Latent membrane protein 1 associated signaling pathways ... - Nature

Oncogene (1999) 18, 7161 ± 7167 ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $15.00 http://www.stockton-press.co.uk/onc

SHORT REPORT

Latent membrane protein 1 associated signaling pathways are important in tumor cells of Epstein-Barr virus negative Hodgkin's disease Hans Knecht*,1, Christoph Berger1, Cathy McQuain1, Sylvia Rothenberger2, Edith Bachmann2, Jennifer Martin3, Christoph Esslinger4, Hans G Drexler5, Yun C Cai1, Peter J Quesenberry1 and Bernhard F Odermatt6 1

LINK Laboratories at the Cancer Center, University of Massachusetts, Worcester, Massachusetts, USA; 2Institute of Pharmacology and Toxicology, University of Lausanne, Switzerland; 3MCD-Biology, University of Colorado, Boulder, Colorado, USA; 4LICR, Lausanne, Switzerland; 5DSMZ, Braunschweig, Germany; 6Laboratory of Immunohistochemistry, University Hospital, Zurich, Switzerland

The latent membrane protein 1 (LMP1) oncogene of Epstein-Barr virus (EBV) is selectively expressed in the Reed-Sternberg (RS) cells of EBV-associated Hodgkin's disease (HD). However, no dierences in clinical presentation and course are found between EBV positive and EBV negative forms of HD suggesting a common pathogenetic mechanism. We have studied the LMP1 associated signaling pathways and their dominant negative inhibition in the myelomonocytic HD-MyZ and the B-lymphoid L-428 HD cell lines. In both EBV negative cell lines expression of LMP1 is associated with the formation of multinuclear RS cells. Dominant negative inhibition of NF-kB mediated signaling at the step of IkB-a phosphorylation results in increased cell death with only a few typical RS cells resistant to overexpression of the dominant negative inhibitor IkB-aND54. However, dominant negative inhibition of NF-kB mediated signaling at the early step of TRAF2 interaction results in the formation of multinuclear cells in both cell lines and, in addition, in clusters of small mononuclear cells in the HD-MyZ cell line. In HD-MyZ cells overexpression of the powerful JBD-inhibitor of the JNK signal transduction pathway is restricted to small cells and never observed in RS cells. These small cells undergo apoptosis as shown by the TUNEL technique. Apoptosis of small cells is still observed after cotransfection of JBD and LMP1 but in addition a few apoptotic HD-MyZ cells with large fused nuclear masses are identi®ed suggesting that speci®c inhibition of JNK leads also to apoptosis of LMP1 induced RS cells. Thus, activation of the JNK signaling pathway is also important in the formation of Reed-Sternberg cells. Our ®ndings are consistent with a model where all three LMP1 associated functions, i.e. NF-kB mediated transcription, TRAF2 dependent signaling, and c-Jun activation act as a common pathogenetic denominator of both EBV negative and EBV positive HD. Keywords: LMP1; Reed-Sternberg cells; TRAF2; JNK pathway

*Correspondence: H Knecht, Division of Hematology, Viollier Institute, Spalenring 145, CH-4002 Basel, Switzerland Received 10 November 1998; revised 19 August 1999; accepted 23 August 1999

Latent membrane protein 1 (LMP1), encoded by the BNLF1 gene of Epstein-Barr virus was identi®ed as a viral oncogene because of its capacity to transform rodent ®broblasts in vitro and to render them tumorigenic in nude mice (Wang et al., 1985; Baichwal and Sugden, 1988; Moorthy and Thorley-Lawson, 1993). In particular, LMP1 is a powerful inducer of NF-kB mediated transcription (HammarskjoÈld and Simurada, 1992; Herrero et al., 1995; Huen et al., 1995; Mitchell and Sugden, 1995) and engages signaling proteins of the tumor necrosis factor receptor-associated factor (TRAF) family (Mosialos et al., 1995). TRAF proteins are a family of recently discovered signal transducers interacting with the cytoplasmic domains of the TNFR family (Rothe et al., 1994). So far, six members called TRAF1-6 have been described (Regnier et al., 1995; Baker and Reddy, 1996; Cao et al., 1996; Takeuchi et al., 1996). TRAF2 has been identi®ed as a common mediator of TNFR1, TNFR2, CD30 and CD40 induced NF-kB activation (Rothe et al., 1995a; Liu et al., 1996; Duckett et al., 1997). TRAF2 mediates about 25% of LMP1 induced NF-kB activation through direct interaction with amino acids 199 ± 231, designated as the c-terminal activation region 1 (CTAR 1) of LMP1 (Devergne et al., 1996; Kaye et al., 1996; Brodeur et al., 1997; Floettmann and Rowe, 1997). However, most of the LMP1 mediated NF-kB activation is dependent on the integrity of the outermost amino acids at positions 383 ± 386 of CTAR2 (Brodeur et al., 1997; Sandberg et al., 1997). This part of the LMP1 mediated NF-kB activation is thought to occur through direct binding of TRADD (tumor necrosis factor associated death domain protein) which then interacts with TRAF2, thus mimicking TNFR1 mediated NF-kB activation (Izumi and Kie, 1997). Moreover, mimicking of CD40 dependent signal transduction by LMP1 (Eliopoulos et al., 1996; Kilger et al., 1998) and CTAR2 dependent activation of the AP-1 family of transcription factors (Whitmarsh and Davis, 1996; Karin et al., 1997) through the c-Jun N-terminal kinase (JNK) pathway have recently been reported (Kieser et al., 1997). Interestingly, the JNK pathway is also activated upon TRAF2 overexpression (Reinhard et al., 1997). The ability of LMP1 to act as a ligand-independent, permanently activated receptor (Gires et al., 1997), its

Signal transduction in Reed-Sternberg cells H Knecht et al

7162

direct and indirect interactions with TRAF2 to bypass physiological regulatory mechanisms, together with its ability to inhibit apoptosis (Henderson et al., 1991; Martin et al., 1993) may therefore result in sustained proliferation or cell death depending on the activity of intracellular signal transducers. Hodgkin's disease (HD) is a malignant lymphoproliferative disorder ordinarily originating in lymph nodes and diagnosed by histological identi®cation of typical bi- or multinucleated giant cells, called ReedSternberg (RS) cells (Urba and Longo, 1992; Kadin, 1994). Immunophenotypic and molecular studies performed on tumor tissue, single RS cells or HD cell lines are consistent with the hypothesis that RS cells are derived from several hematopoietic lineages including the lymphoid, monocytoid, dendritic and granulocytic dierentiation pathway (reviewed by Haluska et al., 1994). Clonal rearrangements of TcRor JH genes have been identi®ed by the Southern technique in about 15% of 602 HD cases (Drexler, 1994). In RS cells engaged in the B cell dierentiation pathway, clonal populations may emerge by escaping antigen-driven apoptosis in the germinal center, as recently documented (Kanzler et al., 1996). However, irrespective of their genotype most primary and cultured RS cells express CD30 and CD40, both belonging to the TNFR family (Stein et al., 1985; DuÈrkop et al., 1992; Carbone et al., 1995). About 40% of all HIV negative HD cases are associated with EBV (reviewed by Glaser et al., 1997). Immunohistochemical and RT ± PCR studies reveal a high amount of the EBV encoded LMP1 in the RS cells from most of these patients (Herbst et al., 1991; Pallesen et al., 1991; Joske et al., 1992). Integrity of LMP1 dependent NF-kB mediated transcriptional activation appears to be biologically relevant in EBVassociated HD, because all naturally occurring LMP1 deletion variants (LMP1-del) isolated from such patients (Knecht et al., 1993) still fully stimulate NFkB mediated transcription (Rothenberger et al., 1997). Constitutive NF-kB activation is also observed in RS cells and HD cell lines and appears to be required for sustained survival of the tumor cells (Bargou et al., 1996, 1997). Moreover, expression of the multifunctional oncoprotein LMP1 in the EBV negative HD cell line L-428 is associated with an increased number of RS cells (Knecht et al., 1996) consistent with the hypothesis that identical signal transduction pathways are relevant in RS cells of EBV negative and EBV positive HD and that consequently the dominant negative inhibitors of LMP1 associated signaling pathways are also acting in LMP1 negative HD. In order to determine whether speci®c LMP1 mediated functions were associated with the formation of RS cells we have transfected the EBV negative HD cell lines HD-MyZ (Bargou et al., 1993) and L-428 (Schaadt et al., 1979) with LMP1 and the dominant negative inhibitors IkB-a-ND54, TRAF2-ND97 and JBD of the JNK interacting protein-1 (JIP-1). LMP1 promotes formation of multinuclear ReedSternberg cells in EBV negative HD cell lines with myelomonocytic and lymphoid features. Here we show the cytologic changes occurring in the adherently growing HD cell line HD-MyZ after Lipofectin based transfection of LMP1. About 20 ± 30% of the cells expressed the oncoprotein after 24 h. At timepoint 48 h

over 40% of the LMP1 positive cells showed bi- or multinuclearity or increased cellular volume (at least twice as large as the average mononuclear nonexpressing HD-MyZ cells (Figure 1a, Table 1). Clusters of LMP1 expressing RS cells were still identi®ed after 168 h (Figure 1b). However, at this timepoint the total number of LMP1 positive cells was less than 30% of the total number identi®ed 48 h post transfection. Decreasing numbers of LMP1 expressing cells were rather due to a loss of oncoprotein expression than to toxicity since no apoptotic cells were observed when analysed by the TUNEL technique. Analogous results were obtained after transfection of the naturally occurring LMP1 deletion variants LMP1-del 10 and LMP1-del 23, which fully maintain NF-kB mediated transcriptional activation (immunohistochemistry and Western blot data not shown). However, when transfecting the empty expression vector, or an EBNA1 or EBNA2 expressing plasmid, no changes in cell size were observed (Figure 1c,d). Electroporation based transfection of L428 cells yielded about 20% of LMP1 expressing cells with an increasing number of multinucleated forms up to 96 h (Table 1) and con®rmed our previous results obtained with the Lipofectin method, much less ecient (eciency 53%) in this cell line (Knecht et al., 1996). These ®ndings demonstrate that LMP1 behaves as an inducer of multinuclearity in EBV negative HD cell lines and suggests that LMP1 couples to signaling pathways already active in EBV negative HD cell lines. Dominant negative inhibition of NF-kB (p50/65) at the level of IkB-a phosphorylation is still consistent with RS cell survival. Considering the high constitutive p50/65 expression identi®ed in tumor derived RS cells and the EBV negative HD cell lines L-428 and HDMyZ (Bargou et al., 1996, 1997) we asked whether overexpression of the dominant negative inhibitor IkBa-ND54 abolished RS cell survival. Strong expresion of the dominant negative inhibitor protein was evident on Western blots and still observed in a few typical RS cells of both HD cell lines up to 72 h after transfection (data not shown). These ®ndings con®rm the xenotransplantation results of Bargou et al. (1997) for the HD-MyZ cell line (in these experiments HD-MyZ cells with permanent IkB-a-ND expression were still able to form slowly growing tumors) and demonstrate analogous signaling in the L-428 cell line. Dominant negative inhibition of TRAF2 mediated NF-kB activation at an early step (prior to IkB-a phosphorylation) favors survival of tumor cells with RS morphology. Expression of TRAF2 mRNA has recently been demonstrated in L-428 cells and RS cells originating from clinical tumor samples (Messineo et al, 1998). Considering the pivotal function of TRAF2 in TNFR related signaling leading to NF-kB activation (reviewed by Baeuerle, 1998) and its interaction with LMP1, we asked whether expression of a dominant negative inhibitor of TRAF2-mediated NF-kB activation at an early step induced morphologically evident changes in HD-MyZ and L-428 cells. Transfection of both cell lines with pcDNA3HA-TRAF2-ND97 resulted in strong protein expression in about 15% of the total number of cells after 24 h and remained essentially unchanged up to 72 h (Table 1, Figure 2a,b). Among the cells expressing the dominant negative inhibitor the percentage of large mononuclear


or multinuclear forms remained moderately increased (Figure 2c) over the whole period up to 168 h. However, at this timepoint strong mutant protein

expression was also identi®ed in several isolated clusters of small mononuclear HD-MyZ cells consistent with a growth advantage of these cells derived

A

C

B

D

Figure 1 Impact of LMP1 expression on morphology of HD-MyZ cells. Preferential association of pSV-LMP1 expression with RS cell morphology is evident. Immunostaining (Knecht et al., 1996) of slide chambers was performed with CS1-4 as primary antibody (a ± c). (a) Forty-eight hours post transfection over 40% of oncoprotein expressing cells are multinuclear or at least twice as large as the average non-expressing HD-MyZ cells (6200). (b) Clusters of LMP1 expressing outreÂ giant cells persist as late as 168 h post transfection (6200). (c) Transfection with the empty vector pKO-neo shows no induction of giant cells (6250). (d) Transfection of pSG5-EBNA1 results in nuclear expression of EBNA1 without changing the morphology of the cells. MoAb against EBNA1 was a kind gift from E Kremmer, Munich, Germany (6480) Table 1 LMP1 expression and TRAF2-ND97 expression promote multinuclearity in HD-MyZ and L-428 cells 24

48

Hours after transfection 72

96

transfection rate positive RS cells negative RS cells transfection rate positive RS cells negative RS cells

24+9 27+7 7+2 9+3 18+5 8+1

24+10 45+9 9+1 15+1 20+6 10+2

13+6 53+8 7+2 10+3 23+6 9+2

9* 61 8 6+2 29+8 8+4

6+1** 60+3 8+1 8+2** 28+2 11+2

transfection rate positive RS cells negative RS cells transfection rate positive RS cells negative RS cells

22+5 26+8 9+2 12+3 21+4 8+2

22+1 49+14 11+5 14+5 29+5 8+2

17+3 50+10 10+2 13+4 35+7 9+3

15+6 58+9 11+2 7+1 29+8 6+2

6+1 68+13 8+3 4+1 31+6 8+1

Cell line and plasmid

168

HD-Myz cell line LMP1 TRAF2-ND97

L-428 cell line LMP1 TRAF2-ND97

Counting of LMP1 and TRAF-ND97 expressing cells was performed on immunostained slide chanbers and cytocentrifuge preparations at a magni®cation of 2006. Fifteen microscopic ®elds were analysed at each time point in every single experiment. Eciency of transfection (transfection rate) was de®ned as % of positive cells (100 cells counted) at each timepoint in every single experiment. RS cells are de®ned as either bi/multinuclear or large mononuclear cells. Large mononuclear cells are polyploid with fused nuclear masses at least twice as large as the average cell. Transfection rates (%+s.d.) and mean percentages +s.d. of LMP1 or TRAF2-ND97 positive/negative RS cells are shown. Positive RS cells represent the percentage of plasmid expressing cells with RS cell morphology. Negative RS cells represent the percentage of RS cells without plasmid expression in the same experiments. Mean percentage+s.d. were calculated from 3 ± 5 independent transfection experiments. Percentages are based on at least 100 plasmid expressing and 100 negative cells at each timepoint in every single experiment. The percentage of RS cells in untransfered cell lines and the vector alone transfection experiments (pKO-neo and pcDNA3) were in the same range (4 ± 9%) as those of the negative RS cells shown in the Table. From one* respectively two** experiments

7163


7164

from one founder cell (Figure 2d). Since such clusters were undetectable up to 96 h, they apparently represent progeny of a single founder cell. Thus, dominant negative inhibition of the NF-kB activation cascade at an early step (prior to IkB-a phosphorylation) appears bene®cial for the survival of RS cells and selected small mononuclear cells which both already constitutively express high levels of NF-kB protein. This at a ®rst glance paradoxical eect may result from still preserved direct or indirect interactions of TRAF2 dominant negative inhibitor protein with other members of the TNFR related signaling complex (Rothe et al., 1997; van Antwerp et al., 1998; Izumi et al., 1999) and shift the balance towards survival. In this context it might be important that the dominant negative TRAF2ND97 protein expressed in the tumor cells maintains its ability to interact with TRADD and has no inhibitory eect on LMP1-mediated JNK activation (Kieser et al., 1999). It also maintains the ability to form heterodimers with cIAP through interaction with the conserved TRAF-N domain (Rothe et al., 1995b). Our ®ndings suggest that dominant negative inhibition

of TRAF2 associated signaling pathways prior to IkBa phosphorylation is bene®cial for RS survival in both cell lines and may confer a growth advantage for some tumor cells regardless of their morphological appearance. Transfection of L-428 cells with plasmid pcDNA3TRAF2, encoding the full length protein, showed signi®cant TRAF2 expression associated with an increase of positive large mononuclear or multinucleated RS cells after 48 h (transfection rate 9+2%; RS cells 46+9%; based on ®ve independent transfection experiments). At 168 h (transfection rate51%) only a very few TRAF2 overexpressing cells (mainly multinuclear) were identi®able. This decrease was due to loss of plasmid expression since apoptotic giant cells were not identi®able when analysed by the TUNEL technique. Thus, TRAF2 overexpression, probably through JNK and NF-kB activation (Reinhard et al., 1997) had the same eects as LMP1 on L-428 cells. In the HD-MyZ cell line, though successful pcDNA3-TRAF2 expression was con®rmed by Western blotting, only few cells showed

A

C B

D

Figure 2 Expression of the dominant negative inhibitor TRAF2-ND97. In this mutant the TRADD-interacting and oligomerization domains are still functional, allowing the interaction and oligomerization of TRADD-TRAF2-ND97-TRAF2. (a) Western blot (10% acrylamide, 26104 cell equivalents per lane) shows low endogenous TRAF2 levels but prominent expression of the dominant negative inhibitor protein in HD-MyZ and L-428 cells. Primary antibody sc-876 directed against the carboxy terminus was used. (b) Western blot (same transfection experiment as shown in (a) immunostained with primary antibody sc-805 against HA-tag con®rms prominent expression of the dominant negative inhibitor. (c) TRAF2-ND97 variant protein expression in mono- and multinuclear L428 cells 48 h post transfection. Primary antibody sc-805 directed against HA-tag was used (6200). (d) Homogeneous cluster of small HD-MyZ cells showing abundant expression of the variant protein 168 h after transfection. Primary antibody sc-805 against HA-tag was used (6400)


A

B

strong protein expression at cytological examination and among the positive cells normally sized elements were frequent. Expression of dominant negative inhibitors of the JNK signaling pathway is associated with small mononuclear HD-MyZ cells and apoptosis. A recently discovered eector function of LMP1 is the induction of the AP-1 family of transcription factors through activation of the JNK pathway resulting in cellular proliferation (Kieser et al., 1997). This eect is independent of NF-kB activation, mediated through the oncoprotein's CTAR2, and eciently blocked through dominant negative mutants of MAP kinase kinase 4 (designated as MKK4 or SEK1) (Kieser et al., 1997; Eliopulos et al., 1998). We have tested the eect of dominant negative expression of MKK4 as well as the expression of JBD (JNK binding domain) of JIP-1 (JNK interacting protein 1), a recently identi®ed speci®c cytoplasmic inhibitor of JNK (Dickens et al., 1997) in HD-MyZ and L-428 cells. Strong cytoplasmic expression of dominant negative MKK4 and JIP-1 was seen in HD-MyZ cells up to 72 h. All expressing cells were small and had often a spindle shaped morphology (Figure 3a,b). No positive multinuclear RS cells were identi®able. Though the total number of JIP-1 inhibitor expressing cells was low (4% at 24 h, 52% at 48 h) a speci®c signal was detected in Western blots (Figure 3c). When analysed with the TUNEL technique, HD-MyZ cells transfected with JBD showed apoptosis limited to small cells accounting for about 2% of the total number of cells. Interestingly, when co-transfected with JBD and LMP1 the number

D C

Figure 4 Schematic representation of LMP1 associated functions and their dominant negative inhibition in the EBV negative myelomonocytic HD-MyZ and B-lymphoid L-428 HD cell lines

Figure 3 Expression of JBD (residues 127 ± 281) of the JNK interacting protein 1 (JIP-1) in HD-MyZ cells. (a) Several small cells with spindle shaped form express the cytoplasmic inhibitor. Large multinuclear cells are consistently negative. Primary antibody sc-807 directed against FLAG epitope was used (6200). (b) Immunostaining is also seen in small cytoplasmic processes. Mitotic ®gures are all negative (6400). (c) Western blot (10% acrylamide, 26104 cell equivalents per lane) con®rms speci®c JBD expression in HD-MyZ cells 24 h post transfection

(lane 5). No JBD expression is found in L-428 cells. Strong JBD expression is observed in transfected COS cells used as a positive control (105 cell equivalents). FLAG-fusion protein sc-4111WB was used as internal control (lane 1). (d) An apoptotic HD-MyZ cell with large fused nuclear masses is seen 48 h post cotransfection with JBD and LMP1. Apoptotic cells were revealed by the in situ TUNEL method (Gavrieli et al., 1992) (6800). As a positive control for apoptosis human thymus and tonsils were used (not shown)

7165


7166

of apoptotic cells remained unchanged but beside the small mononuclear cells a few HD-MyZ cells with large apoptotic nuclear masses were identi®able (Figure 3d). No apoptotic cells were identi®ed in the untransfected HD-MyZ cells, LMP1 transfectants or control vector transfectants. These ®ndings suggest that in the HD-MyZ cell line the JIP-1 expression induces apoptosis by a so far not known mechanism even in LMP1 co-expressing cells on the way from mononuclear cells to multinuclear RS cells. Thus, the JNK pathway appears to be critically involved into the formation of RS cells. In summary our ®ndings demonstrate that LMP1 (TNFR) related signaling pathways are primordial in EBV negative HD cell lines (Figure 4). TRAF2 associated signaling appears as a major mediator of molecular interactions in RS cells by connecting the NF-kB and the JNK pathways. Since LMP1 is selectively expressed in EBV positive HD and since dominant negative inhibition of LMP1-mediated functions results in profound changes in EBV negative HD cell lines, it appears that both forms of Hodgkin's disease, viz. EBV negative and positive, are manifesta-

tions of the same pathogenetical principle, i.e. a particular form of a TNFR related signaling disorder. Acknowledgements The plasmids encoding full length human TRAF2, (pcDNA3-TRAF2), and an HA-tagged, human TRAF2 deletion mutant lacking amino acids 1 ± 97, (pcDNA3HATRAF2-ND97), which acts as a dominant negative inhibitor of TRAF2-mediated NF-kB activation, were a kind gift from David V Goeddel, Tularik Inc. (San Francisco, CA, USA). The plasmid encoding MKK4, (pcDNA3-FlagMKK4), the dominant negative MKK4 expresion vector pcDNA3-Flag-MKK4 (Ala), whose phosphorylation sites Ser257 and Thr261 have been replaced by Ala residues, and the plasmid pcDNA3-Flag-JBD, expressing the JNK binding domain (residues 127 ± 281) of the JNK interacting protein-1 (JIP-1), were a kind gift from Roger Davis (HHMI, Worcester, MA, USA). The authors thank Phyllis Spatrick and Barbara Eddy for vector ampli®cation and sequencing as well as Suzanne King for skilful secretarial assistance in assembling the manuscript. C Berger is a pediatric postdoctoral research fellow supported by the Swiss National Science Foundation. S Rothenberger is supported by the Schweizerische Krebsliga.

References Baeuerle PA. (1998). Curr. Biol., 8, R19 ± R22. Baichwal VR and Sugden B. (1998). Oncogene, 2, 461 ± 467. Baker SJ and Reddy EP. (1996). Oncogene, 12, 1 ± 9. Bargou RC, Mapara MY, Zugck C, Daniel PT, Pawlita M, Dohner H and Dorken B. (1993). J. Exp. Med., 177, 1257 ± 1268. Bargou RC, Leng C, Krappmann D, Emmerich F, Mapara MK, Bommert K, Royer HD, Scheidereit C and Dorken B. (1996). Blood, 87, 4340 ± 4347. Bargou RC, Emmerich F, Krappmann D, Bommert K, Mapara MY, Arnold W, Royer HD, Grinstein E, Greiner A, Scheidereit C and Dorken B. (1997). J. Clin. Invest., 100, 2961 ± 2969. Brodeur SR, Cheng G, Baltimore D and Thorley-Lawson DA. (1997). J. Biol. Chem., 272, 19777 ± 19784. Cao Z, Xiong J, Takeuchi M, Kurama T and Goeddel DV. (1996). Nature, 383, 443 ± 446. Carbone A, Gloghini A, Gattei V, Aldinucci D, Degan M, De Paoli P, Zagonel V and Pinto A. (1995). Blood, 85, 780 ± 789. Devergne O, Hatzivassiliou E, Izumi KM, Kaye KM, Kleijnen MF, Kie E and Mosialos G. (1996). Mol. Cell Biol., 16, 7098 ± 7108. Dickens M, Rogers JS, Cavanagh J, Raitano A, Xia Z, Halpern JR, Greenberg ME, Sawyers CL and Davis RJ. (1997). Science, 277, 693 ± 696. Drexler HG. (1994). Leuk. Lymph., 3, 201 ± 225. Duckett CS, Gedrich RW, Gil®llan MC and Thompson CB. (1997). Mol. Cell Biol., 17, 1535 ± 1542. DuÈrkop H, Latza U, Hummel M, Eitelbach M, Seed B and Stein H. (1992). Cell, 68, 421 ± 427. Eliopoulas AG, Dawson CW, Mosialos G, Floettmann JE, Rowe M, Armitage RJ, Dawson J, Zapata JM, Kerr DJ, Wakelam MJO, Reed JC, Kie E and Young LS. (1996). Oncogene, 13, 2243 ± 2254. Eliopoulos AG and Young LS. (1998). Oncogene, 16, 1731 ± 1742. Floettmann JE and Rowe M. (1997). Oncogene, 15, 1851 ± 1858. Gavrieli Y, Sherman Y and Ben Sasson SA. (1992). J. Cell Biol., 119, 493 ± 501.

Gires O, Zimber-Strobl U, Gonnella R, Ueng M, Marschall G, Zeidler R, Pich D and Hammerschmidt W. (1997). EMBO J., 16, 6131 ± 6140. Glaser SL, Lin RJ, Stewart SL, Ambinder RF, Jarrett RF, Brousset P, Pallesen G, Gulley ML, Khan G, O'Grady J, Hummel K, Preciado MV, Knecht H, Chan JKC and Claviez A. (1997). Int. J. Cancer, 70, 375 ± 382. Haluska FG, Brufsky AM and Canellos GP. (1994). Blood, 84, 1005 ± 1019. HammarskjoÈld ML and Simurada MC. (1992). J. Virol., 66, 6496 ± 6501. Henderson S, Rowe M, Gregory C, Croom-Carter D, Wang F, Longnecker R, Kie E and Rickinson A. (1991). Cell, 65, 1107 ± 1115. Herbst H, Dallenbach F, Hummel M, Nidobitek G, Pileri S, Muller-Lantzsch N and Stein H. (1991). Proc. Natl. Acad. Sci. USA, 88, 4766 ± 4770. Herrero JA, Mathew P and Paya CV. (1995). J. Virol., 69, 2168 ± 2174. Huen DS, Henderson SA, Croom-Carter D and Rowe M. (1995). Oncogene, 10, 549 ± 560. Izumi KM and Kie ED. (1997). Proc. Natl. Acad. Sci. USA, 94, 12592 ± 12597. Izumi KM, McFarland EC, Ting AT, Riley EA, Seed B and Kie ED. (1999). Mol. Cell Biol., 19, 5759 ± 5767. Joske DJL, Emery-Goodman A, Odermatt BF, Bachmann E, Bachmann F and Knecht H. (1992). Blood, 80, 2610 ± 2613. Kadin ME. (1994). Curr. Opin. Oncol., 6, 456 ± 463. Kanzler H, Kuppers R, Hansmann ML and Rajewsky K. (1996). J. Exp. Med., 184, 1495 ± 1505. Karin M, Liu ZG and Zandi E. (1997). Curr. Opin. Cell Biol., 9, 240 ± 246. Kaye KM, Devergne O, Harada JN, Izumi KM, Yalamanchili R, Kie E and Mosialos G. (1996). Proc. Natl. Acad. Sci. USA, 93, 11085 ± 11090. Kieser A, Kilger E, Gires O, Ueng M, Kolch W and Hammerschmidt W. (1997). EMBO J., 16, 6478 ± 6485. Kieser A, Kaiser C and Hammerschmidt W. (1999). EMBO J., 18, 2511 ± 2521.


Kilger E, Kieser A, Baumann M and Hammerschmidt W. (1998). EMBO J., 17, 1700 ± 1709. Knecht H, Bachmann E, Brousset P, Sandvej K, Nadal D, Bachmann F, Odermatt FB, Delsol G and Pallesen G. (1993). Blood, 82, 2937 ± 2942. Knecht H, McQuain C, Martin J, Rothenberger S, Drexler HG, Berger C, Bachmann E, Kittler ELW, Odermatt BF and Quesenberry PJ. (1996). Oncogene, 13, 947 ± 953. Liu ZG, Hsu H, Goeddel DV and Karin M. (1996). Cell, 87, 565 ± 576. Martin JM, Veis D, Korsmeyer SJ and Sugden B. (1993). J. Virol., 67, 5269 ± 5278. Messineo C, Hunter Jamerson M, Hunter E, Braziel R, Bagg A, Irving SG and Cossman J. (1998). Blood, 91, 2443 ± 2451. Mitchell T and Sugden B. (1995). J. Virol., 69, 2968 ± 2976. Moorthy RK and Thorley-Lawson DA. (1993). J. Virol., 67, 1638 ± 1646. Mosialos G, Birkenbach M, Yalamanchili R, VanArsdale T, Ware C and Kie E. (1995). Cell, 80, 389 ± 399. Pallesen G, Hamilton-Dutoit SJ, Rowe M and Young LS. (1991). Lancet, i, 320 ± 322. Regnier CH, Tomasetto C, Moog-Lutz C, Chenard MP, Wendling C, Basset P and Rio MC. (1995). J. Biol. Chem., 270, 25715 ± 25721. Reinhard C, Shamoon B, Shyamala V and Williams LT. (1997). EMBO J., 16, 1080 ± 1092. Rothe M, Wong SC, Henzel WJ and Goeddel DV. (1994). Cell, 78, 681 ± 692.

Rothe M, Sarma V, Dixit M and Goeddel DV. (1995a). Science, 269, 1424 ± 1429. Rothe M, Pan MG, Henzel WJ, Ayres TM and Goeddel DV. (1995b). Cell, 83, 1243 ± 1252. Rothe M, Xiong J, Shu HB, Williamson K, Goddard A and Goeddel DV. (1997). Proc. Natl. Acad. Sci. USA, 93, 8241 ± 8246. Rothenberger S, Bachmann E, Berger C, McQuain C, Odermatt BF and Knecht H. (1997). Oncogene, 14, 2123 ± 2126. Sandberg M, Hammerschmidt W and Sugden B. (1997). J. Virol., 71, 4649 ± 4656. Schaadt M, Fonatsch C, Kirchner H and Diehl V. (1979). Blut, 38, 185 ± 190. Stein H, Mason DY, Gerdes J, O'Connor N, Wainscoat J, Pallesen G, Gatter K, Falini B, Delsol G, Lemke H, Schwarting R and Lennert K. (1985). Blood, 66, 848 ± 858. Takeuchi M, Rothe M and Goeddel DV. (1996). J. Biol. Chem., 271, 19935 ± 19942. Urba WJ and Longo DL. (1992). N. Eng. J. Med., 326, 678 ± 687. Van Antwerp DJ, Martin SJ, Verma IM and Green DR. (1998). Trends Cell Biol., 8, 107 ± 111. Wang D, Liebowitz D and Kie E. (1985). Cell, 43, 831 ± 840. Whitmarsh AJ and Davis RJ. (1996). J. Mol. Med., 74, 589 ± 607.

7167