Atypical protein kinase Ciota plays a critical role in human lung cancer cell growth and tumorigenicity.

Atypical protein kinase C (aPKC) isozymes function in epithelial cell polarity, proliferation, and survival and have been implicated in cellular transformation. However, the role of these enzymes in human cancer is largely unexplored. Here, we report that aPKCiota is highly expressed in human non-small cell lung cancer cell lines, whereas the closely related aPKC isozyme PKCzeta is undetectable in these cells. Disruption of PKCiota signaling reveals that PKCiota is dispensable for adherent growth of non-small cell lung cancer cells but is required for transformed growth in soft agar in vitro and for tumorigenicity in vivo. Molecular dissection of signaling down-stream of PKCiota demonstrates that Rac1 is a critical molecular target for PKCiota-dependent transformation, whereas PKCiota is not necessary for NFkappaB activation in vitro or in vivo. Expression of the PB1 domain of PKCiota (PKCiota-(1-113)) blocks PKCiota-dependent Rac1 activity and inhibits cellular transformation indicating a role for this domain in the transforming activity of PKCiota. Taken together, our data demonstrate that PKCiota is a critical lung cancer gene that activates a Rac1-->Pak-->Mek1,2-->Erk1,2 signaling pathway required for transformed growth. Our data indicate that PKCiota may be an attractive molecular target for mechanism-based therapies for treatment of lung cancer.

Cell Culture, Plasmids, Transfections, and Drug Treatments-Human A549, ChaGo-K-1, H292, H520, H1299, and SK-MES-1 non-small cell lung cancer cell lines as well as the non-transformed HBE4 lung epithelial cell line were obtained from the American Type Culture Collection (Manassas, VA) and maintained as suggested by the supplier. The cells were maintained in a humidified tissue culture incubator at 37°C in 5% CO 2 . A549 cells were stably transfected with recombinant pBabe retroviruses containing FLAG-tagged human full-length wild-type PKC (wtPKC), kinase-dead PKC (kdPKC), or empty vector, as described previously by Lu et al. (22). Expression of FLAG epitopetagged PKC and total PKC was analyzed by immunoblot analysis, as described previously by Murray et al. (17). In some experiments, A549 cells were transiently transfected with either empty LZRS retrovirus or recombinant LZRS retrovirus expressing the PB1 domain of PKC (PKC amino acids 1-113 (PKC-(1-113)) to generate A549/LZRS-and A549/PKC-(1-113) cells, respectively. Expression of PKC-(1-113) was monitored by immunoblot analysis using an antibody directed to the amino-terminal PB1 domain region of aPKCs (Santa Cruz Biotechnology, catalog number 727).
Adherent growth kinetics of A549 cells transfected with empty pBabe, pBabe/kdPKC, and pBabe/wtPKC were determined by plating cells (1 ϫ 10 4 cells/well) into multiwell culture dishes and monitoring cell growth daily over a 7-day period. Each day, cells from triplicate wells were trypsinized and counted using a hemocytometer. In some experiments, A549 cells were maintained in medium containing either 10 or 2% or no fetal bovine serum.
Soft Agar Growth Assays-Anchorage-independent growth was assayed by the ability of cells to form colonies in soft agar. The bottom agar consisted of growth medium containing 10% fetal bovine serum and 0.75% agarose in 60-mm tissue culture dishes. Nine-hundred cells were resuspended in growth medium containing 10% fetal bovine serum and 0.75% agarose and plated on top of the bottom agar. The cells were incubated at 37°C in 5% CO 2 . Cell colonies were visualized and quantified under a dissecting microscope (Olympus) after 4 -6 weeks in culture.
NFB Signaling and Transcriptional Activity Assays-NFB transcriptional activity was assayed using a dual luciferase reporter system (Promega), as described previously (22). In brief, A549 cells stably expressing kdPKC or pBabe vector control were transiently transfected with 500 ng of 3xMHCLuc, a plasmid containing three NFB response elements from the major histocompatibility complex promoter linked to a luciferase reporter gene, and 25 ng of phRL-SV40 using the FuGENE 6 lipofection reagent (Roche Applied Science), as described by the manufacturer. Twenty four hours after transfection, NFB activity was stimulated with 50 ng/ml TNF␣ (R & D Systems) for 2 h. Total cell extracts were prepared for the dual luciferase assay according to the manufacturer's (Promega) instructions. Firefly and Renilla luciferase activity were measured using a Veritas microplate luminometer (Turner BioSystems). The activity of Renilla luciferase was used as an internal control for transfection efficiency.
Tumorigenicity in Nude Mice-The growth of stably infected A549 human lung carcinoma cells as established subcutaneous tumors was studied in athymic nude mice (Harlan-Sprague-Dawley, Indianapolis, IN) in a defined pathogen-free environment. Briefly, A549 cell transfectants were grown in nutrient mixture F-12K containing 10% fetal bovine serum. A549 cell transfectants were harvested and resuspended in serum-containing medium. 4 -6-week-old female nude mice were injected subcutaneously into the flank with 5 ϫ 10 6 cells in 100 l of growth medium. Once palpable tumors were established, tumor size was measured once a week. Tumor growth was quantified by measuring the tumors in three dimensions with calipers. Tumor volume (mm 3 ) was calculated using the formula 0.5236(L ϫ W ϫ H), where L represents the length of the tumor, W represents the width of the tumor, and H represents the height. Animals were individually monitored throughout the experiment. At the conclusion of the study, mice were injected intraperitoneally with 100 g/g 5-bromo-2-deoxyuridine (BrdUrd) 1 h prior to sacrificing the mice by CO 2 asphyxiation. Tumors were excised and divided into sections for protein extraction and tumor fixation. Total tumor extracts were prepared in SDS buffer (2% (w/v) SDS, 4 M urea, 62.5 mM Tris-HCl, pH 6.8, 1 mM EDTA, 5% (v/v) ␤-mercaptoethanol), and equal amounts of protein were subjected to immunoblot analysis. A section of tumor was also fixed in 10% buffered formalin, embedded in paraffin, sectioned (5 m thickness), and stained for appropriate antigens.
Immunoblot Analysis-Cells were harvested by washing with phosphate-buffered saline and scraping off the plate. The cell pellet was lysed in SDS sample buffer. Protein lysates were quantitated using the nitration of tyrosine in nitric acid (57). Equal amounts of protein (ϳ20 g) were loaded for each sample, resolved in 12 or 4 -20% SDS-polyacrylamide gels (Invitrogen), and transferred to polyvinylidene difluoride membrane (Millipore Immobilin-P). A solution of 5% milk and phosphate-buffered saline-Tween 20 was used for blocking. (Tris-buffered saline-Tween 20 was used for phospho-specific antibodies.) Western blot analysis was performed with appropriate antibodies and detected using ECL-Plus (Amersham Biosciences).

PKC Is Required for Human NSCLC Cell Transformation in
Vitro-Little is known about the expression or function of atypical PKC isozymes in human cancers. Because NSCLC is one of the most prevalent and deadliest of human cancers, we assessed the expression of the two atypical PKC isozymes, PKC and PKC, in a panel of human NSCLC cell lines. Each of six human NSCLC cell lines tested (A549, H1299, H292, ChaGoK1, Sk-Mes1, and H520) expressed elevated PKC compared with non-transformed human HBE4 lung epithelial cells (Fig. 1A), whereas none of the cell lines expressed detectable PKC. Thus, PKC is the major aPKC expressed in non-transformed human lung epithelial cells, and PKC is overexpressed in NSCLC cell lines.
We next assessed the role of PKC in the transformed phe-notype exhibited by NSCLC cells. A549 lung adenocarcinoma cells were stably transfected with either wild-type human PKC (wtPKC) or a dominant negative, kinase-deficient PKC mutant (kdPKC) allele (17,23). Immunoblot analysis confirmed expression of recombinant PKC and elevated total PKC in A549 cell transfectants (Fig. 1B). It should be noted that the immunoblot exposure in Fig. 1B was chosen to visualize the expression of the transfected PKC. Interestingly, no significant change in growth rate, saturation density, or survival was observed in any of the cell transfectants grown in adherent culture in 10% serum, 2% serum, or in the absence of serum (Fig. 1C), indicating that PKC signaling is dispensable for adherent growth and survival of A549 cells in culture. These data also demonstrate that the kdPKC allele is not cytotoxic to these cells. Despite having no effect on adherent growth or survival, kdPKC had a dramatic inhibitory effect on the transformed phenotype of A549 cells. A549/pBabe and A549/wtPKC cells form abundant colonies in soft agar, whereas A549/kdPKC cells exhibit significantly impaired anchorage-independent growth (Fig. 1D). Therefore, PKC is necessary for the transformed phenotype of A549 cells. Similar inhibitory effects of kdPKC on soft agar growth were observed in H1299 and ChaGoK lung cancer cells, indicating that these effects are not specific to A549 cells. 2 NFB Is Not Regulated by PKC in A549 Cells-Atypical PKCs regulate the activity of Rac1 and NFB, two critical signaling molecules implicated in oncogenesis (14, 15, 24 -26). Therefore, we assessed the relative importance of NFB and Rac1 in PKCdependent transformation. We first assessed the involvement of NFB in PKC-mediated signaling. NFB protects NSCLC cells from apoptosis through direct transcriptional activation of the anti-apoptotic genes cIAP2 and Bcl-XL (26 -28). Interestingly, neither wtPKC nor kdPKC had an effect on the steady-state levels of cIAP2 or Bcl-XL in A549 cells ( Fig. 2A). Furthermore, neither wtPKC nor kdPKC induced apoptosis in A549 cells as measured by caspase-mediated cleavage of PARP (Fig. 2B) or trypan blue exclusion viability analysis (data not shown). It is well documented that inhibition of NFB transcriptional activity in A549 cells induces apoptosis (26).
We next examined the effect of kdPKC on the ability of A549 cells to activate the NFB pathway in response to TNF␣. Treatment of A549/pBabe and A549/kdPKC cells with 50 g/ml TNF␣ leads to transient phosphorylation of IB␣ at the IK phosphorylation site Ser-32 and a subsequent time-dependent degradation of IB␣ protein (Fig. 2C). The extent and time course of IB␣ phosphorylation and degradation after TNF␣ treatment was identical in A549/pBabe and A549/kdPKC cells. Likewise, TNF␣ induced time-dependent phosphorylation of NFB p65 at Ser-536 to the same extent in A549/pBabe and A549/kdPKC cells (Fig. 2D). Finally, A549 cells exhibited significant basal and TNF␣-stimulated NFB activity that is not affected by expression of kdPKC (Fig. 2E). Taking these data together, we conclude that PKC does not regulate basal or TNF␣-induced NFB signaling in A549 cells.
Rac1 Is a Downstream Target of PKC in NSCLC Cell Transformation-We next assessed the role of Rac1 in PKC-dependent signaling in A549 cells. A549/pBabe cells exhibit significant Rac1 activity, as assessed by the level of GTP-bound Rac1, which is inhibited by the expression of kdPKC but not wtPKC (Fig. 3A). These data strongly indicate that PKC kinase activity is required for Rac1 activation in A549 cells. Treatment of A549 cells with the highly selective, cell-permeant atypical PKC pseudosubstrate peptide inhibitor also blocks Rac1 activity (Fig. 3B), confirming the involvement of PKC activity in Rac1 activation. The PB1 Domain of PKC Is Important for A549 Cell Transformation-PKC couples to Rac1 through interactions involving the PB1 domain of PKC (29). Therefore, we reasoned that the PKC PB1 domain may act as a competitive inhibitor of Rac1 activity in NSCLC cells. Transfection of A549 cells with an LZRS retrovirus expressing the PKC PB1 domain (PKC-(1-113)) inhibits Rac1 activity to levels comparable with those observed after treatment with pseudosubstrate peptide inhibitor (Fig. 3B). Immunoblot analysis confirmed expression of the PKC (1-113) protein in these cells (Fig. 3B). Furthermore, expression of PKC-(1-113) inhibits anchorage-independent growth in soft agar (Fig.  3C), indicating the importance of the PB1 domain in PKC-dependent activation of Rac1 and cellular transformation.
The PKC-Rac1 Signaling Axis Is Required for NSCLC Cell Tumorigenicity in Vivo-We next assessed whether Rac1 is an important downstream effector of PKC-dependent transformation. Expression of a constitutively active Rac1 allele, RacV12, restores transformed growth of A549/kdPKC cells in soft agar (Fig. 4A), indicating that Rac1 is critical for PKC-dependent transformation in vitro. We next assessed the importance of the PKC-Rac1 signaling axis in A549 cell tumorigenicity in vivo. Athymic nude mice were inoculated subcutaneously with A549/ pBabe, A549/kdPKC, or A549/kdPKC/RacV12 cells, and tumor growth was assessed over time. A549/kdPKC cell tumors were significantly growth-inhibited in vivo, whereas A549/kd-PKC cells expressing RacV12 showed tumor growth indistinguishable from A549/pBabe cells (Fig. 4B). Thus, the PKC-Rac1 signaling axis is required for A549-transformed growth in vitro and tumorigenicity in vivo.
To determine the molecular pathway involved in PKC-dependent tumorigenicity, we monitored Rac1 and NFB activity in tumors derived from A549 cell transfectants. Rac1 activity was assessed by monitoring the activity of the downstream Rac1 effector Mek1,2. We recently demonstrated that Mek1,2 is a critical PKC-and Rac1-dependent effector of Ras-mediated transformation of rat intestinal epithelial cells (17). A549/pBabe cell tumors exhibit significant levels of Mek1,2 phosphorylated on the Ser-217/221 Raf activation sites (Fig. 4C) and of active, phosphorylated Erk1,2, indicating Mek/Erk activation (Fig. 4C). It was recently shown that Rac1 targets the Mek/Erk pathway by activating the p21-activated kinase, Pak (30). Activated Pak phosphorylates Mek at Ser-298, thereby facilitating Mek/Erk interactions and Erk activation (30). Interestingly, Mek1,2 is phosphorylated on the Pak-specific phosphorylation site Ser-298 in A549/pBabe tumors (Fig. 4C). A549/kdPKC cell tumors exhibit reduced levels of phospho-Mek at both Raf-and Pak-mediated sites with a concomitant decrease in phospho-Erk levels. Expression of RacV12 in A549/ kdPKC cell tumors restores Pak-and Raf-mediated phosphorylation of Mek and restores Erk phosphorylation to levels indistinguishable from that of A549/pBabe cell tumors. Thus, PKC regulates the Mek/Erk pathway through activation of Rac1 and Pak in A549 cell tumors in vivo.
Although PKC does not regulate NFB in A549 cells in vitro, it is possible that it does so in the in vivo setting. However, expression of the NFB transcriptional targets cIAP2 and Bcl-Xl were not affected by kdPKC or RacV12 in A549 cells in vivo (Fig. 4D). We also found no evidence of apoptosis in tumors expressing kdPKC or RacV12, as measured by caspase-mediated cleavage of PARP (Fig. 4D). Terminal deoxynucleotidyltransferase-mediated dUTP nick-end labeling analysis revealed very low apoptotic indices (Ͻ0.2%) with no significant differences among the three tumor groups (data not shown). We conclude that NFB is not a critical target for PKC-dependent tumorigenicity of A549 cells in vivo.
Protein Kinase C Is Critical for NSCLC Cell Proliferation in Vivo-Our results above indicate that PKC activates Mek/Erk signaling, which has been implicated in tumor cell proliferation. BrdUrd labeling demonstrated that A549/kdPKC tumors exhibit a statistically significant 2.5-3-fold reduction in proliferative index compared with A549/pBabe tumors, which is completely restored by the expression of RacV12 (Fig. 5A), indicating that PKC plays a critical role in A549 cell tumor proliferation. It is possible that the reduced proliferative index observed in A549/kdPKC tumors is the result of reduced tumor vascularization due to decreased angiogenesis. However, immunohistochemical staining for the endothelial cell marker CD31 revealed no change in tumor-associated vessel density in A549 cell tumors in the presence of kdPKC or RacV12 (Fig.  5B). Immunoblot analysis confirmed comparable levels of CD31 protein in A549/pBabe, A549/kdPKC, and A549/kdPKC/ RacV12 tumors (Fig. 5C). We conclude that PKC activates a Rac1/Pak/Mek1,2/Erk1,2 signaling pathway that is necessary for A549 tumor cell growth, although having no demonstrable effect on tumor cell survival or angiogenesis. DISCUSSION The PKC isozymes serve diverse roles in cellular proliferation, differentiation, and survival. Since being identified as the major cellular receptors for the tumor promoting phorbol esters, PKCs have been implicated in carcinogenesis. Indeed, several PKC isozymes have been shown to play critical roles in carcinogenesis in animal models (reviewed in Ref. 31). However, to date, little is known about the role of PKC isozymes in human cancer. Herein, we provide evidence that atypical PKC serves as a critical cancer gene in NSCLC.
aPKCs are thought to be coupled to these diverse cellular functions through direct interactions between aPKC and downstream adaptor molecules mediated through a PB1 domain within the amino-terminal regulatory region of aPKC (29). Two major downstream effectors of aPKC function mediated through PB1 domain interactions that have been identified are the small molecular weight GTPases, Rac1 or cdc42 (29, 40, 41, 43, 46 -48), and the NFB signalosome (38, 39, 49 -51). PKC functions in the establishment of epithelial cell polarity through regulation of Rac1 and/or cdc42, and in cell survival through activation of the canonical NFB pathway. aPKC can activate the NFB pathway by phosphorylating IK␤, which in turn phosphorylates IB␣, targeting it for degradation (52). Loss of IB␣ allows NFB to translocate into the nucleus and activate the transcription of many anti-apoptotic survival genes, including Bcl2 family members (53) and cIAP 1 and 2 (54).
Interestingly, both Rac1 and NFB, similar to aPKC, have been shown to be required for Ras-mediated transformation. However, the question of whether or which of these effectors are involved in PKC-mediated transformation has been unresolved. Our data demonstrate that PKC (but not PKC) is involved in NSCLC transformation. Furthermore, our data provide evidence that the PB1 domain of PKC is important for its ability to support NSCLC transformation. In addition, our data demonstrate that Rac1 is a critical downstream target of PKC-dependent transformation and indicate that the PB1 domain of PKC mediates Rac1 activation. Interestingly, we found no evidence for the involvement of PKC activity in NFB activation of A549 NSCLC cells. Indeed, although basal NFB activity is elevated in these cells and can be further activated by TNF␣, disruption of PKC signaling has no effect on either basal or TNF␣-induced NFB activity. Furthermore, the steady-state levels of the NFB transcriptional targets cIAP and Bcl-Xl, both of which have been shown to be important in A549 cell transformation, are not affected by the disruption of PKC signaling. Our data indicate that, in NSCLC, PKC is tightly coupled to Rac1 signaling but not to NFB activity. Our data are consistent with recent evidence from mouse embryo fibroblasts nullizygous for PKC (the mouse homolog of PKC) that showed that these cells exhibit an intact TNF␣-induced NFB response (55). In contrast, mouse embryo fibroblasts nullizygous for PKC show significant impairment of the NFB pathway in response to TNF␣ and chemotherapeutic agents (56). Taken together with our present data, it appears that, at FIG. 5. PKC is necessary for tumor growth in vivo. A, immunohistochemical staining of A549/pBabe, A549/kdPKC, and A549/kdPKC/ RacV12 cell tumors for BrdUrd. Tumor-bearing animals were injected intraperitoneally with BrdUrd 1 h prior to sacrifice. A labeling index was determined and analyzed for statistical significance, as described under "Experimental Procedures." B, immunohistochemical staining of A549 cell tumors for the endothelial cell marker CD31. C, immunoblot analysis of A549 cell tumors for CD31 and actin. Lanes 1 and 2 indicate extracts from tworepresentative tumors of each genotype. least in fibroblasts and NSCLC cells, PKC does not couple effectively to NFB, whereas PKC does. Interestingly, because A549 cells (or any of the other NSCLC cells we have analyzed) do not express detectable PKC, it appears that NSCLC cells have evolved mechanisms to achieve and sustain high levels of NFB activity independent of aPKC activity. However, we have shown that, in chronic myelogenous leukemia (K562) cells, PKC is required for Bcr-Abl-mediated chemoresistance by activating NFB signaling (22). Therefore, the coupling of PKC to downstream effectors appears to be dependent upon cellular context. It will be of interest to determine whether PKC plays a critical role in the transformed phenotype of other tumor types and, if so, whether Rac1, NFB or perhaps other as yet unidentified downstream targets of PKC are the critical downstream target(s) in diverse tumor types.
In conclusion, our results provide compelling evidence that PKC is required for the transformed phenotype of NSCLC cells both in vitro and in vivo. Interestingly, our data indicate that PKC is dispensable for adherent cell growth, whereas it is required for anchorage-independent transformed growth of NSCLC cells in vitro and tumorigenicity in vivo. These properties make PKC an attractive candidate for the development of mechanism-based therapy against NSCLC.