[CANCER RESEARCH 60, 2555–2561, May 1, 2000]

Aspirin Inhibits Tumor Cell Invasiveness Induced by Epstein-Barr VirusLatent Membrane Protein 1 through Suppression of MatrixMetalloproteinase-9 Expression1

Shigeyuki Murono, Tomokazu Yoshizaki, Hiroshi Sato, Hajime Takeshita, Mitsuru Furukawa, and Joseph S. Pagano2

Lineberger Comprehensive Cancer Center [S. M., J. S. P.] and Departments of Medicine [J. S. P.] and Microbiology and Immunology [J. S. P.], University of North Carolina atChapel Hill, Chapel Hill, North Carolina 27599-7295, and Department of Otolaryngology, School of Medicine [S. M., T. Y., H. T., M. F.], and Department of Molecular Virologyand Oncology, Cancer Research Institute [H. S.], Kanazawa University, Ishikawa 920-8640, Japan

ABSTRACT

Matrix metalloproteinases (MMPs) are thought to play crucial roles intumor invasion and metastasis. Because we have shown that EBV latentmembrane protein 1 (LMP1) enhances MMP-9 expression by activation ofnuclear factor (NF)-kB and activator protein (AP)-1 (T. Yoshizaki, et al.,Proc. Natl. Acad. Sci. USA,95: 3621–3626, 1998), we therefore testedwhether up-regulation of MMP-9 by LMP1 could be correlated withenhanced invasiveness of tumor cellsin vitro. Whether aspirin and sodiumsalicylate could reduce invasiveness and whether LMP1 could enhanceMMP-9 expression in tumors grown in nude mice were also tested. C33Acells stably expressing LMP1 had increased expression of MMP-9 andshowed greater invasion through reconstituted basement membrane com-pared with vector-transfected C33A cells (P< 0.02). Treatment withaspirin or sodium salicylate inhibited invasiveness of the LMP1-express-ing C33A cells (P< 0.03) and suppressed both the LMP1-induced MMP-9expression in zymographic analyses and LMP1-induced MMP-9 promoteractivity in CAT reporter assays (P < 0.01). Endogenous MMP-2 levelswere unaffected by either drug. Both drugs repressed the CAT activity ofthe truncated MMP-9 promoter construct, which only contained a bindingsite for AP-1, to the basal level (P< 0.05). Moreover, EMSA indicated thatthe effects of the salicylates were through the inhibition of not only NF-kBbut also AP-1 binding activity. Inhibitory effect of salicylates could bereversed by p50/p65 subunits of NF-kB or c-Jun overexpression. Theinhibitory effect of aspirin on NF- kB activity was attributable to theinhibition of I kB kinase activity. Finally, tumors derived from C33A cellsstably expressing LMP1 grown in nude mice showed enhanced MMP-9levels compared with tumors derived from vector-transfected C33A cells.This enhancement was inhibited by treatment of the mice with aspirin.These results suggest that aspirin may be able to suppress invasion andmetastasis of EBV-associated tumors that express LMP1 by suppressionof MMP-9.

INTRODUCTION

MMPs3 play a crucial role in degradation of extracellular matrixassociated not only with normal tissue remodeling but also withpathological conditions such as tumor invasion and metastasis. Inva-sion into surrounding tissues and metastasis to distant organs arecharacteristics of many malignant tumors. Degradation of the base-ment membrane is the critical step in the process of both invasion andmetastasis. Among human MMPs reported previously, MMP-2 (gel-

atinase A/Mr 72,000 type IV collagenase) and MMP-9 (gelatinaseB/Mr 92,000 type IV collagenase) are thought to be key enzymes fordegrading type IV collagen, which is a major component of thebasement membrane. Contributions of both enzymes to invasion andmetastasis have been documented in numerous reports (1, 2).

EBV is a ubiquitous human herpesvirus and the causative agent ofinfectious mononucleosis. EBV is also associated with several malig-nancies such as Burkitt’s lymphoma, Hodgkin’s disease (3), non-Hodgkin’s lymphoma in immunocompromised hosts (4), nasal T/nat-ural killer cell lymphoma (5), and NPC (6, 7). In all of these tumors,EBV infection is predominantly latent. EBV latent infection is dividedinto three types by the pattern of expression of six EBNAs (EBNA1,EBNA2, EBNA3A, EBNA3B, EBNA3C, and EBNA-LP) and threeLMPs (LMP1, LMP2A, and LMP2B). Only EBNA1 is expressed inBurkitt’s lymphoma (type I latency). On the other hand, all sixEBNAs and three LMPs are expressed in non-Hodgkin’s lymphomain immunocompromised hosts (type III latency). LMP1, LMP2A, andLMP2B are expressed in NPC, nasal T/natural killer cell lymphomaand Hodgkin’s disease (type II latency) in addition to EBNA1. Thus,expression of LMP1 is common to type II and type III latency. LMP1is considered an oncoprotein because it can transform rodent fibro-blasts (8). In addition to its oncogenic properties, LMP1 was shownrecently to enhance expression of MMP-9 (9), which might contributeto the invasion and metastasis of EBV-associated tumors especially intype II and type III latency.

Enhanced expression of MMP-9 by LMP1 is through the activationof NF-kB and AP-1 (9). There are two regions that activate NF-kB inthe COOH-terminal region of LMP1 (10), and both contribute to theenhanced expression of MMP-9 (11). NF-kB remains sequestered inthe cytoplasm under unstimulated conditions by tightly bound inhib-itory proteins, the IkBs (12). Many of the signals known to activateNF-kB result in phosphorylation of IkBa on serine residues 32 and 36and subsequent degradation of IkBa, allowing NF-kB to translocateinto the nucleus and activate target genes (13). Acetylsalicylic acid(aspirin) and sodium salicylate have been shown to suppress theNF-kB activation induced by TNF-a, lipopolysaccharide, and inter-leukin 1 in certain systems (14). A recent report demonstrated thataspirin and sodium salicylate inhibited the activity of IkB kinaseb,which is responsible for phosphorylation of IkBa (15). Additionally,aspirin was shown recently to inhibit JNK in several systems (16),which resulted in the inactivation of AP-1. Here we demonstrate thatLMP1 increases invasiveness of tumor cellsin vitro and that bothaspirin and sodium salicylate inhibit this invasiveness by suppressionof the induced constitutive MMP-9 expression. In addition, inhibitionof NF-kB and AP-1 binding to the MMP-9 promoter is involved in thesuppression of MMP-9 by both drugs. Furthermore, LMP1 inducesexpression of MMP-9 in tumors grown in nude mice, which issuppressed by aspirin. These results indicate that both aspirin andsodium salicylate may contribute to reduce invasion and metastasis inEBV-associated tumors.

Received 11/1/99; accepted 3/6/00.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by Grant P01CA19014 from the NIH. S. M. is supported inpart by the Sumitomo Life Enterprise Group for Social Welfare, Osaka, Japan.

2 To whom requests for reprints should be addressed, at Lineberger ComprehensiveCancer Center, University of North Carolina, Chapel Hill, NC 27599-7295. Phone:(919) 966-8644; Fax; (919) 966-9673; E-mail; Joseph_Pagano@med.unc.edu.

3 The abbreviations used are: MMP, matrix metalloproteinase; NPC, nasopharyngealcarcinoma; EBNA, EBV nuclear antigen; LMP, latent membrane protein; AP-1, activatorprotein-1; TNF, tumor necrosis factor; TNFR, TNF receptor; JNK, c-Jun NH2-terminalkinase; FBS, fetal bovine serum; CAT, chloramphenicol acetyltransferase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NF-kB, nuclear factor-kB; GST,glutathione S-transferase; CTAR, COOH-terminal activation region; EMSA, electro-phoretic mobility shift assay; IKK, IkB kinase; TRAF, TNFR-associated factor.

2555

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

MATERIALS AND METHODS

Cell Culture. C33A cells, derived from a human papillomavirus-negativecervical cancer cell line, were obtained from the American Type CultureCollection and cultured in DMEM supplemented with 10% FBS and penicillinand streptomycin.

Plasmids. LMP1 expression plasmid was described previously (9). TheIkBa(SS32/36AA) expression plasmid, the p50 subunit of NF-kB (referred toas p50) expression plasmid, and the p65 subunit of NF-kB (referred to as p65)expression plasmid were kindly provided by Dr. Albert Baldwin (University ofNorth Carolina, Chapel Hill, NC). In the IkBa(SS32/36AA) expression plas-mid, both serines at position 32 and 36 were substituted by alanines, resultingin the prevention of degradation of IkBa and subsequent activation of NF-kB.A series of 59flanking sequences of MMP-9 were inserted upstream of theCAT reporter gene as described (17). A c-Jun expression plasmid was de-scribed previously (18).

Transient and Stable Transfection. C33A cells were transfected with 5mg of appropriate reporter and/or effector plasmids using LipofectAMINEreagent (Life Technologies, Inc., Grand Island, NY), following the manufac-turer’s instructions. To establish either the vector-transfected C33A cells(referred to as C33A/pcDNA3 cells) or the C33A cells stably expressing LMP1(referred to as C33A/LMP1 cells), 0.5mg of pcDNA3 (Invitrogen, Carlsbad,CA) or LMP1 expression plasmid, respectively, was transfected into C33Acells and cultivated in the presence of geneticin (700mg/ml Geneticin; LifeTechnologies).

Reagents.Aspirin, sodium salicylate, and MTT were purchased fromSigma Chemical Co. (St. Louis, MO).

MTT Assay. MTT assays were performed to evaluate the cytotoxicity ofaspirin and sodium salicylate. One hundredml of each cell type (13 105

cells/ml) were plated and incubated for 24 h in wells of a 96-well plate. Thenvarious concentrations of each drug were added to the wells. After another 24 hincubation, 10ml of sterile MTT dye were added, and the cells were incubatedfor 6 h at 37°C. Then 100ml of acidic isopropanol (0.04M HCl in isopropanol)were added and thoroughly mixed. Spectrometric absorbance at 595 nm (forformazan dye) was measured with the absorbance at 655 nm for reference.

Conditioned Media. Eight 3 105 cells were plated and cultured for 48 hin DMEM with 10% FBS. Then cells were incubated for 24 h in DMEM withneither FBS nor antibiotics with or without each drug. Media were harvestedand used for gelatin zymography.

Gelatin Zymography. MMP-2 and MMP-9 enzymatic activity were as-sayed by gelatinolysis by means of gelatin zymography as reported previously(9, 19). For mouse tissues, equal amounts of each tissue were homogenized inSDS sample buffer without reducing agent. The mixture was then incubated at37°C for 30 min, and gelatin zymography was performed. MMP-2 is detectedas the clear band appearing atMr 72,000 and MMP-9 atMr 92,000.

Western Blot Analysis. C33A/pcDNA3 cells and C33A/LMP1 cells(8 3 105 cells) were plated and cultured for 96 h in DMEM with 10% FBSwith or without either drug for the last 24 h. Protein extracts were obtained, andthe concentration was calculated as described previously (20), then mixed with23 SDS sample buffer containing reducing agent. For tumors grown in nudemice, equal amounts of each tumor were homogenized in RIPA buffer [25 mM

Tris-Cl (pH 7.5), 150 mM NaCl, 2 mM EDTA, 1% NP40, 0.5% sodiumdeoxycholate, 0.1% SDS, 0.2 mM sodium orthovanadate, 100 mM NaF, 10mg/ml aprotinin, 5mg/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride];then extracts were obtained as reported (21), and the concentration of proteinwas calculated. After SDS/PAGE, immunoblotting was performed as de-scribed previously (9). Antimouse immunoglobulink chain antibody (1:2000dilution; Southern Biotechnology Associates, Birmingham, AL) was used forthe secondary antibody in analyses of tumors grown in nude mice.

CAT Reporter Assay. CAT assays were performed as described previ-ously (9). Construction of the MMP-9 promoter reporter has been described(Ref. 17; Fig. 1C). Cells were incubated 48 h after transfection with or withoutadding either drug 24 h after transfection and then harvested. Acetylated[14C]chloramphenicol was quantified with a PhosphorImager (Molecular Dy-namics, Sunnyvale, CA). Transfection efficiency was monitored by cotrans-fection with b-galactosidase reporter construct.

EMSA. Nuclear extracts were prepared from C33A/pcDNA3 cells orC33A/LMP1 cells with or without treatment of either drug as reported previ-ously with some modifications (22). Synthetic oligonucleotides used for probes

were identical to the NF-kB or AP-1 binding sequences in the promoter regionof MMP-9 (9). Oligonucleotides were annealed with the use of Klenow DNApolymerase and then labeled with [g-32P]ATP with the use of T4 polynucle-otide kinase. The unlabeled oligonucleotides were used for competition. Fivemg of nuclear extracts were incubated with 50,000 cpm labeled probe andanalyzed by 4.8% PAGE; then autoradiography was performed as reported (9).

IkB Kinase Assay.Assays were performed after transient transfection asdescribed previously (23). Briefly, lysates were obtained from C33A cellstransfected with pcDNA3 or with the LMP1 expression construct with orwithout treatment with 5 mM aspirin; then supernatant fluids were collected,and protein concentration was calculated. Five hundredmg of protein wereincubated with 15mg of IKK2 CT antibody (kindly provided by Drs. AlbertBaldwin and Minnie Holmes-McNary [University of North Carolina, ChapelHill, NC]). Incubation was continued with protein A beads (50ml). Theimmunoprecipitates were then vigorously washed and subjected to anin vitrokinase assay. Kinase assays were done with 2mCi of [g-32P]ATP in thepresence of 2mg of GST-IkBa or GST-IkBa(SS32/36TT) (kindly provided byDrs. Albert Baldwin and Minnie Holmes-McNary) for substrates, followed bySDS-PAGE analysis and autoradiography.

In Vitro Invasion Assay. Assays were performed with the use of BiocoatMatrigel Invasion Chambers (Becton Dickinson Labware, Bedford, MA) asdescribed elsewhere. Biocoat Cell Culture Inserts were used for uncoatedassays. We established two indices to evaluatein vitro invasiveness: percentinvasion (% invasion) and invasion index. Both indices were calculated asfollows: % invasion5 number of cells invading through Matrigel-coatedmembrane/number of cells migrating through uncoated control insert mem-brane. Invasion index5 % invasion of tested cells/% invasion of C33A/pcDNA3 cells.

Injection of Cells into Athymic Mice. Either C33A/pcDNA3 cells orC33A/LMP1 cells (23 106 cells) in a volume of 0.1 ml PBS were s.c. injectedinto athymic nude mice. Four weeks after injection, PBS alone or 9 mg/dayaspirin were given s.c. at a distance from the tumor for 3 days, and then micewere sacrificed. Equal amounts of each tumor were homogenized in SDSsample buffer without reducing agent or RIPA buffer for gelatin zymographyor Western blot analysis, respectively.

Statistical Analysis. Significant differences were determined by the pairedt test.P , 0.05 was considered to be statistically significant.

RESULTS

LMP1 Enhances Invasion of C33A Cellsin Vitro. A MatrigelInvasion Chamber system was used to assay tumor cell invasivenessas described in “Materials and Methods.” This system was used toevaluate the contribution of LMP1 to invasiveness. C33A/LMP1 cellswere about twice as invasive in this assay compared with C33A/pcDNA3 cells (P, 0.02; Fig. 1A), suggesting that LMP1 is a possibleinvasiveness-inducing factor.

NF-kB and AP-1 Are Involved in Expression of MMP-9 In-duced by LMP1. We showed previously that transient transfection ofLMP1 into C33A cells enhanced expression of MMP-9 (9). To in-vestigate the contribution of this type IV collagenase to the LMP1-associated invasionin vitro, we first carried out gelatin zymography todetermine MMP-9 expression. The zymographic analysis showed thatC33A/LMP1 cells constitutively enhanced the expression of MMP-9compared with C33A/pcDNA3 cells (Fig. 1B). As before (9), MMP-2gelatinolytic activity was not affected by LMP1 expression. Thisresult suggests that MMP-9 associated with LMP1 expression may beinvolved in cell invasivenessin vitro. Analysis of the effect of LMP1expression on MMP-9 promoter constructs (Fig. 1C) was then carriedout. MMP-9 promoter activity was increased in C33A/LMP1 cells, asshown by CAT reporter assays with the2670 CAT construct(P , 0.01; Fig. 1D, Lanes 1and2). C33A/LMP1 cells also increasedactivity of the 290 CAT construct, which only contains the AP-1binding site and TATA box, compared with the273 CAT constructin which the AP-1 binding site was deleted (P, 0.01; Fig. 1D, Lanes3–6). These results are compatible with previous reports that LMP1

2556

ASPIRIN INHIBITS LMP1-INDUCED INVASION AND MMP-9

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

activates JNK (24) because AP-1 is one of the downstream pathwaysof JNK. EMSA showed increases in nuclear factor binding not only toNF-kB but also to AP-1 binding sequence in the MMP-9 promoter inC33A/LMP1 cells (Fig. 1E).

Influence of Aspirin on Cell Viability and LMP1 Expression.Aspirin and sodium salicylate are cytotoxic in high concentrations;therefore, we investigated the effect of both drugs on cell viability byMTT assay. The effects of 1, 2.5, 5, and 10 mM aspirin or sodiumsalicylate were tested. Absorbance was not reduced, nor was there anyvisible cytotoxic effect by any of the concentrations tested except with10 mM aspirin (data not shown). Therefore, 5 mM aspirin and 10 mMsodium salicylate were used in the experiments. Neither drug affectedthe expression of LMP1 (Fig. 2,Lanes 3–5).

Aspirin Inhibits LMP1-induced Tumor Cell Invasiveness inVitro. The effect of salicylates on cell invasiveness was tested.LMP1-induced invasiveness was clearly inhibited by aspirin(P , 0.03) and sodium salicylate (P, 0.03; Fig. 3,Lanes 5and6).Neither drug affected invasiveness of the C33A/pcDNA3 cells (Fig. 3,Lanes 2and 3). These results suggest that both drugs can reduceLMP1-related invasiveness.

Aspirin Suppresses LMP1-induced MMP-9 Expression throughInhibition of NF- kB and AP-1. To clarify whether expression ofMMP-9 is involved in inhibition of LMP1-inducedin vitro invasionby salicylates, we first evaluated the effect of both drugs on LMP1-induced MMP-9 expression with the use of gelatin zymography.Zymographic analysis demonstrated that aspirin and sodium salicylateclearly abolished the LMP1-induced MMP-9 enhancement (Fig. 4,Lanes 3and4). MMP-2 gelatinolytic activity was affected by neitheraspirin nor sodium salicylate. Direct exposure of collected condi-tioned medium of C33A/LMP1 cells to aspirin or sodium salicylatehad no effect on MMP-9 and MMP-2 gelatinolytic activity (Fig. 4,Lanes 5and6), indicating that inhibition of LMP1-induced MMP-9was not attributable to a direct effect of the drugs on secreted MMP-9itself.

Next, the effects of the drugs on promoter activity were investi-gated. CAT reporter assays also showed that MMP-9 promoter activ-ity was dramatically suppressed by aspirin (P , 0.01) and sodiumsalicylate (P, 0.01; Fig. 5A, Lanes 3and 4). Cotransfection withIkBa(SS32/36AA) also sharply repressed CAT activity (P , 0.01;Fig. 5A, Lane 5). The residual activity might be attributable to the

Fig. 1. Contribution of LMP1 to invasivenessand MMP-9 expression.A, in vitro invasion assay.The invasion index is calculated as in “Materialsand Methods.” Averages with SDs (bars) from twosets of three experiments are shown.pcDNA3,C33A cells stably transfected with pcDNA3;LMP1, C33A cells stably expressing LMP1.B,gelatin zymography. Conditioned media were col-lected from C33A cells stably transfected withpcDNA3 and C33A cells stably expressing LMP1.C, schematic diagram of MMP-9 promoter con-structs.Filled rectangles,AP-1 binding sites;filledoval, NF-kB binding site; open rectangle,SP-1binding site;arrowhead,TATA box. Arrow, RNAstart site.D, CAT reporter assays. CAT reporterconstructs shown inC were transfected into C33Acells stably transfected with pcDNA3 (M) andC33A cells stably expressing LMP1 (f). The rel-ative CAT activities from four experiments areshown;bars,SD. E, EMSA. Nuclear extracts fromC33A cells stably transfected with pcDNA3(pcDNA3) or C33A cells stably expressing LMP1(LMP1) were mixed with either NF-kB (left panel)or AP-1 (right panel)32P-labeled probes. Excess ofnonlabeled NF-kB and AP-1 probes (3100) wereused as competitors (NF-kB, 59-GATCGGGTT-GCCCCAGTGGAATTCCCCAGCCTT-39; AP-1,59-GATCTTCTAGACCGGATGAGTCATAGC-CTT-39). Underlined lettersindicate binding se-quences in the promoter of theMMP-9 gene.NS,nonspecific binding.

2557

ASPIRIN INHIBITS LMP1-INDUCED INVASION AND MMP-9

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

contribution of AP-1, because at least the AP-1 binding site proximalto the start site is essential for activation of MMP-9 promoter byLMP1 (9). Therefore, we investigated the effect of the drugs on AP-1activity with the use of the290 CAT and273 CAT constructs. Theactivity of the290 CAT construct was repressed to the basal level byboth aspirin (P, 0.05) and sodium salicylate (P, 0.05; Fig. 5B,Lanes 4and 5), suggesting that LMP1-induced AP-1 activity wassignificantly repressed by both drugs. However, residual activitycompared with the273 CAT construct was still observed. BecauseC33A/pcDNA3 cells had more activity with the290 CAT than withthe273 CAT construct (Fig. 5B,Lanes 1and6), this residual activityseems to be independent of LMP1. Finally, binding activity of nuclearfactors not only to NF-kB but also to the AP-1 binding sequence in theMMP-9 promoter was reduced by aspirin and sodium salicylate (Fig.6A, Lanes 3, 5,and7; Fig. 6B, Lanes 4, 6,and8). Therefore, inhibitionof LMP1-induced MMP-9 by aspirin and sodium salicylate is prob-ably through the inactivation of both NF-kB and AP-1.

Overexpression of NF-kB p65/p50 and c-Jun Overcomes theInhibitory Effect of Aspirin on LMP1-induced MMP-9 Expres-sion. Overexpression of p50 and p65 can overcome NF-kB inhibitionby IkB(SS32/36AA), confirming that it is NF-kB that is blocked by

IkB(SS32/36AA) (Ref. 25). Therefore, transient transfection assayswere performed to investigate whether the inhibitory effect of aspirinon LMP1-induced MMP-9 expression could be reversed by overex-pression of NF-kB or c-Jun. Zymographic analysis showed that over-expression of p50 and p65 restored MMP-9 expression inhibited byaspirin (Fig. 7A, Lane 4). Overexpression of c-Jun also restoredMMP-9 expression (Fig. 7A, Lane 5). CAT reporter assays with the2670 CAT construct also showed that overexpression of p50 and p65restored MMP-9 promoter activity inhibited by aspirin (Fig. 7B, Lane4). Overexpression of c-Jun also restored MMP-9 promoter activity(Fig. 7B, Lane 5). These results suggest that both NF-kB and c-Jun areinvolved in the inhibitory effect of aspirin on LMP1-induced MMP-9expression.

Aspirin Inhibits LMP1-induced IKK Activity. IKK is responsi-ble for phosphorylation of IkBa on serine residues 32 and 36, whichis a critical step in NF-kB activation. We tested whether LMP1activates NF-kB through IKK, then tested whether aspirin had an

Fig. 2. Immunoblots for LMP1. Lysates were harvested after mock treatment ortreatment with either 5 mM aspirin (Asp) or 10 mM sodium salicylate (NaSal) for 24 h.Fifty mg protein of each sample except for SAV III (20mg) were loaded on a 10%polyacrylamide gel. SAV III is an EBV-positive Burkitt’s lymphoma line exhibiting typeIII latency.

Fig. 3. Association ofin vitro invasiveness with LMP1 and its inhibition by aspirin andsodium salicylate.In vitro invasion assays were carried out as described in Fig. 1.Asp,treatment with 5 mM aspirin;NaSal,treatment with 10 mM sodium salicylate;pcDNA3,C33A cells stably transfected with pcDNA3;LMP1,C33A cells stably expressing LMP1.Bars,SD.

Fig. 4. MMP-2 and MMP-9 gelatinolytic activities were analyzed with gelatin zymog-raphy. Conditioned media were collected from C33A cells stably transfected withpcDNA3 (pcDNA3) and C33A cells stably expressing LMP1 (LMP1) with or withouttreatment with either 5 mM aspirin (Asp) or 10 mM sodium salicylate (NaSal). In someexperiments, collected conditioned media of mock-treated C33A cells stably expressingLMP1 were incubated with either 5 mM aspirin or 10 mM sodium salicylatein vitro (direct)prior to incubation for 30 min at 37°C.

Fig. 5. Stimulation of MMP-9 promoter constructs by LMP1 and repression by aspirinand sodium salicylate.A, stimulation of MMP-9 promoter activity by LMP1 and itsrepression by aspirin, sodium salicylate, and IkB. The relative CAT activities from fourexperiments are shown;bars, SD. B, stimulation of deleted MMP-9 promoter activity(290 CAT) by LMP1 and its repression by aspirin and sodium salicylate. The relativeCAT activities from four experiments are shown;bars, SD. Asp, treatment with 5 mMaspirin; NaSal, treatment with 10 mM sodium salicylate;pcDNA3, C33A cells stablytransfected with pcDNA3;LMP1, C33A cells stably expressing LMP1.

2558

ASPIRIN INHIBITS LMP1-INDUCED INVASION AND MMP-9

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

effect on phosphorylation of IkBa by IKK. In vitro kinase assaysshowed phosphorylation of IkBa in LMP1-transfected cells in con-trast to vector-transfected cells (Fig. 8,Lanes 2and3), suggesting thatIKK is involved in LMP1-induced NF-kB activation. No phosphoryl-ation of IkBa by LMP1 was detected when GST-IkBa(SS32/36TT)was used as substrate (Fig. 8,Lane 7), indicating specificity of thisassay. Furthermore, treatment with aspirin abolished phosphorylationof IkBa (Fig. 8, Lane 4). This result suggests that aspirin inhibitsLMP1-induced NF-kB activation through inhibition of IKK activity.

Aspirin Inhibits LMP1-induced MMP-9 Expression in Vivo.Finally, whether LMP1 induces MMP-9in vivo was investigated.Both C33A/pcDNA3 cells and C33A/LMP1 cells are tumorigenic innude mice. LMP1 was detectable in only C33A/LMP1-derived tumors(Fig. 9A). Gelatin zymography showed that there was enhanced ex-pression of MMP-9 in C33A/LMP1-derived tumors compared withC33A/pcDNA3-derived tumor (Fig. 9B, Lanes 1and 2), indicatingthat LMP1 induces MMP-9 expressionin vivo. Next the effect ofaspirin on expression of MMP-9in vivo was investigated. Treatmentwith aspirin dramatically decreased expression of MMP-9 in C33A/LMP1-derived tumor (Fig. 9B, Lane 3) in contrast to expression ofLMP1, which was not affected (Fig. 9A, Lane 3). Expression ofMMP-2 was not affected by aspirin. These results suggest that aspirinmay be an inhibitor of LMP1-induced MMP-9 expressionin vivo.

DISCUSSION

LMP1 expression is not only essential for B-cell immortalization byEBV, but it is the only EBV protein that transforms nonlymphoid cellssuch as rodent fibroblasts (8), human epithelial cells (26), and humankeratinocytes (27). Consequently, the oncogenic potential of LMP1has been the focus of many investigations. A different aspect of LMP1function, i.e., its ability to induce MMP-9, which might contribute toinvasion and metastasis, is addressed here. LMP1 is expressed in atleast 70% of NPCs and in all EBV-infected preinvasive NPC lesions(28). NPC, which exhibits type II latency, shows clinically aggressivemetastasis not only into regional lymph nodes but distant organs suchas lung and vertebrae, although the primary tumor is still small andbeyond carcinomain situ. Additionally, NPC often shows aggressive

intracranial invasion. Although levels of MMPs including MMP-9 inNPC have not been reported yet, a clinical study showed that LMP1-positive NPCs tended to invade more frequently outside the nasophar-ynx and were more progressive and prone to invade the lymph nodesthan LMP1-negative tumors (29).

The COOH-terminal region of LMP1 contains two functional do-mains named CTAR1 and CTAR2. CTAR1 is needed for binding theTRAF-1, TRAF-2, and TRAF-3 proteins (30, 31), and CTAR2 hasbeen shown to bind the TNFR-associated death domain protein (32),suggesting that LMP1 might act as a TNFR family-like molecule.Indeed, downstream signaling of LMP1 is similar to that of CD40,which is one of the TNFR family members (33). Although LMP1 isnot homologous to any of the known TNFR family members and itsactivity does not appear to depend on the binding of a specific ligand,LMP1 activates NF-kB through a pathway including IKK, as shownhere and previously (34). Both CTAR1 and CTAR2 contribute toNF-kB activation. In addition, LMP1 also activates the JNK pathwaythrough CTAR2, which results in the activation of AP-1 (24).

Aspirin is used as a nonsteroidal anti-inflammatory drug. The rangeof concentrations for the drug is correlated with the amounts in plasma(1–5 mM) for optimal anti-inflammatory effects in patients with rheu-matoid arthritis. Moreover, both epidemiological and clinical studieshave indicated that aspirin and related compounds have considerablepotential as chemopreventive agents for cancers. It has been reportedthat patients with regular aspirin use had a reduced incidence of colon,lung, and breast cancers (35). Aspirin suppresses NF-kB activity orJNK activity induced by certain stimuli such as TNF-a, interleukin 1,and UVB in certain systems (14, 16, 36). It should be emphasized withrespect to activation of NF-kB and JNK that LMP1 is considered to bea constitutively activated TNFR-like molecule (37). Here we showthat aspirin inhibited constitutively active NF-kB and AP-1 and thatinhibition could be reversed by p50/p65 or c-Jun overexpression.Inhibition of LMP1-induced NF-kB activation through inhibition ofIKK is direct evidence of the inhibitory effect of aspirin on constitu-tively active NF-kB. Suppression of LMP1-inducible signaling path-ways may be a potential therapeutic tool for the type II and type IIIEBV-associated diseases.

Fig. 6. Induction of NF-kB (A) and AP-1 (B)DNA binding activity by LMP1 and its suppressionby aspirin and sodium salicylate. Nuclear extractsfrom C33A cells stably transfected with pcDNA3(pcDNA3) or C33A cells stably expressing LMP1(LMP1) in the absence or the presence of the treat-ment with 5 mM aspirin (Asp) or 10 mM sodiumsalicylate (NaSal) were mixed with either NF-kBor AP-1 32P-labeled probes. An excess of nonla-beled NF-kB or AP-1 probe (3100), shown in Fig.1, was used as competitor.NS,nonspecific binding.

2559

ASPIRIN INHIBITS LMP1-INDUCED INVASION AND MMP-9

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

MMP-9 has been shown to have a direct role in tumor metastasis incertain systems (38). Here we demonstrate that LMP1-inducedin vitroinvasiveness is correlated with induction of MMP-9. The possibilitythat suppression of MMP-9 could prevent invasion and metastasis hasbeen explored with the use of anti-MMP-9 ribozyme (39) or ursolicacid (40). Here we show that aspirin can be a possible candidate forrepression of MMP-9 and inhibition of invasiveness. Aspirin com-pletely suppressed MMP-9 promoter activity through its bifunctionaleffect on NF-kB and AP-1. We showed previously that MMP-9promoter activity induced by LMP1 was abolished when either theNF-kB site or the proximal AP-1 site was mutated (9). On the basisof this observation, inhibition of either NF-kB or AP-1 binding byaspirin might be causal for MMP-9 down-regulation. However, thedata in Fig. 5 suggest that the proximal AP-1 site may be responsiblefor mediating down-regulation, although the NF-kB site is essentialfor full induction, even in the presence of the proximal AP-1 site. Apreferential inhibitory effect of aspirin on binding of the upper NF-kBband on EMSA, the one known to contain p65, may be beneficial

because p65 has more transactivation capacity than p50 (41). Over-expression of either p50/p65 or c-Jun restored MMP-9 expression andpromoter activity, suggesting that either one of them may be opera-tive. This possibility may be explained by the previous reports show-ing the cross-coupling of NF-kB p65 and c-Jun (41, 42). Those reportsshowed that p65 activates AP-1 and c-Jun activates NF-kB (41), andthat NF-kB inhibitor blocks AP-1 and AP-1 inhibitor blocks NF-kB(42). Accordingly, we showed previously that LMP1 CTAR1 inducesAP-1 binding activity (11).

Thus, both aspirin and sodium salicylate significantly suppressedLMP1-induced in vitro invasion. Furthermore, aspirin inhibitedLMP1-induced MMP-9 expression in tumors grown in nude mice.The present study is the first report showing that aspirin and sodiumsalicylate suppress not only the expression of MMP-9 in cell culturebut alsoin vitro invasiveness. It is also the first to show induction ofMMP-9 by LMP1 in tumors grown in nude mice and inhibition of thatinduction by treatment of the tumor-bearing mice with aspirin. Thus,hypothetically, aspirin or an aspirin-like effect may have potential toprevent invasion and metastasis, in addition to its anti-inflammatoryand chemopreventive properties. Finally, this study is the first to show

Fig. 7. Restoration of MMP-9 expression and promoter activity. Expression plasmidsof NF-kB subunits p50 (0.2mg) and p65 (1mg) or c-Jun (0.5mg) were cotransfected withLMP1. A, conditioned media with or without treatment with 5 mM aspirin were collectedand analyzed by gelatin zymography.Bars, SD. B, cells treated with or without 5 mMaspirin were harvested, and promoter activity was analyzed by CAT assay with2670CAT construct.Bars,SD.

Fig. 8. In vitro kinase assay of IKK. Whole-cell extracts of vector-transfected andLMP1-transfected C33A cells with or without treatment with 5 mM aspirin were subjectedto in vitro kinase assays using GST-IkBa or GST-IkBa(SS32/36TT) for the substrate asindicated.Arrow, 32P-phosphorylated GST-IkBa (circled P).

Fig. 9. Induction of MMP-9 by LMP1 in tumors grown in nude mice and itssuppression by aspirin.A, immunoblots for LMP1. Lysates were obtained from equalamounts of C33A/pcDNA3 or C33A/LMP1 cell-derived tumors with or without treatmentwith aspirin. One hundred and thirtymg of protein were loaded.B, gelatin zymography.Lysates were obtained from equal amounts of C33A/pcDNA3 or C33A/LMP1 cell-derived tumors with or without treatment with aspirin.

2560

ASPIRIN INHIBITS LMP1-INDUCED INVASION AND MMP-9

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

suppression of downstream signals of LMP1 by aspirin and sodiumsalicylate. Aspirin significantly, but not completely, inhibited LMP1-induced tumor cell invasionin vitro. Unfortunately, tumors caused bys.c. injection into nude mice are generally noninvasive because of theectopic site. However,in vitro invasiveness predictsin vivometastasisin certain systems (43). Therefore, we are now establishing an LMP1-induced tumor invasion and metastasis model in nude mice so as toevaluate the effect of aspirin on invasion and metastasisin vivo.

ACKNOWLEDGMENTS

We thank Drs. Luwen Zhang and Matthew Davenport for helpful discus-sion, Albert Baldwin for the IkBa(SS32/36AA) and p50 or p65 subunit ofNF-kB expression plasmids and reviewing the manuscript, Minnie Holmes-McNary for helping with the IKK assay, and Nancy Raab-Traub and ShannonKenney for reviewing the manuscript.

REFERENCES

1. Liabakk, N. B., Talbot, I., Smith, R. A., Wilkinson, K., and Balkwill, F. Matrixmetalloprotease 2 (MMP-2) and matrix metalloprotease 9 (MMP-9) type IV collag-enases in colorectal cancer. Cancer Res.,56: 190–196, 1996.

2. Parsons, S. L., Watson, S. A., Collins, H. M., Griffin, N. R., Clarke, P. A., and Steele,R. J. Gelatinase (MMP-2 and -9) expression in gastrointestinal malignancy. Br. J.Cancer,78: 1495–1502, 1998.

3. Weiss, L. M., Chen, Y. Y., Liu, X. F., and Shibata, D. Epstein-Barr virus andHodgkin’s disease. A correlativein situ hybridization and polymerase chain reactionstudy. Am. J. Pathol.,139: 1259–1265, 1991.

4. Pallesen, G., Hamilton-Dutoit, S. J., Rowe, M., Lisse, I., Ralfkiaer, E., Sandvej, K.,and Young, L. S. Expression of Epstein-Barr virus replicative proteins in AIDS-related non-Hodgkin’s lymphoma cells. J. Pathol.,165: 289–299, 1991.

5. Harabuchi, Y., Yamanaka, N., Kataura, A., Imai, S., Kinoshita, T., Mizuno, F., andOsato, T. Epstein-Barr virus in nasal T-cell lymphomas in patients with lethal midlinegranuloma. Lancet,335: 128–130, 1990.

6. Raab-Traub, N., Flynn, K., Pearson, G., Huang, A., Levine, P., Lanier, A., andPagano, J. The differentiated form of nasopharyngeal carcinoma contains Epstein-Barr virus DNA. Int. J. Cancer,39: 25–29, 1987.

7. Murono, S., Yoshizaki, T., Tanaka, S., Takeshita, H., Park, C. S., and Furukawa, M.Detection of Epstein-Barr virus in nasopharyngeal carcinoma byin situ hybridizationand polymerase chain reaction. Laryngoscope,107: 523–526, 1997.

8. Wang, D., Liebowitz, D., and Kieff, E. An EBV membrane protein expressed inimmortalized lymphocytes transforms established rodent cells. Cell,43: 831–840,1985.

9. Yoshizaki, T., Sato, H., Furukawa, M., and Pagano, J. S. The expression of matrixmetalloproteinase 9 is enhanced by Epstein- Barr virus latent membrane protein 1.Proc. Natl. Acad. Sci. USA,95: 3621–3626, 1998.

10. Huen, D. S., Henderson, S. A., Croom-Carter, D., and Rowe, M. The Epstein-Barrvirus latent membrane protein-1 (LMP1) mediates activation of NF-kB and cellsurface phenotype via two effector regions in its carboxy-terminal cytoplasmicdomain. Oncogene,10: 549–560, 1995.

11. Takeshita, H., Yoshizaki, T., Miller, W. E., Sato, H., Furukawa, M., Pagano, J. S., andRaab-Traub, N. Matrix metalloproteinase 9 expression is induced by Epstein-Barrvirus latent membrane protein 1 C-terminal activation regions 1 and 2. J. Virol.,73:5548–5555, 1999.

12. Stancovski, I., and Baltimore, D. NF-kB activation: the IkB kinase revealed? Cell,91:299–302, 1997.

13. Verma, I. M., and Stevenson, J. IkB kinase: beginning, not the end. Proc. Natl. Acad.Sci. USA,94: 11758–11760, 1997.

14. Kopp, E., and Ghosh, S. Inhibition of NF-kB by sodium salicylate and aspirin.Science (Washington DC),265: 956–959, 1994.

15. Yin, M. J., Yamamoto, Y., and Gaynor, R. B. The anti-inflammatory agents aspirinand salicylate inhibit the activity of IkB kinase-b. Nature (Lond.),396: 77–80, 1998.

16. Huang, C., Ma, W. Y., Hanenberger, D., Cleary, M. P., Bowden, G. T., and Dong, Z.Inhibition of ultraviolet B-induced activator protein-1 (AP-1) activity by aspirin inAP-1-luciferase transgenic mice. J. Biol. Chem.,272: 26325–26331, 1997.

17. Sato, H., and Seiki, M. Regulatory mechanism of 92 kDa type IV collagenase geneexpression which is associated with invasiveness of tumor cells. Oncogene,8:395–405, 1993.

18. Sato, H., Kita, M., and Seiki, M. v-Src activates the expression of 92-kDa type IVcollagenase gene through the AP-1 site and the GT box homologous to retinoblastomacontrol elements. A mechanism regulating gene expression independent of that byinflammatory cytokines. J. Biol. Chem.,268: 23460–23468, 1993.

19. Yoshizaki, T., Sato, H., Maruyama, Y., Murono, S., Furukawa, M., Park, C. S., andSeiki, M. Increased expression of membrane type 1-matrix metalloproteinase in headand neck carcinoma. Cancer (Phila.).,79: 139–144, 1997.

20. Davenport, M. G., and Pagano, J. S. Expression of EBNA-1 mRNA is regulated bycell cycle during Epstein- Barr virus Type I latency. J. Virol.,73: 3154–3161, 1999.

21. Kulwichit, W., Edwards, R. H., Davenport, E. M., Baskar, J. F., Godfrey, V., andRaab-Traub, N. Expression of the Epstein-Barr virus latent membrane protein 1induces B cell lymphoma in transgenic mice. Proc. Natl. Acad. Sci. USA,95:11963–11968, 1998.

22. Yurochko, A. D., Hwang, E. S., Rasmussen, L., Keay, S., Pereira, L., and Huang,E. S. The human cytomegalovirus UL55 (gB) and UL75 (gH) glycoprotein ligandsinitiate the rapid activation of Sp1 and NF-kB during infection. J. Virol.,71:5051–5059, 1997.

23. Mercurio, F., Zhu, H., Murray, B. W., Shevchenko, A., Bennett, B. L., Li, J., Young,D. B., Barbosa, M., Mann, M., Manning, A., and Rao, A. IKK-1 and IKK-2:cytokine-activated IkB kinases essential for NF-kB activation. Science (WashingtonDC), 278: 860–866, 1997.

24. Kieser, A., Kilger, E., Gires, O., Ueffing, M., Kolch, W., and Hammerschmidt, W.Epstein-Barr virus latent membrane protein-1 triggers AP-1 activity via the c-JunN-terminal kinase cascade. EMBO J.,16: 6478–6485, 1997.

25. Wang, C. Y., Mayo, M. W., and Baldwin, A. S., Jr. TNF- and cancer therapy-inducedapoptosis: potentiation by inhibition of NF-kB. Science (Washington DC),274:784–787, 1996.

26. Dawson, C. W., Rickinson, A. B., and Young, L. S. Epstein-Barr virus latentmembrane protein inhibits human epithelial cell differentiation. Nature (Lond.),344:777–780, 1990.

27. Fahraeus, R., Rymo, L., Rhim, J. S., and Klein, G. Morphological transformation ofhuman keratinocytes expressing theLMP gene of Epstein-Barr virus. Nature (Lond.),345: 447–449, 1990.

28. Pathmanathan, R., Prasad, U., Sadler, R., Flynn, K., and Raab-Traub, N. Clonalproliferations of cells infected with Epstein-Barr virus in preinvasive lesions relatedto nasopharyngeal carcinoma. N. Engl. J. Med.,333: 693–698, 1995.

29. Hu, L. F., Chen, F., Zhen, Q. F., Zhang, Y. W., Luo, Y., Zheng, X., Winberg, G.,Ernberg, I., and Klein, G. Differences in the growth pattern and clinical course ofEBV-LMP1 expressing and non-expressing nasopharyngeal carcinomas. Eur. J. Can-cer,5: 658–660, 1995.

30. Mosialos, G., Birkenbach, M., Yalamanchili, R., VanArsdale, T., Ware, C., and Kieff,E. The Epstein-Barr virus transforming protein LMP1 engages signaling proteins forthe tumor necrosis factor receptor family. Cell,80: 389–399, 1995.

31. Devergne, O., Hatzivassiliou, E., Izumi, K. M., Kaye, K. M., Kleijnen, M. F., Kieff,E., and Mosialos, G. Association of TRAF1, TRAF2, and TRAF3 with an Epstein-Barr virus LMP1 domain important for B-lymphocyte transformation: role in NF-kBactivation. Mol. Cell. Biol.,16: 7098–7108, 1996.

32. Izumi, K. M., and Kieff, E. D. The Epstein-Barr virus oncogene product latentmembrane protein 1 engages the tumor necrosis factor receptor-associated deathdomain protein to mediate B lymphocyte growth transformation and activate NF-kB.Proc. Natl. Acad. Sci. USA,94: 12592–12597, 1997.

33. Eliopoulos, A. G., Dawson, C. W., Mosialos, G., Floettmann, J. E., Rowe, M.,Armitage, R. J., Dawson, J., Zapata, J. M., Kerr, D. J., Wakelam, M. J., Reed, J. C.,Kieff, E., and Young, L. S. CD40-induced growth inhibition in epithelial cells ismimicked by Epstein-Barr virus-encoded LMP1: involvement of TRAF3 as a com-mon mediator. Oncogene,13: 2243–2254, 1996.

34. Sylla, B. S., Hung, S. C., Davidson, D. M., Hatzivassiliou, E., Malinin, N. L.,Wallach, D., Gilmore, T. D., Kieff, E., and Mosialos, G. Epstein-Barr virus-trans-forming protein latent infection membrane protein 1 activates transcription factorNF-kB through a pathway that includes the NF-kB-inducing kinase and the IkBkinases IKKa and IKKb. Proc. Natl. Acad. Sci. USA,95: 10106–10111, 1998.

35. Schreinemachers, D. M., and Everson, R. B. Aspirin use and lung, colon, and breastcancer incidence in a prospective study. Epidemiology,5: 138–146, 1994.

36. Schwenger, P., Bellosta, P., Vietor, I., Basilico, C., Skolnik, E. Y., and Vilcek, J.Sodium salicylate induces apoptosis via p38 mitogen-activated protein kinase butinhibits tumor necrosis factor-induced c-Jun N-terminal kinase/stress-activated pro-tein kinase activation. Proc. Natl. Acad. Sci. USA,94: 2869–2873, 1997.

37. Gires, O., Zimber-Strobl, U., Gonnella, R., Ueffing, M., Marschall, G., Zeidler, R.,Pich, D., and Hammerschmidt, W. Latent membrane protein 1 of Epstein-Barr virusmimics a constitutively active receptor molecule. EMBO J.,16: 6131–6140, 1997.

38. Bernhard, E. J., Gruber, S. B., and Muschel, R. J. Direct evidence linking expressionof matrix metalloproteinase 9 (92-kDa gelatinase/collagenase) to the metastatic phe-notype in transformed rat embryo cells. Proc. Natl. Acad. Sci. USA,91: 4293–4297,1994.

39. Hua, J., and Muschel, R. J. Inhibition of matrix metalloproteinase 9 expression by aribozyme blocks metastasis in a rat sarcoma model system. Cancer Res.,56: 5279–5284, 1996.

40. Cha, H. J., Bae, S. K., Lee, H. Y., Lee, O. H., Sato, H., Seiki, M., Park, B. C., andKim, K. W. Anti-invasive activity of ursolic acid correlates with the reduced expres-sion of matrix metalloproteinase-9 (MMP-9) in HT1080 human fibrosarcoma cells.Cancer Res.,56: 2281–2284, 1996.

41. Stein, B., Baldwin, A. S., Jr., Ballard, D. W., Greene, W. C., Angel, P., and Herrlich,P. Cross-coupling of the NF-kB p65 and Fos/Jun transcription factors producespotentiated biological function. EMBO J.,12: 3879–3891, 1993.

42. Li, J. J., Westergaard, C., Ghosh, P., and Colburn, N. H. Inhibitors of both nuclearfactor-kB and activator protein-1 activation block the neoplastic transformationresponse. Cancer Res.,57: 3569–3576, 1997.

43. Uemura, K., Takao, S., and Aikou, T.In vitro determination of basement membraneinvasion predicts liver metastases in human gastrointestinal carcinoma. Cancer Res.,58: 3727–3731, 1998.

2561

ASPIRIN INHIBITS LMP1-INDUCED INVASION AND MMP-9

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from

2000;60:2555-2561. Cancer Res   Shigeyuki Murono, Tomokazu Yoshizaki, Hiroshi Sato, et al.   Suppression of Matrix Metalloproteinase-9 ExpressionEpstein-Barr Virus Latent Membrane Protein 1 through Aspirin Inhibits Tumor Cell Invasiveness Induced by

  Updated version

  http://cancerres.aacrjournals.org/content/60/9/2555

Access the most recent version of this article at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/60/9/2555.full#ref-list-1

This article cites 42 articles, 23 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/60/9/2555.full#related-urls

This article has been cited by 20 HighWire-hosted articles. Access the articles at:

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/60/9/2555To request permission to re-use all or part of this article, use this link

Research. on April 30, 2021. © 2000 American Association for Cancercancerres.aacrjournals.org Downloaded from