Interferon-Gamma Increases Endothelial Permeability by Causing ...

5 downloads 0 Views 1MB Size Report
Interferon-gamma (IFN-γ) is known to potentiate the progression of ... Consistent with our finding from the permeability assay, IFN-γ induced the formation of.
RESEARCH REPORTS

JOURNAL OF INTERFERON & CYTOKINE RESEARCH Volume 35, Number 7, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/jir.2014.0188

Interferon-Gamma Increases Endothelial Permeability by Causing Activation of p38 MAP Kinase and Actin Cytoskeleton Alteration Chin Theng Ng,1 Lai Yen Fong,1 Mohd Roslan Sulaiman,1 Mohamad Aris Mohd Moklas,2 Yoke Keong Yong,2 Muhammad Nazrul Hakim,1 and Zuraini Ahmad1

Interferon-gamma (IFN-g) is known to potentiate the progression of inflammatory diseases, such as inflammatory bowel disease and atherosclerosis. IFN-g has been found to disrupt the barrier integrity of epithelial and endothelial cell both in vivo and in vitro. However, the mechanisms of IFN-g underlying increased endothelial cell permeability have not been extensively elucidated. We reported that IFN-g exhibits a biphasic nature in increasing endothelial permeability. The changes observed in the first phase (4–8 h) involve cell retraction and rounding in addition to condensed peripheral F-actin without a significant change in the F-/G-actin ratio. However, cell elongation, stress fiber formation, and an increased F-/G-actin ratio were noticed in the second phase (16–24 h). Consistent with our finding from the permeability assay, IFN-g induced the formation of intercellular gaps in both phases. A delayed phase of increased permeability was observed at 12 h, which paralleled the onset of cell elongation, stress fiber formation, and increased F-/G-actin ratio. In addition, IFN-g stimulated p38 mitogen-activated protein (MAP) kinase phosphorylation over a 24 h period. Inhibition of p38 MAP kinase by SB203580 prevented increases in paracellular permeability, actin rearrangement, and increases in the F-/G-actin ratio caused by IFN-g. Our results suggest that p38 MAP kinase is activated in response to IFN-g and causes actin rearrangement and altered cell morphology, which in turn mediates endothelial cell hyperpermeability. The F-/G-actin ratio might be involved in the regulation of actin distribution and cell morphology rather than the increased permeability induced by IFN-g.

Introduction

T

he vascular endothelium is composed of a monolayer of endothelial cells, which tightly control the passage of molecules from the blood to the underlying tissues. There are several key players involved in the maintenance of cell barrier integrity; one of these is the actin cytoskeleton (Yuan and Rigor 2010). The actin cytoskeleton exists in 2 forms, the monomer (G-actin) and the polymer (F-actin). They are dynamic structures that can undergo processes of polymerization and depolymerization in response to challenges, such as growth factors (Gong and others 2004) and cytokines (Frausto-Del-Rio and others 2012). These changes enable the cells to carry out biological processes such as maintenance of cell shape (Lazaro-Dieguez and Egea 2007) and cell migration (Koestler and others 2009). In addition, it was reported that the balance between G- and F-actin also contributes to cell barrier tightness (Waschke and others 2005).

Cytokines are multifunctional proteins that are involved in all biological processes. Clinical studies revealed that the dysregulation of cytokine production significantly contributes to the progression of inflammatory diseases (Barnes 2008; Ait-Oufella and others 2011; Muzes and others 2012). Therefore, cytokine-based therapy has garnered much attention from researchers, not only in basic research fields but also in clinical settings and pharmaceutical industries, since the last decade. Etanercept, a tumor necrosis factor-a (TNF-a) blocker, was approved by the Food and Drug Administration in 1998 for the treatment of rheumatoid arthritis (Kalliolias and Ivashkiv 2009; Schett and others 2013). However, ongoing research is still needed to understand how cytokines act under pathological conditions so that they can be used as a therapeutic target in clinical practice. Interferon-gamma (IFN-g), a pro-inflammatory cytokine, is primarily secreted by natural killer cells, T1 helper cells, and cytotoxic T cells. IFN-g has been comprehensively studied for its impact on barrier function both in vivo and

Departments of 1Biomedical Sciences and 2Human Anatomy, Faculty of Medicine and Health Sciences, Universiti Putra Malaysia, UPM Serdang, Selangor, Malaysia.

513

514

in vitro (Wong and others 1999). The effects of IFN-g on inflammatory diseases such as intestinal diseases (Nakamura and others 1992; Niessner and Volk 1995; Capaldo and Nusrat 2009), asthma (Sedgwick and others 2002), and atherosclerosis (McLaren and Ramji 2009) have also been reported. Previous studies reported that IFN-g acts on endothelial cells by altering cell morphology and actin filaments (Stolpen and others 1986); however, whether these changes lead to endothelial hyperpermeability remain unknown. On the other hand, IFN-g disrupts epithelial barrier function and was found to be correlated with the reduced expression of tight junction proteins (Youakim and Ahdieh 1999) or the internalization of junction proteins dependent on Rho-associated kinase (ROCK) (Bruewer and others 2005; Utech and others 2005). Furthermore, IFN-g not only increases paracellular permeability, but also enhances transcellular permeability of epithelial cells via phosphatidylinositol 3¢-kinase and protein kinase C-dependent pathways (McKay and others 2007). Moreover, IFN-g causes a rapid activation of phosphatidylinositol 3¢-kinase and a subsequent prolonged activation of nuclear factor-kB, which leads to disturbances in barrier integrity (Boivin and others 2009). To date, the role of IFN-g in regulating the barrier function of epithelial cells is well studied but its effects on the disruption of endothelial barrier structure and integrity remain unclear. In this study, we demonstrated that IFN-g disrupted human umbilical vein endothelial cell (HUVEC) barrier integrity by causing alterations in cell morphology, actin reorganization, and increased permeability in a biphasic manner. Western blotting revealed that IFN-g did not affect the F-/G-actin ratio in the first phase (4–8 h) while the second phase (16–24 h) involved F-actin polymerization. In addition, IFN-g activated p38 mitogen-activated protein (MAP) kinase and subsequently led to reassembly of actin, morphological changes, and increased permeability. p38 MAP kinase also regulated the F-/G-actin ratio changes induced by IFN-g in the second phase. However, changes in the F/G-actin ratio are more likely to modulate IFN-ginduced actin rearrangement and cell shape alteration instead of increased permeability.

NG ET AL.

Millipore. Adenosine triphosphate (ATP) assay kit was purchased from BioVision, Inc.

Cell culture HUVECs were grown in 25 cm2 T-flasks and the media was changed every 2 days. The cells were nourished with ECM supplemented with 5% fetal bovine serum, 1% endothelial cell growth supplement, and 1% penicillin/streptomycin and were maintained at 37C, 5% CO2. Cells from passage 3 to 4 were used for all the experiments.

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay HUVECs were plated onto 96-well plates at a concentration of 5 · 103 cells per well ( Jiang and others 2012). After 48 h, the cells were treated with different concentrations of IFN-g for 24 h. We added 10 mL of MTT solution (5 mg/mL) and incubated for 4 h. Later, the solution in each well was removed and 100 mL of dimethyl sulfoxide (DMSO) was added to dissolve the purple salt formazan. The optical densities were measured at 570 nm with a reference wavelength of 650 nm using a microplate reader (Infinite M200; TECAN). The results were expressed as a percentage compared to normal.

Adenosine triphosphate fluorometric assay The protocol was carried out according to the manufacturer’s instructions as previously described (Ang and others 2011). Briefly, the cells at a density of 2 · 105 cells per well were plated in 24-well plates. After 4 days, the cells were treated with different concentrations of IFN-g and incubated for 24 h. Then, the cells were lysed and the supernatant was collected. We incubated 50 mL of the supernatant with 50 mL of reaction mix and incubated the mixture at room temperature for 30 min. The fluorescence intensities of the mixture were measured at an excitation/emission wavelength of 535/587 nm. The ATP concentration for each group was calculated based on the formula provided in the manual. The results were expressed as a percentage compared to control.

Materials and Methods Human recombinant IFN-g was purchased from eBioscience. Human recombinant TNF-a was purchased from PeproTech. HUVECs were purchased from Life Technologies and maintained in phenol red free-endothelial cell medium (ECM) from ScienCell. SB203580 was purchased from Calbiochem. Phalloidin and cytochalasin D were purchased from Sigma. Collagen-coated coverslips (22 mm) were purchased from BD Transduction Laboratory. Phosphatase inhibitor cocktail was purchased from NacalaiTesque. Protease inhibitor cocktail was purchased from Amresco. Rhodamine-phalloidin was purchased from Life Technologies. Prolong Gold antifade reagent was purchased from Molecular Probes. Antirabbit IgG, HRP-linked Ab, phospho-specific p38 MAP kinase (Thr 180/Tyr 182), rabbit polyclonal IgG, p38 MAP kinase Ab, rabbit polyclonal IgG, and 4¢,6-diamidino-2-phenylindole (DAPI) were purchased from Cell Signaling Technology. G/F actin in vivo assay kit was purchased from Cytoskeleton, Inc. In vitro permeability assay kit and Immobilon-P 0.45 mm polyvinylidenedifluoride (PVDF) membrane were purchased from

Permeability assay using fluorescein isothiocyanate-dextran as tracer The assay was carried out according to the in vitro permeability assay kit. Briefly, the cell concentration was adjusted to 1 · 106 cells per mL. Then, 200 mL of cell suspension was seeded on collagen-coated inserts and the bottom wells were filled with 500 mL of ECM. The media was changed every 2 days until cell monolayers were formed. Then, the cells were incubated with various treatments depending on the experimental design. After the indicated treatment time, the inserts were transferred to a new well plate where the bottom wells were filled with new cell basal media. The solutions in each insert were then replaced with 150 mL of fluorescein isothiocyanate (FITC)-dextran (1:50 dilution) (Ang and others 2011). The flux of FITCdextran across the cell monolayers were measured as fluorescent intensities at an excitation/emission wavelength of 485/530 nm. The results were expressed as a percentage compared to control.

INTERFERON-c AND HUVEC HYPERPERMEABILITY

Immunocytochemistry The experiment was carried out according to the actin staining protocol provided by Life Technologies except for the cell seeding density and cell growth duration. The cells were grown as a monolayer on collagen-coated coverslips with a cell seeding density of 1.5 · 105 cells. After 5 days, the cells were challenged with IFN-g or SB203580 or SB203580 followed by IFN-g. The cells were fixed with 3.7% paraformaldehyde/ phosphate-buffered saline (PBS) for 10 min. The cells were then washed in PBS twice and permeabilized with 0.1% Triton X-100/PBS for 4 min. Cells were washed in PBS twice again before being stained with rhodamine-phalloidin/PBS for 20 min at room temperature. The nuclei were stained with DAPI/PBS (0.5 mg/mL) for 2 min. All the coverslips were mounted with Prolong Gold anti-fade reagent and stored at 4C in the dark.

Quantification of F-/G-actin According to the G-/F-actin assay kit protocol, cells have to be lysed with LAS2 buffer and incubated for 10 min at 37C. This was followed by a short centrifugation of lysates to remove cell debris. The samples were spun at 100,000 g, 37C for 1 h and the supernatant was labeled as G-actin. The pellet were resuspended with F-actin depolymerization buffer (equal volume as the supernatant) and incubated on ice for 1 h and then resuspended every 15 min to dissociate the F-actin. Equal volumes of G- and F-actin samples were dissolved in 5 · sodium dodecyl sulfate (SDS) sample buffer and subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), transferred to PVDF membrane (100 V, 90 min), blocked in 5% bovine serum albumin (BSA) for 30 min, and incubated with anti-actin antibody (1:3,000 dilution) overnight. Protein bands were visualized and captured with the Chemi-Smart 3000 system (VilberLourmat) after incubation with chemiluminescent substrate, Luminata Forte Western HRP Substrate (Millipore). We used 10–200 mg of actin to construct a 4 parametric standard curve. The actin level for each group was obtained from the standard curve and the F/G-actin ratio was calculated.

Western blot of p38 MAP kinase phosphorylation Cells were washed with ice-cold PBS before they were scraped with RIPA lysis buffer (containing phosphatase and protease inhibitors) at 4C. Cell lysates were spun at 10,000 rpm for 10 min at 4C. The supernatant was collected and protein concentrations were quantified using the bicinchoninic acid (BCA) protein assay reagent kit (Pierce). We mixed 20 mg of samples with 5 · SDS sample buffer and ran samples on SDS-PAGE, transferred to PVDF membrane (100 V, 90 min), blocked in 5% BSA for 2 h, and incubated with antiphospho-p38 MAP kinase antibody (1:3,000 dilution) overnight. The membranes were washed and incubated with anti-rabbit IgG, HRP-linked antibody (1:10,000 dilution) for 1.5 h. Protein bands were visualized by using the chemiluminescent substrate, Luminata Forte Western HRP Substrate (Millipore). The membranes were stripped and incubated with anti-p38 MAP kinase antibody (1:3,000 dilution). The graph presents fold change compared to control.

515

controls were determined using 1-way analysis of variance followed by the Duncan post hoc test. P < 0.05 was considered as significant. All graphs were constructed using the software GraphPad Prism version 5.0 (GraphPad Software, Inc.).

Results IFN-c does not reduce HUVEC viability We first examined the effect of IFN-g on HUVEC viability. MTT and ATP fluorometric assays showed that IFN-g at doses of 200–3.125 ng/mL did not cause significant cell death after 24 h of incubation (Fig. 1A, B). Previous studies demonstrated that IFN-g exerts an apoptotic effect on immune cells (Cho and Pyo 2010) and cancer cell lines (Detjen and others 2001). In this study, we proved that the dosage of IFN-g and incubation time used in our experiment setting did not affect endothelial cell viability.

IFN-g-induced biphasic increases in HUVEC permeability We next investigated whether IFN-g will cause any significant increase in endothelial permeability. Confluent endothelial cells were exposed to increasing doses of IFN-g and incubated for various times. The first phase of increased permeability induced by 10 ng/mL IFN-g started at 4 h with the peak observed at 8 h (Fig. 1C). Unexpectedly, the increase in permeability returned to a basal level by 12 h. This was then followed by a second phase where the permeability increased again at 16 h and persisted up to 24 h. Thus, we report that IFN-g-induced endothelial hyperpermeability occurred in a biphasic manner with a delayed phase observed at 12 h. A dose– response study (Fig. 1D) showed that 0.1–100 ng/mL of IFN-g caused increases in endothelial permeability with the maximum increase in permeability induced by 10 ng/mL. TNF-a, a well-studied inducer, was used as a reference in our experiment. IFN-g at doses of 10 and 100 ng/mL stimulated increases in permeability that were comparable to 10 ng/mL TNF-a. Interestingly, IFN-g combined with TNF-a has been reported to possess a synergistic effect compared to either single cytokine alone (Fish and others 1999; Bruewer and others 2003).

IFN-g-induced HUVEC hyperpermeability correlates with actin cytoskeleton Increased endothelial permeability is a characteristic of cell barrier impairment and actin has been reported to be involved in the regulation of cell barrier integrity. To correlate the increases in permeability induced by IFN-g with actin, an actin stabilizing agent, phalloidin, was used. Figure 2 showed that endothelial cells challenged with IFN-g for 8 h significantly increased endothelial permeability compared with controls. However, stabilization of actin with phalloidin attenuated IFN-g-induced increases in basal permeability. Cells treated with phalloidin alone did not have altered basal endothelial permeability. These results suggest that IFN-g caused destabilization of actin, which subsequently stimulated endothelial hyperpermeability.

Statistical analysis

IFN-g-induced HUVEC hyperpermeability was due to actin rearrangement and morphological alterations

All the data were reported as the mean – standard error of mean. Significant differences between treatment groups and

To study whether IFN-g-induced hyperpermeability is associated with actin rearrangement and morphological changes,

516

FIG. 1. The effect of IFN-g on HUVECs viability and permeability. (A) The dose and time response of IFN-g on HUVECs viability was assessed by MTT assay and (B), ATP fluorometric assay. The cells were challenged with 3.125–200 ng/ mL of IFN-g and incubated for 24 h. (C) The time–response curve of 0.1–100 ng/mL IFN-g after incubation for 4–24 h. Controls were the cells incubated with IFN-g at time 0. (D) The dose–response curve of IFN-g (0.1–100 ng/mL) at 8 h. TNFa (10 ng/mL, 8 h) was used as control. Three independent experiments were performed (n = 3). *P < 0.05 as compared to control. ATP, adenosine triphosphate; C, control; HUVECs, human umbilical vein endothelial cells; IFN-g, interferon-gamma; RFU, relative fluorescence unit; TNF-a, tumor necrosis factor-a.

NG ET AL.

INTERFERON-c AND HUVEC HYPERPERMEABILITY

FIG. 2. The effect of an actin-stabilizing agent in IFN-ginduced endothelial hyperpermeability. HUVECs were exposed to vehicle (control), IFN-g (10 ng/mL, 8 h), phalloidin (10 - 8 M, 30 min), or pretreated with phalloidin followed by IFN-g. Three independent experiments were performed (n = 3). *P < 0.05 as compared to control. #P < 0.05 as compared to IFN-g only. Pha, phalloidin. F-actin was stained with rhodamine-phalloidin. Under normal conditions, all cells were closely attached with adjacent cells and no prominent gap formation was observed. Actin filaments spanning across the cytoplasm were frequently seen (Fig. 3A). Upon stimulation with IFN-g, actin filaments underwent a dramatic rearrangement. By 4 and 8 h (Fig. 3B, C) of treatment time, all cells displayed a rounded morphology and lost their contacts with adjacent cells. However, they were still connected with each other through the formation of fine retraction fibers. Large gaps were formed between the cells and the actin filaments became condensed in the peripheral region. Interestingly, large gaps were not observed in between the cells at 12 h, instead, there was an increase in stress fiber formation and cell elongation (Fig. 3D). After 16 h of incubation time, the

517

cells remained elongated and stress fiber formation and pronounced intercellular gaps were detected. These changes were sustained for up to 24 h (Fig. 3E–G). The time response of intercellular gap formation observed in actin staining was consistent with the IFN-g-induced increased permeability shown in Fig. 1C. Cytochalasin D was used as a control to display the disruption of F-actin in HUVEC (Fig. 3H). In summary, IFN-g induced cell rounding, cell retraction, and condensed cortical F-actin in the first phase, while cell elongation and stress fiber formation dominated the second phase. These results suggest that IFN-g increased endothelial permeability by causing actin filament disorganization and morphological alterations.

The F-/G-actin ratio affects IFN-c-induced endothelial cell morphological changes and actin rearrangement but not endothelial hyperpermeability To study whether the redistribution of actin filaments induced by IFN-g was associated with the shift in cellular actin pools, the G- and F-actin contents were quantified by immunoblotting using an anti-actin antibody (Fig. 4A). Before addition of IFN-g, cells contained a large proportion of G-actin compared to F-actin. Cells treated with IFN-g for 1, 4, and 8 h did not change in basal F-/G-actin ratio (Fig. 4A). At 12 h, the actin pool of the cells shifted from G-actin to F-actin. F-actin became the dominant actin for the following time points; 16, 20, and 24 h. It is worth mentioning that the increased F-actin elicited by IFN-g began at 12 h, which coincided with cell elongation and stress fiber formation that were first noticed at 12 h and continued until 24 h (Fig. 3). Collectively, IFN-g does not affect the F-/G-actin ratio in the first phase, but rather, it increased the ratio in the

FIG. 3. The effect of IFN-g in endothelial cell shape and actin organization. HUVECs were exposed to either vehicle (A), or 10 ng/mL IFN-g for 4 h (B), 8 h (C), 12 h (D), 16 h (E), 20 h (F), 24 h (G), or cytochalasin D (5 mM, 30 min) (H). Arrowheads indicate the gap formation between the cells; arrows indicate the retraction fiber whereas asterisks indicate the stress fiber formation. Ten confocal Z-stack images (0.4 mm for each optical sections) were projected into 1 image (magnification, 600 · ). Three independent experiments were performed (n = 3). Red and blue colors represent F-actin and nuclei, respectively.

518

NG ET AL.

FIG. 4. The effect of IFN-g on the F-/G-actin ratio and p38 MAP kinase phosphorylation. (A) The time-course study of IFN-g-induced changes in actin pools. HUVECs were treated with either vehicle (control, C) or 10 ng/mL IFN-g for 1–24 h. Phalloidin (pha), an F-actin stabilizing agent, was used to validate the complete separation of F-/G-actin in the cells. Phalloidin-treated group should yield *80% of F-actin. TNF-a (10 ng/mL, 8 h) was used as reference. (B) The time response of IFN-g on p38 MAP kinase phosphorylation. HUVECs were treated with 10 ng/mL IFN-g for 5 min-24 h. TNF-a (10 ng/mL, 15 min) was used as reference. Results represent 3 independent experiments (n = 3). *P < 0.05 as compared to control. MAP, mitogen-activated protein; p-p38, phosphorylated-p38. second phase. These data imply that the F-/G-actin ratio does not appear to directly affect the IFN-g induced increases in permeability occurring as early as 4 h. However, the shifts in G- to F-actin induced by IFN-g might play a part in the regulation of stress fiber formation and altered cell morphology.

IFN-c-induced p38 MAP kinase phosphorylation To explore the molecular mechanisms triggered by IFN-g, the phosphorylation level of p38 MAP kinase was detected. As shown in Fig. 4B, cells exposed to IFN-g for 5–30 min did not demonstrate altered p38 MAP kinase phosphorylation. However, increased p38 MAP kinase phosphorylation was noticed at 4 h and remained phosphorylated until 24 h. TNF-a was used as a reference in our experiment as it has been reported to stimulate HUVEC activation by causing p38 MAP kinase phosphorylation (Petrache and others 2003; Shivanna and others 2010). Their observation is consistent with our result that showed that cells incubated with 10 ng/ mL TNF-a exhibited an increase in p38 MAP kinase phosphorylation. Taken together, these results suggest that the rapid treatment of cells with IFN-g (minutes) does not acti-

vate p38 MAP kinase; however, prolonged exposure (hours) of cells to IFN-g causes p38 MAP kinase phosphorylation.

IFN-c-induced actin rearrangement and endothelial morphological changes in both phases is associated with p38 MAP kinase activation Next, we examined whether IFN-g-induced p38 MAP kinase activation regulates downstream events such as actin rearrangement and cell morphological changes. Figure 5 showed that normal cells displayed a cobblestone morphology where they were closely attached to each other (Fig. 5A). Cells treated with SB203580 alone displayed thin peripheral actin bands, which were more prominent compared to normal cells. In addition, the cytoplasmic F-actin staining intensity was reduced compared with normal cells (Fig. 5B). Upon challenge with 10 ng/mL IFN-g for 8 h (first phase), the cells became retracted and rounded, which was accompanied by large intercellular gap formations (Fig. 5C). Pretreatment of the cells with SB203580 followed by IFN-g for 8 h resulted in a marked reduction in gap formation, cell rounding, peripheral actin bands, and retraction fibers (Fig.

INTERFERON-c AND HUVEC HYPERPERMEABILITY

519

FIG. 5. The effect of p38 MAP kinase inhibition by SB203580 on IFN-g-induced actin rearrangement and morphological changes. HUVECs were grown to confluence and were treated with media only (control) (A), SB203580 (10 mM, 30 min) (B), IFN-g (10 ng/mL, 8 h) (C), SB203580 pretreatment followed by IFN-g for 8 h (D), IFN-g (10 ng/mL, 24 h) (E), or SB203580 pretreatment followed by IFN-g for 24 h (F). Arrowheads indicate the gap formation between the cells; arrows indicate the retraction fibers whereas asterisks indicate the stress fiber formation. Ten confocal Z-stack images (0.4 mm for each optical sections) were projected into 1 image (magnification, 600 · ). Results represent 3 independent experiments (n = 3). Red and blue colors indicate F-actin and nuclei, respectively. 5D). When the cells were exposed to IFN-g for 24 h (second phase), elongation of cells, stress fiber formation, and intercellular gap formation were observed (Fig. 5E), whereas inhibition of p38 MAP kinase attenuated these changes caused by IFN-g at 24 h (Fig. 5F). Collectively, IFN-g-induced actin rearrangement and cell morphological changes are associated with p38 MAP kinase activation. These results are consistent with p38 MAP kinase activation stimulated by IFN-g for 24 h (Fig. 4B).

p38 MAP kinase regulates IFN-c-induced HUVEC hyperpermeability To investigate whether p38 MAP kinase affects the increase in endothelial permeability caused by IFN-g, a p38 MAP kinase inhibitor, SB203580, was used. As shown in Fig. 6A, pretreatment of cells with SB203580 significantly suppressed IFN-g-induced endothelial hyperpermeability. Cells treated with SB203580 alone did not have altered

FIG. 6. The effect of SB203580 on IFN-g-induced HUVECs hyperpermeability and F-/G-actin ratio. (A) The effect of p38 MAP kinase inhibition by SB203580 on IFN-g-induced increased in HUVECs permeability. The cells were pretreated with either vehicle (control), IFN-g alone (10 ng/mL, 8 h), SB203580 (SB) alone (10 mM, 30 min), or SB203580 followed by IFN-g. (B) The effect of p38 MAP kinase inhibition by SB203580 on IFN-g-induced F/G-actin ratio alteration at 24 h. The cells were pretreated with either vehicle (control), IFN-g alone (10 ng/mL, 24 h), SB203580 alone (10 mM, 30 min), or SB203580 followed by IFN-g. Results represent 3 independent experiments (n = 3). *P < 0.05 compared with control. # P < 0.05 as compared to IFN-g. C, control.

520

basal endothelial permeability. These results indicate that p38 MAP kinase is required for IFN-g-induced endothelial hyperpermeability. We suggest that p38 MAP kinase is activated upon IFN-g stimulation and precedes IFN-ginduced endothelial hyperpermeability.

p38 MAP kinase regulates the IFN-c-induced increase in F-/G-actin ratio As shown in Fig. 4A, the increase in F-actin content triggered by IFN-g began at 12 h and was sustained for 24 h. We then extended our study to evaluate the role of SB203580 in the shifts of actin pools elicited by IFN-g (Fig. 6B). Pretreatment of the cells with SB203580 significantly reduced the IFN-g-induced increase in the F-/G-actin ratio whereas SB203580 alone did not alter the basal F-/G-actin ratio. This result indicates that p38 MAP kinase is required for the IFNg-induced increase in the F-/G-actin ratio at 24 h.

Discussion IFN-g is a pleiotropic cytokine that regulates many cellular activities in biological systems. Therefore, it is known that dysregulation of IFN-g may lead to numerous pathologies. Although IFN-g has been shown to cause impaired barrier function and participate in the progression of inflammatory diseases, the underlying molecular mechanisms are still poorly understood. In this study, we were the first to demonstrate that IFN-g elicited increased HUVEC permeability in a biphasic manner. We show that IFN-g induced actin destabilization, which in turn led to hyperpermeability. We further provide evidence showing that actin redistribution and cell morphological changes in response to IFN-g promote increased permeability. In addition, our data also showed that IFN-g increased p38 MAP kinase phosphorylation. By using a p38 MAP kinase inhibitor we report that IFN-g induced p38 MAP kinase activation, which in turn caused actin rearrangement, endothelial morphological changes, and eventually led to endothelial hyperpermeability. p38 MAP kinase is also required for the F-/G-actin ratio changes caused by IFN-g. IFN-g is well recognized for its antiviral (Chesler and Reiss 2002), antimicrobial (Shtrichman and Samuel 2001), and antitumor (Ikeda and others 2002) activities. However, accumulating evidence showing that cytokines disrupted cell barrier function via apoptosis-independent pathways implies the involvement of other mechanisms (Petrache and others 2001; Bruewer and others 2003). Our study demonstrated that IFN-g-induced endothelial hyperpermeability is independent of cell death and was involved in actin cytoskeletal changes and p38 MAP kinase activation. Acute inflammatory mediators such as thrombin and histamine induce a rapid and reversible effect on endothelial hyperpermeability (Bogatcheva and others 2002; Srinivas and others 2006). On the contrary, cytokines such as TNF-a produce a slow onset but prolonged effect in increased cell permeability (McKenzie and Ridley 2007). Our findings showed that IFN-g induced a biphasic increase in HUVEC permeability. The increase in permeability induced by IFNg can be divided into 3 phases: an early phase (4–8 h), a delayed phase (12 h), and a sustained phase (16–24 h). The first phase of increased permeability caused by IFN-g is shorter but more robust than the second phase. In the past,

NG ET AL.

cytokines such as vascular endothelial growth factor (Behzadian and others 2003) and TNF-a (Shivanna and others 2010) have also been reported to disrupt cell barrier integrity in a biphasic manner. However, our results from the permeability assay opposed the observations by others who found IFN-g-induced timeand dose-dependent impairment in cell barrier function (Youakim and Ahdieh 1999; Boivin and others 2009). The difference in the results may be due to different experimental methods and cell types; the other studies assessed monolayer integrity using transendothelial electrical resistance and T84 intestinal epithelial cells. Another possibility is time; previous studies examined the time-response of IFN-g in 24 h intervals (24, 48, 72, and 96 h) (Boivin and others 2009) whereas we assessed the effect of IFN-g on HUVEC permeability by using 4 h intervals for 24 h. Nevertheless, IFN-g caused an increase in endothelial permeability at 24 h in our study, which was consistent with previous reports. Our data clearly showed that 12 h is a transition phase where cell shape changed from round and retracted into elongated. A condensed peripheral actin ring was redistributed across the cell in the direction of cell elongation. At this time point, the sudden increase in the F-/G-actin ratio that was accompanied by cell shape changes and actin rearrangement strongly suggests that the F-/G-actin ratio may be involved in the regulation of cell shape and actin distribution. Indeed, G- and F-actin equilibrium plays a pivotal role in modulating cell shape and motility (Koestler and others 2009). Previous studies also demonstrated that hyperpolymerization or depolymerization of actin increased in vivo microvessel permeability (Waschke and others 2005). However, we showed that the F/G-actin ratio did not directly affect IFN-g-induced increased permeability. It has been reported that stress fibers are a component of the contractile apparatus in endothelial cells (Tojkander and others 2012). In our study, cells exposed to IFN-g displayed a contraction response followed by stress fiber formation suggesting the involvement of the contractile mechanism. Early studies reported that myosin light chain (MLC) phosphorylation by MLC kinase is responsible for thrombininduced actomyosin-based contraction (Goeckeler and Wysolmerski 1995). The latter demonstrated that thrombin induces actin rearrangement and cell contraction via activation of other kinases such as p38 MAP kinase (Borbiev and others 2004), calmodulin kinase II, and extracellular signal-regulated kinase (Borbiev and others 2003). In this study, prolonged exposure (hours) of endothelial cells to IFN-g caused activation of p38 MAP kinase. Interestingly, other inducers such as TNF-a and thrombin induced a rapid phosphorylation of p38 MAP kinase with peaks in minutes (Petrache and others 2003; Borbiev and others 2004; Shivanna and others 2010). Because the short exposure (minutes) of cells to IFN-g did not activate p38 MAP kinase, we suggest that signaling molecules other than p38 MAP kinase control the early event (minutes) induced by IFN-g. A recent study demonstrated that actin remodeling caused p38 MAP kinase activation particularly in hepatic stellate cells (Cui and others 2014). However, in endothelial cells, p38 MAP kinase is often an upstream and key regulatory event for actin reorganization triggered by inducers (Cai and others 2008). We also showed that IFN-g triggers p38 MAP kinase activation and that the p38 MAP kinase mediates

INTERFERON-c AND HUVEC HYPERPERMEABILITY

actin reorganization. The precise mechanism by which p38 MAP kinase leads to this event is not well understood. A recent study reported that IFN-g (100 U/mL, 24 h) induced disruption of the actin cytoskeleton into small, granular structures on HUVECs and that guanylate binding protein 1, an actin remodeling factor, plays a critical role in regulating this event (Ostler and others 2014). How p38 MAP kinase affects actin reorganization is not well understood but it may involve actin-binding proteins such as guanylate binding protein 1. Future work will help elucidate the precise mechanisms by which p38 MAP kinase activation modulates actin cytoskeletal function. Other than p38 MAP kinase, multiple signaling pathways related to permeability coordinate cell morphology and actin cytoskeletal distribution. Studies showed that Rho regulates stress fiber formation and correlates with permeability (Vouret-Craviari and others 1998; Wojciak-Stothard and others 2001). Its target, ROCK, regulates stress fiber formation and the cell elongation induced by TNF-a (McKenzie and Ridley 2007). Rac has been reported to cause accumulation of cortical actin and cell rounding and the appearance of cortical actin is required to cause cell retraction caused by thrombin (Vouret-Craviari and others 1998). In our study, IFN-g induces a condensed peripheral actin band, which appeared concurrently with cell rounding in the first phase followed by cell elongation and stress fiber formation in the second phase. Therefore, it is possible that IFN-g also activates Rac and Rho/ROCK pathways. However, more detailed work should be conducted to fully elucidate the involvement of these pathways in IFN-g-induced endothelial hyperpermeability. In conclusion, we propose that IFN-g increases HUVEC permeability involves actin rearrangement and changes in cell morphology. Changes in the F/G-actin ratio correlate with stress fiber formation and changes in cell morphology induced by IFN-g, but not endothelial hyperpermeability. We provide new evidence showing that p38 MAP kinase phosphorylation is a key regulator that governs IFN-g-induced endothelial hyperpermeability, actin redistribution, cell shape changes, and G-/F-actin changes. All these findings enrich our current understanding of the roles of IFN-g on barrier function and provide a new therapeutic strategy to treat IFN-g-associated vascular diseases. Future studies should focus on the role of actin-binding proteins as these molecules are involved in the modulation of actin dynamics, which subsequently will affect endothelial barrier integrity.

Acknowledgments This research project was supported by Research University Grant Scheme (RUGS), Universiti Putra Malaysia (Project No: 04-02-12-1792RU).

Author Disclosure Statement No competing financial interests exist.

References Ait-Oufella H, Taleb S, Mallat Z, Tedgui A. 2011. Recent advances on the role of cytokines in atherosclerosis. Arterioscler Thromb Vasc Biol 31(5):969–979.

521

Ang KP, Tan HK, Selvaraja M, Kadir AA, Somchit MN, Akim AM, Zakaria ZA, Ahmad Z. 2011. Cryptotanshinone attenuates in vitro oxLDL-induced pre-lesional atherosclerotic events. Planta Med 77(16):1782–1787. Barnes PJ. 2008. The cytokine network in asthma and chronic obstructive pulmonary disease. J Clin Invest 118(11):3546– 3556. Behzadian MA, Windsor LJ, Ghaly N, Liou G, Tsai NT, Caldwell RB. 2003. VEGF-induced paracellular permeability in cultured endothelial cells involves urokinase and its receptor. FASEB J 17(6):752–754. Bogatcheva NV, Garcia JG, Verin AD. 2002. Molecular mechanisms of thrombin-induced endothelial cell permeability. Biochemistry (Mosc) 67(1):75–84. Boivin MA, Roy PK, Bradley A, Kennedy JC, Rihani T, Ma TY. 2009. Mechanism of interferon-gamma-induced increase in T84 intestinal epithelial tight junction. J Interferon Cytokine Res 29(1):45–54. Borbiev T, Birukova A, Liu F, Nurmukhambetova S, Gerthoffer WT, Garcia JG, Verin AD. 2004. p38 MAP kinase-dependent regulation of endothelial cell permeability. Am J Physiol Lung Cell Mol Physiol 287(5):L911–L918. Borbiev T, Verin AD, Birukova A, Liu F, Crow MT, Garcia JG. 2003. Role of CaM kinase II and ERK activation in thrombininduced endothelial cell barrier dysfunction. Am J Physiol Lung Cell Mol Physiol 285(1):L43–L54. Bruewer M, Luegering A, Kucharzik T, Parkos CA, Madara JL, Hopkins AM, Nusrat A. 2003. Proinflammatory cytokines disrupt epithelial barrier function by apoptosis-independent mechanisms. J Immunol 171(11):6164–6172. Bruewer M, Utech M, Ivanov AI, Hopkins AM, Parkos CA, Nusrat A. 2005. Interferon-gamma induces internalization of epithelial tight junction proteins via a macropinocytosis-like process. FASEB J 19(8):923–933. Cai H, Liu D, Garcia JG. 2008. CaM Kinase II-dependent pathophysiological signalling in endothelial cells. Cardiovasc Res 77(1):30–34. Capaldo CT, Nusrat A. 2009. Cytokine regulation of tight junctions. Biochim Biophys Acta 1788(4):864–871. Chesler DA, Reiss CS. 2002. The role of IFN-gamma in immune responses to viral infections of the central nervous system. Cytokine Growth Factor Rev 13(6):441–454. Cho SJ, Pyo S. 2010. Interferon-gamma enhances the apoptosis of macrophages under trophic stress through activation of p53 and the JAK1 pathway. Arch Pharm Res 33(2):285–291. Cui X, Zhang X, Yin Q, Meng A, Su S, Jing X, Li H, Guan X, Li X, Liu S, Cheng M. 2014. F-actin cytoskeleton reorganization is associated with hepatic stellate cell activation. Mol Med Rep 9(5):1641–1647. Detjen KM, Farwig K, Welzel M, Wiedenmann B, Rosewicz S. 2001. Interferon gamma inhibits growth of human pancreatic carcinoma cells via caspase-1 dependent induction of apoptosis. Gut 49(2):251–262. Fish SM, Proujansky R, Reenstra WW. 1999. Synergistic effects of interferon gamma and tumour necrosis factor alpha on T84 cell function. Gut 45(2):191–198. Frausto-Del-Rio D, Soto-Cruz I, Garay-Canales C, Ambriz X, Soldevila G, Carretero-Ortega J, Vazquez-Prado J, Ortega E. 2012. Interferon gamma induces actin polymerization, Rac1 activation and down regulates phagocytosis in human monocytic cells. Cytokine 57(1):158–168. Goeckeler ZM, Wysolmerski RB. 1995. Myosin light chain kinase-regulated endothelial cell contraction: the relationship between isometric tension, actin polymerization, and myosin phosphorylation. J Cell Biol 130(3):613–627.

522

Gong C, Stoletov KV, Terman BI. 2004. VEGF treatment induces signaling pathways that regulate both actin polymerization and depolymerization. Angiogenesis 7(4):313–321. Ikeda H, Old LJ, Schreiber RD. 2002. The roles of IFN gamma in protection against tumor development and cancer immunoediting. Cytokine Growth Factor Rev 13(2):95–109. Jiang W, Jin G, Ma D, Wang F, Fu T, Chen X, Jia K, Marikar FM, Hua Z. 2012. Modification of cyclic NGR tumor neovasculature-homing motif sequence to human plasminogen kringle 5 improves inhibition of tumor growth. PLoS One 7(5):e37132. Kalliolias GD, Ivashkiv LB. 2009. Targeting cytokines in inflammatory diseases: focus on interleukin-1-mediated autoinflammation. F1000 Biol Rep 1:70. Koestler SA, Rottner K, Lai F, Block J, Vinzenz M, Small JV. 2009. F- and G-actin concentrations in lamellipodia of moving cells. PLoS One 4(3):e4810. Lazaro-Dieguez F, Egea G. 2007. Comparative study of the impact of the actin cytoskeleton on cell shape and membrane surface in mammalian cells in response to actin toxins. In: Modern research and educational topics in microscopy, vol. 1. Badajoz, Spain: Formatex. pp. 362–369. McKay DM, Watson JL, Wang A, Caldwell J, Prescott D, Ceponis PM, Di Leo V, Lu J. 2007. Phosphatidylinositol 3¢kinase is a critical mediator of interferon-gamma-induced increases in enteric epithelial permeability. J Pharmacol Exp Ther 320(3):1013–1022. McKenzie JA, Ridley AJ. 2007. Roles of Rho/ROCK and MLCK in TNF-alpha-induced changes in endothelial morphology and permeability. J Cell Physiol 213(1):221–228. McLaren JE, Ramji DP. 2009. Interferon gamma: a master regulator of atherosclerosis. Cytokine Growth Factor Rev 20(2):125–135. Muzes G, Molnar B, Tulassay Z, Sipos F. 2012. Changes of the cytokine profile in inflammatory bowel diseases. World J Gastroenterol 18(41):5848–5861. Nakamura M, Saito H, Kasanuki J, Tamura Y, Yoshida S. 1992. Cytokine production in patients with inflammatory bowel disease. Gut 33(7):933–937. Niessner M, Volk BA. 1995. Altered Th1/Th2 cytokine profiles in the intestinal mucosa of patients with inflammatory bowel disease as assessed by quantitative reversed transcribed polymerase chain reaction (RT-PCR). Clin Exp Immunol 101(3):428–435. Ostler N, Britzen-Laurent N, Liebl A, Naschberger E, Lochnit G, Ostler M, Forster F, Kunzelmann P, Ince S, Supper V, Praefcke GJ, Schubert DW, Stockinger H, Herrmann C, Sturzl M. 2014. Gamma interferon-induced guanylate binding protein 1 is a novel actin cytoskeleton remodeling factor. Mol Cell Biol 34(2):196–209. Petrache I, Birukova A, Ramirez SI, Garcia JG, Verin AD. 2003. The role of the microtubules in tumor necrosis factoralpha-induced endothelial cell permeability. Am J Respir Cell Mol Biol 28(5):574–581. Petrache I, Verin AD, Crow MT, Birukova A, Liu F, Garcia JG. 2001. Differential effect of MLC kinase in TNF-alpha-induced endothelial cell apoptosis and barrier dysfunction. Am J Physiol Lung Cell Mol Physiol 280(6):L1168–L1178. Schett G, Elewaut D, McInnes IB, Dayer JM, Neurath MF. 2013. How cytokine networks fuel inflammation: toward a cytokine-based disease taxonomy. Nat Med 19(7):822–824. Sedgwick JB, Menon I, Gern JE, Busse WW. 2002. Effects of inflammatory cytokines on the permeability of human lung

NG ET AL.

microvascular endothelial cell monolayers and differential eosinophil transmigration. J Allergy Clin Immunol 110(5): 752–756. Shivanna M, Rajashekhar G, Srinivas SP. 2010. Barrier dysfunction of the corneal endothelium in response to TNF-alpha: role of p38 MAP kinase. Invest Ophthalmol Vis Sci 51(3):1575–1582. Shtrichman R, Samuel CE. 2001. The role of gamma interferon in antimicrobial immunity. Curr Opin Microbiol 4(3):251– 259. Srinivas SP, Satpathy M, Guo Y, Anandan V. 2006. Histamineinduced phosphorylation of the regulatory light chain of myosin II disrupts the barrier integrity of corneal endothelial cells. Invest Ophthalmol Vis Sci 47(9):4011–4018. Stolpen AH, Guinan EC, Fiers W, Pober JS. 1986. Recombinant tumor necrosis factor and immune interferon act singly and in combination to reorganize human vascular endothelial cell monolayers. Am J Pathol 123(1):16–24. Tojkander S, Gateva G, Lappalainen P. 2012. Actin stress fibers—assembly, dynamics and biological roles. J Cell Sci 125(Pt 8):1855–1864. Utech M, Ivanov AI, Samarin SN, Bruewer M, Turner JR, Mrsny RJ, Parkos CA, Nusrat A. 2005. Mechanism of IFNgamma-induced endocytosis of tight junction proteins: myosin II-dependent vacuolarization of the apical plasma membrane. Mol Biol Cell 16(10):5040–5052. Vouret-Craviari V, Boquet P, Pouyssegur J, Van ObberghenSchilling E. 1998. Regulation of the actin cytoskeleton by thrombin in human endothelial cells: role of Rho proteins in endothelial barrier function. Mol Biol Cell 9(9):2639– 2653. Waschke J, Curry FE, Adamson RH, Drenckhahn D. 2005. Regulation of actin dynamics is critical for endothelial barrier functions. Am J Physiol Heart Circ Physiol 288(3):H1296– H1305. Wojciak-Stothard B, Potempa S, Eichholtz T, Ridley AJ. 2001. Rho and Rac but not Cdc42 regulate endothelial cell permeability. J Cell Sci 114(Pt 7):1343–1355. Wong RK, Baldwin AL, Heimark RL. 1999. Cadherin-5 redistribution at sites of TNF-alpha and IFN-gamma-induced permeability in mesenteric venules. Am J Physiol 276(2 Pt 2):H736–H748. Youakim A, Ahdieh M. 1999. Interferon-gamma decreases barrier function in T84 cells by reducing ZO-1 levels and disrupting apical actin. Am J Physiol 276(5 Pt 1):G1279– G1288. Yuan SY, Rigor RR. 2010. Regulation of endothelial barrier function. In: Chapter 4: The endothelial barrier. San Rafael, CA: Morgan & Claypool Life Sciences. Available from www.ncbi.nlm.gov/books/NBK54116/

Address correspondence to: Dr. Zuraini Ahmad Department of Biomedical Sciences Faculty of Medicine and Health Sciences Universiti Putra Malaysia UPM Serdang 43400, Selangor Malaysia E-mail: [email protected] Received 15 October 2014/Accepted 9 January 2015