Site-specific Tyrosine Phosphorylation of IκBα Negatively Regulates Its Inducible Phosphorylation and Degradation

The transcription factor NF-κB is retained in the cytoplasm by its interaction with the inhibitory subunit known as IκB. Signal-induced serine phosphorylation and subsequent ubiquitination of IκBα target it for degradation by the 26 S proteasome. Recently, pervanadate, a protein-tyrosine phosphatase inhibitor, was shown to block the degradation of IκBα, thus inhibiting NF-κB activation. We investigated the mechanism by which pervanadate inhibits the degradation of IκBα. Western blot analysis of IκBα from tumor necrosis factor-treated cells revealed a slower migrating IκBα species that was subsequently degraded. However, pervanadate-treated cells also revealed a slower migrating species of IκBα that appeared in a time- and dose-dependent manner and was not degraded by tumor necrosis factor. The slower migrating species of IκBα from pervanadate-treated cells was tyrosine-phosphorylated as revealed by cross-reactivity with anti-phosphotyrosine antibodies, by the ability of the specific tyrosine phosphatase PTP1B to dephosphorylate it, and by phosphoamino acid analysis of IκBα immunoprecipitated from 32P-labeled cells. By site-specific mutagenesis and deletion analysis, we identified Tyr-42 on IκBα as the phosphoacceptor site. Furthermore, in an in vitro reconstitution system, tyrosine-phosphorylated IκBα was protected from degradation. Our results demonstrate that inducible phosphorylation and degradation of IκBα are negatively regulated by phosphorylation at Tyr-42, thus preventing NF-κB activation.

The transcription factor NF-B is retained in the cytoplasm by its interaction with the inhibitory subunit known as IB. Signal-induced serine phosphorylation and subsequent ubiquitination of IB␣ target it for degradation by the 26 S proteasome. Recently, pervanadate, a protein-tyrosine phosphatase inhibitor, was shown to block the degradation of IB␣, thus inhibiting NF-B activation. We investigated the mechanism by which pervanadate inhibits the degradation of IB␣. Western blot analysis of IB␣ from tumor necrosis factor-treated cells revealed a slower migrating IB␣ species that was subsequently degraded. However, pervanadate-treated cells also revealed a slower migrating species of IB␣ that appeared in a time-and dose-dependent manner and was not degraded by tumor necrosis factor. The slower migrating species of IB␣ from pervanadatetreated cells was tyrosine-phosphorylated as revealed by cross-reactivity with anti-phosphotyrosine antibodies, by the ability of the specific tyrosine phosphatase PTP1B to dephosphorylate it, and by phosphoamino acid analysis of IB␣ immunoprecipitated from 32 P-labeled cells. By site-specific mutagenesis and deletion analysis, we identified Tyr-42 on IB␣ as the phosphoacceptor site. Furthermore, in an in vitro reconstitution system, tyrosine-phosphorylated IB␣ was protected from degradation. Our results demonstrate that inducible phosphorylation and degradation of IB␣ are negatively regulated by phosphorylation at Tyr-42, thus preventing NF-B activation.
The transcription factor NF-B regulates the expression of many genes that play essential roles in immune and inflammatory responses including the type I human immunodeficiency virus (1)(2)(3)(4). Like all members of the Rel/NF-B transcription factor family, NF-B has the unique property of being sequestered in its inactive state in the cytoplasm by a noncovalent association with inhibitory proteins called IB (4). In mammalian species, at least seven structural homologs of IB have been identified (4), but only the IB␣ form has been extensively studied. Recently, IB␣-deficient mice have been generated; they exhibit constitutive NF-B activation, severe runting, dermatitis, extensive granulopoiesis, and neonatal death (5,6). However, tumor necrosis factor (TNF), 1 which initiates the degradation of IB␣, causes sustained NF-B activation in embryonic fibroblasts from these knockout mice, suggesting that IB␣ is necessary for the postinduction repression of NF-B activity (5).
Indeed, IB␣ controls the activation of NF-B by masking the nuclear localization signal located on the p50-p65 heterodimer of NF-B (7). In response to a wide variety of stimuli besides TNF, IB␣ undergoes degradation, allowing the p50-p65 heterodimer to migrate to the nucleus (8,9). Since protein synthesis is not required for activation of this transcription factor, induction of target genes can occur within minutes of extracellular stimulus.
Site-specific mutagenesis and peptide mapping have revealed that inducible phosphorylation of IB␣ occurs at both serines 32 and 36 (10 -12). Although all known activators of NF-B induce phosphorylation of IB␣, leading to its degradation, how phosphorylation makes IB␣ susceptible to degradation is not understood (4). Recently, specific proteasome inhibitors have been shown to block the inducible degradation of IB␣, leading to the accumulation of a phosphorylated form of IB␣ while still bound to NF-B (12)(13)(14)(15)(16)(17)(18). These findings suggest that induced phosphorylation of IB␣ is needed but not sufficient for its degradation by the proteasome. Ubiquitination of lysines 21 and 22 (11,19) follows inducible serine phosphorylation, leading to degradation of IB␣ by the 26 S proteasome (8,9). The protein kinase involved in the inducible phosphorylation of IB␣ has been shown to be part of a 700-kDa complex (20), but the molecular identity of this kinase remains elusive.
Recently, reports from our laboratory (21) and another (22) demonstrated that TNF-mediated NF-B activation is completely abolished by the protein-tyrosine phosphatase inhibitor pervanadate (PV), which blocks the degradation of IB␣. In this report, we examined the mechanism by which PV blocks the degradation of IB␣. We demonstrate that inhibition of protein-tyrosine phosphatase activity by PV results in sitespecific tyrosine phosphorylation of IB␣, which prevents its inducible phosphorylation and degradation and hence inhibits the activation of NF-B.

EXPERIMENTAL PROCEDURES
Materials-The cell lines employed in this study included ML-1a, a human myelomonoblastic leukemia cell line kindly provided by Dr. Ken Takeda (Showa University, Showa, Japan). U937, a human histiocytic lymphoma cell line, and HeLa, an epithelial carcinoma cell line, were obtained from the American Type Culture Collection (Rockville, MD). ML-1a and U937 cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum and antibiotics. HeLa cells were cultured in minimum Eagle's medium supplemented with 10% fetal * This work was supported by the Clayton Foundation for Research. 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  bovine serum and antibiotics. Recombinant bacterium-derived purified human TNF (5 ϫ 10 7 units/mg) was a generous gift of Genentech, Inc. (South San Francisco, CA). Pervanadate was prepared fresh each time as described previously (21). Affinity-purified rabbit anti-IB␣ antiserum directed against the N and C termini and full-length IB␣ were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Horseradish peroxidase-conjugated goat anti-rabbit IgG was obtained from Bio-Rad. Calpain inhibitor I (N-acetylleucylleucylnorleucinal (ALLN)), biotinylated anti-phosphotyrosine antibodies, and horseradish peroxidase-conjugated anti-biotin were obtained from Sigma. Anti-FLAG antibody (monoclonal antibody M2) and anti-FLAG antibody-conjugated agarose were obtained from Eastman Kodak Co. Goat anti-mouse IgG conjugated to horseradish peroxidase was obtained from Transduction Laboratories (Lexington, KY). Okadaic acid was obtained from LC Laboratories (Woburn, MA). Calf intestine alkaline phosphatase was obtained from Life Technologies, Inc.
Plasmids Transfections-The FLAG-tagged IB␣ genes cloned into the eukaryotic expression vector pCMV4 were transiently transfected into 50% confluent HeLa cells (1 g of plasmid/100-mm dish) by Lipofectamine (Life Technologies, Inc.) as described by the manufacturer. After 6 h, the medium was replaced by complete medium and left overnight. The transfected cells were pooled from plates and plated in 60-mm dishes overnight at 37°C. Approximately 40 h after transfection, the cells were treated as indicated, and cytoplasmic extracts were prepared by the addition of lysis buffer containing 20 mM HEPES, pH 7.4, 250 mM NaCl, 0.1% Nonidet P-40, 10 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride. After 30 min on ice, the samples were cleared by centrifugation for 10 min. Protein was estimated by the method of Bradford (24). Approximately 40 g of lysate was subjected to 10% SDS-PAGE. Western blot analysis was performed using a monoclonal anti-FLAG antibody at 1 g/ml and horseradish peroxidaseconjugated goat anti-mouse IgG; the blots were visualized by enhanced chemiluminescence (ECL, Amersham Corp.).
Western Blotting of IB␣-ML-1a or U937 cells (2 ϫ 10 6 ) were treated as indicated, collected by centrifugation, and washed two times with cold phosphate-buffered saline. Unless otherwise stated, all treatments with pervanadate were for 30 min at 37°C using a concentration of 100 M. Lysates were prepared as described previously (21). Protein was estimated by the Bradford method (24). Lysates (20 -40 g) were separated by 10% SDS-PAGE and immunoblotted onto nitrocellulose (Bio-Rad) or polyvinylidene difluoride (Millipore Corp., Bedford, MA) membranes. Western blot analysis was performed using the indicated antibodies, and blots were visualized by ECL.
Immunoprecipitation of IB␣ from Orthophosphate-labeled Cells-U937 cells were labeled with [ 32 P]orthophosphate (DuPont NEN) and immunoprecipitated as described earlier (22) except that a mixture of antibodies generated against the N terminus, C terminus, and fulllength IB␣ was used. In vitro transcribed and translated 35 S-labeled IB␣ (see below) was used as a control for the migration of IB␣ on SDS-PAGE. For phosphoamino acid analysis, immunoprecipitates were resolved by 10% SDS-PAGE (a large gel apparatus, 16 ϫ 20 cm) and electroblotted onto polyvinylidene difluoride membranes. The membrane containing phosphorylated IB␣ was identified by autoradiography, excised, and subjected to acid hydrolysis followed by two-dimensional electrophoretic separation on thin-layer chromatography plates as described (25).
In Vitro Reconstitution Assay for IB␣ Degradation-We used an in vitro reconstitution system similar to that described by Chen et al. (8). U937 cells were treated as described in the figure legends, and cytoplasmic extracts were prepared as described for Western blotting of IB␣ except that okadaic acid was added to a final concentration of 3 M and the supernatant was stored at Ϫ80°C. The typical protein concentration of these extracts was 10 -20 g/l. In vitro transcribed and translated 35 S-labeled, FLAG-tagged IB␣ and mutants were generated using the TNT ® coupled reticulocyte lysate system (Promega, Madison, WI) using pBS-IB␣ or pBS-FIB␣ plasmids as the template in a 50-l reaction mixture containing [ 35 S]methionine (Amersham Corp). The reconstitution system contained 1-5 l of 35 S-labeled, FLAG-tagged IB␣ that was incubated with cell extracts (200 -400 g) in the presence of an ATP-regenerating system (50 mM Tris, pH 7.6, 5 mM MgCl 2 , 2 mM ATP, 10 nM creatine phosphate, 3.5 units/ml creatine kinase, and 3 M okadaic acid). In one experiment, we omitted okadaic acid from the lysates and the reaction mixture, but included the kinase inhibitor isopentyladenine (26) at a concentration of 500 M to inhibit the serine phosphorylation of IB␣. 2 The reaction mixtures were incubated at 37°C for the indicated times, and the reactions were stopped by the addition of lysis buffer (20 mM HEPES, pH 7.4, 250 mM NaCl, 0.1% Nonidet P-40, 10 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride) and immunoprecipitated with anti-FLAG antibody-conjugated agarose. The immunoprecipitates were washed twice with lysis buffer, and the protein was eluted by the addition of SDS sample buffer and boiling for 5 min. The samples were subjected to 10% SDS-PAGE, and the dried gels were visualized by a PhosphorImager (Molecular Dynamics, Inc., Sunnyvale, CA) or by autoradiography and quantitated by ImageQuant software (Molecular Dynamics, Inc.).
Purification of PTP1B-The plasmid pGEX-PTP1B-His was used to transform Escherichia coli strain BL21. An overnight culture containing the transformed cells was used to inoculate 250 ml of Luria broth, grown to an A 600 of 0.8, and induced with 0.5 mM isopropylthiogalactoside for 3 h at 37°C. The cells were collected and ruptured in lysis buffer (20 mM Tris, pH 8.0, 0.5% Nonidet P-40, 250 mM NaCl, 10% glycerol, 10 g/ml leupeptin, 10 g/ml aprotinin, and 1 mM phenylmethylsulfonyl fluoride) containing 500 g of lysozyme for 30 min on ice. The lysate was sonicated and clarified by centrifugation at 100,000 ϫ g for 30 min at 4°C. The supernatant was bound to 0.5 ml of Ni 2ϩ -agarose by batch chromatography in the presence of 25 mM imidazole for 1 h at 4°C. The Ni 2ϩ -agarose was washed with 12 ml of lysis buffer containing 1 M NaCl and 25 mM imidazole and then with 2 ml of buffer containing 20 mM HEPES, pH 7.4, and 50 mM NaCl. The protein was eluted with 2 ml of elution buffer (20 mM HEPES, pH 7.4, 50 mM NaCl, 10% glycerol, and 1 M imidazole). The eluted protein (GST-PTP1B-His) was dialyzed overnight at 4°C against storage buffer (50 mM HEPES, pH 7.9, 0.1% 2-mercaptoethanol, and 50% glycerol) and stored at Ϫ20°C.
Protein-tyrosine Phosphatase Assay-U937 cells were treated and lysates prepared as described (21). Lysates (20 g) were mixed with 1 g of purified GST-PTP1B-His or storage buffer and assayed in a 30-l volume of protein-tyrosine phosphatase buffer containing 50 mM Tris, pH 7.4, 50 mM NaCl, and 1 mM dithiothreitol in the presence or absence of 2 mM orthovanadate for 30 min at 37°C. The reactions were stopped with SDS sample buffer, and the samples were boiled for 5 min and subjected to 10% SDS-PAGE and immunoblotting with anti-IB␣ antibodies as described above.

RESULTS AND DISCUSSION
Previously, we (21) and others (22) reported that inhibitors of protein-tyrosine phosphatases block the activation of NF-B by a variety of inducers. In this report, we investigated the mechanism by which PV blocks this activation. We used human ML-1a, U937, and HeLa cells in this study since their response to various inducers of NF-B is well characterized (8,21,22,27). We did not find any differences among these cell lines with regard to the inducible phosphorylation and degradation of IB␣.
PV Treatment Induces Phosphorylated IB␣ That Is Not Degraded-To investigate the mechanism by which PV blocks the degradation of IB␣, we analyzed IB␣ from ML-1a cells by Western blotting with IB␣-specific antibodies. Inducible phosphorylation of serines 32 and 36 of IB␣ appears as a slow migrating band as analyzed by SDS-PAGE and immunoblotting of cytoplasmic extracts (8,12,23). As expected, Western blot analysis showed that TNF caused a slower migrating species of IB␣ to appear within 2 min, to be degraded in 5 min, and to reappear after 30 min (Fig. 1A, upper panel). However, when cells were pretreated with PV, the slower migrating species of IB␣ appeared (Fig. 1A, lower panel), and this form of IB␣ was not degraded upon further stimulation with TNF (lower panel). A dose-response analysis indicated that a PV concentration of 25 M caused the slower migrating band to appear, and 100 M PV converted all IB␣ to the slower migrating species (Fig. 1B, upper panel). A time course analysis with 100 M PV indicated that a 5-min treatment was sufficient to convert all IB␣ to the slower migrating species (Fig. 1B,  lower panel). The retardation of the migration of IB␣ from PV-treated cells was attributed to phosphorylation since calf intestine alkaline phosphatase increased its mobility on SDS-PAGE (Fig. 1C). These observations indicated that PV treatment caused the phosphorylation of IB␣, which thus protected it from TNF-induced degradation.
PV Causes Tyrosine Phosphorylation of IB␣-Since PV inhibits protein-tyrosine phosphatases and increases total protein tyrosine phosphorylation in cells (Ref. 22 and data not shown), it may directly induce the tyrosine phosphorylation of IB␣. To explore this possibility, immunoprecipitated IB␣ from control and PV-treated cells was analyzed by Western blotting using anti-phosphotyrosine antibodies. IB␣ immuno-precipitated from PV-treated cells, but not from control cells, was indeed tyrosine-phosphorylated ( Fig. 2A). The tyrosinephosphorylated band was confirmed to be the slower migrating species of IB␣ by Western blot analysis using anti-IB␣ antibodies (data not shown). Additionally, immunoprecipitation of IB␣ from 32 P-labeled cells revealed a slower migrating species of IB␣ only from cells treated with PV (Fig. 2B). Furthermore, phosphoamino acid analysis revealed that IB␣ was constitutively phosphorylated at serine residues (Fig. 2C, left panel),

FIG. 2. PV treatment causes the tyrosine phosphorylation of IB␣.
A, anti-phosphotyrosine Western blot analysis of IB␣ immunoprecipitated from PV-treated cells. IB␣ was immunoprecipitated from untreated (Control) or PV-treated U937 cells, resolved by 10% SDS-PAGE, and immunoblotted with biotinylated anti-phosphotyrosine antibodies as described under "Experimental Procedures." B, the slower migrating species of IB␣ from PV-treated cells is phosphorylated in vivo. Untreated (Control) or PV-treated U937 cells were labeled with [ 32 P]orthophosphate, and lysates were prepared and immunoprecipitated with either IB␣ antibodies or preimmune serum (PIS), resolved by 10% SDS-PAGE, transferred to polyvinylidene difluoride membranes, and autoradiographed. In vitro transcribed and translated 35 Slabeled IB␣ (Invt.) is shown as a control for the migration of IB␣. C, phosphoamino acid analysis reveals that IB␣ from PV-treated cells is tyrosine-phosphorylated in vivo. After autoradiography of the membrane shown in B, the IB␣ band from the control (n) and the slower migrating band from PV-treated cells (s) were excised from the membrane and subjected to phosphoamino acid analysis and two-dimensional thin-layer electrophoresis as described under "Experimental Procedures." The TLC plate was analyzed by a PhosphorImager, and the migration of phosphoamino acid standards was visualized by ninhydrin as indicated. most likely by casein kinase II (28 -30); but in PV-treated cells, the slower migrating species was additionally phosphorylated at a tyrosine residue (Fig. 2C, right panel). Thus, the pervanadate-induced decrease in the mobility of IB␣ was due to tyrosine (not serine) phosphorylation.
Tyrosine Phosphorylation of IB␣ Prevents TNF-induced Serine Phosphorylation-To further examine the effects of tyrosine phosphorylation on TNF-induced serine phosphorylation, we performed a Western blot analysis. However, like TNF-induced phosphorylation, PV-induced phosphorylation reduces the mobility of IB␣, thus making it difficult to distinguish the serine-and tyrosine-phosphorylated forms by Western blotting. To differentiate the phosphorylation of serine from that of tyrosine, we used a specific tyrosine phosphatase, PTP1B (32), which should reverse only the PV-induced decrease in the mobility of IB␣. The TNF-induced serine phosphorylation of IB␣ was stabilized by pretreatment of the cells with ALLN, a specific protease inhibitor (9,10,14,31), for 1 h prior to TNF stimulation. Additionally, to maintain serine phosphatases in their active state, we omitted okadaic acid in the preparation of these lysates. PTP1B treatment had no effect on IB␣ from untreated or (ALLN ϩ TNF)-treated cells (Fig. 3, lanes 1-6), but reversed the mobility of IB␣ from PVor (PV ϩ TNF)-treated cells (lanes 8 and 11, respectively), suggesting that tyrosine-phosphorylated IB␣ is protected from the TNF-induced serine phosphorylation. We used orthovanadate as a control, and it inhibited the effect of PTP1B (Fig. 3, lanes 9 and 12, respectively).
Besides inducing the tyrosine phosphorylation of IB␣, pervanadate could also either inactivate the IB␣ serine kinase or block an upstream activator of the IB␣ serine kinase. This would require the localization of the tyrosine residue on IB␣ that undergoes phosphorylation in response to PV. The mutation of this tyrosine residue, then, should overcome the inhibition of PV treatment without interfering with the serine kinase.
Tryosine 42 on IB␣ Is Phosphorylated in PV-treated Cells-IB␣ contains eight tyrosine residues, including Tyr-42 at the N terminus; Tyr-181 and Tyr-195 in the fourth ankyrin repeat; and Tyr-248, Tyr-251, Tyr-254, Tyr-289, and Tyr-305 toward the C terminus (Fig. 4A) (33). All except Tyr-289 are conserved among IB␣ species from human, chicken, rat, mouse, and pig (34). To identify the tyrosine residue phosphorylated by PV treatment, we used FLAG-tagged wild-type IB␣ and various mutants in an in vitro reconstitution system as established by Chen et al. (8); we observed no interference of epitope-tagged IB␣ in this assay. Preliminary experiments indicated that incubation of 35 S-labeled IB␣ with lysates from PV-treated cells resulted in its tyrosine phosphorylation as revealed by anti-phosphotyrosine immunoblot analysis (data not shown). Both the wild type (WT) and a mutant lacking residues 243-317 (⌬C) showed a decrease in mobility when treated with PV extracts in the reconstitution system (Fig. 4B); the decrease was reversed by treatment with PTP1B, indicating tyrosine phosphorylation (Fig. 4B). This suggested that the tyrosine phosphorylation site resided in the N terminus, thus eliminating five of the eight tyrosine residues as candidates. Of the three remaining tyrosine residues, only Tyr-42 is a potential phosphorylation site (33). When a Y42F mutant was used, it did not shift in PV extracts in the reconstitution system, thus indicating that it was the potential site of tyrosine phosphorylation (Fig. 4B). Furthermore, we transiently transfected The amino-terminal region (residues 1-72) and ankyrin repeats (residues 73-242; shaded) are indicated. The C-terminal region contains an acidic region, PEST region, and phosphorylation sites for casein kinase II (CK II). All eight tyrosine residues are numbered. The two inducible phosphorylation (serines 32 and 36) and ubiquitination (lysines 21 and 22) sites are shown in an expanded region containing residues 18 -47. A mutant of IB␣ lacking residues 243-317 (⌬C) and a mutant in which Tyr-42 was changed to Phe (Y42F) were used in this study as described under "Experimental Procedures." B, in vitro phosphorylation of Tyr-42 on IB␣ from PV-treated cells. 35 S-Labeled, FLAG-tagged wild-type IB␣ (WT) or mutants (Y42F and ⌬C) were incubated with cell lysates prepared from PV-treated U937 cells in the presence of an ATP-regenerating system at 37°C for 45 min and immunoprecipitated with anti-FLAG antibody-conjugated agarose as described under "Experimental Procedures." Where indicated, the immunoprecipitates were mixed with PTP1B, and a phosphatase assay was performed as described under "Experimental Procedures." Proteins were eluted by the addition of SDS sample buffer, boiled, resolved by 10% SDS-PAGE, and analyzed by a PhosphorImager. C, cell transfection of FLAG-tagged IB␣ implicates Tyr-42 as the phosphorylation site induced by PV in vivo. HeLa cells were transfected with epitope-tagged wild-type IB␣ (WT) or the Y42F mutant as described under "Experimental Procedures." Cells were left untreated or were treated with PV, and cytoplasmic lysates were prepared. Fifty micrograms of lysates was subjected to 8.5% SDS-PAGE (large gel apparatus), immunoblotted onto polyvinylidene difluoride membranes, and probed with anti-FLAG antibodies as described under "Experimental Procedures." s and n represent the slower and normal migrating species of IB␣, respectively. epitope-tagged wild-type and Y42F IB␣ genes into HeLa cells and analyzed the expression of the IB␣ proteins by Western blotting using anti-FLAG antibodies: the FLAG-tagged Y42F IB␣ mutant migrated faster than the wild type on SDS-polyacrylamide gels. Upon PV treatment, FLAG-tagged wild-type IB␣, but not the Y42F mutant, decreased in mobility (Fig. 4C), providing evidence that this site was tyrosine-phosphorylated in vivo. Because of overexpression, however, only a small portion of epitope-tagged IB␣ was shifted in PV-treated cells (Fig.  4C). Interestingly, Tyr-42 is close to the inducible serine phosphorylation sites (serines 32 and 36) and ubiquitination sites (lysines 21 and 22) needed for the degradation of IB␣. Thus, it is possible that the phosphorylation of Tyr-42 stereochemically hindered subsequent phosphorylation by the inducible serine kinase.
Tyrosine-phosphorylated IB␣ Is Protected from Inducible Degradation-Because the IB␣-overexpressing HeLa cells lacked TNF responsiveness, our attempts to utilize these cells expressing epitope-tagged wild-type IB␣ and the Y42F mutant were unsuccessful. Therefore, we used an in vitro reconstitution assay that contained cellular extracts, an ATP-regenerating system, and okadaic acid with wild-type IB␣ or the Y42F IB␣ mutant. 35 S-Labeled, FLAG-tagged wild-type IB␣ was phosphorylated and degraded in a time-dependent manner using this assay (Fig. 5A). In vitro translated 35 S-labeled Y42F IB␣ was mixed with either untreated or PV-treated cell lysates for 60 min as described for Fig. 5A, subjected to SDS-PAGE, and analyzed by a PhosphorImager. The Y42F mutant was degraded similarly in extracts prepared from either untreated or PV-treated cells (Fig. 5B), suggesting a lack of effect of PV on the serine kinase.
To further confirm that tyrosine-phosphorylated IB␣ is protected from degradation, we prepared a tyrosine-phosphorylated form of IB␣ by first incubating 35 S-labeled, epitopetagged IB␣ in extracts from PV-treated cells (similar to Fig.  4B) and then immunoprecipitating it with anti-FLAG antibody-conjugated agarose; we used the immunoprecipitate as the substrate in the in vitro reconstitution system with lysates prepared from untreated (control) or PV-treated cells. Tyrosine-phosphorylated, 35 S-labeled IB␣ was protected from degradation in lysates prepared from PV-treated cells (Fig. 5C). However, Ͼ50% of the 35 S-labeled IB␣ was degraded in control lysates (Fig. 5C), the latter perhaps because tyrosinephosphorylated IB␣ first underwent dephosphorylation at tyrosine, leading to subsequent serine phosphorylation and degradation. These results are consistent with the thesis that site-specific tyrosine phosphorylation of residue 42 on IB␣ protects it from inducible serine phosphorylation and thus inhibits its subsequent degradation. Since phosphorylation of serine residue 32 or 36 is a prerequisite for the subsequent ubiquitination of lysines 21 and 22 (8,11,19), which targets IB␣ for degradation, the mechanism of inhibition by pervanadate appears to be upstream of the ubiquitin machinery. This may explain why ubiquitin-conjugated forms of IB␣ were not observed in cell lysates prepared from (PV ϩ TNF)-treated U937 cells (data not shown).
We have shown here that treatment of cells with PV results in site-specific tyrosine phosphorylation of IB␣ and that this phosphorylation protects it from degradation and hence inhibits the activation of NF-B. While PV blocks the degradation of IB␣ induced by a wide variety of inducers (21,22), the data suggest a common inhibitory role for this tyrosine phosphorylation with all these inducers. It is plausible that a futile cycle occurs in which a protein-tyrosine kinase and phosphatase catalyze the phosphorylation and dephosphorylation of IB␣ in vivo. That would explain why we and others (22) were not able to detect tyrosine-phosphorylated IB␣ in unstimulated cells. It is possible that the amount of tyrosine-phosphorylated IB␣ in unstimulated cells is below the limits of detection due to an active protein-tyrosine phosphatase, but we were able to detect tyrosine-phosphorylated IB␣ in vivo when protein-tyrosine phosphatase activity was abolished.
Although Tyr-42 is conserved among all IB␣ gene products known thus far, IB␤ does not contain an analogous tyrosine positioned near its inducible phosphorylation sites (6). Additionally, IB␤ is not tyrosine-phosphorylated upon treatment with pervanadate. 2 Like IB␣, however, IB␤ also appears to be regulated by phosphorylation and degradation (5,6,11,35). While TNF is able to induce the degradation of IB␣ and IB␤ in HeLa cells (11), 2 it is not able to cause the degradation of IB␤ in U937 cells. 2 Thus, it appears that the ability of inducers to cause the degradation of IB␤ could be cell type-specific (5, 6). The tyrosine phosphorylation of IB␣ may suggest an additional mechanism of regulation, one that affects IB␣, but FIG. 5. In a reconstitution system, tyrosine-phosphorylated IB␣ is protected from degradation. A, FLAG-tagged wild-type IB␣ is phosphorylated and degraded in a time-dependent manner. 35 S-Labeled, FLAG-tagged IB␣ was incubated with U937 cell extracts in the presence of an ATP-regenerating system for various times at 37°C and immunoprecipitated with anti-FLAG antibody-conjugated agarose as described under "Experimental Procedures." Proteins were eluted by the addition of SDS sample buffer and subjected to 10% SDS-PAGE, and the dried gel was visualized by a PhosphorImager and quantitated by ImageQuant software. The remaining undegraded IB␣ is shown as percent of control. B, the inducible serine kinase is active in PV-treated cells. The 35 S-labeled, FLAG-tagged Y42F IB␣ mutant was incubated with extracts from untreated (Control) or PV-treated U937 cells in the reconstitution system for 60 min as described for A. C, tyrosine-phosphorylated IB␣ is protected from degradation in the reconstitution system. To obtain a tyrosine-phosphorylated form of IB␣, 35 S-labeled, FLAG-tagged wild-type IB␣ was first incubated with extracts from PV-treated U937 cells as described for Fig. 4B and immunoprecipitated with anti-FLAG antibody-conjugated agarose. The immunoprecipitated material was then split into tubes containing extracts from untreated (control (C)) or PV-treated U937 cells and processed as described for B. s and n represent the slower and normal migrating species of IB␣, respectively. not IB␤. The block of IB␣ and IB␤ degradation by specific proteasome inhibitors is a viable strategy for targeting the unwanted activation of NF-B in inflammatory and immune responses. The molecular identity of the putative protein-tyrosine kinase and phosphatase suggested here may allow for the design of pharmacological inhibitors of NF-B activation.
Note Added in Proof-While our paper was in press, a report appeared by Imbert et al. (36) confirming our results that treatment of Jurkat cells with pervanadate causes tyrosine 42 phosphorylation of IB␣. They did not examine the effect of this phosphorylation on TNFdependent NF-B activation. Unlike our results, however, Imbert et al. (36) showed that tyrosine 42 phosphorylation of IB␣ accompanies activation of NF-B by pervanadate.