Protein Kinase Cδ Negatively Regulates Hedgehog Signaling by Inhibition of Gli1 Activity*

Constitutive activation of the hedgehog pathway is implicated in the development of many human malignancies; hedgehog targets, PTCH1 and Gli1, are markers of hedgehog signaling activation and are expressed in most hedgehog-associated tumors. Protein kinase Cδ (PKCδ) generally slows proliferation and induces cell cycle arrest of various cell lines. In this study, we show that activated PKCδ (wild-type PKCδ stimulated by phorbol 12-myristate 13-acetate or constitutively active PKCδ) decreased Gli-luciferase reporter activity in NIH/3T3 cells, as well as the endogenous hedgehog-responsive gene PTCH1. In human hepatoma (i.e. Hep3B) cells, wild-type PKCδ and constitutively active PKCδ decreased the expression levels of endogenous Gli1 and PTCH1. In contrast, PKCδ siRNA increased the expression levels of these target genes. Silencing of PKCδ by siRNA rescued the inhibition of cell growth by KAAD-cyclopamine, an antagonist of hedgehog signaling element Smoothened, suggesting that PKCδ acts downstream of Smoothened. The biological relevance of our study is shown in hepatocellular carcinoma where we found that hepatocellular carcinoma with detectable hedgehog signaling had weak or no detectable expression of PKCδ, whereas PKCδ highly expressing tumors had no detectable hedgehog signaling. Our results demonstrate that PKCδ alters hedgehog signaling by inhibition of Gli protein transcriptional activity. Furthermore, our findings suggest that, in certain cancers, PKCδ plays a role as a negative regulator of tumorigenesis by regulating hedgehog signaling.

The Hedgehog (Hh) 3 signal pathway controls a variety of developmental processes such as pattern formation, differentiation, proliferation, and organogenesis (1,2). Hh signaling is restricted in the adult organism where it is implicated in stem cell proliferation and tissue repair (3); persistent signaling or inappropriate reactivation results in cellular hyperproliferation and contributes to the formation and progression of human cancers, including basal cell carcinoma, lung, esophageal and biliary cancer, as well as breast, liver, pancreatic, and prostate cancers (4 -8).
Hh signaling starts with association of the Hh ligand with its receptor Patched (PTCH), which releases PTCH inhibition of Smoothened (Smo), and allows Smo to transduce a signal for the activation and nuclear translocation of a family of transcription factors, cubitus interruptus in Drosophila and Glis (including Gli1, Gli2, and Gli3) in vertebrates, which, in turn, promotes expression of Hh signal target genes. In the absence of Hh ligand, phosphorylation of Gli2/3 targets latent Gli proteins to proteasome-dependent repressor formation (9,10). Compared with Gli2/3, the mechanism of Gli1 regulation is poorly understood. The transcription factors Gli1, Gli2, and Gli3 are critical for the regulation of Hh signaling. Moreover, PTCH1 and Gli1 are transcriptional targets of the Hh signaling pathway expressed in most of these Hh-associated tumors and are used as markers of Hh signaling activation (2,4).
The effect of PKC␦ in cancer cells appears to be related to cell type and context. In some cancers, PKC␦ functions as a proapoptotic factor (27)(28)(29) but appears to have an anti-apoptotic effect in other cancers (30,31). In this study, we found that PKC␦ antagonizes Gli protein transcriptional activity in NIH/ 3T3 cells, and endogenous PKC␦ has a negative effect on Gli activity in human hepatoma cells Hep3B. In addition, we found that PKC␦ expression is markedly decreased or undetectable in hepatocellular cancer (HCC) with active Hh signaling. These results are consistent with the general function of PKC␦ for attenuating proliferation and inducing cell cycle arrest (27)(28)(29)32). Our findings provide the first evidence to suggest the negative regulation of the Hh pathway by PKC␦.
Luciferase Assay-NIH/3T3 cells and Hep3B cells were seeded in 24-well plates at 70% confluence and transfected with the Dual Luciferase Reporter Assay System (Promega) and Gli constructs and different PKC vectors (1-2 g). Luciferase (firefly) and Renilla luciferase activities were determined in lysates of transfected cells with the Dual Luciferase Reporter Assay System as we have previously described (20,23). The relative Gli-luciferase activity was normalized by TK-Renilla activity.
Immunohistochemistry-The status of Hh signaling of HCC tissues has been determined with in situ hybridization (4), and the sections were prepared as described previously (36). Tissue sections were deparaffinized, followed by rehydration with decreased concentrations of ethanol, and immersed in 3% H 2 O 2 for 10 min. Following antigen retrieval in citrate buffer (pH 6.0), the tissue sections were incubated with normal goat serum to block nonspecific antibody binding (20 min at room temperature). The sections were then incubated with primary antibodies (at 1:200 dilution) at 37°C in humid chambers for 2 h. After washing with phosphate-buffered saline three times, the sections were incubated with the biotinylated secondary antibody (goat anti-rabbit IgG) and streptavidin conjugated to horseradish peroxidase for 20 min at 37°C, followed by phosphate-buffered saline wash. The sections were incubated with the diaminiobenzidine substrate for Ͻ30 min. Hematoxylin was used for counterstaining. Negative controls were performed in all cases by omitting the first antibodies.
mRNA Expression Analysis-Total RNA was isolated from treated NIH/3T3 or/and Hep3B cells transfected with different vectors or siRNA with the RNeasy kit and RNase-Free DNase Set (Qiagen, Valencia, CA). An aliquot (1 g) was subjected to reverse transcription with a High Capacity cDNA reverse transcription kit using random primers. One twentieth of the final cDNA was used in each PCR reaction. For mouse PTCH1, cDNA was amplified with CCGTTCAGCTCCGCACAGA (forward primer) and CTCACTCGGGTGGTCCCATAAA (reverse primer), mouse ␤-actin GCTTCTTTGCAGCTCCT-TCGT (forward primer) and CCTTCTGACCCATTCCCACC (reverse primer). For quantitative PCR analyses, we detected transcripts of Gli1 and PTCH1 using Applied Biosystem's assays-by-demand assay mixtures (the sequences for human Gli1 and PTCH1 have been patented by Applied Biosystems) and pre-developed 18 S rRNA (VIC TM -dye labeled probe) Taq-Man assay reagent (P/N 4319413E) was used as an internal control. The levels of Gli1 and PTCH1 were measured in the Real-time PCR Core Facility at the University of Texas Medical Branch.
Protein Preparation, Western Blotting, and Immunoprecipitation-Protein preparation and Western blotting were performed as described previously (17,37). In brief, cells were lysed with 1ϫ cell lysis buffer from Cell Signaling Technology, Inc. (Danvers, MA). Equal amounts of protein were resolved on NuPAGE Bis-Tris gels (Invitrogen) and electrophoretically transferred to polyvinylidene difluoride membranes. Membranes were incubated overnight at 4°C with primary antibodies in TBST buffer with 1% bovine serum albumin, followed with a horseradish peroxidase-labeled secondary antibody for 1 h at room temperature. Membranes were developed using the ECL detection system. Immunoprecipitation was performed as described previously (17). In brief, protein samples (300 g) were incubated with rat anti-HA monoclonal antibody (1 l) and protein A/G beads (20 l) on a shaker at 4°C overnight. Beads were washed three times with lysis buffer, and 4ϫ Trisglycine sample buffer was added. Samples were denatured and analyzed by Western blot.
Immunofluorescence Staining-Hep3B cells were fixed in 4% paraformaldehyde for 20 min at room temperature, permeabilized with 0.2% Triton X-100 cells, incubated in blocking buffer (phosphate-buffered saline with 3% bovine serum albumin) for 15 min, and then with incubated primary antibody for 1 h in blocking solution at room temperature. Samples were then incubated with secondary antibodies for 30 min at room temperature in the blocking solution.
RNA Interference-The SMARTpool PKC␦ siRNA and nontargeting siRNA were obtained from Dharmacon Research in the annealed and purified form. Transfection of siRNA duplexes (40 nM) was carried out using Lipofectamine 2000 following the instructions of the manufacturer. At 72 h after transfection, cells were lysed and subjected to either realtime PCR, immunoblot, or immunofluorescence microscopy analysis.
Cell Proliferation Assay-Hep3B cells were transfected with PKC siRNA or non-targeting siRNA, and 24 h later, replated into 24-wells plates with minimal essential medium containing 2% fetal bovine serum. Either KAAD-cyclopamine or vehicle (DMSO) was added, respectively. Cells were collected by trypsinization daily and counted by using a Coulter counter.
Statistical Analysis-All experiments were repeated at least three times, and data are reported as means Ϯ S.E. For each experiment, real-time PCR values and relative luciferase activity were analyzed using a two-sample t test or analysis of variance for a one-or two-factor experiment according to the structure of factor(s). The factors were vector, drug, and/or dose. Main effects and interactions were assessed at the 0.05 level of significance. Multiple comparisons were conducted using a t statistic with the standard error computed from the residual mean square in the analysis of variance and the comparison-wise error rate with Bonferroni adjustment for the number of comparisons. Statistical computations were carried out using PROC TTEST and PROC GLM in SAS, Release 9.1.

PKC␣ Increases Gli-luciferase Activity in NIH/3T3 Cells-
Neill et al. (48) have shown that PKC␣ is a potent regulator of Gli1 transcriptional activity in 293T cells. To test if the effect of PKC␣ on the regulation of Gli1 is cell type-dependent, we first investigated the effects of PKC␣ on Gli1-luciferase activity in NIH/3T3 cells. NIH/3T3 cells were co-transfected with a wildtype Gli-luciferase reporter or a mutated Gli-luciferase reporter (point mutation that abolishes the binding of Gli), Gli1, and constitutively active PKC␣ AE. PKC␣ AE significantly increased the wild-type Gli-luciferase activity, but not the mutant, indicating the specific regulation of PKC␣ on Gli activity (Fig. 1A, top panel). Western blot analysis was performed in cells overexpressing Gli1 and PKC␣ AE to monitor the status of protein expression (Fig. 1A, bottom panel). Expression of Myctagged Gli1 was not altered by PKC␣ AE compared with the control vector; phosphorylation of overexpressed PKC␣ AE was detected using anti-phospho-PKC␣/␤II (Thr-638/641) antibody to demonstrate the activation of PKC␣. It has been suggested that activation of MEK/ERK signaling is involved downstream of the PKC pathway (30,31); therefore, we probed the membrane with anti-phospho-ERK1/2 (Thr-202/Tyr-204) antibody. ERK1/2 phosphorylation was not altered by PKC␣ AE. To further determine if PKC␣ regulation of Gli1 is mediated by the MEK/ERK pathway, cells were transfected with Gliluciferase reporter, Gli1 and wild-type PKC␣ followed by combination treatment of PMA and PD98059, a specific inhibitor of MEK, for 6 h. The presence of PD98059 completely blocked Gli1 luciferase activity (Fig. 1B, top panel). Results from Western blot shows that PD98059 treatment effectively inhibited ERK1/2 phosphorylation; whereas, either PMA or PD98059 had no effect on Myc-tagged Gli1 protein expression (Fig. 1B,  bottom panel). Our data suggest that PKC␣ plays a positive role in the regulation of Gli1 activity; this effect is mediated by the MEK/ERK pathway.
PKC␦ Down-regulates Gli-luciferase Activity in NIH/3T3 Cells-It has been demonstrated that PKC␦ increased the activity of Gli1 in NIH/3T3 (13) and 293T cells (48). To examine the role of PKC␦ in regulating Gli activity, we co-transfected NIH/ 3T3 cells with Gli luciferase reporter, Gli1, and either wild-type PKC␦ (PKC␦ WT), kinase-dead PKC␦ K376R (PKC␦ KD), constitutively active PKC␦ (PKC␦ ⌬NPS), which is an N-terminal pseudosubstrate domain deletion or empty vector (pcDNA3), and then treated with PMA (200 nM) for 6 h ( Fig. 2A, top panel). Consistent with the previous findings (13, 48), PMA treatment increased Gli-luciferase activity in cells co-transfected with pcDNA3. In contrast, PMA treatment significantly decreased Gli-luciferase activity in cells transfected with PKC␦ WT. In the cells transfected with PKC␦ ⌬NPS, Gli-luciferase activity was further decreased either in the presence or absence of PMA, whereas Gli-luciferase activity was not altered in cells transfected with PKC␦ KD. To further test the specificity of PKC␦-dependent Gli-luciferase activity by PMA treatment, NIH/3T3 cells were co-transfected with Gli luciferase reporter and PKC␦ WT as well as pcDNA3 and treated with various doses of PMA (Fig. 2B, top panel). In the cells transfected with pcDNA3, PMA treatment enhanced Gli1-luciferase activity in a dose-dependent fashion; however, in the presence of PKC␦ WT, PMA treatment decreased Gli-luciferase activity in a dose-dependent manner. The decrease of Gli-luciferase activity in cells transfected with PKC␦ WT, stimulated by PMA, was blocked by rottlerin (Fig. 2C, top panel), further suggesting the effect of PKC␦ on Gli activity. Fig. 2D shows PD98059 decreased Gli-luciferase activity in the cells with PKC␦ WT overexpression either in the absence or presence of PMA. In addition, Western blots were performed (Fig.  2, A-D, bottom panels) showing the effect of PKC␦ on ERK signaling. PMA treatment deleted endogenous phosphorylated and total PKC␦ expression ( Fig. 2A), indicating that the endogenous PKC␦ does not play an important role in PMAstimulated Gli activity in NIH/3T3 cells. PMA treatment did not affect HA-tagged PKC␦ expression in either total or phosphorylation level. Furthermore, ERK1/2 phosphorylation was not altered by overexpression of PKC␦. We also examined the effect of PKC␦ on the expression of the endogenous Gliregulated gene PTCH1. Reverse transcription-PCR (RT-PCR) assays showed that, in cells transfected with pcDNA3 as well as PKC␦ KD, PTCH1 mRNA expression was not altered either with or without PMA treatment. In contrast, PMA treatment decreased PTCH1 mRNA levels in cells transfected with PKC␦ WT in the presence of PMA; PTCH1 mRNA expression was also decreased in cells transfected with PKC␦⌬ NPS either in the presence or absence of PMA (Fig. 2E). Taken together, our data demonstrate that PKC␦ plays a negative role in the regulation of Gli activity stimulated by PMA. Similar results were also observed when Gli2 was used (data not shown).

PKC␦ Affects Hh Pathway Signaling and Gli1 Protein Expression in Human Hepatoma
Hep3B Cells-Our findings imply an antagonistic role for PKC␦ on the activity of Gli, suggesting a role of PKC␦ in regulating tumor cell progression via Hedgehog signaling. Therefore, we next focused on PKC␦ to test whether PKC␦ plays a similar role in Hep3B hepatoma cells, which possess constitutive Hh activity (4). For this purpose, Hep3B cells were co-transfected with Gli luciferase reporter, Gli, and various concentrations of the PKC␦ ⌬NPS; NIH/3T3 cells were used as a positive control (Fig. 3A, top panel). Similar to NIH/ 3T3 cells, overexpression of PKC␦ ⌬NPS significantly decreased Gli-luciferase activity in Hep3B cells in a dose-dependent pattern. To understand how PKC␦ down-regulates Gli1 activity, we tested if overexpression of PKC␦ ⌬NPS affected Myc-tagged Gli1 protein expression by Western blot analysis. The Myc-tagged Gli1 was decreased by the overexpression of PKC␦ ⌬NPS in a dose-dependent fashion in either NIH/3T3 cells (Fig. 3A, bottom panel, left) or Hep3B cells (Fig.  3A, bottom panel, right). Next, we determined whether PKC␦ affects endogenous Hh target genes, PTCH1 and Gli1, in Hep3B cells by real-time RT-PCR. mRNA expression of PTCH1 and Gli1 was decreased in cells transfected with PKC␦ WT and PKC␦ ⌬NPS, whereas PKC␦ KD did not affect the expression of these genes (Fig. 3B, top panel). Furthermore, endogenous Gli1 protein levels in Hep3B cells were decreased by PKC␦ WT and PKC␦ ⌬NPS but not PKC␦ KD as shown by Western blot in Fig.  3B, bottom panel.
To address whether PKC␦ regulates Gli1 through direct protein-protein interaction, we performed co-immunoprecipitation experiments. The results show that both PKC␦ WT and KD associated with Gli1 in cells (Fig. 3C). As a positive control, the interaction between PKC␦ and its upstream kinase PDK1 was readily detected as well. In addition, the time course for Gli1 protein degradation was monitored using pulse-chase experiments. Hep3B cells expressing Gli1 and co-transfected with a control vector or PKC␦ ⌬NPS were treated with cycloheximide, a protein synthesis inhibitor. The rate of Gli1 degradation was similar in the control and PKC␦ ⌬NPS-expressing cells (data not shown), suggesting that PKC␦ is not involved in regulating the protein stability of Gli1 in Hep3B cells. Taken together, our results indicate that overexpression of an active form (either WT or ⌬NPS) but not a kinase-dead mutant of PKC␦ negatively regulates Gli1 expression likely through transcriptional down-regulation of Gli1 mRNA.
Wild-type and Constitutively Active PKC␦ Decreases Nuclear Gli1 Expression-We have shown that PKC␦ down-regulated exogenous and endogenous Gli activity. Modulation of Gli accumulation in the nucleus is a major mechanism for controlling its activity (39). To further understand how PKC␦ downregulates Gli1 activity, we examined the effects of overex- pressed PKC␦ on the distribution of endogenous Gli1 in Hep3B cells by immunofluorescent staining. As shown in Fig. 4, HAtagged PKC␦ WT (green) localized in the cytosol as well as the nucleus, whereas PKC␦ ⌬NPS predominantly localized in the cytosol in an accumulated pattern. Endogenous Gli1 (red) is mainly localized in the nucleus and is only faintly detected in the cytoplasm in non-transfected cells. Cells transfected with PKC␦ WT or PKC␦ ⌬NPS display a significant decrease in the expression levels of endogenous Gli1 (76 -86% and 93-100%, respectively), whereas Gli1 expression levels are not altered in cells transfected with PKC␦ KD. The PKC␦ WT and the PKC␦⌬ NPS have different localization patterns to nucleus but similar effects in the regulation of Gli1 expression, suggesting the cytosolic PKC␦ plays more important role in the regulation of nuclear Gli1 expression. PKC␦ may affect the shuttle of Gli1 between the nucleus and cytosol. This result further suggests that PKC␦ affects Hh signaling by regulating Gli1 protein expression.
PKC␦ Suppression Increases Hh Pathway Signaling-To further confirm that PKC␦ plays a negative role in the regulation of Gli1 activity in Hep3B cells, endogenous PKC␦ was knocked down by PKC␦ siRNA. mRNA levels of PTCH1 and Gli1 were increased in cells transfected with SMARTpool PKC␦ siRNA (Fig. 5A, top panel). Furthermore, PKC␦ siRNA increased Gli1 protein levels compared with NTC siRNA (Fig. 5A, bottom  panel). We also determined if the regulation of Hh signaling by PKC␦ affects cell function, such as cellular proliferation. Treatment with the Smo antagonist KAAD-cyclopamine (2 M) alone decreased Hep3B cell proliferation. In contrast, PKC␦ knockdown with siRNA enhanced the proliferation, and transfection with PKC␦ siRNA significantly blocked the inhibitory effects of KAAD-cyclopamine (Fig. 5B, left panel). The PKC␦ knockdown was monitored by Western blot as shown in Fig. 5B, right panel. The results were confirmed with two individual PKC␦ siRNA (Fig. 5C). Taken together, the loss of PKC␦ increases Hh signaling and Gli1 protein expression and rescues the inhibitory effect of KAAD-cyclopamine on cellular proliferation, demonstrating that PKC␦ negatively regulates Hh sig-naling downstream of Smo. However, we cannot exclude the possibility that PKC␦ indirectly regulates Gli1 by inhibiting Smo.
Reduced PKC␦ Expression Is Associated with Activation of Hh Signaling in HCC Tissue-To determine whether PKC␦ plays a role in regulating Hh signaling in human cancer, we screened a set of HCC clinical specimens for PKC␦ expression. The activation status of Hh signaling in this set of samples has previously been determined by in situ hybridization using probes against Gli1 and PTCH1 (4). The samples with positive staining of these Hh target genes are considered to have activated Hh pathway. The level of PKC␦ expression in this set of HCC specimens was detected using immunohistochemical staining. Among the 11 total cancer samples screened, 6 samples are known to have activated Hh signaling. Interestingly, the expression of PKC␦ was not detected in any of these specimens with activated Hh signaling (three representative cases are shown in Fig. 6, A-C). Nuclear staining with PKC␦ antibody was only observed in two samples screened, and both of these had no detectable Hh signaling ( Fig. 6D and Table 1). The remaining three specimens with undetectable Hh signaling did not demonstrate expression of PKC␦ (data not shown). These results suggest that decreased expression of PKC␦ may account for activation of Hh signaling in certain HCC specimens, underscoring the importance of PKC␦ mediated negative regulation in suppressing the oncogenic Hh signaling.

DISCUSSION
The Hh signaling pathway, acting through transcriptional factors such as Gli1, has been identified as critical for the initiation and growth of a number of cancers (4 -8). PKC␣ generally stimulates cell growth, but PKC␦ slows proliferation and induces cell cycle arrest of various cell lines (32,40,41). The data from reporter assay reveal that PKC␣ and PKC␦ have different effects on Gli activity. Activation of PKC␣ (by PMA or the constitutively active mutant) increased Gli-luciferase activity in NIH/3T3 cells. Activation of PKC␦ decreased Gli-luciferase activity and down-regulated the mRNA levels of the Hh target gene. In Hep3B cells, wild-type PKC␦ and constitutively active PKC␦ decreased the expression levels of the endogenous genes Gli1 and PTCH1, whereas PKC␦ siRNA increased the expression of these Hh target genes.
Using a different PKC␦ plasmid (pCO2-PKC␦-cat) and Gli reporter plasmid (pGL3P-6GB), Neill et al. (48) reported that constitutively active PKC␦ appeared to increase Gli1 transcriptional activity in HEK293 cells. We found PMA increased the activity of overexpressed Gli1 in NIH/3T3 cells as previously reported by Riobo et al. (13). Based on the finding that rottlerin prevented the stimulation of endogenous Hh signal by PMA in LIGHT2 cells, these investigators concluded that PKC␦ plays a positive role in Hh signaling. We found that rottlerin decreased overexpressed Gli activity, but in combination with overexpression of wild-type PKC␦, rottlerin partially rescued the decrease of Gli-luciferase activity by PMA. Although rottlerin was initially described as a selective PKC␦ inhibitor (32), subsequent studies showed that it is an inappropriate and ineffective inhibitor of PKC␦, and can uncouple mitochondrial respiration from oxidative phosphorylation and exert an inhibitory effect in PKC␦ Ϫ/Ϫ cells (32, 49 -51). The nonspecific effect of rottlerin may play a role in blocking Hh signaling stimulated by PMA. In addition, long term PMA treatment will deplete all of the conventional and novel PKC isoforms (13,52). Thus, it is difficult to form conclusions regarding the function of PKC␦ in the regulation of Gli activity by either rottlerin or PMA.
Phosphorylation of Gli2/3 by protein kinase A, GSK3 and casein kinase-I targets latent Gli proteins to proteasome-dependent repressor formation (2). As an Hh signaling target gene, Gli1 is regulated mainly at the transcriptional level (2). Recent reports identify protein kinase A regulation of Gli1 localization (14), and that ␤-Trcp and Numb regulate Gli1 degradation (39,53). We found that PKC␦ significantly decreased the activity of overexpressed Gli1, indicating that the function of PKC␦ is rate limiting for Gli1 function. In an attempt to understand the mechanism by which Gli1 is down-regulated by active forms of PKC␦, we performed immunoprecipitation studies and found that Gli1 interacts with PKC␦ in Hep3B cells. However, the binding does not depend on the kinase activity of PKC␦, because both the WT and KD PKC␦ co-immunoprecipi- tated with Gli1. Given the fact that overexpression of PKC␦ has no effect on the protein stability of Gli1 (data not shown), it is likely that PKC␦ induces phosphorylation of Gli1 specifically or possibly other regulatory proteins resulting in a decrease of the transactivation activity of Gli1. Thus, the down-regulation of Gli1 protein, as observed with overexpression of PKC␦, can be attributed to the decrease of Gli1 mRNA levels. In addition, PKC␦ may phosphorylate other unknown proteins of Hh signaling, which then regulate the activity of Gli. Future studies will further elucidate these potential mechanisms.
Riobo et al. (13) and Kasper et al. (16) showed that MEK/ERK signaling plays a positive role in the regulation of overexpressed Gli activity. PKC␦ activates the MEK/ERK pathway (30,31), thus it would be reasonable to conclude that PKC␦ should increase Gli activity. However, with overexpression and knockdown of PKC␦, we found that PKC␦ decreased Gli-luciferase activity and mRNA levels of Hh target genes. Moreover, the inhibition of Gli-luciferase activity by PKC␦ could not be rescued by the MEK/ERK pathway inhibitor PD98059; overex-pression or knockdown of PKC␦ did not affect MEK/ERK signaling in NIH/3T3 and Hep3B cells. Therefore, our findings, using complementary approaches, suggest that PKC␦ plays a negative role in the regulation of Gli activity, which is independent of the MEK/ERK pathway. Consistent with a previous report (13), our data indicate that activation of PKC by PMA increased Gli-luciferase activity through MEK/ERK signaling. Individual PKC isoforms have specific functions (42)(43)(44). For example, PKC␣ and PKC␤ stimulate growth, whereas PKC␦ generally slows proliferation and induces cell cycle arrest (32,40,41). Our data are consistent with the anti-apoptotic function of PKC␣ and the inhibitory function of PKC␦ in cell proliferation. The effects of PMA depend on the specific intracellular isoforms of PKC (32). Recently, Lasfer et al. (45) reported that PKC␦ is required for apoptosis induction in Hep3B cells, and Huang et al. (4) found that constitutive Hh activity is important in these cells. PMA treatment induced a marked translocation of PKC␣, PKC␦, and PKC⑀ in Hep3B cells (46). Our finding, that the negative regulation of Hh signaling by PKC␦, is in agreement with a role for PKC␦ in apoptosis induction of certain cancer cells.
We demonstrate that PKC␦ was weakly expressed in all of the HCC sections with detectable Hh signaling, and high expression of PKC␦ was found in only two HCCs, both of them had no detectable Hh signaling. It has been reported that nuclear retention of full-length PKC␦ can be cleaved to form the catalytic fragment (47). In this study, we found nuclear staining of PKC␦ in HCCs without detectable Hh signaling, suggesting PKC␦ may directly regulate Gli in the nucleus and play a role in regulating Hh signaling in hepatocarcinogenesis. However, this does not appear to be the only mechanism by which Hh signaling activation is regulated in HCC. Three HCC specimens with no detectable Hh signaling also had no detectable expression of PKC␦, suggesting down-regulation of PKC␦ is not sufficient to activate Hh signaling. A large cohort study on the correlation between Hh signaling and PKC␦ will be useful to show the significance contribution of PKC␦ to Hh signaling activation in HCCs. Nevertheless, our data from limited tumor specimens are in favor of our hypothesis that PKC␦ negatively regulates Hh signaling in HCC. Thus, elevated expression of PKC␦ may cause reduced Hh signaling, whereas loss of PKC␦ may contribute to Hh signaling activation.
In summary, we have shown that PKC␦ down-regulates Hh signaling by inhibiting the activity of the transcription factor FIGURE 6. Immunohistochemical analysis of PKC␦ expression in HCC sections. Immunohistochemical detection of PKC␦ was performed with specific antibodies from Abcam (cat# ab47473). A total of 11 HCC specimens with known Hh signaling status (4) were used in this study (see Table 1 for details). We found that all tumors with active Hh signaling demonstrated no or weak PKC␦ expression (arrow in A and B; A is from specimen #2, B is from specimen #11). Positive staining of PKC␦ was noted in the stromal compartment in one specimen (arrowhead in C), although the tumor was negative for PKC␦ expression (C is from specimen #7). Two specimens with positive PKC␦ staining (arrow in D) were negative for Hh signaling activation (see Table 1 for details; D is from specimen #6).
Gli in NIH/3T3 cells and Hep3B cells. In fact, the balance between PKC␣ and PKC␦ is important in the regulation of Gli activity by PMA. When PKC␣ is dominant, the negative effect of PKC␦ is small, and PMA increases Gli activity through the PKC␣/MEK/ERK pathway. However, when PKC␦ is dominant, PMA treatment decreases Gli activity through the activation of PKC␦. Our study provides a better understanding of the complex cross-talk between the Hh and the PKC pathways. Studies detailing the molecular mechanism for PKC␦ regulation of Gli activity will provide a basis for the design of more efficient targeted therapies for Hh responsive cancers.