Non-Small Cell Lung Carcinoma Cell Motility, Rac Activation and Metastatic Dissemination Are Mediated by Protein Kinase C Epsilon

Background Protein kinase C (PKC) ε, a key signaling transducer implicated in mitogenesis, survival, and cancer progression, is overexpressed in human primary non-small cell lung cancer (NSCLC). The role of PKCε in lung cancer metastasis has not yet been established. Principal Findings Here we show that RNAi-mediated knockdown of PKCε in H358, H1299, H322, and A549 NSCLC impairs activation of the small GTPase Rac1 in response to phorbol 12-myristate 13-acetate (PMA), serum, or epidermal growth factor (EGF). PKCε depletion markedly impaired the ability of NSCLC cells to form membrane ruffles and migrate. Similar results were observed by pharmacological inhibition of PKCε with εV1-2, a specific PKCε inhibitor. PKCε was also required for invasiveness of NSCLC cells and modulated the secretion of extracellular matrix proteases and protease inhibitors. Finally, we found that PKCε-depleted NSCLC cells fail to disseminate to lungs in a mouse model of metastasis. Conclusions Our results implicate PKCε as a key mediator of Rac signaling and motility of lung cancer cells, highlighting its potential as a therapeutic target.


Introduction
Lung cancer is one of the major causes of cancer-related mortality, with more than 220,000 cases diagnosed and 157,000 deaths estimated for 2010 [1,2]. The genesis of lung cancer involves the progressive appearance of genetic and epigenetic alterations both in oncogenes and tumor-suppressor genes, ultimately leading to deregulated activation of mitogenic and survival signaling pathways. Common alterations in non-small cell lung cancer (NSCLC), the most prevalent form of lung cancer, include mutations in K-Ras, overexpression of epidermal growth factor receptor (EGFR) and Bcl2, as well as inactivation/down-regulation of p53, Rb, and Pten tumor suppressor genes [3]. Accumulating evidence indicates that protein kinase C (PKC) expression and/or activity is considerably altered in human lung cancer [4], although a causal relationship with disease progression remains to be established. The PKC family consists of three classes of serine-threonine kinases with distinct biochemical and regulatory properties: classical/conventional (cPKCs a, b, and c), novel (nPKCs d, e, g, and h), and atypical (aPKCs f and i). cPKCs and nPKCs are responsive to phorbol esters and the second messenger diacylglycerol (DAG), a product of PIP2 breakdown by phospholipase C (PLC) in response to activation of tyrosine-kinase and G-protein-coupled receptors. PKC isozymes have paradoxical functional roles depending on the cellular context, ranging from apoptosis to survival, and from mitogenesis to cell cycle inhibition [4,5].
Despite our extensive knowledge on PKC in cancer development, there is surprisingly limited information regarding the role of individual PKC isozymes in lung cancer progression compared to other epithelial cancers such as skin, breast or prostate cancer, where the involvement of PKCs has been extensively documented [6]. A few available studies established that PKC isozymes display unique functional properties in lung cancer cells. Our laboratory reported that NSCLC cells undergo arrest upon activation of either PKCa or PKCd. Whereas PKCa induces G2/M arrest and senescence in NSCLC cells, PKCd induces p21 cip1 in the G1 phase of the cell cycle [7,8,9]. PKCe, on the other hand, drives G1 to S progression of the cell cycle and has a pro-survival role in NSCLC cells, as also described for other cancer cell models [10,11,12,13,14]. Remarkably, PKCe is abnormally up-regulated in .90% of primary human NSCLCs compared to normal lung epithelium, an effect not observed in small cell lung cancer (SCLC) [10]. It is noteworthy that PKCe has been implicated in oncogenesis in various cancer types. Most notably, ectopic overexpression of PKCe in non-transformed cell lines confers growth advantage or can even lead to malignant transformation [14,15,16].
While a large body of evidence links PKC to the promotion of cancer cell invasion and metastatic dissemination, the specific roles of individual PKCs remain only partially understood. [17,18,19,20,21]. Most notably, phorbol esters promote actin cytoskeleton reorganization and stimulate cell motility in a number of cell lines [22,23]. Cell type-specific associations between PKC signaling and small Rho GTPases family have been reported. Rac1, a member of the Rho family, has been widely implicated in cytoskeleton rearrangements and cell migration, and play important roles in tumorigenesis and metastasis [24]. PKC has been also implicated in the secretion and expression of basement membrane remodeling proteins, such as matrix metalloproteinases (MMPs) and urokinase-type plasminogen activator receptor (uPAR) [17,18,25]. The relationship between individual PKCs and cell motility/invasiveness in lung cancer cells remains to be elucidated.
In this study we focused on PKCe and its potential role in the control of Rac activation and NSCLC cell motility. By means of RNAi and pharmacological approaches we established a key role for PKCe in Rac activation in response to growth factor receptor stimulation, as well as in migration and invasiveness of NSCLC cells. Our studies may have significant implications for the understanding of the molecular mechanisms underlying lung cancer cell metastasis and hopefully underscore novel therapeutic opportunities for treatment of the disease.

Cell culture
Human NSCLC cell lines (H358, H322, H1299 and A549) were obtained from ATCC and cultured in RPMI 1640 medium supplemented with 10% FBS, 2 mM glutamine, 100 U/ml penicillin, and 100 mg/ml streptomycin. Immortalized human bronchioepithelial cells (HBEC) were a kind gift from Dr. Trevor Penning (University of Pennsylvania) and were cultured on gelatincoated plates in KSFM medium supplemented with 0.05 mg/ml bovine pituitary extract and 5 ng/ml EGF.

RNAi
Two different RNAi sequences were used for each target. For transient depletion of PKCe we used ON-TARGET Plus RNAi duplexes purchased from Dharmacon (Catalog # J-004653-06, J-004653-08). For EGFR depletion we used ON-TARGET plus SMARTpool (L-003114-00-0005). Silencer Control RNAi #3 (Ambion) was used as control. siRNAs were transfected with LipofectamineRNAi/MAX (Invitrogen). After 48 h cells were serum starved for 24 h and used for the indicated experiments. For stable depletion of PKCe, we used shRNA lentiviral particles from Sigma (Catalog #SHCLNV-NM_005400, clone ID x-741s1c1 and x-375s1c1). Upon infection, cells were selected for 2 weeks with puromycin (1-2 mg/ml depending on the cell line), as previously described [26].

Rac-GTP pull-down assays
Determination of Rac-GTP levels was carried out essentially as previously described [27]. Briefly, cells growing at low confluency were serum starved for 24 h and stimulated with PMA, FBS or EGF at the concentrations indicated. Cells were lysed in pull-down buffer (20 mM Tris-ClH pH 7.4, 150 mM NaCl, 5 mM MgCl2, 0.5% NP40, 5 mM b-glycerophosphate, 1 mM DTT and protease inhibitors) containing 15 mg/ml of PBD-GST. Lysates were cleared by centrifugation (10 min at 4uC, 13,0006g) and incubated with glutathione-Sepharose beads (GE Healthcare) for 45 min at 4uC. After centrifugation, Rac-GTP bound to the beads was washed twice with pull-down buffer and run on SDS-PAGE. Rac1 was detected by Western blot using an anti-Rac1 antibody (Upstate Biotechnology, 1:1000 dilution).

Cytoskeleton morphology assays
Cytoskeleton morphology was assessed as described previously [26]. Briefly, cells growing on glass coverslides at low confluency were serum starved for 24 h and stimulated with PMA or FBS at the concentrations indicated. Following fixation with 4% formaldehyde, F-actin was stained with phalloidin-rhodamine and nuclei were counter-stained with DAPI. Slides were visualized by fluorescence microscopy and 5 random fields were scored for presence/absence of ruffles.

Wound assays
Experiments were carried out in triplicate plates essentially as previously described [26].Three micrographs were obtained at time = 0 h and 9-24 h later depending on the cell line. Wound closure measurements were normalized to the maximum initial scratch area for each well.

Migration in Boyden chambers
Cells were serum starved for 24 h, trypsinized, suspended in 0.1% BSA/RPMI, and seeded (1.5610 4 cells/well) in the upper compartments of the chemotaxis chamber (NeuroProbe). Polycarbonate membranes of 8 mM pore diameter were used to separate the upper and lower compartments. In the lower chamber, RPMI medium containing 10% FBS was used. After an incubation of 18 h at 37uC, membranes were recovered and cells on the upper side (non-migratory) wiped off the surface. Cells on the lower side of the membrane were fixed and stained with DIFF Quik Stain Set (Dade Behring). Migratory cells in each well were counted by contrast microscopy in 5 independent fields.

Invasion assays and determination of proteases
Invasion assays using Matrigel-coated polycarbonate membranes were conducted as previously described [5]. The mRNA levels of extracellular matrix (ECM) proteases and protease inhibitors were determined by qPCR as described previously [28]. Briefly, RNA was isolated with the QIAGEN RNeasy kit, reverse transcribed, and assayed for simultaneous detection of ECM-related genes and 2 housekeeping genes by qPCR using RT2 Profiler PCR Array plates (QIAGEN) and a RT2 SYBR Green/5-carboxy-X-rhodamine (ROX) qPCR master mix. Data were normalized using GAPDH and b-actin as housekeeping genes, and the relative levels of mRNA calculated according to the DCt method. Genes that displayed .1.5-fold change in expression as a consequence of PKCe depletion were selected for further analysis.
For the determination of proteases in conditioned medium (CM), cells (3610 6 ) were incubated for 48 h in serum-free medium and the supernatant concentrated using an Amicon Ultra-4 filtration device with a 30 kDa cut-off (Millipore). The concentrated CM was used for detection of proteases with a RayBiotech antibody array following the manufacturer's instructions, or for gelatin zymography, as described previously [16].

Cell adhesion
Plates were coated with 500 mg/ml Matrigel, blocked with 0.5% BSA and cells (3610 4 cells/well in 0.1% BSA/RPMI) were seeded in triplicate. FBS (10%) was added to stimulate adhesion and after incubation for 18 h at 37uC unattached cells were washed away with PBS, fixed in methanol and stained with crystal violet solution (3.5 mg/ml crystal violet, 19% ethanol, 1% methanol). After solubilization in 2% SDS, optical density was measured at 570 nm.

Experimental metastasis assays and survival in nude mice
Studies were carried out in strict accordance to the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of Pennsylvania (Protocol number 802268) and The Wistar Institute (Protocol number 112318). Cells were grown to 80% confluency and suspended in PBS. Male athymic mice (6-8 weeks, Harlan Laboratories) were injected i.v. with 5610 5 cells in 0.1 ml (10 mice/group). After 100 days, mice were sacrificed and the lungs removed, fixed in 4% formaldehyde/PBS, paraffin embedded, dissected and stained with hematoxilin/eosin (H&E). The incidence of metastasis was determined as described [29,30].
To determine the area of metastatic lesions, digital images of 2 sections from the 5 lobules from each lung were processed with ImagePro 6.2, as described previously [31]. For survival studies, nude mice (10 mice/group) were injected with 2610 6 cells and the time of death of each animal was recorded.

Western blots
Western blots and densitometric analyses were carried out as previously described [9,27]. Bands were visualized by Enhanced Chemiluminescence (ECL) Western blotting detection system. Images were captured using a FUJIFILM LAS-3000 system and the LAS-2000 software.

Statistical analysis
Averages of at least 3 independent experiments were compared with a t-test using GraphPad Prism 3.0. For metastasis experiments, the distribution of each group was compared against the distribution of the control, using a Mann-Withney non-parametric test. In all cases, a p,0.05 was considered statistically significant.

PKCe mediates PMA-induced Rac activation and Racmediated responses
To determine if PKCs play a role in the regulation of Rac activity in NSCLC cells, we first examined the effect of phorbol 12-myristate 13-acetate (PMA) on Rac activation in four different NSCLC cell lines (A549, H358, H1299, and H322). Cells were treated with 100 nM PMA for different times (0-30 min) and the levels of Rac1-GTP (active Rac) assessed using a PBD pull-down assay. As shown in Fig. 1A, PMA induced a fast activation of Rac1 in all four NSCLC cell lines that begins at 1-2 min and is sustained for at least 30 min. As PMA can directly activate targets unrelated to PKC that modulate the activity of small GTPases (such as chimaerins, MRCK or RasGRPs), we examined the effect of the pan-PKC inhibitor GF19203X (bisindolmaleimide I) [32]. Rac1 activation by PMA was indeed sensitive to the pan-PKC inhibitor GF109203X, particularly in A549, H358, and H322 cells, suggesting that the effect of the phorbol ester is mediated primarily through PKC activation (Fig. 1B).
Rac mediates the reorganization of the actin cytoskeleton into ruffles and lamellipodia, a critical step for cell motility [33,34]. When A549 cells were treated with PMA, a marked induction of circular ruffles was observed. This effect was essentially abolished by GF109203X (Fig. 1C). Similar results were observed in H358, H1299 and H322 cells (Fig. S1). To determine if PMA activates NSCLC cell motility we used a wound assay. In comparison to vehicle treatment, which closed the wound by 50-70% depending on the NSCLC cell line, full closure of the wound was observed in response to PMA (Fig. S2). PMA-induced NSCLC cell motility was markedly reduced by the PKC inhibitor GF109203X ( Fig. 1D and Fig S2). These results argue for the involvement of phorbol ester/DAG responsive PKCs, namely cPKCs and/or nPKCs, in Rac activation in NSCLC cells.
NSCLC cells express very high levels of PKCe compared to normal lung epithelial cells (Fig. S3), as also observed in human NSCLC specimens [10]. To determine if PKCe is implicated in the regulation of Rac activation and cell motility in NSCLC cells we knocked-down PKCe using RNAi. Two different siRNA duplexes (e1 and e2) were used, which led to .80% depletion in all NSCLC cell lines used ( Fig. 2A). Notably, activation of Rac1 by PMA in PKCe-depleted NSCLC cells was markedly reduced relative to their corresponding control cells transfected with a non-target RNAi sequence (

PKCe is required for growth factor-induced Rac activation and NSCLC cell motility
To assess the involvement of PKCs in Rac activation in NSCLC cells in response to physiological stimuli, we first examined the effect of the PKC inhibitor GF109203X on serum responses. A549, H358 and H322 cells were serum-starved for 24 h and Rac-GTP levels determined in response to acute stimulation with serum. Addition of FBS activated Rac1 in all NSCLC cell lines, and this effect was essentially abolished by GF109203X (Fig. 3A). The requirement of Rac for NSCLC cell migration was establishedby means of expressing either the RacGTPase activating protein (Rac-GAP) b2-chimaerin [35] or a dominant-negative Rac mutant (N17-Rac1) using adenoviral means. Either b2chimaerin or N17-Rac1 impaired NSCLC cell motility. The effect was proportional to the expression levels achieved by increasing the multiplicities of infection (MOI) of the corresponding AdVs (Fig. 3Band 3C).
Next, we wished to examine if PKCe mediates the activation of Rac1 by serum. As shown in Fig. 3D, depletion of PKCe with either e1 or e2 RNAi duplexes impaired Rac1 activation induced by FBS in A549 cells. In addition, NSCLC cell migration induced by serum was essentially abolished when PKCe was knocked-down ( Fig. 3E) without affecting cell viability (Fig. S5). GF109203X also inhibited cell migration, although prolonged incubation with the ''pan'' PKC inhibitor, as required in the motility assay, reduced cell viability (Fig. S5). The involvement of PKCe in NSCLC cell motility was further confirmed by means of an overexpression approach. Adenoviral delivery of PKCe to H358 or A549 cells enhanced cell motility. This effect was not observed upon overexpression of PKCa or PKCd (Fig. S6). These results clearly implicate PKCe as a mediator of Rac responses in NSCLC cells.
EGF signals to Rac via a PLCc-PKCe-dependent pathway EGF and PDGF are potent activators of Rac, and they induce the formation of circular ruffles in a similar manner than PMA and FBS (see Fig. 1D and also [36,37]). To determine if these growth factors could be mediating NSCLC cell motility induced by FBS, we used EGFR or PDGFR pharmacological inhibitors. Rac pull-down assays revealed that the EGFR inhibitor AG1478 greatly reduced FBS-induced Rac1 activation in A549 cells. On the other hand, the PDGFR inhibitor PRI-4 was ineffective (Fig. 4A). To confirm the involvement of EGFR we knocked it down using RNAi duplexes. We found that Rac1 activation in response to FBS was markedly inhibited in EGFR-depleted A549 cells (Fig. 4B), thus reinforcing the conclusions obtained with the EGFR inhibitor. Both AG1478 and EGFR RNAi also inhibited FBS-induced A549 cell motility (Fig. 4C). Furthermore, treatment of A549 cells with EGF (100 ng/ml, 2 min) caused a ,5-fold induction in Rac-GTP levels, and this effect was essentially abolished upon PKCe depletion (Fig. 4D). Likewise, ruffle formation in response to EGF was inhibited as a consequence of PKCe depletion (Fig. 4E).
PKCe is a downstream effector of EGFR, and it translocates to the plasma membrane in response to EGF in a DAG-dependent manner ( [38], and data not shown). As EGFR couples to the DAG-generating enzyme PLCc in response to stimulation [39], we reasoned that inhibition of PLC should affect Rac activation by EGF. A549 cells were treated with the PLC inhibitor U73122 or its inactive analogue U73343 and stimulated with EGF. U73122 treatment prevented Rac1 activation by EGF, whereas its inactive analogue did not. As an additional control we used PMA, which activates Rac1 regardless of PLC inhibition (Fig. 4F). Altogether, these results argue for the requirement of a PLCc-dependent activation of PKCe for the activation of Rac1 by EGF. Thus, PKCe is a downstream effector of EGFR in NSCLC cells.

Pharmacological inhibition of PKCe affects NSCLC cell migration
To further establish the requirement of PKCe for Rac1 activation in NSCLC cells we used eV1-2, a specific inhibitor of PKCe translocation that has been widely used in cellular and animal models [40,41,42,43]. A549 and H358 cells were incubated with either eV1-2 or the control peptide (TAT) and analyzed for changes in motility. In agreement with results observed using PKCe RNAi, pharmacological inhibition of PKCe significantly reduced the migratory ability of NSCLC cells in response to FBS (Fig. 5A and Fig. S7) without affecting cell viability (Fig. S5). eV1-2 also inhibited ruffle formation in response to FBS (Fig. 5B). Finally, elevation of Rac-GTP levels in response to FBS or EGF was diminished by pharmacological inhibition of PKCe (Fig. 5C).

PKCe is implicated in invasiveness and metastasis of NSCLC cells
The ability of cancer cells to invade and metastasize is associated with the production and release of proteases that remodel the extracellular matrix (ECM) as well as changes in cell adhesion [44]. To determine if PKCe is involved in NSCLC cell invasiveness we assessed migration through Matrigel using a Boyden chamber. Fig. 6A shows that PKCe-depleted A549 cells essentially lost their ability to migrate through Matrigel (Fig. 6A). On the other hand, knocking down PKCe did not affect the adhesion of A549 cells to Matrigel (Fig. 6B). These findings suggest that PKCe plays a critical role in the remodeling of the matrix and prompted us to determine the expression of relevant ECM proteases and protease inhibitors. As a first approach we used a commercial qPCR array. Interestingly, mRNA levels of ADAMTS1, MMP13, and MMP16 metalloproteinases were down-regulated in PKCe-depleted A549 cells. PKCe depletion also caused a marked induction in mRNA levels of overall the protease inhibitors TIMP1 and TIMP2, suggesting an imbalance in protease activity (Fig. 6C). As a complementary approach, we also determined levels of metalloproteinases secreted to the medium using an Antibody Array. Secretion of MMP3, MMP9, MMP10, and MMP13 from PKCe-depleted cells to conditioned medium was markedly reduced compared to control (Fig. 6D). The reduction in MMP9 activity by PKCe depletion was confirmed by zymography (Fig. S8). Thus, PKCe may affect the synthesis and/or release of specific proteases from NSCLC cells.
To establish the relevance of our findings in a metastasis model in vivo, we examined the ability of A549 cells to form lung tumors after i.v. inoculation into athymic nude mice. For these experiments we generated stable PKCe-depleted A549 cell lines (puromycin-selected) using two different shRNA lentiviruses. A control cell line was generated by infection with a non-targeting shRNA lentivirus (Fig. 7A). Mice were sacrificed 100 days after inoculation. Notably, whereas control cells readily form macroscopic lung tumors in nude mice (7.561.5 tumor nodules/lung), PKCe-depleted cells failed to form tumor nodules (0.460.2 and 1.560.5 tumor nodules/lung for e1 and e2, respectively) (Fig. 7B). Quantification of the tumor areas in histological sections of the lungs also showed remarkable differences between control and PKCe-depleted A549 cells (Fig. 7C). Representative pictures of lung tumors and H&E staining are depicted in Fig. 7D.
Finally, we carried out a survival experiment in athymic nude mice. A549 cells subject to stable PKCe-depletion or control cells were injected into the tail vein of athymic mice, and survival of the animals was followed for .8 months. As shown in Fig. 7E, whereas 100% of mice injected with control A549 cells died at ,200 days after inoculation, those injected with PKCe-depleted cells have overall survival rates of 80% (PKCe shRNA #1) and 50% (PKCe shRNA #2) at 250 days. Altogether, these experiments suggest an important role for PKCe in NSCLC cell invasion and metastatic dissemination.

Discussion
Although the involvement of PKCs in carcinogenesis has been recognized for decades, the specific roles of individual PKC isoforms during specific stages of cancer progression remain only partially understood, particularly in metastatic dissemination. This is evident in lung cancer, where PKCs, mainly those isoforms responsive to DAG/phorbol esters (cPKCs and nPKCs), have been poorly studied. Here we demonstrate that PKCe plays an essential role in lung cancer cell motility and invasiveness. Our studies identified PKCe as a key regulator of Rac1 activation in response to stimuli in NSCLC cells. Notably, RNAi depletion or pharmacological inhibition of PKCe impairs Rac1 activation as well as Rac-dependent responses, including ruffle formation, migration, and invasion. Rac1, a member of the Rho family of small GTPases plays crucial roles in the reorganization of actin cytoskeleton and motility, and inhibition of Rac1 or its downstream effectors severely impairs cell migration and dissemination of cancer cells [45]. Studies have shown that Rac is required for migration of NSCLC cells [46], which we validated in this study using a dominant-negative Rac mutant and by expressing the Rac-GAP b2-chimaerin. Aberrant expression of Rac exchange factors (Rac-GEFs) Ect2 and Vav1/2/3 has been reported in lung cancer [47,48,49], suggestive of a crucial role for the Rac signaling pathway in the progression of the disease.  Emerging evidence in the last years strongly argues for a connection between PKC isozymes and the invasive phenotype. Several reports have directly or indirectly suggested a link between PKCe and invasiveness, and PKCe was found to be highly overexpressed in lung cancer as well as in other epithelial cancers [14], prompting us to reason that this PKC may be highly relevant for NSCLC metastatic dissemination. Indeed, early studies showed that NIH 3T3 fibroblasts overexpressing PKCe acquire a polarized morphology and form invadopodial-like structures [20], and that PKCe signals downstream of b1-integrin to modulate cell spreading [14]. PKCe regulates motility and invasion in models of breast as well as head and neck cancer, at least partially through RhoA and RhoC GTPases [19,50]. Moreover, skin transgenic overexpression of PKCe in mice leads to the development of metastatic squamous carcinomas [51]. A link between PKCe and Rac has been reported in lung endothelial cells downstream of the sphingosine-1-phosphate receptor [52]. Experiments presented in our study show that there are major changes in the expression of proteases related to invasiveness upon depletion of PKCe from NSCLC cells.  protease inhibitors TIMP1 and TIMP2 are up-regulated in PKCe knockdown NSCLC cells. Furthermore, the overall secretion of several MMPs from NSCLC cells was markedly reduced by PKCe RNAi. Although the requirement for Rac in the regulation of protease/protease inhibitors in NSCLC cells remains to be determined, recent studies in different cellular models identified Rac signaling as a requirement for the secretion of MMP9, MMP2 and MMP1 [53,54,55]. Moreover, in lung cancer cells Rac mediates PKCi-induced secretion of MMP10 [56], suggesting common downstream targets for PKC isozymes controlling invasion. Thus, it is conceivable that PKCe overexpression signals via Rac to impact on the expression of multiple proteases, leading ultimately to enhanced invasiveness of lung cancer cells.
Our results identified PKCe as an essential mediator of Rac1 activation downstream of the EGFR, a receptor that couples to PLCc to generate DAG in response to its activation [39]. In PKCe-depleted NSCLC cells EGF fails to activate Rac or Racdependent responses, including ruffle formation and motility. Pharmacological inhibition of PLC impairs Rac activation by EGF, arguing that DAG-mediated recruitment of PKCe to the plasma membrane plays a crucial role in the activation of the Rac signaling pathway. How PKCe controls Rac activation downstream of EGFR (or eventually other tyrosine-kinase receptors) remains unexplored, hence identifying such mechanisms merits further investigation. One attractive mechanism may involve the activation of exchange factors for Rac (Rac-GEFs) through phosphorylation. There are at least 25 Rac-GEFs in the genome, and their relative expression in lung cancer has not been fully established. Using a Rac-GEF qPCR array we recently found that NSCLC cell lines express high levels of Vav1/2, Ect2, DEPC2 and DEPC1 compared to normal cells (M.C.C. and M.G.K., unpublished studies). Elegant studies by Fields and coworkers demonstrated that Ect2 is required for the oncogenic activity of PLCi, a member of the atypical PKC family. Ect2 is overexpressed in lung cancer and is an essential component of the PKCi-Par6 complex that controls cell polarity. RNAi-mediated knockdown of Ect2 in A549 cells inhibits Rac1 activity as well as it impairs growth, tumorigenicity and invasion of NSCLC cells [48]. However, upon RNAi depletion of Ect2 from A549 cells we could not find any significant changes in Rac1 activation induced by EGF (C.L.H. and M.G.K., unpublished studies). These results argue for major differences in the regulation of Rac signaling by DAG-responsive and DAG-unresponsive PKCs. Vav isoforms, Rho/Rac-GEF overexpressed in nearly half of human primary lung tumors [49], are candidate PKCe effectors. Vav1 activation is regulated by tyrosine phosphorylation, and it may be also regulated by serine/threonine kinases [57]. Very recently, Vav1 was found to be associated with PKC h, a member of the nPKC family closely related to PKCe [58]. Other small GTPase GEFs regulated by serine/threonines kinases include STEF, b1-Pix, and GEF-H1 [59,60,61]. Moreover, phorbol esters can activate Rho GTPases through mechanisms that involve Rho-GEF activation [62,63]. Conceivably, similar mechanisms may take place on Rac-GEFs via specific phosphorylation by PKCe. PKCe may also modulate Rac-GAPs that control the rate of GTP hydrolysis from the GTPase. For example, phorbol esters stimulate the phosphorylation of the Rho-GAP DLC1, leading to its relocalization and suppression of GAP activity [64]. Very recently, our laboratory identified a PKC-mediated phosphorylation mechanism that modulates the association of the Rac-GAP b2-chimaerin to the plasma membrane [6]. The identification of those substrates of PKCe implicated in the regulation of the Rac signaling pathway will certainly provide important insights into the mechanisms of lung cancer metastasis.
Finally, our studies may have significant therapeutic implications. Several modulators of PKC activity have been clinically tested for different types of malignancies with variable degrees of success [32,65,66]. Despite the established oncogenic and invasive activity of PKCe, to our knowledge PKCe inhibitors have not been tested as anti-cancer agents. This may be largely due to the failure to develop selective inhibitors directed against the ATP-binding site of PKCe, a shortcoming also observed for several other members of the PKC family. Nonetheless, here we found that the specific PKCe translocator inhibitor eV1-2 impairs Rac1 activation and migration in a similar manner as observed upon delivery of PKCe RNAi duplexes. In preliminary studies we found that systemic delivery of eV1-2 fused to the permeable peptide TAT inhibits NSCLC xenograft growth in nude mice (M.C.C., C.L.H., and M.G.K., unpublished observations). Our studies may establish a framework for designing PKCe inhibitors as anti-cancer agents, particularly for those malignancies such as lung cancer for which there are currently limited therapeutic options available.