Aberrant NF-Jb2/P52 Expression in Hodgkin/Reed–Sternberg Cells and CD30-Transformed Rat ﬁbroblasts

Aberrant NF-jB2/p52 expression in Hodgkin/Reed–Sternberg cells and CD30-transformed rat ﬁbroblasts

Mizuho Nonaka1, Ryouichi Horie2, Kinji Itoh3, Toshiki Watanabe4, Naoki Yamamoto1 and Shoji Yamaoka*,1

1Department of Molecular Virology, Graduate School of Medicine, Tokyo Medical and Dental University, 1-5-45 Bunkyo-ku, Tokyo 113-8510, Japan; 2Fourth Department of Internal Medicine, School of Medicine, Kitasato University, Kanagawa, Japan; 3Second Department of Pathology, School of Medicine, Toho University, Tokyo, Japan; 4Division of Pathology, Department of Cancer Research, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan

Overexpression of CD30and constitutive nuclear factor- containing NF-kB essential modulator (NEMO, also jB (NF-jB) activation are hallmarks of the malignant calledIKK g) for NF-kB activation (Yamaoka et al., Hodgkin Reed–Sternberg (H-RS) cells. Previous investi- 1998; Makris et al., 2002; Ben-Neriah andSchmitz, gations have demonstrated that both proliferation and 2004). The canonical NF-kB signaling pathway involves survival of H-RS cells require constitutive NF-jB activity, IKK2 (also calledIKK b) andleadsto nuclear translo- which is comprised of the p50and RelA subunits. We cation of RelA-containing NF-kB dimers (Ben-Neriah report here enhanced expression of NF-jB2/p52 and andSchmitz, 2004). In contrast, B-cell activating factor RelB-containing NF-jB DNA-binding activity in Epstein– receptor (BAFF-R), CD40, or TNF-like weak inducer Barr virus-negative H-RS cells. Kinetic studies revealed of apoptosis (TWEAK) were reportedto activate the that a proteasome inhibitor MG132 induced p100 noncanonical NF-kB signaling pathway, which did accumulation with reduced p52 expression in H-RS cells, not necessarily require NEMO for NF-kB activation suggesting proteasome-dependent processing of p100. In (Claudio et al., 2002; Coope et al., 2002; Kayagaki et al., addition, treatment with a protein synthesis inhibitor 2002; Saitoh et al., 2003). The noncanonical NF-kB cycloheximide rapidly downregulated inhibitor of NF-jB signaling pathway involves IKK1 (also calledIKK a) (IjB) kinase activity in H-RS cells. We also demonstrate andleadsto the phosphorylation andprocessing of that overexpression of CD30in rat fibroblasts at levels p100, resulting in nuclear translocation of RelB-contain- comparable to those in H-RS cells results in constitutive ing NF-kB dimers (Solan et al., 2002; Derudder et al., IjB kinase activation, proteasome-dependent p100 pro- 2003; Ben-Neriah andSchmitz, 2004). cessing, and NF-jB-dependent cell transformation. Our Hodgkin’s lymphoma (HL) is characterized by a low results thus indicate that CD30triggers the noncanonical frequency of malignant Hodgkin and Reed–Sternberg NF-jB activation pathway, and suggest that deregulated (H-RS) cells in a reactive backgroundof non-neoplastic CD30signaling contributes to the neoplastic features of cells. (Horie et al., 2003; Thomas et al., 2004). The H-RS cells. peculiar activatedphenotype of H-RS cells andtheir Oncogene (2005) 24, 3976–3986. doi:10.1038/sj.onc.1208564 pattern of cytokine secretion, which contribute to the Publishedonline 14 March 2005 pathological appearance andclinical manifestations of the disease, are believed to be consequences of Keywords: Hodgkin/Reed–Sternberg cells; CD30; NF- constitutive NF-kB activation (Skinnider and Mak, kB2; IKK 2002). Inhibition of constitutive NF-kB ledto decreased proliferation, enhancedapoptotic response, and strongly impairedtumor growth in immuno-deficient mice (Bargou et al., 1997; Hinz et al., 2002). These Introduction observations suggestedan important role for NF- kBin the pathogenesis of HL. HL is often associatedwith The nuclear factor-kB (NF-kB) family of transcription Epstein–Barr virus (EBV), andseveral research groups factors plays a pivotal role in inflammation, prolifera- including ours have recently demonstrated that the tion, andprevention of apoptosis (Li andVerma, 2002; EBV-encoded latent membrane protein-1 activates the Ben-Neriah andSchmitz, 2004). It has been reported canonical andnoncanonical NF- kB signaling pathways that stimulation by cytokines, such as tumor necrosis (Atkinson et al., 2003; Eliopoulos et al., 2003; Saito factor (TNF)-a or IL-1b, requires the canonical et al., 2003). Of note, constitutive IKK activation was inhibitor of NF-kB(IkB) kinase (IKK) complex observedboth EBV-positive and-negative HL cell lines (Krappmann et al., 1999; Mathas et al., 2003). *Correspondence: S Yamaoka; E-mail: [email protected] CD30 is one of the TNF receptor (TNFR) family Received6 September 2004; revised24 January 2005; accepted24 members implicatedin lymphoidcell maturation as well January 2005; publishedonline 14 March 2005 as in the development of hematologic malignancies CD30 triggers noncanonical NF-jB activation M Nonaka et al 3977 including HL, anaplastic large-cell lymphoma (ALCL), p100 andp52 expression by immunoblotting (Figure 1c). andimmunoblastic lymphoma, but the biological As positive andnegative controls, we usedMT-2 and consequences of CD30 expression in the pathogenesis Jurkat cells, respectively. MT-2 cells were established in of these diseases have not been well understood (Opat vitro andexpress the HTLV-I Tax protein, which is andGaston, 2000; SchneiderandHu ¨ binger, 2002; Horie known to induce canonical NF-kB activation andthe et al., 2003). We previously showedthat overexpression processing of p100 through its physical binding to p100 of CD30 ledto recruitment andself-aggregation of TNF (Xiao et al., 2001; Hironaka et al., 2004). As shown in receptor-associatedfactor (TRAF)2 andTRAF5, re- Figure 1c, MG132 treatment induced transient p100 sulting in constitutive NF-kB activation, independently accumulation in L540 andL428 cells andreducedp52 of the CD30 ligand(Horie et al., 2002). A high level of expression in MT-2, HDLM-2, L428, andL540 cells. CD30 expression is a phenotype sharedwith ALCL The effect of MG132 on p100 andp52 levels was not so (Stein et al., 2000). However, we have quite recently striking in tumor cells when comparedto cells stimu- demonstrated that constitutive NF-kB activation is not latedwith cytokines (Coope et al., 2002; Dejardin et al., observedin ALCL cells because the nucleophosmin 2002), because of the presence of p52 prior to the (NPM)-anaplastic lymphoma kinase (ALK) chimeric MG132 treatment. These results indicate that the protein, which is characteristic of most systemic/nodal generation of p52 in H-RS cells involves proteasome- ALCL, abrogates CD30-driven NF-kB activation (Horie dependent processing of NF-kB2 p100. et al., 2004). In this study, we report enhanced proteasome- Protein synthesis inhibition impairs IKK activity in H-RS dependent processing of p100 to p52 in EBV-negative cells in correlation to the p100 status H-RS cell lines. We also show that overexpression of CD30 results in transformation of rat fibroblasts as To investigate the mechanism of IKK activation in H- well as constitutive NF-kB activation similar to that in RS cells, we assessedeffects of a protein synthesis H-RS cells, suggesting contribution of deregulated inhibitor cycloheximide (CHX) on IKK activity in H- CD30 signaling to malignant cell transformation of RS cells, because it was previously shown that prior H-RS cells. treatment with CHX inhibitedsignal-inducedprocessing of p100, but not the canonical NF-kB activation (Coope et al., 2002; Mu¨ ller andSiebenlist, 2003). Whole-cell extracts were preparedfrom MT-2 andH-RS cells Results treatedwith or without CHX for 4 h, andthen IKK1- Aberrant expression of NF-kB p52 and RelB in H-RS associatedkinase activity was determinedby in vitro cells kinase assay (Figure 2). Addition of CHX led to a markedreduction of IKK activity in L540 cells, while It was reportedthat H-RS cell lines as well as primary HDLM-2 andL428 cells showedmoderate reduction of tumor cells from Hodgkin disease patients showed IKK activity in the presence of CHX. CHX hadno constitutive NF-kB DNA-binding activity comprised appreciable effect on IKK activity in MT-2 cells, of p50 andRelA (Bargou et al., 1997). We analysedthe whereas Tax activates IKK in a NEMO-dependent DNA-binding NF-kB components in Jurkat, human T- manner. Parallel immunoblotting revealedequivalent cell leukemia virus type I (HTLV-I)-infectedMT-2 and amounts of IKK1 in the immunoprecipitates from these EBV-negative H-RS cell lines by supershift assay cells. Thus, the relative contribution of the CHX- (Figure 1a). The poor NF-kB DNA-binding activity of sensitive pathway to IKK activation was varied Jurkat cells containedmainly p50 andc-Rel, while all depending on the H-RS cell lines, and the levels of the examinedcomponents were includedinthe strong p100 expression in these cells (Figure 1b) appearedto be NF-kB-DNA-binding activity of MT-2 cells. In the H- correlatedwith their CHX-resistant IKK activity. RS cells, the DNA-binding NF-kB components were composednot only of p50 andRelA but also of RelB Transforming property of CD30 and p52 in the H-RS cells. Indeed, an immunoblot study revealedaberrant expression of p100 as well as its We recently reportedthat overexpression of CD30 in processedform p52 in H-RS cells (Figure 1b). Parallel 293 cells induced self-association of CD30, leading to immunoblotting of 293 T cells expressing p100 and/or persistent NF-kB activation independently of the CD30 NF-kB-inducing kinase (NIK), which is known to ligand(Horie et al., 2002). To examine if overexpression trigger IKK1 activation andsubsequent phosphoryla- of CD30 results in cell transformation in an NF-kB- tion-dependent p100 processing to p52, served as dependent manner, CD30 was expressed in a subline of positive andnegative controls (Figure 1b). Although Rat-1, B5, that carries integratedform of pIg k2bsrH expression of p100 andp52 in L540 cells was relatively (Yamaoka et al., 1998) andbecomes resistant to weak among H-RS cells, L540 cells expressedmore p52 blasticidin S under constitutively elevated NF-kB acti- than Jurkat or vector-transfected293 T cells, in which vity (Chinanonwait et al., 2002). B5 cells were infected p52 was undetectable. To know whether the processing with retrovirus capable of expressing human CD30 and of NF-kB2 p100 to p52 is proteasome-dependent in H- culturedin the presence of puromycin for a couple of RS cells, we useda proteasome inhibitor MG132 and days to establish a pool of infected cells. At this analysedH-RS cells (HDLM-2, L540, andL428) for stage, pooledcells didnotshow apparently elevated

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3978

Figure 1 Aberrant NF-kB2/p52 expression in H-RS cells. (a) The subunit composition of constitutive NF-kB activity in Jurkat, MT- 2, andH-RS cells were analysedby electrophoretic mobility shift assay (EMSA) with preimmune (PI) serum, anti-p50 antibody,anti- RelA serum, anti-RelB serum, anti-c-Rel antibody, or anti-p52 serum. (b) 293 T cells were transfectedwith pCn vector, pCn100, pRK-5 vector, and/or pRK-NIK by the calcium phosphate precipitation method. Whole-cell extracts (30 mg) of 293 T, Jurkat, MT-2, andH- RS cells were subjectedto SDS–PAGE andimmunoblotting (IB) with anti-p52, anti-NIK, or anti-actin antibody.( c) Jurkat, MT-2, andH-RS cells were treatedfor 4 or 8 h with MG132 (20 mM) or DMSO. After treatment, whole-cell extracts were subjectedto SDS– PAGE and IB with anti-p52 or anti-actin antibody (left panels). The nonspeciﬁc band indicated above the p52 band was more intense in L428, L540, andJurkat cell results because demonstration of relatively weaker expression of p100/p52 in these cells requireda longer exposure time. The bar graphs (right panels) show relative bandintensity of p100 (ﬁlledbar) andp52 (open bar) normalizedto the corresponding actin bandintensity. A value of 1.0 represents normalizedp100 or p52 bandintensity at time 0 h

NF-kB-speciﬁc DNA-binding activity or CD30 expres- This was not causedby a nonspeciﬁc retroviral sion comparable to that seen in H-RS cell lines (data not mutagenesis, because there was absolutely no survival shown). Retrovirus-transduced B5 cells were then with vector-transduced B5 (B5-vector 1) cells after subjectedto selection with blasticidineS, which resulted blasticidin S selection. Therefore, we used in the in survival of CD30-transduced B5 (B5-CD30) cells. following experiments a puromycin-resistant pool of

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3979 reduction of IKK activity as observed for L540 H-RS cells treatedwith CHX (Figure 2). NF- kB DNA-binding activity in B5-CD30 clone 6 cells was not enhanced further in response to stimulation for 8 h by the CD30’s ligandCD153 expressedon fixedmouse cells (Figure 5e and f), while this stimulation induced ligand-dependent NF-kB activation in B5 cells expressing CD30 at lower Figure 2 IKK activity in H-RS cells is sensitive to protein levels (B5-CD30 pool) (Figure 5dandf). Similar results synthesis inhibition. Cytoplasmic extracts of MT-2 andH-RS cells were obtainedwith other B5-CD30 cell clones (datanot treatedfor 4 h with or without cycloheximide(CHX; 50 mg/ml) shown). NF-kB DNA-binding components induced by were immunoprecipitatedwith IKK1-specific antibody(556532) CD153 stimulation in B5-CD30 pool cells included andsubjectedto in vitro kinase assay using GST-IkBa as substrate. IKK1 in the immunoprecipitates was detected with IKK1-specific relatively more RelA andless RelB comparedto those in antibody (H-744) B5-CD30 clone 6 cells (Figure 5g), due to the initial canonical NF-kB activation anddelayedtriggering of the noncanonical pathway. Another piece of evidence supporting the noncanonical NF-kB activation by CD30 is that transient CD30 expression activatedan vector-transduced B5 (B5-vector 1) cells as a control. NF-kB-dependent promoter in an NEMO-deficient B5-CD30 cells isolatedfollowing blasticidineS selection subline of Rat-1, 5R, in which latent membrane protein exhibitedconstitutive NF- kB DNA-binding activity and 1 of EBV, but not TNFa, activatedNF- kB (Saito et al., aberrant p52 expression (Figure 3a). Flow cytometric 2003) (Figure 6). analysis revealedthat B5-CD30 cells expressedCD30 at levels comparable to those for H-RS cells (Figure 3b and c). Since B5-CD30 cells showedalteredmorphologies Discussion andloss of the contact inhibition on monolayer (Figure 4), we testedthe ability of B5-vector and High levels of CD30 expression, constitutive NF-kB B5-CD30 cells to form colonies in soft agar (Table 1). activation, and cell transformation B5-vector 1 cells did not form significant number of colonies larger than 60 mm in diameter, while B5-CD30 Constitutive NF-kB activity in H-RS cells was reported cell clones efficiently formedlarge colonies in soft agar to contribute to the aberrant cytokine gene expression (Figure 4 and Table 1). These data indicate that andneoplastic features of H-RS cells (Hinz et al., 2002; overexpression of CD30 induces transformation of rat Skinnider and Mak, 2002). Enhanced expression of fibroblasts. This transformedphenotype was completely many NF-kB target genes including chemokines, cyto- reversedby retroviral expression of a super-repressor kines, andapoptosis regulators was shown in H-RS cells form of IkBa (SR-IkBa) (Figure 4 andTable 1), (Skinnider and Mak, 2002; Ku¨ ppers et al., 2003) and indicating that this transformation depends on the abrogation of NF-kB activity induced apoptosis in H- constitutive NF-kB activity induced by CD30. As RS (HDLM-2) cells (Hinz et al., 2002). These observa- expected, the high NF-kB DNA-binding activity was tions suggestedan important role for NF- kB in the lost in SR-IkBa-transduced B5-CD30 cells (Figure 3a), pathogenesis of HL. As for mechanistic explanations, while the expression levels of CD30 on SR-IkBa- Krappmann et al. (1999) revealedconstitutive IKK transduced B5-CD30 cells remained similar to those activity in the HDLM-2, L1236, andKMH-2 cell lines, on B5-CD30 cells (Figure 3b). In addition, p100 and p52 while mutations in the IkBa or IkBe genes as well as virtually disappeared in SR-IkBa-transduced B5-CD30 amplification of the NF-kB/REL locus were also cells, presumably as a consequence of loss of NF-kB- reported(Cabannes et al., 1999; Joos et al., 2002; dependent p100 production. Supershift analyses re- Martı´ n-Subero et al., 2002; Barth et al., 2003; Emmerich vealedthat NF- kB DNA-binding components induced et al., 2003). In fact, among the three EBV-negative cell in B5-CD30 clone 6 cells included RelB (Figure 5a). We lines usedin this study,L428 expresses inactive I kBa were unable to examine the presence of p52 in the andI kBe, while the other two express intact IkB NF-kB DNA-binding components, because antibody proteins. Loss of functional IkBs may facilitate translo- capable of surpershifting rat p52 was not available. cation of RelA-containing NF-kB complexes, but the We next examinedeffects of MG132 on the p100 and processing of p100 causedby constitutively elevated p52 levels in B5-CD30 clone 6 cells by immunoblotting. IKK1-associatedkinase activity in H-RS cells including B5-CD30 clone 6 cells aberrantly expressedp100 as well L428 (Figures 1 and2) cannot be explainedsolely by as p52 compared with control cells. Addition of MG132 inactivation of the IkB proteins. While elevatedexpres- resultedin a markeddecrease in p52, while the level of sion of mRNA for p100 in H-RS cells was previously p100 remainedhigh (Figure 5b). We further assessedif reported(Hinz et al., 2002), this report has demon- CHX treatment affects IKK activity in B5-CD30 clone 6 stratedaberrant p52 expression in H-RS cells. The cells. Cells were treatedwith or without CHX for 4 h, elevatedIKK1-associatedkinase activity andprocessing andthen IKK1-associatedkinase activity was deter- of p100 in H-RS cells (Figures 1 and2) are compatible minedby in vitro kinase assay (Figure 5c). Addition of with the recently highlightednoncanonical NF- kB acti- CHX to B5-CD30 clone 6 cells ledto a remarkable vation, which is characterized by NEMO-independent

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3980

Figure 3 Overexpression of CD30 induces constitutive NF-kB activity in rat fibroblasts. (a) Upper panels: nuclear extracts (5 mg) from B5 cells infectedwith pMRX-puro (B5-vector) or pMRX-CD30-puro (B5-CD30 clones 1, 5, and6) andSR-I kBa-transduced B5-vector or B5-CD30 clones (1, 5, and6) were incubatedwith 32P-labeledNF- kB or Oct-1 probe andsubjectedto EMSA. Lower panels: whole- cell extracts (30 mg) of B5 infectants were subjectedto SDS–PAGE andimmunoblotting (IB) with anti-p52, anti-I kBa, or anti-actin antibody. (b) FACS analyses of CD30 expression on B5 infectants. Cells were incubatedwith FITC-labeledanti-CD30 antibody.Open histogram with thick line represents B5-vector or B5-CD30 clones, filledhistogram represents SR-I kBa-transduced B5-vector or B5- CD30 clones, andopen histogram with thin line represents pMX-IRES-Zeocin-transducedB5-vector or B5-CD30 clones. ( c) FACS analyses of CD30 expression on H-RS cells. Cells were incubatedwith FITC-labeledirrelevant mouse IgG (filledhistogram) or anti- CD30 antibody (open histogram)

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3981

Figure 4 Overexpression of CD30 causes NF-kB-dependent transformation of rat ﬁbroblasts. (a–f) Phase-contrast micrographs of B5-vector (a and e) andCD30-transducedB5-CD30clones 1 ( b and f), 5 (c and g), and6 ( d and h) cells culturedon monolayer ( a–d)or in soft agar for 2 weeks (e–h). (i–p) Phase-contrast micrographs of SR-IkBa-transduced B5-vector (i and m) or B5-CD30 clones 1 (j and n), 5 (k and o), and6 ( l and p) cells culturedon monolayer ( i–l) or in soft agar for 2 weeks (m–p)

Table 1 Overexpression of CD30 causes NF-kB-dependent transfor- We previously demonstrated that overexpressed mation of rat fibroblasts CD30 triggeredself-multimerization andrecruitment Cells Colony-forming Average size of of TRAF proteins, leading to ligand-independent NF- efficiency (%) colonies (mm) kB activation (Horie et al., 2002). The present report demonstrates for the first time that overexpression of B5-vector 1 0.3 61.5 B5-CD30 clone 1 6.5 186.5 CD30 in rat fibroblasts at levels similar to those in H-RS B5-CD30 clone 5 12.6 402.5 cells leads to ligand-independent transformation accom- B5-CD30 clone 6 4.3 94.5 paniedby CHX-sensitive constitutive IKK activation andinduction of RelB-containing NF- kB complexes, B5-vector 1-vector 2 0.5 60.8 the features that can also be seen in H-RS cells (Table 1, B5- CD30 clone 1-vector 2 11.3 97.5 B5-CD30 clone 5-vector 2 7.4 148.0 Figures 3, 4 and5). These observations suggest that B5-CD30 clone 6-vector 2 12.0 117.0 CD30 is a likely candidate that participates in the constitutive NF-kB activation anddevelopment of the B5-vector 1-SR-IkBa 0.2 60.5 oncogenic phenotype foundin EBV-negative H-RS B5-CD30-clone 1-SR-IkBa 0.7 62.0 B5-CD30 clone 5-SR-IkBa 0.5 61.5 cells. Since two other TNF receptor family proteins B5-CD30 clone 6-SR-IkBa 0.5 62.5 CD40 andRANK are also reportedto be expressedin H-RS cells (Carbone et al., 1995; Annunziata et al., Cells were culturedin 0.33% agarose containing regular mediumfor 2 2000; Fiumara et al., 2001; Ku¨ ppers et al., 2003) and weeks. Colonies larger than 60 mm were countedmicroscopically and implicatedin the processing of p100 (Coope et al., 2002; the average size of colonies is indicated Novack et al., 2003), it wouldbe intriguing to examine possible contribution of these molecules to the NF-kB induction of RelB-containing DNA-binding complexes, activity in H-RS cells. Several lines of evidence suggest long duration, and susceptibility to protein synthesis that CD30 supports survival of H-RS cells in multiple inhibition (Ben-Neriah andSchmitz, 2004). ways. CD30 stimulation induced upregulation of the

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3982

Figure 5 CD30 transformation leads to NF-kB activation similar to H-RS cells. (a) Nuclear extract (5 mg) of B5-CD30 clone 6 cells was incubatedwith preimmune (PI), anti-p50, anti-RelA, or anti-RelB serum andthen subjectedto EMSA. The positions of supershifted bands are indicated. (b) Whole-cell extracts were preparedfrom B5-vecor 1 or B5-CD30 clone 6 cells treatedfor 4 h with MG132 or DMSO, andsubjectedto SDS–PAGE andimmunoblotting with anti-p52 or anti-actin antibody.( c) Cytoplasmic extracts of B5-vector 1 andB5-CD30 clone 6 cells treatedfor 4 h with or without cycloheximide(CHX; 50 mg/ml) were immunoprecipitatedwith IKK1-specific antibody (H-744) and subjected to in vitro kinase assay using GST-IkBa as substrate. IKK1 in the immunoprecipitates was detected by immunoblotting with IKK1-specific antibody (IMG-136). (d) FACS analyses of CD30 expression on B5 infectants. Cells were incubatedwith FITC-labeledanti-CD30 antibody.Filledhistogram with thick line represents B5-vector cells; open histogram with thin line represents B5-CD30-pool cells; open histogram with thick line represents B5-CD30 clone 6 cells. (e) FACS analyses of CD153 expression on mouse SV-T2 cells. Cells were incubatedwith irrelevant mouse IgG (filledhistogram) or anti-CD153 antibody (open histogram) and then stained with FITC-conjugated anti-mouse IgG antibody. (f) Nuclear extracts (5 mg) preparedfrom B5-vector, B5-CD30-pool, andB5-CD30 clone 6 cells stimulatedfor 8 h with PFA-fixedSV-T2-vector or SV-T2-CD153 cells were incubatedwith 32P-labeledNF- kB or Oct-1 probe andsubjectedto EMSA. ( g) Nuclear extracts (5 mg) of B5-CD30-pool (left panel) and B5-CD30 clone 6 cells (right panel) after stimulation with PFA-fixedSV-T2-CD153 cells were incubatedwith PI, anti-p50, anti-RelA, or anti-RelB serum andthen subjectedto EMSA, using 32P-labeledNF- kB probe

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3983 to p100. Annunziata et al. (2000) reportedthat CD40- induced NF-kB activation was mediated by proteolysis of TRAF3 in H-RS (KM-H2) cells. These reports suggest that TRAF3 may be involvedin constitutive NF-kB activity in H-RS cells. Obviously, further efforts will be requiredto elucidateroles of TRAF3 andNIK in the regulation of IKK andp100 processing in H-RS cells. In summary, this study has demonstrated that overexpression of CD30 induces ligand-independent cell transformation andtriggers the noncanonical NF- kB signaling, which is also observedin H-RS cells. Our results suggest that overexpression of CD30 may contribute to the neoplastic phenotype of HL.

Figure 6 CD30 activates NF-kB in the absence of NEMO. Cells were transiently cotransfectedwith 0.25 mgofIgkCona-luc, 0.25 mg of pEF-1 LacZ, and1 mg pME (vector) or 1 mg pME-CD30 (CD30). Materials and methods Luciferase activity was determined 36 h after transfection and normalizedon the basis of b-galactosidase activity. Each column Cells and viruses represents the mean with s.d. of three independent experiments HDLM-2, L428, andL540 cell lines were purchasedfrom the German Collection of Micro-organisms andCell Cultures (Braunschweig, Germany). MT-2 is a human T cell line inhibitor of apoptosis genes cIAP1 andcIAP2 in transformed in vitro with HTLV-I (Sugamura et al., 1984). HDLM-2 cells andALCL cells (Hu¨ binger et al., 2004). These cells were maintainedin RPMI 1640 supplementedwith Activation of CD30, CD40, or RANK in H-RS cells by 10% fetal bovine serum andantibiotics. Rat-1, 5R, B5, SV-T2, their respective ligands triggered MEK/ERK pathway 293T, andPLAT-E packaging cells were previously described andpromotedH-RS cell survival (Zheng et al., 2003). and maintained in Dulbecco’s modified Eagle’s medium Thus, CD30 appears to support survival of H-RS cells (DMEM) supplementedwith 10% fetal bovine serum and antibiotics. (Yamaoka et al., 1998; Morita et al., 2000; not only through persistent NF-kB activation but also Takahashi et al., 2001; Chinanonwait et al., 2002). For through activation of the MEK/ERK pathway. production of retroviruses, PLAT-E cells were transfected using FuGENE 6 (Roche Diagnostics Co., IN, USA) How does the processing of p100 contribute to constitutive according to the manufacturer’s instructions. Culture super- NF-kB activity? natants of PLAT-E cells were collected48 h after transfection. Approximately 3 Â 105 B5 cells were infectedwith retrovirus The processing of p100 to p52 is often associatedwith capable of expressing human CD30 or empty vector (vector 1). constitutive NF-kB activity in human malignancies Cells were selectedin the presence of 2 mg/ml of puromycin and andcytokine-inducedlong-lasting NF- kB activation pools of infectedcells (B5-vector 1 or B5-CD30-pool) were (Dejardin et al., 1999; Cogswell et al., 2000; Ben-Neriah established. Infected cells were then subjected to selection with 5 mg/ml of Blasticidin S. Individual 11 cell clones were isolated andSchmitz, 2004; Chen andGreene 2004; Hironaka (B5-CD30 clones) andusedfor further experiments. B5-vector et al., 2004), suggesting that p100 processing contributes 1 andB5-CD30 clone cells were infectedwith retrovirus to the temporal control of NF-kB activity. In mamma- capable of expressing a super-repressor form of IkBa (SR- lian cells, p100 was shown to have powerful IkB activity IkBa) or empty vector (vector 2) with a similar procedure. on RelA andRelB (Solan et al., 2002). Derudder et al. Pools of SR-IkBa-transduced B5-vector 1 or B5-CD30 cells (2003) demonstrated that p100 interacted with and (B5-vector 1-SR-IkBa or B5-CD30-SR-IkBa) were established inhibitedDNA bindingof RelB-containing complexes following selection of infectedcells with 400 mg/ml of Zeocin. quickly after TNFa stimulation, while lymphotoxin b SV-T2-vector andSV-T2-CD153 cells were establishedby signaling involved p100 processing and did not induce infection of SV-T2 cells with retrovirus preparedwith pMRX- such inhibition by p100 (Derudder et al., 2003). The puro andpMRX-CD153-puro, respectively. SV-T2-vector and SV-T2-CD153 cells were treatedwith a buffer (20 m M HEPES expression of p100 is induced upon NF-kB activation, (pH7.4), 100 mM NaCl, 10 mM EDTA, 0.5% BSA) for 20 min while its processing is usually regulatedtightly in a NIK- at 41C and10 min at 37 1C, dislodged by vigorous pipetting, , IKK1- and proteasome-dependent manner. This helps fixed in PBS containing 4% paraformaldehyde (PFA) for to ensure the rapiddownregulation of NF- kB activity 15 min at room temperature, andthen washedthree times by following NF-kB activation by TNFa,IL1b, or LPS. PBS before being usedfor stimulation. PFA-fixedSV-T2- Thus, abrogation of this inducible autoregulated in- vector or SV-T2-CD153 cells were added to the culture of hibitory pathway is likely to be requiredfor persistent B5-vector1, B5-CD30-pool, or B5-CD30 clone cells, andthen NF-kB activation foundin tumor cells. Regarding the cells were lysedafter 8 h of incubation. 293 T cells were regulation of IKK activity leading to p100 processing, transfectedby the calcium phosphate precipitation method. Liao et al. (2004) recently showedthat inductionof p100 processing by extracellular signals involves degradation Reagents of TRAF3 accompaniedby enhancedexpression of Anti-p52 serum (06-413) usedfor supershift assay was NIK, which serves as a docking molecule recruiting IKKa purchasedfrom Upstate Biotechnology, Inc., NY, USA.

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3984 Anti-p50 recognizing mouse p50, anti-RelA, andanti-RelB Preparation of cell extracts sera were kindly provided by Drs N Rice and A Israe¨ l (Institut For preparation of whole-cell extracts, cells were suspended in Pasteur de Paris, France). Anti-p50 (H-119), anti-c-Rel (N), RIPA buffer (20 mM Tris–HCl (pH 8.0), 137 mM NaCl, 1 mM anti-IKK1 (H-744), anti-IkBa (C-21), anti-actin (C-2), anti- MgCl ,1mM CaCl , 10% glycerol, 1% Nonidet P-40, 0.5% NIK (H-248), andanti-p52 antibodies(C-5) were purchased 2 2 deoxycholate, and 0.1% SDS), supplemented with 0.5 mM from Santa Cruz Biotechnology, CA, USA. Anti-IKK1 phenylmethylsulfonyl ﬂuoride (PMSF), 1 mg/ml leupeptin and antibody (IMG-136) was purchased from IMGENEX (CA, 1 mg/ml aprotinin. Extracts were clearedby centrifugation. USA). FITC-conjugatedanti-CD30 antibodyandmouse IgG1 Cytoplasmic andnuclear extracts were preparedas described usedfor Flow Cytometric analysis were purchasedfrom previously (Yamaoka et al., 1998). DAKO (Glostrup, Denmark). FITC-conjugatedanti-mouse IgG for Flow Cytometric analysis was purchasedfrom American Qualex manufactures (CA, USA). Anti-IKK1 Western blot analysis monoclonal (556532) andanti-CD153 antibodieswere pur- Whole-cell extracts or cytoplasmic extracts were fractionated chasedfrom Becton Dickinson Pharmingen (CA, USA). on 8–12% SDS–polyacrylamide gels and transferred onto Puromycin andBlasticidinS were purchasedfrom Funakoshi Immobilon membranes (Millipore, MA, USA). Blots were (Tokyo, Japan) andZeocin were from Invitrogen (CA, USA). revealedwith an enhancedchemiluminescence detection CHX and paraformaldehyde (PFA) were purchased from system (ECL, Perkin Elmer, MA, USA). All the experiments Wako (Osaka, Japan). All the other reagents, including were repeatedat least twice. Densitometry was performed MG132, were purchasedfrom Sigma (MO, USA) unless using ATTO Densitograph (Bioscience & Biotechnology otherwise noted. Cells were treated with 20 mM of MG132 or ATTO, Tokyo, Japan). The relative bandintensity of p100 50 mg/ml of CHX, if necessary. or p52 was normalizedto that of actin.

Plasmids EMSA A ScaI/BglII DNA fragment, containing the cytomegalovirus Nuclear extracts (5 mg) were incubatedfor 30 min at room early enhancer-promoter, U3-deleated murine leukemia virus- temperature in binding buffer (10 mM HEPES (pH 7.8), long terminal repeat andpackaging signal, was excisedfrom 100 mM NaCl, 1 mM EDTA, 1 mM DTT, 2.5% glycerol, 1 mg pRxhCD25iN (Hanada et al., 1996) andsubclonedinto the of poly (d(I-C))) with 0.5 ng of 32P-labeled kB probe derived same enzymatic site of pMX-puro (Kawakami et al., 1997), from the H-2Kb promoter (Yamaoka et al., 1998) or 32P- generating pMRX-puro. pCn, pCn-100 (Yamaoka et al., labeledOct-1 probe (Mori et al., 2000). For supershift assays, 1996), pRK-5 vector, andpRK-NIK (Hsu et al., 1996) (kind nuclear extracts were incubatedwith speciﬁc antibodiesfor gifts of Dr D Goeddel, Amgen, CA, USA) were described 30 min on ice before incubation with the labeledprobe. previously. An MluI/NotI DNA fragment containing the entire Samples were run on a polyacrylamide gel containing 2.5% coding region of human CD30 was excised from pME-CD30 glycerol in 0.5 Â TBE. All the experiments were performedat (Horie et al., 2002) andsubclonedinto the BglII/NotI site of least twice. pMRX-puro, generating pMRX-CD30-puro. A SalI DNA fragment containing the Zeocin-resistance gene derived from pZeoSV (Invitrogen, CA, USA) andan internal ribosome Immunoprecipitation and kinase assay entry site (IRES) sequence was subclonedinto the SalI sites of Cytoplasmic extracts were subjectedto immunoprecipitation pMX-puro, generating pMX-IRES-Zeocin. An EcoRI frag- with anti-IKK1 antibody in TNT buffer (20 mM Tris–HCl (pH ment containing the coding sequence of IkBa with S32A S36A 7.5), 200 mM NaCl, 1% Triton X-100, 0.5 mM PMSF, 100 mM mutations was excisedfrom pRc-CMV-I kBa (Whiteside et al., Na3VO4,20mM b-glycerophosphate, 1 mg/ml leupeptin, and 1995) andsubclonedinto the EcoRI site of pMX-IRES- 1 mg/ml aprotinin). Immunoprecipitates were collectedon Zeocin, generating pMX-SR-IkBa-IRES-Zeocin. A cDNA of protein G-sepharose beads (Pierce, IL, USA), which were then human CD153 was ampliﬁedby PCR using cDNAs prepared washedthree times with TNT buffer andthree times with kinase from MT-2 cells (Sugamura et al., 1984) as a template. A reaction buffer (20 mM HEPES (pH 7.5), 10 mM MgCl2,50mM BamHI/NotI fragment containing the entire coding region of NaCl, 100 mM Na3VO4,20mM b-glycerophosphate, 2 mM DTT, human CD153 was subclonedbetween the BamHI and NotI and20 mM ATP). The immunoprecipitates were usedin in vitro sites of pMRX-puro, generating pMRX-CD153-puro. Full kinase assay andimmunoblotting for detection of immunopre- construction details are available upon request. cipitatedIKK1. Kinase reaction was performedfor 30 min at 301Cusing5mCi of [g-32P]ATP andglutathione- S-transferase Flow cytometric analysis (GST)-IkBa (amino acids 1–72) as substrate (Yamaoka et al., 1998). The reaction products were separated on 12% SDS– For detection of cell surface CD30, approximately 5 Â 105 cells polyacrylamide gels and revealed by autoradiography. were stainedwith FITC-conjugatedanti-CD30 antibodyor FITC-conjugatedmouse IgG1 at a concentration of 5 mg/ml Reporter gene assays for 30 min on ice. For detection of cell surface CD153, SV-T2 cells were stainedwith anti-CD153 monoclonal antibodyor IgkCona-luc (Munoz et al., 1994) was usedto determineNF- mouse IgG1 followedby a secondstaining with FITC- kB-dependent transcriptional activation in Rat-1 and 5R cells. conjugatedanti-mouse IgG. After staining, cells were washed, Cells were transfectedwith pME or pME-CD30 (Horie et al., resuspended, and subjected to FACScalibur analysis. The data 2004) along with IgkCona-luc andpEF1-LacZ (Yamaoka were analysedwith the CellQuest program. et al., 1998) by a calcium phosphate co-precipitation method. Assays for luciferase and b-galactosidase were performed 36 h after transfection in standard methods. Luciferase activity was Tumorigenicity assay normalizedon the basis of b-galactosidase activity. Each Anchorage-independent cell growth was examined essentially experiments was repeatedat least three times, andthe results as described previously (Yamaoka et al., 1996). are expressedas a mean with s.d.

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3985 Abbreviations NIK plasmids and Dr T Kitamura (University of Tokyo) for EBV, Epstein–Barr virus; EMSA, electrophoretic mobility shift the pMX vector andPLAT-E cells. We also thank the assay; HL, Hodgkin lymphoma; H-RS cells, Hodgkin/Reed– members of Department of Molecular Virology for helpful Sternberg cells; IkB, inhibitor of NF-kB; IKK, IkB kinase; NF- discussion. This work was supported by Grant-in-Aid kB, nuclear factor-kB; NEMO, NF-kB essential modulator; for Scientiﬁc Research on Priority Areas from the Ministry NIK, NF-kB-inducing kinase; TNF, tumor necrosis factor; of Education, Culture, Sports, Science and Technology to TNFR, TNF receptor; TRAF, TNF receptor-associatedfactor NY andSY, grant from the Ministry of Health, Labor and Welfare of Japan andthe Human Science Foundationto Acknowledgements NY, andGrant-in-Aidfor Scientiﬁc Research (C) from the We thank Drs N Rice andA Israe¨ l (Institut Pasteur de Paris) Ministry of Education, Culture, Sports, Science and Techno- for plasmids and antisera, Dr D Goeddel (Amgen, CA) for logy to SY.

References

Annunziata CM, Saﬁran YJ, Irving SG, KasidUN and Hironaka N, Mochida K, Mori N, Maeda M, Yamamoto N Cossman J. (2000). Blood, 96, 2841–2848. andYamaoka S. (2004). Neoplasia, 6, 266–278. Atkinson PGP, Coope HJ, Rowe M andLey SC. (2003). Horie R, Higashihara M andWatanabe T. (2003). Int. J. J. Biol. Chem., 278, 51134–51142. Hematol., 77, 37–47. Bargou RC, Emmerich F, Krappmann D, Bommert K, Horie R, Watanabe M, Ishida T, Koiwa T, Aizawa S, Itoh K, Mapara MY, ArnoldW, Royer HD, Grinstein E, Greiner Higashihara M, Kadin ME and Watanabe T. (2004). Cancer A, Scheidereit C and Do¨ rken B. (1997). J. Clin. Invest., 100, Cell, 5, 353–364. 2961–2969. Horie R, Watanabe T, Morishita Y, Ito K, Ishida T, Kanegae Barth TFE, Martin-Subero JI, Joos S, Menz CK, Hasel C, Y, Saito I, Higashihara M, Mori S, Kadin ME and Mechtersheimer G, Parwaresch RM, Lichter P, Siebert R Watanabe T. (2002). Oncogene, 21, 2493–2503. andMo¨ ller P. (2003). Blood, 101, 3681–3686. Hsu H, Shu HB, Pan MG and Goeddel DV. (1996). Cell, 84, Ben-Neriah Y andSchmitz ML. (2004). EMBO Rep, 5, 299–308. 668–673. Hu¨ binger G, Schneider C, Sto¨ hr D, Ruff H, Kirchner D, Cabannes E, Khan G, Aillet F, Jarrett RF andHay RT. Schwa¨ nen C, SchmidM, Bergmann L andMu ¨ ller E. (2004). (1999). Oncogene, 18, 3063–3070. Exp. Hematol., 32, 382–389. Carbone A, Gloghini A, Gattei V, Aldinucci D, Degan Joos S, Menz CK, Wrobel G, Siebert R, Gesk S, Ohl S, M, Paoli P, De Zagonel V andPinto A. (1995). Blood, 85, Mechtersheimer G, Tru¨ mper L, Mo¨ ller P, Lichter P and 780–789. Barth TFE. (2002). Blood, 99, 1381–1387. Chen LF andGreene WC. (2004). Nat. Rev. Mol. Cell. Biol., 5, Kawakami Y, Miura T, Bissonnette R, Hata D, Khan WN, 392–401. Kitamura T, Maeda-Yamamoto M, Hartman SE, Yao L, Chinanonwait N, Miura H, Yamamoto N andYamaoka S. Alt FW andKawakami T. (1997). Proc. Natl. Acad. Sci. (2002). FEBS Lett., 531 (3), 553–560. USA, 94, 3938–3942. Claudio E, Brown K, Park S, Wang H and Siebenlist U. Kayagaki N, Yan M, Seshasayee D, Wang H, Lee W, French (2002). Nat. Immunol., 3, 958–965. DM, Grewal IS, Cochran AG, Gordon NC, Yin JP, Cogswell PC, Guttridge DC, Funkhouser WK and Baldwin Starovasnik MA andDixit VM. (2002). Immunity, 17, AS. (2000). Oncogene, 19, 1123–1131. 515–524. Coope HJ, Atkinson PGP, Huhse B, Belich M, Janzen J, Krappmann D, Emmerich F, Kordes U, Scharschmidt E, Holman MJ, Klaus GGB, Johnston LH andLey SC. (2002). Do¨ rken B andScheidereitC. (1999). Oncogene, 18, 943–953. EMBO J., 21, 5375–5385. Ku¨ ppers R, Klein U, Schwering I, Distler V, Bra¨ ninger A, Dejardin E, Deregowski V, Chapelier M, Jacobs N, Gielen Cattoretti G, Tu Y, Stolovitzky GA, Califano A, Hansmann J, Merville MP andBours V. (1999). Oncogene, 18, M-L andDalla-Favera R. (2003). J. Clin. Invest., 111, 2567–2577. 529–537. Dejardin E, Droin NM, Delhase M, Haas E, Cao Y, Makris C, Liao G, Zhang M, Harhaj EW andSun SC. (2004). J. Biol. Li ZW, Karin M, Ware CF andGreen DR. (2002). Chem., 279, 26243–26250. Immunity, 17, 525–535. Li Q andVerma ML. (2002). Nat. Rev. Immunol., 2, 725–734. Derudder E, Dejardin E, Pritchard LL, Green DR, Ko¨ rner M Makris C, Roberts JL andKarin M. (2002). Mol. Cell. Biol., andBaudV. (2003). J. Biol. Chem., 278, 23278–23284. 22, 6573–6581. Eliopoulos AG, Caamano JH, Flavell J, Reynolds GM, Martı´ n-Subero JI, Gesk S, Harder L, Sonoki T, Tucker PW, Murray, PG, Poyet JL andYoung LS. (2003). Oncogene, Schlegelberger B, Grote W, Novo FJ, Calasanz MJ, 22, 7557–7569. Hansmann ML, Dyer MJS andSiebert R. (2002). Blood, Emmerich F, Theurich S, Hummel M, Haeffker A, Vry MS, 99, 1474–1477. Do¨ hner K, Bommert K, Stein H andDo ¨ rken B. (2003). Mathas S, Lietz A, Janz M, Hinz M, Jundt F, Scheidereit J. Pathol., 201, 413–420. C, Bommert K andDo ¨ rken B. (2003). Blood, 102, Fiumara P, Snell V, Li Y, Mukhopadhyay A, Younes M, 1028–1034. Gillenwater AM, Cabanillas F, Aggarwal BB andYounes A. Mori N, Fujii M, Iwai K, Ikeda S, Yamasaki Y, Hata T, (2001). Blood, 98, 2784–2790. Yamada Y, Tanaka Y, Tomonaga M and Yamamoto N. Hanada K, Tsunoda R and Hamada H. (1996). J. Leukoc. (2000). Blood, 95, 3915–3921. Biol., 60, 181–190. Morita S, Kojima T andKitamura T. (2000). Gene Ther., 7, Hinz M, Lemke P, Anagnostopoulos I, Hacker C, Krappmann 1063–1066. D, Mathas S, Do¨ rken B, Zenke M, Stein H andScheidereit Mu¨ ller JR andSiebenlist U. (2003). J. Biol. Chem., 278, C. (2002). J. Exp. Med., 196, 605–617. 12006–12012.

Oncogene CD30 triggers noncanonical NF-jB activation M Nonaka et al 3986 Munoz E, Courtois G, Veschambre P, Jalinot P andIsrae¨ lA. Sugamura K, Fujii M, Kannagi M, Sakitani M, Takeuchi M (1994). J. Virol., 68, 8035–8044. andHinuma Y. (1984). Int. J. Cancer., 34, 221–228. Novack DV, Yin L, Hagen-Stapleton A, Schreiber RD, Takahashi Y, Tanaka Y, Yamashita A, Koyanagi Y, Goeddel DV, Ross FP and Teitelbaum SL. (2003). J. Exp. Nakamura M andYamamoto N. (2001). J. Virol., 75, Med., 198, 771–781. 6748–6757. Opat S andGaston JS. (2000). Autoimmunity, 33, 45–60. Thomas RK, Re D, Wolf J andDiehl V. (2004). Lancet Oncol., Saito N, Courtois G, Chiba A, Yamamoto N, Nitta T, 5, 11–18. Hironaka N, Rowe M, Yamamoto N andYamaoka S. Whiteside ST, Ernst MK, LeBail O, Laurent-Winter C, Rice N (2003). J. Biol. Chem., 278, 46565–46575. andIsrae¨ l A. (1995). Mol. Cell. Biol., 15, 5339–5345. Saitoh T, Nakayama M, Nakano H, Yagita H, Yamamoto Xiao G, Cvijic ME, Fong A, Harhaj EW, Uhlik MT, N andYamaoka S. (2003). J. Biol. Chem., 278, 36005– WaterﬁeldM andSun SC. (2001). EMBO J., 20, 6805–6815. 36012. Yamaoka S, Courtois G, Bessia C, Whiteside ST, Weil R, Schneider C and Hu¨ binger G. (2002). Leuk. Lymphoma, 43, Agou F, Kirk HE, Kay RJ andIsrae ¨ l A. (1998). Cell, 93, 1355–1366. 1231–1240. Skinnider BF and Mak TW. (2002). Blood, 99, 4283–4297. Yamaoka S, Inoue H, Sakurai M, Sugiyama T, Hazama M, Solan NJ, Miyoshi H, Carmona EM, Bren GD andPaya CV. Yamada T and Hatanaka M. (1996). EMBO J., 15, 873–887. (2002). J. Biol. Chem., 277, 1405–1418. Zheng B, Fiumara P, Li YV, Georgakis G, Snell V, Younes M, Stein H, Foss HD, Du¨ rkop H, Maraﬁoti T, Delsol G, Pulford Vauthey JN, Carbone A andYounes A. (2003). Blood, 102, K, Pileri S andFalini B. (2000). Blood, 96, 3681–3695. 1019–1027.

Oncogene