Oncogene (2001) 20, 2171 ± 2177 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc SHORT REPORTS homozygous deletion or MDM2 overexpression in Burkitt lymphoma lines carrying wild type

Mikael S LindstroÈ m1, Ulf Klangby2 and Klas G Wiman*,1

1Karolinska Institute, Department of Oncology-Pathology, Cancer Center Karolinska (CCK), Karolinska Hospital, SE-171 76 Stockholm, Sweden; 2AB Sangtec Medical, Box 20045, SE-161 02 Bromma, Sweden

The hallmark of Burkitt lymphoma (BL) is a constitu- al., 1992) and 60-70% of BL cell lines (Wiman et al., tively activated c-myc that drives tumor cell growth. 1991; Farrell et al., 1991; Vousden et al., 1993; A majority of BL-derived cell lines also carry mutant Cherney et al., 1997). Mutations in Bcl-6 and Bax p53. In addition, the p16INK4a promoter is hypermethy- and the presence of Epstein-Barr virus (EBV) have lated in most BL biopsies and BL cell lines, leading to also been suggested to contribute to the molecular silencing of this gene. Activation of c-myc and/or cell pathogenesis of BL (Klein, 1993; Gutierrez et al., cycle dysregulation can induce ARF expression and p53- 1997, 1999). dependent . We therefore investigated the The ARF encoded by the CDKN2A is p14ARF-MDM2-p53 pathway in BL cell lines. a bona ®de tumor suppressor that is frequently lost in p14ARF was expressed and localized to nucleoli in all various human tumors (reviewed by Sharpless and BL carrying mutant p53. Three out of seven BL carrying DePinho, 1999; Sherr and Weber, 2000). ARF binds wt p53 had a homozygous deletion of the CDKN2A locus and inhibits the p53 antagonist MDM2 that targets that encodes both p14ARF and p16INK4a. Three BL p53 for proteasome-mediated degradation through carrying wild type p53 retained the CDKN2A locus and nuclear export (Kamijo et al., 1998; Pomerantz et al., overexpressed MDM2. DNA sequencing revealed a point 1998; Zhang et al., 1998; Tao and Levine, 1999; Zhang mutation in CDKN2A 2 in one of these BL, and Xiong, 1999; Weber et al., 1999). Both mouse Seraphine. However, this point mutation did not a€ect p19ARF and human p14ARF have been shown to p14ARF's nucleolar localization or ability to induce p53. sequester MDM2 in nucleoli and inhibit MDM2's The Bmi-1 protein that negatively regulates the p14ARF ubiquitin ligase activity, thus triggering p53 accumula- promoter and co-operates with c-myc in tumorigenesis tion (Honda and Yasuda, 1999; Weber et al., 1999; was expressed at low to moderate levels in all BL Weber et al., 2000). The CDKN2A locus also encodes analysed. Our results indicate that inactivation of the the tumor suppressor p16INK4a, an inhibitor of the ARF-MDM2-p53 pathway is an essential step during the CDK4/cyclin D complex that phosphorylates the pRb development of Burkitt lymphoma, presumably as a protein. Thus, homozygous deletion of CDKN2A will mechanism to escape c-myc induced apoptosis. Oncogene simultaneously disrupt both the pRb and p53 regula- (2001) 20, 2170 ± 2177. tory pathways (see Sherr and Weber, 2000). Selective inactivation of human p14ARF via deletion of exon 1b Keywords: p14ARF; MDM2; p53; nucleolus; Bmi-1; has been found in some metastatic cell lines Burkitt lymphoma (Kumar et al., 1998), and in acute T-cell lymphocytic leukemia (Gardie et al., 1998). A germline exon 1b deletion has been identi®ed in a rare melanoma-neural Introduction system tumor syndrome family (Randerson-Moor et al., 2001). p14ARF can potentially also be inactivated Burkitt lymphoma (BL) is a high grade B-cell by overexpression of the Bmi-1 or Twist transcription lymphoma that regardless of geographical origin carry factors that repress the ARF promoter (Jacobs et al., reciprocal chromosomal translocations that juxtapose 1999a; Maestro et al., 1999). Bmi-1 overexpression has the c-myc proto-oncogene on 8 to the been shown to collaborate with c-myc through averting immunglobulin heavy chain locus on chromosome 14 apoptosis (Jacobs et al., 1999b). In addition, ARF or the kappa or lambda light chain locus on inactivation may occur through methylation of the 2 or 22 (for a review see Klein, 1993). ARF promotor (Esteller et al., 2000). In addition, p53 mutations have been found in at least Mice expressing a c-myc transgene under the 30% of BL biopsies (Gaidano et al., 1991; Bhatia et control of the immunoglobulin heavy chain enhancer (Em) develop clonal pre-B and B-cell lymphoma, in analogy to human BL (Adams et al., 1985). A majority of the emerging tumors carry p53 mutation *Correspondence: KG Wiman Received 8 November 2000; revised 8 January 2001; accepted 24 or INK4A/ARF deletion, and/or overexpress MDM2 January 2001 protein (Eischen et al., 1999). Loss of INK4A/ARF p14ARF and MDM2 alterations in Burkitt lymphoma MS LindstroÈm et al 2172 abrogates B-cell apoptosis in mice carrying an Em- pathway is invariably inactivated in BL, a c-myc myc transgene, and greatly accelerates lymphomagen- driven tumor. We therefore examined the status of esis (Eischen et al., 1999; Schmitt et al., 1999). These p14ARF, MDM2 and Bmi-1 in BL cell lines carrying results led us to ask whether the ARF-MDM2-p53 wild type or mutant p53.

Figure 1 (a) Analysis of p14ARF homozygous deletion by PCR. Homzygous deletion of p14ARF exon 1b and exon 2 was observed in cell lines BL2, BL28 and Ew36 whereas other lines had at least one allele intact. Status of p53 is shown in Table 1. Exon 1b and exon 2 were always co-deleted and exon1b speci®c deletions were not found. (b) RT-PCR analysis of p14ARF revealed that mRNA is abundantly expressed in BL lines. No signal was detected in BL with homozygous deletion. Materials and methods: BL suspension cells were centrifuged, washed 16 in PBS, resuspended in digestion bu€er (100 mM NaCl, 10 mM Tris-Cl, 25 mM EDTA, 0.5% w/v SDS, 0.1 mg/ml proteinase K, 20 mg/ml pancreatic RNase) and incubated in a shaker for 12 h at 508C. Samples were extracted using phenol/chloroform/isoamyl alcohol (25 : 24 : 1) and DNA was precipitated using 7.5 M ammonium acetate and 95% EtOH/dH2O. Pellets were washed twice in 70% EtOH/dH2O and resuspended in Tris-EDTA bu€er. (2) PCR was carried out using an iCyclerTM thermocycler (Bio-Rad laboratories, Hercules, CA, USA). Reactions were performed in a volume of 50 ml containing a minimum of 400 ng DNA template; 10 pmoles of each primer, 10% DMSO, 1.5 mM MgC12, 2.5 mM dNTP's using Taq polymerase and bu€er provided with the enzyme (GIBCO). Primers used for sequencing of p14ARF were for exon 1b: (F) 5' TGCAGTTAAGGGGGGCAGGAG-3' and (R) 5'-TTATCTCCTCCTCCTCCTAGCCTG-3' and for exon 2: (551R) 5'- TCTGAGCTTTGGAAGCTCT-3' and (42F) 5'-GGAAATTGGAAACTGGAAGC-3'. The PCR products were puri®ed using Qiagen Gelextraction kit (Qiagen) and subsequently sequenced using the same set of primers and the ABI PRISM BigDye Terminator Cycle Sequencing Kit (PE Biosystems, Foster City, CA, USA). Samples were run on an ABI PRISM Sequencer (PE Biosystems). The quality of genomic DNA was con®rmed by parallel PCR ampli®cation of a GAPDH gene exon 8 fragment. p14ARF cDNA was synthesized from total RNA using M-MLV Reverse Transcriptase according to the manufacturer's instructions (GIBCO). PCR was performed for a total of 35 cycles using an annealing temperature of 618C and the following primers for p14ARF cDNA: (F) 5'-AAGGATCCATGGTGCGCAGGTTCTTGG-3' and (R) 5'-GTTGTGGCGGGGGCAGTTGT-3'. Ten microliters from each reaction was run on 1% agarose gels and transferred onto Hybond N+membranes (Amersham, Buckinghamshire, UK). Membranes were probed with 32P-dCTP-labeled probes corresponding to exon 1b. Hybridizing bands were visualized using PhosphoImager (Molecular Dynamics). GAPDH cDNA was ampli®ed by RT-PCR using the primers: 5'- TGCCTCCTGCACCACCAACTG-3' (GAPDHE7F) and 5'-ATGCCAGCCCCAGCGTCAAAG-3' (GAPDHE8R)

Oncogene p14ARF and MDM2 alterations in Burkitt lymphoma MS LindstroÈm et al 2173 First, RT ± PCR revealed that p14ARF mRNA was homozygous deletion of this locus (results for a subset abundantly expressed in all BL lines that had retained of BL lines are shown in Figure 1b). Thus, p14ARF CDKN2A locus but not detected in BL lines carrying expression is not downregulated by hypermethylation

Figure 2 Immun¯uorescence staining analysis of p14ARF, MDM2 and Bmi-1 in BL lines. Bmi-1 was detected in the nucleus of human diploid ®broblasts and in U2OS cells (a ± d, i). Note the close association of Bmi-1 with the nucleolus (i). DG75 and BL41 lines displayed strong accumulation of Bmi-1 in intranuclear dots (e,f,j), whereas Seraphine and majority of other BL lines showed a more homogenous weak nucleoplasmic Bmi-1 staining (g,h, see text for details). p14ARF was detected in majority of Burkitt lymphoma lines, here exempli®ed by Namalwa and BL41 (k±n). p14ARF was localized to nucleoli, as evidenced by B23 and p14ARF double immunostaining (o±q) and phase contrast microscopy (r,s). MDM2 protein was easily detected by mAb SMP14 in Seraphine cells where it was localized to discrete nuclear blobs (x). Seraphine only showed weak p53 immunoreactivity with the polyclonal anti-p53 antibody FL393 (t). Scale bars; a±h, k±n 10 mM; i,j, o±s, x 5 mM and t±v 20 mM. Materials and methods: The MDM2 speci®c mouse monoclonal antibody SMP14 (diluted 1 : 200) and mouse monoclonal anti-p14ARF (14P02) were purchased from Neomarkers/Labvision Corp., Fremont, CA, USA and diluted 1 : 100. Our p14ARF speci®c rabbit polyclonal antiserum 271 (diluted 1 : 100) has been described (LindstroÈ m et al., 2000). The Bmi-1 F6 mouse monoclonal (diluted 1 : 100) was a kind gift from Maarten van Lohuizen (Netherlands Cancer Institute, Amsterdam, The Netherlands) and has been described (Alkema et al., 1997). Nucleolus speci®c marker antibody B23/nucleophosmin (1 : 200) was a generous gift from Pui K Chan (Baylor College, TX, USA). Goat-polyclonal antibody FL-393 (1 : 400) against human p53 was obtained from Santa Cruz. Secondary ¯uorescent antibody kits were from VECTOR (Vector Laboratories, Inc. Burlingame, CA, USA) and included horse anti-mouse lg antibodies conjugated with FITC, Texas Red and AMCA and also goat anti-rabbit lg (FITC, Texas Red and AMCA). FITC conjugated rabbit anti-goat lg was purchased from Sigma (St Louis, MO, USA). Secondary antibodies were diluted 1 : 50-1 : 100. For immuno¯uorescence studies, smears from exponentially growing suspension cells were prepared on glass slides using a cytospin centrifuge (usually 1500 r.p.m. for 5 min) and then ®xed in 4% paraformaldehyde for 10 min, followed by permeabilization in 0.2% NP40 in PBS for 3 min. Adherent cell lines were seeded onto glass slips in 6-well tissue culture dishes and grown for 24 ± 28 h. Cells at 70 ± 90% con¯uence were ®xed in 4% paraformaldehyde for 10 min and permeabilized in 0.2% NP40 for 5 min. Antibodies and normal sera were diluted in blocking bu€er (2% BSA, 0.2% Tween-20, 10% glycerol, 0.05% NaN3 in sterile PBS). For DNA staining, Hoechst dye 33258 was included in the BSA-blocking bu€er during the last antibody incubation. Procedures for immunostaining, image capturing and analysis have been described (LindstroÈ m et al., 2000)

Oncogene p14ARF and MDM2 alterations in Burkitt lymphoma MS LindstroÈm et al 2174 in BL, in contrast to p16INK4a (Klangby et al., 1998; expressed and localized to nucleoli in BL lines carrying Baur et al., 1999). Using a p14ARF monoclonal mutant p53. Thus, the previously described correlation antibody we found that all BL lines harboring mutant between mutant p53 status and overexpression of ARF p53 expressed low to moderate levels of p14ARF that (Quelle et al., 1995; Stott et al., 1998) is applicable to was localized to nucleoli Figure 2k ± n) as evidenced by BL lines, in keeping with the proposed ARF-p53 feed phase-contrast microscopy and double immunostaining back loop (Stott et al., 1998), and suggesting that the with an antibody against the nucleolar protein B23/ presence of nucleolar p14ARF is a good indicator for nucleophosmin (Figure 2o ± s). p14ARF was not the p53 gene status in BL lines. detected by Western blotting (Wb) nor immuno¯uor- Next, a subset of BL lines carrying wt p53 were escence (IF) staining in our BL lines encoding wt p53, examined with regard to bi-allelic deletion of in agreement with the notion that p53 represses the CDKN2A using PCR primers speci®c for exon1b and ARF promoter (Robertson and Jones, 1998). The BL exon 2. The wild type p53-carrying BL28, Ew36 and line BL41 had the highest levels of p14ARF (Figures BL2 lines all had homozygous deletion of CDKN2A 2k and 3a). Other lines expressed lower but clearly (Table 1, Figure 1a). On the other hand, Ly47, Ly91, detectable levels of p14ARF (Table 1). In summary, and BL7 that also carry wild type p53 retained at least our data demonstrate that p14ARF protein is one wild type CDKN2A allele according to PCR and DNA sequencing (Table 1, Figure 1a). A re-evaluation of our previous data with focus on p14ARF status showed that all of 19 primary BL biopsies had retained Table 1 Pattern of CDKN2A (p14ARF and p16INK4a), p53 and intact CDKN2A and expressed p14ARF mRNA MDM2 alterations in Burkitt lymphoma lines (Klangby et al., 1998). Of a total of 47 established EBV p53 p14ARF p16INK4a MDM2 BL lines only three (6%) had lost p14ARF by homozygous deletion (Klangby et al., 1998; this study Ew36 7 wt del del + BL2 7 wt del del ++ Table 1). In a study from another laboratory, one out BL28 7 wt del del + of eight primary BL biopsies had a deletion of Seraphine + wt mut, 7 mut +++ CDKN2A (Stranks et al., 1995). Thus, loss of Ly47 + wt wt, 7 ND +++ p14ARF by homozygous deletion is infrequent in BL Ly91 + wt wt, 7 ND +++ BL7 + wt wt, 7 ND + biopsies and cell lines but occurs in a subset of lines BL41 7 R248Q wt, nucleolar ++ met + that harbor wild type p53. Interestingly, all three wild BL60 + R282W wt, nucleolar + met + type p53-carrying BL lines that had homozygous DG75 7 R283H wt, nucleolar ++ met + deletion of CDKN2A were EBV-negative, whereas Namalwa + R248Q wt, nucleolar ++ met + the four BL that carried wild type p53 and had Raji + T234H/R213Q wt, nucleolar + met + Ramos 7 I254D wt, nucleolar + met + retained CDKN2A were EBV-positive (Table 1). BL49 7 C176Y/R248W wt, nucleolar + ND + Mouse plasmacytomas (MPCs) invariably harbor c- BL30 7 M246T wt, nucleolar + ND + myc translocations similar to those found in BL. Our BL37 + R237I wt, nucleolar + ND + screening of a panel of primary MPC tumors and CA46 7 R248Q wt, nucleolar ++ met + Akata + Frameshift 190 wt, nucleolar + met + derived cell lines indicated infrequent homozygous CDKN2A deletion (LindstroÈ m et al., unpublished p53 gene status was determined by direct DNA sequencing, p14ARF results). Zhang et al. (1998) have shown that the status by indirect immuno¯uorescence (IF), Western blotting (Wb) p16INK4a promotor is hypermethylated and p19ARF and direct DNA sequencing. MDM2 levels were investigated using IF and mRNA expression retained in majority of MPCs, and Wb. Abbreviations and notes: wt, wildtype; del, deleted; ND, not determined; mut, mutant; met, methylated; presence or absence of like in BL. Thus, homozygous CDKN2A deletion EBV is indicated by (+) or (7); levels of MDM2 and p14ARF are appears to be much less frequent in both MPC and BL indicated with numbers of (+); p16INK4a methylation data were than in lymphomas arising in Eu-myc transgenic mice adapted from Klangby et al., 1998. Materials and methods: BL cell where CDKN2A deletion occurs in around 20% of the lines were provided by George Klein (Microbiology and Tumor Biology canter, Karolinska Institute, Stockholm, Sweden) and tumors (Eischen et al., 1999). However, the ARF- Gilbert Lenoir (IARC, Lyon, France). Control cell lines used were: MDM2-p53 pathway appears functional in a majority MCF7, HeLa, Sw480, U2OS, Saos2 and human diploid ®broblasts of MPC cell lines (Zhang et al., 2001), in contrast to (for details see LindstroÈ m et al., 2000 and Stott et al., 1998). Cells BL lines (this study). were grown at 378C (5% CO ) in Iscove's modi®ed Dulbecco's 2 DNA sequencing of p14ARF (exon 1b, exon 2 and medium (IMDM) supplemented with penicillin/streptomycin 100 U/ ml, 2 mM L-glutamine and 15% fetal calf serum (GIBCO, Grand the immediate upstream promotor region of p14ARF) Island, NY, USA). The p53 status of the BL lines (except BL28) has in all BL lines carrying an intact CDKN2A locus in been reported (Farrell et al., 1991; Vousden et al., 1993; Cherney et this study revealed no point mutations or small al., 1997 and Capoulade et al., 1998). cDNA from BL28 was deletions that could a€ect p14ARF (Table 1). Pre- ampli®ed from total RNA using Advantage RT-PCR kit according to the manufacturer's instructions (Clontech, Palo Alto, CA, USA). viously, we had found a point mutation creating a Gly p53 was sequenced from pooled PCR reactions using the following to Val change in p14ARF at codon 114 in the BL line set of primers: Seraphine. To test the functional signi®cance of this point mutation, Seraphine p14ARF cDNA was cloned p53F1, 5'-GAAGGATCCGAGGAGCCGCAGT-3'; p53F781, 5'CTGAGGTTGGCTCTGACTGTACCACCATCC-3'; into the pEGFP-C1 vector and transfected into MCF-7 p53R789, 5'AACAAGCTTATTACCACTGGAGTCTTC-3'; cells that carry wild type p53 and express MDM2. The p53R1179, 5'-AGGGAATTCAGTCTGAGTCAGGC-3' Seraphine p14ARF mutant localized in nucleoli and

Oncogene p14ARF and MDM2 alterations in Burkitt lymphoma MS LindstroÈm et al 2175

Figure 3 Western blot analysis of p14ARF, MDM2 and Bmi-1 protein in BL lines. (a) Faint bands representing p14ARF were detected in BL lines BL41, DG75, Ramos and Namalwa. Antibody 14P02 (diluted 1 : 500) did not show reactivity in the p14ARF negative cell line MCF-7 (or in BL2, data not shown). HeLa and SW480 were used as strong positive controls (Stott et al., 1998). (b) MDM2 is overexpressed in BL2 and Seraphine (in comparison with BL28 and Ew36) but is also detected in other lines such as DG75 using monoclonal antibody 2A10 (1 : 500). Thirty mg of protein lysate was loaded in each well. (c) Bmi-1 protein was detected using the F6 monoclonal (1 : 300) in all BL lines tested. HeLa cells were used as a positive control. Materials and methods: Total cell extracts were prepared in lysis bu€er (100 mM Tris pH 8.0, 150 mM NaCl, 1% NP40, 1 mM phenylmethylsulfonyl ¯uoride). Protein concentration was determined using the Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, CA, USA). Samples containing equal amounts of protein (usually 30 mg) were separated on 8% (MDM2), 12% (Bmi-1) and 17.5% (p14ARF) SDS polyacrylamide gels and transferred to nitrocellulose membranes (Amersham). Blocking and incubation with primary and secondary antibodies were performed in 5% milk and 0.1% Tween20 in PBS. Immunodetection using the ECL system was carried out according to manufacturer's instructions (Amersham). Membranes were stripped in 2% SDS, 50 mM DTT and 50 mM Tris-HCl (pH 7.0) for 30 min at 708C. Equal loading was veri®ed when reprobing membranes with a mouse monoclonal antibody against b-actin (1 : 500) (Sigma)

induced p53 and MDM2 as eciently as the wt EGFP- (Jacobs et al., 1999a,b). Using the Bmi-1 monoclonal C1-p14ARF fusionprotein, when overexpressed (data antibody F6 (Alkema et al., 1997) we examined Bmi-1 not shown). We conclude that the Gly to Val expression in our BL lines and in human primary cells substitution at codon 114 does not signi®cantly a€ect and tumor cell lines. In good agreement with ®ndings the function of Seraphine p14ARF. by others (Alkema et al., 1997; Voncken et al., 1999) MDM2 is overexpressed in certain tumors carrying we found an intense nuclear Bmi-1 staining in certain wild type p53, for instance sarcomas (Oliner et al., distinct sub-nuclear domains in U20S, HeLa and 1992; Cordon-Cardo et al., 1994) and lymphomas/ MCF7 cells in addition to a ®ne grained nuclear leukemias (Watanabe et al., 1994). Overexpression of staining excluding the nucleolus (Figure 2c,d,i data not MDM2 is thought to inactivate wild type p53 through shown). Early passage human diploid foreskin ®bro- binding to the transcriptional transactivation domain blasts displayed a weaker ®ne-grained nuclear staining of p53 and by increased nucleo-cytoplasmic shuttling without the large nuclear domains observed in the and degradation. A previous study demonstrated that tumor cell lines (Figure 2a,b). MDM2 is overexpressed in some BL lines that carry The majority of BL lines expressed low to moderate wild type p53, probably due to enhanced translation amounts of nuclear Bmi-1 protein (exempli®ed in (Capoulade et al., 1998). We found that Seraphine, Figure 2g,h). Two BL (BL41 and DG75) displayed BL2, Ly47 and Ly91 expressed readily detectable levels an intense `classical' Bmi-1 staining pattern similar to of MDM2 protein as shown by Western blotting and that observed in U2OS cells (Figure 2e,f). Western immuno¯uorescence staining (Figures 3b and 2u ± x, blotting showed that a majority of the cell lines data not shown). A few other BL lines, i.e. BL60, Raji expressed equally low levels of Bmi-1 protein without and DG75, also expressed MDM2 protein but at any clear correlation to CDKN2A or p53 status signi®cantly lower levels. Most BL lines expressed very (Figure 3c). We conclude that Bmi-1 protein is not low or undetectable amounts of MDM2. overexpressed in any of the BL lines. However, we Finally, we asked whether expression of Bmi-1 was cannot rule out the possibility that Bmi-1 co-operates altered in any of the BL lines. Bmi-1 belongs to a with c-myc during BL development in vivo. group of Polycomb gene silencers and collaborates In conclusion, our data demonstrate that the ARF- with c-myc in lymphoma development (Haupt et al., MDM2-p53 pathway is frequently inactivated in BL 1993; Jacobs et al., 1999b). The CDKN2A locus is cell lines. In our panel, all but one, BL7, had either p53 proposed to be a key target of Bmi-1 repression mutation, homozygous ARF deletion or overexpressed

Oncogene p14ARF and MDM2 alterations in Burkitt lymphoma MS LindstroÈm et al 2176 MDM2. This is consistent with the notion that rapidly in vitro than those with only MDM2 over- ablation of ARF-p53 dependent apoptosis in response expression or no alteration at all in agreement with to c-myc activation provides a strong selective observations by others (Gronbaek et al., 2000; Pinyol advantage during BL development and probably is an et al., 2000; Schmitt et al., 1999). This may be due to essential genetic change. Further support of this idea incomplete inactivation of p53-dependent apoptosis by comes from studies of non-Hodgkin lymphomas MDM2 overexpression. Indeed, some studies have showing that p53 mutation and CDKN2A deletion indicated that p53 may still respond to anticancer are mutually exclusive events (Pinyol et al., 2000). drugs, such as etoposide, in BL cells that overexpress However, CDKN2A deletion is infrequent in BL MDM2 (Gao et al., 1999). biopsies (Klangby et al., 1998; Stranks et al., 1995) Our results emphasize the need to examine all and cell lines (this study) compared to other B-cell components of the ARF-MDM2-p53 pathway in a neoplasms. The reason for the bias for p53 mutation in given tumor in order to assess its p53 function. They BL is unclear. It is conceivable that hypermethylation also stress the importance of developing therapeutic of the promoter, which occurs in a majority of BL methods aiming at pharmacological rescue of this biopsies and BL lines, reduces further selection for pathway in human tumors, either by reintroduction homozygous deletion of the entire CDKN2A locus. of p14ARF, reactivation of mutant p53, or disruption This would favor inactivation of the ARF-p53 pathway of the MDM2-p53 protein complex. by other mechanisms, e.g. p53 mutation. Moreover, the presence of EBV and expression of EBV-encoded may have an impact on the selection pressure Acknowledgments for homozygous CDKN2A (p14ARF) deletion versus We thank George Klein, Karolinska Institute, Stockholm other mechanisms of inactivation of ARF-p53 depen- and Gilbert Lenoir, IARC, Lyon, for providing BL cell dent apoptosis. Although our series of wild type p53- lines and Maarten van Lohuizen, Netherlands Cancer carrying BL lines is too small to draw ®rm conclusions, Center, Amsterdam, Pui K Chang, Baylor College of it is noteworthy that all BL lines that had deleted both Medicine, Houston, and Arnold J Levine, Rockefeller CDKN2A alleles were EBV-negative while the remain- University, New York, for their generous gifts of anti- der were EBV positive. This raises the possibility that bodies. We also acknowledge Joseph Lawrence at the core sequencing facility at Cancer Center Karolinska (CCK) for EBV somehow tips the balance towards retention of skilful technical assistance and Katja Pokrovskaja, Kar- the CDKN2A locus, perhaps through direct inhibition olinska Institute, for expertise in immuno¯uorescence of p16 and/or p14ARF. microscopy techniques. This study was supported by the Strikingly, we have observed that BL lines having Swedish Cancer Society, the Cancer Society of Stockholm, CDKN2A deletions or p53 mutations grow more and the Karolinska Institute.

References

AdamsJM,HarrisAW,PinkertCA,CorcoranLM, GaidanoG,BallerininP,GongJZ,InghiramiG,NeriA, Alexander WS, Cory S, Palmiter RD and Brinster RL. Newcomb EW, Magrath IT, Knowles DM and Dalla- (1985). Nature, 318, 533 ± 538. Favera R. (1991). Proc.Natl.Acad.Sci.USA,88, 5413 ± Alkema MJ, Bronk M, Verhoeven E, Otte A, van't Veer LJ, 5417. Berns A and van Lohuizen M. (1997). Dev., 11, Gardie B, Cayuela JM, Martini S and Sigaux F. (1998). 226 ± 240. Blood, 91, 1016 ± 1020. Baur AS, Shaw P, Burri N, Delacretaz F, Bozman FT and Gronbaek K, de Nully Brown P, Moller MB, Nedergaard T, Chaubert P. (1999). Blood, 94, 1773 ± 1781. Ralfkiaer E, Moller P, Zeuthen J and Guldberg P. (2000). Bhatia KG, Gutierrez MI, Huppi K, Siwarski D and Leukemia, 14, 1727 ± 1735. Magrath IT. (1992). Cancer Res., 52, 4273 ± 4276. Gutierrez MI, Cherney B, Hussain A, Mostowski H, Tosato Capoulade C, Bressac-de Paillertes B, Lefrere I, Ronsin M, G, Magrath I and Bhatia K. (1999). Cancer Res., 59, 696 ± Feunteun J, Tursz T and Wiels J. (1998). Oncogene, 16, 703. 1603 ± 1610. Gutierrez MI, Bhatia K, Cherney B, Capello D, Gaidano G Cherney BW, Bhatia KG, Sgadari C, Gutierrez MI, and Magrath I. (1997). Ann. Oncol., 10, 987 ± 994. MostowskiH,PikeSE,GuptaG,MagrathITandTosato Haupt Y, Bath ML, Harris AW and Adams Jm. (1993). G. (1997). Cancer Res., 57, 2508 ± 2515. Oncogene, 11, 3161 ± 3164. Cordon-Cardo C, Latres E, Drobnjak M, Oliva MR, Pollack Honda R and Yasuda H. (1999). EMBO J., 18, 22 ± 27. D, Woodru€ JM, Marechal V, Chen J, Brennan MF and Jacobs JJ, Kieboom K, Marino S, DePinho RA and van Levine AJ. (1994). Cancer Res., 54, 794 ± 799. Lohuizen M. (1999a). Nature, 397, 164 ± 168. Eischen CM, Weber JD, Roussel MF, Sherr CJ and Jacobs JJ, Scheijen B, Voncken JW, Kieboom K, Berns A Cleveland JL. (1999). Genes Dev., 13, 2658 ± 2669. and van Lohuizen M. (1999b). Genes Dev., 13, 2678 ± 2690. Esteller M, Tortola S, Toyota M, Capella G, Peinado MA, Kamijo T, Weber JD, Zambetti G, Zindy F, Roussel MF and Baylin SB and Herman JG. (2000). Cancer Res., 60, 29 ± Sherr CJ. (1998). Proc. Natl. Acad. Sci. USA, 95, 8292 ± 33. 8297. Farell PJ, Allan GJ, Shanahan F, Vousden KH and Crook T. Klangby U, Okan I, Magnusson KP, Wendland M, Lind P (1991). EMBO J., 10, 2879 ± 2887. and Wiman KG. (1998). Blood, 91, 1680 ± 1687. Gao C, Nakajima T, Taya Y and Tsuchida N. (1999). Klein G. (1993). Gene, 135, 189 ± 196. Biochem. Biophys. Res. Commun., 264, 860 ± 864.

Oncogene p14ARF and MDM2 alterations in Burkitt lymphoma MS LindstroÈm et al 2177 Kumar R, Sauroja I, Punnonen K, Jansen C and Hemminki Stranks G, Height SE, Mitchell P, Jafadyel D, Yuille MAR, K. (1998). Genes Chrom. Cancer, 27, 273 ± 277. DeLord C, Clutterbuck RD, Treleaven JG, Powles RL, LindstroÈ m MS, Klangby U, Inoue R, Pisa P, Wiman KG and Nacheva E, Oscier DG, Karpas A, Lenoir GM, Smith SD, Asker C. (2000). Exp. Cell Res., 256, 400 ± 410. Millar JL, Catovsky D and Dyer MJS. (1995). Blood, 85, Maestro R, Dei Tos AP, Hamamori Y, Karsnokutsky S, 893 ± 901. Sartorelli V, Kedes L, Doglioni C, Beach DH and Hannon Tao W and Levine AJ. (1999). Proc. Natl. Acad. Sci. USA, GJ. (1999). Genes Dev., 13, 2207 ± 2217. 12, 6937 ± 6941. Oliner JD, Kinzler KW, Meltzer PS, George DL and Voncken JW, Schweizer D, Aagaard L, Sattler L, Jantsch Vogelstein B. (1992). Nature, 358, 80 ± 83. MR and van Lohuizen M. (1999). J. Cell Sci., 112, 4627 ± Pinyol M, Hernandez L, Martinez A, Cobo F, Hernandez S, 4639. Bea S, Lopez-Guillermo A, Nayach I, Palacin A, Nadal A, Vousden K, Crook T and Farrell PJ. (1993). Journal of Gen. Fernandez PL, Montserrat E, Cardesa A and Campo E. Virology, 74, 803 ± 810. (2000). Am.J.Pathol.,156, 1987 ± 1996. Watanabe T, Hotta T, Ichikawa A, Kinoshita T, Nagai H, Pomerantz J, Schreiber-Agus N, Liegeois NJ, Silverman A, Uchida T, Murate T and Saito H. (1994). Blood, 84, 3158 ± Alland L, Chin L, Potes J, Chen K, Orlow I, Lee HW, 3165. Cordon-Cardo C and DePhino RA. (1998). Cell, 92, 713 ± Weber JD, Taylor LJ, Roussel MF, Sherr CJ and Bar-Sagi 723. D. (1999). Nat. Cell Biol., 1, 20 ± 26. Quelle DE, Zindy F, Ashmun RA and Sherr CJ. (1995). Cell, Weber JD, Kuo ML, Bothner B, DiGiammarino EL, 83, 993 ± 1000. Kriwacki RW, Roussel MF and Sherr CJ. (2000). Mol. Randerson-Moor JA, Harland M, Williams S, Cuthbert- Cell. Biol., 7, 2517 ± 2528. Heavens D, Sheridan E, Aveyard J, Sibley K, Whitaker L, Wiman KG, Magnusson KP, Ramqvist T and Klein G. Knowles M, Newton-Bishop JN and Bishop DT. (2001). (1991). Oncogene, 6, 1633 ± 1639. Hum. Mol. Genet., 10, 55 ± 62. Zhang S, Ramsay ES and Mock BA. (1998). Proc. Natl. Robertson KD and Jones PA. (1998). Mol. Cell. Biol., 18, Acad.Sci.USA,95, 2429 ± 2434. 6457 ± 6473. Zhang SL, DuBois W, Ramsay ES, Bliskovski V, Morse 3rd Schmitt CA, McCurrach ME, de Stanchina E, Wallace- HC, Taddesse-Heath L, Vass WC, DePinho RA and Mock Brodeur RR and Lowe SW. (1999). Genes Dev., 13, 2670 ± BA. (2001). Mol. Cell. Biol., 21, 310 ± 318. 2677. Zhang Y, Xiong Y and Yarbrough WG. (1998). Cell, 92, Sharpless NE and DePinho RA. (1999). Curr. Opin. Genet. 725 ± 734. Dev., 9, 22 ± 30. Zhang Y and Xiong Y. (1999). Mol. Cell, 3, 579 ± 591. Sherr CJ and Weber JD. (2000). Curr.Opin.Genet.Dev.,10, 94 ± 99. StottFJ,BatesS,JamesMC,McConnellBB,StarborgM, Brookes S, Palmero I, Ryan K, Hara E, Vousden KH and Peters G. (1998). EMBO J., 17, 5001 ± 5014.

Oncogene