Blockade of AP-1 Activity by Dominant-NegativeTAM67 Can Abrogate the Oncogenic Phenotypein Latent Membrane Protein 1-Positive HumanNasopharyngeal Carcinoma

Xin Jin, Xin Song, Lili Li, Zhenlian Wang, Yongguang Tao, Lin Deng, Ming Tang, Wei Yi, and Ya Cao*

Cancer Research Institute, Xiangya School of Medicine, Central South University, Changsha, Hunan, PR China

Although activating protein-1 (AP-1) transcription factors play an important role in mediating metastasis fornasopharyngeal carcinoma (NPC), the biological and physiological functions of AP-1, in relation to the oncogenicphenotype of NPC, are not fully understood. Our previous study showed that the latent membrane protein 1 (LMP1)

mediated a primary dimer form of c-jun and jun B. In this study, we used a NPC cell line that express a specific inhibitorof AP-1, a dominant-negative c-jun mutant (TAM67), to investigate the role of AP-1 in regulating the NPC oncogenicphenotype. First, we observed that TAM67 inhibited cell growth in vitro and in vivo. Next, with Western blotting, we

discovered that TAM67 impaired the cyclin D1/cdk4 complex but had little effect on the cyclin E/cdk2 complex,concomitantly with inhibiting Rb phosphorylation. RT-PCR and luciferase assay results demonstrated that the levels ofcyclin D1 mRNA and the promoter activity in TAM67 transfectants were reduced as compared with control cells.

Thereby, we show that blockade of AP-1 transcriptional activity has a negative impact on cyclin D1 transcription. Weobtained the first evidence that TAM67 prevented NPC growth both in vitro and in vivo. AP-1 appears to be a noveltarget for treating or preventing LMP1-positive NPC effectively. � 2007 Wiley-Liss, Inc.

Key words: Epstein–Barr virus; latent membrane protein 1; heterodimer; TAM67; jun B; c-jun; nasopharyngeal

carcinoma; cell cycle

INTRODUCTION

Epstein–Barr virus (EBV), a gamma herpes virus,is associated with many malignancies, includingBurkitt’s lymphoa, nasopharyngeal carcinoma (NPC),and Hodgkin’s disease [1], The EBV-encoded latentmembrane protein 1 (LMP1) is an integral membraneprotein that plays an important role in promotingmalignant cell transformation and is detected at theprotein level in 35–65% of all NPCs [2–4]. LMP1consists of a short, highly charged cytoplasmicamino terminus of 23 amino acids; 6 hydrophobictransmembrane domains, required for protein aggre-gation; and a long acidic cytoplasmic carboxylterminus of 200 amino acids [5]. The carboxylterminus comprises two functional domains: themembrane proximal C-terminal activation region-1(carboxyl terminal activator region [CTAR1], resi-dues 194–232) and the membrane distal CTAR2(residues 351–386) [6,7]. CTAR2 activates activatingprotein-1 (AP-1) signaling pathways. A critical eventin LMP1 signaling is AP-1 activation, as shown by thefindings that AP-1 activation has caused abnormalmorphological cell change [8,9].

The AP-1 transcription factor consists of a large setof dimer combinations formed between the jun, FOS,and ATF families of proteins. They contain transacti-vation domains, DNA-binding domains, and leucinezippers. The leucine regions permit the formation of

AP-1 member dimers, which are the functionaltranscriptional units. The jun family, consisting ofc-jun, jun B, and jun D, can form both heterodimersand homodimers among themselves, whereas theFOS family members (c-FOS, FOS B, Fra1, and Fra2)can heterodimerize only with jun proteins. AP-1proteins are usually activated not only in response tostress signals, such as UV irradiation but also by thegrowth factor pathways, promoting the mitogen-induced cell cycle progression or regulating apopto-sis by modulating cyclin D1 and p53 expression.

Recent findings show that the overexpression ofc-jun makes tumors invasive [10,11]. On the otherhand, it also triggers apoptosis [12,13]. In contrast,

MOLECULAR CARCINOGENESIS 46:901–911 (2007)

� 2007 WILEY-LISS, INC.

Abbreviations: EBV, Epstein–Barr virus; NPC, nasopharyngealcarcinoma; LMP1, latent membrane protein 1; CTAR, carboxylterminal activator region; AP-1, activating protein-1; FCS, fetal calfserum; RT, reverse transcription; PCR, polymerase chain reaction;pRb, hypophosphorylated form of Rb.

Xin Jin’s present address is Institute of Health Sciences, ShanghaiInstitutes for Biological Sciences, Chinese Academy of Sciences andShanghai Jiao Tong University School of Medicine, Shanghai, China.

*Correspondence to: Cancer Research Institute, Xiangya Schoolof Medicine, Central South University, 110 Xiangya Road, Changsha410078, Hunan, PR China.

Received 21 November 2006; Revised 18 January 2007; Accepted5 February 2007

DOI 10.1002/mc.20319

jun B is always classified as the inhibiting componentof the AP-1 transcription factor [14,15], and it canregulate erythroid and myogenic differentiation[16]. However, some findings demonstrate that junB promotes carcinogenesis [17,18]. What is impor-tant to this discussion is that the balance of jun Bwith c-jun activity is involved in regulating the keysteps in the proliferation and differentiation process[19]. Therefore, not only can AP-1 complexespromote differentiation and trigger apoptosis, butthey may also induce cellular proliferation andtransformation, depending upon their composition.In our previous study, with a Tet-on LMP1 HNE2 cellline, we showed that LMP1 induced high AP-1activity mainly including jun B and c-jun proteins,and that jun B could efficiently form a new hetero-dimeric complex with c-jun protein under theregulation of LMP1. We also found that this hetero-dimeric form can bind to the AP-1 consensussequence [20,21]. Although we found that LMP1mediated high AP-1 activity, little is known aboutthe biological and physiological functions of the AP-1 transcription factors in NPC.

In the present study, we used a specific inhibitor ofAP-1, a dominant-negative c-jun mutant, TAM67.We found that TAM67 could block AP-1 activity. Wealso showed that TAM67 inhibited NPC growth bothin vivo and in vitro. With Western blotting toelucidate the exact molecular mechanism, we foundthat TAM67 inhibited LMP1-positive cell growth anddecreased the expression of cyclin D1 proteins, thusreducing cdk4 activity, but it had little effect on thecomplex of cyclin E and cdk2, which in turn causedthe hypophosphorylation of retinoblastoma (Rb)that blocked the release of E2F1 from the Rb/E2Fcomplex. Furthermore, the downregulation of cyclinD1 was seen at both RNA and protein levels. Theseresults demonstrate that AP-1 is a critical transcrip-tion factor for cyclin D1 transcription. In conclusion,we obtained for the first time evidence that theblockade of AP-1 activity by dominant-negativeTAM67 prevented NPC growth both in vitro and invivo. AP-1 appears to be a novel target for treating orpreventing LMP1-positive NPC effectively.

MATERIALS AND METHODS

Cell Lines and Culture Conditions

HNE2-LMP1 cells, which can stably express LMP1and were transfected with vector pSG5, werecultured in RPMI 1640 medium containing 10%fetal calf serum (FCS) and 100 U/mL penicillin/streptomycin. In addition the HNE2-LMP1-TAM67NPC cell line, which can express the dominant-negative mutant of c-jun (TAM67), was established[22] and cultured in RPMI 1640 medium containing10% FCS and 100 U/mL penicillin/streptomycin. Allcell lines were grown at 378C under 5% CO2 and 95%air at 99% humidity.

Materials

Cyclin D1 promoters luciferase reporter plasmidwas kindly provided by Prof. Strauss [23]. The AP-1luciferase reporter plasmid, whichcontains a sequen-ce of the collagenase promoter region (�73 toþ67 bp) including one AP-1-binding site, was kindlyprovided by Dr. Li JJ [51]. The pGL2Basic-E2F1-lUCreporter plasmid contains three E2F1-binding sites[24]. The h-gal expression plasmid (pRSVh-gal) waskindly provided by Prof. Perkins, and was used as anintrinsic control. Antibodies used were as follows:Cyclin D1 (DCS-6, Dako, Glostrup, Denmark), cyclinE (pc438, Oncogene, Cambridge, MA), cdk2 (sc-163,Santa, CA), cdk4 (sc-260, Santa), E2F1 (sc-193,Santa), Rb (sc-50, Santa), a-tubulin (sc-5286, Santa),c-jun (sc-1694, Santa), p21 (sc-756, Santa), p27(sc-1641, Santa), phosphorylated Rb-821 (44–582,Biosource, Camarillo, California), phosphorylatedRb-826 (44–586, Biosource), c-jun (sc-44, Santa), andjun B (sc-8051, Santa).

Cell Growth Assay

The MTT assay was used to assess the effects ofTAM67 on cell growth. Cells (2� 104) were platedinto each well of a 96-well microtiter plate. Every dayfor the following 4 d, 80 mL of MTT solution (5 mg/mL in PBS) was added to each well. Cells wereincubated at 378C for 4 h. MTT crystals were thensolubilized in 800 mL of DMSOþ2.5% medium.Absorbance was read at 570 nm. All experimentswere performed at least three times.

Flow Cytometry Analysis

Cells were collected after being cultured in 10%fetal bovine serum for 36 h. They were then rinsedwith PBS and suspended in 75% ethanol at �708Covernight. Fixed cells were centrifuged and washedwith PBS twice. To detect DNA content, cells wereincubated in the dark with 50 mg/mL of PI and 0.1%RNase A in 400 mL of PBS at 258C for 30 min. Stainedcells were analyzed on a FAC sort (Becton Dickinson,San Jose, CA). The percentage of cell cycle phases wasdetermined with the Cell Quest software program.

Western Blotting

Whole cell extracts were prepared as follows: aftertreatment with different media, cells were washedwith ice-cold PBS and lysed by incubating 10-min onice with lysis buffer (10 mM Tris-HCl pH 8.0, 1 mMEDTA, 2% SDS, 5 mM DTT, 10 mM PMSF, proteinaseinhibitors mix). The lysate was centrifuged for 5 minat 14,000 rpm in a microfuge. After protein quanti-fication by BCA assay reagents (Pierce, Rockford, IL),50-mg aliquots of protein lysates were mixed withsample buffer and boiled for 5 min. The samples werethen resolved on 5% and 10% polyacrylamide gelscontaining SDS, and transferred to a nitrocellulosemembrane by electroblotting. The membrane was

902 JIN ET AL.

Molecular Carcinogenesis DOI 10.1002/mc

incubated in blocking buffer (TBS containing 5%skim milk and 0.1% Tween-20) for 2 h, followedby incubation with primary antibody diluted inthe same buffer. Antibody was diluted 1:1,000. Themembrane was washed in TBS containing 0.1%Tween-20. Specific secondary antibody was detectedwith peroxidase-conjugated anti-IgG at 1:10,000.Relative proteins were detected by the supersignalchemiluminescence system (ECL, Pierce, Rockford,IL) followed by exposure to autoradiographic film.

Transient Transfection and Luciferase Assay

Transfections were performed with the SuperFecttransfection reagent (Qiagen, Hilden, Germany)method. Before transfection, cells were seeded in6-well plates overnight, while both plasmid DNA(1.0 mg) and SuperFect transfection Reagent (10 mL)were diluted in serum-free medium (100 mL), respec-tively. DNA and SuperFect were then mixed andincubated for 10 min at room temperature to formthe transfection complex. Cells were rinsed twicewith PBS. We then added 600 mL of cell growthmedium (containing serum and antibodies) to thetransfection complex, mixed, and immediatelytransferred the entire volume to the cells. After 3 hof incubation at 378C in the CO2 incubator, DNA-containing medium was replaced with fresh mediumcontaining 10% serum, and cells were furtherincubated for 24 h before use. Luciferase activitywas detected with a luciferase assay kit (Promegacorporation, Madison, WI).

Immunoprecipitation

Five microliters of antibody and Sepharose beads(Pharmacia, Uppsala, Sweden) were mixed with500 mg of total extract in 1 mL of lysis buffer plusprotease and phosphatase inhibitors at 48C over-night. On the following day, beads were washed fourtimes with lysis buffer plus inhibitors and boiled inSDS–polyacrylamide gel electrophoresis (PAGE)sample buffer (50 mM Tris-HCl [pH 6.8], 5% 2-mercaptoethanol, and 10% glycerol, 1% SDS) forloading onto a gel for 10%.

Tumorigenicity in Nude Mice

Exponentially growing cells were trypsinized,washed, and resuspended in RPMI medium at aconcentration of 1.2�107 cells/150 mL. Cell suspen-sions were injected subcutaneously into the left flankof 4- to 6-wk-old female athymic nude mice (Balb/c,Nu/Nu mice bred in the Laboratory Animal Unit,Central South University). Mice were maintained ina pathogen-free environment and monitored dailyfor growth at the injection site (three mice per cellline). Once tumors were established, tumor volumemeasurements were taken every 6 d with calipersalong two major axes. Tumor volume was calculatedas follows: 0.5� (long dimension)� (short dimen-sion)2. At the end of the experiments, tumors were

excised. The Committee on the Use of Live Animalsin Teaching and Research at Central South Univer-sity in Changsha, China, approved of this protocol.The data shown are mean values of at least threedifferent experiments and are expressed asmean� SD.

RNA Extraction and Semiquantitative RT-PCR

Total RNA was extracted by a single step methodwith TRIzol reagent (Life Technologies, Gaithers-burg, MD). Reverse transcription (RT)- polymerasechain reaction (PCR) was conducted with a SUPER-SCRIPT One-Step RT-PCR system (Life Technology).Initially, cDNA was generated at 508C for 30 min.PCR was then conducted, and the optimal cyclenumber for quantification (24 cycles) was deter-mined. Each amplification cycle consisted of 0.5 minat 948C for denaturation, 0.4 min at 548C for primerannealing, and 1 min at 728C for extension. Thesequences of PCR primers product used are asfollows: cyclin D1 sense primer, 50-CCC TCG GTGTCC TAC TTC AAA-30; cyclin D1 antisense primer, 50-CAC CTC CTC CTC CTC CTC TTC; b-actin senseprimer, 50-CCA GGC ACC AGG GCG TGA TG-30; andb-actin antisense primer, 50-CGG CCA GCC AGGTCC AGA CG-30 [25]. The sizes of the amplicons forcyclin D1 and b-actin were 726 and 436 bp,respectively.

Colony-Forming Assay

Different cells (either HNE2-LMP1 or HNE2-LMP1-TAM67) were seeded into 6-well dishes in 2 mL ofmedium and cultured for 14 d to allow colonies toform. Colonies (>50 cells) on dishes were visualizedby staining with 0.0125% crystal violet (w/v in 75%ethanol; Sigma–Aldrich, St. Louis, MO). Platingefficiency was determined as the ratio of the numberof colonies and the number of cells seeded. Each datapoint represents the mean of at least three indepen-dent experiments.

Statistical Analysis

The data shown were mean values of at least threedifferent experiments and expressed as mean� SD.Student’s t-test was used for comparison. P<0.05 isconsidered statistically significant.

RESULTS

Expression of c-jun Mutant in NPC Cells

To investigate the role of the AP-1 transcriptionfactors in LMP1-positive NPC cell growth, the HNE2-LMP1-TAM67 cell line which can stably express thedominant-negative mutant of c-jun (TAM67) wascreated and the HNE2-LMP1 cell line was establishedas control. TAM67 lacks the transactivation domain(Fig. 1a) and inhibits both c-jun-induced transactiva-tion and oncogenic transformation when stablytransfected into cells [26–28]. The expression of

POSITIVE HUMAN NASOPHARYNGEAL CARCINOMA 903

Molecular Carcinogenesis DOI 10.1002/mc

TAM67 was confirmed by Western blotting with thec-jun antibody (sc-44) recognizing the conservedDNA-binding domain of wide-type c-jun (Fig. 1a). Asshown in Figure 1b, the TAM67 encoded dominantnegative form of c-jun was only expressed byHNE2-LMP1-TAM67 cells. The wild-type c-jun wasexpressed in both HNE2-LMP1-TAM67 and HNE-LMP1 cells, while the TAM67 expression resulted inthe downregulation of the endogenous c-jun expres-sion and did not reduce levels of jun B. As illustrat-ed in Figure 1c, TAM67 was able to decrease theinteraction between c-jun and jun B proteins withimmunoprecipitation (IP)-Western blotting. First,an antibody of c-jun was used to immunoprecipitatethe total proteins from the transfectant and controlcells and jun B expression was analyzed with Westernblotting. Meanwhile, c-junwas detected with Westernblotting as a loading control. The data showed thatjun B decreased in the transfectant as compared tothe control cells. On the other hand, we used theantibody of jun B to immunoprecipitate the totalproteins and to analyze the c-jun expression, and junB was detected with Western blotting as a loadingcontrol. The data demonstrated that c-jun alsodecreased in the transfectant as compared to thecontrol cells, suggesting that TAM67 can decreasethe formation of the dimer of c-jun and jun B.

Effect of TAM67 on NPC Cell Proliferation In Vitro

To investigate the effect of TAM67 on NPC cellgrowth, we tested the effect of TAM67 on cell mor-phology. As seen in Figure 2a, HNE2-LMP1-TAM67cells partially lost contact inhibition and showedround-up morphology. In addition, we compared aTAM67 transfectant (clone 1) with the control withMTT (Fig. 2b). We determined that TAM67 inhibitedNPC cell growth under different conditions of serum.When the growth rates of the TAM67 transfectantand the control were compared in medium contain-ing 10%, 5%, or 1% FCS, the expression of TAM67decreased the growth rate of NPC cells. On day 4, asshown in Figure 2b, the proliferation rates of clone 1in medium containing 10%, 5%, or 1% FCS weresuppressed significantly to 20%, 34%, and 45%,respectively, compared with HNE2-LMP1 cells (P<0.05). We also compared another clone (clone 2)with the control with MTT and a similar result wasobtained (data not shown). The analysis of HNE2-LMP1-TAM67 cells and control cells in mediumcontaining 10%, 5%, or 1% FCS revealed doublingtimes (meanþ SD) of 47.1þ1.4, 108.7þ 7.8, 192.3þ12.5 h versus 36.3þ1.8, 57.2þ 4.6, 120.5þ6.2 h,respectively (Fig. 2c, P<0.05). The comparisonbetween clone 2 and the control produced a similarresult to that between clone 1 and the control (Fig.2c, P<0.05). The results indicate that HNE2-LMP1cells expressing TAM67 showed a significantly pro-longed doubling times compared with control cells.

Figure 1. Expression of c-jun mutants in HNE2-LMP1-TAM67 cells.Two cell lines (clone 1 and clone 2) expressing a stable dominant-negative c-jun construct, HNE2-LMP1-TAM67, were created, asdetailed in Materials and Method Section. (a) Graphical representationof c-jun mutant and antibodies of c-jun or jun B utilized in theseinvestigations. (b) Immunoblotting for wild-type c-jun and TAM67was performed with protein extracts from HNE2-LMP1-TAM67 cellsresolved on 12.5% SDS–PAGE gels and probed with an anti-c-junantibody (sc-44). (c) IP-Western blot to demonstrate TAM67 was ableto inhibit the interaction between c-jun and jun B. The nuclear proteinof HNE2-LMP1 and HNE2-LMP1-TAM67 clone 1 cells was extracted.(A) Precleared, the extracts were immunoprecipitated with 5 mg ofc-jun antibody and analyzed by immunoblotting with jun B antibodies.Blots were stripped and reprobed with anti-c-jun antibody as a loadingcontrol. (B) Precleared, the nuclear extracts were immunoprecipitatedwith 5 mg of jun B antibody and analyzed by Western blotting withc-jun antibodies. Blots were stripped and reprobed with anti-jun Bantibody as a loading control.

904 JIN ET AL.

Molecular Carcinogenesis DOI 10.1002/mc

Expression of TAM67 Inhibits ColonyFormation of NPC Cells

The effect of TAM67 on the colony growth of NPCcells was assessed. As shown in Figure 3a, significantreductions in cloning efficiency were observed inclone 1 and clone 2 cells as compared with controlcells. The quantitation of data demonstrated that thestable transfection of TAM67 inhibited the colonyformation ability of HNE2-LMP1 cells by �83% and80% compared with control cells, respectively,(Fig. 3b). The results suggest that TAM67 can inhibitthe colony formation of NPC cells.

TAM67 Suppresses the Growth of XenograftedTumors in Nude Mice

To assess the effect of TAM67 on xenograftedtumor growth, we utilized one group of HNE2-LMP1-

TAM67 cells (clone 1). These cells were subcuta-neously injected intothenudemice,andtumordeve-lopment was monitored. As illustrated in Figure 4b,large tumors developed very rapidly in mice injectedwith HNE2-LMP1 cells, consistent with the pre-viously documented behavior of LMP1 in this assay.In contrast, clone 1, which overexpressed the TAM67protein, gave rise to only very small tumors. On days26 and 32, tumor volume was reduced by �70% and�78% in the TAM67 expressing xenografts com-pared with HNE2-LMP1 group. The mice xeno-grafted with TAM67 transfectants and HNE2-LMP1cells were killed (Fig. 4a) on day 38 after injection andthe weight of all tumors was determined (when theexperiment was terminated because of the formationof large tumors in the control group).

As shown in Figure 4c, the TAM67 expressingxenografts caused significant suppression of tumorweight, reaching only 35% of that of the controltumors by day 38. This reduction in tumor growthsuggests that TAM67 significantly suppresses thegrowth of xenografted LMP1-positive NPC tumors.

Characteristics of AP-1 Activity inHNE2-LMP1-TAM67 Cells

In our previous study, we showed that LMP1induced high AP-1 activity. To determine whetherTAM67 inhibits AP-1 activity in LMP1-positive NPCcells, we performed transfection with an AP-1reporter vector system containing synthetic AP-1

Figure 2. The effects of TAM67 on the growth rate of NPC cellsin vitro. (a) Photographs illustrating the morphology of the three celllines (10�). (b) TAM67 transfectant and control cells were cultured indifferent serum. MTT experiments were detailed in Materials andMethods Section. Cells (2�104) were plated into each well, eachcondition was run in triplicate, and the graphs represent the meansand standard deviations. * Indicates statistical significant difference.(P< 0.05 vs. HNE2-LMP1 cells.) (c) Doubling time for all cells indifferent serum was calculated * Indicates statistical significantdifference. (P< 0.05 vs. HNE2-LMP1 cells.)

Figure 3. Stable TAM67 expression inhibited colony formation oftransfected HNE2-LMP1 cells. (a) Cells (1� 103) were seeded intoeach well, colonies were visualized and counted after staining withcrystal violet. (b) Colonies were counted, and the actual number ofcolony/plate was plotted (mean� SD of three separate experiments,each performed in three 6-replicate wells). * Indicates statisticalsignificant difference. (P< 0.05 vs. HNE2-LMP1 cells.)

POSITIVE HUMAN NASOPHARYNGEAL CARCINOMA 905

Molecular Carcinogenesis DOI 10.1002/mc

DNA-binding sites, and the b-galactosidase reporterconstruct was cotransfected as an internal control.The results showed that LMP1-induced AP-1 tran-scriptional activity was significantly lower in HNE2-LMP1-TAM67 cells (clone 1 and clone 2) than incontrol cells (Fig. 5).

TAM67 Arrests Cells in G0/G1 Phase

A flow cytometric analysis was performed toevaluate the underlying mechanism of inhibitionin NPC cell proliferation by TAM67. Figure 6a depictsrepresentative histograms for cell cycle distributionin HNE2-LMP1-TAM67 cells and HNE2-LMP1 cells.As can be seen in Figure 6b, TAM67 transfectantswere blocked in cell cycle progression and accumu-lated in G0/G1 phase, and �61.8% and 59.2% of thecells were arrested in G0/G1 phase, whereas thepercentage of cells in G0/G1 phase was �31.6% in

HNE2-LMP1 cells. Flow cytometry showed that theexpression of TAM67 cells reduced the proportion ofcells in S and G2/M phases, while increasing theproportion of cells in the G0/G1 phase the overalldata revealed that the expression of TAM67 blocksthe cell cycle by causing a G1 arrest.

Effects of TAM67 on the Expression of Cell Cycle Proteins

The cell cycle is regulated by cyclins and cdks. Tofurther investigate the effect of TAM67 on the cellcycle, we analyzed levels of cyclins and cdks in theTAM67 transfectant and in control cells throughimmunoblotting (Fig. 7a). We found the amountsof cyclin E and cdk2 were essentially unchangedbetween cells infected with TAM67 and control cells.The expression of cyclin D1 and cdk4 protein wasdownregulated in HNE2-LMP1-TAM67 cells. Thequantitation of data showed that cyclin D1decreased by up to �60% and cdk4 decreased by upto �40% compared with HNE2-LMP1 cells (Fig. 7c).We also tested whether TAM67 increased theexpression of a variety of cyclin–cdk inhibitors. Wechose two CDK inhibitors, p21Cip1 and p27Kip1,which bind and inhibit a wide range of cyclin/cdkcomplexes, especially the cdk2 complex. The proteinlevels of p21Cip1 and p27Kip1 were similar betweencells. The results suggest that TAM67 might have aneffect on the cyclin D1/CDk4 complex, but it haslittle effect on the cyclin E/cdk2 complex.

Figure 4. Tumorigenicity in nude mice of TAM67-overexpressingcells. For HNE2-LMP1 and clone 1 cell lines, six nude mice wereinjected subcutaneously with 1.2� 107 cells suspended in 150 mL ofPBS. (a) Shown are photographs of tumors in each group beforenecropsy at day 38 postinjection, the bar indicates 10 mm. (b) Tumorsize was measured in two dimensions and tumor volumes werecalculated as described above. Each value represents the meanþ SD(n¼ 3). * Indicates statistical significant difference. (P< 0.05 vs.HNE2-LMP1 cells.) (c) Tumor weights for all mice in each set weremeasured at day 38. Shown are the means and standard deviationsfor the tumors derived from each cell line. * Indicates statisticalsignificant difference. (P< 0.05 vs. HNE2-LMP1 cells.)

Figure 5. Inhibition of AP-1 activity in TAM67 expression cells.Transcriptional repression by TAM67 in the stable cell lines. HNE2-LMP1 and HNE2-LMP1-TAM67 were cotransfected with 0.5 mg AP-1-Luc and 0.5 mg of b-galactosidase reporter plasmids. Values werenormalized to b-galactosidase activity, which reflects transfectionefficiency. The graphs represent the means and standard deviationsfrom three individual experiments. * Indicates statistical significantdifference. (P< 0.05 vs. HNE2-LMP1 cells.)

906 JIN ET AL.

Molecular Carcinogenesis DOI 10.1002/mc

It is well known that the retinoblastoma tumorsuppressor protein Rb and the transcription factorE2F are the key regulators of the cell cycle. Wedetected the Rb phosphorylation status in differentcell lines. Western blotting analysis revealed thathyperphorylation was present in LMP1-positive NPCcells in a few Rb proteins. In contrast, not much Rbprotein was hyperphorylated in the TAM67 transfec-tant. On the other hand, the protein expression ofE2F1 decreased in the TAM67 transfectant comparedwith the control cells. The quantitation of datashowed that E2F1 decreased by up to �30% com-pared with the control cells (Fig. 7c). Rb binds manyproteins including E2F1, thus preventing cell cycleprogression towards the S phase. The change in theRb phosphorylation status will lead to different E2Factivity. More ‘‘free’’ E2F1 factors should be releasedfrom Rb/E2F complex when Rb protein was hyper-phorylated, thus the E2F1 transactivity wouldbe increased. E2F-1 reporter vector system andb-galactosidase reporter constructs were cotrans-fected in different cells; the results demonstrate thatTAM67 can inhibit E2F1 transactivity (Fig. 7b).

The observed hypophosphorylation of Rb inducedby TAM67 suggests a defect in the activation ofessential G1 cdks. Accordingly, the activities of cdk2and cdk4 were measured. The primary substrates ofcdk4 and cdk2 in G1 progression are the members ofthe retinoblastoma protein—Rb. Rb contains at least16 consensus sequences for cdk phosphorylation.The phosphorylation of Rb threonine 826 (phos-phor-RbThr826) can be catalyzed only by cdk4, andthe phosphorylation of Rb threonine 821 (phosphor-RbThr821) can be catalyzed only by cdk2 [29–31]. Toobtain further insights into cdk–cyclin complex

activity, the levels of Rb threonine 826 and 821phosphorylation were assayed (Fig. 8a). Quantita-tion of data showed that the levels of Rb threonine826 phosphorylation decreased by up to �60%compared with HNE2-LMP1 cells whereas the pro-tein levels of Rb threonine 821 phosphorylation weresimilar in different cells (Fig. 8b). Thus, the levels ofRb threonine 826 phosphorylation were decreased inTAM67 transfectant cells.

Expression of TAM67 Inhibits Cyclin D1 Expression

To investigate whether the dimer of c-jun and jun Bis directly involved in transcriptional regulation ofcyclin D1, we analyzed the activation of transfectedreporter promoter constructs for cyclin D1 (CCND1-luc). In our previous study, we showed that LMP1activated the cyclin D1 promoter and increased theexpression of cyclin D1 protein [32]. When theinteraction of c-jun and jun B protein was reduced byTAM67, the activity of the cyclin D1 promotersignificantly decreased in contrast to the LMP1-positive NPC cells (Fig. 9a). Subsequently, RT-PCRassays were conducted to detect cyclin D1 mRNA inTAM67 cells and in the control cells. Figure 9b showsthat a single band of 726 bp corresponding to thecoding region of cyclin D1 was readily detected inHNE2-LMP1 cells. In contrast, this band was nearlyundetectable in HNE2-LMP1-TAM67 cells thatexpress TAM67. Quantitation of data showed thatthe levels of cyclin D1 mRNA decreased by up to�67% compared with HNE2-LMP1 cells, demon-strating that cyclin D1 transcription in TAM67-expressing cells was strongly inhibited. In combina-tion, these experiments indicate that AP-1 can

Figure 6. TAM67 arrests NPC cell progress in the G0/G1 phase. (a) Representative histograms depicting cell cycledistribution in either HNE2-LMP1 or HNE2-LMP1-TAM67 cells. Cells were cultured in 10% FCS for 36 h and thenstained with propidium iodide and subjected to cytoflorimetric analysis. (b) Results shown are mean� SD fromthree independent experiments. The number of cells in the G0/G1, S, and G2/M phases are indicated.

POSITIVE HUMAN NASOPHARYNGEAL CARCINOMA 907

Molecular Carcinogenesis DOI 10.1002/mc

specifically affect the regulation of cyclin D1 expres-sion from a transcriptional level.

DISCUSSION

NPC is the human tumor displaying the mostconsistent association with the EBV gene-LMP1 [33].Although well studied in LMP1 and B lymphocytes,there is still a lack of knowledge about the effects ofLMP1 on NPC. Previous studies have demonstratedthat the AP-1 transcription factor is an importantregulator of proliferation, transformation, and apop-tosis in many cell types. In these reports, accumulat-ing evidence suggests that AP-1 is associated with thepromotion of colon and breast cancers [34–36].Interestingly, some publications report an importantrole for AP-1 in mediating metastasis for NPC [37,38].However, the biological and physiological functionsof AP-1 in relation to the growth of NPC are not fullyunderstood.

Our previous study showed that LMP1 triggeredAP-1 via the c-jun N-terminal kinase cascade. Afurther study showed that LMP1 induced high basallevels of AP-1 transcriptional activity in contrast toLMP1-negative NPC cells and mediated a primarydimer form of c-jun and jun B. These findings areimportant for understanding the role of AP-1 inmediating the proliferation of NPC.

To demonstrate the role of AP-1 in NPC, weinvestigated whether we could block AP-1 activitywith the c-junmutant (TAM67) acting as a dominant-negative mutant. Our results showed that theexpression of TAM67 was accompanied by thedownregulation of endogenous c-jun which is con-sistent with other findings [39], presumably as c-juncan be autoregulated through its AP-1 sites. The exactmolecular action of TAM67 has two possiblemechanisms. The first mechanism is ‘‘quenching’’or ‘‘squelching’’ the formation of inactive hetero-

Figure 7. G1 arrested by TAM67 affects G1 cyclin-cdk complexcomponents. (a) Western blotting analysis of cell cycle proteins inHNE2-LMP1, TAM67 expressing cell lines. Equivalent amounts oftotal-cell extracts were visualized with indicated antibodies. (b) Effectof TAM67 on the transactivity of E2F1. Cultures were transientlycotransfected with E2F1-luc and b-galactosidase reporter plasmids.Values were normalized to b-galactosidase activity, which reflects

transfection efficiency. The graphs represent the means andstandard deviations from three individual experiments. (c) Aquantification of the western blots was assessed. Blots werequantified by scanning densitometry and normalized with theexpression of a-tubulin as the control. * Indicates statisticalsignificant difference. (P< 0.05 vs. HNE2-LMP1 cells.)

908 JIN ET AL.

Molecular Carcinogenesis DOI 10.1002/mc

dimers between TAM67 and AP-1 proteins [40,41]. Asecond mechanism is ‘‘blocking’’ in which TAM67functions through the formation of homodimers[40,41]. IP experiments suggested that TAM67 coulddecrease the interaction of c-jun and jun B. Whileanother possibility of blocking AP-1 activity lies inthat TAM67 decrease the level of c-jun expression,thereby leading to the inhibited availability to bindjun B or TPA elements.In keratinocytes, TAM67inhibits both AP-1 and NF kappa B activity[27,50,51]; we also show LMP-1 not only activatesAP-1 signaling, but also NF-kappa B signaling in EBV-infected NPC cell lines [52]. While the transactiva-tion of NF-kappa B induced by LMP1 was not beeninhibited after expression of TAM67 [22]. Futurestudy should be done about the crosstalk betweenAP-1 signaling and NF-kappa B signaling.

We have analyzed the molecular basis for growth-inhibiting effects by TAM67. It is well known thatduring G1, there are two active cyclin–cdk com-plexes, comprising D-type cyclins and cdks, and thatcyclin D1/cdk4 and cyclin E/cdk2 complexes areconstitutively active throughout the early G1 phaseof dividing cells. In contrast, cyclin E-cdk2 com-plexes are activated with the transition across the lateG1 restriction [42]. These results suggested that theAP-1 blockade caused a reduced expression of cyclinD1 and cdk4, the main regulator in the early G1 phaseof the cell cycle, whereas TAM67 expression did notreduce levels of cyclin E and cdk2. Similarly, wefound that TAM67 had little effect on p21Cip1 and

p27Kip1 levels. Furthermore, cdk4 and cdk2 activitieswere measured, and we reported that cdk4 activitywas a deregulation of TAM67 transfectant cells.Conversely, the activity of cdk2 was not altered byTAM67 expression. The activities of cdk4 and cdk2were assessed by the phosphorylation of Rb threo-nine 826 and 821, which is consistent with otherstudies [43,44]. The abnormal expression patterns ofcyclins, cdks, and cdkI may lead to alterations inhypophosphorylated form of Rb (pRb) proteinexpression. Retinoblastoma tumor suppressor (Rb)is a well-known regulator of the cell cycle. The pRbprevents cell cycle progression toward the S phase byinteracting with many proteins, including E2F1 [45].In our study, Western blotting analysis showed thatthe AP-1 blockade caused the hypophosphorylation

Figure 8. Phosphorylation of Rb was analyzed in HNE2-LMP1 andHNE2-LMP1-TAM67 cells. Extracts were prepared from HNE2-LMP1and TAM67 expressing cell lines. (a) Immunoblotting was performedwith antibodies to Rb threonine 821 phosphorylation and Rbthreonine 826 phosphorylation. (b) A quantification of the Westernblots was assessed. Blots were quantified by scanning densitometryand normalized with the expression of a-tubulin as the control. *Indicates statistical significant difference. (P<0.05 vs. HNE2-LMP1cells.)

Figure 9. The inhibitory effect of TAM67 on the transcription ofcyclin D1 gene. (a) Cyclin D1 promoter analysis. Cyclin D1 promoterluciferase reporter CCND1 were transiently transfected into theHNE2-LMP1-TAM67 and HNE2-LMP1 cells, along with b-galactosi-dase reporter plasmids as an internal control. Forty hours aftertransfection, cells were harvested and luciferase activities weredetermined. Values were normalized to b-galactosidase activity,which reflects transfection efficiency. The graphs represent themeans and standard deviation from three individual experiments.* Indicates statistical significant difference. (P< 0.05 vs. HNE2-LMP1cells.) (b) RT-PCR analysis to b-actin mRNA and cyclin D1 mRNAexpression. Bands were quantified by scanning densitometry. Resultsare shown from three independent experiments and b-actin hasbeen included as a loading control, all bands have been normalizedby b-actin.

POSITIVE HUMAN NASOPHARYNGEAL CARCINOMA 909

Molecular Carcinogenesis DOI 10.1002/mc

of Rb and reduced the expression of E2F1, which inturn inhibited E2F transactivity, ultimately inducinga G1 cell cycle blockage. TAM67 inhibition of NPCcell growth was attributed to the deregulation ofcyclin D1/cdk4 activity and the subsequent cell cyclearrest in the G1 phase. Similar to other studies, alinkage existed between AP-1 and G1 restriction inthe regulation of mitogenic cell progression [46].Therefore, the present study provides further evi-dence that AP-1-dependent cyclin D1/cdk4 activityupregulation participates in the LMP1-induced pro-liferation of NPC.

Results from our previous studies suggested thatcyclin D1 transcription is activated by LMP1; how-ever, little is known about the molecular mechanisminvolved. Because the cyclin D1 promoter contains atypical AP-1-binding site [47,48], the direct effect ofAP-1 transactivation on cyclin D1 transcription wasdemonstrated in HNE2-LMP1 cells that express adominant-negative mutant of c-jun, TAM67. Wecompared the levels of cyclin D1 mRNA andpromoter activity in stable clones of TAM67 withcontrol cells. The results showed that the expressionsof cyclin D1 mRNA and the promoter activity inTAM67 clones were lower than those in controlHNE2-LMP1 cells, demonstrating that the blockadeof AP-1 transcriptional activity has a negative impacton cyclin D1 transcription.

Another perspective of this work is the opening ofa field of investigation for the search of drugsdirected against LMP1-positive NPC. During theprocess of several human diseases (including cancer),protein–protein interactions are usually altered [49]and dominant-negative mutants can bind and affecta particular target. In the present study as well as inour previous studies [20,21], we showed that LMP1mediated high AP-1 activity the blockade of whichby TAM67 abrogated the oncogenic phenotype andgrowth of NPC. Our work provides not only a newinsight into the molecular mechanism underlyingLMP1-induced NPC proliferation but also a newtherapeutic approach for targeting AP-1 in LMP1-positive NPC.

ACKNOWLEDGMENTS

This work was supported by the DevelopmentProject of National Critical Basic Research (973)‘‘Basic Research of Tumorigenesis and MalignantDevelopment’’ (No. 2004CB518703); the NationalNature Science Foundation of China (No.30300403); and the National Nature Science Foun-dation for Distinguished Young Scholars of China(No. 39525022).

References

1. Rickinson AB. Epstein-Barr virus in action in vivo. N Engl J Med1998;338:1461–1463.

2. Brooks L, Yao QY, Rickinson AB, Young LS. Epstein-Barr viruslatent gene transcription in nasopharyngeal carcinoma cells:Coexpression of EBNA1, LMP1, and LMP2 transcripts. J Virol1992;66:2689–2697.

3. Fahraeus R, Fu HL, Ernberg I, et al. Expression of Epstein-Barrvirus-encoded proteins in nasopharyngeal carcinoma. Int JCancer 1988;42:329–338.

4. Young LS, Dawson CW, Clark D, et al. Epstein-Barr virus geneexpression in nasopharyngeal carcinoma. J Gen Virol 1988;69:1051–1065.

5. Hennessy K, Fennewald S, Hummel M, Cole T, Kieff E. Amembrane protein encoded by Epstein-Barr virus in latentgrowth-transforming infection. Proc Natl Acad Sci USA1984;81:7207–7211.

6. Huen DS, Henderson SA, Croom-Carter D, Rowe M. TheEpstein-Barr virus latent membrane protein-1 (LMP1) med-iates activation of NF-kappa B and cell surface phenotype viatwo effector regions in its carboxy-terminal cytoplasmicdomain. Oncogene 1995;10:549–560.

7. Izumi KM, Kaye KM, Kieff ED. The Epstein-Barr virus LMP1amino acid sequence that engages tumor necrosis factorreceptor associated factors is critical for primary B lympho-cyte growth transformation. Proc Natl Acad Sci USA 1997;94:1447–1452.

8. Behrens A, Jochum W, Sibilia M, Wagner EF. Oncogenictransformation by ras and fos is mediated by c-Jun N-terminalphosphorylation. Oncogene 2000;19:2657–2663.

9. Maslyar DJ, Aoki M, Vogt PK. The growth-promoting activityof the Bad protein in chicken embryo fibroblasts requiresbinding to protein 14-3-3. Oncogene 2001;20:5087–5092.

10. Wisdom R, Johnson RS, Moore C. c-Jun regulates cell cycleprogression and apoptosis by distinct mechanisms. EMBO J1999;18:188–197.

11. Smith LM, Wise SC, Hendricks DT, et al. c-Jun overexpressionin MCF-7 breast cancer cells produces a tumorigenic, invasiveand hormone resistant phenotype. Oncogene 1999;18:6063–6070.

12. Bossy-Wetzel E, Bakiri L, Yaniv M. Induction of apoptosis bythe transcription factor c-Jun. EMBO J 1997;16:1695–1709.

13. Wang N, Verna L, Hardy S, et al. c-Jun triggers apoptosis inhuman vascular endothelial cells. Circ Res 1999;85:387–393.

14. Passegue E, Wagner EF. JunB suppresses cell proliferation bytranscriptional activation of p16(INK4a) expression. EMBO J2000;19:2969–2979.

15. Jacobs-Helber SM, Abutin RM, Tian C, Bondurant M,Wickrema A, Sawyer ST. Role of JunB in erythroiddifferentiation. J Biol Chem 2002;277:4859–4866.

16. Chalaux E, Lopez-Rovira T, Rosa JL, Bartrons R, Ventura F.JunB is involved in the inhibition of myogenic differentiationby bone morphogenetic protein-2. J Biol Chem 1998;273:537–543.

17. Mao X, Orchard G, Lillington DM, Russell-Jones R, Young BD,Whittaker SJ. Amplification and overexpression of JUNB isassociated with primary cutaneous T-cell lymphomas. Blood2003;101:1513–1519.

18. Robinson CM, Prime SS, Huntley S, et al. Overexpression ofJunB in undifferentiated malignant rat oral keratinocytesenhances the malignant phenotype in vitro without alteringcellular differentiation. Int J Cancer 2001;91:625–630.

19. Schlingensiepen KH, Schlingensiepen R, Kunst M, et al.Opposite functions of jun-B and c-jun in growth regulationand neuronal differentiation. Dev Genet 1993;14:305–312.

20. Wang C, Deng X, Li X, et al. Matrix metalloproteinase 9expression is induced by Epstein-Barr virus LMP1 via NF-kappa B or AP-1 signaling pathway in nasopharyngealcarcinoma cells. Zhonghua Zhong Liu Za Zhi 2002;24:9–13.

21. Song X, Tao YG, Deng XY, et al. Heterodimer formationbetween c-Jun and Jun B proteins mediated by Epstein-Barrvirus encoded latent membrane protein 1. Cell Signal2004;16:1153–1162.

910 JIN ET AL.

Molecular Carcinogenesis DOI 10.1002/mc

22. Deng L, Yang J, Zhao XR, et al. Cells in G2/M phase increasedin human nasopharyngeal carcinoma cell line by EBV-LMP1through activation of NF-kappaB and AP-1. Cell Res 2003;3:187–194.

23. Hinz M, Krappmann D, Eichten A, Heder A, Scheidereit C,Strauss M. NF-kappaB function in growth control: Regulationof cyclin D1 expression and G0/G1-to-S-phase transition.Mol Cell Biol 1999;19:2690–2698.

24. Rotheneder H, Geymayer S, Haidweger E. Transcriptionfactors of the Sp1 family: Interaction with E2F and regulationof the murine thymidine kinase promoter. J Mol Biol 1999;293:1005–1015.

25. Masuda M, Suzui M, Yasumatu R, et al. Constitutiveactivation of signal transducers and activators of transcrip-tion 3 correlates with cyclin D1 overexpression and mayprovide a novel prognostic marker in head and necksquamous cell carcinoma. Cancer Res 2002;62:3351–3355.

26. Dong Z, Birrer MJ, Watts RG, Matrisian LM, Colburn NH.Blocking of tumor promoter-induced AP-1 activity inhibitsinduced transformation in JB6 mouse epidermal cells. ProcNatl Acad Sci USA 1994;91:609–613.

27. Dong Z, Crawford HC, Lavrovsky V, et al. A dominantnegative mutant of jun blocking 12-O-tetradecanoylphor-bol-13-acetate-induced invasion in mouse keratinocytes.Mol Carcinog 1997;19:204–212.

28. Cooper SJ, MacGowan J, Ranger-Moore J, Young MR,Colburn NH, Bowden GT. Expression of dominant negativec-jun inhibits ultraviolet B-induced squamous cell carcinomanumber and size in an SKH-1 hairless mouse model. MolCancer Res 2003;1:848–854.

29. Knudsen ES, Wang JY. Differential regulation of retinoblas-toma protein function by specific Cdk phosphorylation sites.J Biol Chem 1996;271:8313–8320.

30. Knudsen ES, Wang JY. Dual mechanisms for the inhibition ofE2F binding to RB by cyclin-dependent kinase-mediated RBphosphorylation. Mol Cell Biol 1997;17:5771–5783.

31. Kaelin WG Jr. Functions of the retinoblastoma protein.Bioessays 1999;21:950–958.

32. Zhao XR, Gu HH, Weng XX, Yi W, Deng XY, Cao Y. Theprimary study on expression and function of D-type cyclins innasopharyngeal carcinoma cell lines. Sheng Wu Hua Xue YuSheng Wu Wu Li Xue Bao (Shanghai) 2000;32:192–196.

33. Hu LF, Chen F, Zheng X, et al. Clonability and tumorigenicityof human epithelial cells expressing the EBV encodedmembrane protein LMP1. Oncogene 1993;8:1575–1583.

34. Liu Y, Lu C, Shen Q, Munoz-Medellin D, Kim H, Brown PH.AP-1 blockade in breast cancer cells causes cell cycle arrest bysuppressing G1 cyclin expression and reducing cyclin-dependent kinase activity. Oncogene 2004;23:8238–8246.

35. Suto R, Tominaga K, Mizuguchi H, et al. Dominant-negativemutant of c-Jun gene transfer: A novel therapeutic strategyfor colorectal cancer. Gene Ther 2004;11:187–193.

36. Liu Y, Ludes-Meyers J, Zhang Y, et al. Inhibition of AP-1transcription factor causes blockade of multiple signaltransduction pathways and inhibits breast cancer growth.Oncogene 2002;21:7680–7689.

37. Kondo S, Wakisaka N, Schell MJ, et al. Epstein-Barr viruslatent membrane protein 1 induces the matrix metallopro-teinase-1 promoter via an Ets binding site formed by a single

nucleotide polymorphism: Enhanced susceptibility to naso-pharyngeal carcinoma. Int J Cancer 2005;115:368–376.

38. Murono S, Yoshizaki T, Sato H, Takeshita H, Furukawa M,Pagano JS. Aspirin inhibits tumor cell invasiveness induced byEpstein-Barr virus latent membrane protein 1 throughsuppression of matrix metalloproteinase-9 expression. Can-cer Res 2000;60:2555–2561.

39. Kielosto M, Nummela P, Katainen R, Leaner V, Birrer MJ,Holtta E. Reversible regulation of the transformed phenotypeof ornithine decarboxylase- and ras-overexpressing cells bydominant-negative mutants of c-Jun. Cancer Res 2004;64:3772–3779.

40. Brown PH, Kim SH, Wise SC, Sabichi AL, Birrer MJ. Dominant-negative mutants of c-Jun inhibit AP-1 activity throughmultiple mechanisms and with different potencies. CellGrowth Differ 1996;7:1013–1021.

41. Thompson EJ, Gupta A, Stratton MS, Bowden GT. Mechan-ism of action of a dominant negative c-jun mutant ininhibiting activator protein-1 activation. Mol Carcinog 2002;35:157–162.

42. Ezhevsky SA, Ho A, Becker-Hapak M, Davis PK, Dowdy SF.Differential regulation of retinoblastoma tumor suppressorprotein by G(1) cyclin-dependent kinase complexes in vivo.Mol Cell Biol 2001;21:4773–4784.

43. Glassford J, Holman M, Banerji L, et al. Vav is required forcyclin D2 induction and proliferation of mouse B lymphocytesactivated via the antigen Receptor. J Biol Chem 2001;27:41040–41048.

44. Parada Y, Banerji L, Glassford J, et al. BCR-ABL andinterleukin 3 promote haematopoietic cell proliferation andsurvival through modulation of cyclin D2 and p27Kip1expression. J Biol Chem 2001;276:23572–23580.

45. Bremner R, Cohen BL, Sopta M, et al. Direct transcriptionalrepression by pRB and its reversal by specific cyclins. Mol CellBiol 1995;15:256–265.

46. Hennigan RF, Stambrook PJ. Dominant negative c-juninhibits activation of the cyclin D1 and cyclin E kinasecomplexes. Mol Biol Cell 2001;12:2352–2363.

47. Herber B, Truss M, Beato M. Increased expression of cyclin D1is an early event in multistage colorectal carcinogenesis.Oncogene 1994;9:2105–2107.

48. Arber N, Hibshoosh H, Moss SF, et al. Increased expression ofcyclin D1 is an early event in multistage colorectalcarcinogenesis. Gastroenterology 1996;110:69–674.

49. Blum JH, Dove SL, Hochschild A, Mekalanos JJ. Isolation ofpeptide aptamers that inhibit intracellular processes. ProcNatl Acad Sci USA 2000;97:2241–2246.

50. Cooper S, Ranger-Moore J, Bowden TG. Differential inhibi-tion of UVB-induced AP-1 and NF-kappaB transactivation bycomponents of the jun bZIP domain. Mol Carcinog 2005;43:108–116.

51. Li JJ, Rhim JS, Schlegel R, Vousden KH, Colburn NH.Expression of dominant negative Jun inhibits elevated AP-1and NF-kappaB transactivation and suppresses anchorageindependent growth of HPV immortalized human keratino-cytes. Oncogene 1998;16:2711–2721.

52. Yin L, Liao W, Yi W, et al. LMP1 activates NF-kappa B viadegradation of I kappa B alpha in nasopharyngeal carcinomacells. Chin Med J (Engl) 2001;114:718–722.

POSITIVE HUMAN NASOPHARYNGEAL CARCINOMA 911

Molecular Carcinogenesis DOI 10.1002/mc