Background
While considerable evidence has shown that immunoglobulins (Igs) "unexpectly" expressed in malignant tumors of epithelial origin [
1‐
10], much less is known about the molecular mechanisms of nonlymphoid cells expressing Igs. In our previous work, we have also demonstrated that nonlymphoid NPC cells express immunoglobulin kappa light chain. In addition, we have found that EBV-encoded latent membrane protein 1 (LMP1) can upregulate the expression of kappa light chain in NPC cells and both NF-κB and AP-1 signaling pathways are involved in LMP1-augmented kappa light chain expression [
1]. These results promote us using of NPC cell lines as model to further explore the mechanisms underlying the expression of Ig kappa in nonlymphoid cells.
Expression of kappa light chain gene is under the control of distinct cis-regulatory elements, including the kappa intron enhancer (iE
κ) and the kappa 3' enhancer (3'E
κ) [
11,
12], which are located within the J
κ-C
κ region and downstream of C
κ region, respectively. Both enhancers are inactive at the pro-B and pre-B cell stages and active at the Igκ-expressing mature B cell and plasma cell stages. The activity of these enhancers in other non-kappa-producing cell lineages, such as T-lymphoid cells, epithelial cells and NIH3T3 fibroblasts, is generally silent [
11,
13]. Base on these, it is generally believed that the activation of iE
κ and 3'E
κ is required for immunoglobulin kappa gene expression and is B cell lineage-restricted events [
14,
15]. An interesting feature of kappa gene transcription is its inducibility. Certain agents, such as cycloheximide (CYC), phorbol esters and bacterial product lipopolysaccharide (LPS) can induce the activation of kappa enhancers and result in kappa gene expression at the pre-B cell stage [
16]. Nucleation of transcription factors PU.1, PIP, c-Fos and c-Jun on the kappa 3' enhancer core can cause a very dramatic induction in 3'E
κ activity in NIH3T3 fibroblasts, a cell in which the enhancer is normally silent [
13]. These findings reinforce the possibility of nonlymphoid cells expressing Ig kappa by certain unidentified mechanisms and suggest that other extracellular factors, such as gene products encoded by viruses, are also likely to induce kappa enhancers' activation, finally result in kappa gene transcription and expression.
One viral protein, latent membrane protein 1, is considered as a major oncogenic protein encoded by EBV for its transform and tumorigenic activities and is found to be able to transform cell lines and alter the phenotype of cells due to its oncogenic potential [
17]. Biologically, LMP1 is an integral membrane protein with six transmembrane segments that facilitate self-aggregation in the plasma membrane and transduces ligand-independent signals, including NF-κB, c-Jun-NH
2-terminal kinase (JNK), p38/MAPK, Ras/MEK/ERK/MAPK, PI3K/Akt and JAK/STAT [
18‐
23]. The nuclear factor κB (NF-κB) and c-Jun N-terminal kinase (JNK) signaling pathways are the most important, since their activation results in the overexpression of most LMP1 target genes [
24]. LMP1 can mimic CD40 signaling to induce B cell activation and differentiation in vivo. They share some molecules such as TRAF1, 2, 3, and 5 as signal transducers as well as some pathways such as NF-κB, JNK, p38/MAPK, PI3K/Akt and JAK/STAT pathways [
25]. In normal B cells, an important mechanism of Ig production is CD40 ligation-triggered cellular signaling pathways [
26]. In addition, it has been found that CD40 signaling can increase IgH 3' enhancer activity [
27]. These studies, in combination with our previous finding that kappa light chain is significantly higher in LMP1-positive than in LMP1-negative NPC cells [
1], we thus speculate that upregulation the expression of kappa light chain by LMP1 may be the result of LMP1-induced kappa enhancer's activation in NPC cells.
The function of enhancers is mediated by DNA binding proteins that recruit to the enhancer [
12]. Multiple protein binding sites have been identified and characterized in each of the kappa enhancers. A κB binding site within the iE
κ and the activity of iE
κ is mainly dependent on the nuclear factor NF-κB binding to κB element [
28,
29]. Deletion or mutation of the κB site abolishes the activity of iE
κ, suggesting it might as a crucial enhancer element [
16,
30]. In addition, the human kappa gene J
κ-C
κ region also contains a perfect consensus AP-1 site, which located ~320 bases downstream of the κB site. The κAP-l site in the context of the iE
κ can positively regulate the iE
κ activity and kappa expression in B cells, suggests that it plays a role in kappa gene regulation [
31]. However, in Igκ-expressing nonlymphoid cells, whether these two binding sites play roles in functional activation of iE
κ is still unknown.
Since kappa enhancers' activation is required for Ig kappa gene expression and their activations are generally considered as B cell lineage-restricted events, and since NF-κB and AP-1 binding sites exist within and downstream the iEκ enhancer, and on the basis of our previous findings that both NF-κB and AP-1 pathways are involved in LMP1-augmented Ig kappa expression in human NPC cells, we therefore focus on the iEκ enhancer and attempt to study further whether it is active in Igκ-expressing NPC cells and whether LMP1-upregulated kappa expression is correlated with the activation of iEκ via NF-κB and AP-1 pathways. In this study, luciferase reporter analysis demonstrate that the iEκ whose activation is required for immunoglobulin kappa gene expression indeed activates in Igκ-expressing NPC cells and stable or transient LMP1 expression can upregulate the activity of iEκ in NPC cells. Moreover, mutation analysis of κB or AP-1 binding site within or downstream the iEκ, inhibition of LMP1-mediated NF-κB and AP-1 signaling pathways by using specific chemical inhibitors and dominant inhibitory molecules indicate that both sites are functional and LMP1-enhanced iEκ activity is regulated, to some extent, through these two sites. Gel shift assays show that LMP1 promotes NF-κB subunits p52 and p65 as well as AP-1 family members c-Jun and c-Fos binding to the κNF-κB and the κAP-1 motifs in vitro, respectively. Both chemical inhibitors and dominant negative mutants targeting for NF-κB and AP-1 pathways can attenuate theLMP1-enhanced bindings. Co-IP assays using nuclear extracts from HNE2-LMP1 cells reveal that p52 and p65, c-Jun and c-Fos proteins interact with each other at endogenous levels. ChIP assays further demonstrate p52 and p65 binding to the κB motif as well as c-Jun and c-Fos binding to the AP-1 motif of Ig kappa gene in vivo. Based on the findings reported here, we conclude that the iEκ enhancer is active in NPC cells and is further activated by LMP1 via NF-κB and AP-1 pathways, which contributes to the upregulation of Ig kappa by LMP1 in NPC cells.
Discussion
In this article, we showed that the aberrant expression of Ig kappa light chain in NPC cells. Recent studies have demonstrated that the expression of Igs is widespread in epithelial cancers from many organs and includes basically all kinds of isotypes. Among heavy chains, α chain for IgA and γ chain for IgG are the mostly identified; but in light chain, only κ chain but not λ chain is confirmed. Moreover, several studies indicated that tumor-derived Igs have certain biological functions. Qiu et al [
5] found induction of cancer cell apoptosis and inhibition of cancer growth by blocking tumor-derived IgG, whose light chain is kappa, using either antisense oligodeoxynucleotide or anti-human IgG, thus confirming that IgG secreted by epithelial cancers has some unidentified capacity to promote the growth and survival of tumor cells. We also found that blockade of cancer-derived Ig alpha suppresses the growth and viability of cancer cells. Furthermore, we have demonstrated that cancer-derived Ig alpha promotes the malignant proliferation ability of cancer cells and increases the access percentage of S phase from the early mitosis of synchronized cancer cells [
38]. These findings support the important role of cancer-derived Ig as a growth factor of cancer cells. In addition, By
in situ hybridization to analyze kappa constant region mRNA in different stages of cervical tissue samples, we found that the expression of kappa constant region mRNA is markedly increased in uterine cervical epithelia with dysplasia and carcinoma, as compared with cervicitis, thus suggesting a closely associated of kappa light chain expression with cell malignancy and is associated with increasing tumor grades [
10]. Recently, we analyzed the ADCC immuno-activity of Ig derived from cancer cells and found that cancer-derived Ig is capable of reacting with FcR of monocytes and NK cells by its Fc region as does normal Ig, and to accomplish ADCC with effector cells (unpublished data). Based on these findings, it may be hypothesized that cancer-derived Ig could compete with B cell-derived Ig for the FcR on effector cells, thus inhibits ADCC and favors tumor immune escape. The potential biological functions of the tumor-derived Igs and the finding that nonlymphoid cells expressing Igs reported by different research groups revealed that this phenomenon is not a happenchance. However, the mechanisms underlying the expression of Igs in nonlymphoid cells are still unknown. In present study, we focus mainly on exploring the possible mechanisms by which nonlymphoid cells expressed Ig kappa and found that in Igκ-expressing NPC cells, kappa intron enhancer is activated. The activity of iE
κ can be further activated by LMP1-stimulated NF-κB and AP-1 aberrant activation. It could be concluded that LMP1 stimulates transcription factors NF-κB and AP-1 binding to the corresponding site in kappa gene via NF-κB, JNK/MAPK signal pathways and finally upregulates kappa light chain induction. Such mechanisms would explain, at least in part, LMP1-positive human epithelial cancer cells produce immunoglobulins.
The activation of kappa enhancers, whose function is mediated by proteins binding to the enhancers, is required for Ig kappa gene expression [
12,
14,
15]. We found the iE
κ is active in both LMP1-negative and LMP1-positive NPC cells. In LMP1-negative HNE2 cells, the iE
κ activity is relatively low and is in accord with low kappa expression level. LMP1 can further activate the activity of iE
κ and contributes to the upregulation of Ig kappa in NPC cells. Our results indicated that mutant of either NF-κB or AP-1 biding site did not completely abolish the basal and LMP1-induced iE
κ activities (Fig.
2B). In addition to NF-κB and AP-1 motifs to modulate the enhancer's activity, other positive regulatory elements have been identified within the iE
κ, including κA and E-box motifs, these sequences could potentially regulate the activity of iE
κ [
33,
34]. Therefore, other transcription factors bind to kappa gene through various signaling pathways to regulate kappa expression in NPC cells can not be excluded at this time.
In B cells, functional analyses of motifs within iE
κ performed using isolated enhancers to activate reporter genes in transfection assays have shown that B cell-specific activity of iE
κ depends substantially on the κB element [
39]. Moreover, mutations of E-box motifs have variable and weaker effects on transcription compared to mutation of the κB site [
16]. These observations indicated NF-κB acts as the master and commander of kappa gene expression via the κB motif in iE
κ in B cells. Similarly, our result indicated mutation of the NF-κB motif displayed a more inhibitory effect on LMP1-increased iE
κ activity compared to mutation of the AP-1 site, suggested that of NF-κB and AP-1 pathways, NF-κB pathway may play a leading role in LMP1-augmented iE
κ activity in NPC cells.
Transcripition factor NF-κB comprised of homo- and heterodimers of the p65 (RelA), RelB, c-Rel, p50/p105 (NF-κB1) and p52/p100 (NF-κB2) polypeptides can both induce and repress gene expression by binding to discrete κB elements in promoters and enhancers [
40]. NF-κB is found in the cytoplasm of pre-B cell lines as an inactive complexes associated with an IκB inhibitor, whereas in mature and transformed B cells, NF-κB is active and localized in the nucleus. NF-κB DNA binding activity and nuclear relocalization can be activated by a variety of stimuli. In our previous study, IκBα phosphorylation accompanying by IκBα degradation has been found in NPC cells and LMP1 can further induced IκBα phosphorylation and degradation [
1]. Our results presented here indicated LMP1 increased the released-NF-κB translocating freely to the nucleus (Fig.
4E) and binding to the κB motif of iE
κ (Fig.
4A, lane 3). We characterized the NF-κB/DNA complex containing p52 and p65 subunits by Gel Super-shift assay (Fig.
4C, lanes 5 and 6). We also found LMP1 induced the processing of p100 to p52 (Fig.
4F) and the nuclear translocation of p52 (Fig.
4G). Generally, p50/p65 is considered as a 'classical' heterodimers. p52 forms heterodimers with other NF-κB subunits, such as p65 and RelB, or as a homodimer has also been found [
41]. However, in our experiments, we failed to detect p50, c-Rel and RelB subunits in NF-κB/DNA complex. We also confirmed the interaction of p52 and p65 at endogenous levels by co-IP assay (Fig.
5). Moreover, both p52 and p65 could directly bind to the NF-κB binding region within the iE
κ enhancer (Fig.
8B). Perkins [
42] found that p52/p65 preferentially activates HIV-1 gene expression relative to the p50/p65 heterodimers, which is similar to our results. The results suggest that a heterodimer of p65 with p52 subunit binding to κB site within the iE
κ may play an important role in upregulating the activity of iE
κ and kappa light chain production in HNE2-LMP1 NPC cells.
We reported earlier that NPC cells express activated forms of JNK (pJNK) whether LMP1-negative or LMP1-positive and LMP1 can increase the phosphorylation level of JNK [
1]. JNK is one of the kinases that regulated the activation of AP-1 transcription factor. Upon stimulation, this protein kinase enters the nucleus to induce or phosphorylate subunits of AP-1 and the resultant enhanced AP-1 activity can then participate in the regulation of gene expression. The AP-1 transcription factor is a dimeric complex that comprises a group of structurally and functionally related members of the Jun family (c-Jun, JunB and JunD), Fos family (c-Fos, FosB, Fra-1 and Fra-2), ATF (ATFa, ATF-2 and ATF-3) and JDP (JDP-1 and JDP-2) subfamilies, which can bind to AP-1 consensus sequence 5'-TGAG/CTCA-3' [
43]. Different types of AP-1 complexes are functionally distinct and may activate different target genes [
44]. By EMSA analysis, we showed that nuclear extracts of both HNE2 and HNE2-LMP1 cells could bind κAP-1 motif and LMP1 was able to increase this binding (Fig.
6A, lanes 2 and 3). Super-EMSA further characterized the protein/DNA complex containing c-Jun and c-Fos transcription factors (Fig.
6C). Moreover, our results demonstrated LMP1-induced JNK phosphorylation level coincided with the phosphorylation level of c-Jun at Ser63 and Ser73 in the nucleus (Fig.
6D) and this c-Jun phosphorylation was much closely related to the DNA binding activity of the c-Jun/c-Fos heterodimer. Similar results that the phosphorylation level of c-Jun (ser63, ser73) is related to the DNA binding activity of c-Jun/JunB heterodimer was reported [
23]. Our results suggest that LMP1 can act through activation of JNK, a c-Jun N-terminal kinase needed for AP-1 activation and induce formation of the c-Jun/c-Fos/DNA complex to upregulate the activity of iE
κ in NPC cells. We also found stable expression of TAM67 almost completely blocked LMP1-induced AP-1 DNA binding in HNE2-LMP1 cells (Fig.
6A, lane 4). Similar results that stable TAM67 expression completely inhibited MKK6-induced AP-1 binding in MCF-7 cells [
45] and an inhibition of nickel-induced AP-1 element binding by TAM67 in human bronchial epithelial cells [
46] were recently reported. Although we have demonstrated the heterodimerization of c-Jun and c-Fos (Fig.
7) and this heterodimer can directly bind to the AP-1 site located near the iE
κ enhancer (Fig.
8C), we have used only c-Jun and c-Fos in this report, therefore, other dimeric forms of AP-1 transcription factor involved in regulating the iE
κ activity in NPC cells can not be excluded at this time.
Methods
Cell lines and cell culture
HNE2, HNE2-LMP1, HNE2-LMP1-DNMIκBα and HNE2-LMP1-TAM67 cell lines used were as previously described [
1]. All the cell lines were maintained in RPMI1640 (GIBCO) supplemented with 10% FBS (GIBCO), 1% glutamine, and 1% antibiotics at 37°C in humidified atmosphere with 5% CO2.
Chemicals and cell treatments
The selective JNK inhibitor SP600125 (Cat No.420119, Calbiochem) and NF-κB inhibitor Bay11-7082 (Cat No.196870, Calbiochem) were prepared as a stock solution of 20 mM in dimethylsulfoxide (DMSO, Sigma). Subconfluent cells were treated with the compound at indicated concentrations for indicated time. Detailed treatment procedures were described in figure legends. The final concentration of DMSO in the culture media was kept less than 0.1% which had no significant effect on the cell growth.
Plasmid constructs
The human Iα promoter was a 342 bp promoter fragment identical to that used previously [
32], obtained by amplification from human HNE2 cells genomic DNA. The sense primer 5'-
gagctc ctctgtctcggggtctctga-3' used in this reaction was carrying
Sac I cloning site whereas the antisense primer 5'-
aagctt ccgtctgtccttagcagagc-3' had
Hind III site. Italic nucleotides represent restriction endonuclease recognition sites. This fragment was inserted into the
Sac I/
Hind III sites of the pGL3-Basic vector (Promega) and the plasmid was designated as pGL3-α.
A 575 bp fragment containing the intact human iE
κ and the AP-l binding site at the 3' flank of iE
κ was cloned. Briefly, a 575 bp DNA fragments containing human kappa light chain genomic sequences were amplified from HNE2 cells genomic DNA by PCR using specific primers from the human Ig kappa gene (GenBank accession no. NG_000834): 5'-
ggatcc ctgacttctccctatctgtt-3'(sense), which contains an artificial
BamH I site, and 5'-
gtcgac ccattctgagggctttgc-3'(antisense), which contains an artificial
Sal I site. Italic nucleotides represent restriction endonuclease recognition sites. The PCR-amplified fragments were then digested with
BamH I/
Sal I and inserted into the corresponding restriction sites of the pGL3-α plasmid described above to generate pα-iE
κwt. The PCR products were confirmed by their size, as determined by electrophoresis and by DNA sequencing. The NF-κB motif and the AP-1 motif mutants (designated as pα-iE
κ-mtκB and pα-iE
κ-mtAP-1, respectively) from pα-iE
κwt were generated by PCR based on an overlap extension technique [
47]. The primers used for generating mutations were: 5'-ccccagag
a g a gatt g cc aagaggccacctg-3' and 5'-tt
gg c aatc t c t ctctgggggattc-3' (for NF-κB site), 5'-gaggctttcct
g gactca gccgctgcc-3' and 5'-gc
tgagtc c aggaaagcctccg-3' (for AP-1 site). PCR-amplified fragments carrying the desired mutations were then cloned into
BamH I/
Sal I sites of the pGL3-α plasmid. Bold nucleotides represent sequences of κNF-κB (
ggggatttcc) and κAP-1(
tgactca) motifs and bold italic nucleotides represent mutated nucleotides. The expected mutations and the integrity of the enhancer were confirmed by automated sequencing using an Applied Biosystems sequencer and software (Foster City, CA).
The pSG5-based expression vector for wild-type LMP1 derived from B95.8 EBV strain was kindly provided by Dr. Izumi (Brigham and Women's Hospital). Expression plasmid of dominant negative mutant of IκBα (DNMIκBα), which had a deletion of 71 amino acids at the N terminus and was cloned into the multiple cloning sites of pcDNA3, was kindly provided by Dr. Krappmann (Max-Delbruck-center for Molecular Medicine, Berlin, Germany). Expression plasmid of mutant c-Jun (TAM67) was constructed by inserting the TAM67 sequence into the vector pGem3z which contains a human keratin 14 promoter and a human growth hormone segment, was kindly provided by Dr. J. Li (NCI, Frederich, U.S.A.).
Luciferase reporter assays
The pGL3-α, pα-iEκwt, pα-iEκ-mtκB and pα-iEκ-mtAP-1 luciferase reporter plasmids described above were used in conjunction with the control pGL3-Basic vector (Promega) and the internal control plasmid pRL-SV40 (Promega). Cells were cultured in 24-well plates at a density of 1 × 105 per well overnight and were transfected with Lipofectamine™ 2000 (Invitrogen) as per the manufacturer's instructions. Each transfection contained 800 ng/well of firefly luciferase reporter and 80 ng/well of internal control pRL-SV40 or contained 400 ng/well of firefly luciferase reporter and 80 ng/well of internal control pRL-SV40 together with 200 ng/well of each expression plasmid or blank expression plasmid necessary to normalize the amount of DNA transfected. 24 hrs after transfection, cells were either left untreated or treated with 20 μM Bay11-7082, 20 μM SP600125 or 0.1% DMSO for 12 hrs. Cells were harvested at 36 h after transfection and lysates were analyzed for luciferase activity using the Dual Luciferase Reporter assay (Promega) according to the manufacturer's directions with a GloMax™ Microplate Luminometer (Promega). The luciferase reporter plasmids were co-transfected with pRL-SV40 to correct for variations in transfection efficiency. The relative luciferase activity normalized to the value of pRL-SV40 activity. Results were expressed as fold induction of pGL3-Basic activity, which was assigned a value of 1. The data represent the mean ± SD of the three independent experiments performed in triplicate.
Western blot analysis
Whole cell lysates preparation and western blot analysis were performed according to the method previously described [
1]. Nuclear or cytoplasmic extracts were prepared by the use of NE-PER Nuclear and Cytoplasmic Extraction Kit (Cat. No.78833, Pierce) in accordance with the manufacturer's protocol. Protein concentration was determined by BCA Assay Reagent (Cat. No.23228, Pierce). The following antibodies were used for immunodetection with appropriate dilutions: mouse LMP1 monoclonal antibody (CS.1-4, DAKO); p52(sc-298), p65(sc-8008), c-Jun(sc-44), c-Fos(sc-52), nucleolin(sc-8031), α-tubulin(sc-5286), goat anti-rabbit IgG-HRP (sc-2004), goat anti-mouse IgG-HRP (sc-2005) and donkey anti-goat IgG-HRP (sc-2020) (all from Santa Cruz); phospho-JNK(Thr183/Tyr185)(9251), phospho-c-Jun(Ser63)(9261S) and phospho-c-Jun (Ser73) (9164S) (all from Cell Signaling Technology).
Electrophoretic mobility shift assay (EMSA)
Nuclear extracts were prepared by the use of NE-PER Nuclear and Cytoplasmic Extraction Kit (Cat. No.78833, Pierce) in accordance with the manufacturer's protocol. The protein concentration in nuclear extracts was determined using the BCA protein assay reagent (Cat. No.23228, Pierce) and EMSAs were carried out using aliquots containing equal amounts of protein. EMSA analysis was performed using the LightShift™ Chemiluminescent EMSA Kit (Cat. No.20148, Pierce) following the manufacturer's instructions. The reaction mixtures (20 μl) containing about 10 μg nuclear extracts were incubated with 2 nmol/L of the biotin-labeled double-stranded oligonucleotide probes in reaction buffer (Pierce) for 20 min at room temperature. Samples were subjected to electrophoresis in 5% nondenaturing polyacrylamide gel and transferred to Biodyne™ B Nylon membrane (Cat. No.77016, Pierce). For competition analyses, 200-fold excess of the unlabeled wild-type or mutant or nonspecific probe was included in the binding reaction. For antibody supershift experiments, the reaction mixtures were preincubated with 2 μg of p50(sc-8414X), p52(sc-298X), p65(sc-8008X), c-Rel(sc-272X), RelB(sc-226X), c-Jun(sc-44X), c-Fos(sc-52X) and rabbit IgG(sc-2027) antibody (all from Santa Cruz) at room temperature for 1 hr. The complementary oligonucleotides used as probes or competitors were listed below: the human κNF-κB oligonucleotides used were 5'-ccagagggggatttcc aagaggcca-3' and 5'-tggcctcttggaaatcccc ctctgg-3', derived from the sequence of the NF-κB site within the human kappa intron enhancer. The human κAP-l oligonucleotides used were 5'-gctttccttgactca gccgctgcc-3' and 5'-ggcagcggctgagtca aggaaagc-3', derived from the AP-1 sequence ~320 bp 3' of the κNF-κB site. The nonspecific oligonucleotides used as competitor DNA for κNF-κB and κAP-l were κAP-l and κNF-κB oligonucleotides, respectively. The mutated κNF-κB oligonucleotides used were 5'-ccagaga g a gatt g cc aagaggcca-3' and 5'-tggcctcttgg c aatc t c t ctctgg-3' (designated as mutκB). The mutated κAP-l oligonucleotides used were 5'-gctttcctg gactca gccgctgcc-3' and 5'-ggcagcggctgagtc c aggaaagc-3' (designated as mutAP-l). Binding sites were shown in bold type and mutated nucleotides were shown in bold italic. The mutated oligo probes for NF-κB and AP-1 binding sites in EMSAs were identical to those of the mutated sequences in the reporter gene constructs.
Co-immunoprecipitation (Co-IP)
Non-denatured nuclear proteins were purified using NE-PER Nuclear and Cytoplasmic Extraction Kit (Cat. No.78833, Pierce) according to the manufacturer's instructions. Protein concentration was determined by BCA Assay Reagent (Cat. No.23228, Pierce). 200 μg of nuclear extracts prepared from HNE2-LMP1 cells were mixed with 40 μl protein A-Sepharose beads (Sigma) in the immunoprecipitation assay buffer (1× PBS, 0.5% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS), incubated at 4°C for 2 h with gentle agitation and centrifuged for 2 min at 2000 rpm for preclearing. The recovered supernatant was incubated with 2 μg of an antibody to a member of the complex in the presence of 1× protease inhibitors at 4°C overnight with mild shaking. Then, 50 μl of protein A-Sepharose beads was added, and the incubation was continued for 2 h at 4°C with gentle shaking. Protein A-precipitated protein complex was recovered by brief centrifugation, followed by three times washes with immunoprecipitation assay buffer. The harvested beads resuspended in 30 μl of 2× SDS PAGE sample buffer were boiled for 5 min to release the bound protein. The samples were then analyzed by Western blot with a specific antibody to another member of the complex. A 20 μg aliquot of nuclear extract was used as an input control. The same membrane was stripped by incubating at 50°C for half an hour in stripping buffer [100 mM β-mercaptoethanol, 2% (wt/vol) sodium dodecyl sulfate and 62.5 mM Tris-HCl (pH 6.8)] and reprobed with the corresponding IP antibody.
Chromatin immunoprecipitation (ChIP) assay
ChIP was performed using the ChIP assay kit (Upstate Biotechnology, Lake Placid, NY) and was then conducted according to the manufacturer's recommendations. Briefly, formaldehyde solution was added directly to HNE2-LMP1 cells at a final concentration of 1% at room temperature for 10 min. Then the cells was neutralized with glycine at room temperature for 5 min and washed twice with ice-cold 1× phosphate-buffered saline containing protease inhibitors. The cells were lysed by SDS lysis buffer with protease inhibitors. Chromatin in the lysate (350 μl) was sheared by sonication with a Branson Sonifier Cell disruptor B15 (output control 4, duty cycle 40%), with 14 cycles of 20-second pulses and 20-second intervals to an average length of ~500 bp as determined by 2% agarose gel electrophoresis. The suspension was precleared with salmon sperm DNA/protein A/agarose-50% slurry for 1 h at 4°C. After "precleared" the chromatin, a small aliquot (10 μl) was saved as "input DNA" for PCR analysis later. Other each 100 μl aliquots of sheared cross-linked chromatin were incubated with 2 μg each of antibodies p50(sc-8414X), p52(sc-298X), p65(sc-8008X), c-Rel(sc-272X), RelB(sc-226X), c-Jun(sc-44X), c-Fos(sc-52X), rabbit IgG(sc-2027) (Santa Cruz), or no Ab overnight at 4°C with mild shaking. The immune complexes were incubated with salmon sperm DNA/protein A/agarose-50% slurry with mild shaking for 2 h at 4°C, washed, and eluted. Cross-links were reversed by 5 M NaCl. After proteinase K digestion, DNA in samples was phenol extracted, ethanol precipitated, and resuspended in 50 μl of ddH2O. Two microliters of DNA solution was used for 36 cycles of PCR amplification. PCR products were analyzed by electrophoresis on a 2% agarose gel and visualized by ethidium bromide staining. The following primers were used in the ChIP assays: human iEκ enhancer including the NF-κB-binding region, 5'-ctactgctctcccacccaac-3' and 5'-tgcagcaattttcagccata-3'(159 bp); the AP-1-binding region located downstream the human iEκ enhancer, 5'-gcctgttatcccagcacagt-3' and 5'-tgcatgcttttctgaccttg-3'(188 bp).
Statistical analysis
All statistical calculations were performed with the statistical software program SPSS ver.12.0. Differences between various groups were evaluated by the Student's t test. The difference was of statistical significance, when P < 0.05.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
HDL carried out all experiments and drafted the manuscript. HZ, ZD, DSH participated in the design of the study and assisted with the reporter gene analysis. SFL, ZJL participated in the statistical analysis. ML, XYD contributed in the design of the study and data interpretation. ZLW, MT, YS, WY assisted with the construction of luciferase reporter plasmids and participated in the sequence alignment. YC conceived of the study and participated in the design and coordination. All authors read and approved the final manuscript.