A Role for NF-κB in the Induction of β-R1 by Interferon-β

Previous experiments have suggested that induction of the β-R1 gene by interferon (IFN)-β required transcription factor ISGF-3 (IFN-stimulated gene factor-3) and an additional component. We now provide evidence that nuclear factor-κB (NF-κB) can serve as this component. Site-directed mutagenesis of an NF-κB binding site in the β-R1 promoter or over-expression of an IκBα super-repressor abrogated IFN-β-mediated induction of a β-R1 promoter-reporter. IFN-β treatment did not augment abundance of NF-κB but did lead to phosphorylation of the p65 NF-κB subunit. It is proposed that IFN-β-mediated enhancement of the transactivation competence of NF-κB components is required for inducible transcription of the β-R1 promoter. These results provide a novel insight into the role of NF-κB in the transcriptional response to IFN-β.

Interferons (IFNs) 1 elicit multiple biological responses mediated by the proteins encoded by interferon-stimulated genes (ISGs) (1). IFN-␣ and -␤ activate transcription of ISGs through the transcription factor interferon-stimulated gene factor-3 (ISGF-3), which interacts with the interferon-stimulated response element (ISRE) present in the promoters of ISGs (2). We previously reported the selective induction of ␤-R1 by IFN-␤ but not IFN-␣ in human fibrosarcoma cells (3,4). ISGF-3 was essential but not sufficient for IFN-␤-dependent induction of ␤-R1 (3). In the current report we show that the transcription factor nuclear factor-B (NF-B) is also involved in the induction of this gene by IFN-␤.
NF-B is a dimer of members of the Rel family of proteins, including p50, p52, c-Rel, RelA/p65, and RelB. NF-B dimers bind to DNA segments collectively referred to as B elements (5). In unstimulated cells, NF-B is sequestered in the cytosol with oligomeric inhibitory components, termed IB. Stimulation of cells with appropriate inducers results in the rapid degradation of IB and translocation of NF-B to the nucleus, where it binds target sequences to initiate transcription. Increasing evidence suggests that a second signaling pathway, independent of IB␣ degradation (6 -10), culminates in phosphorylation of the p65/RelA subunit in its transactivation domain (11).
A role for NF-B in the induction of ISG in response to IFNs has not been reported. While studying IFN-␤ activation of ␤-R1 promoter, we found that modification of pre-existing NF-B by phosphorylation was essential for IFN-␤-induced transcription of ␤-R1.
Promoter-Reporter Plasmid Construction and Mutagenesis-Site-directed mutagenesis of the ISRE and B sequence in pGL3-wt-␤-R1 (4) was achieved by multiple rounds of polymerase chain reaction using pGL3-wt-␤-R1 as template and appropriate primers (14) . Three rounds of polymerase chain reaction were performed using two sets of primers to obtain mISRE-␤-R1. The first pair of primers used to make the pGL3-mISRE mutant were 5Ј-GAAGAGAACAccACAtAAACTCTTGG-AAGC-3Ј (forward primer with ISRE sequence underlined and mutagenized nucleotides in lowercase) and BglII reverse primer 5Ј-TTGGA-AGATCTAGTAGAAATG-3Ј. The second pair of primers were SacI primer 5Ј-ATACGAGCTCTCCGCTGC-3Ј and 5Ј-GCTTCCAAGAGTTT-aTGTggTGTTCTCTTC-3Ј (backward primer with ISRE sequence underlined and mutagenized nucleotides in lowercase). The first pair of primers used to make pGL3-mB mutant were 5Ј-GCATGACTCAAA-GAGtGAAATTaCTGTGCCAT-3Ј (forward primer with NF-B sequence underlined and mutagenized nucleotides in lowercase) and BglII reverse primer 5Ј-TTGGAAGATCTAGTAGAAATG-3Ј. The second pair of primers were SacI primer 5Ј-ATACGAGCTCTCCGCTGC-3Ј and 5Ј-ATGGCACAGtAATTTCaCTCTTTGAGTCATGC-3Ј (backward primer with B binding sequence underlined and mutagenized nucleotides in lowercase). Plasmid DNA from two clones for each set of cloning was sequenced to verify the nucleotide sequence.
Cell Extracts and EMSA-For electrophoretic mobility shift assays (EMSA) the NF-B binding site (5Ј-GAGCAGAGGGAAATTCCGTAAC-* This work was supported by National Institutes of Health Grant 2PO1 CA62220 (to G. R. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. TT-3Ј) from the hIP-10 gene was used as a probe. Nuclear extracts were prepared and gel-shift assays performed as described earlier (18). Metabolic Labeling of Cells and Immunoprecipitation of p65/RelA-U1.wt cells were labeled with [ 32 P]orthophosphate (200 Ci/ml) as described (7). Cells were stimulated for 5 min with IFN-␤ (5000 units/ ml) or TNF-␣ (50 ng/ml). Lysis and immunoprecipitations were performed using anti-p65 (Santa Cruz Biotechnology, Santa Cruz, CA) as described (7,11). Phosphorylated proteins were separated by 10% SDS-PAGE and autoradiographed.
In Vitro Phosphorylation of p65 Transactivation Domain-The carboxyl terminus of p65 fused to GST obtained from Dr. H. Sakurai (8) was expressed in E. coli and cells lysed in 0.5% Nonidet P-40 lysis buffer as described (19) and bound to glutathione-Sepharose beads (Amersham Pharmacia Biotech, Uppsala, Sweden). In vitro phosphorylation was performed using ϳ1 g of GST-fused p65 protein in 23 l of kinase buffer containing 3 Ci [␥-32 P]ATP (19) . Nuclear extracts from IFN-␤-, IFN-␣-, or TNF-␣-treated cells (3 g of protein in a volume of 2 l) were added to the reaction and incubated at 30°C for 30 min (8,19). Following the kinase reaction, phosphorylated protein was separated by 10% SDS-PAGE and autoradiographed. Western analysis with anti-GST antibody (Amersham Pharmacia Biotech) was used to quantitate the total amount of p65 protein.
We proposed that a second cis-element might be involved in the induction of the ␤-R1 gene (3). We had previously observed 3 synergistic induction of ␤-R1 by IFN-␤ and TNF-␣, the latter being a well characterized activator of transcription factor NF-B. Hence, a salient choice for analysis was the B element on the promoter of ␤-R1.
Using transient transfection assays, wt-␤-R1 or mB-␤-R1 promoter-reporters were analyzed for IFN-␤ responsiveness. Mutation of the B element on the promoter of ␤-R1 markedly reduced IFN-␤ induction of the promoter, by 70% (Fig. 1b). Induction of the wild-type and mB-␤-R1 constructs by IFN-␥ differed only slightly (Fig. 1b). This result indicated that the B element was specifically required for IFN-␤ but not IFN-␥mediated induction.
Requirement of Basal NF-B Nuclear Activity for Induction of ␤-R1 Gene by IFN-␤-We used EMSAs to address whether the requirement for the B binding site was associated with IFN-␤ induced nuclear translocation and DNA binding by NF-B. Analysis of extracts of U1.wt cells revealed, as previously reported (20,21), basal NF-B DNA binding activity (Fig. 2a) by densitometric analysis of the ratios of the specific NF-B complexes to a nonspecific band (Fig. 2b).
To address the possibility that basal nuclear NF-B activity might be essential for induction of ␤-R1 by IFN-␤, we analyzed two sibling cell lines that differed in basal NF-B. Wild-type 293 cells did not contain basal nuclear NF-B as detected by EMSA, whereas Z5 cells contained constitutive NF-B (Fig. 2c). The major B binding complex in Z5 cells consisted of p65 and p50 subunits (Fig. 2d). IFN-␤ treatment did not produce a detectable change in NF-B in either cell line. RNase protection analysis documented the induction of ␤-R1 by IFN-␤ in Z5 cells but not in 293 cells (Fig. 2e), although another type I IFN-inducible gene (6 -16) could be induced equally well in both cell lines. TNF-␣, in combination with IFN-␤, mediated synergistic induction of ␤-R1 in both Z5 and 293 cells, confirming the existence of an intact NF-B pathway in both cell lines.
We also over-expressed super-repressor mutant isoform of IB␣ in U1.wt cells, resulting in diminished induction of the wt-␤-R1 promoter-reporter by IFN-␤ (Fig. 2f). To exclude nonspecific effects of IB␣ over-expression, we analyzed IFN-␤mediated induction of a promoter-reporter construct derived from the p561 gene (a type I IFN-induced gene). Induction of the 561 promoter-reporter (which lacks a B site) was unaffected by IB␣ (Fig. 2f). In transient transfection assays, a 6xB promoter-reporter construct was not activated by IFN-␤, but TNF-␣ gave a 6-fold induction (Fig. 2g).
Roles of p65 and p50 Proteins in the Induction of ␤-R1 by IFN-␤-We asked whether homodimers of p50 (which lack a transactivation domain and are transcriptionally inactive) were sufficient to promote ␤-R1 transcription by virtue of occupancy at the B site. For these experiments, p50 was overexpressed in U1.wt cells along with the wt-␤-R1 promoterreporter, and reporter gene activation in response to IFN-␤ was examined. IFN-␤ responsiveness of the promoter was reduced by 60% to a level similar to that obtained by disruption of the NF-B binding site (Table I). This result indicated that p50 homodimers were insufficient to support transcription of ␤-R1 and suggested the requirement of transcriptionally competent NF-B complexes.
Co-expression of p65 with the wt-␤-R1 promoter increased the basal activity of the promoter by 9-fold (Table I). IFN-␤ treatment further increased promoter activity by 6-fold. Taken together these results indicated that transcriptionally competent NF-B complexes were required for maximal transcription of the ␤-R1 gene in response to IFN-␤.
Basal NF-B activity was essential for IFN-␤-mediated transcription of ␤-R1, although IFN-␤ treatment did not lead to increased NF-B (by EMSA). Recent reports (6, 8 -11, 27) have shown that phosphorylation of p65 augments NF-B-dependent gene transcription. To address IFN-␤-mediated phosphorylation of NF-B proteins, cells were metabolically labeled with [ 32 P]orthophosphate, and whole cell extracts were immu-noprecipitated with anti-p65 antibodies, which co-precipitated p65, p50, and IB (Fig. 3a). The identity of the phosphorylated proteins was confirmed by Western analysis (results not shown). This analysis revealed increase in p65 phosphorylation in cells treated with 1FN-␤ or TNF-␣ (Fig. 3a). Phosphorylation of p65 in untreated U1.wt cells was anticipated, as these cells contain basal NF-B nuclear activity. We also detected IFN-␤mediated increases in p50 and IB phosphorylation.
Strikingly, phosphorylated IB␣ remained intact in IFN-␤treated cells but was rapidly degraded in TNF-␣-treated cells. Between 5 min and 1 h, we observed rapid IB␣ degradation in response to TNF-␣ but not to IFN-␤ (Fig. 3c).
Experiments using GST-fused p65 revealed that an inducible p65 phosphorylating activity was present in nuclear extracts of IFN-␤-or TNF-␣ treated cells (Fig. 3b) but not in IFN-␣-treated cells (Fig. 3d). DISCUSSION Here, we address the role of NF-B in the induction of ␤-R1 by IFN-␤. We had previously hypothesized that ISGF-3 was  b Luciferase activity on transfection of wt-␤-R1 in the absence of treatment was set at 1. This value was used to normalize the effects of treatment with IFN-␤ or over-expression of NF-B subunits. c 0.25 g of pCMV-65 was used along with 10 g of wt-␤-R1 promoter-reporter construct. Basal activity was higher for cells transfected with p65 in the absence of treatment. d -Fold induction in luciferase activity by IFN-␤ was significantly less upon co-transfection of wt-␤-R1 and pCMV-50 compared to wt-␤-R1 with empty vector (p Ͻ 0.001, paired t test). essential but not sufficient for the induction of the gene and proposed that an additional component might be required (3). We studied NF-B because of the striking synergy between TNF-␣ and IFN-␤ for induction of ␤-R1, despite the fact that neither TNF-␣ nor interleukin-1␤ alone induced ␤-R1.
Site-directed mutagenesis of the B element in the ␤-R1 promoter or over-expression of a super-repressor mutant of IB␣ blocked the induction of the ␤-R1 gene by IFN-␤, indicating that NF-B components were required. However, IFN-␤ did not induce increased NF-B DNA binding activity by EMSA. Further IFN-␤ treatment did not activate transcription from a NF-B reporter-construct.
To address the question of whether the physical presence of B binding activity was sufficient to enhance the transcriptional response of the ␤-R1 gene to IFN-␤, we individually over-expressed transcriptionally inert p50 or transcriptionally active p65. Over-expression of p50 suppressed ␤-R1 transcription, whereas over-expression of p65 elevated basal ␤-R1 promoter activity but supported a further 6-fold increase in response to IFN-␤. These results confirmed a requirement for active NF-B complexes to drive maximal transcription of the ␤-R1 gene.
In Daudi and other cell lines, type I IFN was reported to signal to NF-B via a pathway involving PI3K and serinethreonine kinase Akt (22). The output of this pathway comprised markedly increased NF-B. IFN-mediated endogenous NF-B-dependent gene expression was not documented in these or prior reports (22)(23)(24). The signaling pathway described in this report (22), implicated STAT3 as a docking site for PI3K components. Such signaling is not observed in HT1080 cells, where STAT3 activation is dispensable for generating PI3K lipid kinase activity and where physical interaction between STAT3 and PI3K does not occur (25). Given these dissimilar signaling characteristics of type I IFN receptor activation in HT1080 cells as compared with Daudi cells (26), it was expected that increased NF-B DNA binding activity would not be observed after IFN-␤-treatment of HT1080 cells.
Recent reports highlighted the importance of p65 phosphorylation for NF-B-dependent gene transcription (6, 8 -11, 27). Using in vivo metabolic labeling and in vitro phosphorylation of GST-p65, we showed that p65 was phosphorylated in IFN-␤-treated cells. A number of kinases have been implicated in the phosphorylation of p65 (8,9,11).
This report adds to the multiple roles of NF-B. Because there is no cytokine-dependent increase in NF-B DNA binding activity, IFN-␤-treated cells represent an excellent system in which to investigate post-translational modification of p65.