Activation of Nuclear Factor κB (NF-κB) in Prostate Cancer Is Mediated by Protein Kinase C ϵ (PKCϵ)*

Background: PKCϵ, a potential oncogene, is up-regulated in prostate cancer. Results: PKCϵ facilitates the formation of TNFR-I complex to regulate the NF-κB pathway via a C1 domain/diacylglycerol-dependent mechanism. Conclusion: PKCϵ is an upstream regulator of NF-κB signaling in prostate cancer. Significance: Mechanisms identified here may reveal novel PKCϵ effectors that contribute to prostate cancer progression and highlight the potential relevance of this pathway for therapeutic purposes. Protein kinase C ϵ (PKCϵ) has emerged as an oncogenic kinase and plays important roles in cell survival, mitogenesis and invasion. PKCϵ is up-regulated in most epithelial cancers, including prostate, breast, and lung cancer. Here we report that PKCϵ is an essential mediator of NF-κB activation in prostate cancer cells. A strong correlation exists between PKCϵ overexpression and NF-κB activation status in prostate cancer cells. Moreover, transgenic overexpression of PKCϵ in the mouse prostate causes preneoplastic lesions that display significant NF-κB hyperactivation. PKCϵ RNAi depletion or inhibition in prostate cancer cells diminishes NF-κB translocation to the nucleus with subsequent impairment of both activation of NF-κB transcription and induction of NF-κB responsive genes in response to the proinflammatory cytokine tumor necrosis factor α (TNFα). On the other hand, PKCϵ overexpression in normal prostate cells enhances activation of the NF-κB pathway. A mechanistic analysis revealed that TNFα activates PKCϵ via a C1 domain/diacylglycerol-dependent mechanism that involves phosphatidylcholine-phospholipase C. Moreover, PKCϵ facilitates the assembly of the TNF receptor-I signaling complex to trigger NF-κB activation. Our studies identified a molecular link between PKCϵ and NF-κB that controls key responses implicated in prostate cancer progression.

The NF-B family of transcription factors plays a crucial role in inflammation as well as in the development and progression of cancer. Extensive evidence indicates that the NF-B pathway is implicated in controlling the expression of genes involved in cell survival, proliferation, angiogenesis, and invasion (1,2). NF-B is a dimer formed by proteins of the Rel family (RelA/ p65, RelB, c-Rel, NF-B1/p50, and NF-B2/p52) that is retained in the cytoplasm as a complex with inhibitory IB proteins. In the canonical pathway, external stimuli such as proinflammatory cytokines promote the dissociation of the ternary complex (mainly that composed of IB␣-p50-p65), an event triggered by phosphorylation of IB␣ at Ser-32 and Ser-36 by IB␣ kinase (IKK), 2 followed by proteasomal degradation of IB␣. The released NF-B dimer is subsequently translocated into the nucleus where it binds specific elements in the promoters of NF-B-responsive genes (3). Abnormally high NF-B activity and aberrant expression of NF-B-regulated gene products are clinical hallmarks of chronic inflammation and have been widely linked to the cancer phenotype. Inflammation has indeed been shown to contribute to prostate cancer development via multiple mechanisms such as oxidative stress, genomic instability, and DNA damage or indirectly by increasing levels of proinflammatory factors such as tumor necrosis factor ␣ (TNF␣), which themselves affect cancer risk (4). NF-B hyperactivation may result from enhanced production of tumor-promoting cytokines, enhanced stimulation of growth factor receptors, and/or aberrant expression/activation of upstream NF-B kinases such as IKK and NF-B-inducing kinase (NIK) (5). The functional association of NF-B with oncogenic and tumor suppressor signaling networks, including the Ras/ERK, PI3K/Akt, and p53 pathways, argues for a high level of complexity in the mechanisms leading to NF-B hyperactivation (6,7).
Several lines of evidence strongly suggest that the NF-B pathway is dysregulated in prostate cancer and has been implicated in the progression to the androgen-independent state that ultimately leads to patient death (8). Constitutive NF-B activation has been reported in prostate tumors and has prognostic importance as it correlates with poor outcome in a subset of primary tumors (5). Nuclear p65 NF-B can be observed in organ-confined prostate tumors, suggesting that constitutive NF-B activation may be an early event in prostate cancer development (5,9). NF-B hyperactivation is a feature of prostate cancer cell lines lacking androgen receptor expression (such as PC3 and DU145 cells), whereas androgen-dependent cell lines (such as LNCaP and CWR22Rv1 cells) generally display low levels of basal NF-B activity (10). NF-B has been shown to mediate the effect of proinflammatory cytokines in prostate cancer cells, including TNF␣, IFN␥, and IL-1␤ (11). Binding of TNF␣ to its receptor TNFR-I in cancer cells, including prostate cancer cells, leads to the recruitment of adaptor proteins (TRADD, TRAF2, RIP) and the formation of a signaling complex that regulates NF-B activation and transcriptional activation of inflammatory, survival, and anti-apoptotic genes such as BCL2, BCL2L1, PTGS2 (COX2), and XIAP (3,(11)(12)(13).
There is ample evidence that PKC serine-threonine kinases are involved in the activation of NF-B. Although many studies highlighted the relevance of the atypical PKCs and as NF-B modulators (14,15), Diacylglycerol (DAG)/phorbol ester responsive PKCs also emerged as potential modifiers of NF-B signaling (16 -18). Both classical/conventional cPKCs (␣, ␤, and ␥) and novel nPKCs (␦, ⑀, , and ) have been implicated as regulators of apoptosis, survival, differentiation, mitogenesis, and transformation in a strict cell-type dependent manner. Studies from several laboratories, including ours, revealed that PKC␦ generally acts as a negative regulator of proliferation and/or mediates apoptotic responses, whereas PKC⑀ is a prosurvival and mitogenic kinase (19 -22). In prostate cancer cells, activation of PKC⑀ accelerates G 1 /S transition, mediates survival through Bad-dependent and Bad-independent mechanisms, and confers androgen independence (23)(24)(25). Most interestingly, PKC⑀ emerged as a potential oncogene and cancer biomarker and it is up-regulated not only in prostate cancer but also in several other epithelial cancers including lung, breast, and thyroid cancer (19,26,27). PKC⑀ up-regulation can be observed in Ͼ 95% of human prostate tumors and is common in advanced stages of the disease (19,28,29). Notably, overexpression of PKC⑀ in normal immortalized RWPE-1 prostate cells to levels observed in prostate cancer cells confers growth advantage and causes ERK and Akt activation (30). Our laboratory recently demonstrated that transgenic overexpression of PKC⑀ but not PKC␣ or PKC␦ in the mouse prostate induces prostatic intraepithelial neoplasia (PIN) (30). These findings thereby suggest a crucial role of PKC⑀ in prostate cancer development. However, little is known regarding the potential mechanisms underlying the effects of PKC⑀ in prostate tumorigenesis.
By means of cellular and animal models, in this study we identified a key role for PKC⑀ as a mediator of NF-B signaling in prostate cancer. PKC⑀ turned out to be an essential effector of TNF␣ and mediates constitutive activation of NF-B in androgen-independent prostate cancer cells. PKC⑀ regulates the expression of NF-B-responsive gene products implicated in prostate cancer development and progression. Interestingly, transgenic overexpression of PKC⑀ in mice conferred NF-B hyperactivation in preneoplastic lesions, arguing for a critical role for this nPKC in NF-B signaling.

EXPERIMENTAL PROCEDURES
Materials-TNF␣ was purchased from Pepro Tech (Rocky Hill, NJ). PMA was procured from LC Laboratories (Woburn, MA). The pan-PKC inhibitor GF 109302X (bisindolylmaleimide I) was obtained from BioMol (Plymouth Meeting, PA). The PKC⑀ inhibitor peptide ⑀V1-2 (Tat-fused) and the carrier Tat peptide were kindly provided by Dr. Daria Mochly-Rosen (Stanford University, CA). [ 32 P]␣-deoxy adenosine triphosphate (dATP) was from PerkinElmer Life Sciences (Santa Clara, CA). Fetal bovine serum was purchased from Hyclone (Logan, UT). Keratinocyte serum-free medium was purchased from Invitrogen. Other cell culture reagents and media were from the ATCC.
Transfection of Mammalian Expression Vectors-RWPE-1 and LNCaP cells were transfected with various PKC⑀ expression vectors or empty vector using Lipofectamine Plus (Invitrogen). Experiments were carried out 24 h after transfection.
Adenoviral Infections-Generation of the PKC⑀ adenovirus (AdV) has been described previously (32). Subconfluent LNCaP or RWPE-1 cells in 6-well plates were infected with the PKC⑀ AdV (multiplicity of infection ϭ 1 pfu/cell) in RPMI 1640 supplemented with 2% FBS or in keratinocyte serum-free medium without supplements, respectively. A LacZ AdV was used as control. After 4 h, viral particles were removed, and cells were incubated for 24 h in complete medium. Expression of PKC⑀ was readily detected 24 h after infection and remained stable for several days (Ref. 24) and data not shown).
Analysis of PKC⑀ Translocation by Ultracentrifugation-PKC⑀ translocation was determined by Western blot analysis of soluble and particulate fractions obtained by ultracentrifugation, as described previously (33). Briefly, LNCaP cells were harvested into a lysis buffer containing 20 mM Tris-HCl (pH 7.4), 5 mM EGTA, 5 g/ml 4-(2-aminoethyl)-benzenesulfonylfluoride, 1 g/ml pepstatin A, 5 g/ml aprotinin, and 5 g/ml leupeptin, and sonicated. The cytosolic (soluble) fraction was obtained by collection of the supernatant after centrifugation of the cell lysate (1 h at 100,000 ϫ g at 4°C), and the remaining pellet represented the particulate fraction.
Luciferase Reporter Assays-Cells in 12-well plates (7 ϫ 10 4 cells/well) were transfected with 0.5 g of a NF-B firefly luciferase reporter (kind gift from Dr. Dave Manning, University of Pennsylvania) vector using Lipofectamine Plus. The Renilla luciferase expression vector pRL-TK (50 ng, Promega, Madison, WI) was cotransfected for normalization of transfection efficiency. After 48 h, cells were stimulated with TNF␣ or vehicle and lysed 6 h later with passive lysis buffer (Promega). Luciferase activity was determined in cell extracts using the dualluciferase reporter assay system (Promega). Results were expressed as the ratio between firefly and Renilla luciferase activities.
Preparation of Nuclear Extracts and EMSA-Nuclear and cytosolic fractions were obtained after cell lysis using the NE-PER nuclear protein extraction kit (Pierce Biotechnology Inc., Rockford, IL). NF-B DNA-binding in nuclear extracts was determined by EMSA as described previously (34). Briefly, the NF-B oligonucleotide probe (5Ј-agcttGAGGGGATTC-CCTTA-3Ј) was labeled with [␣-32 P]deoxyadenosine triphosphate using Klenow enzyme and purified on a Sephadex G-25 column. The binding reaction was carried out at 25°C for 10 min with or without 5 g of nuclear proteins, 1 g of poly(dI-dC), and 10 6 cpm of labeled probe in a final volume of 20 l of binding buffer (10ϫ buffer) (100 mM Tris-HCl (pH 7.5), 500 mM NaCl, 50 mM MgCl 2 , 100 mM EDTA, 10 mM DTT, 1% Triton X-100, and 50% glycerol). Binding specificity was confirmed by cold competition. Decreased signal intensity was observed with 25-and 50-fold molar excess of cold NF-B probe, whereas the signal did not diminish with a cold AP-1 probe (Ref. 35 and data not shown). DNA-protein complexes were separated on a 6% non-denaturing polyacrylamide gel at 200 V. The gel was fixed and dried, and DNA-protein complexes were visualized by autoradiography.
RNA Isolation and cDNA Synthesis-Subconfluent LNCaP cells were treated with either TNF␣ (10 ng/ml) or vehicle, and 6 h later RNA was extracted using the RNeasy kit (Qiagen, Valencia, CA). 2 g of RNA/sample were reverse-transcribed using random hexamers as primers and the Taqman reverse transcription reagent kit (Applied Biosystems, Branchburg, NJ).
Real-time PCR-PCR primers and fluorogenic probes for COX2, VEGF, MMP9, and IL6 were purchased from Applied Biosystems. The probes were 5Ј end-labeled with 6-carboxyfluorescein. PCR amplifications were performed using an ABI PRISM 7700 detection system in a total volume of 12.5 l containing Taqman Universal PCR MasterMix (Applied Biosystems), commercial target primers (300 nM), the fluorescent probe (200 nM), and 1 l of cDNA. PCR product formation was continuously monitored using the Sequence detection system software version 1.7 (Applied Biosystems). The 6-carboxyfluorescein signal was normalized to endogenous 18 S.
Immunofluorescence and Confocal Microscopy-For NF-B nuclear translocation studies, LNCaP cells were plated on coverslides in 12-well plates and 48 h later stimulated with either TNF␣ (10 ng/ml) or vehicle. At the indicated times, cells were washed twice with PBS, fixed for 10 min with precooled methanol, washed three times for 5 min with PBS, and permeabilized for 15 min with 0.5% Triton X-100 in PBS followed by a 10-min incubation in 100 mM glycine in PBS. After a blocking incubation with goat serum in PBS for 30 min, sections were incubated with a rabbit anti-NF-B p65 antibody (1:250) overnight. The following day, sections were washed twice for 5 min with PBS and incubated with a CY3-conjugated anti-rabbit antibody (1:2000, Jackson ImmunoResearch Laboratories, Inc.) for 1 h. After additional washings, DNA was stained using DAPI (0.1 g/ml, 10 min). Coverslides were washed three times with PBS, mounted with Vectashield, and visualized by confocal microscopy.
For live cell imaging, cells were transfected with pEGFP plasmids using the Amaxa nucleofector, grown in glass-bottomed culture dishes (MatTek, Ashland, MA), and cultured for 24 h before being subjected to various treatments. Living cells were visualized with a confocal laser scanning fluorescence microscope (LSM 410 or 510, Carl Zeiss).
Lysates were vortexed in IP buffer for 20 min at 4°C to solubilize the membrane fraction and centrifuged, and the cleared extracts were incubated with 1 g of mouse monoclonal antibody to p55 TNFR-I (Upstate Biotechnology, Billerica, MA) overnight at 4°C with gentle rocking. After addition of antimouse IgG-agarose beads (Sigma), the complexes were incubated for 1 h at 4°C. Immunoprecipitated proteins were washed three times with cold IP buffer. The samples were eluted by incubation at 65°C for 15 min in 2ϫ SDS-PAGE sample buffer, resolved through SDS-PAGE, and subsequently immunoblotted with PKC⑀, RIP, TRAF2, or TRADD antibodies. For reverse coimmunoprecipitation, cell lysates were immunoprecipitated with 1 g of anti-rabbit PKC⑀ (Santa Cruz Biotechnology, Inc.). The samples were then analyzed by immunoblotting with an anti-TNFR-I antibody.
Generation of Prostate PKC⑀ Transgenic Mice and Animal Care-The generation and phenotypic characterization of probasin (PB)-driven PKC⑀ transgenic mice are described elsewhere (30). The standard nomenclature for this line is FVB/ N-Tg (Pbsn-Prkce) and is referred to as PB-PKC⑀. FVB/N inbred mice (used in maintenance of transgenic lines) were acquired from Charles River Laboratories, Inc. (strain code 207). Mice were housed in individually ventilated cages on autoclaved hardwood bedding in an Association for Assessment and Accreditation of Laboratory Animal Care-accredited facility at the MD Anderson Cancer Center, Science Park, Smithville, TX. All procedures were performed in compliance with the Public Health Service Guide for the Care and Use of Laboratory Animals, 8 th edition, 2010.
Immunohistochemistry-Immunohistochemical analysis was carried on 12-month-old homozygous Tg/Tg males on a pure FVB/N background and WT FVB/N males. Complete necropsy, macroscopic examination, tissue collection, and processing were carried out as described previously (30). Briefly, mice were sacrificed with CO 2 , and their entire genitourinary tract was removed. Prostates were submitted "en bloc" and processed in standard formalin-fixed, paraffin-embedded sections. For staining we used anti-phospho NF-B p65 antibody (1:100, Cell Signaling Technology, Inc.) and Envision plus labeled polymer anti-rabbit-HRP (Dako, Carpinteria, CA), followed by incubation with Dako 3,3Ј-diaminobenzidine (DAB) substrate.
Statistical Analysis-Results were compared by analysis of variance using GraphPad Prism 5.0. In all cases, p Ͻ 0.05 was considered statistically significant.

Prostate-specific Expression of PKC⑀ Correlates with NF-B
Hyperactivation-Extensive evidence suggests a key role for PKC⑀ in mitogenic, survival, and oncogenic signaling (27). PKC⑀ is distinctively up-regulated in human prostate cancer and mediates anti-apoptotic responses in prostate cancer cell lines in culture (28,29). As the NF-B pathway also plays a prominent role in human prostate cancer development, we decided to investigate a potential association between these two pathways. The PC3 and DU145 androgen-independent prostate cancer cell lines display elevated PKC⑀ levels relative to a normal immortalized prostate epithelial cell line (RWPE-1) or androgen-dependent LNCaP prostate cancer cells (Fig. 1A), as reported previously (30). A clear translocation of NF-B from the cytosol to the nuclear fraction could be seen in PC3 and DU145 prostate cancer cell lines, indicating the constitutive activation of NF-B in these cells, as reported (10). Notably, there is a strong correlation between PKC⑀ levels and nuclear NF-B expression (Fig. 1A) as well as with NF-B DNA-binding activity (as determined by EMSA) (Fig. 1B) across different prostate cells. These results argue for a potential relationship between PKC⑀ and NF-B activation status in prostate cancer cells.
To determine whether an association between PKC⑀ and NF-B occurs in an in vivo setting, we took advantage of a mouse transgenic model that we recently developed in which prostate-specific expression of PKC⑀ is driven by the PB promoter. PB-PKC⑀ mice develop a preneoplastic phenotype char-acterized by hyperplasia and PIN, with a penetrance of 100% at 12 months (30). Interestingly, preneoplastic lesions in PB-PKC⑀ mice display a characteristic nuclear NF-B staining. This effect was particularly prominent in PIN lesions relative to normal areas or regions with mild hyperplasia and could not be detected in prostates from age-matched WT control FVB/N male mice (Fig. 1C).
PKC⑀ Mediates NF-B Activation in Prostate Cancer Cells-First, we examined whether NF-B hyperactivation in androgen-independent PC3 cells is mediated by PKC⑀. When we silenced PKC⑀ in PC3 cells using two different RNAi duplexes, we noticed a significant reduction in phospho-IB␣ levels. Moreover, total IB␣ levels in PKC⑀-depleted cells were higher than in parental PC3 cells or cells subjected to control RNAi ( Fig. 2A). To further assess the role of PKC⑀ in the maintenance of NF-B hyperactivation, we took advantage of a NF-B lucif- erase reporter. As shown in Fig. 2B, silencing of PKC⑀ from PC3 cells led to a pronounced inhibition of NF-B luciferase activity. Similar results were observed in androgen-independent DU145 cells (supplemental Fig. S1). Therefore, PKC⑀ mediates NF-B constitutive activation in androgen-independent prostate cancer cells.
We then examined whether PKC⑀ plays a role in the activation of the NF-B pathway in prostate cancer cells. Phorbol 12-myristate 13-myristate (PMA), a prototypical PKC activator, is a known activator of NF-B responses (11). As expected, treatment of LNCaP cells with PMA (100 nM) induced a timedependent phosphorylation of IB␣, a well established readout of NF-B activation, with concomitant degradation of IB␣ (supplemental Fig. S2). Both PMA-induced IB␣ phosphorylation and degradation were sensitive to the "pan" PKC inhibitor GF 1090203X (Fig. 2C), suggesting that it is mediated by PKC and not by other phorbol ester receptors unrelated to PKC that have been described in prostate cancer cells (36 -38). To determine whether PKC⑀ mediates PMA-induced NF-B activation in prostate cancer cells, LNCaP cells were subjected to PKC⑀ RNAi using two different duplexes. A scrambled RNAi duplex was used as control. PKC⑀ depletion of Ͼ 80% was achieved with either PKC⑀ duplex (Fig. 2D) and did not affect the expression of other PKCs (see Refs. 24,39). IB␣ phosphorylation by PMA was markedly reduced in PKC⑀-depleted cells. Moreover, PKC⑀ RNAi rescued the degradation of IB␣ in response to PMA (Fig. 2D).
PKC⑀ Is Required for TNF␣-induced Activation of NF-B in LNCaP Cells-Inflammation is an important factor in the development of prostate cancer. NF-B is an established mediator of responses by inflammatory cytokines such as TNF␣, IFN␥, and IL-1␤, including in prostate cancer cells (11). Stimulation of LNCaP cells with TNF␣ (10 ng/ml) led to a time-dependent elevation in phospho-IB␣ levels with a concomitant decrease in total IB␣ levels (supplemental Fig. S3). TNF␣induced phosphorylation of IB␣ as well as the subsequent degradation of total IB␣ were inhibited by GF 1090203X (Fig. 2E, left panel), indicating that it is PKC-mediated. Furthermore, silencing PKC⑀ with two different RNAi duplexes significantly attenuated TNF␣-induced IB␣ phosphorylation and IB␣ degradation (Fig. 2E, center panel). To further establish the requirement of PKC⑀ in this response, we used a pharmacological approach. The cell-permeable, Tat-fused peptide ⑀V1-2 selectively blocks the translocation and activation of PKC⑀ but not of other PKCs (40). As shown in Fig. 2E (right panel), ⑀V1-2 significantly inhibited TNF␣-induced IB␣ phosphorylation compared with the Tat control peptide.
As a complementary approach to determine the involvement of PKC⑀ in TNF␣-induced activation of NF-B, we used a luciferase reporter assay. Upon transfection of a NF-B luciferase reporter plasmid into LNCaP cells, a significant activation of luciferase activity was observed in response to TNF␣, which was maximal at 6 h (data not shown) and was essentially blunted by expression of the IB␣ super-repressor (IB␣ rep ). Pretreatment with the PKC inhibitor GF 1090203X impaired TNF␣-induced activation of NF-B luciferase reporter activity (Fig. 2F, left panel). Moreover, RNAi silencing of PKC⑀ (Fig. 2F,  center panel) or pretreatment with the PKC⑀ inhibitor ⑀V1-2 (right panel) caused a marked reduction in NF-B luciferase reporter activity induced by TNF␣. Neither PKC␣ nor PKC␦ RNAi depletion reduced NF-B luciferase activity in response to TNF␣ (supplemental Fig. S4). From these results we concluded that PKC⑀ is required for the activation of NF-B by both PMA and TNF␣ in prostate cancer cells.
PKC⑀ Is Required for Nuclear Translocation of NF-B and Activation of NF-B DNA-binding Activity-Upon activation, NF-B p65 dissociates from phosphorylated IB and translocates to the nucleus where it binds to specific elements in NF-B-responsive gene promoters (2,3). The effect of PKC⑀ depletion on NF-B nuclear translocation was examined using a fractionation assay. In LNCaP cells, TNF␣ promotes the translocation of NF-B to the nucleus, as judged by the disappearance of NF-B from the cytosolic fraction and the concomitant elevation of NF-B in the nuclear fraction. This effect was reduced in PKC⑀ depleted cells (Fig. 3A). Similar conclusions were obtained using immunocytochemistry. Indeed, as shown in Fig. 3B, nuclear relocalization of NF-B can be readily observed in response to TNF␣ treatment. However, this effect was barely detected in PKC⑀-depleted LNCaP cells. Similar results were observed with PMA (data not shown). Moreover, NF-B nuclear translocation was significantly lower in LNCaP cells treated with the ⑀V1-2 peptide inhibitor compared with the control Tat carrier peptide (Fig. 3C).
To determine whether PKC⑀ was implicated in the activation of NF-B DNA-binding activity by TNF␣, we used an EMSA approach. TNF␣ caused a pronounced activation of NF-B DNA-binding activity in nuclear extracts of LNCaP cells. NF-B DNA-binding activity was reduced upon treatment with GF 109203X (data not shown). Most notably, PKC⑀-depleted LNCaP cells show defective activation of NF-B DNA binding activity in nuclear extracts (Fig. 3D). A similar effect was observed by treatment with ⑀V1-2 (Fig. 3E). Overall, these results strongly advocate a role of PKC⑀ as a mediator of TNF␣induced NF-B activation in prostate cancer cells.
PKC⑀ Controls the Expression of NF-B-regulated Genes-NF-B regulates the expression of genes implicated in survival, proliferation, metastasis, and invasion (3,11). To establish the biological significance of the regulation of NF-B transcriptional activity by PKC⑀, we examined the induction of NF-Bresponsive genes using quantitative PCR. Treatment of LNCaP cells with TNF␣ caused a marked elevation in COX2, VEGF, MMP9, and IL6 mRNA levels. When we carried out similar experiments in PKC⑀-depleted LNCaP cells, the induction of all these NF-B-responsive genes was significantly diminished (Fig. 5A). A similar reduction was observed upon treatment with the ⑀V1-2 peptide inhibitor (Fig. 5B). Conversely to PKC⑀ silencing/inhibition, adenoviral overexpression of PKC⑀ in LNCaP cells potentiated the induction of COX2, VEGF, MMP9, and IL6 mRNA levels by TNF␣ (Fig. 5C). Therefore, PKC⑀ mediates the induction of NF-B-responsive genes by this inflammatory cytokine.
PKC⑀ Translocation to the Plasma Membrane Is a Requirement for the Activation of NF-B by TNF␣-PKC translocation to membranes is a hallmark of enzyme activation. PKC isozymes can distinctly relocalize to multiple intracellular compartments, ultimately leading to a differential access to substrates (41). We raised the question whether TNF␣ could induce changes in the subcellular localization of PKC⑀ in LNCaP cells. In the first set of experiments, we examined the redistribution of PKC⑀ using an ultracentrifugation approach (42). Fig. 6A shows that TNF␣ treatment caused the disappearance of PKC⑀ immunoreactivity from the cytosolic (soluble) fraction with a concomitant increase in the particulate fraction, as well established for PMA (31,34).
Translocation of PKC⑀ is mediated by binding of the ligand (either phorbol ester or DAG generated by stimulation of mem- brane receptors) to their C1a and C1b domains located in the regulatory N-terminal region. Cys residues in C1 domains of PKCs are essential for folding and thus required for proper ligand binding and enzyme translocation (41,43). A PKC⑀ C1a-C1b domain double mutant was generated (C204A/C276A-PKC⑀) (Fig. 6B) and expressed in LNCaP cells. Unlike WT-GFP-PKC⑀, the C1 domain PKC⑀ mutant failed to translocate from the cytosolic to the particulate fraction in response to PMA or TNF␣ (Fig. 6C). These findings were further corroborated by real-time microscopy using GFP-fused constructs. As shown in Fig. 6D, a time-dependent translocation of WT-PKC⑀ to the cell periphery was readily detected 5 min after TNF␣ treatment. On the other hand, the C1 domain PKC⑀ mutant failed to translocate in response to TNF␣ (Fig. 6D), even at longer times (data not shown). The C1 domain requirement for translocation by TNF␣ argues for the involvement of DAG in the activation of PKC⑀ in response to the cytokine.
To authenticate the requirement of PKC⑀ peripheral translocation in NF-B activation by TNF␣, a number of constructs to target PKC⑀ to specific intracellular compartments were generated. PKC⑀ was fused at the N terminus to different tags that direct the kinase to plasma membrane, nucleus, cytoplasm, endoplasmic reticulum or mitochondria (31). Remarkably, at similar levels of expression, only myr-PKC⑀, which is targeted to the plasma membrane (31), was capable of inducing IB␣ phosphorylation and degradation, similar to TNF␣ in nontransfected cells (Fig. 6E). Likewise, only myr-PKC⑀ activated NF-B binding activity (Fig. 6F) and induced NF-B luciferase  reporter activity when expressed in LNCaP cells (G), an indication that targeting PKC⑀ to the plasma membrane, is sufficient to promote NF-B activation. Together with the translocation experiments, our data strongly argue that translocation of PKC⑀ to the plasma membrane is required for the activation of NF-B by TNF␣.
PKC⑀ Activation by TNF␣ Is Mediated by PC-PLC-G-protein-coupled receptors activate PKCs through the production of DAG via phosphatidylinositol (PI)-or phosphatidylcholine (PC)-specific PLCs (44,45). To determine which mechanism(s) mediate(s) TNF␣-induced activation of PKC⑀ in prostate cancer cells, we used the PI-specific PLC inhibitor U73122 and the PC-specific PLC inhibitor D609. TNF␣-mediated PKC⑀ membrane translocation in LNCaP cells was significantly attenuated by D609 but not by U73122 (Fig. 7A), suggesting the involvement of PC-PLC. This conclusion was confirmed using realtime microscopy, which revealed that translocation of GFP-PKC⑀ to the plasma membrane is sensitive to D609 but not affected by U73122 (Fig. 7B). These results also authenticate the involvement of DAG in the activation of PKC⑀ by TNF␣.
Next, we wished to determine whether PC-PLC was required for NF-B activation by TNF␣. As shown in Fig. 7C, activation of NF-B binding activity by TNF␣ can be essentially blunted by pretreatment with D609 but not by U73122. Moreover, both the induction of IB phosphorylation by TNF␣ and its potentiation by PKC⑀ overexpression (using a PKC⑀ AdV) were impaired by D609 but not by U73122 (Fig. 7D). Therefore, PC-PLC mediates the activation of the PKC⑀-NF-B axis by TNF␣ in prostate cancer cells.
PKC⑀ Is Implicated in the Formation of the TNFR-I Complex-The mobilization of PKC⑀ to the plasma membrane upon TNF␣ stimulation led us to speculate on its potential association with TNFR-I, the receptor for TNF␣. Interestingly, when we immunoprecipitated TNFR-I from LNCaP cells, we observed that endogenous PKC⑀ coimmunoprecipitates with this receptor (Fig. 8A). The association was enhanced by treatment with TNF␣. A reverse coimmunoprecipitation assay using an anti-PKC⑀ antibody led to a similar conclusion (Fig.  8B).
TNFR-I recruits adaptor proteins TRADD, TRAF2 and RIP upon stimulation to form a signaling complex that promotes the activation of NF-B (12,46). To determine whether PKC⑀ modulates the formation of the complex we examined the association of adaptor proteins to TNFR-I by coimmunoprecipitation. These experiments revealed that in PKC⑀-depleted LNCaP cells the association of the adaptors with the receptor is markedly diminished. Specifically, the association of RIP, TRAF2, and TRADD with TNFR-I was reduced by 41%, 56%, and 39%, respectively, in PKC⑀-depleted cells compared with control cells (Fig. 8C). These results thus suggest a role for PKC⑀ in the formation of the TNFR-I complex.

DISCUSSION
NF-B is an inducible transcription factor activated by a wide array of stimuli, including cytokines, bacterial endotoxins, and cytotoxic stimuli such as chemotherapeutic agents, oxidative stress, and ionizing radiation (1,11). There is mounting evidence that NF-B is constitutively activated in prostate cancer and several other cancer types and that this pathway is a key mediator of inflammatory responses associated with disease initiation and progression (5,8,9). Inhibition of NF-B activity in prostate cancer cell lines drastically reduces their ability to form colonies in soft agar and suppresses both growth and development of metastatic lesions in vivo (47). Moreover, expression of oncogenes or loss of tumor suppressor genes in the mouse prostate leads to the development of invasive prostate carcinoma through NF-B (6,48,49). Despite the growing evidence for a role of NF-B in prostate tumorigenesis and resistance to therapy, the mechanisms underlying the activation of NF-B in prostate cancer remain only partially understood.
The involvement of PKC in regulating NFB has been known for several years (14,16). However, the implication of individual PKC isozymes in the control of this pathway, and in particular in prostate cancer, remained to be elucidated. Studies highlighted the involvement of phorbol ester/DAG unresponsive ("atypical") PKCs in prostate cancer progression (14). For example, atypical PKC promotes prostate cancer growth and transcription of the IL6 gene through an NF-B-dependent pathway (15). PKC also promotes the phosphorylation of IKK in response to TNF␣ and is required for prostate cancer cell survival (50). PKC⑀, a member of the "novel" DAG-responsive PKC isozymes, has emerged as an important player in prostate  cancer progression (27). Genetic deletion of the PKC⑀ gene inhibits the formation of prostate tumors in transgenic adenocarcinoma of mouse prostate (TRAMP) mice (51). Most importantly, PKC⑀ is in the vast majority of human prostate cancer specimens (19,28,29). PKC⑀ emerged as an oncogenic kinase and, when ectopically expressed in fibroblasts, it promotes growth advantage and transformation (52). Similarly, when we overexpressed PKC⑀ in non-transformed RWPE-1 epithelial prostate cells to levels similar to those in prostate cancer cells, we observed a manifest growth advantage with concomitant elevations in phospho-ERK and phospho-Akt (30). Notably, in this study we found that ectopic overexpression of PKC⑀ in RWPE-1 cells potentiates the activation of NF-B and the induction of NF-B-responsive genes by TNF␣. We recently reported that prostate-specific transgenic overexpression of PKC⑀ in mice induces preneoplastic lesions characterized by hyperplasia and PIN. These lesions display elevated phospho-Akt, phospho-S6, and phospho-Stat3 levels and resistance to apoptotic stimuli (30). As shown here, PINs from PB-PKC⑀ transgenic mice also display high nuclear NF-B staining, a hallmark of NF-B pathway activation. We have found recently that PKC⑀ cooperates with Pten deficiency to promote prostate cancer. The resulting adenocarcinomas in prostates from mice overexpressing PKC⑀ and haplodeficient in Pten display remarkable NF-B hyperactivation, even stronger than in PINs 3 . Thus, PKC⑀ drives the activation of signaling pathways implicated in the development and progression of prostate cancer in conjunction with other oncogenic alterations.
This study provides evidence that in prostate cancer cells, PKC⑀ depletion or inhibition diminishes TNF␣-induced IB␣ phosphorylation and degradation, NF-B nuclear translocation, and transactivation potential. Moreover, PKC⑀ is required for the maintenance of constitutive NF-B activation in androgen-independent cell lines such as PC3 and DU145 cells. It is interesting that depletion of other DAG-responsive PKCs, namely PKC␣ or PKC␦, did not significantly affect NF-B transcriptional activity, arguing for a remarkable PKC isozymespecificity for the activation of the pathway in prostate cancer cells. In support of this conclusion, a study in 293T cells revealed that NF-B activating kinase (NAK), an IKK kinase, mediates IKK and NF-B activation in response to growth factors in a manner that is dependent on PKC⑀ but not PKC␣ or PKC (53). Notably, unlike PB-PKC⑀, PB-PKC␣ and PB-PKC␦ mice do not develop preneoplastic lesions (30). Our study also provides evidence that TNF␣ is a bona fide stimulus for PKC⑀ activation. TNF␣ promotes the translocation of PKC⑀ from the cytosol to the plasma membrane in LNCaP cells to activate the NF-B pathway, an effect mediated by DAG generated by PCbut not PI-PLC. Although death receptors do not generally couple directly to DAG generation, PKC activation in response to TNF␣ has been shown in various cellular models (18,54,55). It should be noted that PKC⑀ redistributes to different intracellular compartments in response to distinct stimuli. For example, translocation of PKC⑀ to mitochondria and the endoplasmic reticulum/Golgi has been reported in prostate cancer cells (56,57). The characteristic redistribution of PKC⑀ to the plasma membrane that we observed in prostate cancer cells in response to TNF␣ results in its association with the TNFR-I complex and is consistent with the idea that allosteric activation of PKC⑀ by DAG facilitates its access to substrates and binding partners at the plasma membrane (41,57). Binding of TNF␣ to its receptor in cancer cells triggers the rapid assembly of a TNFR-I-TRADD-TRAF2-RIP complex at the plasma membrane (12,13). This complex triggers a NF-B response but no apoptosis (46). Subsequently, a second cytosolic complex is formed that lacks TNFR-I but includes FADD and procaspases 8 and 10. Apoptosis would be activated by complex II, provided the signal from complex I fails to activate NF-B. Our results show that in the absence of PKC⑀ there is an evident reduction in the association of TNFR-I with the adaptor proteins required for NF-B signaling. PKC⑀ may phosphorylate the TNF␣ receptor in prostate cancer cells to differentially modulate the association of adaptor proteins, as shown previously for PKC␦ in neutrophils (55). Another possibility is that PKC⑀ phosphorylates adaptor proteins to promote the association with the TNF␣ receptor. Alternatively, PKC⑀ may serve as a docking protein for different components of the TNFR-I complex. These mechanisms are currently under investigation in our laboratory.
In summary, our study provides evidence that PKC⑀ plays a key role in constitutive and cytokine-mediated NF-B signaling activation in prostate cancer. PKC⑀ modulates the expression of NF-B-regulated genes relevant for cell survival, angiogenesis, and invasiveness (illustrated in Fig. 9). The identification of PKC⑀ as a crucial upstream regulator of NF-B signaling in prostate cancer argues for a central role of this kinase in the control of pathways involved in prostate cancer development and progression. It is important to highlight that there has been significant interest in the development of PKC⑀ inhibitors as anti-inflammatory and anti-cancer agents (58,59). Selective PKC⑀ inhibitors can be effective candidates for sensitization of cancer cells to chemotherapeutic agents and radiotherapy, as shown previously for NF-B inhibitors. Proof of principle has been recently provided in lung cancer models, where inhibition or depletion of PKC⑀ impairs the tumorigenic and invasive capacity of these cells and "normalizes" the expression of apoptotic and survival genes implicated in disease progression, including well established NF-B-regulated genes (60,61). Thus, targeting the PKC⑀-NF-B pathway may provide novel means for the treatment of prostate cancer or other cancers where this pathway may prove to be relevant.