Induction of cIAP-2 in Human Colon Cancer Cells through PKCδ/NF-κB*

Activation of protein kinase C (PKC) prevents apoptosis in certain cells; however, the mechanisms are largely unknown. Inhibitors of apoptosis (IAP) family members, including NAIP, cIAP-1, cIAP-2, XIAP/hILP, survivin, and BRUCE, block apoptosis by binding and potently inhibiting caspases. Activation of NF-κB contributes to cIAP-2 induction; however, the cellular mechanisms regulating cIAP-2 expression have not been entirely defined. In this study, we examined the role of the PKC and NF-κB pathways in the regulation of cIAP-2 in human colon cancers. We found that cIAP-2 mRNA levels were markedly increased in human colon cancer cells by treatment with the phorbol ester, phorbol-12-myristate-13-acetate (PMA), or bryostatin 1. Inhibitors of the Ca2+-independent, novel PKC isoforms, but not inhibitors of MAPK, PI3-kinase, or PKA, blocked PMA-stimulated cIAP-2 mRNA expression, suggesting a role of PKC in PMA-mediated cIAP-2 induction. Pretreatment with the PKCδ-selective inhibitor rottlerin or transfection with an antisense PKCδ oligonucleotide inhibited PMA-induced cIAP-2 expression, whereas cotransfection with a PKCδ plasmid induced cIAP-2 promoter activity, which, taken together, identifies a role for PKCδ in cIAP-2 induction. Treatment with the proteasome inhibitor, MG132 or inhibitors of NF-κB (e.g. PDTC and gliotoxin), decreased PMA-induced up-regulation of cIAP-2. PMA-induced NF-κB activation was blocked by either GF109203x, MG132, PDTC, or gliotoxin. Moreover, overexpression of PKCδ-induced cIAP-2 promoter activity and increased NF-κB transactivation, suggesting regulation of cIAP-2 expression by a PKCδ/NF-κB pathway. In conclusion, our findings demonstrate a role for a PKC/NF-κB-dependent pathway in the regulation of cIAP-2 expression in human colon cancer cells. These data suggest a novel mechanism for the anti-apoptotic function mediated by the PKCδ/NF-κB/cIAP-2 pathway in certain cancers.

Colorectal cancer is the third leading cause of cancer-related deaths in the United States with ϳ130,000 new cases diagnosed per year (1). Despite advances in surgical management and multimodality therapy, ϳ50% of patients with cancers of the colon and rectum die from their disease (2). A significant obstacle for the successful management of patients with colo-rectal cancer is intrinsic drug resistance or, in patients who initially respond to chemotherapy, acquired drug resistance (3). Alterations in genes that confer a survival and anti-apoptotic advantage and the related signaling transduction pathways involved in the regulation of the cell cycle or in DNA damage repair may result in the resistance of populations of cancer cells to chemotherapy (3).
Programmed cell death, often referred to as apoptosis, is a physiological cell suicide program that is critical for the development and maintenance of healthy tissues. Caspases are synthesized initially as single polypeptide chains representing latent precursors that undergo proteolytic processing at specific aspartic acid residues to produce subunits that form the active heterotetrameric protease (4 -6). Recent studies have shown that inhibitors of apoptosis (IAP) 1 protein family members can bind and potently inhibit the proteolytic activities of caspases 3, 6, and 7, block activation of pro-caspase-9, and thus prevent apoptosis induced by TNF, Fas signaling, etoposide, and growth factor withdrawal (4,(7)(8)(9)(10)(11). To date, six human IAP family proteins have been identified: NAIP, cIAP-1/HIAP-2, cIAP-2/HIAP-1, XIAP/hILP, Survivin, and BRUCE (4). IAPs have been shown to be involved in the resistance of certain tumor cells to chemotherapeutic drugs or other apoptotic agents (12). For example, overexpression of cIAP-2 appears to be responsible for sustained neutrophilia in some cases of chronic neutrophilic leukemia (13). cIAP-2 has been functionally implicated in TNF induction of the ubiquitous transcription factor, NF-B, and protection from apoptosis (10,14).
The protein kinase C (PKC) family, some isoforms of which are stimulated by phorbol esters such as phorbol-12-myristate-13-acetate (PMA), is responsible for transducing many cellular signals during a variety of cellular processes such as mitogenesis, cellular metabolism, differentiation, tumor promotion, and apoptosis (15)(16)(17)(18). For example, in Fas-induced apoptosis, phorbol esters produce a marked attenuation of the cell death pathway (18). Moreover, other cellular models of apoptosis have been used to demonstrate that, during the transduction of cell death signals, there is selective inhibition/activation of PKC isotypes, depending on cell type and apoptotic stimuli (19,20). PMA can protect cells from stimuli-induced apoptosis by a mechanism dependent on the activation of PKC (21,22).
The conventional PKCs ␣, ␤I, ␤II, and ␥ are Ca 2ϩ -dependent and PMA-responsive (23,24). The novel PKCs ␦, ⑀, , and are Ca 2ϩ -independent but PMA-responsive, whereas the atypical PKC isoforms and do not depend on Ca 2ϩ or respond to PMA (24). PKC␦, a member of the novel PKCs, has been associated with apoptosis, cell transformation, growth arrest, and differentiation of various cell types (25)(26)(27)(28). For example, PKC␦ acts as a pro-survival factor in human breast tumor cells (29). Inhibition of PKC␦ with rottlerin or transfection of a kinasedead mutant of PKC␦ increased apoptosis and potentiated chemotherapy-induced apoptosis in non-small cell lung cancer cells (30). In contrast, overexpression of PKC␦ inhibited the proliferation of fibroblasts (31), induced monocytic differentiation of the myeloid progenitor cell line (17), and enhanced enterocyte-like differentiation of colon cancer cell line Caco-2 (28). PKC␦ has been reported to undergo tyrosine phosphorylation in response to various stimuli such as PMA, epidermal growth factor, and PDGF (17,32,33). TNF␣ has been shown to induce NF-B activation through PKC␦ in human neutrophils (34).
Our laboratory is interested in the regulation of proteins and signaling pathways which can contribute to the inhibition of cell death in resistant cancer cells. Previously, we have shown that inhibition of NF-B or PI 3-kinase sensitizes colon cancer cells to TRAIL-induced apoptosis (35,36). The purpose of our present study was to better define the cellular mechanisms regulating the expression of the IAP family members in human colon cancer cells. Here, we show that expression of one of the IAP family member, cIAP-2, is preferentially induced in human colon cancer cells following PMA treatment; inhibition of Ca 2ϩindependent, novel PKC isoforms, but not inhibition of MAPK, PI3-kinase and PKA, blocked PMA-stimulated cIAP-2 mRNA expression, suggesting a role for PKC in PMA-mediated cIAP-2 induction. Further studies utilizing complementary approaches identified PKC␦ as a critical isoform mediating these effects. Finally, a role for NF-B activation, mediated through PKC, acts ultimately to induce cIAP-2 expression. Thus, these results demonstrate that the PKC␦/NF-B pathway plays an important role in the regulation of the anti-apoptosis protein cIAP-2 in human colon cancer cells.

EXPERIMENTAL PROCEDURES
Materials-PMA, wortmannin, and actinomycin D were purchased from Sigma Chemical Company. Bryostatin 1 was from Biomol Research Laboratories Inc. (Plymouth meeting, PA). Bis-indolylmaleimide (GF109203x), PD98059, Gö6983, Gö-6976, rottlerin, H89, MG132, pyrrolidine dithiocarbamate (PDTC), and gliotoxin were from Calbiochem (San Diego, CA). The NF-B oligonucleotide was purchased from Promega (Madison, WI). Anti-cIAP-2 antibody was purchased from R&D Systems (Minneapolis, MN). Antibodies against p50, p65 and PKC isoforms ␣, ␤I, ␤II, ␥, ␦, ⑀, , , , , and myelin basic protein (MBP) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-actin antibody was purchased from Sigma. The rat monoclonal FIG. 1. PMA treatment induces cIAP-2 expression in Caco-2 cells. A, Caco-2 cells were treated with PMA (100 nM) for various times and total RNA extracted for cIAP-2 mRNA expression by RPA using the hAPO-5 multiprobe. The probe set includes cDNAs for L32 and GAPDH to control for RNA loading. B, to determine whether induction of cIAP-2 mRNA by PMA occurs in a dose-dependent manner, Caco-2 cells were treated with various concentrations of PMA for 4 h; RNA was extracted and RPA performed as above. C, cells were treated with 0 or 100 nM PMA and 10 g/ml actinomycin D for 4 h. Total cellular RNA was extracted, and RPA was performed as described above. D, upper panel, Caco-2 cells were treated with 100 nM PMA for various times and whole cell protein extracted for cIAP-2 protein detection by Western blot using anti-cIAP-2 antibody. Lower panel, the same membrane hybridized with anti-␤-actin antibody to indicate relative amounts of protein per lane.
RNA Isolation and RNase Protection Assays-RNA was isolated from cells using Ultraspec RNA according to the manufacturer's protocol. RiboQuant MultiProbe RNase Protection Assay (RPA) System was used for the detection of multiple, specific mRNA species as we have previously described (38). 32 P-labeled antisense RNA probes were prepared using the Human Apoptosis hAPO-5 Template Set and hybridization performed according to the manufacturer's protocol.
Protein Preparation and Western Immunoblot-Western immunoblot analyses were performed as described previously (39). Cells were lysed with TNN buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5 mM Nonidet P-40, 50 mM NaF, 1 mM sodium orthovanadate, 1 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride, and 25 g/ml each of aprotinin, leupeptin, and pepstatin A) at 4°C for 30 min. Lysates were clarified by centrifugation (10,000 ϫ g for 30 min at 4°C) and protein concentrations determined using the method of Bradford (40). Briefly, total protein was resolved on a 10% polyacrylamide gel and transferred to polyvinylidene difluoride membranes. Filters were incubated overnight at 4°C in blotting solution (Tris-buffered saline containing 5% nonfat dried milk and 0.1% Tween 20). Protein expression was detected with antibodies to various PKC isoforms, cIAP-2, HA, or to ␤-actin following blotting with a horseradish peroxidase-conjugated secondary antibody and visualized using ECL detection.
PKC␦ Kinase Assay-PKC␦ activity was determined in cell extracts as described (41). Briefly, total PKC␦ was determined by measuring the incorporation of 32 P into MBP. Extracts from Caco-2 cells treatment with or without PMA were incubated with PKC␦ antibody overnight and with protein A beads for 3 h at 4°C by gentle rocking. Immunocomplexed beads were washed twice with cell lysis buffer and twice with kinase buffer (25 mM Tris, pH 7.4; 2 mM dithiothreitol; 0.1 mM Na 3 VO 4 ; 10 mM MgCl 2 ; and 5 Ci of [␥-32 P]ATP). Immunocomplexes were resuspended in 40 l of kinase buffer supplemented with 10 M ATP and 5 g of MBP and incubated for 30 min at 30°C. Kinase reactions were terminated with SDS sample buffer. Samples were size-fractionated by SDS-PAGE and 32 P-labeled MBP was quantified autoradiographically after fixing and drying the gel.

Preparation of Nuclear Extracts and Electrophoretic Mobility Shift Assays (EMSAs)-
The nuclear extracts were prepared from Caco-2 cells according to the procedure described by Han et al. (42). EMSAs were performed as described previously (43) with minor modifications. Nuclear extracts (10 g) were incubated with 40,000 cpm of 32 P-labeled NF-B consensus oligonucleotide (5Ј-AGTTGAGGGGACTTTCCCAGG-3Ј) and 2 g of poly(dA⅐dT) in a buffer containing 8% glycerol, 100 mM NaCl, 5 mM MgCl 2 , 5 mM dithiothreitol, and 0.1 g/ml phenylmethylsulfonyl fluoride in a final volume of 20 l, for 15 min at room temperature. For supershift studies, 2 l of antiserum was added to the nuclear protein for 20 min at room temperature prior to the addition of labeled probe. The complexes were fractionated on 6% native polyacrylamide gels run in 1ϫ TBE buffer (89 mM Tris, 89 mM boric acid, and 2.0 mM EDTA), dried, and exposed to Kodak X-AR film at Ϫ70°C. Competition binding experiments were performed by the addition of the nonradioactive oligonucleotide, in 100-fold molar excess, at the time of addition of radioactive probe.
Transient Transfections, Luciferase Assays-Caco-2 cells (0.1 ϫ 10 6 / well) were seeded in 6-well plates 24 h prior to transfection. Cells were then transfected with 1 g of pTB701-HA vector or pTB701-HA-PKC␦ and 0.3 g of the cIAP-2 promoter-luciferase construct or pNF-B-luc construct using LipofectAMINE following the supplier's instructions. The pRL-Tk-luc plasmid (0.05 g per well) was co-transfected to normalize for variation in transfection efficiency. After 12 h, the cells were washed with phosphate-buffered saline and then maintained in fresh

cIAP-2 Regulation in Human Colon Cancers
medium for 36 h prior to harvest. In each experiment, the pGL2 plasmid was also transfected in separate wells to compare the specific activity of promoter-reporter constructs with the basic activity of the promoterless plasmid. Luciferase assays were performed as described previously (43). Briefly, 48 h after transfection, the cells were rinsed with phosphate-buffered saline, harvested and lysed with 1ϫ cell culture lysis reagent. Luciferase activity in 20 l of extract was assayed with the dual luciferase assay system. Light emissions were integrated for the initial 10 s of emission by using a Monolight 2010 luminometer (Analytical Luminescence Laboratory). Transient transfection of cells with PKC␦ oligonucleotides was performed using antisense (5Ј-CCCACCAT-GGCGCCGTTC-3Ј) and sense oligonucleotide sequences to the translation start site of PKC␦ (44). Oligonucleotides were dissolved in sterile deionized water to a final concentration of 1 mM, aliquoted, and stored at Ϫ20°C until use. Antisense and sense oligonucleotides to PKC␦ (1 M) were transfected into Caco-2 cells using LipofectAMINE. After transfection, cells were incubated 48 h before treatment.

RESULTS
PMA Treatment Induces cIAP-2 mRNA and Protein Expression in Caco-2 Cells-PKC isoforms are involved in the regulation of certain anti-apoptotic proteins (23,24). The expression of cIAP-2, a member of the IAP family of anti-apoptotic proteins (4), can be induced by NF-B resulting in the protection of certain cells from apoptosis (10,14,37). To better delineate upstream signaling pathways responsible for cIAP-2 induction, the human colon cancer cell line Caco-2 was treated with the phorbol ester PMA, and RNA extracted for RPA using a multiprobe (hAPO-5; BD Pharmingen) containing cDNAs for a number of anti-apoptotic genes as well as the housekeeping genes L32 and GAPDH (Fig. 1).
Treatment with PMA (100 nM) markedly induced cIAP-2 mRNA expression over a time course with maximal induction occurring 4 h after treatment (Fig. 1A). In contrast, xIAP, another IAP family member, was not affected by PMA treatment. The expression of TRAF-4, a member of the tumor necrosis factor (TNF) receptor-associated factor (TRAF) family of putative signal-transducing proteins, and cIAP-1 was slightly increased with PMA. The remainder of the apoptosis-related genes in this multiprobe template were either not expressed in Caco-2 cells or the level was not affected by PMA. The expression of L32 and GAPDH remained constant thus demonstrat-ing equal loading. Caco-2 cells were then treated with varying concentrations of PMA and harvested at 4 h (Fig. 1B). Induction of cIAP-2 was noted with the lowest concentration (5 nM); maximal expression occurred with dosages between 20 and 50 nM. Once again, expression of TRAF-4 was minimally affected and expression of L32 and GAPDH remained constant.
Steady-state mRNA levels may be modulated by transcriptional or post-transcriptional mechanisms. To assess whether the induction of cIAP-2 was due to an increase in transcription, Caco-2 cells were treated with PMA (100 nM) for 4 h in the presence or absence of actinomycin D (10 g/ml), which inhibits transcription by inhibiting DNA-primed RNA polymerase (45), and assessed by RPA (Fig. 1C). As expected, treatment with PMA alone induced cIAP-2 expression; this induction was blocked by actinomycin D suggesting that PMA increases cIAP-2 expression levels by increasing transcription. There was no effect of either PMA or actinomycin D on the expression of TRPM-2 (testosterone repressed prostate message 2), an apoptotic gene, or the housekeeping genes L32 and GAPDH.
We next determined whether the induction of cIAP-2 mRNA levels by PMA also resulted in the corresponding induction of cIAP-2 protein expression (Fig. 1D). Treatment with PMA (100 nM) increased cIAP-2 protein expression. The blot was stripped and reprobed with ␤-actin to ensure equal protein loading. Taken together, these results identify induction of cIAP-2 mRNA and protein levels with PMA treatment.
Regulation of PMA-stimulated cIAP-2 Expression Through the PKC Pathway-To better delineate the signaling pathway leading to PMA-mediated cIAP-2 induction, Caco-2 cells were pretreated for 30 min with inhibitors to PKC (GF109203x), MEK (PD98059), PI 3-kinase (wortmannin), or PKA (H89) prior to addition of PMA (100 nM). Cells were harvested 4 h later and RNA extracted for RPA ( Fig. 2A). As expected, PMA increased cIAP-2 expression (lane 2); this induction was completely blocked by treatment with GF109203x (lane 3). In contrast, treatment with PD98059, wortmannin, or H89 did not block cIAP-2 induction by PMA (lanes 4 -6). Treatment with the inhibitors alone had no effect on cIAP-2 expression (lanes 7-10). Similar to PMA, the non-phorbol ester PKC agonist bryostatin-1, induced cIAP-2 expression; this induction was completely blocked by GF109203x treatment (Fig. 2B). Minimal to no effect on xIAP or TRAF-4 expression was noted with either PMA, bryostatin 1 or the inhibitors. In a similar fashion, L32 and GAPDH expression remained constant indicating equal loading. Bryostatin stimulates various PKC isoforms in certain cell lines (46,47). Treatment with bryostatin induced cIAP-2 mRNA expression and this induction was completely blocked by GF109203x treatment (Fig. 2B). Collectively, these results suggest a role for the PKC pathway, but not MEK/ MAPK, PI 3-kinase, or PKA, in PMA-mediated cIAP-2 induction.
Expression of PKC Isoforms in Caco-2 Cells-To further determine the function of the different PKC isoforms in the regulation of cIAP-2 expression, Caco-2 whole cell extracts were analyzed by Western blot. Caco-2 cells contained proteins immunoreactive with antibodies to PKC␣, ␤I, ␤II (conventional PKCs [cPKC]), PKC␦, , (novel PKCs [nPKC]), and PKC, , (atypical PKC [aPKC]) (Fig. 3A). Expression of PKC␥, ⑀, was not detected. Treatment with PMA (1 M) for 48 h significantly decreased the basal expression of PKC␣, ␤I, ␦, and . In contrast, expression of PKC␤II and was increased whereas expression of PKC and was unchanged.
Chronic treatment of Caco-2 cells with PMA reduces PKC␣, ␤I, and ␦ but not PKC␤II and . To address the potential role of these PKC isoforms in the induction of cIAP-2 by PMA, Caco-2 cells were treated with PMA (1 M) for 24 h followed by addition of PMA (100 nM) for 4 h; RNA was extracted and analyzed by RPA (Fig. 3B). The induction of cIAP-2 was blocked by chronic PMA treatment suggesting that the PKCs, ␣, ␤I, ␦, , may be involved in PMA-induced cIAP-2 expression. It was noted that there was a slight increase of cIAP-2 after chronic treatment with PMA that might come from the undegraded PKC.
Inhibition of PKC␦ Attenuates PMA-mediated cIAP-2 Induction-To further delineate the specific isoforms involved in cIAP-2 regulation, we used Gö6983, which inhibits many of PKC isoforms but not PKC at the concentration used (48). As shown in Fig. 4A, pretreatment with Gö6983 completely abolished PMA-induced cIAP-2 expression, suggesting that the PKC isoenzyme is not involved in PMA-induced cIAP-2 expression.
To determine whether PKC␦ may be involved in the induction of cIAP-2, Caco-2 cells were pretreated with the PKC␦selective inhibitor, rottlerin, or the conventional PKC and PKC inhibitor Gö6976 for 30 min prior to the addition of PMA; cells were harvested 4 h later and RNA analyzed by RPA (Fig. 4B). Pretreatment with rottlerin resulted in a complete inhibition of PMA-mediated cIAP-2 induction in Caco-2 cells cultured in serum-free medium for 24 h prior to treatment (Fig.  4B(i)); partial inhibition of cIAP-2 induction was noted with Gö6976 pretreatment (Fig. 4B(ii)). The fact that cIAP-2 induc-

cIAP-2 Regulation in Human Colon Cancers
tion is partially blocked with Gö6976 suggests involvement of other isoenzymes in addition to PKC␦. Consistently, pretreatment with rottlerin or Gö6983 completely abolished PMA-induced cIAP-2 protein expression (Fig. 4C). Activation of PKC␦ after PMA treatment was also examined using MBP as the substrate. As shown in Fig. 4D, PKC␦ activity was increased after PMA treatment. These results suggest a critical role for the PKC␦ isoenzyme in PMA-dependent cIAP-2 induction.
PKC␦ Overexpression Induces cIAP-2 Promoter Activity Whereas PKC␦ Antisense Oligonucleotide Attenuates cIAP-2 Induction-The inhibition of PMA-induced cIAP-2 mRNA expression by rottlerin suggests that the PKC␦ isoenzyme plays a role in PMA-mediated cIAP-2 induction. To further confirm this observation, Caco-2 cells were transfected with a PKC␦ construct together with a cIAP-2 promoter reporter construct (37). Initially, Caco-2 cells were transfected with the HA-tagged PKC␦ construct (PKC␦) or control vector; PKC␦ protein expression was examined 48 h later by Western blotting using anti-HA or anti-PKC␦ antibody (Fig. 5A). The 80-kDa band corresponding to endogenous PKC␦ and a slower migrating band corresponding to HA-tagged PKC␦, were noted after transfection, thus confirming PKC␦ overexpression. Overexpression of PKC␦ resulted in an induction of cIAP-2 promoter activity (Fig. 5B).
To further confirm the role of PKC␦ in cIAP-2 mRNA induction, Caco-2 cells were transfected with phosphorothioate-modified PKC␦ sense or antisense oligonucleotides, which have been utilized by other investigators to inhibit PKC␦ in other cell models (44), for 48 h followed by PMA treatment for 4 h. As shown in Fig. 5C, transfection with the PKC␦ antisense oligonucleotide suppressed cIAP-2 induction with PMA. Expression of endogenous PKC␦ was decreased thus confirming that the antisense oligonucleotide functions to decrease PKC␦ expression. Taken together, these results, using complementary approaches (i.e. chemical inhibition of PKC␦, overexpression of PKC␦ and transfection with a PKC␦ antisense oligonucleotide) demonstrate a contributory role for PKC␦ in PMA-mediated cIAP-2 induction in Caco-2 cells.
PKC-dependent Induction of cIAP-2 mRNA Expression Occurs in other Human Colon Cancer Cell Lines-We have shown that PKC regulates cIAP-2 expression in the Caco-2 human colon cancer cell line. In order to determine whether this regulation is limited to Caco-2 cells or occurs in other colon cancer cells, KM12C, KM20, HCT-116, DLD1, Lovo, and SW480 cells were incubated in the presence or absence of PMA (100 nM) for 4 h with or without the PKC inhibitor Gö6983 (2 M). Total RNA was extracted and RPA performed (Fig. 6). PMA induced cIAP-2 mRNA expression in KM20, KM12C, and HCT-116 cells (Fig. 6, A-C) as well as DLD1, Lovo and SW480 (data not shown). Pretreatment with Gö6983 blocked cIAP-2 induction in all cell lines. Similar to Caco-2 cells, PMA treatment resulted in minimal alteration of either xIAP or TRAF-4 expression. These results confirm our findings in Caco-2 cells and, moreover, suggest a general regulation of cIAP-2 expression by the PKC pathway in colon cancer cells.
The Expression of cIAP-2 Depends on the Activation of NF-B-The anti-apoptotic function of NF-B is believed to be mediated through the induction of anti-apoptotic genes (49). Among the anti-apoptotic genes, cIAP-2 expression is preferentially up-regulated by stimuli that activate NF-B (4,37). To investigate whether NF-B activation is involved in PMA-induced cIAP-2 induction, we used the proteasome inhibitor MG132 and the PKC inhibitor GF109203x. Caco-2 cells were preincubated 30 min, with or without the inhibitors, followed by the addition of PMA (100 nM) in the presence or absence of inhibitors. cIAP-2 expression was determined by RPA and the NF-B DNA binding activity was assessed by EMSAs using nuclear extracts. Treatment with the proteasome inhibitor MG132 inhibited PMA-mediated induction of cIAP-2 mRNA expression (Fig. 7A). Moreover, PMA-induced NF-B activation and was blocked by either GF109203x or MG132 (lanes 2-4,  Fig. 7B). Supershift analysis demonstrated contribution of both p50 and p65 NF-B subunits following PMA treatment ( lanes  7-8, Fig. 7B). To confirm our results using the MG132 compound, Caco-2 cells were treated with either PDTC, an antioxidant, which functions as an NF-B inhibitor by blocking the dissociation of the NF-B⅐IB complex (50), or gliotoxin, a potent and specific inhibitor of NF-B (51), for 30 min prior to addition of PMA for 4 h. Cells were harvested for either RNA or nuclear protein and analyzed by RPA or EMSA, respectively. Similar to MG132, either PDTC or gliotoxin inhibited PMAmediated cIAP-2 mRNA induction (Fig. 7C) as well as PMAmediated induction of NF-B binding activity (Fig. 7D). Finally, we determined whether PKC␦ directly increases NF-B transactivation in Caco-2 cells by transfection of the NF-Bluciferase plasmid, which contains four tandem copies of the or PKC␦ construct (lane 3). After 48 h, cell lysates were prepared and assessed by immunoblot analysis using antibodies to HA, PKC␦, or actin. B, Caco-2 cells were co-transfected with 0.3 g of a plasmid containing the cIAP-2 promoter fragment linked to the luciferase reporter gene and 1 g of PKC␦ construct or empty vector. After a 48 h incubation, the transfected cells were harvested and luciferase activity measured in the crude cell lysates as described under "Experimental Procedures." All results were normalized for transfection efficiency using the pRL-Tk-luc plasmid (Promega) (*, p Ͻ 0.05 compared with control group using the Student's t test). C, Caco-2 cells were transfected with sense or antisense oligonucleotides (1 M) to PKC␦ as described under "Experimental Procedures" and then treated with 100 nM of PMA for 4 h. Cells were harvested, and the lysates resolved by SDS-PAGE (10% polyacrylamide) and assayed for PKC␦ expression by immunoblotting with anti-PKC␦ antibody. Total RNA was extracted for assessment of cIAP-2 mRNA expression by RPA using the hAPO-5 multiprobe.

cIAP-2 Regulation in Human Colon Cancers
NF-B consensus sequence, along with the plasmid expressing PKC␦. Contransfection with PKC␦ significantly increased NF-B transactivation compared with contransfection of the empty vector (PTB701) (Fig. 7E). Together, these findings strongly support the regulation of cIAP-2 induction with PMA through PKC␦-mediated NF-B activation. DISCUSSION In our current study, we have shown that human colon cancer cells express members of the IAP family, including xIAP, cIAP-2, and cIAP-1. Among these members, we have demonstrated that PMA treatment of human colon cancer cells selectively induces cIAP-2 expression. In addition, inhibition of PKC␦ with the chemical inhibitor rottlerin or the PKC␦ antisense oligonucleotide attenuated cIAP-2 induction, whereas PKC␦ overexpression increased cIAP-2 promoter activity, thus further demonstrating that PKC␦ is involved in the regulation of cIAP-2 expression. Furthermore, PMA-induced cIAP-2 expression and NF-B activation were blocked by either GF109203x (a PKC inhibitor), MG132 (a proteasome inhibitor that blocks NF-B activation) or inhibitors of NF-B. Moreover, inhibition of NF-B activation abolished cIAP-2 induction by PMA, whereas overexpression of PKC␦ resulted in NF-B transactivation. Together, these data suggest a role for a PKC␦/ NF-B-dependent pathway in the regulation of cIAP-2 expression in human colon cancer cells.
The tumor-promoting phorbol ester, PMA, can prevent FasL or TRAIL-induced apoptosis via activation of NF-B (52,53). Moreover, in the absence of NF-B activation, PMA is a strong inducer of apoptosis through stimulation of the upstream caspases-8 and -9 as well as of the effector caspase-3 (52). In the absence of NF-B activation, these survival influences are markedly abrogated revealing the apoptotic effect of PMA. These results suggest that NF-B activation is a critical step in the tumor promoting effect of PMA. Despite these important findings, the signaling pathways involved in PMA regulation of genes that are directly involved in cell death or death resistance remains largely unknown. Our current study has shown that cIAP-2, an anti-apoptosis gene, is regulated though PKCstimulated NF-B activation. Compared with other member of the IAP family, the regulation is specific for cIAP-2, since neither xIAP nor cIAP-1 was significantly induced in PMAtreated human colon cancer cells. On the other hand, the upregulation of cIAP-2 by PMA appears to be a universal response in human colon cancer cells as noted by a marked induction in all of the colon cancer cell lines analyzed.
Specifically, we have demonstrated a role for the PKC␦ isoform in the activation of NF-B and subsequent induction of cIAP-2. Recent studies have suggested that the transcription factor complex PKC␦/NF-B plays a key role in the regulation of various genes, which affect cell survival and proliferation. For example, PKC␦ regulates ICAM-1 expression via NF-B activation in HU-VEC cells (54), and PKC␦ also negatively regulates airway smooth muscle cell cyclin D1 expression in part by activation of NF-B (55). Selective activation of PKC␦ not only increased cIAP-2 promoter activity, but also increased transcription of other NF-B-dependent genes, such as GM-CSF, RANTES, and ICAM-1 promoters (56). Consistent with the activation of NF-B by PKC␦, the promoters of cIAP-2, IL-8, GM-CSF, RANTES, and ICAM-1 each contain NF-B-responsive elements (37,(57)(58)(59)(60). Activation of PKC␦ was associated with the binding of nuclear proteins to the NF-B oligonucleotide binding sequences as well as the trans activation of an NF-B reporter plasmid (56). These results are consistent with our findings demonstrating the presence of p65 RelA and p50 NF-B1 in PMA-induced DNA binding complex and overexpression of PKC␦ significantly increasing NF-B transactivation.
Although PKC␦ has been implicated to act as a prosurvival factor in some tumor cells (29,30), other studies suggest a pro-apoptotic effect of PKC␦, thus demonstrating that the role of PKC␦ is highly dependent on cell type and cellular context. For instance, PKC␦ overexpression inhibits Sindbs virus-induced apoptosis but enhances etoposide-induced apoptosis in the same cell line (61,62). A recent study showed that increased expression of the PKC␦ isoform enhanced the rate of apoptosis in the Caco-2 human colon cancer cell line, which was further augmented by phorbol ester treatment (28). Future studies will be required to elucidate the cellular effect (i.e. proor anti-apoptotic) of PMA-mediated cIAP-2 induction in human colon cancer cells. Moreover, although our results indicate a role for PKC␦ in cIAP-2 regulation, we cannot rule out the possibility that other PKC isoforms, in addition to PKC␦, may be involved in the induction of cIAP-2 since pretreatment with Gö6976, which inhibits conventional PKCs, partially attenuated PMA-mediated cIAP-2 induction. Also, a previous report surveying the specificity of commonly used protein kinase inhibitors showed that 20 M rottlerin had no effect on PKC␦

cIAP-2 Regulation in Human Colon Cancers
activity, but instead inhibited a number of other kinases (63). However, in our study inhibition of PKC␦ expression by transfection with an antisense PKC␦ oligonucleotide decreased cIAP-2 mRNA expression while overexpression of PKC␦ increased transcription from the cIAP-2 promoter, suggesting that PKC␦ plays a critical role in the regulation of cIAP-2 in our system.
In addition to PKC, PMA can stimulate the MAPK/ERK and PI 3-kinase pathways which are associated with induction of cell proliferation and survival (64,65). MAPK and PI 3-kinase have also been reported to regulate the expression of IAP in some cells. MEK blockade results in sensitization to spontaneous and drug-induced apoptosis, and these effects correlate with modulation of the expression of the IAP and Bcl-2 families (66). However, selective inhibition of MEK did not inhibit the action of PMA in our system. PI 3-kinase is a critical regulator of cell proliferation, survival, and differentiation (67,68) and PI 3-kinase regulates the action of PMA (65). Recently, Sade et al. (69) have found that overexpression of constitutively active Akt elevated IAP-2 and blocked dexamethasone-induced apoptosis in a mature T cell line. However, PI3-kinase may not be involved in PMA-mediated cIAP-2 induction in colon cancer cells, since wortmannin, a PI3-kinase inhibitor, had no effect on PMA-induced cIAP-2. Apart from activation of PKC and PI3kinase, PMA is also an activator of PKA (70). The cAMP/PKA pathway is an alternative pathway of transcriptional regulation of IAP-1 expression and may constitute a powerful protective agent in cells exposed to stress conditions (71). Treatment with the selective PKA inhibitor, H89, did not attenuate the PMA-induced cIAP-2 expression, suggesting that PKA is not FIG. 7. PMA-induced cIAP-2 expression acts through NF-B activation. A, Caco-2 cells were preincubated for 30 min with the proteosome inhibitor MG132 (15 M) and then treated with PMA (100 nM) for 4 h in the presence or absence of the inhibitor. Total RNA was isolated for RPA. B, cells were treated with MG132 (15 M) or the PKC inhibitor GF109203x (2 M) and NF-B DNA binding activity was assessed by EMSA using nuclear extracts. In addition, nuclear extracts were incubated with 32 P-labeled NF-B specific DNA probe alone (lane 5) or in the presence of unlabeled wild type (wt) NF-B oligonucleotide (lane 6) or specific antibodies to p50 (lane 7) and p65 (lane 8). C, Caco-2 cells were preincubated for 30 min with PDTC (50 M) or gliotoxin (0.2 M) and then treated with PMA (100 nM) for 4 h in the presence or absence of inhibitors. Total RNA was isolated for RPA. D, cells were treated as described in C, and NF-B DNA binding activity was assessed by EMSA using nuclear extracts. E, Caco-2 cells were co-transfected with 0.3 g of a plasmid containing the consensus NF-B binding site linked to the luciferase reporter gene and 1 g of PKC␦ construct or empty vector. After a 48-h incubation, the transfected cells were harvested and luciferase activity measured in the crude cell lysates as described under "Experimental Procedures." All results were normalized for transfection efficiency using the pRL-Tk-luc plasmid (Promega) (*, p Ͻ 0.05 compared with control group using the Student's t test). involved in the PMA-mediated cIAP-2 induction. Thus, cIAP-2 induction noted in colon cancer cells by PMA does not appear to involve regulation by MEK, PI3-kinase or PKA, suggesting selective activation through PKC.
In conclusion, we demonstrate selective induction in the expression of the anti-apoptotic protein cIAP-2 in human colon cancer cells by the tumor-promoting phorbol ester, PMA. Importantly, we show that the induction by PMA is mediated through a PKC/NF-B pathway, and, specifically, we identify PKC␦ as a critical component for this induction. Together with our recent findings demonstrating that inhibition of NF-B can sensitize resistant colon cancers to chemotherapy-induced apoptosis (35), our current study further delineates potential upstream regulators of NF-B, which may play a role in the enhanced survival of certain colon cancer cells. Collectively, these results add to our current understanding of the signaling pathways contributing to the induction of cell survival and anti-apoptotic proteins in colon cancer cells.