NF-κB only partially mediates Epstein Barr virus latent membrane protein 1 activation of B cells

Transcription

1 Journal of General Virology (1998), 79, Printed in Great Britain NF-κB only partially mediates Epstein Barr virus latent membrane protein 1 activation of B cells Sara Liljeholm, 2 Kate Hughes, 1 Thomas Grundstro m 1 and Peter Brodin 2 Division of Tumour Biology 1, Department of Cell and Molecular Biology 2, Umea University, S Umea, Sweden The latent membrane protein 1 (LMP1) of Epstein Barr virus (EBV) is required for EBV-induced immortalization of human B cells and causes tumorigenic transformation of cell lines. LMP1 expression induces phenotypic changes resembling B cell activation, such as cell size increase and up-regulation of cell surface activation markers. LMP1 contains two domains that activate the transcription factor NF-κB, one through interactions with TRAF proteins and the other with the TRADD protein. The purpose of the present study was to investigate the importance of NF-κB induction in the up-regulation of the B cell activation markers ICAM-1 and CD71 by LMP1. This study shows that expression of LMP1 activates transcription from p50/p65- and c-rel-responsive promoters, and that this activity can be completely inhibited by expression of a dominant inhibitory IκB mutant. ICAM-1 and CD71 are nevertheless upregulated by LMP1 in primary B cells and cell lines expressing the dominant IκB. Furthermore, LMP1- induced cell size increase of primary B cells was unaffected by IκB expression. It was concluded that even when LMP1 is unable to activate NF-κB, it is still capable of inducing certain characteristics of activated B cells, strongly suggesting that LMP1 can also activate cells independently of NF-κB. Introduction Epstein Barr virus (EBV) is a herpesvirus carried by the majority of the human population. The virus immortalizes human primary B cells in vitro, and the resulting lymphoblastoid cell lines (LCLs) express a small subset of the EBV genome (reviewed in Sugden, 1989; Kieff & Liebowitz, 1990), including the latent membrane protein 1 (LMP1). Expression of LMP1 is crucial for immortalization to take place (Kaye et al., 1993). LMP1 can, by itself, transform cell lines (Wang et al., 1985; Baichwal & Sugden, 1988) and LMP1 is frequently found in EBV-associated malignancies such as Burkitt s lymphomas (Epstein & Achong, 1979), nasopharyngeal carcinomas (Fa hraeus et al., 1988), T cell leukaemias (Chen et al., 1993; Anagnostopoulos et al., 1992; Su et al., 1991) and in Reed Sternberg cells of Hodgkin s lymphomas (Herbst et al., 1991). LMP1 expression induces many phenotypic changes resembling EBV-immortalization of B lymphocytes in a variety of cell types. The expression of LMP1 in type I Burkitt s Author for correspondence: Peter Brodin. Present address: Department of Molecular Biology, Preclinical Research & Development, Astra Ha ssle AB, Tvisteva gen 48, S Umea, Sweden. Fax peter.brodin hassle.se.astra.com lymphoma cell lines, which have several characteristics of resting B cells (Rowe et al., 1987), leads to cell size increase, upregulation of cell surface activation markers such as ICAM-1, LFA-1, CD23, CD21 and CD71 (Wang et al., 1988), as well as expression of the bcl-2 oncogene (Henderson et al., 1991; Martin et al., 1993; Rowe et al., 1994). Importantly, transient expression of LMP1 in human primary B cells also leads to induction of activation markers, cell size increase and DNA synthesis (Peng & Lundgren, 1992). LMP1 is a multiple transmembrane-domain protein oriented in the plasma membrane in such a way that both the C and N termini are cytoplasmic (Liebowitz et al., 1986; Fennewald et al., 1984; Hennessy et al., 1984). LMP1 is distributed as discrete patches in the plasma membrane, and associates with the cytoskeleton in lymphocytes and transfected fibroblasts (Liebowitz et al., 1987; Martin & Sugden, 1991). The short N-terminal part is responsible for patching of the LMP1 molecules, but has not been shown to be important for the function of LMP1 in transformation assays (Izumi et al., 1994) or B cell activation (Peng-Pilon et al., 1995). The 200 amino acid cytoplasmic C-terminal domain appears to contain most of the elements necessary for LMP1 function, and deletion of this domain abolishes the ability of LMP1 to transform Rat-1 fibroblasts (Moorthy & Thorley-Lawson, SGM CBBH

2 S. Liljeholm and others 1993), immortalize B cells (Kaye et al., 1995) and induce upregulation of surface activation markers in B cells (Huen et al., 1995; Peng-Pilon et al., 1995). LMP1 triggers the activation of the transcription factor NFκB (Hammarskjo ld & Simurda, 1992; Laherty et al., 1992), and the NF-κB proteins p50, p65 and c-rel have been shown to mediate activation of the surface marker ICAM-1 (Jahnke & Johnson, 1994). The C-terminal domain of LMP1 contains two NF-κB-activating sub-domains, CTAR1 and CTAR2 (C-terminal activating regions 1 and 2). CTAR1 is membraneproximal, located between amino acids 194 and 232, and CTAR2 is at the C terminus between amino acids 351 and 386 (Huen et al., 1995; Mitchell & Sugden, 1995). CTAR1 interacts directly with TNF receptor-associated factor (TRAF) proteins of the TNF receptor signalling pathway and activates NF-κB to a low degree (Mosialos et al., 1995; Devergne et al., 1996). CTAR2, which does not interact directly with TRAFs, seems to be the major site for NF-κB activation (Huen et al., 1995; Mitchell & Sugden, 1995). This domain has recently been shown to activate NF-κB through binding of the TNF receptorassociated death domain (TRADD) protein (Izumi & Kieff, 1997). The present study was designed to investigate if, or to what degree, up-regulation of B cell activation markers upon transient expression of LMP1 is dependent on NF-κB activation. We show that expression of LMP1 activates NF-κB, and that expression of a dominant mutant of IκBα, an inhibitor of NF-κB, totally abolishes LMP1-induced NF-κB activity in DG75 Burkitt s lymphoma and K562 pre-erythromyelocytic cell lines. Flow cytometric analysis of B cell activation markers shows that the markers CD71 and ICAM-1 are to a large extent up-regulated by LMP1 even when the NF-κB activity is decreased below basal levels. Furthermore, the activation of primary B cells by LMP1 as measured by cell size increase could not be inhibited by expression of the dominant IκB. The results strongly suggest that B cell activation by LMP1 is dependent on an additional signal(s) separate from NF-κB. Methods Cells. DG75 is an EBV-negative Burkitt s lymphoma cell line (Rowe et al., 1986) and K562 is a pre-erythromyelocytic cell line (Gauwerky & Golde, 1980). CBM1 Ral-STO is an LCL (Ernberg et al., 1989). Human primary B cells were isolated from donated blood as previously described (Pilon et al., 1991). Yields were cells per 500 ml blood sample. At 18 h prior to transfection, the primary B cells were treated with 5 ng ml PDB (phorbol 12,13-dibutyrate, Sigma). All cells were maintained at 1 10 ml in a growth medium that consisted of RPMI 1640 supplemented with 5% FCS (10% for primary B cells) and antibiotics at 37 C in a humidified atmosphere containing 6% carbon dioxide. Plasmids. The LMP1 expression plasmid peflmp1 was obtained by ligating the LMP1 gene from pcmvlmp1 (Peng & Lundgren, 1992) as a blunted XbaI SmaI fragment into XbaI-digested and blunted pefc- X, carrying the strong polypeptide chain elongation factor 1α promoter. This plasmid is a derivative of pef-bos (Mizushima & Nagata, 1990), where a ClaI SalI BamHI XbaI polylinker is introduced between the two Xba1 sites. The p50 expression plasmid pcmv50 consists of the cdna encoding amino acids of p105 (Kieran et al., 1990) cloned into prc CMV (Invitrogen). The p65 expression plasmid pcmv65 has been described previously (Mercurio et al., 1992). The c-rel cdna was amplified by PCR from an LPS- and PDB-stimulated murine B cell library (a kind gift from P. Sideras, Division of Tumour Biology), with primers designed to add HindIII and NotI sites flanking the cdna. The PCR product was cloned between the HindIII and NotI sites of prc CMV. The isolated cdna was identical to the sequence described by Bull et al. (1990). The IκBα cdna was amplified by PCR from a B6 (EBV-immortalized human lymphoblastoid) cell line library (a kind gift from P. Sideras), with primers designed to add HindIII and NotI sites flanking the cdna. The PCR product was identical to the MAD-3 sequence described by Haskill et al. (1991), except for a C instead of a T at nucleotide position 400, which does not change the amino acid sequence. The PCR product was cloned between the HindIII and NotI sites of prc CMV, and it was moved as a HindIII (blunted) XbaI fragment to SalI (blunted) and XbaIdigested pefc-x to generate pefiκbα. pefiκbαdm was derived from pefiκbα by mutating serines 32 and 36 to alanines by PCR-directed mutagenesis. pgl2(igκb) carries a luciferase gene downstream of two κb elements derived from the immunoglobulin kappa light chain enhancer (Pierce et al., 1987). The BglII SpeI fragment carrying a hexameric µe5 µe2 site (Corneliussen et al., 1994) of a derivative of pgl2-basic was replaced by the sequence 5 GGGACTTTCCGAGAGGCTCGAGCCTCTC GGAAAGTCCCGCGG 3. The underlined sequences indicate the two IgκB sites arranged as in Pierce et al. (1987). This κb site preferentially binds p50 p65 heterodimers (Zabel et al., 1991). pgl2(κb1), a kind gift from A. Antonsson (Division of Tumour Biology), was constructed in an analogous way, replacing the BglII SpeI fragment in the µe5 µe2 reporter with the sequence 5 GGTTTACAGGGGACTTCCCTCGGGTTGC- GATCGGTTTACAGGGGACTTCCCTCGGGTTCCGCGG 3. The underlined sequences indicate the two κb1 sites, arranged in tandem as in Ohmori et al. (1994). The κb1 site has been shown to be c-rel-specific in a transient transfection assay (Ohmori et al., 1994). pefcd2 expresses a part of the rat CD2 (rcd2) and is used as a cotransfection marker (Peng & Lundgren, 1992). The plasmid was constructed by isolating a BamHI HindIII cdna fragment from percd2-2 (He et al., 1988) and ligating it to the BamHI and Xba1 sites in pefc-x (the HindIII and XbaI sites were blunted). The rcd2 cdna encodes only the transmembrane and extracellular domains, and is unable to transduce signals (He et al., 1988; Ohno et al., 1991). Transfection by electroporation. Electroporations were performed under optimized conditions using 10 cells in 500 µl culture medium in 1 ml cuvettes with a 0 4 cm space between the electrodes (Bio- Rad). After the addition of DNA, the samples were gently mixed and kept at room temperature for 5 min. Electroporation was performed at 950 µf using 300 V for DG75, 330 V for K562 and 340 V for primary B cells. After electroporation, the samples were kept at room temperature for 5 min before being diluted with 10 ml culture medium. The amount of DNA per transfection was always adjusted to µg using puc19 as a carrier when necessary. The number of viable cells was determined 24 h after electroporation using the trypan blue exclusion method and was 67% for DG75, 70% for K562 and 13% for human primary B cells. Viability did not differ significantly between transfections with the various constructs. Transfection efficiency was determined by flow cytometric analysis, where cells bearing the co-transfection marker CD2 CBBI

3 NF-κB only partially mediates LMP1 effects antigen were scored as positive, and was 48% for DG75, 58% for K562 and 3% for primary B cells. Transfection efficiency did not vary significantly between transfections with the various constructs. Western blotting. Proteins from total cell extracts of 2 10 transfected DG75 cells, prepared 24 h after transfection with the indicated plasmids, were separated by SDS PAGE according to the method of Laemmli (1970) and blotted onto nitrocellulose filters (Amersham), basically according to the method of Towbin et al. (1979). Excess proteinbinding sites were blocked by incubation of the filters for 1 h in blocking reagent (Boehringer Mannheim). The filters were incubated overnight with an appropriate dilution of the anti-lmp1 monoclonal antibody CS1 4 (Dako). The protein antibody complexes were reacted with horseradish peroxidase-conjugated anti-mouse secondary antibody and visualized using the chemiluminescence Western blotting kit (Boehringer Mannheim) according to the manufacturer s instructions. Assay for NF-κB activity. NF-κB activity was determined by quantifying luciferase expressed from the transfected NF-κB reporter plasmids. At 24 h post-transfection, cells were washed in PBS and lysed in 100 µl cell culture lysis buffer (Promega). Luciferase activity was measured from 10 µl cell lysate in 100 µl luciferase assay reagent (Promega) in a BioOrbit 1250 luminometer. Flow cytometric analyses. Live-cell fluorescence was carried out using saturating amounts of FITC-conjugated anti-icam-1, anti-cd71 and an isotype-matched control antibody (all from Immunotech) and with saturating amounts of biotinylated anti-rcd2 antibody followed by streptavidin phycoerythrin to specifically label transfected cells. Prior to analysis, propidium iodine was added to a final concentration of 10 mg ml. Binding was quantified using a flow cytometer (FACSCalibur, Becton Dickinson) gated on side and forward light scatter as well as for propidium iodine-emitted fluorescence to monitor only live cells, and gated on phycoerythrin-emitted fluorescence to monitor only rcd2- positive cells (i.e. successfully transfected cells). For each sample, at least 3000 live rcd2-positive cells were counted. Results LMP1 activates transcription from p50/p65- and c-rel-specific κb sites NF-κB belongs to a family of transcription factors with five known mammalian members, including the founding member c-rel and the classic NF-κB heterodimer of p50 and p65. In order to quantify the effects of LMP1 expression on NF-κB activity, we designed luciferase reporter plasmids with promoters preferred by certain NF-κB family members. The p50 p65 reporter pgl2(igκb) contains two copies of the κb site from the immunoglobulin kappa light chain enhancer (Pierce et al., 1987), and pgl2(κb1) contains two copies of a κb site that has been shown to be c-rel-specific in a transient transfection assay (Ohmori et al., 1994). The effect of LMP1 expression on activation of transcription from these reporters was investigated. DG75 cells were used because Burkitt s lymphoma type I cells share many characteristics with resting B cells (Rowe et al., 1987), and K562 was used since this cellline shows clear effects upon LMP1 expression (Peng & Lundgren, 1993), while having a lower basal NF-κB activity than B cells. Fig. 1. LMP1 activation of NF-κB in K562 and DG75 cells. NF-κB activity was measured using luciferase reporter plasmids (2 µg) with p50/p65- responsive IgκB or c-rel-responsive κb1 sites (see Methods) 24 h after co-transfection with plasmids expressing LMP1 (peflmp1, 0 25 µg) or the parental vector, pefc-x (VC, vector control, 0 25 µg). The results from the luciferase measurements are represented in luciferase units. The mean SD of three transfections is shown. Fig. 1 shows that transient expression of LMP1 from the expression vector peflmp1 activates transcription of the luciferase gene from both the p50 p65 and the c-rel reporter in K562 and DG75 cells. The induction was 6-fold and 11-fold from the IgκB and κb1 sites, respectively, in K562 and 2 5-fold and 6-fold from the IgκB and κb1 sites, respectively, in DG75. Titration showed that maximal NF-κB induction was achieved with between 0 1 and 0 5 µg peflmp1 in both cell lines (see below). Transfection of plasmids expressing the p50 p65 or c-rel proteins also increased transcription from their corresponding luciferase reporter (data not shown). It should be noted that the basal level (vector control) of NF-κB is substantially higher in cells of B cell origin, like DG75, than in K562 cells (Miyamoto et al., 1994; Liou et al., 1994), and that the IgκB site gives a higher basal luciferase gene transcription than the κb1 site in both cell types (Fig. 1). LMP1-induced NF-κB activity can be completely inhibited by overexpression of IκBα One objective of this study was to find a way to selectively inhibit NF-κB induction in LMP1-expressing cells. NF-κB forms a complex with an inhibitory κb protein, IκB, which inactively sequesters NF-κB in the cytoplasm. Activation of NF-κB requires the phosphorylation and subsequent degradation of IκB, releasing NF-κB to translocate to the nucleus and activate its target genes. We therefore co-transfected peflmp1 together with vectors expressing the NF-κB inhibitor IκBα.We used vectors expressing both wild-type IκBα (pefiκbα) and a dominant IκBα mutant, IκBαDM, with the inducible serine phosphorylation sites at positions 32 and 36 mutated (pefiκbαdm). This mutant cannot be phosphorylated and is consequently not proteolysed upon NF-κB stimulation (Brown et al., 1995). This mutant efficiently retains NF-κB in the cytoplasm, unable to bind to the κb sites in the nucleus and induce transcription. CBBJ

4 S. Liljeholm and others (a) (b) Fig. 2. Inhibition of LMP1-induced NF-κB activity from p50/p65-responsive IgκB sites (a) and c-rel-responsive κb1 sites (b) by overexpression of IκBα. K562 and DG75 cells were transfected with 0 25 µg peflmp1 together with the IgκB orκb1 luciferase reporter (2 µg) and, where indicated, the IκBα-expressing plasmids pefiκbα (3 µg) or sufficient amounts of pefiκbαdm to decrease the NF-κB activity below vector control levels (0 5 µg in K562 and 2 µg in DG75). The amounts of transfected promoter-carrying plasmids were equalized using pefc-x. Fold induction after 24 h is compared to the pefc-x vector control (VC) normalized to 1. The mean SD of three transfections is shown. Expression of IκBαDM inhibits the LMP1-induced NF-κB activity from IgκB sites (Fig. 2a) and κb1 sites (Fig. 2b). The dominant IκBαDM resulted in NF-κB activities below vector control levels when co-expressed with LMP1 in K562 and DG75 (Fig. 2). NF-κB activation by LMP1 was completely inhibited with 0 5 µg pefiκbαdm in K562, whereas 2 µg was needed for total inhibition in DG75. This difference probably reflects the higher basal level of NF-κB in DG75 (Fig. 1). Low transfection efficiency and viability (see Methods) render the corresponding analysis very difficult in primary B cells. However, co-transfections of 4 µg pefiκbαdm together with peflmp1 (0 25 µg) resulted in total inhibition of the LMP1- induced NF-κB activity (data not shown). These inhibitions in primary B cells resulted in NF-κB activity at or below vector control levels. Wild-type IκBα also inhibited the LMP1-induced NF-κB activity. In Fig. 2(a), 3 µg pefiκbα was used, resulting in partial inhibition. Addition of higher amounts of this plasmid also leads to total inhibition (data not shown). Transfection of wildtype or mutant IκBα also totally inhibited the NF-κB activity obtained by transfection of plasmids expressing p50 p65 or c-rel proteins (data not shown). Fig. 3. Western blot analysis of LMP1 expression in DG75 cells. peflmp1 (0 25 µg) was transfected either alone or together with 2 µg pefiκbαdm into DG75 cells. pefc-x was transfected as vector control (VC). The cells were harvested after 24 h, lysed and blotted as described in Methods. The LCL Ral-STO was used as a positive LMP1 control. IκB expression does not influence LMP1 protein levels A potential problem with overexpression of IκBα could be alterations in transcription from the expression vectors due to possible κb sites in promoters or enhancers. We have encountered such problems with the human cytomegalovirus (CMV) promoter (unpublished observations), but as seen in Fig. 3, LMP1 expression from the polypeptide chain elongation factor 1α promoter of peflmp1 was unaffected by the simultaneous IκBαDM expression from pefiκbαdm. The same was seen for the co-transfection marker rcd2, expressed from pefcd2 (Fig. 4). LMP1 up-regulates cell surface activation markers Expression of LMP1 has earlier been shown to lead to upregulation of cell surface activation markers in cell lines and human primary B cells (Wang et al., 1988; Peng & Lundgren, 1992, 1993). Here we have analysed the expression of ICAM- 1 and CD71 (transferrin receptor) in DG75, K562 and human primary B cells after transient expression of LMP1 from peflmp1. The rcd2 expression plasmid, pefcd2, was cotransfected to allow specific dual-colour flow cytometric analysis of successfully transfected rcd2-positive cells, as previously described (Pilon et al., 1991; Peng & Lundgren, 1992, 1993). Representative flow cytometric data are presented in Fig. 4 as dot-plot profiles with logarithmic scales, and expression of the cell surface markers in K562 and DG75 cells is summarized in Table 1. Table 2 shows the corresponding data for human primary B cells, and also the cell size as measured by FSC (light forward scattering count). Values are shown as the percentage of cells scoring positive above an arbitrarily set intensity threshold at which approximately 10% of the vector control CBCA

5 NF-κB only partially mediates LMP1 effects (a) (b) (c) Fig. 4. Cell surface activation marker expression. (a) K562 cells co-transfected with 5 µg pefcd2 plus 2 µg peflmp1 and/or 4 µg pefiκbαdm. (b) DG75 cells co-transfected with 5 µg pefcd2 plus 2 µg peflmp1 and/or 12 µg pefiκbαdm. Expression was analysed 48 h after electroporation of the plasmids indicated in column 1. Columns 2 and 3 show dual-colour flow cytometric data on viable rcd2-positive cells (i.e. successfully transfected cells) double-stained with anti-rcd2 (vertical axis), and anti-icam-1 antibody (column 2) or anti-cd71 (column 3) (horizontal axis). The scales are logarithmic, and the vertical lines indicate the threshold level above which lies approximately 10% of the cells that were transfected with the vector control pefc-x. (c) Cell surface activation marker expression and cell size increase of human primary B cells co-transfected with 20 µg pefcd2 plus 4 µg peflmp1 and/or 10 µg pefiκbαdm. Column 3 shows FSC (light forward scattering count) on the horizontal axis (linear scale) plotted against CD71 staining (logarithmic scale). The amounts of transfected promoter-carrying plasmids were equalized using pefc-x in all transfections. Table 1. Summary of flow cytometric data on the effects of LMP1 and IκBαDM expression constructs on cell surface marker expression in K562 and DG75 cells after transfection with the plasmids indicated in column 1 Each value is the average of three or more independent transfections performed as in Fig. 4, and indicates the percentage of the rcd2-positive cells that stained stronger than the threshold level described in Fig. 4. The SD values are given in parentheses. K562 DG75 Plasmids ICAM-1 CD71 ICAM-1 CD71 pefc-x 10 5 (0 2) 10 8 (0 4) 10 7 (0 2) 10 5 (0 2) peflmp (4 7) 8 0 (1 6) 54 4 (1 8) 32 0 (3 0) peflmp1 pefiκbαdm 38 7 (10 0) 8 5 (1 5) 26 1 (2 7) 13 1 (2 0) pefiκbαdm 6 9 (1 1) 11 0 (1 5) 4 5 (1 4) 7 6 (1 6) (pefc-x) transfectants scored positive. These analyses show that LMP1 induces up-regulation of ICAM-1 cell surface expression in K562 (Fig. 4a; Table 1), and that both CD71 and ICAM-1 are up-regulated in DG75 (Fig. 4b; Table 1) and primary B cells (Fig. 4c; Table 2) (compare peflmp1 in row 2 with pefc-x in row 1). Expression of LMP1 in primary B cells also resulted in cellular enlargement (Fig. 4c; Table 2). The surface markers are up-regulated in a dose-dependent manner, as judged from the correlation between vertical axis (rcd2) and horizontal axis intensity. It should be noted that the scale is logarithmic, so for example, the strongest ICAM-1-staining cells among the peflmp1 transfectants show almost 10-fold more staining than vector control transfectants. LMP1 expression did not affect the binding of a control antibody isotype matched with the antibodies toward the cell surface markers (data not shown). We found that 8-fold more peflmp1 (2 µg) was needed to obtain full cell surface marker induction compared to the CBCB

6 S. Liljeholm and others Table 2. Summary of flow cytometric data on the effects of LMP1 and IκBαDM expression constructs on cell surface marker expression and cell size in human primary B cells after transfection with the plasmids indicated in column 1 Each value is the average of three independent transfections performed as in Fig. 4, and indicates the percentage of the rcd2-positive cells that stained stronger than the threshold level described in Fig. 4. FSC (light forward scattering count) is a measure of cell size. The SD values are given in parentheses. Primary B cells Plasmids ICAM-1 CD71 FSC pefc-x 9 6 (0 3) 10 3 (0 3) 10 3 (0 5) peflmp (5 1) 32 4 (0 8) 28 2 (4 7) peflmp1 pefiκbαdm 38 8 (11 9) 25 6 (5 5) 23 8 (3 7) pefiκbαdm 4 8 (3 8) 6 9 (2 2) 7 9 (2 2) activation by co-expressing IκBαDM from pefiκbαdm as described above. As can be seen in row 4 of Fig. 4, expression of IκBαDM alone does not affect the expression of the cotransfection marker rcd2, but slightly decreases the expression levels of the analysed LMP1-inducible surface markers in K562, DG75 and primary B cells. This is probably due to the decrease in basal NF-κB activity seen in Fig. 2. Co-expression of LMP1 with IκBαDM leads to reduction of the LMP1-induced up-regulation of both the analysed cell surface markers (Fig. 4; Tables 1 and 2). The amounts of pefiκbαdm used (4 µg for K562, 12 µg for DG75 and 10 µg for primary B cells) were sufficient to decrease NF-κB activity below basal levels as measured by both IgκB and κb1 reporters (data not shown). Nevertheless, a substantial part of the LMP1- induced cell surface marker induction is retained in both the cell lines and in primary B cells. In DG75 cells, ICAM-1 is still upregulated to 40 70% compared to vector control levels, and not even extreme overexpression using up to 26 µg pefiκbαdm could abolish the LMP1-induced up-regulation of ICAM-1 (data not shown). Furthermore, the LMP1-induced cellular enlargement of primary B cells is not significantly affected by inhibition of NF-κB activation (Fig. 4c; Table 2). Fig. 5. Dose-response curves for NF-κB activation ( ) and ICAM-1 upregulation ( ) by peflmp1. K562 cells were transfected with various amounts of peflmp1, and NF-κB activation and ICAM-1 up-regulation were measured as described in Methods. amount needed for maximal NF-κB induction (0 25 µg) as measured using the luciferase reporter plasmids. Fig. 5 shows a comparison of the dose-response of NF-κB and ICAM-1 activation by peflmp1 in K562 cells. The optimal amounts of peflmp1 for these activations are different, with full ICAM-1 up-regulation occurring only at LMP1 levels where the NF-κB activity is already decreased to less than half its maximal level. Up-regulation of cell surface activation markers and cellular enlargement by LMP1 is only partially dependent on NF-κB activation In order to investigate if the LMP1-induced up-regulation of the cell surface markers described above is dependent on NF-κB activation, we inhibited the LMP1-induced NF-κB Discussion The EBV latent membrane protein LMP1 has profound influences on growth control and other phenotypes of human B cells and cell lines. Apart from acting as an oncogene in mouse fibroblasts (Wang et al., 1985; Baichwal & Sugden, 1988), it activates the expression of numerous cellular proteins, several of them markers found on activated B cells (Wang et al., 1988; Peng & Lundgren, 1992) or involved in protection against apoptosis (Henderson et al., 1991; Martin et al., 1993; Rowe et al., 1994). The present study was designed to investigate the importance of NF-κB induction in the upregulation of the B cell activation markers ICAM-1 and CD71 by LMP1. We found that LMP1 activates transcription under the control of κb-sites preferentially binding p50 p65 heterodimers as well as sites binding c-rel in both K562 and DG75 cells (Fig. 1). The p50 p65 preferred site gave the highest transcription in both cell types, but LMP1 gave the highest activation from the c-rel preferred site. This indicates that LMP1 activation relative to the basal level is higher for c-rel. NF-κB induction by LMP1 could be reduced by expressing the NF-κB inhibitor IκBα, and completely inhibited by a dominant mutant, IκBαDM, defective in inducible degradation (Fig. 2). Expression of IκBαDM in cells expressing LMP1 decreased the NF-κB activity to below vector control levels. The ability of this mutant to totally abolish LMP1-induced NFκB activation gave us a tool to investigate to what degree the LMP1-induced up-regulation of cell surface activation markers and cell size increase are dependent on NF-κB activation. LMP1-induced up-regulation of the surface markers CD71 and ICAM-1 were found to be reduced when IκBαDM was co- CBCC

7 NF-κB only partially mediates LMP1 effects Fig. 6. A schematic model for the up-regulation of cell surface markers by LMP1. expressed (Fig. 4; Tables 1 and 2). Nevertheless, the surface markers were still up-regulated in all cell types tested. In primary B cells, ICAM-1 was up-regulated to approximately 60% of the level obtained when LMP1 was expressed alone. The up-regulation of CD71 was substantially inhibited by IκBαDM in DG75 cells, but only to a low degree in primary B cells. We also found that IκBαDM was unable to significantly inhibit the LMP1-induced cellular enlargement in primary B cells (Table 2). Since we know from the luciferase reporter experiments that the dominant IκBα mutant inhibits detectable NF-κB induction upon LMP1 transfection, and also significantly reduces basal levels of NF-κB, the results strongly indicate that the induction of at least some of the cell surface activation markers by LMP1 is dependent on an additional signal(s) separate from NF-κB. The relatively strong IκBαDM inhibition of CD71 in DG75 cells may agree with a simple model where the LMP1-induced NF-κB activity directly influences the expression of the cell surface molecule at the transcriptional level or by controlling its transport to the cell surface. However, to explain the inability of dominant IκB to inhibit a substantial part of the LMP1-induced up-regulation of CD71 and cell size increase in primary B cells, and expression of ICAM-1 in K562 and DG75, we propose a scheme whereby LMP1 induces several signals, at least one of which is independent of NF-κB (Fig. 6). This interpretation is strengthened by the observation that several-fold more LMP1 expression plasmid is needed to obtain full up-regulation of the cell surface markers than is needed to fully activate NF-κB. As can be seen in Fig. 5, the dose-response curves for ICAM-1 upregulation and NF-κB activation by transfection of peflmp1 are strikingly different. NF-κB is activated to the highest levels with µg peflmp1, and higher amounts of plasmid result in substantially lower activity. In contrast, maximal cell surface expression of ICAM-1 is achieved with approximately 10-fold higher amounts of peflmp1, a level of LMP1 expression resulting in less than half maximal NF-κB activation. The failure of LMP1 to up-regulate CD71 in K562 cells indicates that the LMP1-induced up-regulation of CD71 in DG75 and primary B cells involves a signal transduction pathway(s) where a component is lacking or is not active in the non-b-cell line K562. Our results are in line with growing evidence that LMP1 signalling is rather complex, with two domains, CTAR1 and CTAR2, that signal through different members of the TNF receptor signalling pathway. CTAR1 interacts directly with TRAF proteins (Mosialos et al., 1995), while CTAR2 signals via interaction with the TRADD protein (Izumi & Kieff, 1997). Both domains mediate activation of NF-κB (Huen et al., 1995). It has been shown that mutants lacking CTAR1 or CTAR2 activate ICAM-1 to the same degree in B cell lines (approximately 50% of wild-type LMP1), whereas their ability to activate NF-κB differ and is cell-line-dependent (Huen et al., 1995). Furthermore, it has been demonstrated that LMP1 can induce the epidermal growth factor (EGF) receptor and the A20 zinc finger protein via two different pathways. EGF receptor up-regulation is TRAF-mediated but NF-κB-independent, whereas A20 induction seems to be mediated directly through NF-κB (Miller et al., 1996). From the results reported here we conclude that even when LMP1 is unable to activate NF-κB, due to an excess of a dominant IκBα mutant, it is still capable of up-regulating cell surface activation markers, strongly suggesting that LMP1 can activate cells independently of NF-κB (Fig. 6). The ICAM-1 and CD71 promoters contain many transcription factor binding sites, including both NF-κB and AP-1 sites (Ouyang et al., 1993; Jahnke & Johnson, 1994; Mun oz et al., 1996). In the case of ICAM-1, activation of the promoter by the antioxidant pyrrolidinedithiocarbamate has been shown to involve the AP- 1 site (Mun oz et al., 1996). This makes the recently described LMP1-mediated activation of AP-1 via the JNK kinase (Kieser et al., 1997; Kilger et al., 1998) a strong candidate for an additional signalling pathway involved in the up-regulation of the cell surface activation markers. However, the high number of transcription factor sites in addition to NF-κB and AP-1 in the ICAM-1 and CD71 promoter regions makes it possible that additional signal transduction pathways are involved. This work was funded by grants from the Swedish Cancer Society to P.B. and T.G., and by a grant from the Swedish Natural Science Research Council to T.G. References Anagnostopoulos, I., Hummel, M., Finn, T., Tiemann, M., Korbjuhn, P., Dimmler, C., Gatter, K., Dallenbach, F., Parwaresch, M. R. & Stein, H. (1992). Heterogeneous Epstein Barr virus infection patterns in peripheral T-cell lymphoma of angioimmunoblastic lymphadenopathy type. Blood 80, Baichwal, V. R. & Sugden, B. (1988). Transformation of Balb 3T3 cells by the BNLF-1 gene of Epstein Barr virus. Oncogene 2, Brown, K., Gerstberger, S., Carlson, L., Franzoso, G. & Siebenlist, U. (1995). Control of IκB-α proteolysis by site specific, signal induced phosphorylation. Science 267, CBCD

8 S. Liljeholm and others Bull, P., Morley, K. L., Hoekstra, M. F., Hunter, T. & Verma, I. M. (1990). The mouse c-rel protein has an N-terminal regulatory domain and a C-terminal transcriptional transactivation domain. Molecular and Cellular Biology 10, Chen, C. L., Sadler, R. H., Walling, D. M., Su, I. J., Hsieh, H. C. & Raab Traub, N. (1993). Epstein Barr virus (EBV) gene expression in EBVpositive peripheral T-cell lymphomas. Journal of Virology 67, Corneliussen, B., Holm, M., Waltersson, Y., Onions, J., Hallberg, B., Thornell, A. & Grundstro m, T. (1994). Calcium calmodulin inhibition of basic-helix-loop-helix transcription factor domains. Nature 368, Devergne, O., Hatzivassiliou, E., Izumi, K. M., Kaye, K. M., Kleijnen, M. F., Kieff, E. & Mosialos, G. (1996). Association of TRAF1, TRAF2, and TRAF3 with an Epstein Barr virus LMP1 domain important for B-lymphocyte transformation: role in NF-κB activation. Molecular and Cellular Biology 16, Epstein, M. A. & Achong, B. G. (editors) (1979). The Epstein Barr virus. Berlin: Springer Verlag. Ernberg, I., Falk, K., Minarovits, J., Busson, P., Tursz, T., Masucci, M. G. & Klein, G. (1989). The role of methylation in the phenotypedependent modulation of Epstein Barr nuclear antigen 2 and latent membrane protein genes in cells latently infected with Epstein Barr virus. Journal of General Virology 70, Fa hraeus, R., Fu, H. L., Ernberg, I., Finke, J., Rowe, M., Klein, G., Falk, K., Nilsson, E., Yadav, M., Busson, P., Tursz, T. & Kallin, B. (1988). Expression of Epstein Barr virus-encoded proteins in nasopharyngeal carcinoma. International Journal of Cancer 42, Fennewald, S., van Santen, V. & Kieff, E. (1984). Nucleotide sequence of an mrna transcribed in latent growth-transforming virus infection indicates that it may encode a membrane protein. Journal of Virology 51, Gauwerky, C. & Golde, D. W. (1980). Hormonal effects on cell proliferation in a human erythroleukemia cell line (K562). Blood 56, Hammarskjo ld, M. L. & Simurda, M. C. (1992). Epstein Barr virus latent membrane protein transactivates the human immunodeficiency virus 1 long terminal repeat through induction of NF-κB activity. Journal of Virology 66, Haskill, S., Beg, A. A., Tompkins, S. M., Morris, J. S., Yurochko, A. D., Sampson-Johannes, A., Mondal, K., Ralph, P. & Baldwin, A. S., Jr (1991). Characterization of an immediate-early gene induced in adherent monocytes that encodes I kappa B-like activity. Cell 65, He, Q., Beyers, A. D., Barclay, A. N. & Williams, A. F. (1988). A role in transmembrane signaling for the cytoplasmic domain of the CD2 T lymphocyte surface antigen. Cell 54, Henderson, S., Rowe, M., Gregory, C., Croom Carter, D., Wang, F., Longnecker, R., Kieff, E. & Rickinson, A. (1991). Induction of bcl-2 expression by Epstein Barr virus latent membrane protein 1 protects infected B cells from programmed cell death. Cell 65, Hennessy, K., Fennewald, S., Hummel, M., Cole, T. & Kieff, E. (1984). A membrane protein encoded by Epstein Barr virus in latent growthtransforming infection. Proceedings of the National Academy of Sciences, USA 81, Herbst, H., Dallenbach, F., Hummel, M., Niedobitek, G., Pileri, S., Muller-Lantsch, N. & Stein, H. (1991). Epstein Barr virus latent membrane protein expression in Hodgkin and Reed Sternberg cells. Proceedings of the National Academy of Sciences, USA 88, Huen, D. S., Henderson, S. A., Croom-Carter, D. & Rowe, M. (1995). The Epstein Barr virus latent membrane protein-1 (LMP1) mediates activation of NF-κB and cell surface phenotype via two effector regions in its carboxy terminal cytoplasmic domain. Oncogene 10, Izumi, K. M. & Kieff, E. D. (1997). The Epstein Barr virus oncogene product latent membrane protein 1 engages the tumor necrosis factor receptor-associated death domain protein to mediate B lymphocyte growth transformation and activate NF-κB. Proceedings of the National Academy of Sciences, USA 94, Izumi, K. M., Kaye, K. M. & Kieff, E. D. (1994). Epstein Barr virus recombinant molecular genetic analysis of the LMP1 amino-terminal cytoplasmic domain reveals a probable structural role, with no component essential for primary B-lymphocyte growth transformation. Journal of Virology 68, Jahnke, A. & Johnson, J. P. (1994). Synergistic activation of intercellular adhesion molecule 1 (ICAM-1) by TNF-α and IFN-γ is mediated by p65 p50 and p65 c-rel and interferon-responsive factor Stat1 α (p91) that can be activated by both IFN-γ and IFN-α. FEBS Letters 354, Kaye, K. M., Izumi, K. M. & Kieff, E. (1993). Epstein Barr virus latent membrane protein 1 is essential for B-lymphocyte growth transformation. Proceedings of the National Academy of Sciences, USA 90, Kaye, K. M., Izumi, K. M., Mosialos, G. & Kieff, E. (1995). The Epstein Barr virus LMP1 cytoplasmic carboxy terminus is essential for B- lymphocyte transformation. Fibroblast cocultivation complements a critical function within the terminal 155 residues. Journal of Virology 69, Kieff, E. & Liebowitz, D. (1990). Epstein Barr virus and its replication. In Fields Virology, 2nd edn, pp Edited by B. N. Fields & D. M. Knipe. New York: Raven Press. Kieran, M., Blank, V., Logeat, F., Vandekerckhove, J., Lottspeich, F., Le-Bail, O., Urban, M. B., Kourilsky, P., Baeuerle, P. A. & Israel, A. (1990). The DNA binding subunit of NF-κB is identical to factor KBF1 and homologous to the rel oncogene product. Cell 62, Kieser, A., Kilger, E., Gires, O., Ueffing, M., Kolch, W. & Hammerschmidt, W. (1997). Epstein Barr virus latent membrane protein-1 triggers AP-1 activity via the c-jun N-terminal kinase cascade. EMBO Journal 16, Kilger, E., Kieser, A., Baumann, M. & Hammerschmidt, W. (1998). Epstein Barr virus-mediated B-cell proliferation is dependent upon latent membrane protein 1, which simulates an activated CD40 receptor. EMBO Journal 17, Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, Laherty, C., Hu, H., Opipari, A., Wang, F. & Dixit, V. (1992). The Epstein Barr virus LMP1 gene product induces A20 zinc finger protein expression by activating nuclear factor, κb. Journal of Biological Chemistry 267, Liebowitz, D., Wang, D. & Kieff, E. (1986). Orientation and patching of the latent infection membrane protein encoded by Epstein Barr virus. Journal of Virology 58, Liebowitz, D., Kopan, R., Fuchs, E., Sample, J. & Kieff, E. (1987). An Epstein Barr virus transforming protein associates with vimentin in lymphocytes. Molecular and Cellular Biology 7, Liou, H.-C., Sha, W. C., Scott, M. L. & Baltimore, D. (1994). Sequential induction of NF-κB Rel family proteins during B-cell terminal differentiation. Molecular and Cellular Biology 14, Martin, J. & Sugden, B. (1991). Transformation by the oncogenic latent membrane protein correlates with its rapid turnover, membrane localization, and cytoskeletal association. Journal of Virology 65, Martin, J. M., Veis, D., Korsmeyer, J. & Sugden, B. (1993). Latent membrane protein of Epstein Barr virus induces cellular phenotype independently of expression of Bcl-2. Journal of Virology 67, CBCE

9 NF-κB only partially mediates LMP1 effects Mercurio, F., Didonato, J., Rosette, C. & Karin, M. (1992). Molecular cloning and characterization of a novel Rel NF-κB family member displaying structural and functional homology to NF-κB p50 p105. DNA and Cell Biology 11, Miller, W. E., Mosialis, G., Kieff, E. & Raab-Traub, N. (1996). Epstein Barr virus LMP1 induction of the epidermal growth factor receptor is mediated through a TRAF signaling pathway distinct from NF-kB activation. Journal of Virology 71, Mitchell, T. & Sugden, B. (1995). Stimulation of NF-κB-mediated transcription by mutant derivatives of the latent membrane protein of Epstein Barr virus. Journal of Virology 69, Miyamoto, S., Chiao, P. J. & Verma, I. M. (1994). Enhanced IκB-α degradation is responsible for constitutive NF-κ B activity in mature murine B-cell lines. Molecular and Cellular Biology 14, Mizushima, S. & Nagata, S. (1990). pef-bos, a powerful mammalian expression vector. Nucleic Acids Research 18, Moorthy, R. K. & Thorley-Lawson, D. A. (1993). All three domains of the Epstein Barr virus-encoded latent membrane protein LMP1 are required for transformation of Rat-1 fibroblasts. Journal of Virology 67, Mosialos, G., Birkenbach, M., Yalamanchili, R., VanArsdale, T., Ware, C. & Kieff, E. (1995). The Epstein Barr virus transforming protein LMP1 engages signaling proteins for the tumor necrosis factor receptor family. Cell 80, Mun oz, C., Castellanos, M. C., Ara ntzazu, A., Vara, A., Esteban, M. E., Redondo, J. M. & Landa zuri, M. O. (1996). Transcriptional upregulation of intracellular adhesion molecule-1 in human endothelial cells by the antioxidant pyrrolidine dithiocarbamate involves the activation of activating protein-1. Journal of Immunology 157, Ohmori, Y., Tebo, J., Nedospasov, S. & Hamilton, T. A. (1994). Kappa B binding activity in a murine macrophage-like cell line. Journal of Biological Chemistry 269, Ohno, H., Nakamura, T., Yagita, H., Okamura, K., Taniguchi, M. & Saito, T. (1991). Induction of negative signal through CD2 during antigen-specific T cell activation. Journal of Immunology 147, Ouyang, Q., Bommakanti, M. & Miskimins, W. K. (1993). A mitogenresponsive promoter region that is synergistically activated through multiple signalling pathways. Molecular and Cellular Biology 13, Peng, M. & Lundgren, E. (1992). Transient expression of the Epstein Barr virus LMP1 gene in human primary B cells induces cellular activation and DNA synthesis. Oncogene 7, Peng, M. & Lundgren, E. (1993). Transient expression of Epstein Barr virus LMP1 gene in B-cell chronic lymphocytic leukemia cells, T-cells, and hematopoetic cell lines: cell type independent-induction of CD23, CD21 and ICAM-1. Leukemia 7, Peng-Pilon, M., Ruuth, K., Lundgren, E. & Brodin, P. (1995). The cytoplasmic C-terminal domain but not the N-terminal domain of latent membrane protein 1 from Epstein Barr virus is essential for B cell activation. Journal of General Virology 76, Pierce, J. W., Leonardo, M. & Baltimore, D. (1987). Oligonucleotide that binds nuclear factor NF-κB acts as a lymphoid-specific and inducible enhancer element. Proceedings of the National Academy of Sciences, USA 85, Pilon, M., Gullberg, M. & Lundgren, E. (1991). Transient expression of the CD2 cell surface antigen as a sortable marker to monitor high frequency transfection of human primary B cells. Journal of Immunology 146, Rowe, M., Rooney, C. M., Edwards, C. F., Lenoir, G. M. & Rickinson, A. B. (1986). Epstein Barr virus status and tumor cell phenotype in sporadic Burkitt s lymphoma. International Journal of Cancer 37, Rowe, M., Rowe, D. T., Gregory, C. D., Young, L. S., Farrel, P. J., Rupani, H. & Rickinson, A. B. (1987). Differences in B cell growth phenotype reflects novel patterns of Epstein Barr virus latent gene expression in Burkitt s lymphoma cells. EMBO Journal 6, Rowe, M., Peng-Pilon, M., Huen, D. S., Hardy, R., Croom-Carter, D., Lundgren, E. & Rickinson, A. B. (1994). Upregulation of bcl-2 by the Epstein Barr virus latent membrane protein LMP1: a B-cell specific response that is delayed relative to NF-κB activation and to induction of cell surface markers. Journal of Virology 68, Su, I.-J., Hsieh, H.-C., Lin, K.-H., Uen, W.-C., Kao, C.-L., Chen, C.-J., Cheng, A.-L., Kadin, M. E. & Chen, J.-Y. (1991). Aggressive peripheral T-cell lymphomas containing Epstein Barr viral DNA: a clinicopathological and molecular analysis. Blood 77, Sugden, B. (1989). An intricate route to immortality. Cell 57, 5 7. Towbin, H., Steahelin, T. & Gordon, J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proceedings of the National Academy of Sciences, USA 76, Wang, D., Liebowitz, D. & Kieff, E. (1985). An EBV membrane protein expressed in immortalized lymphocytes transforms established rodent cells. Cell 43, Wang, D., Liebowitz, D., Wang, F., Gregory, C., Rickinson, A., Larson, R., Springer, T. & Kieff, E. (1988). Epstein Barr virus latent membrane protein alters the human B-lymphocyte phenotype: deletion of the amino terminus abolishes activity. Journal of Virology 62, Zabel, U., Schreck, R. & Baeuerle, P. A. (1991). DNA binding of purified transcription factor NF-κB. Affinity, specificity, Zn + dependence, and differential half-site recognition. Journal of Biological Chemistry 266, Received 12 February 1998; Accepted 19 May 1998 CBCF