Raji and Daudi are EBV-positive BL cell lines. In contrast, BJAB and Ramos are EBV-negative BL cell lines. B95/Ramos is Ramos infected with the B95-8 strain of EBV. L428, KM-H2, HDLM-2 and L540 are HL cell lines. These cell lines were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium supplemented with 10 or 20% fetal bovine serum (FBS) and antibiotics. Human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% FBS and antibiotics.

Total RNA was extracted from various cell cultures by TRIzol (Invitrogen Life Technologies, Carlsbad, CA, USA) according to the protocol provided by the manufacturer. The first-strand cDNA was synthesized from 1 μg cellular RNA using a PrimeScript RT-PCR kit (Takara Bio Inc., Otsu, Japan) with random primers. The sequences of the primers used are summarized in Table I.

Cells (293T) were transfected by the calcium phosphate DNA coprecipitation method. The expression plasmids pSG5-LMP-1, pSG5-LMP-1Δ187-351, pSG5-LMP-1Δ349 and pSG5-LMP-1Δ194–386 were previously described (24,25). For CD30 expression, the plasmids wild-type human CD30 (pME-hCD30) and its mutant [pCR-hCD30(Δ95)] were used (12). The wild-type and various mutants of IκB-ζ, and pcDNA3-RelA were described previously (26,27). The dominant-negative mutants of IκBα, IκBβ, IκB kinase (IKK) α, IKKβ, IKKγ and NF-κB-inducing kinase (NIK) have been previously described (28–31). The plasmid for truncated TRAF2 protein with retention of only the TRAF domain, ΔTRAF2, has been described previously (32). The human IκB-ζ promoter-luciferase gene constructs have already been described (23,33). The single and combined internal deletion mutants of NF-κB sites were constructed by deletion of the NF-κB sites of the plasmid pGL3-hIκB-ζ(−853) (23). A reporter plasmid, expressing luciferase through a minimal promoter linked to five copies of the typical NF-κB responsive element from the IL-2 receptor α chain (IL-2Rα) gene (κB-LUC), was used to measure the NF-κB transcription competence (34). Two copies of the IL-8 activator protein-1 (AP-1) binding site were inserted upstream of the IL-8 enhancer-less core promoter linked to luciferase gene (AP-1-LUC) (35). Plasmids containing the IL-8 promoter (−133 to +44 bp) and the IL-6 promoter (−225 to +14 bp) linked to luciferase expression vectors were constructed from luciferase expression vectors (35,36). Bcl-3 luciferase reporter construct was described previously (37). In all cases, phRL-TK was cotransfected to correct for transfection efficiency. After 24 h, luciferase assays were conducted using the dual luciferase reporter system (Promega Corp., Madison, WI, USA), in which the relative luciferase activity was calculated by normalizing transfection relative to the Renilla luciferase activities. Data were expressed as mean ± SD of three experiments.

Nuclear proteins were extracted and transcription factors bound to specific DNA sequences were examined by EMSA, as previously described (38). The top strand sequences of the oligonucleotide probes or competitors were as follows: for the NF-κB element (κB1) of the IκB-ζ gene, 5′-GATCCGACGGGAATGTCCGGGACT-3′; for the mutated κB1 sequence, 5′-GATCCGACGtGtATGaCCGGG ACT-3′; for the NF-κB element (κB2) of the IκB-ζ gene, 5′-GATCGGTCTGGGAATTTCCAGTG-3′; for the mutated κB2 sequence, 5′-GATCGGTCTGtGtATaaCCAGTG; for the NF-κB element of the IL-2Rα gene, 5′-GATCCGGCAGGGG AATCTCCCTCTC-3′; and for the AP-1 element of the IL-8 gene, 5′-GATCGTGATGACTCAGGTT-3′. The above underlined sequences are the NF-κB and AP-1 binding sites, respectively. The sites of mutation are indicated in lowercase letters. In competition experiments, the nuclear extract was pre-incubated with 100-fold excess of unlabeled oligonucleotides for 15 min. To identify NF-κB proteins in the DNA-protein complex shown by EMSA, we used antibodies specific for various NF-κB family proteins, including p50, RelA, c-Rel, p52 and RelB (Santa Cruz Biotechnology Inc., Santa Cruz, CA, USA). These antibodies were incubated with the nuclear extracts for 45 min at room temperature before incubation with radiolabeled probes.

Lymph node biopsy samples were obtained from patients with BL and HL. IκB-ζ immunohistochemistry was performed using an anti-IκB-ζ antibody (Cell Signaling Technology, Inc., Beverly, MA, USA) after pretreatment of deparafinized tissue sections with ready-to-use proteinase K (Dako, Carpinteria, CA, USA). The sections were counterstained with methyl green, hydrated in ethanol, cleaned in xylene and mounted. Informed consent was obtained from all tissue donors.

To investigate the role of IκB-ζ in the pathogenesis of BL and HL, we assessed IκB-ζ mRNA expression levels in established BL and HL cell lines using reverse-transcription polymerase chain reaction (RT-PCR). We found that IκB-ζ mRNA expression was limited to EBV-infected BL cell lines but not in uninfected cells (Raji, Daudi and B95–8/Ramos) (Fig. 1A). On the other hand, all HL cell lines showed IκB-ζ mRNA levels (Fig. 1B). All EBV-infected BL cell lines and HL cell lines constitutively expressed LMP-1 and CD30, respectively (Fig. 1A and B). Immunohistochemical staining of BL cells, and Hodgkin and Reed-Sternberg cells in lymph nodes showed abundant IκB-ζ protein in the nuclei of these cells (Fig. 1C).

To investigate the induction of IκB-ζ in BL and HL, we performed transient expression assays using mammalian expression vectors for LMP-1 and CD30 in 293T cells. After the transfection, RNA was extracted from the cells and the IκB-ζ mRNA levels were analyzed by RT-PCR. IκB-ζ mRNA induction was observed at 48 and 24 h after LMP-1 and CD30 transfection, respectively. In contrast, IκB-ζ mRNA was hardly detected in cells transfected with empty vectors (pSG5 and pME18S) (Fig. 2A and B).

To determine whether LMP-1 and CD30 regulate IκB-ζ promoter activity, transient expression assays were performed using the reporter plasmids, pGL3-hIκB-ζ(-11k) and pGL3-hIκB-ζ(−853), and expression vectors for LMP-1 and CD30. LMP-1 and CD30 transactivated the −11 kb and −853 IκB-ζ promoter fragments, but this effect was lost in pGL3-basic. LMP-1 and CD30 induced relative levels of IκB-ζ promoter-directed luciferase expression in a dose-dependent manner, suggesting that LMP-1 and CD30 functionally activate minimum IκB-ζ promoter between −853 and −17 bp (Fig. 2C and D).

To map the regions in the LMP-1 and CD30 proteins that mediate activation of IκB-ζ promoter, LMP-1 and CD30 mutants were expressed and their effect on IκB-ζ promoter activity was investigated. The LMP-1 mutants used included LMP-1Δ187–351 (which contains only CTAR-2 in the carboxyl-terminus), LMP-1Δ349 (which lacks CTAR-2) and LMP-1Δ194–386 (in which the entire carboxyl-terminal cytoplasmic region is deleted) (Fig. 3A). In cells that expressed CTARs-free LMP-1Δ194–386, IκB-ζ promoter activity was not increased. In contrast, activation of pGL3-hIκB-ζ( −11k) and pGL3-hIκB-ζ( −853) was observed by both CTAR-1-free LMP-1Δ187–351 and CTAR-2-free LMP-1Δ349, although to a lesser extent than by wild-type LMP-1 (Fig. 3B, lower panel). These results suggest that LMP-1 activates IκB-ζ expression via the cooperative activity of CTAR-1 and CTAR-2 signaling motifs. As measured in an NF-κB-dependent luciferase reporter gene assay, LMP-1 increased NF-κB activation via CTAR-1 and CTAR-2 (Fig. 3B, upper panel).

The carboxyl-terminal region of CD30 is also essential for signal transduction (Fig. 4A) (12). Next, we investigated whether the carboxyl-terminal region of CD30 plays a role in the induction of IκB-ζ. As shown in Fig. 4B, lower panel, IκB-ζ-driven reporter gene activity was not increased by CD30Δ95, which lacks the carboxyl-terminal region of CD30. Notably, the effect of the structural context of the carboxyl-terminal region of LMP-1 and CD30 on IκB-ζ promoter activation correlated with NF-κB activation as reporter analyses by mutant constructs (Figs. 3B and 4B, upper panels). These results suggest that LMP-1 and CD30 activate IκB-ζ promoter through the NF-κB signaling pathway.

LMP-1 and CD30 are constitutively aggregated pseudo-TNF receptors that activate NF-κB through their carboxyl-terminal cytoplasmic domains associated with TRAF2 (7–12). Aggregated TRAF2 activates NIK and its downstream target, the IKK complex, which is composed of two catalytic subunits, IKKα and IKKβ, and a regulatory subunit, IKKγ (39,40). The IKK complex phosphorylates the inhibitory IκB proteins, which are bound to NF-κB in the cytosol. Their phosphorylation is followed by their degradation, dissociation of NF-κB from the inhibitors, and NF-κB translocation into the nucleus (41). In order to determine the role of TRAF2/NIK/IKK/IκB proteins in mediating IκB-ζ activation induced by LMP-1 and CD30, 293T cells were cotransfected with LMP-1 or CD30 expression plasmid and plasmids expressing dominant-negative forms of TRAF2, NIK, IKKα, IKKβ, IKKγ, IκBα or IκBβ. All dominant-negative mutants reduced IκB-ζ promoter activation by LMP-1 and CD30 (Fig. 5A). These data indicate that the activation of NF-κB through TRAF2/NIK/IKK plays a role in the activation of IκB-ζ promoter by LMP-1 and CD30.

The human IκB-ζ promoter contains two NF-κB motifs (κB1 and κB2) (Fig. 5B, top panel) (33). To determine the involvement of κB1 and κB2 in the induction of IκB-ζ gene expression by LMP-1 and CD30, we investigated the activity of IκB-ζ promoter with deletions in κB1 and κB2 sites. As shown in Fig. 5B, LMP-1 and CD30 activated the wild-type promoter pGL3-hIκB-ζ( −853) activity. A single deletion of the κB2 site from the IκB-ζ reporter plasmid (κB2D) markedly inhibited LMP-1- or CD30-induced transactivation, whereas a single deletion of the κB1 site (κB1D) resulted in moderate activation. These data indicate that IκB-ζ κB2 site was necessary for transcription of IκB-ζ. Furthermore, double deletions (κB1/κB2D) completely abolished the LMP-1- or CD30-induced transactivation. Expression of RelA is sufficient to induce IκB-ζ expression. The predominant role of κB2 in the induction of IκB-ζ expression was further supported by the finding that a promoter with a deleted κB2 site failed to respond to overexpressed RelA (Fig. 5B, bottom panel). In contrast, a promoter with a deleted κB1 was slightly activated by RelA. Considered together, these data suggest that upregulation of IκB-ζ by LMP-1 and CD30 requires both the κB1 and κB2 sites of the IκB-ζ promoter.

Since the deletional analysis of the IκB-ζ promoter indicated that LMP-1 and CD30 activated transcription through both the κB1 and κB2 sites, it was important to identify the nuclear factors that bind to these sites. Using the κB1 and κB2 sequences in the IκB-ζ promoter as the probes in EMSA, NF-κB binding was detected in 293T cells transfected with LMP-1 (Fig. 6A, lane 2) and CD30 (Fig. 7A, lane 2). In addition, formation of these complexes was competed with excess of unlabeled wild-type κB1, κB2 and consensus IL-2R κB oligonucleotides (Figs. 6A and 7A, lanes 3, 5 and 7). In contrast, mutated oligonucleotides, κB1 mut and κB2 mut, and irrelevant consensus IL-8 AP-1 oligonucleotide, did not compete with the labeled probes (Figs. 6A and 7A, lanes 4, 6 and 8). Supershift analyses demonstrated that the κB1 and κB2 complexes contained both p50 and RelA subunits of the NF-κB family (Figs. 6A and 7A, lanes 9 and 10). To determine the role of LMP-1 and CD30 on endogenous NF-κB binding to DNA, we measured NF-κB binding to respective NF-κB sites in the IκB-ζ promoter in LMP-1-expressing Daudi cells and CD30-expressing L428 and HDLM-2 cells. As expected, protein complexes bound to both κB1 and κB2 sites were detected in nuclear extracts from Daudi, L428 and HDLM-2 cells (Figs. 6B and 7B, lane 1). The specificity of DNA-protein complexes in these extracts was determined by competition studies using unlabeled competitors. As observed in nuclear extracts from 293T cells transfected with LMP-1 and CD30 expression plasmids, unlabeled κB1 and κB2 oligonucleotides, and consensus NF-κB site from the IL-2Rα promoter, but not the mutated κB1 and κB2 oligonucleotides, and consensus AP-1 element, efficiently competed with the labeled probes (Figs. 6B and 7B, lanes 2–7). Antibodies against p50, RelA, c-Rel and RelB induced a supershift of the DNA-protein complexes in nuclear extracts of Daudi cells (Fig. 6B, lanes 8–10 and 12), whereas the κB1 and κB2 complexes contained p50, RelA, c-Rel, p52 and RelB in nuclear extracts of L428 and HDLM-2 cells (Fig. 7B, lanes 8–12). Taken together, the results indicate that NF-κB proteins bind to both κB elements of the IκB-ζ promoter in Daudi, L428 and HDLM-2 cells.

Unlike other IκB family members, IκB-ζ has dual opposite functions on the expression of different cellular genes activated by NF-κB (19–21,27). To identify the genes whose expression is regulated by IκB-ζ, we examined the promoter activities of several NF-κB target genes in 293T cells transfected with LMP-1, CD30 or RelA, and IκB-ζ. IL-6, IL-8 and Bcl-3 genes are known to be activated by NF-κB (35–37). Luciferase reporter analyses indicated that promoters of IL-6, IL-8 and Bcl-3 were activated by transfection of LMP-1, CD30 and RelA as expected (Fig. 8A–C). Cotransfection of IκB-ζ dose-dependently inhibited the LMP-1-, CD30- and RelA-induced activation of promoters of IL-6, IL-8 and Bcl-3. In addition, we analyzed whether IκB-ζ overexpression leads to downregulation of known NF-κB targets on mRNA level. To this end, we determined the effect of IκB-ζ overexpression on IL-6 and IL-8 mRNA expression in the presence of LMP-1 by RT-PCR. These analyses demonstrated that IL-6 and IL-8 mRNA levels were downregulated after IκB-ζ overexpression (Fig. 9), suggesting that IκB-ζ plays a role in negatively regulating NF-κB targets.

We also analyzed the effect of IκB-ζ overexpression on the activation of its promoter induced by LMP-1, CD30 and RelA. LMP-1-, CD30- and RelA-induced IκB-ζ promoter activation was dose-dependently repressed by IκB-ζ over-expression (Fig. 10A–C). Thus, IκB-ζ can repress its own transcription. IκB-ζ expression itself was regulated by NF-κB, suggesting that its activity is controlled through a negative feedback loop.

To confirm the role of IκB-ζ in NF-κB activity, we transfected 293T cells with IκB-ζ, and LMP-1, CD30 or RelA, and measured the activity of κB-LUC, an NF-κB reporter construct. As expected, we found that NF-κB reporter activity induced by LMP-1, CD30 and RelA was repressed by IκB-ζ overexpression in a dose-dependent manner (Fig. 11A, left panel, and 11B). However, the results showed that IκB-ζ did not affect AP-1 reporter activity induced by LMP-1 (Fig. 11A, right panel).

Mutants of IκB-ζ truncated from the amino- and carboxyl-termini were expressed in the presence of LMP-1 in 293T cells, and the NF-κB reporter activity was measured (Fig. 12A). The amino-terminal truncated mutants (153–728 and 188–728) as well as the full-length IκB-ζ (1–728), showed inhibitory activities against LMP-1-induced NF-κB activation, whereas the mutants consisting of the amino-terminus to amino acid 456 (1–456) and the amino-terminal truncated mutant (457–728) exhibited less activity than the full-length IκB-ζ (1–728) (Fig. 12B). These results indicate that the region between amino acids 188–728 harbors a domain with transcriptional inhibitory activity.