Group I p21–activated kinases (PAK) are important effectors of the small GTPases Rac and Cdc42, which regulate cell motility/migration, survival, proliferation, and gene transcription. Hyperactivation of these kinases have been reported in many tumor types, making PAKs attractive targets for therapeutic intervention. PAKs are activated by growth factor–mediated signaling and are negatively regulated by the tumor suppressor neurofibromatosis type 2 (NF2)/Merlin. Thus, tumors characterized by NF2 inactivation would be expected to show hyperactivated PAK signaling. On the basis of this rationale, we evaluated the status of PAK signaling in malignant mesothelioma, an aggressive neoplasm that is resistant to current therapies and shows frequent inactivation of NF2. We show that group I PAKs are activated in most mesotheliomas and mesothelioma cell lines and that genetic or pharmacologic inhibition of PAKs is sufficient to inhibit mesothelioma cell proliferation and survival. We also identify downstream effectors and signaling pathways that may contribute mechanistically to PAK-related tumorigenesis. Specifically, we show that inhibition of PAK results in attenuation of AKT and Raf–MAPK signaling and decreased tumor cell viability. Collectively, these data suggest that pharmacologic inhibition of group I PAKs may have therapeutic efficacy in tumors characterized by PAK activation. Mol Cancer Res; 10(9); 1178–88. ©2012 AACR.

The p21-activated kinases (PAK) are serine/threonine kinases activated by the small GTPases Cdc42 and Rac in response to a variety of cell stimuli (1–3). The PAK family is grouped into 2 classes; group I, which includes PAK1-3, and group II, which contains PAK4-6 (4–7). When inactivated, group I PAKs are thought to exist as an autoinhibitory homodimer, whereby the PAK inhibitory domain (PID) of 1 PAK monomer binds to the kinase domain of a second monomer, thereby inactivating catalytic activity (8). PAKs are activated via binding of their p21-GTPase–binding domain (PBD) to small GTPases (8). Binding to small GTPases through the PBD relieves the interaction between the PID and kinase domains, permitting phosphorylation at PAK threonine 423 within the kinase domain, preventing reformation of the autoinhibitory dimer state (8). Once serine 423 is phosphorylated, PAKs undergo autophosphorylation of serine 141 and other serine residues within the C-terminal PID/PBD domains to keep PAKs catalytically active by preventing autoinhibitory homodimerization (7). In addition, phosphoinositides were recently shown to be essential cofactors for PAK1 activation (9). Group II PAKs, with the exception of PAK5, lack a recognizable PID domain and, thus, were initially thought to be constitutively active in cells (7). However, recent studies have showed that autophosphorylation of group II PAKs occurs during growth factor receptor–mediated signaling (7).

Once activated, PAKs can phosphorylate a myriad of downstream effectors regulating cellular processes such as motility, proliferation, and survival. Initially discovered through their interaction with Cdc42 and Rac, PAKs were later shown to signal downstream of these small GTPases to regulate cytoskeletal dynamics and cell motility via direct phosphorylation of substrates such as paxillin, cortactin, LIMK, and myosin light-chain kinase (2). More recently, other group I PAK substrates have been elucidated and shown to regulate other cellular processes. PAK1 can directly phosphorylate both Raf-1 and MEK1, thereby positively regulating mitogen-activated protein kinase (MAPK) signaling to promote cell proliferation and survival (10). PAKs were further shown to be important mediators and regulators of growth factor receptor–induced MAPK activation (11). In addition, PAKs are required for oncogenic Ras- and ErbB2-regulated MAPK signaling and tumorigenesis (12, 13).

Previous work has revealed that Merlin, the product of the neurofibromatosis type 2 (NF2) tumor suppressor gene, is a substrate of group I PAKs (3, 14). Phosphorylation of Merlin at Serine 518 by PAK1/2 thwarts the inhibitory effect of Merlin and, thus, constitutes an alternate means to abolish this tumor suppressor axis in proliferating cells (3, 14). Interestingly, Merlin was also shown to negatively regulate PAKs by directly interacting with the PBD domain and preventing recruitment of PAKs to focal adhesions (15). Moreover, the NF2/Merlin status of a cell was found to inversely correlate with PAK activity (15). The latter finding suggests that in tumor cells with inactivation of NF2, PAK activity exists in a hyperactivated state, potentially contributing to the proliferation, survival, and invasiveness of cancer cells.

Malignant mesotheliomas are aggressive, diffuse neoplasms arising from the serosal lining of the pleura, peritoneal, or pericardial cavities. Although exposure to asbestos is considered a causal event in most patients with mesothelioma, tumor latency following initial exposure is typically 20 to 40 years (16, 17). During this time, affected mesothelial cells acquire multiple genetic alterations contributing to mesothelioma pathogenesis. One such somatic genetic change that occurs in approximately 50% of mesotheliomas is biallelic inactivation of the NF2 tumor suppressor gene by a combination of point mutation and allelic loss (18). Concordantly, targeted deletion of 1 copy of Nf2 in mice was sufficient to accelerate mesothelioma formation in mice exposed to asbestos (19, 20). Thus, inactivation of the NF2 tumor suppressor gene is thought to be an important event in the pathogenesis of many mesotheliomas.

Although loss of NF2 in mesothelioma and other tumor types, including schwannoma, meningioma, melanoma, and renal cell carcinoma, contributes to tumorigenesis, restoration of NF2 expression as a therapy is unattainable currently because of difficulties of long-term gene expression and immune responses associated with viral-mediated gene therapy (21–24). Thus, targeting pathways that are normally negatively regulated by NF2, and whose activity or signaling becomes aberrant when this tumor suppressor is inactivated, may represent a more achievable treatment strategy. In this investigation, we evaluated PAKs as potential targets for therapeutic intervention in mesothelioma. We determined that PAK1 and PAK2 are phosphorylated and activated in most human and murine mesothelioma tumor specimens and cell lines tested. We also show that genetic or pharmacologic inhibition of PAKs signaling is sufficient to inhibit mesothelioma cell viability, proliferation, and survival. Furthermore, we show that hyperactivated PAK signal to a variety of downstream effectors, including the AKT and Raf–MAPK signaling axes, contribute to tumor cell survival and proliferation. Collectively, these findings provide strong preclinical evidence supporting group I PAK–targeted therapy as a potential intervention for the treatment of mesothelioma and other neoplasms.

Immunohistochemistry

Slides of formalin-fixed, paraffin-embedded samples of human and murine mesothelioma specimens were antigen retrieved with citrate and incubated overnight with anti-phospho-PAK1/2/3 (pSer141, 1:100; Sigma-Aldrich). Sections were stained with 3,3′-diaminobenzidine and counterstained with hematoxylin. A tissue microarray (TMA) of human mesothelioma specimens was obtained through Fox Chase Cancer Center's Histopathology Core Facility. To show antibody specificity, murine mesothelioma tissue was treated or not with lambda phosphatase (NEB) for 3 hours and then subjected to immunohistochemistry (IHC) analysis with anti-phospho-PAK1/2/3.

Primary cell cultures

Primary mouse mesothelioma cells were isolated from ascitic fluid and/or peritoneal lavage, as described elsewhere (19). Patient-derived mesothelioma cell lines were established and characterized as previously reported (25, 26).

Two-dimensional gel electrophoresis and immunoblot analysis

Briefly, actively proliferating placebo-treated or IPA-3–treated tumor cells were harvested and lysed in a nondenaturing buffer (7 mol/L urea, 4% CHAPS). Protein extracts were separated in the first dimension by isoelectric focusing (IEF) on 7 cm/pH 4-7 IEF strips (Biorad) for 2 hours at 8,000 V-h. IEF strips were then reduced in SDS buffer and embedded into the top well of a 4% to 12% gradient SDS-PAGE gel for separation in the second dimension, and then proteins were transferred to nitrocellulose membrane. Antibodies specific for total PAK1 and PAK1/2/3 (Cell Signaling) were used to probe the membrane and determine where specific PAK isoforms ran on an SDS-PAGE gel.

siRNA against PAK1 and PAK2

Stealth siRNA pools against human PAK1 and 2 (Invitrogen) were nucleofected into human Meso 22 cells using an Amaxa Cell Line Kit R and program T20 of a Nucleofector System (Lonza AG). After 48 hours, the cells were harvested, protein was extracted, and immunoblot analysis was conducted.

Lentiviral short hairpin RNA virus production and infection of mesothelioma cell lines

The pLKO.1 shGFP, shPAK1A, shPAK1B, shPAK2A, and shPAK2B vectors were purchased from the RNAi Consortium through Sigma-Aldrich. For all experiments, 70% confluent 293 HEK cells (10 cm plates) were transfected with 5 μg of each of the vectors individually plus 3.75 μg and 1.25 μg of psPAX2 packaging and pMD2G envelope vectors, respectively. After transfecting for 24 hours, the media were removed and fresh media were added to the 293 cells. Media were then collected 24 and 48 hours later, and the virus-containing media supplemented with 8 μg/mL of polybrene were used to infect ME12 and Meso 22 cell lines. Twenty-four hours after infection, the ME12 and Meso 22 cells were selected in media containing 2 μg/mL puromycin and passaged under continuous selection for at least 2 passages before use in experiments.

Immunoblotting

Immunoblots were prepared with 50 μg of protein lysate/sample, as previously described (27). Antibodies against phospho-PAK (P-PAK1/2/3, pSer141—Sigma Aldrich), total PAK1/2/3, P-AKT (Ser-473), total AKT, p44/42 MAPK (Erk1/2; 137F5; Cell Signaling), P-ERK1/2 (E4), and β-actin (Santa Cruz Biotechnology) were used for immunoblot analysis.

pcDNA–PID vector construction

The untagged PID domain of PAK1 was subcloned from a pBMN–PID–GFP plasmid via restriction digestion with XhoI and BamHI. The gel-purified fragment was ligated into pcDNA3.1 (pcDNA–PID) and subsequently confirmed via restriction enzyme digestion analysis and DNA sequencing.

Clonogenic assays

Two human mesothelioma cell lines (Meso 22, ME12) were nucleofected with pcDNA or pcDNA–PID vectors (10 μg/nucleofection) using an Amaxa Cell Line Kit R and program T20 of the Nucleofector System. A third human mesothelioma line (Meso-17) was transfected with pcDNA or pcDNA–PID vectors (20 μg/transfection) using the Xfect Transfection Reagent (Clontech). Cells were harvested 24 or 72 hours after nucleofection/transfection for immunoblot analysis and cell viability assays, respectively. For stable clonogenic assays, cells were nucleofected/transfected with pcDNA, pcDNA–PID, pcDNA–PID + pcDNA-HA-myr-AKT1, or pcDNA-PID + myc-Raf1(BXB) for 48 hours and then selected with 400 μg/mL of G418 (neomycin) for 2 weeks. Colonies were fixed with 4% paraformaldehyde, stained with Diff-Quik (Fisher Scientific), and then counted.

For short hairpin RNA (shRNA) knockdown clonogenic assays, mesothelioma cells were infected and selected with puromycin (2 μg/mL) for at least 2 passages. Cells were then counted, and 500 cells were plated per well in 6-well plates. After 1 to 2 weeks, the resulting cell colonies were fixed with 4% paraformaldehyde and stained with 0.2% crystal violet in 20% ethanol.

Cell number assay

To evaluate cell proliferation, ME12 and Meso 22 cells stably expressing shGFP, shPAK1A, shPAK1B, shPAK2A, or shPAK2B (2 × 104 cells each, at passage 3 of selection) were seeded on a 12-well plate in triplicate (Day 1). Cells were counted each day for 3 days and plotted as a line graph with SDs.

IPA-3 treatment of mesothelioma cell lines

Mesothelioma cells were treated with varying concentrations of IPA-3, a small-molecule inhibitor of PAK or dimethyl sulfoxide (DMSO) control, for 24, 48, or 72 hours for immunoblot analysis of antibody arrays, or for cell viability, cell cycle, and apoptosis assays.

Cell viability assays

Mesothelioma cells were each seeded onto 96-well plates at a density of 5,000 cells per well. After 24 hours, cells were treated with 25 μmol/L IPA-3 or DMSO vehicle control for 72 hours. MTS reagent was added and absorbance was determined at 490 nm as a read out of cell viability. Mesothelioma cells were nucleofected with pcDNA or pcDNA–PID vectors (10 μg/nucleofection); 24 hours after nucleofection, cells were seeded onto 96-well plates at a density of 5,000 cells per well. MTS reagent was added to the plate 72 hours after nucleofection, and absorbance was determined at 490 nm as a read out of cell viability. As a control, ME12 and Meso 22 cells were also treated with control compound PIR3.5 at increasing concentrations and analyzed for cell viability via MTS assay.

Cell-cycle analysis

Mesothelioma cells were treated with 25 μmol/L IPA-3 or DMSO vehicle control for 24 hours, harvested, and fixed in cold 70% ethanol overnight. The next day, cells were pelleted and washed in PBS before being treated with RNase in PBS (200 μg/mL) for 30 minutes at room temperature. Cells were then pelleted and stained with propidium iodide (10 μg/mL). DNA content was determined on a FACscan Flow Cytometer (Beckton Dickinson) and quantitated using FlowJo analysis software.

Apoptosis assay

Tumor cells were treated with 25 μmol/L IPA-3 or vehicle control for 48 hours. The cells were harvested and analyzed for apoptosis using the Cell Death Detection ElisaPLUS Kit (Roche Applied Biosciences). All experiments were conducted in triplicate in replicate independent experiments.

Antibody array analysis

Human PhosphoKinase PathwayScan Antibody Arrays were purchased from R&D Systems. Mesothelioma cell lines were treated with 25 μmol/L IPA-3 of vehicle control or nucleofected with pcDNA or pcDNA–PID as described above. Cells were harvested and lysed for protein extraction according to the manufacturer's protocol. For each treatment condition, 500 μg of total protein extract was incubated with membrane A or membrane B overnight at 4°C. Membranes were developed as recommended by the manufacturer.

Group I PAKs are activated in mesothelioma tumors and cell lines

To evaluate the phosphorylation and activation status of group I PAKs in mesothelioma specimens, we used a 3-pronged approach. First, we used a phospho-specific antibody against serine 141 of PAK1/2/3 to assess PAK phosphorylation via IHC on a TMA spotted with archival human mesothelioma tumors. All 15 pleural mesothelioma tumor specimens tested stained more positively for active PAK than did normal pleural lining surrounding the lung (Fig. 1A). Concordantly, asbestos-induced murine mesothelioma tumors also stained positively for active PAK (Fig. 1B). Lamda phosphatase treatment of murine mesothelioma tissues abolished phospho-PAK1/2/3 staining, demonstrating the specificity of the antibody used in our IHC analyses (Supplementary Fig. S3). Second, we evaluated the phosphorylation status of group I PAKs in human mesothelioma cell lines by immunoblotting using the same antibody employed for the IHC analysis. As shown in Fig. 1C, hyperactivation of PAK was found in most human and murine mesothelioma cell lines tested. In all, PAK phosphorylation was observed in approximately 80% of human and murine mesothelioma cell lines tested in which NF2 expression was lost. It should be noted that not all mesothelioma cell lines showed strong PAK phosphorylation, even though all were NF2 deficient, suggesting that other pathways may control PAK activity in tumor cells.

The antibody used for both IHC and immunoblot analyses recognize phosphorylated forms of all 3 group I PAKs. To determine if a specific group I PAK is consistently activated in mesothelioma, we used a third approach to assess specific group I PAK phosphorylation: 2D gel electrophoresis/immunoblot analysis. Seven human mesothelioma cell line extracts were separated by IEF followed by second-dimension SDS-PAGE electrophoresis. Proteins from the gels were transferred to nitrocellulose and then blotted with antibodies against PAK1, PAK1/2/3, and P-PAK1/2/3 antibodies. As shown in Fig. 1D and Supplementary Fig. S1, PAK1 was expressed and appeared to be phosphorylated on the basis of the presence of multiple forms migrating toward the more acidic isoelectric point (pI). In addition, immunoblot analysis with the PAK1/2/3 antibody revealed a second, slower migrating isoform of group I PAKs that was also phosphorylated (Fig. 1D and Supplementary Fig. S1). For further clarification, we used RNAi against PAK1 or PAK2, which revealed that the slower migrating band is PAK1 and the faster migrating band is PAK2 (Supplementary Fig. S2). Immunoblot analysis of the 2D Western blot with an antibody against phosphorylated PAK1/2/3 (P-PAK1/2/3) appeared to recognize PAK 2 with higher affinity than PAK1 (Fig. 1D and Supplementary Figs. S1 and S2). Together, these data show that PAK1 and PAK2 are phosphorylated and active in human mesothelioma cells.

Expression of a dominant-negative PAK–PID is sufficient to inhibit PAK phosphorylation and tumor cell viability and proliferation

We next tested whether or not PAK activity is required for mesothelioma cell viability. To inhibit all 3 group I PAKs simultaneously, we expressed the PID (a.a. 83–149—which inhibits all group I PAKs) exogenously in mesothelioma cells. Transient expression of PID of mesothelioma cell lines by nucleofection for 72 hours in ME12 and Meso 22 cells or by transfection for 24 hours in Meso 17 cells resulted in decreased phospho-PAK levels, based on immunoblot analysis (Fig. 2A and Supplementary Fig. S4). Total PAK levels were also decreased in ME12 and Meso 22 cells after 72 hours but were unchanged when PID was expressed for 24 hours (Fig. 2A and Supplementary Fig. S6). To test whether PAK activity is required for tumor cell viability, mesothelioma cells were transiently nucleofected/transfected with PID and cell viability was evaluated 72 hours later. As shown in Fig. 2B, expression of PID was sufficient to inhibit cell viability by at least 2-fold in ME12 and Meso 22 cells. Similar results were observed in transfected Meso 17 cells (Supplementary Fig. S4B). In light of the fact that transfection efficiencies are generally low in mesothelioma cell lines, we decided to stably express PID and evaluate its effect on cell proliferation/survival via colony formation assay. Mesothelioma cells were nucleofected or transfected with PID or control pcDNA plasmids and then selected with G418 beginning 24 hours after nucleofection/transfection and lasting 2 weeks, at which point the cells were fixed and stained. As shown in Fig. 2B and Supplementary Fig. S4B, colony formation was undetectable in all 3 mesothelioma cell lines expressing the PID sequence; however, colony formation was robust in the neomycin-selected control cells transfected with pcDNA. Together, these data suggest that group I PAK activity is required for mesothelioma cell survival and proliferation.

Knockdown of PAK1 and PAK2 decreases mesothelioma cell proliferation

In a complementary experiment, we addressed whether expression of PAK1 and/or PAK2, the 2 predominant group I PAKs expressed in mesothelioma, are required for mesothelioma cell proliferation. Using lentiviruses expressing shRNAs against PAK1, PAK2, or GFP (control), we infected ME12 and Meso 22 cell lines and selected with puromycin to make stable cell lines. As shown in Fig. 3A, we were able to knock down both PAK1 and PAK2 effectively with 2 different shRNAs to each isoform, when compared with shGFP control cells. The stable cell lines were then seeded for cell proliferation and clonogenic assays. In Fig. 3B, knock down of PAK1 had a modest effect on mesothelioma cell proliferation, with the exception of shPAK1A knockdown, which essentially blocked Meso 22 cell proliferation. Knockdown of PAK2 consistently caused proliferation arrest in ME12 cells, whereas only 1 shRNA against PAK2 (shPAK2B) caused obvious changes in Meso 22 cell proliferation (Fig. 3B). Similar results were found in clonogenic assays in which cells were sparsely seeded and allowed to grow for a week, followed by subsequent staining with crystal violet to document colony formation (Fig. 3C). ME12 cell lines grew much slower that Meso 22 cells in this assay, such that ME12 colonies were not large enough for visualization with crystal violet. To circumvent this issue, the ME12 cell lines seeded and counted for cell doubling assays were allowed to grow for a week and stainable colonies were obtained for analysis (Fig. 3C, left). Together, these data and our results with the PID construct (Fig. 2) implicate PAK1 and PAK2 in the regulation of mesothelioma cell proliferation.

A novel small-molecule inhibitor of PAKs (IPA-3) inhibits PAK1 and PAK2 phosphorylation, mesothelioma cell viability, proliferation, and survival

Although genetic attenuation of PAK activity via exogenous expression of a dominant-negative PID sequence and knockdown with shRNA was sufficient to inhibit mesothelioma cell viability and proliferation, targeted expression of PID or shRNA expression in cancer cells is not a feasible or rational cancer therapy. On the other hand, small-molecule inhibitors of kinases have been effectively exploited as targeted therapies to inhibit cancer cell proliferation and survival. However, many of the small-molecule compounds targeting kinases are actually ATP-competitive inhibitors and, as such, can be promiscuous and target multiple kinases due to the fact that ATP-binding sites in kinases are evolutionarily conserved (8). Recently, a small-molecule inhibitor of PAK1, and to lesser extent PAK2/3/5, was discovered that effectively inhibits PAK activation by locking or stabilizing PAK homodimers in an autoinhibitory state via covalent binding to the PID sequence of PAKs (28, 29). In addition, this inhibitor, dubbed inhibitor of PAK activation 3 (IPA-3), was showed to be cell permeable and capable of inhibiting platelet-derived growth factor–mediated activation of MAPK signaling, similar to genetic perturbation of PAK1 (11, 28).

To test whether IPA-3 can inhibit mesothelioma cell viability akin to exogenous expression of the PID sequence of PAK1, we tested the effect of increasing concentrations of IPA3, or DMSO control, on mesothelioma cell viability. As shown in Fig. 4A, addition of IPA-3 for 72 hours inhibited mesothelioma cell viability in a dose-dependent manner, with an IC50 of approximately 25 μmol/L. Treatment of mesothelioma cell lines with control compound PIR3.5 produced little change in cell viability, suggesting the effect observed with IPA-3 is mediated through inhibition of group I PAKs (Supplementary Fig. S5). To evaluate whether IPA-3 inhibits PAK phosphorylation, we treated ME12 and Meso 22 cells for 72 hours and evaluated PAK1/2/3 phosphorylation via immunoblot analysis (Fig. 4B). IPA-3–treated mesothelioma cells showed a marked reduction in PAK phosphorylation and slightly decreased total PAK levels (Fig. 4B). Although IPA-3 was initially identified as an inhibitor of PAK1, we also found that IPA-3–treated ME12 cells displayed decreases in both PAK1 and PAK2 phosphorylation with both forms shifting from more acidic (or phosphorylated) to more basic based on 2D gel electrophoresis immunoblot analysis (Fig. 4C). Together, these data suggest that IPA-3 treatment can inhibit group I PAK phosphorylation and mesothelioma cell viability as effectively as overexpression of PID.

To assess mechanistically how IPA-3 treatment decreases tumor cell viability, we evaluated the effect of PAK inhibition on cell-cycle progression/cell proliferation. Mesothelioma cells were treated with IPA-3 or DMSO control for 48 hours and then fixed and stained with propidium iodide. IPA-3–treated cells were then analyzed by flow cytometry for DNA content. As shown in Fig. 4D, IPA-3–treated mesothelioma cells showed a marked increase in the percentage of cells in G2/M, suggesting arrest or delay in that stage of the cell cycle. This finding is consistent with other reports implicating a role for PAKs in mitosis (30–32). To address the effect of IPA-3–mediated PAK inhibition on mesothelioma cell survival, ME12 and Meso 22 cells were treated with the drug for 48 hours and assessed for apoptosis. As shown in Fig. 4E, IPA-3–treated ME12 and Meso 22 showed a significant increase in DNA fragmentation when compared with DMSO-treated control cells, suggesting a role for PAKs in mesothelioma cell survival as well. An increase in the sub-G1 cell population was also observed in propidium iodide–stained Meso 22 cells treated with IPA-3, consistent with an increase in apoptotic cells (Fig. 4D, panel 2). These data suggest that inhibiting group I PAKs with the small-molecule inhibitor IPA-3 causes a decrease in PAK phosphorylation in association with decreased mesothelioma cell viability, proliferation, and survival, consistent with the effects observed for dominant-negative expression of PID or knockdown of PAK1 and PAK2 in mesothelioma cells.

Group I PAKs promote tumor cell survival and proliferation via the AKT1 and Raf1–MAPK signaling pathways

Our data implicate Group I PAKs in mesothelioma cell viability, proliferation, and survival. To identify downstream effector pathways that mediate oncogenic PAK signaling, we evaluated the status of the AKT and MAPK signaling pathways. In Fig. 5A, IPA-3 treatment reduced the phosphorylation of both AKT and ERK1/2, suggesting that PAKs play a significant role in regulating these signaling molecules in mesothelioma cells. A similar effect on AKT and ERK1/2 phosphorylation was upon exogenous expression of PID (Supplementary Fig. S6). Expression of PID or treatment with IPA-3 also inhibited the phosphorylation of multiple kinases, based on results from reverse-phase antibody arrays, implicating these signaling axes as bona fide targets of PAKs (Supplementary Fig. S7). Some of these kinases, including p70S6 kinase, p90RSK, STAT5a/b, STAT1, and MSK1/2, have not been previously linked to PAK signaling but are implicated in carcinogenesis (33–36).

To determine if inhibition of AKT and Raf1 are required for PAK-mediated mesothelioma cell viability, constitutively active forms of AKT1 and Raf1 were coexpressed with PID. Both constitutively active AKT1 and Raf1 were able to rescue cell viability in clonogenic assays in the presence of PAK inhibition by PID in 2 different mesothelioma cell lines tested (Fig. 5B). These data implicate AKT1 and Raf1–MAPK signaling as important downstream effectors of PAK for mesothelioma cell viability. In a complementary experiment, we inhibited the AKT and MAPK signaling pathways with LY294002 and PD98509, respectively. In Fig. 5C, inhibition of both AKT and MAPK with LY294002 and PD98509, respectively (bar 5), was sufficient to phenocopy the effect of PAK inhibition (IPA-3 treatment—bar 2) on tumor cell viability, whereas inhibition of AKT or MAPK signaling alone was not (bars 3 and 4). Interestingly, inhibiting AKT, but not MAPK, sensitized mesothelioma cells to PAK inhibition (Fig. 5C, bars 6 and 7), suggesting that the AKT signaling node may be a more dominant effector of PAK signaling than the MAPK pathway for mesothelioma cell survival. Collectively, these data implicate the AKT and MAPK signaling pathways as important effector pathways of PAKs in regulating tumor cell viability.

In this report, we show that group I PAKs are frequently activated in mesothelioma, a tumor type with frequent inactivation of NF2/Merlin. This finding is in agreement with the notion that Merlin acts as a tumor suppressor by restraining PAK signaling. We show for the first time that inhibition of group I PAKs by genetic or pharmacologic means is effective in restraining mesothelioma cell viability, proliferation, and survival. We also provide proof-of-concept that use of novel small-molecule inhibitors of group I PAKs may serve as important therapeutic agents for treating tumor types that show frequent hyperactivation of PAK. Moreover, we show that inhibition of PAK results in attenuated signaling via AKT and MAPK, 2 central nodes in tumorigenesis.

Several reports have showed that group I PAKs is important in cellular transformation and tumorigenesis. For example, PAK1 was recently shown to be hyperactivated in schwannomas from patients with NF2 syndrome and was required for rat schwannoma cell proliferation (37). Independently, PAK activity was found to be required for tumor cell growth and invasion in a cell-based model of NF2 (38). Both groups showed that targeting more than 1 of the group I PAKs was required to inhibit NIH-3T3-NF2-BBA–transformed cell proliferation and growth in vivo (37, 38). Consistent with these results, we show that both PAK1 and PAK2 are activated in NF2-deficient mesothelioma cells and that genetic (shRNA) or pharmacologic inhibition of both kinases was sufficient to inhibit tumor cell viability, although our shRNA data showed that targeting either isoform could affect mesothelioma cell proliferation (Fig. 3). In addition, other studies have shown that PAK activity is required for both Ras- and ERBB2-mediated transformation in cell models of breast cancer (12, 39).

To our knowledge, this is the first report demonstrating that pharmacologic inhibition of PAKs (with IPA-3) is efficacious in patient-derived mesothelioma cells, although PAK kinases have been shown to be effective in other preclinical cancer models (40). IPA-3 was recently shown to inhibit PAK2-mediated cell spreading but not viability of primary human schwannoma cells (41). Although this lack of effect on the viability of these benign tumor cells appears to contradict our findings and those of others who have inhibited PAK by genetic means (37, 38), the schwannoma cells in the aforementioned study were treated with IPA-3 for only 24 hours, whereas in our study, mesothelioma cells were treated for 72 hours before the cell viability analysis. In addition, schwann cells and mesothelial cells arise from different embryonic lineages; thus, the resulting neoplasms may have different tissue-specific dependency on PAK activity for cell survival and proliferation. Moreover, other investigations have found that dissimilar cell types respond differently to PAK inhibition. For example, Yi and colleagues (37) and Tang and colleagues (13) found that although PAK1 perturbation had no effect in NIH3T3 cells, PAK1 ablation/inhibition in the rat schwannoma cell line RT4 resulted in growth arrest.

Although our data suggest group I PAKs are important for tumor cell viability, it is not clear what downstream substrates or effector pathways regulated by PAKs are critical for regulating mesothelioma cell proliferation and survival. We found that inhibition of PAKs in mesothelioma cells, via dominant-negative or pharmacologic means, had a modest but reproducible negative effect on both AKT and ERK signaling pathways regulated in part by PAKs and that promote tumor cell proliferation and survival. Interestingly, dominant expression of the PID sequence did not affect AKT or ERK activity in a cell-based model of NF2 (38), although these experiments were conducted in vivo, where feedback loops could account for these differences. We showed that AKT1 and Raf–MAPK signaling pathways are potentially important mediators of group PAK's prosurvival and proproliferation signaling in mesothelioma cells. A similar result with dual inhibition of AKT and ERK kinases was shown to specifically affect colon cancer cell growth and invasion (42). In addition, PAK inhibition affected the phosphorylation of multiple kinases involved in various signal transduction pathways, based on results from reverse-phase antibody arrays, implicating these signaling axes as bona fide targets of PAK activity. The kinases that were altered by inhibition of PAK via expression of PID and by treatment with IPA-3 represent novel effectors because both of these methods inhibit PAK activity in a specific manner and are unlikely to have off-target effects. Phosphorylation of paxillin, a known substrate of PAKs, was also diminished in both PID-expressing and IPA-3–treated mesothelioma cells, further supporting the global signaling antibody array approach used herein. In addition, phosphorylation of p70S6 kinase, p90RSK, STAT5a/b, STAT1, and MSK1/2, proteins previously not considered as direct PAK substrates or effectors, were also decreased in PID-expressing and IPA-3–treated mesothelioma cells. These data suggest that many effectors of PAK signaling could be important, either independently or in concert, for the proliferation and survival of tumor cells characterized by hyperactivation of PAK. Importantly, we show that activation of AKT1 and Raf–MAPK signaling appears to be key downstream effects that contribute to oncogenic PAK signaling in mesothelioma.

Together, these data implicate group I PAKs as important regulators of tumor cell proliferation and survival in mesothelioma and other types of cancer exhibiting hyperactivation of PAK and identifies new potential effector pathways of PAKs in tumor cells. PAK inhibition was sufficient to attenuate both the AKT and MAPK signaling pathways in mesothelioma cells, thus simultaneously inhibiting 2 important oncogenic effectors critical to tumor cell survival and proliferation, respectively. Thus, in some tumor types, PAK inhibitors could prove to be as effective as single agents as combinatorial AKT and MAPK inhibitors (summarized in Fig. 5D). This report reinforces the importance of developing clinically relevant small-molecule inhibitors of PAK as a targeted therapy for multiple malignancies.

No potential conflicts of interest were disclosed.

Conception and design: C.W. Menges, J. Chernoff, J.R. Testa

Development of methodology: C.W. Menges, E. Sementino

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C.W. Menges, J. Talarchek, J. Chernoff

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C.W. Menges

Writing, review, and/or revision of the manuscript: C.W. Menges, J. Chernoff, J.R. Testa

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Xu, J. Peterson, J.R. Testa

Study supervision: C.W. Menges, J. Peterson, J.R. Testa

The authors thank Drs. S. Jhanwar and H. Pass for providing mesothelioma cell lines.

This study was supported by NIH awards CA-114047 (J.R. Testa), CA-148805 (J. Chernoff, J.R. Testa), GM083025 (J.R. Peterson), and CA-06927 (Fox Chase Cancer Center), an appropriation from the Commonwealth of Pennsylvania, an award from the Children's Tumor Foundation (2011-15-012; J. Chernoff), and by a gift from the Local #14 Mesothelioma Fund of the International Association of Heat and Frost Insulators & Allied Workers in memory of Hank Vaughan and Alice Haas (J.R.T.).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Dummler
B
,
Ohshiro
K
,
Kumar
R
,
Field
J
. 
Pak protein kinases and their role in cancer
.
Cancer Metastasis Rev
2009
;
28
:
51
63
.
2.
Molli
PR
,
Li
DQ
,
Murray
BW
,
Rayala
SK
,
Kumar
R
. 
PAK signaling in oncogenesis
.
Oncogene
2009
;
28
:
2545
55
.
3.
Xiao
GH
,
Beeser
A
,
Chernoff
J
,
Testa
JR
. 
p21-activated kinase links Rac/Cdc42 signaling to merlin
.
J Biol Chem
2002
;
277
:
883
6
.
4.
Eswaran
J
,
Soundararajan
M
,
Knapp
S
. 
Targeting group II PAKs in cancer and metastasis
.
Cancer Metastasis Rev
2009
;
28
:
209
17
.
5.
Eswaran
J
,
Soundararajan
M
,
Kumar
R
,
Knapp
S
. 
UnPAKing the class differences among p21-activated kinases
.
Trends Biochem Sci
2008
;
33
:
394
403
.
6.
Jaffer
ZM
,
Chernoff
J
. 
p21-activated kinases: three more join the Pak
.
Int J Biochem Cell Biol
2002
;
34
:
713
7
.
7.
Wells
CM
,
Jones
GE
. 
The emerging importance of group II PAKs
.
Biochem J
2010
;
425
:
465
73
.
8.
Yi
C
,
Maksimoska
J
,
Marmorstein
R
,
Kissil
JL
. 
Development of small-molecule inhibitors of the group I p21-activated kinases, emerging therapeutic targets in cancer
.
Biochem Pharmacol
2010
;
80
:
683
9
.
9.
Strochlic
TI
,
Viaud
J
,
Rennefahrt
UE
,
Anastassiadis
T
,
Peterson
JR
. 
Phosphoinositides are essential coactivators for p21-activated kinase 1
.
Mol Cell
2010
;
40
:
493
500
.
10.
Eblen
ST
,
Slack
JK
,
Weber
MJ
,
Catling
AD
. 
Rac-PAK signaling stimulates extracellular signal-regulated kinase (ERK) activation by regulating formation of MEK1-ERK complexes
.
Mol Cell Biol
2002
;
22
:
6023
33
.
11.
Beeser
A
,
Jaffer
ZM
,
Hofmann
C
,
Chernoff
J
. 
Role of group A p21-activated kinases in activation of extracellular-regulated kinase by growth factors
.
J Biol Chem
2005
;
280
:
36609
15
.
12.
Arias-Romero
LE
,
Villamar-Cruz
O
,
Pacheco
A
,
Kosoff
R
,
Huang
M
,
Muthuswamy
SK
, et al
A Rac-Pak signaling pathway is essential for ErbB2-mediated transformation of human breast epithelial cancer cells
.
Oncogene
2010
;
29
:
5839
49
.
13.
Tang
Y
,
Chen
Z
,
Ambrose
D
,
Liu
J
,
Gibbs
JB
,
Chernoff
J
, et al
Kinase-deficient Pak1 mutants inhibit Ras transformation of Rat-1 fibroblasts
.
Mol Cell Biol
1997
;
17
:
4454
64
.
14.
Kissil
JL
,
Johnson
KC
,
Eckman
MS
,
Jacks
T
. 
Merlin phosphorylation by p21-activated kinase 2 and effects of phosphorylation on merlin localization
.
J Biol Chem
2002
;
277
:
10394
9
.
15.
Kissil
JL
,
Wilker
EW
,
Johnson
KC
,
Eckman
MS
,
Yaffe
MB
,
Jacks
T
. 
Merlin, the product of the Nf2 tumor suppressor gene, is an inhibitor of the p21-activated kinase, Pak1
.
Mol Cell
2003
;
12
:
841
9
.
16.
Ramos-Nino
ME
,
Testa
JR
,
Altomare
DA
,
Pass
HI
,
Carbone
M
,
Bocchetta
M
, et al
Cellular and molecular parameters of mesothelioma
.
J Cell Biochem
2006
;
98
:
723
34
.
17.
Yang
H
,
Testa
JR
,
Carbone
M
. 
Mesothelioma epidemiology, carcinogenesis, and pathogenesis
.
Curr Treat Options Oncol
2008
;
9
:
147
57
.
18.
Bianchi
AB
,
Mitsunaga
SI
,
Cheng
JQ
,
Klein
WM
,
Jhanwar
SC
,
Seizinger
B
, et al
High frequency of inactivating mutations in the neurofibromatosis type 2 gene (NF2) in primary malignant mesotheliomas
.
Proc Natl Acad Sci U S A
1995
;
92
:
10854
8
.
19.
Altomare
DA
,
Vaslet
CA
,
Skele
KL
,
De Rienzo
A
,
Devarajan
K
,
Jhanwar
SC
, et al
A mouse model recapitulating molecular features of human mesothelioma
.
Cancer Res
2005
;
65
:
8090
5
.
20.
Fleury-Feith
J
,
Lecomte
C
,
Renier
A
,
Matrat
M
,
Kheuang
L
,
Abramowski
V
, et al
Hemizygosity of Nf2 is associated with increased susceptibility to asbestos-induced peritoneal tumours
.
Oncogene
2003
;
22
:
3799
805
.
21.
Dalgliesh
GL
,
Furge
K
,
Greenman
C
,
Chen
L
,
Bignell
G
,
Butler
A
, et al
Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes
.
Nature
2010
;
463
:
360
3
.
22.
Sinkre
P
,
Perry
A
,
Cai
D
,
Raghavan
R
,
Watson
M
,
Wilson
K
, et al
Deletion of the NF2 region in both meningioma and juxtaposed meningioangiomatosis: case report supporting a neoplastic relationship
.
Pediatr Dev Pathol
2001
;
4
:
568
72
.
23.
Sterman
DH
,
Molnar-Kimber
K
,
Iyengar
T
,
Chang
M
,
Lanuti
M
,
Amin
KM
, et al
A pilot study of systemic corticosteroid administration in conjunction with intrapleural adenoviral vector administration in patients with malignant pleural mesothelioma
.
Cancer Gene Ther
2000
;
7
:
1511
8
.
24.
Wilson
JM
. 
Lessons learned from the gene therapy trial for ornithine transcarbamylase deficiency
.
Mol Genet Metab
2009
;
96
:
151
7
.
25.
Pass
HI
,
Stevens
EJ
,
Oie
H
,
Tsokos
MG
,
Abati
AD
,
Fetsch
PA
, et al
Characteristics of nine newly derived mesothelioma cell lines
.
Ann Thorac Surg
1995
;
59
:
835
44
.
26.
Taguchi
T
,
Jhanwar
SC
,
Siegfried
JM
,
Keller
SM
,
Testa
JR
. 
Recurrent deletions of specific chromosomal sites in 1p, 3p, 6q, and 9p in human malignant mesothelioma
.
Cancer Res
1993
;
53
:
4349
55
.
27.
Menges
CW
,
Chen
Y
,
Mossman
BT
,
Chernoff
J
,
Yeung
AT
,
Testa
JR
. 
A phosphotyrosine proteomic screen identifies multiple tyrosine kinase signaling pathways aberrantly activated in malignant mesothelioma
.
Genes Cancer
2010
;
1
:
493
505
.
28.
Deacon
SW
,
Beeser
A
,
Fukui
JA
,
Rennefahrt
UE
,
Myers
C
,
Chernoff
J
, et al
An isoform-selective, small-molecule inhibitor targets the autoregulatory mechanism of p21-activated kinase
.
Chem Biol
2008
;
15
:
322
31
.
29.
Viaud
J
,
Peterson
JR
. 
An allosteric kinase inhibitor binds the p21-activated kinase autoregulatory domain covalently
.
Mol Cancer Ther
2009
;
8
:
2559
65
.
30.
Banerjee
M
,
Worth
D
,
Prowse
DM
,
Nikolic
M
. 
Pak1 phosphorylation on t212 affects microtubules in cells undergoing mitosis
.
Curr Biol
2002
;
12
:
1233
9
.
31.
Bompard
G
,
Rabeharivelo
G
,
Frank
M
,
Cau
J
,
Delsert
C
,
Morin
N
. 
Subgroup II PAK-mediated phosphorylation regulates Ran activity during mitosis
.
J Cell Biol
2010
;
190
:
807
22
.
32.
Thiel
DA
,
Reeder
MK
,
Pfaff
A
,
Coleman
TR
,
Sells
MA
,
Chernoff
J
. 
Cell cycle-regulated phosphorylation of p21-activated kinase 1
.
Curr Biol
2002
;
12
:
1227
32
.
33.
Klampfer
L
. 
Signal transducers and activators of transcription (STATs): Novel targets of chemopreventive and chemotherapeutic drugs
.
Curr Cancer Drug Targets
2006
;
6
:
107
21
.
34.
Ulivi
P
,
Arienti
C
,
Amadori
D
,
Fabbri
F
,
Carloni
S
,
Tesei
A
, et al
Role of RAF/MEK/ERK pathway, p-STAT-3 and Mcl-1 in sorafenib activity in human pancreatic cancer cell lines
.
J Cell Physiol
2009
;
220
:
214
21
.
35.
Pon
YL
,
Zhou
HY
,
Cheung
AN
,
Ngan
HY
,
Wong
AS
. 
p70 S6 kinase promotes epithelial to mesenchymal transition through snail induction in ovarian cancer cells
.
Cancer Res
2008
;
68
:
6524
32
.
36.
Perez-Cadahia
B
,
Drobic
B
,
Espino
PS
,
He
S
,
Mandal
S
,
Healy
S
, et al
Role of MSK1 in the malignant phenotype of Ras-transformed mouse fibroblasts
.
J Biol Chem
2011
;
286
:
42
9
.
37.
Yi
C
,
Wilker
EW
,
Yaffe
MB
,
Stemmer-Rachamimov
A
,
Kissil
JL
. 
Validation of the p21-activated kinases as targets for inhibition in neurofibromatosis type 2
.
Cancer Res
2008
;
68
:
7932
7
.
38.
Chow
HY
,
Stepanova
D
,
Koch
J
,
Chernoff
J
. 
p21-Activated kinases are required for transformation in a cell-based model of neurofibromatosis type 2
.
PLoS One
2010
;
5
:
e13791
.
39.
Vadlamudi
RK
,
Adam
L
,
Wang
RA
,
Mandal
M
,
Nguyen
D
,
Sahin
A
, et al
Regulatable expression of p21-activated kinase-1 promotes anchorage-independent growth and abnormal organization of mitotic spindles in human epithelial breast cancer cells
.
J Biol Chem
2000
;
275
:
36238
44
.
40.
Zhu
G
,
Wang
Y
,
Huang
B
,
Liang
J
,
Ding
Y
,
Xu
A
, et al
A Rac1/PAK1 cascade controls beta-catenin activation in colon cancer cells
.
Oncogene
2012
;
31
:
1001
12
.
41.
Flaiz
C
,
Chernoff
J
,
Ammoun
S
,
Peterson
JR
,
Hanemann
CO
. 
PAK kinase regulates Rac GTPase and is a potential target in human schwannomas
.
Exp Neurol
2009
;
218
:
137
44
.
42.
Huynh
N
,
Liu
KH
,
Baldwin
GS
,
He
H
. 
P21-activated kinase 1 stimulates colon cancer cell growth and migration/invasion via ERK- and AKT-dependent pathways
.
Biochim Biophys Acta
2010
;
1803
:
1106
13
.

Supplementary data