Tpl2 (Tumor Progression Locus 2) Phosphorylation at Thr 290 Is Induced by Lipopolysaccharide via an I (cid:1) -B Kinase- (cid:2) -dependent Pathway and Is Required for Tpl2 Activation by External Signals*

The serine-threonine protein kinase encoded by the tumor progression locus 2 ( Tpl2 ) proto-oncogene transduces Toll-like receptor and death receptor signals in a variety of cell types. Here we show that Tpl2 undergoes phosphorylation at Thr 290 both in cells overexpressing Tpl2 and in cells stimulated with lipopolysaccharide (LPS) or tumor necrosis factor- (cid:3) and that phosphorylation on this site parallels Tpl2 activation. Reconstitution of Tpl2 (cid:4) / (cid:4) macrophages with wild type Tpl2 or Tpl2 T290D restored ERK activation by LPS, whereas reconstitution of the same cells with Tpl2 T290A did not, suggesting that phosphorylation at Thr 290 is required for the physiological activation of Tpl2 by external signals. Both the wild type Tpl2 and the kinase-inactive mutant Tpl2 K167M undergo Thr 290 phosphorylation, suggesting that Thr 290 may be a site of trans-phosphorylation rather than auto-phosphorylation. Pretreatment of 293 cells and primary macrophages with the I (cid:1) -B kinase- (cid:2) (IKK and Tpl2 mutants at potential activation loop phosphorylation sites were probed anti-phospho-ERK, anti-total mutant failed induce phosphorylation of T290D mutant induced phosphorylation, phosphorylation required Tpl2 posttrans- lational modifications responsible for the Tpl2 Tpl2.HA

The serine-threonine protein kinase encoded by the tumor progression locus 2 (Tpl2) proto-oncogene transduces Toll-like receptor and death receptor signals in a variety of cell types. Here we show that Tpl2 undergoes phosphorylation at Thr 290 both in cells overexpressing Tpl2 and in cells stimulated with lipopolysaccharide (LPS) or tumor necrosis factor-␣ and that phosphorylation on this site parallels Tpl2 activation. Reconstitution of Tpl2 ؊/؊ macrophages with wild type Tpl2 or Tpl2 T290D restored ERK activation by LPS, whereas reconstitution of the same cells with Tpl2 T290A did not, suggesting that phosphorylation at Thr 290 is required for the physiological activation of Tpl2 by external signals. Both the wild type Tpl2 and the kinase-inactive mutant Tpl2 K167M undergo Thr 290 phosphorylation, suggesting that Thr 290 may be a site of trans-phosphorylation rather than auto-phosphorylation. Pretreatment of 293 cells and primary macrophages with the I-B kinase-␤ (IKK␤) inhibitor PS-1145 blocked Tpl2 phosphorylation at Thr 290 , suggesting that phosphorylation depends on IKK␤, an obligatory positive regulator of Tpl2. We conclude that Tpl2 phosphorylation at Thr 290 is induced by LPS, depends on IKK␤, and is required for the physiological activation of Tpl2 by external signals.
Tumor progression locus 2 (Tpl2) 1 encodes a serine-threonine protein kinase that is activated by provirus integration in Moloney murine leukemia virus-induced rodent lymphomas and in mouse mammary tumor virus-induced mammary adenocarcinomas (1,2). Provirus integration always occurs in the last intron of the gene and gives rise to a truncated mRNA that is translated into a protein in which 43 C-terminal amino acids encoded by the last exon are replaced by seven amino acids encoded by intron sequences (1). The latter protein exhibits enhanced kinase activity and is highly oncogenic (3).
Overexpression of Tpl2 in a variety of cell types activates all of the mitogen-activated protein kinase pathways (4 -6), nuclear factor-activated T cells, and NF-B (7-10) and promotes cell proliferation (1,11). Moreover, the expression of the tumorspecific truncated form of Tpl2 transforms cells in culture (6) and is highly oncogenic in animals (3). Despite the profound effects of Tpl2 overexpression, Tpl2 knock-out mice appear healthy. However, these mice are resistant to LPS/D-galactosamine-induced shock (12), a rapidly developing syndrome that depends on TNF-␣, and to high-dose LPS-induced shock, 2 a slowly developing inflammatory syndrome that is mediated by TNF-␣, IL-1␤, interferon-␥, and other proinflammatory molecules (13,14). The same mice are resistant to TNF-␣-induced inflammatory bowel disease (15). These finding suggested that Tpl2 plays an obligatory role in the transduction of LPS and TNF-␣ signals, a hypothesis that has now been amply confirmed. Studies from this laboratory have indeed shown that Tpl2 is required for the transduction of LPS and CD40L signals that activate ERK in macrophages and B cells and TNF-␣ signals that also activate ERK in macrophages (12,16,17). Our more recent studies have also shown that, in mouse embryo fibroblast, Tpl2 is required for the transduction of TNF-␣ signals that activate in addition to ERK, c-Jun NH 2 -terminal kinase, and NF-B. 3 Studies based on these observations confirmed that Tpl2 is activated by both LPS (Refs. 18,19, and this report) and TNF-␣ (20). 3 Tpl2 activation by LPS or anti-CD40 antibodies depends on TRAF6. Thus, anti-CD40 antibodies promote the association of endogenous Tpl2 and TRAF6 with CD40. Moreover, overexpression of TRAF6 in Tpl2 ϩ/ϩ fibroblasts and keratinocytes promotes the activation of ERK, whereas overexpression of TRAF6 in Tpl2 Ϫ/Ϫ cells does not (17). More recent studies addressed the role of RIP1 and TRAF2 on Tpl2 activation by TNF-␣ and the role of IKK␤ on Tpl2 activation by either TNF-␣ or LPS. These studies revealed that the activation of Tpl2 by TNF-␣ depends on RIP1 and TRAF2, because treatment of RIP1 Ϫ/Ϫ and TRAF2 Ϫ/Ϫ mouse embryo fibroblasts with TNF-␣ does not activate Tpl2. 3 In addition, they showed that Tpl2 activation by LPS or TNF-␣ depends on the activation of IKK␤ (19,20). Because inactive Tpl2 in unstimulated cells is stoichiometrically bound to NF-B1/p105 and Tpl2 activation by external signals depends on its dissociation from p105 (9,18), these findings were interpreted to suggest that the function of IKK␤ is to promote the phosphorylation and degradation of p105 and the subsequent release of activated Tpl2 (19,20).
In this report, we present evidence that Tpl2 undergoes phosphorylation at Thr 290 in vivo and that phosphorylation is required for the enzymatic activation of the Tpl2 kinase. We also show that phosphorylation is induced by LPS and TNF-␣ and that it is required for the physiological activation of Tpl2 by external signals. Finally, we present evidence that Thr 290 is a site of trans-phosphorylation rather than auto-phosphorylation and that its modification depends on IKK␤, an obligatory positive upstream regulator of Tpl2 (19,20). Of the 52-(p52) and the 58-kDa (p58) isoforms of Tpl2, the p58 isoform is targeted preferentially for phosphorylation at Thr 290 , suggesting that the two isoforms are functionally distinct. These data collectively show that phosphorylation at Thr 290 is required for Tpl2 activation by external signals. In addition, they show that IKK␤ is required for the activation of Tpl2, not only because it phosphorylates p105 and targets it for degradation but also because it promotes the phosphorylation of Tpl2 at Thr 290 . Therefore, these data provide a novel link between IKK␤ and Tpl2 activation by external signals.

MATERIALS AND METHODS
Cell Culture-Bone marrow-derived macrophages (BMDM) from Tpl2 ϩ/ϩ and Tpl2 Ϫ/Ϫ mice (12) were infected with a pZIP-Neo/SV40 large T antigen retrovirus harvested from virus-producing Y2 cells (kindly provided by Anthony DeFranco, University of California San Francisco). To infect primary BMDMs, 1 ϫ 10 5 cells were resuspended in 2 ml of virus supernatants. 2 h later, they were cultured in macrophage media (Dulbecco's modified Eagle's medium supplemented with 20% fetal bovine serum, 30% L929-conditioned media, non-essential amino acids, penicillin, and streptomycin). Because Tpl2 Ϫ/Ϫ cells express constitutively the G418 resistance gene, they could not be G418-selected. Therefore, to isolate cultures of large T antigen-expressing cells, we selected for cells surviving long term culture. Two months were generally sufficient to isolate fully infected immortalized lines. The immortalized macrophage cell lines were maintained in macrophage media. RAW264.7 cells, Y2 cells, 293 cells, L929 cells (12), and Phoenix cells (kindly provided by Garry Nolan, Stanford University) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, nonessential amino acids, penicillin, and streptomycin.
To inhibit IKK␤, 293 cells transiently transfected with Tpl2 constructs were treated with PS-1145 (Millennium Pharmaceuticals Inc.) 1.5 h prior to harvesting and RAW264.7 cells were treated with PS-1145 (50 M) 1.5 h prior to LPS (1 g/ml from Salmonella enteritidis, purchased from Sigma) stimulation. Alternatively, IKK␤ was knocked down by siRNA (see below).
Expression Plasmids, Site-directed Mutagenesis, and Retrovirus Constructs-Wild type Tpl2, C-terminally tagged with the HA epitope, was cloned by PCR in the polylinker of pCMV5 between EcoRI and BamHI. Tpl2 mutants were generated by site-directed mutagenesis of this plasmid using the QuikChange site-directed mutagenesis kit (Stratagene). All of the mutations were confirmed by sequencing. Wild type Tpl2 and Thr 290 mutants of Tpl2 (Thr3 Ala and Thr3 Asp) were also cloned in the EcoRI site of the pMSCV-based retrovirus vector pMigR1 (kindly provided by Warren Pear, University of Pennsylvania).
Transfections, Retrovirus Infections, and Small Interfering RNA Experiments-Transfection of 293 cells was carried out in 6-well plates (1 g of DNA/plate) or 100-mm plates (8 g of DNA/plate). Plasmid DNA was mixed with FuGENE 6 (Roche Applied Science) in Opti-MEM I (Invitrogen), and following 25 min of incubation at room temperature, the mixture was added to 40% confluent 293 cell monolayers. Cells were harvested 48 h later. Cultures were serum-starved for 18 h prior to harvesting in all of the experiments in which ERK phosphorylation was the end point. Retrovirus packaging and infection were carried out as follows. pMigR1 retrovirus constructs of wild type Tpl2.HA, Tpl2 T290A.HA, and Tpl2 T290D.HA were transfected into 293-Phoenix cells in combination with pVSVG and pCL Ampho. The culture medium was changed 24 h later. Viral supernatants were collected at the 72-h time point, and they were concentrated 30-fold by ultracentrifugation (20,000 rpm in the SW 40Ti rotor for 3 h at 4°C). Immortalized macrophages were infected with the concentrated viruses using the spinfect method (21). Efficiencies of infection achieved by this method were reproducibly higher than 50%. Cells were serum-starved 48 h after infection, and they were stimulated with LPS at the 72-h time point.
Three double-stranded Stealth TM IKK␤ RNA oligonucleotides and one control oligonucleotide were synthesized (Invitrogen) and tested in RAW264.7 macrophages. One of them (sense 5Ј-GCCAAGGAGGAGAU-CUCCGAAGAUA-3Ј) was successful in knocking down mouse IKK␤ by 70%. This siRNA was used for subsequent experiments. All of the siRNAs (150 nM) were transfected into RAW264.7 cells using Lipofectamine 2000 (Invitrogen). Efficiencies of siRNA transfer were monitored using siGLO siRNA purchased from Dharmacon. Cell lysates were harvested 72 h after transfection.
Phosphopeptide Mapping-Tryptic phosphopeptide mapping was performed as described previously (22,23). pCMV5 constructs of wild type Tpl2.HA, Tpl2-T290A.HA, and Tpl2-T290D.HA were transfected into 293 cells, and they were metabolically labeled with [ 32 P]orthophosphate (0.5 mCi/ml) (PerkinElmer Life Sciences). Immunoprecipitated 32 P-labeled Tpl2 proteins were resolved in SDS-PAGE, transferred to polyvinylidene difluoride membranes, and visualized by autoradiography. Following elution from the membranes by treatment with 0.5% polyvinylpyrrolidone-360 in 0.1 M acetic acid for 30 min at 37°C, the samples were digested with 20 g of TPCK-treated trypsin (Sigma) in 50 mM NH 4 HCO 3 , pH 8.0, for 5 h at 37°C. The resulting peptides were dried and, following a 2-h treatment with performic acid (1 volume of 30% H 2 O 2 , 9 volumes of formic acid), they were re-dried and resuspended in a buffer containing 2.2% formic acid and 7.8% acetic acid in H 2 O, pH 1.9. The resuspended peptides were spotted onto nitrocellulose plates, and they were separated in the first dimension by electrophoresis at pH 1.9 and in the second dimension by thin layer chromatography in phosphochromatography buffer (37.5% n-butyl alcohol, 25% pyridine, and 7.5% acetic acid in H 2 O).
Antibodies, Immunoprecipitations, and Western Blotting-Phospho-p44/p42 MAPK (Thr 202 /Tyr 204 ) and phospho-MEK1/2 (Ser 217 /Ser 221 ) antibodies as well as the corresponding antibodies that recognize "total" i.e. both phosphorylated and non-phosphorylated forms of these kinases, were purchased from Cell Signaling Technology. Anti-HA monoclonal and polyclonal antibodies were from Covance. Anti-NF-B1/p105 antibody was from Delta Biolabs. Anti-Tpl2 (M20) antibody and anti-Tpl2-agarose-conjugated antibody were from Santa Cruz Biotechnology. The phospho-Thr 290 -specific anti-Tpl2 antibody was raised by injecting rabbits with the peptide-CKDLRGT(PO 4 )EIYMSPE. Anti-rabbit IgG horseradish peroxidase and anti-mouse IgG horseradish peroxidase were purchased from Amersham Biosciences. Protein G-Sepharose was purchased from Invitrogen. Phosphospecific antibodies were diluted in 5% bovine serum albumin in TBS (Tris-buffered saline) supplemented with 0.1% v/v Tween 20 (TBS-T), whereas all of the other antibodies were diluted in 5% milk in TBS-T.
Cells were lysed in Triton X-100 lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM Na 2 EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerophosphate, 1 mM Na 3 VO 4, 1 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride). The protein concentration in the soluble fraction, isolated by centrifugation of the total lysates (12,000 rpm for 10 min at 4°C), was determined using the Bradford assay (Bio-Rad). For immunoblotting, 50 -100 g of protein was separated by SDS-PAGE, transferred onto polyvinylidene difluoride membranes (0.45 M, Millipore), and blocked for 1 h at room temperature with 5% nonfat milk dissolved in TBS-T. Following three washes with TBS-T, membranes were incubated overnight at 4°C with primary antibody, washed in TBS-T, and incubated for 1 h at room temperature with secondary antibody followed by enhanced chemiluminescence (ECL, Amersham Biosciences). Protein bands were quantitated using a densitometer (Bio-Rad GS-800, Quantity one). For immunoprecipitation, lysates (500 g-1.5 mg) were precleared with Sepharose G-beads and immunoprecipitated overnight using a monoclonal anti-HA (Tpl2.HA) antibody. Immunoprecipitates were washed three times with lysis buffer and re-suspended in 50 l of loading buffer containing ␤-mercaptoethanol. Aliquots of 20 l were resolved in SDS-PAGE, and blots were probed for Thr 290 -phosphorylated or total Tpl2.
In Vitro Kinase Assays-293 cells transiently transfected with Tpl2.HA constructs and immortalized Tpl2 ϩ/ϩ murine macrophages and RAW264.7 cells expressing stably near physiological levels of Tpl2.HA were lysed in Triton X-100 lysis buffer. Tpl2 immunoprecipitated from cell lysates was washed three times in lysis buffer and once in kinase buffer (25 mM Tris-HCl, pH 7.5, 5 mM ␤-glycerophosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 , and 10 mM MgCl 2 ). The washed immunoprecipitates were incubated with 20 M ATP, 1 Ci of [␥-32 P]ATP (PerkinElmer Life Sciences) and 0.5 g of GST-MEK1 (U. S. Biological) in kinase buffer for 30 min at 25°C. Phosphorylated GST-MEK1 was visualized by autoradiography. In some of the experiments, the kinase reaction was carried out in kinase buffer containing 200 M cold ATP and 0.5 g of GST-MEK1. In these experiments, phosphorylated GST-MEK1 was detected by probing Western blots of the products of the kinase reaction with the phospho-MEK1/2 antibody.

RESULTS
The Tpl2 Kinase Is Activated by LPS and TNF-␣-Our earlier studies had shown that Tpl2 transduces LPS signals that activate ERK in macrophages and that these signals are required for the induction of TNF-␣, cyclooxygenase-2, and other proinflammatory molecules (12,16). These data suggested that Tpl2 may be activated by LPS. In vitro kinase assays on Tpl2 immunoprecipitated from LPS-stimulated RAW264.7 macrophages indeed showed that the activity of the Tpl2 kinase, which is very low prior to stimulation (0 min), increases rapidly following stimulation (Fig. 1A). More recent studies have shown that Tpl2 is also activated by TNF-␣ in both macrophages and primary mouse embryo fibroblasts (20). 3 Most of our earlier studies on Tpl2 had been carried out on primary peritoneal or BMDM. Recently, we established permanent lines by infecting wild type and Tpl2 Ϫ/Ϫ BMDMs with an SV40 large T antigen retrovirus construct. Both the Tpl2 ϩ/ϩ (Tpl2(ϩ)Mac) and the Tpl2 Ϫ/Ϫ (Tpl2(Ϫ) Mac) lines express Mac-1. However, only the Tpl2(ϩ)Mac lines undergo ERK phosphorylation and activation when treated with LPS (Fig. 1B1). Therefore, these cell lines exhibit properties similar to the properties of primary macrophages. Stimulation of one of the Tpl2(ϩ)Mac lines with LPS induced the activation of Tpl2. The kinetics of activation were similar to those observed in LPStreated RAW264.7 macrophages (Fig. 1B2).
Tpl2 Is Phosphorylated at Thr 290 in Vivo-To address the potential role of activation loop phosphorylation in Tpl2 activation, we generated Tpl2 mutants in which phosphorylatable residues in this region were replaced one at a time with alanine or with the phosphomimetic aspartic acid (Fig. 2A). The expression constructs of these mutants, fused at their C terminus to a HA epitope tag along with expression constructs of HAtagged wild type Tpl2, were transiently transfected into 293 cells. Transfected cells were metabolically labeled with [ 32 P]orthophosphate (0.5 mCi/ml for 3 h). The labeled Tpl2 proteins were immunoprecipitated with an anti-HA antibody, and after SDS-PAGE, they were digested with trypsin. The resulting tryptic peptides were subjected to two-dimensional separation (electrophoresis versus thin layer chromatography), and they were detected by autoradiography. The results were interpreted to suggest that Tpl2 undergoes phosphorylation at multiple sites. Because the Tpl2 T290A mutant lacked phosphopeptide 8 (Fig. 2B), we conclude that Thr 290 is likely to be one of the Tpl2 phosphorylation sites.
Tpl2 T290A Is Inactive, whereas Tpl2 T290D Exhibits Weak Kinase Activity-Western blots of 293 cells transiently transfected with HA-tagged constructs of wild type Tpl2, Tpl2 K167M (kinase-inactive), and the activation loop mutants described in the preceding paragraphs were probed with anti-HA (Tpl2.HA), anti-phospho-ERK, and anti-ERK antibodies. The results showed that substitution of Ser 274 and Tyr 282 with either alanine or aspartic acid gave rise to Tpl2 kinase mutants unable to activate ERK, whereas substitution of Thr 278 with either residue gave rise to mutants that were as active as the wild type protein. On the other hand, substitution of Thr 290 with alanine had the opposite effect than its substitution with aspartic acid. Thus, whereas a T290A mutant failed to induce phosphorylation of ERK, a T290D mutant induced ERK phosphorylation, although weakly, suggesting that phosphorylation at Thr 290 may contribute to the activation of the Tpl2 kinase (Fig. 2C). To address this hypothesis, we carried out Tpl2 in vitro kinase assays using Tpl2 immunoprecipitated from 293 cells transiently transfected with HA-tagged expression constructs of wild type Tpl2, Tpl2 K167M, Tpl2 T290A, and Tpl2 T290D. Kinase activity was determined by measuring the ability of the immunoprecipitated proteins to trans-phosphorylate GST-MEK1 and to undergo autophosphorylation. The results showed that, whereas Tpl2 T290A is inactive, Tpl2 T290D is active. However, its specific activity was lower than that of wild type Tpl2 (Fig. 2D). These data collectively indicate that Thr 290 is Tpl2 phosphorylation site.
Tpl2 T290D Transduces LPS Signals in Reconstituted Tpl2 Ϫ/Ϫ Macrophages, whereas Tpl2 T290A Does Not-To confirm the biological significance of Tpl2 phosphorylation at Thr 290 , we also examined whether Tpl2 T290A and Tpl2 T290D restore ERK phosphorylation by LPS in immortalized Tpl2(Ϫ)Mac macrophages. To address this question, Tpl2(Ϫ)Mac macrophages were infected with pMigR1 retrovirus constructs expressing wild type Tpl2, Tpl2 T290A, or Tpl2 T290D. Alternatively, they were infected with the empty pMigR1 vector. Infected cells were stimulated with LPS, and they were harvested 30 min later. The results showed that wild type Tpl2 fully restores and Tpl2 T290D partially restores ERK activation by LPS, whereas Tpl2 T290A is inactive (Fig. 3). We conclude that phosphorylation at Thr 290 plays an obligatory role in the physiological activation of Tpl2 by LPS in macrophages. Because in the absence of LPS stimulation Tpl2 T290D did not activate ERK, we also conclude that phosphorylation at Thr 290 is required but not sufficient for the activation of Tpl2.
A Thr 290 -specific Phospho-Tpl2 Antibody-Phosphorylation at Thr 290 Targets Preferentially the p58 Isoform of Tpl2 and Does Not Depend on the Activity of Tpl2 Kinase-To further explore the biological significance of Tpl2 phosphorylation at Thr 290 , we raised a rabbit polyclonal antibody against the phosphoryl-FIG. 2. Catalytically active wild type Tpl2 is phosphorylated at Thr 290 . Tpl2 T290A is inactive, whereas Tpl2 T290D exhibits weak kinase activity. A, potential phosphorylation sites and phosphorylation site mutants in the Tpl2 activation loop. B, two dimensional separation of phosphopeptides generated by trypsin digestion of wild type Tpl2 (WT Tpl2) and Tpl2 T290A transfected into 293 cells and labeled metabolically with [ 32 P]orthophosphate. Phosphopeptide 8 was not detected in tryptic digests of the T290A mutant. The same phosphopeptide was not detected in tryptic digests of the T290D mutant (data not shown). C, Western blots (IB) of 293 cells transiently transfected with HA-tagged constructs of wild type Tpl2.HA (WT), kinase-inactive Tpl2 (K167M.HA), and Tpl2 mutants at potential activation loop phosphorylation sites were probed with anti-HA (Tpl2), anti-phospho-ERK, and anti-total ERK antibodies. Whereas a T290A mutant failed to induce phosphorylation of ERK, a T290D mutant induced ERK phosphorylation, although weakly, suggesting that phosphorylation at Thr 290 may be required for Tpl2 activation. The posttranslational modifications responsible for the slow migrating Tpl2 bands have not been determined. D, 293 cells were transiently transfected with wild type Tpl2.HA (WT), Tpl2 K167M.HA, Tpl2 T290A.HA, and Tpl2 T290D.HA or the empty vector. Cell lysates were probed with antibodies against HA (Tpl2) and against phosphorylated or total ERK. The same lysates were used to carry out in vitro kinase assays. The end point of these assays was the auto-phosphorylation of Tpl2 and phosphorylation of GST-MEK1. IP, immunoprecipitated. ated peptide-CKDLRGT(PO 4 )EIYMSPE, which spans the phosphorylation site. The antibody was first used to probe Western blots of Tpl2 immunoprecipitated from lysates of 293 cells transiently transfected with expression constructs of HAtagged wild type Tpl2, Tpl2 K167M, Tpl2 T290A, Tpl2 T290D, or the empty vector. The same immunoprecipitates were probed with a polyclonal anti-HA antibody. The phosphospecific antibody found two bands (Fig. 4A, upper panel) that appeared to correspond to the 52-(p52) and 58-kDa (p58) Tpl2 isoforms detected with the anti-HA antibody (Fig. 4B, lower  panel). Of these bands, p58 band was detected with the phosphospecific antibody in cells transfected with wild type Tpl2, Tpl2 K167M, and Tpl2 T290D but it was missing from cells transfected with the empty vector or it was very faint in cells transfected with Tpl2 T290A. In contrast, the p52-like band was detected in the lysates of all the cells, including those transfected with the vector and the Tpl2 290A construct (Fig.  4A, upper panel). To determine whether the p52-like band was indeed the p52 Tpl2 isoform, Tpl2.HA immunoprecipitated from lysates of 293 cells transfected with wild type Tpl2 or with the empty vector was probed with the phosphospecific antibody and re-probed after stripping the first antibody with a polyclonal anti-HA (Tpl2) antibody. The results (Fig. 4B) showed that the p52-like band was not p52 and that the phosphospecific antibody detects only the p58 isoform. Therefore, these data showed that, of the p58 and p52 isoforms of Tpl2, only the p58 isoform undergoes phosphorylation at Thr 290 .
To confirm that the antibody is indeed specific for Tpl2 phosphorylated at Thr 290 , we used it to probe a Western blot of Tpl2 immunoprecipitated from lysates of 293 cells transfected with expression constructs of wild type Tpl2, Tpl2 T290A, Tpl2 T290D, or with the empty vector. The same blot was stripped, and following treatment with alkaline phosphatase, it was reprobed with the same antibody. The results confirmed that the antibody detects only the p58 isoform and that treatment with alkaline phosphatase practically eliminates the immunoreactivity of wild type Tpl2 but not the immunoreactivity of Tpl2 T290D (Fig. 4C). We conclude that the antibody indeed detects specifically Tpl2 phosphorylated at Thr 290 .
Tpl2 Phosphorylation at Thr 290 Is Induced by LPS and Depends on the Activity of IKK␤-To determine whether phosphorylation of Tpl2 at Thr 290 is induced by LPS, RAW264.7 cells engineered to stably express Tpl2.HA were LPS-stimulated. Tpl2.HA was immunoprecipitated from cell lysates harvested at the indicated time points before and after LPS stimulation. Probing the immunoprecipitates with the anti-phospho-Thr 290 antibody revealed that Tpl2 is not phosphorylated at Thr 290 in unstimulated cells but becomes phosphorylated in response to LPS. Probing the same lysates with the rabbit polyclonal anti-HA antibody showed that the p58 isoform is rapidly degraded following treatment with LPS, as expected. Finally, probing for phosphorylated and total ERK showed that ERK phosphorylation correlates with the phosphorylation of Tpl2 at Thr 290 (Fig. 5). Similar experiments with cells stimulated with TNF-␣, rather than LPS, revealed that TNF-␣ also induces phosphorylation of Tpl2 at Thr 290 (data not shown).
Recent studies provided evidence that IKK␤ is required for Tpl2 activation by LPS and TNF-␣ (19,20). To determine whether IKK␤ is required for Tpl2 phosphorylation at Thr 290 , we first treated RAW264.7 Tpl2.HA cells with the potent IKK␤ inhibitor PS-1145 (Fig. 6A1). Alternatively, the cells were transfected with the IKK␤ siRNA oligonucleotide or with a control double-stranded RNA oligonucleotide. The siRNA-mediated down-regulation of IKK␤ was monitored by Western blotting (Fig. 6A2). PS-1145-treated and non-treated cells and siRNA or control oligonucleotide-treated cells were stimulated with LPS, and they were harvested 30 min later. Tpl2.HA was immunoprecipitated from cell lysates using a monoclonal anti-HA antibody, and the immunoprecipitates were probed with phospho-Thr 290 and anti-HA polyclonal antibodies. The results (Fig. 6A) showed that pretreatment with PS-1145 or IKK␤ siRNA blocks Tpl2 phosphorylation at Thr 290 . In parallel experiments, we transiently transfected wild type Tpl2 into 293 cells and we examined the phosphorylation of the wild type protein before and after treatment with PS-1145. Control 293 cells were transfected with Tpl2 T290A.HA. The results confirmed that overexpressed Tpl2, which is constitutively active, is also constitutively phosphorylated at Thr 290 . In addition, they demonstrated that treatment with PS-1145 blocks the phosphorylation of overexpressed Tpl2 (Fig. 6B). Therefore, phosphorylation of Tpl2 at this site depends on the activity of IKK␤. DISCUSSION Evidence presented in this report showed that Tpl2 undergoes phosphorylation at Thr 290 in vivo and that phosphorylation at this site is required for the enzymatic activity of overexpressed Tpl2, as well as for the physiological activation of Tpl2 by external signals. In unstimulated cells expressing physiological levels of Tpl2, the Tpl2 kinase was inactive and lacked detectable phosphorylation at Thr 290 . However, both kinase activity and phosphorylation were induced upon stimulation with LPS or TNF-␣. When overexpressed in 293 cells, wild type Tpl2 underwent phosphorylation at Thr 290 and it was catalytically active in the absence of stimulation. Upon transient transfection into 293 cells, the kinase-inactive mutant Tpl2 K167M also underwent phosphorylation at Thr 290 , suggesting that Thr 290 is a site of trans-phosphorylation rather than auto-phosphorylation. Additional studies revealed that inhibition of IKK␤, a positive upstream regulator of Tpl2, blocks the phosphorylation of Tpl2 at Thr 290 , suggesting that phosphorylation depends on the activity of IKK␤. The finding that Tpl2 undergoes phosphorylation at Thr 290 in vivo is in agreement with a recent report (24) that appeared while this work was being prepared for publication. However, the published report (24) failed to show that phosphorylation is induced by LPS or TNF-␣ and that it is required for the physiological activation of Tpl2 by external signals. Moreover, it did not address the relationship of Tpl2 phosphorylation at Thr 290 with upstream signaling. All of these issues are addressed in the present study.
In unstimulated cells, endogenous Tpl2 is stoichiometrically bound to NF-B1/p105 (9). Bound Tpl2 is stable but inactive (18,25). When overexpressed in transiently transfected 293 cells, a fraction of Tpl2 remains unbound. Despite the fact that this applies to wild type Tpl2 as well as to the mutants Tpl2 T290D and Tpl2 T290A, only wild type Tpl2 and The T290D mutant were catalytically active upon overexpression. Given the fact that, upon overexpression, Tpl2 or perhaps the unbound fraction of Tpl2 undergoes phosphorylation at Thr 290 , these findings suggest that phosphorylation at Thr 290 is required for the catalytic activity of the Tpl2 kinase.
Expression of wild type Tpl2 and Tpl2 T290D to near physiological levels in Tpl2 Ϫ/Ϫ macrophages restored ERK activation by LPS, whereas expression of Tpl2 T290A did not. These data demonstrate conclusively that phosphorylation of Tpl2 at Thr 290 is required for the physiological activation of Tpl2 by external signals. However, because the expression of Tpl2 T290D to near physiological levels in Tpl2 Ϫ/Ϫ macrophages did not activate ERK in the absence of LPS stimulation, these data also suggest that phosphorylation at Thr 290 is necessary but FIG. 6. Phosphorylation of Tpl2 at Thr 290 is IKK␤-dependent. A1, RAW264.7 Tpl2.HA cells were treated with PS-1145 (50 M) for 1.5 h prior to stimulation with LPS (1 g/ml). Tpl2 was immunoprecipitated (IP) with a monoclonal anti-HA antibody from cell lysates harvested 30 min later. Immunoprecipitates were probed with the phospho-Thr 290 antibody (upper panel). Lanes 4, 5, and 6 show data from an experiment done in triplicate. After stripping, the blot was reprobed with a polyclonal anti-HA antibody (middle panel). A Western blot was probed with anti-p105 antibody (lower panel). IB, immunoblotting. A2, RAW264.7 cells were transfected with the IKK␤ Stealth RNA oligonucleotide (sense 5Ј-GCCAAGGAGGAGAUCUCCGAAGA-UA-3Ј) or with a control double-stranded RNA oligonucleotide prior to stimulation with LPS. siRNA knocked down the expression of IKK␤ by 70% (lanes 1 and 2). Tpl2.HA immunoprecipitated from cell lysates harvested before and 30 min after the start of the stimulation was probed with the anti-phospho-Thr 290 antibody. Following the removal of the first antibody, the same blots were probed with a polyclonal anti-HA antibody. B, wild type Tpl2 HA, Tpl2 T290A.HA, and the empty vector were transiently transfected into 293 cells. 48 h later, both cultures were treated with PS-1145 for 1.5 h. Tpl2 was immunoprecipitated from lysates of the transfected cells with a monoclonal anti-HA antibody. Immunoprecipitates were probed with the phospho-Thr 290 antibody or with a polyclonal anti-HA antibody. not sufficient for Tpl2 activation.
Translational initiation of the Tpl2 mRNA from three consecutive ATGs gives rise to 58-(p58) and 52-kDa (p52) isoforms of the kinase. Interestingly, earlier studies had shown that p58 is preferentially released from p105 upon LPS stimulation (18). As a result, p58 is also preferentially activated and degraded in LPS-treated cells (18). Given the data presented in this report showing that p58 is phosphorylated at Thr 290 more efficiently than p52, we hypothesized that phosphorylation at Thr 290 may also promote the release of Tpl2 from p105. Data presented elsewhere (26) indeed confirmed that phosphorylation at Thr 290 is not only required for the catalytic activation of the Tpl2 kinase but also promotes the release of Tpl2 from p105 and is required for the degradation of Tpl2 via the proteasome in response to LPS.
When overexpressed in 293 cells, both the wild type and the kinase-inactive mutant of Tpl2 (K167M) undergo phosphorylation at Thr 290 . This finding suggests that phosphorylation at this site is mediated by another kinase rather than Tpl2 itself. Recent studies revealed that activation of the IKK complex is required for Tpl2 activation by upstream signals (19,20). Therefore, we addressed the question whether IKK is required for the phosphorylation of Tpl2 at Thr 290 . Data presented in Fig. 6 clearly show that both the inhibition of IKK␤ with PS-1145 and the knocking down of IKK␤ with the IKK␤ siRNA prevent phosphorylation of Tpl2 at Thr 290 . We conclude that IKK␤ may phosphorylate Tpl2 at Thr 290 directly or it may regulate the kinase that phosphorylates Tpl2 at this site. Given that the Thr 290 site and the consensus IKK␤ phosphorylation site are distinctly different (27), we favor the latter hypothesis. IKK␤ is also known to phosphorylate p105 and to promote its degradation (28). Based on this information, it has been hypothesized that IKK␤ activates Tpl2 by promoting the phos-phorylation and degradation of p105, which ultimately leads to the release at Tpl2 from its inhibitor. The data in this report provide evidence that IKK␤ promotes the activation of Tpl2 by promoting not only the phosphorylation of p105 but also the phosphorylation of Tpl2. Therefore, these data provide an additional link between IKK␤ and Tpl2 activation in cells stimulated with LPS. Overall, the data presented in this report identify Tpl2 phosphorylation at Thr 290 as a required step in Tpl2 activation by external signals. Based on these data, we propose the model of Tpl2 activation and signaling shown in Fig. 7. According to this model, IKK␤ contributes to the activation of Tpl2 by external signals not only by promoting the phosphorylation and degradation of p105, as previously suggested, but also by promoting the phosphorylation of Tpl2 at Thr 290 . Based on data in this report and the results of additional studies from this laboratory (26), Tpl2 phosphorylation at Thr 290 is required for the enzymatic activation of Tpl2, the release of Tpl2 from p105, and the subsequent degradation of Tpl2 via the proteasome in response to external signals. FIG. 7. Tpl2 phosphorylation at Thr 290 depends on IKK␤ and is required for Tpl2 activation by LPS. LPS signals promote the assembly and activation of the IKK complex. IKK␤, a component of this complex, phosphorylates p105 and promotes its degradation. In addition, it phosphorylates Tpl2 at Thr 290 , an event that is required for Tpl2 activation. Therefore, IKK␤ is required for the activation of Tpl2, not only because it phosphorylates p105 but also because it promotes the phosphorylation of Tpl2 at Thr 290 .