Role for Protein Kinase Cθ (PKCθ) in TCR/CD28-mediated Signaling through the Canonical but Not the Non-canonical Pathway for NF-κB Activation*

NF-κB is a family of essential transcription factors involved in both embryonic development and inflammatory responses of the immune system. NF-κB can be activated by two pathways, i.e. the canonical (NF-κB1) pathway, which acts through the catalytic components of the IκB kinase complex and leads to IκB phosphorylation, degradation, and subsequent NF-κB nuclear translocation, or the non-canonical (NF-κB2) pathway, which involves NF-κB-induced kinase-dependent proteolytic processing of p100/p52 to yield translocation-competent p52-containing NF-κB complexes. We examined the relative roles of the NF-κB1 and NF-κB2 pathways in TCR/CD28 costimulation. We found that TCR/CD28 costimulation activates the canonical but not the non-canonical NF-κB pathway and that the serine/threonine kinase protein kinase Cθ (PKCθ) is essential for TCR/CD28-mediated canonical NF-κB activation in T cells. Importantly, TCR/CD28 costimulation induces higher p52 protein levels in T cells, but this effect is secondary to enhanced de novo synthesis of p100, not to enhanced processing of extant p100; PKCθ deficiency impairs signal-dependent p52 accumulation because of defects in p100 production. Finally, we found that TCR/CD28 costimulation induces IκBα, IκBβ, and IκBϵ degradation, and PKCθ is required for IκBα and IκBϵ but not IκBβ degradation. PKCθ acts solely within the canonical pathway to activate NF-κB, and PKCθ deficiency impacts upon p100/p52 processing in a manner that is independent of NF-κB-induced kinase.

Rel, share a structural motif called a Rel homology domain that contains regions permitting protein dimerization, DNA binding, and nuclear localization. Limited proteolytic processing of the p100 and p105 proteins generates p52 and p50, respectively, and NF-B DNA binding and transactivation is carried out by heterodimers of p50 or p52 with one of the transactivation-domain containing Rel proteins (RelA, RelB, or c-Rel) (1). Activation (nuclear translocation and DNA binding) of NF-B heterodimers is stimulated by upstream signaling events that impact on proteins of the inhibitor of B (IB) kinase (IKK) complex, which consists of two catalytic components, IKK␣ (IKK1) and IKK␤ (IKK2), and the adaptor protein IKK␥ (NF-B essential modulator (NEMO)). Activated IKK␣ or IKK␤ phosphorylates proteins of the IB family, which is comprised of IB␣, IB␤, IB⑀, p100, and p105 (1). IB␣, IB␤, and IB⑀ bind to NF-B heterodimers in the cytoplasm, masking one (IB␣ and IB⑀) or both (IB␤) of the NF-B nuclear localization sequences. IB␣ and IB⑀ also contain nuclear export sequences that ensure rapid export of IB-bound NF-B molecules from the cell nucleus; therefore, when NF-B complexes are bound to IB, NF-B is predominantly localized in the cytoplasm (2,3). Importantly, the carboxyl-terminal portions of p100 and p105 bear significant homology to other IB proteins, contributing to the cytoplasmic retention of unprocessed p100 and p105 (4). The phosphorylation of IB␣ or IB⑀ proteins at amino-terminal serine residues, or of carboxyl-terminal residues in p100, by IKK complex proteins leads to degradation of IB sequences and the subsequent nuclear localization of NF-B complexes. Conversely, proteolytic processing and removal of IB-like sequences from p105 to generate p50 appears to be a constitutive process, and additionally, IB␤ appears to be largely insensitive to signal-dependent processing (1). Therefore, the majority of stimulus-dependent NF-B signaling appears to depend on the degradation of IB␣ (and/or IB⑀) or on p100/p52 processing.
There are two major pathways for NF-B activation, known as the canonical (NF-B1) and non-canonical (NF-B2) activation pathways. In the canonical pathway, cytoplasmic sequestration of preformed NF-B heterodimers is relieved by signaldependent stimulation of IKK␤ or IKK␣ activity and the subsequent phosphorylation and degradation of IB proteins (1). In contrast, in the non-canonical pathway, signaling by surface receptors leads to the activation of NF-B-induced kinase (NIK), a mitogen-activated protein kinase kinase kinase (MAP3K)-like enzyme that specifically phosphorylates and activates IKK␣ (5). NIK can also directly (but independently) bind both IKK␣ and p100, and upon activation by upstream signals, NIK recruits IKK␣ into a multiprotein complex with p100. Formation of this complex leads to the phosphorylation of the carboxyl-terminal region of p100, stimulating proteolytic processing of p100 to p52 (6). This processing releases p52 from inhibition imposed by its carboxyl-terminal IB-like sequences, allowing p52-containing NF-B complexes to translocate to the nucleus and bind DNA. Together, the canonical and non-canonical pathways mediate the stimulus-dependent NF-B activation, although their relative importance in a given cell type depends on the constellation of surface receptors on a given cell and upon the nature of the activating stimulus received.
The relative roles of the canonical and non-canonical NF-B activation pathways in signaling by the T cell receptor (TCR) and the costimulatory receptor CD28 have yet to be elucidated. Here we present genetic data indicating that the serine/threonine kinase PKC acts downstream of TCR/CD28 signaling to mediate the activation of NF-B through the canonical pathway. Furthermore, we show that although NIK and IKK␣ are activated downstream of TCR/CD28 signaling, there is no signal-dependent increase in p100/p52 processing downstream of TCR/CD28 signaling. Instead, we elucidate a role for PKC in the canonical pathway for NF-B activation, and indirectly in p52 generation, by demonstrating that PKC deficiency leads to the selective impairment of the canonical NF-B activation pathway.
Cell Culture, Transfection, and Reporter Assays-Primary T cells were isolated from pooled spleens and lymph nodes of wild-type or PKC Ϫ/Ϫ C57BL/6 mice by standard procedures and enriched to Ͼ85% purity using a mouse T cell enrichment column (R&D Systems, Minneapolis, MN). The cells were stimulated for the indicated times with hamster anti-CD3 (10 g/ml) and/or anti-CD28 (2 g/ml) mAbs, which were cross-linked using a goat anti-hamster Ig (10 g/ml), or with phorbol myristate acetate (PMA; 50 ng/ml). Culture and transfection of human leukemic Jurkat-E6.1 or Jurkat-TAg cells, as well as reporter assays, were described elsewhere (9 -11). Reporter assays were performed at least three times with similar results.
Immunoprecipitation, Immunoblotting, and Kinase Assay-Immunoprecipitation, immunoblotting, and in vitro kinase assays were performed as described (9,12,13). For in vitro kinase assays, target kinases were immunoprecipitated from cell lysates prepared with radioimmune precipitation buffer. The immunoprecipitated kinases were incubated alone or with substrate at 30°C for 20 min in kinase buffer (20 mM HEPES, 20 mM ␤-glycerophosphate, 10 mM MgCl 2 , 10 mM p-nitrophenyl phosphate, 100 M Na 3 VO 4 , 2 mM dithiothreitol, 20 M ATP, 10 g/ml aprotinin, 150 mM NaCl, pH 7.5) containing [␥-32 P]ATP. Phosphorylated proteins were resolved by SDS-PAGE, transferred onto nitrocellulose membranes, and visualized by autoradiography. The membrane was used subsequently for immunoblotting to analyze the expression of protein substrates. Substrate phosphorylation and protein level were quantitated using NIH Image 1.61 densitometry software (rsb.info.nih.gov/nih-image/index.html).
Metabolic Labeling and Pulse-Chase-Jurkat-E6.1 cells in log phase were washed once with methionine-free RPMI 1640 medium, resuspended in the same medium containing 5% dialyzed fetal bovine serum at a final concentration of 10 ϫ 10 6 cells/ml, and cultured at 37°C for 30 min. For Fig. 5C below, 100 Ci/ml [ 35 S]methionine (ICN Biomedicals, Costa Mesa, CA) was added to the cell culture for 1 h at 37°C. The cells were washed twice with cold phosphate-buffered saline and resuspended in RPMI 1640 medium containing 10% fetal bovine serum. Cells were chased for different time periods with "cold" methionine in the presence of anti-CD3/CD28 costimulation. Cells were harvested at the indicated times, and lysates were precleared with protein G-Sepharose for 30 min and immunoprecipitated with anti-p100/p52 mAb. Immune complexes were resolved by SDS-PAGE, and the gels were dried and subjected to autoradiography. For Fig. 5B below, after culturing cells in methioninefree RPMI 1640 medium, the cells were pulsed in the presence of anti-CD3/CD28 costimulation for the indicated times, harvested, washed with cold phosphate-buffered saline, and processed as above. The radiolabeled protein bands were quantified using NIH Image 1.61 software.
Preparation of Nuclear Extracts-This experiment was performed by following a published procedure (14). Primary T cells were either left unstimulated or stimulated with anti-CD3 or anti-CD3 plus anti-CD28 mAbs for the indicated times. Nuclear extracts were prepared by first incubating the cells with a buffer containing 10 mM HEPES (pH 7.9), 10 mM KCl, 2 mM MgCl 2 , 0.1 mM EDTA, 0.2% Nonidet P-40, 1 mM dithiothreitol, and protease inhibitors for 30 s on ice. Cells were spun down at 6,000 rpm for 5 min in a benchtop centrifuge. The supernatant was kept as cytosol fraction. The nuclear pellets were resuspended in an extraction buffer containing 20 mM HEPES (pH 7.9), 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.63 M NaCl, 25% glycerol, 0.5 mM dithiothreitol, and protease inhibitors and then vigorously vortexed for 30 min at 4°C. The lysed nuclei were centrifuged at maximum speed for 30 min. Equal protein amounts of nuclear and cytosolic extracts were resolved by SDS-PAGE, transferred to nitrocellulose, and subjected to immunoblot analysis.

RESULTS
The Canonical NF-B Activation Pathway Is Impaired in PKC Ϫ/Ϫ T Cells-PKC is a serine/threonine kinase with a well known role in NF-B activation downstream of TCR/CD28 signaling in T cells (15). Studies using overexpression of wildtype PKC or its mutants (constitutively active PKC-A148E or kinase-dead PKC-K/R) first placed PKC downstream of TCR/ CD28 signaling in the activation of NF-B reporter constructs and demonstrated that PKC specifically activates IKK␤ but not IKK␣ (9,16) to phosphorylate IB proteins (which target IB proteins for degradation). Subsequently, two independently generated PKC gene knockout mice displayed defects in NF-B activation following TCR/CD28 stimulation in peripheral T cells (7,17), confirming an essential role for PKC in TCR/CD28-mediated NF-B activation. Accordingly, we chose to investigate the mechanism of NF-B impairment exhibited in T cells from PKC Ϫ/Ϫ mice (7).
We examined whether PKC deficiency leads to the impairment of IKK␤ activation by comparing the in vitro kinase activity of endogenous IKK␤ immunoprecipitated from wild-type and PKC Ϫ/Ϫ T cells following PMA or anti-CD3/anti-CD28 stimulation. We found that immunoprecipitated IKK␤ has little activity in resting T cells but that it efficiently phosphorylates GST⅐IB␣ following PMA (Fig. 1A) or anti-CD3/CD28 (Fig. 1B) stimulation in wild-type peripheral mouse T cells. In contrast, IKK␤ activity was severely impaired in PKC Ϫ/Ϫ T cells under these same conditions (Fig. 1, A and B), indicating a requirement for PKC in the initiation of stimulation-dependent IKK␤ catalytic activation.
In the canonical pathway for NF-B activation, stimulation of IKK leads to the phosphorylation and degradation of IB proteins. At present, it is unclear which IB proteins play a principal role in the TCR/CD28-mediated activation of NF-B, so we examined which IB isoforms are degraded in response to TCR/CD28 signaling. Additionally, because PKC Ϫ/Ϫ T cells exhibit impaired IKK␤ activity, we examined whether PKC deficiency also impacts stimulation-dependent degradation of specific IB proteins. We observed that either PMA ( Fig. 2A) or anti-CD3/CD28 (Fig. 2B) stimulation could induce significant degradation of both IB␣ and IB⑀ and to a lesser extent, IB␤, in wild-type T cells.
This finding is consistent with previous reports indicating that IB␣ and IB⑀ are degraded in a stimulus-dependent manner but that IB␤ is relatively stimulus-insensitive (18). However, probably due to the defect in IKK␤ activation (Fig. 1), we observed that T cells from PKC Ϫ/Ϫ mice show impaired stimulus-dependent (PMA or ␣-CD3/CD28) degradation of both IB␣ and IB⑀ but not IB␤ (Fig. 2). Furthermore, although ligation of TCR/CD28 surface receptors results in nuclear translocation of both p50 and p65/RelA in wild-type T cells, we observed a block in nuclear translocation of p50 and p65/RelA in PKC Ϫ/Ϫ T cells (Fig. 3), which likely resulted from impaired signal-dependent IB degradation. Together, these data supply genetic evidence that PKC acts through the canonical pathway for NF-B activation by causing the activation of IKK␤, resulting in the degradation of IB proteins and NF-B nuclear translocation.
NIK and IKK␣ Are Activated by TCR/CD28 Costimulation-In addition to the signal-dependent degradation of IB proteins described above, other pathways may exist in T cells for regulation of NF-B activation, for example, through the non-canonical pathway of NF-B activation, which involves the MAP3K-like protein NIK. Although NIK overexpression has been demonstrated to specifically activate IKK␣ in T cells, and IKK␣ can be activated downstream of TCR/CD28 signaling in T cells (22), the operation of such a non-canonical pathway has not been demonstrated in T cells in response to TCR/CD28 signaling.
BecauseNIKactivityisrequiredfornon-canonicalstimulationdependent p52 generation (5), we examined whether NIK is activated following TCR/CD28 signaling, as represented by the autophosphorylation of immunoprecipitated endogenous NIK in vitro. We found that the strongest stimulation of NIK activity was obtained by subjecting T cells to a combination of PMA treatment and CD28 cross-linking (Fig. 4A). However, we also found that anti-CD3 stimulation alone could induce increased NIK autophosphorylation in Jurkat-E6.1 cells after 30 min of stimulation. Importantly, anti-CD3/CD28 costimulation induced NIK activity sooner than anti-CD3 stimulation alone, with NIK activity peaking by 10 min after stimulation (Fig.  4A). Therefore, NIK kinase activity is stimulated following TCR/CD28 signaling in T cells.
Another signature event in the non-canonical p100/p52 processing pathway is the catalytic activation of IKK␣, which is FIG. 1. Impaired IKK␤ activation in PKC ؊/؊ T cells. Primary T cells purified from PKC ϩ/ϩ or PKC Ϫ/Ϫ mice were stimulated for 5 min with 50 ng/ml PMA (A) or for 1 h with plate-bound anti-CD3 mAb plus soluble anti-CD28 mAb (10 and 2 g/ml, respectively) (B) as indicated. Endogenous IKK␤ was immunoprecipitated (IP) and assayed for kinase activity (KA) against its specific substrate, GST⅐IB␣ 1-54 . The same membranes were immunoblotted (IB) with an anti-IKK␤ antibody. As a control, whole cell lysates (WCL) were also immunoblotted with an anti-PKC antibody. The numbers under the panels represent the -fold increase in IKK␤ activity when compared with unstimulated cells (ϭ 1) as determined by densitometry.  Fig. 1 for the indicated times. Nuclear and cytosolic extracts were prepared, and p50 or p65 expression in each fraction, as well as PKC expression in the cytosol fraction, was determined by immunoblotting with specific antibodies. Two nonspecific bands were used as a nuclear protein loading control, and actin was used as a loading control for the cytosol fraction.
indispensable for processing of p100 (19). Accordingly, we investigated whether IKK␣ activity is stimulated following TCR/ CD28 ligation in T cells by examining the autophosphorylation of endogenous IKK␣. Fig. 4A shows that anti-CD3 or anti-CD3/ CD28 stimulation induced IKK␣ activity that corresponded in intensity and timing with the activation of NIK described above. This result is consistent with an earlier report demonstrating that NIK could specifically activate IKK␣ in T cells (22). Finally, we found that overexpression of a dominantnegative (kinase-inactive) form of NIK (DN-NIK) could block anti-CD3/CD28-mediated stimulation of an NF-B reporter gene in Jurkat-TAg cells, although in our hands, DN-NIK had to be expressed at very high levels to achieve near complete blockade of NF-B reporter activation (Fig. 4B). Taken together, these data show that TCR/CD28 signaling can stimulate NIK and IKK␣ activity, which are principal components of the non-canonical NF-B activation pathway downstream of TCR/CD28 signaling, in T cells.
The Non-canonical Pathway for NF-B Activation Is Not Operative Downstream of TCR/CD28 Signaling-In the case of lymphotoxin ␤ receptor (LT␤R) signaling, receptor-mediated stimulation of p100/p52 processing requires NIK and IKK␣, and increases in p52 levels correspond with decreasing cellular p100 protein levels, indicative of the induced processing of p100 to form p52 (20). Because both NIK and IKK␣ are activated downstream of TCR/CD28 signaling, we examined T cells for evidence of p100/p52 processing following TCR/CD28 signaling and found that stimulation of murine primary T cells through anti-CD3/CD28 treatment resulted in increased levels of p52 (Fig. 5A). However, this increase in p52 levels paralleled a time-dependent increase in p100 protein levels. Furthermore, the TCR/CD28-stimulated increases in both p52 and p100 protein levels were blocked by the addition of cycloheximide, indicating a requirement for de novo protein synthesis. These findings raised the possibility that the increase in p52 observed after TCR/CD28 stimulation is secondary to increases in p100 protein levels. To address this possibility, we stimulated [ 35 S]methionine-labeled Jurkat-E6.1 T cells with anti-CD3/ CD28 mAbs for different durations and examined the level of immunoprecipitated p100 and p52 protein in the cells by autoradiography. Fig. 5B shows that p100 steadily accumulates in unstimulated cells and that there appears to be a constitutive level of p52 processing ongoing in these cells. Nonetheless, increased p100 levels were detected as early as 1 h after stimulation and were quite marked by 5 h after stimulation, whereas p52 increased in a delayed fashion, with the first detectable increase observed at 5 h after stimulation. In contrast, when the cells were pulsed for 1 h with [ 35 S]methionine and then chased for increasing durations in the presence of TCR/CD28 costimulation, no changes were observed in either p100 or p52 protein levels (Fig. 5C). Therefore, unlike LT␤R signaling (20), but similar to lipopolysaccharide (LPS) signaling (21), TCR/CD28 costimulation does not induce detectable p100/p52 processing. Instead, the increases in p52 (Fig. 5, A  and B) do not appear to result from enhanced p100 degradation but rather are likely secondary to TCR/CD28-stimulated increases in p100 protein levels. Similar experiments in primary T cells were inconclusive due to technical difficulties in labeling these cells with [ 35 S]methionine to a sufficiently high specific activity (data not shown).
PKC Operates Independently of NIK to Activate NF-B-Because both PKC-dependent NF-B activation and NIK activation occur downstream of TCR/CD28 signaling, it was important to determine whether these two proteins cooperate in the activation of NF-B signaling or whether they act in separate pathways. As demonstrated by others (22), we found that DN-NIK could block the activation of an NF-B reporter construct following anti-CD3/CD28 stimulation (Fig. 4B). However, DN-NIK failed to block NF-B activation induced by expression of a constitutively active PKC mutant (PKC-A148E) in Jurkat-TAg cells (Fig. 6A), indicating that NIK does not act downstream of PKC in the PKC-mediated activation of NF-B. Moreover, consistent with earlier reports (9, 16), we found that PKC-A148E specifically activated IKK␤ but not IKK␣ in an in vitro kinase assay (Fig. 6B). Furthermore, overexpression of DN-NIK failed to block the PKC-A148E-induced activation of IKK␤ as demonstrated by the fact that FLAGtagged IKK␤ immunoprecipitated from cells expressing both PKC-A148E and DN-NIK phosphorylated GST⅐IB␣ in vitro to the same extent as FLAG-IKK␤ immunoprecipitated from cells expressing PKC-A148E alone (Fig. 6C). These data, together with the finding from another group (16) that NIK specifically activates IKK␣ but not IKK␤, indicate that NIK does not act downstream of PKC to activate NF-B.
Since PKC and NIK are separately activated by TCR/ CD28 signaling, we hypothesized that NIK activation should be intact in PKC Ϫ/Ϫ T cells. Accordingly, we observed that endogenous NIK from murine primary T cells exhibited a significant basal level of kinase activity against myelin basic protein in vitro but that stimulation of the cells through ␣-CD3/CD28 (Fig. 7A) or through PMA (Fig. 7B) caused an increase in NIK-mediated myelin basic protein phosphoryla- tion. This signal-dependent increase in NIK activity was unaffected by PKC deficiency, as evidenced by the fact that the in vitro kinase activity of NIK following PMA or anti-CD3/CD28 stimulation of PKC Ϫ/Ϫ T cells was identical to that observed in wild-type T cells (Fig. 7, A and B). Of note, the finding that PMA-mediated NIK activation is unaffected by PKC deficiency (Fig. 7B) rules out a role for PKC in the PMA-stimulated NIK activation observed here and in Fig. 4A.
Taken together, the above data indicate that T cell activation signaling separately regulates PKC and NIK. Both NIK and IKK␣ are activated downstream of TCR/CD28 signaling, and NIK activation is normal downstream of TCR/CD28 signaling in PKC Ϫ/Ϫ cells. However, although NIK activation and IKK␣ activation are required for stimulation of the non-canonical NF-B pathway, the experiments in Fig. 5 demonstrated that TCR/CD28 signaling does not drive signal-dependent increases in p100/p52 processing. Therefore, NIK activation and IKK␣ activation are apparently not sufficient to drive signal-dependent p100/p52 processing; instead, p52 generation in response to FIG. 5. TCR/CD28 costimulation increases p100 synthesis and the de novo p100 synthesis-dependent processing of p100/p52. A, primary wild-type T cells were stimulated for the indicated times with anti-CD3 plus anti-CD28 mAbs in the absence or presence of cycloheximide (CHX, 10 g/ml). Cell lysates were immunoblotted with an anti-p100/p52 antibody. B, Jurkat-E6.1 cells were labeled with [ 35 S]methionine and stimulated in parallel with anti-CD3/CD28 mAbs for the indicated times. p100/p52 immunoprecipitates were analyzed by SDS-PAGE and autoradiography. C, Jurkat-E6.1 cells were pulsed with [ 35 S]methionine for 1 h and chased in medium containing cold methionine in the presence of anti-CD3/CD28 costimulation for the indicated time periods. p100/p52 immunoprecipitates were analyzed as in B. These experiments were repeated three times with similar results.

FIG. 6. NIK is not involved in PKC-induced NF-B signaling pathway.
A, Jurkat-TAg cells were cotransfected with Xpress-PKC-A/E, NF-B-Luc, and ␤-galactosidase reporter plasmids, plus empty vector (pEF) or the indicated amount of HA-DN-NIK constructs. After 24 h, cell extracts were prepared, and normalized NF-B activity levels (upper panel) or protein expression levels (lower panels) were determined. This result is representative of at least three similar experiments. B, Jurkat-TAg cells were cotransfected with Xpress-PKC-A/E and FLAG-IKK␤ or FLAG-IKK␣ constructs as indicated. After 24 h, IKK␤ and IKK␣ were immunoprecipitated (IP) with an anti-FLAG mAb and subjected to an in vitro kinase assay. The same membrane was immunoblotted (IB) with the anti-FLAG mAb. Cell lysates were also immunoblotted with an anti-Xpress antibody to determine PKC-A/E expression. WCL, whole cell lysates. C, Jurkat-TAg cells were cotransfected with the indicated plasmids. After 24 h, FLAG-IKK␤ was immunoprecipitated and subjected to an in vitro kinase assay. Expression of immunoprecipitated FLAG-IKK␤, Xpress-PKC-A/E, and HA-DN-NIK was analyzed by immunoblotting with the corresponding antibodies.
TCR/CD28 signaling appears to be secondary to de novo synthesis of p100.
The p100 gene promoter is itself a target of NF-B (23), so it was possible that the increase in p100 level and the generation of p52 in response to TCR/CD28 signaling is an indirect effect of PKC-mediated NF-B activation. We therefore examined whether p100/p52 processing was affected by PKC deficiency by comparing the extent of TCR/CD28-stimulated p52 generation in wild-type versus PKC Ϫ/Ϫ murine peripheral T cells. We found that the TCR/CD28-dependent increase in p100 and p52 expression was impaired in PKC Ϫ/Ϫ T cells (Fig. 7C), indicating that PKC activity is required for p52 generation by TCR/ CD28 signaling. This requirement most likely reflects an indirect role of PKC in up-regulating p100 in an NF-B-dependent manner (Fig. 8). Of note, the basal level of p100/p52 in PKC Ϫ/Ϫ T cells was higher than in wild-type T cells, although p100 was not inducible in PKC Ϫ/Ϫ T cells. This result suggests that sustained basal activity of PKC may negatively regulate p100 through an unknown pathway other than the canonical NF-B pathway.

DISCUSSION
Here we provide genetic evidence for the specific activation of IKK␤ by PKC and subsequent phosphorylation and degradation of IBs. Although the specific action of PKC in IKK␤ activation has been demonstrated before (9,16), this earlier work relied on overexpression studies, so our data provide significant additional evidence that PKC acts within the canonical NF-B activation pathway. Furthermore, although both independently generated PKC knockout mice show deficiencies in TCR/CD28-dependent p50 nuclear mobilization (7,17), we extend these findings by showing that this defect in p50 nuclear mobilization results from impairments in ␣CD3/CD28stimulated IKK␤ activation and in the subsequent signal-dependent degradation of IB isoforms (primarily IB␣ and IB⑀), definitively placing the action of PKC downstream of TCR/CD28 and within the canonical pathway for NF-B generation.
Some surface signaling events have been demonstrated to stimulate p52 generation, including signaling through the LT␤R (20,21), LPS receptor (21), and CD40 (24). In each of these cases, signaling through the surface receptor induces an increase in the expression of p52 that is dependent on de novo protein synthesis; blockade of protein neosynthesis by cycloheximide prevents increases in stimulus-dependent cellular p52 levels even at very early time points (24). There are two mechanisms whereby signal-dependent increases in cellular p52 levels may be achieved. The first is direct and involves a stimulated increase in the rate of proteolytic processing of p100/p52, which is dependent on the catalytic activity of NIK (5) and the subsequent recruitment of a ubiquitin ligase to a  7. Intact TCR/CD28-induced NIK activation but impaired p100/p52 expression in PKC ؊/؊ T cells. Primary T cells purified from wild-type or PKC Ϫ/Ϫ mice were stimulated for 15 min with anti-CD3/CD28 mAbs (A) or for 5 min with PMA (B) as indicated. Endogenous NIK was immunoprecipitated (IP) and subjected to an in vitro kinase assay using myelin basic protein (MBP) as a substrate. IB, immunoblotting. C, primary T cells purified from wild-type or PKC Ϫ/Ϫ mice were stimulated with anti-CD3 plus anti-CD28 mAbs for the indicated times. Expression of p100/p52 was analyzed by immunoblotting, and the same membrane was reprobed with anti-PKC or -actin antibodies. lysine residue in p100 (25). In this scenario, for example, with LT␤R signaling, p52 generation may correlate with a decrease in the amount of p100 precursor (20). The second means by which p52 levels may become elevated in cells is not through enhanced processing of p100 but simply through increases in p100 protein levels. In this case, the rate of p100/p52 processing remains unchanged, but p100/p52 processing takes place co-translationally on p100 (21) so that the level of p52 protein in the cell is elevated due to the increased amount of p100 precursor being generated. This appears to be the case for TCR/CD28-stimulated and also for LPS-stimulated p52 increases.
Our data do not rule out a role for NIK and IKK␣ in TCR/ CD28-mediated p52 generation, but it does seem unlikely that NIK or IKK␣ plays a major role in this process because although NIK and IKK␣ activity is elevated downstream of TCR/ CD28 signaling, the rate of p100/p52 processing is not. This idea is consistent with the finding that NIK is not required for constitutive p100/p52 processing in mouse embryo fibroblasts, but both are required for LT␤R-mediated increases in p100/p52 processing (20,26). It is unclear whether IKK␣ protein expression is required for constitutive p100/p52 processing; however, IKK␣ kinase activity appears to be dispensable for constitutive p100/p52 processing (19,26). Interestingly, we found that overexpression of DN-NIK could incompletely block TCR/CD28mediated NF-B activation (complete blockade of TCR/CD28induced NF-B signaling could only be achieved when DN-NIK was drastically overexpressed) but that DN-NIK could not block PKC-A148E-mediated NF-B activation. We speculate that under normal conditions, TCR/CD28 signaling activates both IKK␣ and IKK␤, which may cooperate to induce canonical NF-B activation (and subsequent p100 protein synthesis). However, in the presence of DN-NIK, IKK␣ may be sequestered, leading to the impairment of NF-B induction. Overexpression of PKC-A148E may bypass this effect of DN-NIK by specifically overdriving the activation of IKK␤, which may be informative with respect to the relative roles of PKC and IKK␤ (when compared with IKK␣) in canonical NF-B activation.
Although it is generally accepted that TCR/CD28 signaling activates NF-B through the canonical pathway, generation of p52 downstream of TCR/CD28 signaling has never been directly addressed. Others have described stimulation of p52 processing by different combinations of costimulatory signals (e.g. CD2/CD28 costimulation) and weak, delayed activation of p52 after CD3 signaling (27), whereas another report indicated that signaling through CD28 alone leads to the nuclear translocation of pre-existing p52/RelA heterodimers but not to p100/ p52 processing or the generation of additional p100 (28). Our data indicate that a combination of TCR and CD28 signaling up-regulates p100 expression and co-translational processing of p52 much more strongly than either signal alone reportedly does (27,28). The data presented here therefore represent the first effort to directly address the mechanism of p52 up-regulation by TCR/CD28 costimulation.
Some surface receptor signaling events are capable of stimulating both the canonical and non-canonical NF-B activation pathways. For example, both pathways of NF-B activation are stimulated downstream of signaling through the LT␤R, by LPS signaling or, as we demonstrate here, by TCR/CD28 costimulation. In the case of LT␤R or LPS signaling, RelA-containing (canonical) NF-B nuclear complexes predominate in the early phase of the response, but these are gradually replaced by p52and RelB-containing complexes that result after delayed p100/ p52 processing (20,21). The generation of p52-containing nu-clear NF-B complexes results in long term NF-B activation that is required in developmental events such as the generation and maintenance of lymphoid architecture. However, the stimulation of p100 generation by TCR/CD28 signaling may have special consequences for T cell activation signaling and survival because the constellation of genes bound and activated by p52-containing hetero- (20,29) or homodimers (30) could be different from those genes that are activated by the p50/RelA heterodimers that are active early on in the response to TCR/ CD28 signaling. One potential gene target of p52 is the antiapoptotic gene Bcl-2, the transcription of which can be stimulated by p50 or p52 homodimers in combination with Bcl-3 (30). This interaction could be of particular significance if early Bcl-2 gene induction is mediated by p50 homodimers, but long term Bcl-2 production is mediated by p52 homodimers. This intriguing possibility has yet to be investigated, but it could provide an explanation for how TCR/CD28 costimulation mediates some of its antiapoptotic effects and the role of PKC in regulating these effects.