Introduction

The p21-activated protein kinases (PAKs) are serine/threonine protein kinases activated by binding to the Rho family small GTPases, Rac, and Cdc42. PAKs are categorized into two subgroups (A and B) based on their amino acid sequences and functions [1]. Group B PAKs (PAK4, 5, and 6) do not contain the highly conserved auto-inhibitory domain that is found in the Group A PAKs members (PAK1, 2, and 3). Thus, while Group A members have 80-90% sequence identity within their catalytic domains, Group B PAKs have only approximately 50% identity to the kinase domains of the Group A PAKs [2, 3]. Unlike Group A PAKs, which are activated upon binding to Cdc42 and Rac, Group B PAKs are not strongly activated by GTPase binding [4].

Apart from playing an important role in cytoskeleton organization and cell morphology by regulating the actin cytoskeleton [5–10], PAKs have also recently been shown to regulate apoptosis. Both group A and group B PAKs have been shown to play key roles in regulating the apoptotic response, although their effects may vary, depending on the circumstances and type of cells. For example, PAK2 has pro-apoptotic effects in cells that are sensitive to signaling through the c-Jun N-terminal kinase (JNK) and p38 mitogen-activated protein (MAPK) pathways [11]. It is likely that PAK2 is cleaved by caspase-3 during apoptosis, activating PAK2 by releasing its kinase domain. The activated kinase is thought to contribute to membrane and morphological changes that occur during apoptosis [12–14]. On the other hand, PAK1 and PAK4 have been reported to protect cells from apoptosis without cleavage by caspases. PAKs have also been shown to increase the phosphorylation of Ser112 and Ser136 of Bad [15–17], to activate MAPK pathways [18], and to block the caspase cascade by acting upstream of the mitochondrial pathway and effector caspases [18]. Overexpression of PAK5 activates JNK pathway to decrease apoptosis [19]. In addition, it is found that PAK5 shuttles from the mitochondria to the nucleus, and that the mitochondrial localization of PAK5, which leads to phosphorylation of Ser112 and Ser136 of Bad, is vital to its effects on cell survival [20, 21]. PAK5 can also activate Raf-1, an important effector of Ras-mediated signaling, and target it to the mitochondria, regulating its kinase activity and controlling Raf-1 dependent signaling at mitochondria [22].

Given the fact that PAK5 can protect cells from apoptosis, we reason that PAK5 may be associated with human carcinogenesis. In a previous study, we found that PAK5 expression increased with colorectal carcinoma (CRC) progression from normal colon mucosa to metastatic carcinoma, with remarkable increases in invasive and metastatic CRCs. Furthermore, we found that PAK5 decreased CRC cell adhesion but promoted their migration, which may contribute to CRC metastasis [23]. However, the functional role of PAK5 in CRC progression remains to be further studied. In this study, we aim to examine whether PAK5 can protect chemotherapeutic drug-induced apoptosis and to uncover the signaling pathway involved. Our results showed that overexpression of PAK5 inhibited camptothecin-induced apoptosis in SW480 cells, while knockdown of PAK5 expression resulted in increased apoptosis in Lovo cells. Furthermore, we found that overexpression of PAK5 inhibited activity of Caspase-8. Consistent with the previous study, we also found that PAK5 directly phosphorylated Bad on serine 112 and indirectly led to phosphorylation of serine 136 via the Akt pathway in CRC cells. Taken together, in this study, we present data indicating that PAK5 inhibits apoptosis induced by chemotherapeutic agent and provide evidence of its potential mechanism of action.

Materials and methods

Cell transfection and selection

PEGFP-C3-PAK5 plasmid was constructed as described previously [23]. The PAK5-shRNA plasmid (pGFP/Neo-PAK5) and empty plasmid vector (pGFP/Neo-control) were purchased from GenePharma (Shanghai, China). Cells were seeded in six-well plate with 40–60% confluence 20 h before transfection and PEGFP-C3-PAK5 or control PEGFP-C3 were transfected by Lipofectamine 2000 (Invitrogen, Guangzhou, China) according to the user’s protocol. For RNA interference, PAK5-shRNA and control vector (GenePharma, Shanghai, China) were transfected into the targeted cells by Oligofectamine (Invitrogen) according to the user’s protocol. To make stable cell line, transfected cells were selected by 1,000 μg/ml G418 (Merk, Guangzhou, China) for 2 weeks and positive clones were confirmed by Western blotting.

Reagents and antibodies

AKT Inhibitor IV was obtained from Merck. The anticancer agents used in this study were oxaliplatin (Sigma), 5-fluorouracil (5-Fu, KeyGEN), and camptothecin (Sigma). Anti rabbit PAK5 antibody was made as described previously [23]. Rabbit polyclonal anti-Bad, anti-phospho Ser112-Bad, anti-phospho Ser136-Bad (185D10), mouse monoclonal anti-Caspase-8 (1C12), rabbit monoclonal anti-poly (ADP-ribose) polymerase (PARP; 46D11), mouse monoclonal anti-GFP (4B10), and anti-phospho Ser 473-AKT antibodies were from Cell Signaling. The anti-actin antibody and secondary antibodies conjugated to horseradish peroxidase were from Santa Cruz Biotechnology.

Apoptosis assays

Cells were plated in six-well plates at 70–80% confluence 24 h before exposure to anticancer agents. DMSO stock solutions of anticancer agents were diluted with cell culture medium then added to cells. Cells were exposed to anticancer agents for 0, 12, 24, or 48 h, then were collected for flow cytometry or used to prepare lysates. Cells were collected by trypsinization, combined with the floating cells, and then labeled with Annexin V-PE and 7-AAD Viability Staining Solution according to the manufacturer’s recommendations (BioVision) for fluorescence-activated cell sorting analysis. For analysis of apoptosis by nuclear staining with Hoechst 33258 (Sigma), cells were treated as above, washed once with phosphate-buffered saline (PBS) and then fixed with ice-cold methanol (500 ml/well) for 10 min. After fixing, cells were washed twice with PBS, stained with 1 mM Hoechst 33258 for 10 min, and then extensively washed twice with PBS and distilled water. Apoptosis was indicated by the presence of condensed or fragmented nuclei which bound the Hoechst dye with high affinity.

For analysis of cell morphology by photo-microscopy, coverslips were washed once with PBS and then inverted and mounted onto glass slides. Cells were visualized under an Olympus microscope. Two hundred cells in three randomly chosen fields were counted and scored for the incidence of apoptotic chromatin changes under fluorescence microscopy.

Western blotting analysis

Cells were exposed to 10, 50, or 100 μM camptothecin for 0, 12, 24, or 48 h, and then collected for Western blotting analysis. Cells were washed twice with ice-cold PBS then lysed with ice-cold lysis buffer (20 mM Tris-HCl, 1 mM EDTA, 1 mM EGTA, 1 mM sodium vanadate, 0.2 mM phenylmethylsulph-ony1 fluoride, 0.5% NP-40, 1 mg/ml leupeptin, 1 mg/ml aprotinin, 1 mg/ml Pepstatin), and PMSF (1:100) for 30 min on ice. For the detection of phosphorylated Bad, cells were lysed with ice-cold lysis buffer supplemented with protease and phosphatase inhibitors. The lysates were transferred to 1.5 ml Eppendorf tubes to incubate on ice for 30 min, with occasional vortexing. Insoluble material was removed by centrifugation. Soluble protein was removed to a new tube and stored at −20°C until required. Protein samples (30-50 μg) were separated by SDS/PAGE (10% acrylamide gel) using a Bio-Rad Mini-Protean system (100 V for about 3 h). Proteins were transferred to PVDF membranes, 4 MA/cm2 for 1 h. Then the membranes were blocked for 1 h at room temperature with 5% (w/v) skim milk powder in Tris buffered saline-Tween 20 (0.1% by volume, TBS-T). The membranes were then incubated for 1-2 h at room temperature with primary antibodies in 5% (w/v) skim milk powder dissolved in TBS-T (1:500 dilution). Primary antibodies were removed, and the membranes were extensively washed with TBS-T four times for 20 min. The membranes were then incubated for 1 h at room temperature with the secondary antibodies (goat anti-mouse or anti-rabbit antibody coupled to horseradish peroxidase, 1:3,000-5,000 dilution) in 5% (w/v) skim milk powder dissolved in TBS-T. Following removal of the secondary antibody, the membranes were extensively washed as above for an hour and developed using the Enhanced Chemiluminescence Detection System (Pierce).

Statistical analysis

All results were confirmed in multiple independent experiments, with each time point or condition assayed in triplicate within each experiment. Data were expressed as means ± SD and analyzed by one-way ANOVA, followed by Dunnett’s multiple comparison tests, with significance defined as P < 0.05.

Results

Overexpression of PAK5 inhibits camptothecin-induced apoptosis

To investigate the function of PAK5 in apoptosis, we up-regulated PAK5 expression in CRC cell lines by gene transfection. We observed that exogeneous PAK5 protein increased significantly in SW480 cells transfected with the PEGFP-C3-PAK5 plasmid compared to untransfected cells and cells transfected with the empty plasmid vector PEGFP-C3 (Fig. 1a). Nuclear morphological analysis revealed that exposure of cells to oxaliplatin or 5-FU induced apoptosis to a similar extent in the PAK5 overexpressing SW480 cells as it did in the parental cells. However, apoptosis following exposure to camptothecin was lower in the PAK5 overexpressing SW480 cells compared to the parental cells (Fig. 1b). To further examine the effect of PAK5 overexpression on the induction of apoptosis by camptothecin, parental, and PAK5-overexpressing SW480 cells were exposed to various concentrations of the agent (10, 50, 100 μM) for different durations (0, 12, 24, 48 h), and the extent of apoptosis were analyzed by flow cytometry. Analysis of apoptosis by AnnexinV-PE staining confirmed that camptothecin-induced apoptosis was abrogated by the overexpression of PAK5 in SW480 cells (Fig. 1c), with a significant decrease in apoptosis observed in cells exposed to 100 μM camptothecin for 48 h compared with empty vector SW480 cells and parental cells (Fig. 1d).

Fig. 1
figure 1

Chemotherapeutic agents-induced apoptosis in SW480 cells. a Overexpression of PAK5 in SW480 cells. SW480 cells were transfected with PEGFP-c3-PAK5 or empty plasmid vector PEGFP-c3, and PAK5 expression was assessed by Western blotting. b Assessment of apoptosis by Hoechst 33258 staining. Equal numbers of SW480 parental cells, empty vector-transfected cells, or cells overexpressing PAK5 were plated in six-well plates. The cells were exposed to oxaliplatin (120 μM), 5-Fu (50 μM), or camptothecin (100 μM) for 48 h, then subjected to fluorescence microscopy analysis. c Flow cytometric analysis of apoptosis induced by camptothecin. Equal numbers of stably transfected SW480 cells were plated in six-well plates 24 h before exposure to 100 μM camptothecin for 0, 12, 24, or 48 h. After treatment, cells were collected and evaluated by flow cytometric analysis of Annexin V-PE and 7-AAD Viability Staining. The percentages of positive cells were labeled. d SW480 cells were exposed to various concentrations of camptothecin (10, 50, and 100 μM) for 48 h, then evaluated for apoptosis. These results reflect the data from three independent experiments. *p < 0.05

Full size image

Camptothecin-induced activation of caspase-8 and PARP is inhibited in PAK5-overexpressing SW480 cells

Since PAK5 could protect cells from camptothecin-induced apoptosis as detected by flow cytometry; next, we were interested in determining what steps in the camptothecin-induced death pathway are abrogated by PAK5. We first examined the possibility that PAK5 could affect initiator caspases and thereby function upstream of, or independently of, the mitochondrial pathway. To test this possibility, cleavage of the initiator caspase-8 and the effector PARP was examined following camptothecin treatment. As expected, after a 24-h exposure to camptothecin, considerable amounts of intermediate and active caspase-8 fragments (p43/41) could be detected in the control cells (Fig. 2a). Activation of PARP was observed to follow a similar time course (Fig. 2b). In contrast, in cells overexpressing PAK5, cleavage of both caspase-8 and PARP was dramatically reduced and delayed (Fig. 2a and b). Caspase-8 and PARP both remained completely uncleaved after 24 h in the presence of camptothecin, and they were only partially cleaved at 48 h, indicating that induction of caspase-8-like activity was delayed in the PAK5-overexpressing SW480 cells. These results confirm the decrease (and the delay in induction) of apoptosis in cells overexpressing PAK5 (Fig. 2).

Fig. 2
figure 2

PAK5 inhibits Camptothecin-induced activation of caspase-8 and PARP in SW480 cells. SW480 cells were exposed to 100 μM camptothecin for various times (0, 12, 24, and 48 h), then assessed for cleavage of caspase-8 (a) and PARP (b) by Western blotting. The results presented were representative of three different experiments

Full size image

Camptothecin-induced apoptosis is increased in PAK5-deficient LoVo cells

To investigate whether the effects of PAK5 on apoptosis are exclusive to SW480 cells, or whether PAK5 exerts similar effects in other cell lines, LoVo cells were transfected with a PAK5-shRNA plasmid (Fig. 3a). When the PAK5-shRNA LoVo cells were exposed to 100 μM camptothecin for various times, there was an increase in apoptosis compared to parental cells (Fig. 3b). After 24 h of stimulation, considerable amounts of intermediate and active caspase-8 fragments (p43/41 and p18) could be detected in the PAK5-siRNA LoVo cells, and similar effects were observed for PARP activation (Fig. 3c).

Fig. 3
figure 3

Knockdown of PAK5 promotes camptothecin-induced apoptosis as well as activation of caspase-8 and PARP in LoVo cells. a shRNA mediated knockdown of PAK5 in LoVo cells. Twenty-four hours after transfection of LoVo cells with PAK5-shRNA, PAK5-NC, GAPDH, or control, stably transfected cells were selected by 1,000 μg/ml G418 for 2 weeks and expression of PAK5 was examined by Western blotting (*p < 0.05). b Apoptosis in PAK5 knockdown LoVo cells. Equal numbers of stably transfected LoVo cells were plated in six-well plates 24 h before treatment with 100 μM camptothecin for 48 h. The extent of apoptosis was evaluated by flow cytometric analysis of Annexin V-PE and 7-AAD Viability Staining. c Activation of caspase-8 and PARP in PAK5 knockdown LoVo cells. Control and PAK5-siRNA LoVo cells were exposed to 100 μM camptothecin for various times (0, 12, 24, or 48 h) then assessed for cleavage of caspase 8 and PARP by Western blotting. The results presented were representative of three different experiments

Full size image

PAK5 phosphorylates Bad

Because other members of the PAK family have been shown to phosphorylate Bad, we examined the effect of PAK5 on the phosphorylation status of Bad in SW480 cells transfected with either empty vector PEGFP-C3 or PEGFP-C3-PAK5. After treating the cells with 100 μM camptothecin for 48 h, we observed that PAK5 overexpression leads to an increased in the phosphorylation of Bad specifically on serine 112 and 136 (Fig. 4a). However, when cells were exposed to the AKT Inhibitor IV for 48 h, there was no increase in the phosphorylation of Ser 136 (Fig. 4b and c). These experiments indicate that PAK5 may directly increase the phosphorylation of Ser 112, but indirectly increase the phosphorylation of Ser 136 via the Akt pathway.

Fig. 4
figure 4

PAK5 regulates Bad phosphorylation both directly and indirectly. a Parental, GFP control, GFP-PAK5, and PAK5-siRNA SW480 cells were treated with 100 μM camptothecin for 48 h, and total Bad and the phosphorylation of Bad (on Ser 112 and 136) were examined by Western blotting. Right panel: Photodensitometric analysis of relative amount of phosphorlated Bad with actin as internal control. The Bad/Actin ratio was set as 100 in parental SW480 cells. b Parental, GFP control, GFP-PAK5 and PAK5-siRNA SW480 cells were plated in six-well plates at 60–70% confluence 24 h before a 48 h treatment with Akt Inhibitor IV, and total Bad and the phosphorlation of Bad (on Ser 112 and 136) were examined by Western blotting. Right panel: Photodensitometric analysis of relative amount of phosphorlated Bad with actin as internal control. The Bad/Actin ratio was set as 100 in parental SW480 cells. c PAK5-overexpressing SW480 cells were incubated in the presence or absence of 50 μM AKT Inhibitor IV for 48 h, then expression of total (general) and phospho-AKT was evaluated by Western blotting. The results in A and B reflect the data from three independent experiments; *p < 0.05; c was representative of three different experiments

Full size image

Discussion

PAK5 is a recently identified member of group B PAK kinase family. Group B PAKs are highly expressed in the brain [13, 14] and are localized in the mitochondria where they inhibit apoptosis by phosphorylating BAD in CHO cells [17, 18]. PAK5 shares some common activities with other group B PAKs. Like PAK4, PAK5 can confer protection against apoptotic stimuli and has the ability to phosphorylate Bad on Ser-112 [18, 20]. Both PAK1 and PAK5 interacted with Raf-1 and phosphorylated it at Ser 338, which targeted it to the mitochondria [22, 24]. In our prior study, we found that PAK5 was overexpressed in a variety of CRC cells, and that PAK5 expression was increased with CRC progression from the adenoma to carcinoma, with the highest expression observed in highly invasive and metastatic CRCs [23]. Expression of PAK5 was also found to be increased during the development of CRC from lower to higher Duke’s stages. Moreover, the expression of PAK5 was increased from well to poorly differentiated CRCs. These findings imply that PAK5 plays an important role in the progression of CRC via its regulation of apoptosis and cell survival. Therefore, the aim of the present study was to determine whether PAK5 can protect cells from apoptosis and the signal pathway involved.

To determine whether PAK5 can protect cells from apoptosis, stably transfected cell lines with PEGFP–c3 empty vector, PEGFP-c3-PAK5 or PAK5-shRNA plasmids were developed. These cell lines provide a good model system to study the function of PAK5 in apoptosis and cell survival. In this study, we determined the response of the SW480 cells overexpressing PAK5 and the parental cells to a variety of anticancer agents, including oxaliplatin, 5-FU, and camptothecin. Our results showed that the oxaliplatin/5-FU-induced apoptosis exhibited no significant difference between the PAK5 overexpressing cells and the parental cells after the same treatment. However, a moderate decrease in camptothecin-induced apoptosis was observed in the PAK5 overexpressing cells compared with parental cells. Interestingly, previous study showed that SW480 cell lines had low sensitivity to 5-Fu-induced apoptosis and high sensitivity to camptothecin-induced apoptosis [25], our present study further demonstrated that apoptosis induced by camptothecin was significantly affected in PAK5-overexpressing SW480 cells, suggesting that PAK5 activity was negatively correlated with camptothecin-induced apoptosis.

Our results showed that the overexpression of PAK5 in SW480 cells led to the inhibition of camptothecin-induced apoptosis and cleavage of the initiator caspase-8 and the effector PARP. Conversely, shRNA-mediated deletion of PAK5 in LoVo cells strongly increased camptothecin-induced apoptosis and cleavage of initiator caspase-8 and the effector PARP. Thus, PAK5 appears to be a novel regulator responsible for cellular resistance to apoptosis. It is still unclear how PAK5 inhibits caspase-8 activity and whether these effects are direct or indirect. The most attractive possibility is that PAK5 mediated phosphorylation of Bad promotes the interaction between Bad phosphorylated at either Ser-112 or Ser-136 and 14-3-3 proteins. As a result, Bad is sequestrated in the cytoplasm and could not translocate into the mitochondria to trigger the apoptosis pathway and the activation of caspases including caspase-8. However, we could not rule out the possibility that PAK5 functions by causing changes in the protein level of FLIP, which can affect the activity of caspase-8 [26–28]. On the other hand, PAK1 has been shown to promote osteoclast survival by modulating expression of the IAP family member survivin and the expression of survivin was found to be positively correlated to clinicopathologic parameters of colorectal carcinoma [29, 30]. It is therefore attractive to speculate that PAK5 may upregulate the expression of survivin to antagonize camptothecin-induced apoptosis. Regardless of the mechanism, the regulation of caspase-8 activity demonstrated here is a new function of PAK5, and our results provide novel insights into the mechanism by which colorectal carcinoma cells develop resistance to apoptosis.

Current data support the concept that PAK5 appears to induce resistance to apoptosis via the phosphorylation of Bad. While PAK1 and PAK4 have been reported to directly phosphorylate Bad on Ser 112 and 136, and perhaps additional Ser residues [13, 17, 18, 31], our data suggest that the PAK5 phosphorylation of BAD at Ser112 was a direct action. In contrast, the PAK5-stimulated phosphorylation of BAD at Ser136 may be mediated through Akt. Future studies will examine the relationship between PAK5 and the Akt pathway, and how their expression and activity affect apoptosis and cell survival.

In summary, here we demonstrate that PAK5 is involved in cellular resistance to apoptosis, and that this effect likely occurs via alterations in Bad phosphorylation leading to a decrease in the activity of caspase 8. Further examination of PAK5 level and activity in human tumor and normal tissues will help determine whether PAK5 represents a novel target to improve the response to conventional cancer therapies.