Activation of Nuclear Transcription Factor NF-κB by Interleukin-1 Is Accompanied by Casein Kinase II-mediated Phosphorylation of the p65 Subunit*

In fibroblasts and hepatoma cells, interleukin-1 (IL-1) treatment results in the rapid nuclear accumulation of the transcription factor NF-κB, present largely as p65 (RelA)/p50 heterodimers. It is well established that this process is dependent in large part upon the phosphorylation and subsequent degradation of the cytosolic inhibitor IκB. We looked for other IL-1-induced modifications of NF-κB components and found that, in both cell types, IL-1 stimulation led, within minutes, to phosphorylation of both NF-κB p65 and p50. Phosphorylation of p65 was sustained for at least 30 min after addition of the cytokine and occurred principally upon serine residues. Immunoprecipitates of NF-κB complexes contained an associated protein kinase, the biochemical characteristics of which were indistinguishable from casein kinase II (CKII). Purified CKII efficiently phosphorylated p65 in vitro, apparently on the same major sites that became phosphorylated in intact IL-1-treated cells. Although IL-1 treatment caused little apparent stimulation of total cellular CKII activity, the fraction that was specifically associated with NF-κB complexes was markedly elevated by the cytokine. The association of CKII with NF-κB occurred in the cytoplasm, suggesting that this phosphorylation might be involved either in control of translocation of the activated complex or in modulation of its DNA binding properties.

Interleukin-1 (IL-1) 1 receptors are widely distributed on cells of many lineages (1). Thus IL-1 can exert effects on connective tissues, for example inducing mitogenesis through secretion of PDGF; effects on lymphoid cells, such as inducing up-regulation of surface immunoglobulin and secretion of other cytokines; and effects on hepatocytes, for example inducing secretion of acute phase proteins such as C-reactive protein. All of these responses are triggered through a single type of receptor (IL-1R1) (2), which is the archetype for a diverse family of proteins. Included in this group are the insect Toll proteins (3), which function in the establishment of dorsal-ventral polarity during embryogenesis, and the product of the tobacco N-gene (4), which mediates resistance to the pathogen tobacco mosaic virus. Many IL-1-responsive genes are regulated through activation of Rel family transcription factors, in particular NF-B p65/p50 heterodimers, and indeed, ligation of Toll ultimately signals the activation of Dorsal, a Rel-related protein (5).
NF-B is a term used to refer to a group of transcription factors/DNA binding activities that recognize a consensus motif, 5Ј-GGGRNNYYCC-3Ј. All of these factors are dimers composed of subunits that share a common domain prototypically found in c-Rel (6). The transactivating activity of the NF-B factors is regulated by a distinct set of proteins termed IBs, which are characterized by a series of ankyrin repeats. IBs function by sequestering NF-B dimers in the cytosol. (7). A variety of stimuli can induce activation of NF-B (8), including viruses, lipopolysaccharides, and pro-inflammatory cytokines such as IL-1 and tumor necrosis factor (TNF) ␣. Activation of NF-B involves its translocation from the cytoplasm to the nucleus, following dissociation from one or more IBs. Cytokines bring about this dissociation by activating a newly identified protein kinase complex (9) that phosphorylates IB at two critical residues located at the amino terminus (residues Ser-32 and Ser-36 in the case of IB␣, 10 -11). The phosphorylated form of IB, although still bound to p65/p50 heterodimers (10,(12)(13), is then subject to rapid proteolysis via the ubiquitin-proteasome pathway (14). Removal of IB exposes a nuclear localization sequence present in NF-B, allowing the factor to translocate to the nucleus. In addition, several reports have shown that p50/p50 homodimers, which appear to bind to NF-B sites but not transactivate, are regulated by a mechanism not involving cytoplasmic/nuclear translocation but rather are resident in the nucleus and regulated by a nuclear protein Bcl-3, which when complexed prevents association of this form of NF-B with its recognition site (15). Recently, it was reported that other subunits of NF-B, including p65 and p50 are phosphoproteins (16). Furthermore, activation of NF-B in HeLa cells treated with H 2 O 2 , phorbol 12-myristate 13-acetate, or TNF␣ (17) or in endothelial cells treated with TNF␣ (18) is accompanied by an increased level of p65 phosphorylation. In this report, we show that treatment of fibroblasts or hepatoma cells with IL-1 also results in phosphorylation of p65 on serine residues, and we identify the kinase responsible as casein kinase II, raising the possibility that this is an additional level at which its activity is regulated.

EXPERIMENTAL PROCEDURES
HepG2 hepatoma cells (ATCC HB-8065) and human diploid gingival fibroblasts (19) were maintained as described previously (20). Carrierfree [ 32 P]orthophosphate was obtained from NEN Life Science Products while [␥-32 P]ATP (3000 Ci/mol) was from Amersham Corp. Purified polyclonal rabbit anti-peptide antibodies specific for NF-B p65, NF-B p50, and IB␣/MAD-3 together with corresponding immunizing peptides were obtained from Santa Cruz Biotechnology, Inc. Rabbit polyclonal antisera raised against synthetic peptides corresponding to * 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. amino acids 198 -215 of the ␤-subunit and amino acids 376 -391 of the ␣-subunit of human casein kinase II have been described previously (21) and were the kind gift of Minoo Ahdieh, Department of Biochemistry, Immunex Corporation. Recombinant human IL-1␣ was expressed and purified as described (22). Recombinant human NF-B p65 was a gift from Dr. Guido Franzoso (Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, Bethesda, MD). Casein kinase II, purified to greater than 95% from Pisaster ochraceus, was purchased from Upstate Biotechnology, Inc.
Cell Labeling and Immunoprecipitation-10-cm dishes of HepG2 cells (approximately 80% confluent) or post-confluent fibroblasts were pre-incubated for 30 min in serum-free RPMI 1640 medium lacking phosphate and containing 20 mM HEPES, pH 7.4. This medium was then replaced with 3 ml/dish of the same medium supplemented with 0.4 -1.0 mCi/ml [ 32 P]orthophosphoric acid and incubation continued for 4 h. IL-1 was added to a final concentration of 20 ng/ml at the end of the labeling period in a minimal volume (less than 0.5%) of phosphate-free medium. Addition of vehicle alone had no effect on any of the reported parameters. After stimulation, cell monolayers were rinsed three times with ice-cold phosphate-buffered saline and then lysed in 1 ml of a buffer containing 50 mM Tris-HCl, pH 8.0, 1% Nonidet P-40, 150 mM NaCl, 20 mM ␤-glycerophosphate, 10 mM NaF, 2 mM EGTA, 1 mM Na 3 VO 4 , 50 M leupeptin, 50 M pepstatin A, 1 mM PMSF (IP buffer). Lysates were passed 10 times through a 25 gauge hypodermic needle, incubated on ice for 10 min, then centrifuged at 13,000 ϫ g for 15 min. Clarified lysates were diluted 2-fold into IP buffer supplemented with 0.2% SDS and 1% sodium deoxycholate. 1-ml aliquots were pre-cleared for 30 min with 20 l of agarose-conjugated rabbit IgG (Sigma), and the supernatants were mixed overnight with 20 l of a suspension of anti-p65-agarose (2.5 g of antibody). Control incubations were carried out in the presence of a 40-fold molar excess of immunizing peptide. Immunoprecipitates were washed four times with IP buffer containing 0.1% SDS and 0.5% deoxycholate, twice with IP buffer containing 0.5 M NaCl, and once with 50 mM Tris-HCl, pH 7.0. They were then resuspended in Laemmli sample buffer (24), boiled for 5 min, and resolved by electrophoresis on 8 -16% SDS-polyacrylamide gels (SDS-PAGE). The gels were dried and quantitated by phosphorimaging. In some cases, labeling with [ 32 P]orthophosphate was omitted, and the washed NF-B immunoprecipitates were further processed for measurement of associated casein kinase activity as described below.
Casein Kinase Assays-To determine casein kinase activity associated with cytoplasmic and nuclear NF-B, cytoplasmic and nuclear extracts were prepared from IL-1-treated HepG2 cells as described for EMSA above except that buffers included phosphatase inhibitors (20 mM ␤-glycerophosphate, 10 mM NaF, 0.1 mM Na 3 VO 4 ). 1-ml aliquots of both types of extract were adjusted to contain 1% Nonidet P-40, 0.1% SDS, and 500 mM NaCl in a final volume of 1.4 ml and were then immunoprecipitated with anti-p65-agarose as described above. Before assay, NF-B immunoprecipitates were washed twice with 50 mM Tris-HCl, pH 7.5, 500 mM LiCl, 0.5 mM dithiothreitol, 20 mM ␤-glycerophosphate, 5 mM NaF, then twice with 50 mM Tris-HCl, pH 7.5, 20 mM ␤-glycerophosphate, 15 mM MgCl 2 , 5 mM NaF, 0.5 mM dithiothreitol. Purified casein kinase II (15 ng/reaction) was diluted in 15 l of 50 mM Tris-HCl, pH 7.5, 1 mM dithiothreitol, 2 mM EDTA, 10% glycerol. Phosphorylation reactions were initiated by adding 30 l of a kinase assay buffer containing 50 mM Tris-HCl, pH 7.5, 100 mM KCl, 15 mM MgCl 2 , 30 M ATP, 2.5-4.5 Ci [␥-32 P]ATP, 1 mM dithiothreitol, and substrate (dephosphorylated ␣and ␤-caseins (Sigma) at 0.5 mg/ml or recombinant NF-B p65 at approximately 5 g/ml). When present, heparin was included at 20 g/ml. Samples were incubated for 20 min at 30°C with frequent mixing, and the reactions were terminated by addition of 10 l of 4 ϫ Laemmli sample buffer. Incorporation of labeled ATP into substrate proteins was visualized and quantitated by SDS-PAGE and phosphoimaging. For determination of total cellular casein kinase II activity, cells were treated with IL-1 as necessary, washed in PBS, and extracted in 50 mM HEPES, pH 7.9, 500 mM NaCl, 1% Triton X-100, 0.1% SDS, 20 mM ␤-glycerophosphate, 10 mM NaF, 1 mM EGTA, 1 mM EDTA, 1 mM dithiothreitol, 1 mM PMSF, 50 M pepstatin A, 50 M leupeptin. After centrifugation (13,000 ϫ g, 15 min), lysates were precleared with agarose-conjugated rabbit-IgG and then mixed for 4 h at 4°C with a mixture of polyclonal antisera raised against human casein kinase II ␣and ␤-subunits (1:200 dilution). Immunoprecipitates were collected on protein A-agarose (Sigma), washed three times with extraction buffer, and then processed for casein kinase assay as described above. Alternatively, confluent cells in 175-cm 2 flasks (typically 4 -10 flasks/treatment group) were treated with or without IL-1 for 15 min, washed with cold PBS, and scraped into lysis buffer (20 mM Tris-HCl, pH 8.5, 20 mM ␤-glycerophosphate, 50 mM NaF, 0.1 mM Na 3 VO 4 , 2 mM dithiothreitol, 0.5 mM EDTA, 0.5 mM EGTA, 0.1% Nonidet P-40, 1 mM PMSF, 50 M pepstatin A, 50 M leupeptin). The cell suspensions were lysed by passage through a 25 gauge hypodermic needle, incubated at 0°C for 15 min, and clarified by centrifugation (25,000 ϫ g, 15 min). Lysates were immediately loaded onto a Mono-Q HR 5/5 anion exchange column (Pharmacia) equilibrated in lysis buffer lacking Nonidet P-40 and containing only 10 mM NaF. The column was washed with 10 volumes of equilibration buffer and eluted with a gradient of 0 -0.5 M NaCl in equilibration buffer at a flow rate of 1 ml min Ϫ1 . Thirty 1-ml fractions were collected, and 10-l aliquots of the fractions were assayed for casein kinase activity.
Phosphopeptide Mapping and Phosphoamino Acid Analysis-Samples of radiolabeled ␣and ␤-caseins or NF-B p65 were eluted from polyacrylamide gel slices, TCA-precipitated, and oxidized (25). Casein samples, in 100 l of 50 mM NH 4 HCO 3 , were proteolytically digested by addition of 2 g of sequencing grade chymotrypsin (Boehringer Mannheim). After 8 h, a second aliquot of chymotrypsin was added and digestion continued for 18 h. NF-B p65 was incubated, as above, with 1 g of chymotrypsin for 5 h, after which 1 g of sequencing grade modified trypsin (Promega) was added, and incubation was continued for 18 h. The resulting peptides were washed several times by lyophilization from distilled water and applied to thin-layer cellulose plates for electrophoresis at pH 1.9 (1.0 kV for 45 min) followed by ascending chromatography in n-butyl alcohol:pyridine:acetic acid:water (7.5:5:1.5:6) for 6 h. Radiolabeled phosphoamino acid compositions were determined by partial acid digestion of gel-purified proteins followed by two-dimensional electrophoresis as described (26).
Western Blot Analysis-Cellular extracts were separated on 10% SDS-polyacrylamide gels (Novex). Following SDS-PAGE, proteins were electrophoretically transferred to nitrocellulose. After overnight blocking in Tris-buffered saline containing 5% non-fat dried milk and 0.1% Tween-20, blots were incubated with anti-NF-B p65 (diluted to 2 g/ml in Tris-buffered saline containing 1% non-fat dried milk) for 1 h. Immune complexes were detected by sequential incubation of the blots with horseradish peroxidase-conjugated goat anti-rabbit IgG (Jackson Laboratories) and visualized using ECL reagent (Amersham Corp.) coupled with autoradiography on X-Ray films (XAR-5, Kodak).

RESULTS
The aim of this study was to determine if stimulation of HepG2 hepatoma cells or human gingival fibroblasts (HGF) with IL-1 leads to phosphorylation of other components of NF-B besides the IB inhibitor family. In both of these cell types, as shown by EMSA supershift analysis (Fig. 1), all of the IL-1-inducible nuclear NF-B complexes contain p65, and a majority contain p50. After treatment with IL-1 for various times, lysates from cells metabolically labeled with [ 32 P]orthophosphate were prepared and immunoprecipitated with anti-p65 antibody. To control for the presence of nonspecifically recognized proteins, we carried out duplicate immunoprecipitations, one of which contained an aliquot of the peptide against which the anti-p65 antibody was raised. As reported by others (16), NF-B/Rel family members and the IBs are physically associated phosphoproteins and are readily co-precipitated as shown in Fig. 2. From lysates of untreated HepG2 cells and (less prominently) from HGF, a phosphoprotein of 40 kDa was specifically immunoprecipitated. We performed Western blot analysis with a specific antibody (not shown), and we found that this phosphoprotein corresponded to IB␣. After 5 min of IL-1 treatment, the 40-kDa species disappeared and was replaced by a slightly slower migrating species (shown as IB(P) in Fig. 2) which also reacted with anti-IB␣. At subsequent times, IB was undetectable as either a labeled phosphoprotein (Fig. 2) or by Western blotting (not shown). These data are entirely consistent with previous reports (27)(28) in which activators of NF-B were shown to increase the phosphorylation of IB and thereby decrease its electrophoretic mobility. A protein that we tentatively identify as NF-B p50 was transiently phosphorylated in both cell types following IL-1 treatment. Interestingly, Li et al. (16) have reported that p50 phosphorylation was increased by treatment of Jurkat cells with phorbol 12-myristate 13-acetate and phytohemagglutinin and that phosphorylated p50 exhibited more stable DNA binding than the unphosphorylated form. Our most striking observation was an IL-1-mediated increase in the level of p65 phosphorylation. Although barely detectable in unstimulated HepG2 cells or fibroblasts, p65 was the most prominent phosphorylated species in immunoprecipitates made from cells treated with IL-1 for 5 min. Phosphorylation of P65 was maximal after 15 min and still evident for as long as 60 min after cytokine addition (not shown). IL-1-stimulated phosphorylation of an unidentified co-precipitated protein of about 130 kDa was consistently observed in HepG2 cells but was not investigated further. We isolated p65 from IL-1 stimulated HepG2 cells for phosphoamino acid analysis and found that phosphorylation occurred exclusively upon serine residues (Fig. 3).
IL-1 is known to activate a number of cytosolic serine/threonine kinases, including members of the mitogen-activated protein kinase family (20), stress-activated protein kinase/p38/ JNK family (21, 29 -31), and a novel ␤-casein kinase (32)(33). One of the preferred phosphorylation sites of the ␤-casein kinase purified from lungs of IL-1-injected rabbits is serine 57 of ␤-casein (34). 2 ) The primary sequence surrounding this residue bears some resemblance to a sequence in human p65 immedi-ately COOH-terminal to the Rel-homology domain, surrounding a serine at position 340. Specifically, both sequences have prolines at the Ϫ6, ϩ4, ϩ6, ϩ8, and ϩ10 positions relative to the serine residue and hydrophobic residues at the Ϫ8 and ϩ1 positions. p65 also contains two potential serine phosphorylation sites (35) for casein kinase II (minimal recognition sequence: Ser [Thr]-X-X-Glu [Asp]) located at amino acid positions 276 and 539. We reasoned that the IL-1-stimulated kinase present in NF-B immunoprecipitates might be one of these activities and that if it remained complexed with NF-B during immunoprecipitation, it should be possible to detect it based on its ability to phosphorylate exogenously added casein. Fig. 4a shows that extensively washed p65 immunoprecipitates from untreated HepG2 cells contain a kinase activity capable of phosphorylating both ␣and ␤-caseins and that the level of this activity is markedly stimulated if the cells are first treated with IL-1. The association of kinase activity with NF-B in the immunoprecipitates is specific since it is prevented by inclusion of excess immunizing peptide in the immunoprecipitation mixtures (Fig. 4a). Notably, kinase activity was completely abolished when the reactions were performed in the presence of 20 g/ml heparin, a potent and characteristic inhibitor of casein kinase II (36). None of the co-precipitated kinase activity could, therefore, be attributable to the cytokine-activated ␤-casein kinase, which is not inhibited by heparin at these concentrations and selectively phosphorylates ␤-casein (32). 3  jected to chymotryptic digestion and two-dimensional phosphopeptide mapping. These maps were compared with maps of casein phosphorylated by authentic casein kinase II (Fig. 4b). Although some of the digestion products (particularly in the case of ␣-casein) remained at the origin, the resolvable phosphopeptides were found to be essentially identical. Both kinases phosphorylated ␣-casein predominantly on serine and ␤-casein on threonine residues, with less phosphorylation on serine (data not shown).
Taken together, these data confirm the identity of the NF-B-associated kinase as casein kinase II. Because the anti-p65 antibody co-precipitates p50, and possibly other, associated proteins, it is not possible to determine from our data if p65 directly associates with the kinase. It is perhaps notable that the primary sequence of human p50 (37) also contains a potential casein kinase II phosphorylation site (SDLE at position 338 -341) that is conserved in both the rabbit and mouse p50 genes (38).
It was important to determine if the kinase associated with NF-B in cell extracts was the same as that responsible for its phosphorylation in intact cells. Accordingly, we compared the phosphopeptide patterns of p65 phosphorylated in IL-1-stimulated HepG2 cells with recombinant p65 phosphorylated in vitro by purified casein kinase II. Fig. 4c shows that casein kinase II efficiently phosphorylated recombinant p65 (samples of IL-1-induced rabbit ␤-casein kinase were also able to phosphorylate the recombinant protein, but at less than one-tenth the rate of casein kinase II, data not shown). To generate conveniently small fragments for mapping, phosphorylated p65 was digested sequentially with chymotrypsin and trypsin. In both samples, two major phosphopeptides (labeled x and y in Fig. 4d) exhibited no net charge, but migrated to a similar extent in the chromatographic dimension. A minor spot (z) was present in the in vivo labeled sample, but was barely visible in the in vitro phosphorylated material and may result from additional phosphorylation of p65 by a different kinase. In both samples, variable amounts of radioactive material remained at the origin, but the difficulty of preparing sufficient labeled p65 from cell cultures precluded further detailed analysis. It is not clear if casein kinase II phosphorylates p65 at either or both of the potential consensus phosphorylation sites described above, or indeed at a different site. Casein kinase II did not phosphorylate, in vitro, a recombinant p65 truncated to contain only the Rel homology domain, 4 and so it is likely that the phosphorylation site(s) lie in the COOH-terminal portion of the molecule.
It is now appreciated that casein kinase II activity is acutely regulatable by components of signal transduction pathways. Thus, modest but rapid stimulation of casein kinase II activity can occur in response to growth factors including EGF (39),

FIG. 3. Basal and IL-1-mediated phosphorylation of NF-B p65
occurs on serine residues. 32 P-labeled p65 from untreated HepG2 cells (top panels) or cells treated for 15 min with IL-1 (bottom panels) was carefully excised from SDS-PAGE gels, extracted, and hydrolyzed. Samples were subjected to two-dimensional phosphoamino acid analysis and visualized using a PhosphorImager (left panels). Internal phosphoamino acid standards were detected using ninhydrin and are shown in the panels on the right (PS, phosphoserine; PT, phosphothreonine; PY, phosphotyrosine).
FIG. 4. IL-1-treatment causes the association of casein kinase II with NF-B complexes. a, NF-B was immunoprecipitated from unstimulated or IL-1-treated (20 ng/ml, 15 min) HepG2 cells in the absence (Ϫ) or presence (ϩ) of immunizing peptide as indicated. Equal amounts of the washed immunoprecipitates were assayed for associated kinase activity using ␣and ␤-caseins as substrates. Reaction products were analyzed by SDS-PAGE. The assays were carried out in the absence (left panel) or presence (right panel) of 20 g/ml heparin. b, twodimensional chymotryptic phosphopeptide maps were prepared from caseins phosphorylated by the NF-B co-precipitated kinase (right panels) or casein kinase II (left panels). Sample application points are indicated (ori.); electrophoresis was in the horizontal dimension (anode at left) and chromatography in the vertical dimension. c, phosphorylation of recombinant NF-B p65 (150 ng) by purified casein kinase II (15 ng) in vitro. d, comparison of phosphopeptides generated by trypsin/chymotrypsin digestion of recombinant p65 phosphorylated in vitro by purified casein kinase II (left panel) and endogenous p65 immunoprecipitated from orthophosphate-labeled HepG2 cells stimulated for 15 min with IL-1 (right panel).
insulin, and insulin-like growth factor 1 (40). Although the underlying mechanism is not completely understood, it may involve the phosphorylation of either casein kinase II itself or accessory factors (39). We therefore wished to test whether increased phosphorylation of p65 was attributable to an IL-1stimulated increase in casein kinase II activity. After stimulation of HepG2 cells for various times with IL-1, whole-cell detergent lysates were prepared and immunoprecipitated with antibodies to casein kinase II. The washed immunoprecipitates were assayed for their ability to phosphorylate casein in an immune-complex kinase assay. The assay specifically measures casein kinase II since no activity could be detected in the presence of 20 g/ml heparin or if non-immune serum was used (data not shown). IL-1 caused a somewhat variable but small increase (to 150% of control) in immunoprecipitable casein kinase II activity, evident after 5 min of stimulation (Fig. 5a). TNF-␣ was recently reported to increase casein kinase II activity measured in crude cytosolic extracts of Swiss 3T3 and L929 cells (41), the kinetics and magnitude of the reported effects were similar to our findings with IL-1. We also directly measured casein kinase II activity in cytosolic extracts of HGF after separation on an anion exchange column (Fig. 5b). Casein kinase II is detected as a sharp peak of predominantly ␣-casein phosphorylating activity that elutes at about 380 mM NaCl (fractions [21][22][23][24] and that is inhibited by low concentrations of heparin or unlabeled GTP (data not shown). In contrast to the immunoprecipitation assay data described above, and in agreement with data reported by others (33), a 15-min pretreatment with IL-1 did not increase this peak of casein kinase II activity. Similar results were consistently obtained with HepG2 and a variety of other cell lines (data not shown). The peak of ␤-casein-specific, IL-1-stimulated kinase which elutes in fractions 7-12 ( Fig. 5b) corresponds to the activity described by Guesdon et al. (32)(33). Taken together, the above experiments suggest that the small IL-1-induced increase in total casein kinase II activity is restricted to that fraction of the enzyme (presumably associated with nuclei, membranes, or other organelles) which is only extractable by detergent; although, we cannot rule out the possibility that IL-1 induces an alteration in a subset of the kinase molecules such that they are more reactive with our antisera. In either case, it seems unlikely that such a small effect on casein kinase II activity is alone sufficient to account for the dramatic increase in p65 phosphorylation.
The subcellular location of casein kinase II has been reported to vary between the cytoplasm and nucleus at specific phases of the cell cycle (42) or to be constitutively nuclear (43). In HepG2 cells and fibroblasts, NF-B p65-containing complexes migrate into the nucleus upon IL-1 stimulation (Fig. 1). It is, therefore, possible that in resting cells, cytosolic NF-B would be in a different subcellular location from casein kinase II and that their association occurs simply as a consequence of nuclear translocation of the transcription factor. Accordingly, cytosolic and nuclear extracts were prepared from HepG2 cells after various periods of IL-1 stimulation and immunoprecipitated with anti-p65 for determination of associated casein kinase II activity (Fig. 6, a and c). A casein kinase II-associated pool of NF-B is established first in the cytoplasm and is detectable between 2 and 5 min after addition of IL-1. Accumulation of this form of NF-B in the nucleus is delayed until the 15 min time point is reached, and peak levels of casein kinase II activity associated with both nuclear and cytoplasmic NF-B are reached after 30 min of IL-1 treatment. The level of casein kinase II detectable in the nuclear fraction correlates well with the appearance of immunoreactive p65 in the nucleus, as shown by Western blotting (Fig. 6b). These data are consistent with a model in which casein kinase II associates with (and phosphorylates) cytoplasmic NF-B p65 which then translocates to the nucleus. An obvious mechanism, although difficult to directly test at the present time, would be one in which dissociation of IB from p50/p65 heterodimers exposes a phosphorylation site on one or both subunits for constitutively active, cytoplasmic casein kinase II. DISCUSSION The results of the experiments described above suggest that NF-B p65 is one of a number of transcriptional regulators that are physically associated with and phosphorylated by casein  5. IL-1 has a minimal effect on total cellular casein kinase II activity. a, detergent lysates were prepared from HepG2 after treatment for the indicated times with IL-1. Casein kinase II was immunoprecipitated with a mixture of polyclonal antisera that recognize the ␣and ␤-subunits and assayed using ␤-casein as a substrate. The inset shows the PhosphorImager output from a typical experiment. Mean results (Ϯ S.E.) from three independent experiments are plotted. b, after treatment of HGF cells with (bottom panel) or without (top panel) IL-1 for 15 min, cytosolic extracts were prepared and applied to a Mono-Q anion exchange column. The column was eluted with a linear NaCl gradient, and fractions were collected as described under "Experimental Procedures." Each of the fractions was assayed for ␣and ␤-casein kinase activities. The positions of IL-1stimulated ␤-casein kinase and casein kinase II are indicated. Similar results were obtained with HepG2 cells. kinase II. These include Sp1, Jun (44 -45), Fos, Myc/Max (46 -48), Myb (46 -48), and serum response factor (49). In some instances, phosphorylation by casein kinase II has been reported to have a negative effect upon binding of transcription factors to their cognate DNA recognition sequences; this is the case for Max homodimers and c-Jun. In the case of serum response factor, the transcription factor/DNA exchange rate is increased without a change in equilibrium binding affinity.
Does phosphorylation of p65 affect the DNA binding or transcriptional activity of NF-B? Definitive answers to this question must await identification of the phosphorylation site and its replacement with a non-phosphorylatable residue. Hayashi et al. (50) reported that the NF-B⅐IB complex isolated from human T cells was associated with a kinase activity that could directly phosphorylate p65 and p50 during an in vitro kinase assay and that phosphorylation correlated with increased DNA binding activity. These authors did not identify the kinase but showed that it was distinct from MAP kinase, Mos, PKC, and cAMP-dependent protein kinase. In light of our data, it is possible that the activity identified by this group was casein kinase II. Indeed, a 43-kDa moiety was affinity labeled using an ATP analog. This is close to the size expected for the catalytic subunits of casein kinase II. Naumann and Scheidereit (17) isolated NF-B complexes, containing phosphorylated p65, from TNF-stimulated HeLa cells and demonstrated that dephosphorylation of the complexes in vitro resulted in reduced DNA-binding capacity.
Another important component of the regulation of p65 is interaction with cytosolic IB, which both inhibits the DNA binding activity of the complexes and causes retention in the cytoplasm. It is therefore possible, since we have shown that phosphorylation of p65 by casein kinase II in intact cells takes place in the cytosol, that this phosphorylation might facilitate IB dissociation and/or translocation of NF-B into the nucleus. Alternatively, phosphorylation of p65 may inhibit the re-binding of IB␣ or other inhibitors. Finally, it is of interest that casein kinase II has recently been proposed by several groups as the kinase most likely responsible for constitutive phosphorylation of the COOH-terminal domain of IB␣ (51)(52)(53). This phosphorylation takes place on serines 283, 289, and 293 and on threonine 291. A requirement for these phosphorylations in both normal turnover and HIV-induced degradation of IB was suggested by the observation (53) that immunodepletion of casein kinase II abrogated both processes in in vitro IB degradation assays. Thus, it is possible that CKII may exert functionally opposite effects on NF-B and its inhibitor.
FIG. 6. Association of casein kinase II with NF-B occurs in the cytosolic compartment. a, nuclear and cytoplasmic extracts were made from HepG2 cells after treatment for increasing times with 20 ng/ml IL-1 as indicated. NF-B was immunoprecipitated using anti-p65 antibody. The washed immunoprecipitates were assayed for associated ␣and ␤-casein kinase activities in the presence (bottom panels) or absence (top panels) of heparin. Similar results were obtained in two independent experiments. b, aliquots of the nuclear and cytoplasmic extracts shown in panel a were separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotted with anti-NF-B p65 antibody. Immune complexes were visualized using an enhanced chemiluminescence (ECL, Amersham Corp.) procedure. c, quantitative analysis of the data for ␤-casein phosphorylation presented in panel a. Quantitation of ␣-casein phosphorylation gave essentially identical results but has been omitted for clarity.