Effect of P21‐activated kinase 1 (PAK‐1) inhibition on cancer cell growth, migration, and invasion

Abstract P21‐activated kinase‐1 (PAK‐1) is a serine/threonine kinase involved in multiple signaling pathways that mediate cellular functions such as cytoskeletal motility, cell proliferation, and survival. PAK‐1 expression is altered in various cancers, including prostate and breast. Our recent studies showed that prostate cancer cells expressing higher levels of PAK‐1 were resistant to the cytotoxic effects of the PAK‐1 inhibitor, inhibitor targeting PAK‐1 activation‐3 (IPA‐3), compared to those with lower expression. This study expanded these findings to other cancers (breast and melanoma) by testing the hypothesis that genetic and pharmacological inhibition of PAK‐1 alters cell growth, migration, and invasion in prostate, breast, and skin cancer cell lines. We also tested the specificity of IPA‐3 for PAK‐1 and the hypothesis that gene silencing of PAK‐1 altered the efficacy of sterically stabilized liposomes (SSL) containing IPA‐3 (SSL‐IPA‐3). PAK‐1 expression was identified in four different breast cancer cell lines, and in a melanoma cell line. The expression of PAK‐1 correlated to the IC50 of IPA‐3 as measured by MTT staining. PAK‐1 inhibition using shRNA correlated with decreased cell migration and invasion in prostate cancer DU‐145 and breast cancer MCF‐7 cells. Decreased migration and invasion also correlated to decreased expression of E‐cadherin and alterations in C‐X‐C Chemokine Receptor type 4 and Homing Cell Adhesion Molecule expression. PAK‐1 inhibition increased the cytotoxicity of IPA‐3, and the cytotoxicity of SSL‐IPA‐3 to levels comparable to that of free drug. These data demonstrate that both pharmacological and molecular inhibition of PAK‐1 decreased growth in prostate, breast, and melanoma cancer cell lines, and increased the toxicity of IPA‐3 and its liposomal formulation. These data also show the specificity of IPA‐3 for PAK‐1, are some of the first data suggesting that IPA‐3 is a therapeutic treatment for breast cancer and melanoma, and demonstrate the efficacy of liposome‐encapsulated IPA‐3 in breast cancer cells.


| INTRODUC TI ON
P21-activated kinases (PAKs) are serine/threonine kinases that mediate multiple signal transduction pathways, including those that control cellular functions such as cytoskeletal motility, cell proliferation, and survival. 1,2 Upregulation of PAKs in tumors is suggested to increase cellular transformation, motility, and invasion in surrounding tissues leading to cancer metastasis. 3,4 In addition, overactivation of PAKs results in downregulation of proapoptotic pathways and promotion of cell survival. 3 While it has been reported that PAK-6, a group II PAK, is overexpressed in prostate cancer, 5,6 recent studies from our laboratory and others have shown that PAK-1 is also overexpressed in prostate tumor tissues and in metastasized sites in the human lung. 2,7 PAK-1 is a major downstream effector of Rac1, which mediates cytoskeletal remodeling during prostate cancer invasion. [8][9][10] Previous studies showed that PAK-1 mediated the growth of prostate PC-3 cell tumor xenografts in athymic nude mice as well as the transforming growth factor-β (TGFβ)-induced prostate cancer cell epithelial-mesenchymal transition (EMT). 11 These studies suggested that PAK-1 plays a major role in prostate cancer progression and is a potential target for prostate cancer therapy. PAK-1 has also been suggested to be involved in the early stages of breast cancer and may partially participate in the mechanisms mediating the transformation of mammary epithelial cells into mesenchymal malignant cells. 12 Studies have also shown that overexpressed or hyperactivated PAK-1 mediates the anchorage independence of transformed epithelial cells during the progression of breast cancer. 12,13 PAK-1 is also essential for AKT-and Ras-induced oncogenic transformations in both prostate and breast cancer cells. 14,15 In addition to their catalytic active site, most PAKs, including PAK-1, have critical conformations that are required for their functions. These critical conformational sites may prove useful for the design of allosteric small molecule inhibitors whose efficacy do not depend on targeting the catalytic site of PAKs, which are ATP-binding domain common to many kinases. 16 Such precision targeting may lower off-target toxicity and increase specificity. This hypothesis is supported by data derived from an allosteric small molecule inhibitor of group I PAKs called "inhibitor targeting PAK-1 activation-3" (IPA-3). 17 We provided further support for this hypothesis by showing that IPA-3 decreased prostate tumor growth in vitro and in vivo. 11,18 Despite the promising effect of IPA-3 on prostate tumor growth, this compound has some drawbacks that limit its pharmacological potential. The primary one of these is that IPA-3 is metabolically unstable and required daily injections for its efficacy. 11 We addressed this limitation by developing a novel liposomal formulation of IPA-3 using sterically stabilized liposomes (SSL-IPA-3) and determined that these nanoparticles decreased prostate cancer tumor growth in vivo as compared to free IPA-3, with less frequent dosing (every 3 days). 18 Our data also showed the novel finding that both free IPA-3 and that encapsulated in SSL decreased the viability of breast cancer cells. 18 Not surprisingly, the efficacy of both free IPA-3 and SSL-IPA-3 correlated to PAK-1 expression in that IPA-3 was more efficacious at limiting viability in cells with lower levels of PAK-1 compared to high PAK-1 expressing cells. This raised concerns whether the efficacy of IPA-3 was truly dependent on PAK-1 expression.
This study used both pharmacological and molecular approaches to investigate the hypothesis that the efficacy of free IPA-3 and SSL-IPA-3 is dependent on the expression of PAK-1 in diverse cancer cells, including prostate and breast cancer, as well as melanoma cells.
These studies showed that cancer cells with high PAK-1 expression were less responsive to the cytotoxicity of IPA-3 and SSL-IPA-3 compared to cells with low PAK-1 expression. This current study also investigated the ability of PAK-1 gene silencing to alter prostate and breast cancer cell growth and alter markers of EMT. These data demonstrate that the toxicity of IPA-3 is partially mediated by PAK-1 expression, and support the clinical potential of SSL-IPA-3 for the treatment of cancers with altered PAK-1 expression.

| Cell lines and cell culture
The human breast cancer cell lines BT-474, MCF-7, MDA-231, MDA-468, the melanoma-derived cell line, MDA-435, and the immortalized breast epithelial cell line, MCF-10A, were purchased from ATCC. MCF-10A cells were grown in F12/DMEM (50/50) medium, and the rest of the cells were all cultured in RPMI medium. All culture media were supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) penicillin/streptomycin antibiotics (ATCC). The human prostate cancer cell line, DU-145, was purchased from ATCC, and was also maintained in RPMI medium supplemented with 10% FBS and 1% penicillin/streptomycin. All the cells were maintained in a humidified atmosphere at 37°C in incubators with 5% CO 2 . The DU-145 cells were chosen based on our previous studies that showed the cells to be resistant to the activity of both free and encapsulated IPA-3. 18

| Lentiviral transduction
The transduction of PAK-1 knockdown (KD) and control lentiviral

| Immunoblot analysis
Cell lysates from different cell lines were collected in RIPA buffer, which contained a protease inhibitor cocktail (Santa Cruz Biotechnology, Inc). The BCA assay was used to determine protein concentrations. Samples of 40 µg of protein were separated using SDS-PAGE and then transferred to nitrocellulose membranes that were then blocked in 5% (w/v) nonfat dry milk in Tris-buffered saline-Tween 20 (TBST). After 2 hours of blocking, the membranes were incubated with a rabbit PAK-1 antibody (Cell Signaling Technology) at a dilution of 1:1000 in 1% (w/v) BSA TBST overnight. The antibody against GAPDH (Santa Cruz Biotechnology Inc) was used at a dilution of 1:4000 in 1% (w/v) BSA in TBST for 1 hour. Membranes were then incubated with the appropriate peroxidase-conjugated secondary antibodies (Promega,) used at a dilution of 1:2500. The membranes were then washed with TBST three times for 10 minutes each. Bands were visualized using chemiluminescent substrates (Thermo Scientific) and imaged with a FluorChem SP digital Imager (Alpha Innotech). Immunoblot analysis was performed on protein samples from at least three different passages (n = 3) of control and KD cells. Densitometry was performed using National Institutes of Health Image J software.

| Determination of cell growth
Cell growth was measured using crystal violet staining. DU-145 and MCF-7 cells were seeded in 4000 and 10 000 cells, respectively, into each well of 6-well plates in triplicate. Cells were maintained at 37°C for 10 days and media were changed every 3 days. Cells were fixed with 4% formaldehyde for 20 minutes and then stained using 1% (v/v) crystal violet at room temperature for 20 minutes. Excess crystal violet staining solution was then removed and the plates were washed with water and left to air dry at room temperature. Cells were then imaged using a Canon EOS Rebel T3i camera.

| Measurement of cell migration
Cells were seeded in 6-well plates at a density of 1 × 10 6 cells/well under the above conditions. Cell migration was assessed once cells reached a confluency of at least 90% (after 48 hours) using the scratch wound healing assay as previously described. 19 Briefly, a "wound gap" was created in the cell monolayer by scratching using 1-mL pipette tips. The migration ability of the different cells was monitored, imaged, and quantified after 24-72 hours. Cells were fixed with 4% (v/v) formaldehyde for 20 minutes followed by 20 minutes staining with 1% (v/v) crystal violet in PBS. The excess of crystal violet was washed with water three times. Plates were left to air dry and then imaged using an inverted fluorescence microscope equipped with an AxioCam MRc5 digital camera (Carl Zeiss MicroImaging Inc).
The gaps from the scratch were imaged, measured, and normalized to the calculated control wound closure.

Quantification of invasive cells was accomplished by counting cells
in at least three different fields.

| Preparation of sterically stabilized IPA-3 liposomes
Sterically stabilized IPA-3 (SSL-IPA-3) liposomes were prepared as described in our previous study 18 using the thin lipid hydration method followed by freeze-thaw cycles and a high-pressure extrusion. 20,21 Our previous study also described the physical characteristics, stability, and composition of these liposomes. 18 Briefly, cholesterol (5 µmol/mL), phospholipids, including DSPC (9 µmol/mL) and DSPE-PEG (1 µmol/mL) in chloroform, and IPA-3 (4 µmol/mL) in ethanol were added into a round bottom flask, the solvents were then evaporated under vacuum in a water bath at 65°C using a rotary evaporator (Buchi Labortechnik AG). The formed thin film was then hydrated and suspended in PBS to achieve a final lipid concentration of 10 µmol/mL. The formulation then underwent five liquid nitrogen freeze-thaw cycles above the phase transition temperature of the primary lipid, prior to passing five times through a Lipex extruder (Northern Lipids, Inc) at 65°C using double stacked polycarbonate membranes (80 nm, GE Osmonics). Excess unencapsulated IPA-3 and lipids were eliminated using dialysis in 10% (w/v) sucrose for at least 20 hours with three changings of the dialysis media. Liposome suspensions were stored at 4°C, protected from light, and used within 24-48 hours of preparation. Empty SSL (made without IPA-3 encapsulation) were also formulated and used as vehicle controls.
Quantification of IPA-3 was evaluated using methods previously described by us. 18

| MTT staining and cell viability
The MTT assay was used to determine the IC 50 of free IPA-3 in breast and melanoma cells, following treatment with free and lipo- and their respective controls. 22 Cells were seeded in 48-well tissue culture plates at 5 x 10 4 cells/ml and incubated at 37°C in a 5% CO 2 incubator for 24 hours to allow the cells to attach and grow.

| Assessment of annexin V and propidium iodide staining
Cells in which PAK-1 was inhibited were exposed to free IPA-3 and SSL-IPA-3 dosing for 48 hours. This was followed by assessment of annexin V-FITC (marker of apoptosis) and propidium iodide (PI, marker of necrosis) staining using flow cytometry. DU-145 PAK-1 KD and MCF-7 PAK-1 KD cells were seeded and allowed to grow for 24 hours prior to treatment with free IPA-3, SSL-IPA-3, and the controls of DMSO or empty liposomes. Cells were collected and then stained according to the manufacturer protocol using the FITC annexin V apoptosis detection kit (Fisher Scientific). Staining was quantified using a Dako Cyan flow cytometer. For each measurement, 20 000 events were counted. The different populations corresponding to viable and non-apoptotic (annexin V-PI-), apoptotic (annexin V+PI-), and late apoptotic (annexin V+PI+) cells, as well as necrotic cells (annexin V-PI+) (Q4-Q1, respectively) were shown using the plots of annexin V FITC vs PI from gated cells.

| Statistical analysis
All experiments were repeated at least three times (n = 3). Results are shown as the average of all replicates ± SEM. An unpaired twotailed Student's t test was used to compare data sets with normal distribution. A nonparametric test such as the Mann-Whitney test was used if data did not have Gaussian distribution using GraphPad Prism software. The significance level (alpha) was set at .05 (marked with symbols (*) wherever differences are statistically significant).  Figure S1) and IC 50 values were estimated from the dose-response curves ( Figure 1C,D). There was an excellent correlation between the expression of PAK-1 and the IC 50 of free IPA-3 ( Figure 1D).

| PAK-1 knockdown inhibited the migration and invasion of DU-145 and MCF-7 cells
We used the scratch and Transwell ® assays to assess the role of

| PAK-1 knockdown altered the expression of other proteins
We further assessed the effect of PAK-1 inhibition on DU-145 and MCF-7 cell growth by determining differences in the expression of E-cadherin, N-cadherin, CXCR-4, and HCAM using immunoblot analysis ( Figure 6). These proteins were chosen as many studies have shown their involvement in cancer proliferation and progression. 23

| PAK-1 inhibition altered the toxicity of IPA-3 to DU-145 and MCF-7 cells
We studied the effect of PAK-1 inhibition on the activity of free IPA-3 and IPA-3 encapsulated in liposomes (SSL-IPA-3) in both DU-145 and MCF-7 cells using the MTT assay. As shown in Figure 7A and MCF-7 PAK-1 KD cells were exposed to SSL-IPA-3; however, the level of staining was not as high as that seen in free IPA-3 treated cells. Increases in PI staining were only seen at the highest doses of IPA-3 and SSL-IPA-3 used, suggesting that the primary mechanism of cell death was apoptosis.

| D ISCUSS I ON
PAK-1, a serine/threonine kinase, was suggested to play a major role in various cancers. 27 Studies have even suggested that PAK-1 is a tumor oncogene and a therapeutic target for the treatment of cancer. 13,15 Compared to prostate cancer, there are limited studies on the role of PAK-1 in other cancers. Studies that used mouse knockout models demonstrated a role of PAK-1 in cell migration. 2 Data from our study support that PAK-1 mediated cell growth in several different cancer cell lines, including those derived from breast and melanoma. Our previous study also suggested that prostate cancer cells with the highest expression of PAK-1 (DU-145) as well as the breast cancer cells (MCF-7) were the least susceptible to IPA-3 18, raising the concern that the toxicity of IPA-3 may not be dependent on PAK-1 expression. We tested this possibility using a dual approach of pharmacological and molecular inhibition. As expected, both approaches decreased the growth of prostate and breast cancer cells. These data support the conclusion that toxicity of IPA-3 to cancer cells is mediated in part by PAK-1, and support the hypothesis that PAK-1 acts as an oncogene in not only prostate cancer cells, 15 but extends this hypothesis to breast cancer cells and melanoma.
Data from this study also demonstrated that PAK-1 inhibition decreased the expression of E-cadherin, CXCR-4, and HCAM. E-cadherin plays a crucial role in the epithelial adherens junctions, where several proteins interact, including α-and β-catenin, to mediate the actin cytoskeleton. 28 Inhibition of E-cadherin expression is required for EMT and plays a role in cancer migration and metastasis. 29 These data align with previous reports showing that PAK-1 regulates cytoskeletal organization and cell-cell interactions. 30 These data may also explain why PAK-1 inhibition changes the morphology of DU-145 cells. CXCR-4 is a G-protein-coupled receptor that has been shown to be overexpressed in multiple cancers and is involved in cell adhesion, survival, and growth. 31 Studies have shown that suppression of CXCR-4 in vitro inhibited cell invasion. 32 Our data agree with these findings and report the novel finding that PAK-1 regulates the F I G U R E 9 Effect of PAK-1 inhibition on the efficacy of IPA-3 and SSL-IPA-3 in MCF-7 PAK-1 KD cells using annexin V and PI staining. referred to as CD44 antigen, is a cell-surface glycoprotein that has been shown to be involved in cell adhesion, cellular interactions, and migration and was suggested as a potential diagnostic and prognostic marker of malignancy in breast and ovarian cancers. [33][34][35] Our data show that the expression of HCAM was inhibited following PAK-1 inhibition. To our knowledge, this is the first report that PAK-1 may mediate the expression of HCAM in any cell type.
PAK-1 expression is increased during the early stages of human breast cancer progression. 12,36 Studies suggest that PAK-1 overexpression can predict tumor recurrence and resistance to tamoxifen, which is a selective estrogen receptor modulator commonly used for the treatment of hormone-receptor-positive, early stage breast cancer. 37  but are now believed to be melanoma in origin. However, these cells are also ER-negative 38 and had a relatively low IC 50 as compared to ER-positive cells. The higher IC 50 in MCF-10A is interesting given that these cells are ER-negative, however, these are not cancer cells, and are considered models for normal breast cell function.
The overexpression of PAK-1 is believed to be responsible for the phosphorylation of the estrogen receptor, creating promiscuous phosphorylated receptors resistant to tamoxifen treatment. [39][40][41] Our studies further support the link between ER expression and PAK-1 expression in breast cancer cells and suggest a link between the efficacy of PAK-1 inhibitors and ER status.
We previously showed that IPA-3, in both free and in liposomal forms, inhibits prostate cancer growth in vitro and in vivo. 18 However, the efficacy of IPA-3 correlated to the expression of PAK-1, with cells that express higher levels of PAK-1 demonstrating decreased susceptibility. 18 This phenomenon extended to liposome-encapsulated IPA-3 in that SSL-IPA-3 did not decrease cell viability in DU-145 cells in our previous studies at any dose. 18 This is of concern because it suggests that these liposomal formulations may not be as efficacious in the clinic in high-grade cancers that overexpress PAK-1. This also suggests the possibility that the efficacy of both IPA-3 and SSL-IPA-3 was not specific to PAK-1. Data in this study definitely show that PAK-1 KD alters the toxicity of cells to IPA-3, however, the data do not unequivocally show that IPA-3 toxicity is totally dependent on PAK-1. It is possible that IPA-3 may also inhibit other PAK's, such as PAK-2 at higher concentrations. 17,42,43 The loss of PAK-1 in MCF-7 cells may have also altered the kinetics of inhibition. Further studies are needed to confirm this hypothesis.
The mechanisms mediating the increased sensitivity of prostate and breast cancer cells to SSL-IPA-3 during PAK-1 inhibition are probably not drastically different than those mediating the toxicity of free IPA-3. Most likely, the decrease in PAK-1 is shifting the dose curve to the left, essentially increasing the potency of IPA-3. This hypothesis is supported by the fact that PAK-1 inhibition resulted in similar mechanisms of cell death (apoptosis) that was seen in free IPA-3, as determined using flow cytometry. Alterations in annexin V and PI staining also confirm the morphological data and data derived from MTT staining.
While our data suggest that PAK-1 is a promising potential therapeutic target for some cancers, the IC 50 value of IPA-3 (~15 µM) is not optimal for clinical translation. This is one reason that liposomal encapsulation was used for IPA-3, which is typically limited by its stability. Nevertheless, the data also suggest that more studies are needed to develop more potent PAK-1 inhibitors. Such studies are already under way in our laboratory.
In summary, our data show that the pharmacological effect of IPA-3 is mediated, in part, by PAK-1, demonstrate some of the first data suggesting that IPA-3 is a potential therapeutic treatment for breast cancer and melanoma, and demonstrate the efficacy of liposome-encapsulated IPA-3 in breast cancer cells. This is, as far as we know, the first report of a direct correlation between PAK-1 expression and efficacy of IPA-3 in breast cancer.

ACK N OWLED G EM ENTS
This project was supported in part with funds from the National

D I SCLOS U R E S
The authors declare no conflict of interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.