PAK4 suppresses TNF-induced release of endothelial microparticles in HUVECs cells

TNF induces EMP release both in HUVECs and in mice

First, we analyzed the ability of TNF to induce the EMP release in vitro and in vivo. In vitro, the ability of TNF to induce EMP was analyzed in human umbilical vein endothelial cells (HUVECs) by flow cytometry using annexin V-FITC, which binds to phosphatidylserine exposed on EMP. As shown in Figure 1A, TNF between 20 ng/mL and 100 ng/mL, dose-dependently induced the EMP release in comparison with untreated control cells; stimulation with a serine protease, thrombin, was used as a positive control. The level of released EMP in response to thrombin was about half of that observed in response to 100 ng/ml TNF (Figure 1A). The release of EMP was also time dependent, with a significant increase between 3 and 24 hours (Figure 1B). Thus, the following experiments with HUVECs were performed using a 24 h stimulation with 100 ng/mL TNF.

Figure 1. TNF induces EMP release in vitro and in vivo. (A) EMP release analyzed by flow cytometry (FC) in HUVECs treated 24 h with increasing TNF concentrations, or thrombin (2 IU/ml). The data represent the mean values based on 1x10⁶ cells; three independent experiments were performed. (B) EMP release analyzed by FC in untreated HUVECs, or HUVECs treated with 100 ng/ml TNF for indicated times. The data represent the mean values of EMP per 1x10⁶ cells from three independent experiments. (C) EMP release analyzed by FC in sera of TNF-treated mice. Male C57BL/6 mice (10 weeks old) were randomly assigned to 5 groups (n=10 mice per group) and treated with PBS or increasing doses of TNF for 24 hours. The data represent the mean values of EMP/μl; * P<0.05, ** P<0.01, *** P<0.001, NS indicates no significance.

Next, we investigated whether TNF induces EMP release also in vivo. Male C57BL/6 mice (10 weeks old) were randomly assigned to 5 groups (n=10) and treated 24 h with PBS or increasing doses of TNF (4, 20, 100, or 500 μg/kg). As shown in Figure 1C, EMP in mice sera increased significantly when treated with 20 and 100 μg/kg TNF.

TNF induces cytoskeleton rearrangement and down-regulates PAK4

EMP shed by endothelial cells are associated with cytoskeleton disorders [15]. Thus, PAK4 as a regulator of cytoskeletal dynamics and cell motility [19] may also play a role in EMP generation. To test this hypothesis, we treated HUVECs with 100 ng/ml TNF for 24 hours, and stained the cells with Phalloidin-FITC or anti-PAK4 polyclonal antibody. Without TNF treatment, the F-actin bundles mainly localized at cell border, but in TNF-treated cells, the F-actin bundles were throughout the cells, and the dense peripheral actin band was replaced by thicker stress fibers (Figure 2B, lower panel). In untreated cells, PAK4 could be observed at the cell-cell contacts, while after TNF stimulation, PAK4 staining could hardly be detected (Figure 2B, upper panel). Immunoblotting indicated a time dependent decrease in the approximately 64 kDa PAK4 protein band, and the upstream cdc42 protein (Figure 2A). Although F-actin rearrangement was observed by confocal microscopy, TNF did not change the total levels of β-actin (Figure 2A); tubulin was used as a loading control. These data demonstrate that TNF down-regulates PAK4 in HUVECs.

Figure 2. TNF down-regulates PAK4 and cdc42 expression in HUVECs. (A) Western blotting of HUVECs incubated with 100 ng/ml TNF for up to 24 hours, analyzed using anti-PAK4, anti-cdc42, anti-β-actin, and control anti-tubulin antibodies. (B) Confocal microscopy of HUVECs incubated 0 and 24 h with 100 ng/ml TNF, and stained for PAK4 and F-actin; scale bar = 10 μm. Cells were counter stained with DAPI.

To investigate how TNF down-regulates the PAK4 levels, HUVECs were treated with TNF alone, or with TNF + soluble TNFR1, the soluble part of TNF receptor 1 that interacts with TNF and inhibits its binding to TNFR1 on the cell membrane. After 6 h stimulation with TNF, the PAK4 expression decreased, both at protein and mRNA levels (Figure 3A left panel, and Figure 3B). However, soluble TNFR1 stabilized the PAK4 protein levels in TNF-treated cells (Figure 3A, right panel). In addition, TNF induced a rapid degradation of IκBα, the inhibitor of NF-κB, in HUVECs (Figure 3A, right panel), but this degradation was prevented by the soluble TNFR1 (Figure 3A, left panel). These results indicate that TNF decreases PAK4 expression through the TNFR1-dependent pathway that might also involve NF-κB signaling.

Figure 3. TNF decreases PAK4 gene and protein levels in HUVECs. (A) Western blot analysis of PAK4 and IκBα in HUVECs stimulated with 100 ng/ml TNF without and with 300 ng/ml soluble TNFR1. (B) Quantitative real-time PCR of PAK4 mRNA in HUVECs treated with 100 ng/ml TNF. All data are expressed as the mean + SD; *** P<0.001, compared with untreated cells.

Involvement of PAK4 in TNF-induced EMP release

To investigate the role of PAK4 in TNF-induced EMP generation, we first inhibited PAK4 function in TNF-stimulated (100 ng/ml) HUVECs by PAK4 inhibitor PF-3758309 or PAK4 siRNAs. PF-3758309 is an ATP-competitive, small-molecule pyrrolopyrazole inhibitor, blocking the PAK4 kinase activity and downstream pathways [24]. HUVECs were treated with 10 μM PF-3758309 overnight or transfected with PAK4 siRNAs for 48 h, before TNF stimulation. To stably knockdown PAK4 and avoid off-target effects, two distinct small interfering RNAs targeting PAK4 were used. As shown in Figure 4A, PF-3758309 had no obvious effect on PAK4 expression, but PAK4 specific siRNAs effectively reduced the PAK4 protein levels in HUVECs, with an estimated efficiency of ~85% knockdown (immunoblotting quantification using ImageJ). Immunofluorescent staining of PAK4 in PF-3758309 or PAK4 siRNAs treated cells showed similar results (Figure 4B).

Figure 4. PAK4 inhibition sensitizes HUVECs to TNF-induced EMP release. (A) Western blotting of HUVECs transfected 48 h with PAK4 or control siRNAs, or treated 24 h with PAK4 inhibitor PF-3758309, and analyzed by PAK4 and control tubulin antibodies. (B) Immunofluorescence of HUVECs transfected with PAK4 or control siRNAs, or treated with PF-3758309, and stained by anti-PAK4 antibody; scale bar = 10 μm. (C) EMP release measured by FC in HUVECs transfected 48 h with PAK4 or control siRNAs, or treated 24 h with PF-3758309, and stimulated 24 h with TNF (100 ng/ml). (D) Time course of EMP release in TNF-treated HUVECs transfected 48 h with PAK4 or control siRNAs, or incubated 24 h with PF-3758309; * P<0.05, ** P<0.01.

Without TNF stimulation, PF-3758309 or PAK4 siRNAs treatment did not show any significant effect on the basal EMP release, compared to siRNA control (Figure 4C, lanes 1, 2, 3). TNF (24 h) stimulation significantly increased the EMP release, but this release was not substantially affected by PF-3758309 or PAK4 suppression by siRNA (Figure 4C, lanes 4, 5, 6). Next, we examined the time dependent EMP release upon 100 ng/ml TNF stimulation in PF-3758309 or PAK4 siRNAs treated HUVECs. Three hours after TNF treatment, PAK4 knockdown (siPAK4+TNF) or PAK4 inhibition (PF-3758309+TNF) markedly increased the EMP release (Figure 4D, blue and orange curves vs purple curve, P<0.01), indicating that PAK4 knockdown or inhibition sensitizes HUVECs to TNF-induced EMP release.

Overexpression of PAK4 decreases TNF-induced EMP release

To determine whether activation of PAK4 decreases the TNF-induced EMP release, we transiently overexpressed GFP-tagged PAK4 in HUVECs. Western blot analysis revealed a highly expressed GFP-PAK4 protein band in addition to the endogenous PAK4 band (Figure 5A). Overexpression of GFP-PAK4 reduced the TNF-induced EMP release (Figure 5B, lane 5 vs lane 3, P<0.01), but had no effect on the basal EMP release (Figure 5B, lane 2 vs lane 1). However, in contrast to PAK4, the PAK4 kinase mutant (GFP-PAK4KM) did not significantly decrease the TNF-induced EMP release (Figure 5B, lane 8 vs lane 3). Pre-incubation of HUVECs with z-VAD-FMK (2 μM), a pan-caspase inhibitor, was sufficient to abrogate the EMP release induced by TNF (Figure 5B, lane 7 vs lane 3, P<0.01). The effect of z-VAD-FMK on TNF-induced EMP release was similar to the overexpressed PAK4 (Figure 5B, lane 7 vs lane 5). Overexpression of GFP-PAK4 also markedly prevented the EMP release induced by TNF in PAK4 knockdown cells (Figure 5C, lane 6 vs lane 4, P<0.01).

Figure 5. PAK4 overexpression decreases TNF-induced EMPs release. (A) Western blotting of HUVECs transfected 48 h with GFP, GFP-PAK4 plasmid, siPAK4 or GFP-PAK4+siPAK4, and analyzed by PAK4 and control tubulin antibodies. (B) EMP release in HUVECs transfected with GFP, GFP-PAK4, GFP-PAK4KD, siPAK4 or GFP-PAK4+siPAK4, and stimulated 24 h with TNF (100 ng/ml) or control vehicle. Overexpression of GFP-PAK4 significantly inhibits TNF-induced EMP release (lane 5 vs lane 3). The TNF-induced EMP is also inhibited by caspase inhibitor z-VAD (lane 7). PAK4 kinase mutant (GFP-PAK4KM) does not reduce the TNF-induced EMP release (lane 8). The values represent the mean of three independent experiments. ** P<0.01; NS, not significant.

PAK4 prevents TNF-induced caspase activation and apoptosis

Caspases mediate the membrane blebbing associated with apoptosis [25]. Since the TNF-induced EMP release was reduced by the caspase inhibitor z-VAD-FMK (Figure 5C, lane 5), we examined the PAK4 involvement in caspase activation and apoptosis. GFP-PAK4 or PAK4 siRNA were transiently transfected into HUVECs and after 48 h, cells were treated with 100 ng/ml TNF overnight. Flow cytometry analysis using FITC-conjugated active caspase-3 antibody revealed that GFP-PAK4 or PAK4 siRNA transfection had no effect on active caspase-3 in the absence of TNF (Figure 6A, upper panels). In the presence of TNF, about 40% of un-transfected cells stained as active caspase-3 positive; however, GFP-PAK4 overexpression reduced the percentage of active caspase-3 positive cells to about 15%, and PAK4 knockdown increased it to about 66%, (Figure 6A, lower panels). The TNF-induced apoptosis was significantly reduced by PAK4 overexpression, but increased by PAK4 suppression (Figure 6B). In addition, PAK4 suppression in TNF-treated cells increased the activation of caspase-8 analyzed by immunoblotting (Figure 6C).

Figure 6. PAK4 inhibits caspase activation and apoptosis in HUVECs. (A) Active caspase-3-FITC staining analyzed by FC in HUVECs transfected 48 h with GFP-PAK4 or PAK4 siRNA, and stimulated with TNF (100 ng/ml). (B) Quantitative evaluation of apoptotic cells analyzed by FC in panel A. The data represent the mean ± s.d. of three independent experiments. P-values were calculated using Student’s t-test. * P<0.05, ** P<0.01. (C) Western analysis of total caspase-8, active caspase-8, and control tubulin in TNF-stimulated HUVECs transfected with GFP-PAK4 or PAK4 siRNA.

Animals

Male C57BL/6 mice (weight about 25-29g, 10 weeks old) were kept under a conventional 12 hour light-dark cycle in a temperature-controlled room with free access to water and food. Mice were randomly assigned to 5 groups (10 mice per group) and treated with PBS or different doses of TNF (4 μg/kg, 20 μg/kg, 100 μg/kg or 500μg/kg) for 24 hours. All experimental protocols were approved by the Institutional Animal Care and Use Committee of the Shanghai Tongji University and performed in accordance with the National Institutes of Health Guidelines for the use of experimental animals.

Cell culture

The primary human umbilical vein endothelial cells (HUVECs) were purchased from ATCC (PCS-100-010) and were grown in vascular cell basal media supplemented with endothelial cell growth kit components.

Antibodies, proteins and drugs

Primary antibody against PAK4 was purchased from Proteintech. IκBα antibody was from Cell Signalling. Antibodies against tubulin, cdc42, β-actin, and IκBα, phalloidin-FITC, and horseradish peroxidase (HRP)-conjugated secondary antibody were purchased from Sigma. Alexa 488 and 568-conjugated secondary antibodies were from Molecular Probes. Caspase-8 and active caspase-8 antibodies were from Cell Signalling. Purified TNF protein and sTNFR1 (soluble part of TNFR1) were purchased from Immunotools (Friesoythe, Germany); thrombin was from Sigma. The PAK4 inhibitor PF-3758309 (Pfizer) was used at 10 μM.

Measurements of EMP

Untreated or TNF treated HUVECs (7x10⁵ cells in 12-well plates) were collected by centrifugation (5000g for 5 minutes). EMP was measured in supernatants by flow cytometry (Beckman Coulter, Fullerton, CA). 10 μl of centrifuged media or sera were labelled by annexin V-FITC and 1 μm latex beads (Beckman Coulter, flowcount) were used as gating parameters. Particles less than or equal to 1 μm were defined as EMPs. Number of EMP was calculated as described5.

HUVEC transfections

For transfection, HUVEC cells were seeded at 85%~90% confluency and allowed to adhere overnight. Small interfering RNA (siRNA), GFP-PAK4, or GFP-PAK4KM plasmid were transfected using Lipofectamine 2000 (Invitrogen) as described by the manufacturers. For studies of RNA interference, transfections were performed with PAK4 short interfering (si) RNA duplexes (two different siRNAs mixed, final concentration 50 nM, Life Technologies) or a nonsilencing siRNAs as negative controls (Life Technologies). At 48 hours after siRNA transfection, cells were harvested to determine the level of PAK4 proteins by immunoblotting, or were stimulated with 100 ng/mL TNF.

Cloning strategy

Human PAK4 DNA fragment was amplified from cDNA bank by PCR and cloned into pGEMTeasy (Promega) vector. GFP-PAK4 was cloned into the EcoRI/SalI site of EGFP-C2 vector. The GFP fusion protein was generated to carry GFP at its N-terminus. GFP-PAK4KM was cloned into the EcoRI/SalI site of EGFP-C2 vector. This PAK4 kinase mutant contains a lysine to methionine mutation at amino acid 350. The mutation was generated using site-directed mutagenesis kit (Stratagene).

Immuno-fluorescence microscopy

HUVECs were cultured to subconfluence on 8 chamber glass slides (BD Falcon, Bedford, MA). After treatment, they were fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) and permeabilized with 0.1% Triton X-100. Actin cytoskeleton was stained with Phalloidin-FITC. Cells were incubated with 1:400 diluted polyclonal anti-PAK4 antibody, followed by a sequential incubation with a 1:1000 diluted Alexa 488 or 568-conjugated secondary antibody. Nuclei were counter stained with DAPI (Sigma-Aldrich). The cells were imaged using a ZEISS 780 inverted confocal laser scanning microscope (ZEISS, Cambridge, UK) and analyzed by using Zen software.

Western blotting

Cells were washed with cold PBS and lysed with 1 ml RIPA buffer (NEB, Hertfordshire, UK) containing protease inhibitor cocktail (Roche) on ice. Homogenized samples were microcentrifuged at full speed for 15 min at 4°C. The supernatants were subjected to SDS-PAGE. Separated proteins were transferred onto nitrocellulose membrane and probed with primary antibodies and HRP-conjugated secondary antibodies.

Quantitative real time PCR

Total RNA was extracted from cultured HUVECs using Trizol solution (Invitrogen) and reverse transcribed into cDNA with a Reverse Transcriptase kit (Invitrogen) according to manufacturer’s instructions. The expression of PAK4 were detected using a TagMan mRNA assay kit (Biosystems). The relative expression was determined by the 2^-ΔΔCt method with GAPDH as the internal reference. Samples were analysed in triplicate. The primer sequences used were the following. PAK4-F: ATGTGGTGGAGATGTACAACA, PAK4-R: GTTCATCCTGGTGTGGGTGAC, GAPDH-F: GGAGCGAGATCCCTCCAAAAT, GAPDH-R: GGCTGTTGTCATACTTCTCATGG

Apoptotic assay

Cells were fixed with the Cytofix/Cytoperm fixation and permeabilization solution kit (BD Pharmingen), stained with FITC-conjugated anti-caspase-3 antibody (BD Pharmingen), and analyzed using CyAn flow cytometer (Beckman Coulter, Fullerton, CA). Data were evaluated using Flowjo (Tree Star Inc., Ashland, Oregon).

Statistics

Results are expressed as mean ± SD. The statistical significance was assessed using analysis of variance (ANOVA), followed by Student’s t-test. A P value of <0.05 was considered significant.