 13808–13813.
[28] P.M. Voorhoeve, R. Agami, The tumor-suppressive functions of thehuman INK4A locus, Cancer Cell 4 (2003) 311–319.
[29] D.R.de Bruijn, S.V. Allander, A.H. van Dijk, M.P. Willemse,J. Thijssen, J.J.M. van Groningen, P.S. Meltzer, A. Geurts van Kessel,The synovial-sarcoma-associated SS18-SSX2 fusion proteininduces epigenetic gene (de)regulation, Cancer Res. 66 (2006)9474–9482.
[30] S. Carreira, J. Goodall, I. Aksan, S.A. La Rocca, M.D. Galibert,L. Denat, L. Larue, C.R. Goding, Mitf cooperates with Rb1 andactivates p21Cip1 expression to regulate cell cycle progression,Nature 433 (2005) 764–769.
[31] D.M. Scolnick, T.D. Halazonetis, Chfr defines a mitotic stresscheckpoint that delays entry into metaphase, Nature 406 (2000)430–435.
[32] D.A. Brito, C.L. Rieder, Mitotic checkpoint slippage in humansoccurs via cyclin B destruction in the presence of an activecheckpoint, Curr. Biol. 16 (2006) 1194–1200.
[33] V. Dulic, G.H. Stein, D.F. Far, S.I. Reed, Nuclear accumulation ofp21Cip1 at the onset of mitosis: a role at the G2/M-phasetransition, Mol. Cell Biol. 18 (1998) 546–557.
[34] T.A. Chan, P.M. Hwang, H. Hermeking, K.W. Kinzler, B. Vogelstein,Cooperative effects of genes controlling the G(2)/M checkpoint,Genes Dev. 14 (2000) 1584–1588.
[35] A.K. Godwin, M.W. Lieberman, Early and late responses toinduction of rasT24 expression in Rat-1 cells, Oncogene 5 (1990)1231–1241.
[36] X. Lu, D.P. Lane, Differential induction of transcriptionally activep53 following UV or ionizing radiation: defects in chromosomeinstability syndromes? Cell 75 (1993) 765–778.
[37] L.A. Allan, T. Duhig, M. Read, M. Fried, The p21(WAF1/CIP1)promoter is methylated in Rat-1 cells: stable restoration ofp53-dependent p21(WAF1/CIP1) expression after transfection of agenomic clone containing the p21(WAF1/CIP1) gene, Mol. Cell Biol.20 (2000) 1291–1298.
[38] R.A. Weinberg, The retinoblastoma protein and cell cycle control,Cell 81 (1995) 323–330.
[40] W. Du, M. Vidal, J.E. Xie, N. Dyson, RBF, a novel RB-related genethat regulates E2F activity and interacts with cyclin E inDrosophila, Genes Dev. 10 (1996) 1206–1218.
[41] D.J. Riley, E.Y. Lee, W.H. Lee, The retinoblastoma protein: morethan a tumor suppressor, Annu. Rev. Cell Biol. 10 (1994)1–29.
[42] S. Waga, G.J. Hannon, D. Beach, B. Stillman, The p21 inhibitor ofcyclin-dependent kinases controls DNA replication by interactionwith PCNA, Nature 369 (1994) 574–578.
[43] V.V. Ogryzko, P. Wong, B.H. Howard, WAF1 retards S-phaseprogression primarily by inhibition of cyclin-dependent kinases,Mol. Cell Biol. 17 (1997) 4877–4882.
[44] S.H. Khan, G.M. Wahl, p53 and pRb prevent rereplication inresponse to microtubule inhibitors by mediating a reversibleG1 arrest, Cancer Res. 58 (1998) 396–401.
[45] J.S. Lanni, T. Jacks, Characterization of the p53-dependentpostmitotic checkpoint following spindle disruption, Mol. CellBiol. 18 (1998) 1055–1064.
2409E X P E R I M E N T A L C E L L R E S E A R C H 3 1 5 ( 2 0 0 9 ) 2 3 9 9 – 2 4 0 9
[47] M. Motwani, X. Li, G.K. Schwartz, Flavopiridol, a cyclin-dependentkinase inhibitor, prevents spindle inhibitor-inducedendoreduplication in human cancer cells, Clin. Cancer Res. 6(2000) 924–932.
[48] C. Vogel, A. Kienitz, I. Hofmann, R. Muller, H. Bastians, Crosstalk ofthe mitotic spindle assembly checkpoint with p53 to preventpolyploidy, Oncogene 23 (2004) 6845–6853.
[50] S.L. Lessnick, C.S. Dacwag, T.R. Golub, The Ewing's sarcomaoncoprotein EWS/FLI induces a p53-dependent growth arrest inprimary human fibroblasts, Cancer Cell 1 (2002) 393–401.
[51] S.M. Nijman, E.M. Hijmans, S. El Messaoudi, M.M. van Dongen,C. Sardet, R. Bernards, A functional genetic screen identifiesTFE3 as a gene that confers resistance to the anti-proliferativeeffects of the retinoblastoma protein and transforminggrowth factor-beta, J. Biol. Chem. 281 (2006)21582–21587.
