Caveolin-1 scaffolding domain promotes leukocyte adhesion by reduced basal endothelial nitric oxide-mediated ICAM-1 phosphorylation in rat mesenteric venules

Abstract

Exogenously applied caveolin-1 scaffolding domain (CAV) has been shown to inhibit inflammatory mediator-induced nitric oxide (NO) production and NO-mediated increases in microvessel permeability. However, the effect of CAV on endothelial basal NO that prevents leukocyte adhesion remains unknown. This study aims to investigate the roles of exogenously applied CAV in endothelial basal NO production, leukocyte adhesion, and adhesion-induced changes in microvessel permeability. Experiments were conducted in individually perfused rat mesenteric venules. Microvessel permeability was determined by measuring hydraulic conductivity (Lp). NO was quantified with fluorescence imaging in DAF-2-loaded vessels. Perfusing venules with CAV inhibited basal NO production without affecting basal Lp. Resuming blood flow in CAV-perfused vessels significantly increased leukocyte adhesion. The firmly adherent leukocytes altered neither basal Lp nor adherens junction integrity. Increases in Lp occurred only upon formyl-Met-Leu-Phe application that induces release of reactive oxygen species from the adherent leukocytes. The application of NO synthase inhibitor showed similar results to CAV, and NO donor abolished CAV-mediated leukocyte adhesion. Immunofluorescence staining showed increases in binding of ICAM-1 to an adhesion-blocking antibody concurrent with a Src-dependent ICAM-1 phosphorylation following CAV perfusion. Pre-perfusing vessels with anti-ICAM-1 blocking antibody or a Src kinase inhibitor attenuated CAV-induced leukocyte adhesion. These results indicate that the application of CAV, in addition to preventing excessive NO-mediated permeability increases, also causes reduction of basal NO and promotes ICAM-1-mediated leukocyte adhesion through Src activation-mediated ICAM-1 phosphorylation. CAV-induced leukocyte adhesion was uncoupled from leukocyte oxidative burst and microvessel barrier function, unless in the presence of a secondary stimulation.

nitric oxide (NO), in addition to being a potent vasodilator, has also been considered a “double-edged sword” mediator in inflammation. Excessive NO production induced by inflammatory mediators contributes to an increase in microvessel permeability and plays a pro-inflammatory role (15, 32, 41). Constitutive (or basal) NO production, on the other hand, has been reported to prevent leukocyte adhesion and platelet aggregation on the vessel wall, therefore, playing an anti-inflammatory role (21, 22, 37).

Caveolin-1 (CAV), the major structural protein in the caveolae of endothelial cells, is a recognized endogenous inhibitor of endothelial NO synthase (eNOS) (11, 19). The discovery of the scaffolding domain of CAV that directly interacts with eNOS, and the use of Antennapedia homeodomain (AP) fusions to facilitate CAV uptake by living cells, provided a valuable tool for investigating the mechanisms involved in the regulation of eNOS activity through protein-protein interaction and NO-dependent vascular functions in vivo. Our previous studies demonstrated that exogenously applying AP-CAV to a perfused intact microvessel inhibits platelet activating factor (PAF)-induced NO production and NO-mediated increases in microvessel permeability (44, 50). In whole animal studies, the administration of AP-CAV suppressed tissue edema and vascular leakage (2). These studies suggest an anti-inflammatory therapeutic potential for AP-CAV. However, whether the exogenously applied CAV peptide affects basal endothelial NO production has not been studied. If AP-CAV inhibits basal NO, whether it causes leukocyte adhesion and increases microvessel permeability is unknown. This study directly addresses these questions by measuring basal NO, microvessel permeability, and leukocyte adhesion in AP-CAV-perfused intact microvessels.

Experiments were conducted in individually perfused venular microvessels in rat mesenteries. We first quantitatively measured the effect of AP-CAV on endothelial basal NO production with DAF-2 using fluorescence imaging. We then examined the effect of AP-CAV on leukocyte adhesion, the mechanisms involved in the increases in adhesive activity of endothelial cells, and their impact on microvessel permeability. Permeability was determined by measuring hydraulic conductivity (Lp). The changes in endothelial junction integrity and adhesive capacity of adhesion molecules were evaluated by confocal imaging of immunofluorescence staining on AP-CAV-perfused microvessels.

METHODS

Animal preparation.

Experiments were carried out on venular microvessels in rat mesenteries. All procedures and animal use were approved by the Animal Care and Use Committee at West Virginia University. Female Sprague-Dawley rats (2- to 3-mo-old; body weight of 220–250 g; Hilltop Laboratory Animals; Scottdale, PA) were anesthetized with pentobarbital sodium given subcutaneously. The initial dosage was 65 mg/kg of body wt, with an additional 3-mg dose given as needed. The trachea was intubated, and a midline surgical incision (1.5–2 cm) was made in the abdominal wall. The mesentery was gently moved out of the abdominal cavity and spread over a glass coverslip attached to an animal tray. The upper surface of the mesentery was continuously superfused with mammalian Ringer solution. The temperature of the superfusate was maintained at 37°C. All experiments were carried out on venular microvessels, with diameters ranging between 40 and 50 μm. Each experiment was performed on a single microvessel with one experiment per animal.

Measurement of Lp in individually perfused rat mesenteric microvessels.

Modified Landis technique was used to measure the volume flux of water across the microvessel wall. The methods have been evaluated in detail (5, 20). In brief, a single microvessel was cannulated with a micropipette and perfused with albumin-Ringer solution (control), containing 1% (vol/vol) hamster red blood cells as markers. A hydrostatic pressure (range of 40–60 cmH₂O) was applied through the micropipette to the microvessel lumen to maintain a continuous flow. Water flux was measured when the vessel was briefly occluded downstream. Lp was calculated as the slope of the relationship between the initial water flux per unit area of microvessel wall and the pressure difference across the microvessel wall. In each experiment, the baseline Lp and the Lp after application of testing solutions were measured in the same vessel, and the changes in Lp were expressed as the ratio of Lp_test/Lp_control.

Fluorescence imaging of endothelial NO production.

Endothelial NO production was measured on DAF-2 DA-loaded vessels using fluorescence imaging. Experiments were performed on a Nikon Diaphod 300 microscope equipped with a 12-bit digital charge coupled device camera (ORCA; Hamamatsu) and a computer-controlled shutter (Lambda 10-2; Sutter Instrument; Novato, CA). A 75-W xenon lamp was used as the light source. The excitation wavelength for DAF-2 was selected by an interference filter (480/40 nm), and emission was separated by a dichroic mirror (505 nm) and a band-pass barrier (535/50 nM). All the images were acquired and analyzed using Metafluor software (Universal Imaging). To minimize the photo-bleaching, a neutral density filter (0.5 ND) was positioned in front of the interference filter, and the exposure time was minimized to 0.12 s at 1-min intervals.

During the experiments, each vessel was cannulated and perfused with albumin-Ringer solution containing DAF-2 DA (5 μM) throughout the experimental duration. Image collection was started at the beginning of DAF-2 DA perfusion. All images were collected from a group of endothelial cells located in the same focal plane of the vessel wall using a Nikon Fluor lens (×20; numerical aperture, 0.75). A steady-state of endothelial DAF-2 fluorescence (FI_DAF) with continuous perfusion of DAF-2 DA was indicated by the rate change of the increased fluorescence intensity (FI). Quantitative analysis was conducted at the individual endothelial cell level using selected regions of interest (ROIs) along the vessel wall. Each ROI covers the area of one individual cell as indicated by the fluorescence outline. The tissue autofluorescence was subtracted from all of the measured FIs. The basal NO production rate was calculated from the slope of the FI_DAF increase during albumin-Ringer perfusion after DAF-2 loading reached the steady-state. FI_DAF was expressed in arbitrary units (AU). The rate of FI_DAF increase (df/dt) derived by first differential conversion of cumulative FI_DAF over time indicates the rate of NO production. Details have been described previously (45).

Immunofluorescence staining and confocal imaging.

VE-cadherin and adherent leukocyte staining was conducted in vessels after CAV-induced leukocyte adhesion. Phosphorylated ICAM-1 at Y526 staining was conducted in vessels after 30 min of AP-CAV perfusion in the presence or absence of the Src kinase inhibitor PP1. The rat mesentery bearing the perfused vessel was fixed with 1% paraformaldehyde and treated with Triton X-100 before exposure to primary antibodies against VE-cadherin (C-19, sc-6458, Santa Cruz Biotechnology) and CD45 (clone OX-1, Biolegend), or phospho-(Y526)-ICAM-1 (ab51033, Abcam), and followed by incubation with their corresponding secondary antibodies (Invitrogen).

ICAM-1 staining was conducted in intact microvessels. Each venule was first perfused with albumin-Ringer solution for 10 min and then perfused with ICAM-1 primary antibodies (1A29 BD Pharmingen, 0.1 mg/ml) for 20 min. The secondary antibody (Invitrogen) was applied to each vessel for 15 min after washout of the primary antibody from the vessel lumen. Nuclei were labeled by perfusion vessels with DRAQ5 (Biostatus). The confocal images were acquired after all of the fluorescence markers were washed out from the vessel lumen.

All of the images were acquired using Leica confocal microscope. Each image stack was collected by optical sectioning at successive x-y focal planes with a vertical depth of 0.3 μm [phospho-(Y526)-ICAM-1] and 0.5 μm (ICAM-1) using Leica ×25 objective (HC Plan APO, NA 0.95) and 1,024 × 1,024 scanning format. Leica software was used for image acquisition and image analysis.

The FI of labeled ICAM-1 and phospho-(Y526)-ICAM-1 was quantified from a segment of the vessel wall. The total FI was calculated as area × depth × mean amplitude, where the area is the selected ROI per vessel section and the depth is the total number of images at z dimension. Because ICAM-1 was expressed on endothelial cell surface, FI was quantified as total intensity per square micrometer of vessel wall (FI/A). Assuming a cylindrical geometry, surface area of the vessel wall was calculated as 2π × r × L, where r is the radius of the microvessel and L is the length of selected ROI from the vessel.

Solutions and reagents.

Mammalian Ringer solution (16) was used for the experiments. All perfusates contained albumin-Ringer solution (BSA; 10 mg/ml). AP-CAV, the fusion peptide of CAV scaffolding domain with AP, the Antennapedia internalization sequence from Drosophila Antennapedia homeodomain, and AP-CAV-X, the fusion peptide of scrambled CAV with AP, were synthesized by Tufts University (2). The chemotactic peptide formyl-Met-Leu-Phe-OH (fMLP) was purchased from Calbiochem (San Diego, CA). PP1 [4-amino-5-(4-methylphenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine], sodium nitroprusside, and N(Ω)-monomethyl-l-arginine (l-NMMA) were purchased from Sigma (St. Louis, MO). The stock solutions were prepared with 100% DMSO, and each final solution was achieved by >1:1,000 dilution with albumin-Ringer solution.

Data analysis and statistics.

All values are means ± SE. Paired t-tests were used for paired data analysis. ANOVA was used to compare data between groups. A probability value of P < 0.05 was considered statistically significant.

RESULTS

AP-CAV inhibits basal NO production without affecting basal Lp in intact venules.

The effect of AP-CAV on basal NO was examined in four vessels. A steady-state of FI_DAF was reached at 39.5 ± 1.3 min with continuous DAF-2 DA perfusion. The FI_DAF accumulation rate, an indication of basal NO production rate, was 0.13 ± 0.01 AU/min. After perfusion of each vessel with AP-CAV (10 μM), FI_DAF accumulation rate significantly decreased to 0.02 ± 0.01 AU/min within 1 min (P < 0.01). To confirm that DAF-2 was still functional after the application of AP-CAV, a NO donor, SNP (10 μM), was added to the perfusate in the presence of AP-CAV in two of the vessels. The FI_DAF increased at a relatively linear rate of 0.15 ± 0.01 AU/min, which is comparable to that under control conditions (Fig. 1, A and C). Based on these results, we estimate that the level of NO generated by 10 μM SNP is close to basal NO production rate. To compare the inhibitory effect of AP-CAV on basal NO with that of the NOS inhibitor, basal NO production was measured in another eight vessels in the absence or presence of l-NMMA (500 μM, four vessels per group). The mean FI_DAF accumulation rate of four vessels was 0.13 ± 0.02 AU/min and was significantly decreased to 0.01 ± 0.002 AU/min in the presence of l-NMMA (Fig. 1, B and C), demonstrating a similar magnitude of basal NO reduction to that of AP-CAV.

Fig. 1.

The application of Antennapedia homeodomain (AP)-caveolin-1 (CAV) inhibits basal nitric oxide (NO). A: the time-dependent changes in cumulative DAF-2 fluorescence intensity (FI_DAF) from one representative experiment before (○) and after (●) treatment with 10 μM AP-CAV, followed by 10 μM SNP (△) perfusion for 40 min. B: time-dependent changes in cumulative FI_DAF from two representative experiments in the absence (●) or presence of (○) of 500 μM l-NMMA. C: summary of NO production rate derived from cumulative FI_DAF in each group of vessels. The dashed arrow indicates the time when DAF-2 fluorescence reached steady state. The arrow indicates the time when testing solution was added. †Significant decreases from control (P < 0.05).

To examine whether reduced basal NO via perfusion of AP-CAV has a direct effect on basal permeability, we measured Lp in four microvessels. The mean baseline Lp of the four vessels was 1.8 ± 0.2 × 10⁻⁷ cm·s⁻¹·cmH₂O⁻¹. After vessels were perfused with AP-CAV (10 μM) for 30 min, the mean Lp was 1.8 ± 0.3 ×10⁻⁷ cm·s⁻¹·cmH₂O⁻¹, which was not significantly different from that of the control. Figure 2 shows a single experiment and the data summary. Our previous study demonstrated that perfusion of mesenteric venules with AP-CAV at 1 μM for 30 min also did not affect basal Lp (44).

Fig. 2.

Perfusion of AP-CAV for 30 min has no effect on basal hydraulic conductivity (Lp). A: a representative experiment shows the time course of Lp changes when the vessel was perfused with AP-CAV (10 μM). B: summary results of four experiments.

AP-CAV perfusion induces leukocyte adhesion without increasing microvessel permeability unless in the presence of a secondary stimulation.

The effect of AP-CAV-induced basal NO reduction on leukocyte adhesion was examined in eight microvessels. Each vessel was perfused with 1 or 10 μM of AP-CAV for 30 min, followed by resuming blood flow in the perfused microvessel for 10 min. When each vessel was recannulated with albumin-Ringer perfusate, we observed a dose-dependent leukocyte adhesion. In vessels perfused with AP-CAV at 1 μM, the adherent leukocytes increased from baseline levels of 0.7 ± 0.1 to 8.4 ± 0.1 per 100 μm of vessel length (n = 3; P < 0.05). In 10 μM AP-CAV-perfused vessels, the mean adherent leukocytes increased from baseline levels of 1.3 ± 0.2 to 26.3 ± 3.2 per 100 μm of vessel length (n = 5). Figure 3A, left, demonstrates the representative images of leukocyte adhesion before and after AP-CAV (10 μM) perfusion, followed by resumed blood flow from a single vessel. When Lp was measured in the presence of CAV-induced adherent leukocytes, the mean value was 2.1 ± 0.1 × 10⁻⁷ cm·s⁻¹·cmH₂O⁻¹, which was not significantly different from the mean baseline Lp of 1.7 ± 0.2 ×10⁻⁷ cm·s⁻¹·cmH₂O⁻¹.

Fig. 3.

Reduction of basal NO by AP-CAV induced significant leukocyte adhesion without increasing Lp in the absence of a secondary stimulation. A: paired video images from two rat venules. The two images on the left show the same vessel under control conditions and after AP-CAV (10 μM)-induced leukocyte adhesion. The two images on the right show that the application of sodium nitroprusside (SNP) abolished AP-CAV-induced leukocyte adhesion. B: the time course of Lp changes from a representative experiment showing that AP-CAV (10 μM)-induced leukocyte adhesion (25/100 μm of vessel length) did not increase Lp, unless formyl-Met-Leu-Phe-OH (fMLP) (10 μM) was added to the perfusate. C: summary graph showing the correlation between the number of adherent leukocytes (per 100 μm vessel length) and the changes in Lp in four group of studies. Perfusion vessels with AP-CAV at 1 μM (n = 3) and 10 μM (n = 5) show AP-CAV dose-dependent increases in leukocyte adhesion and fMLP-induced increases in Lp. The application of SNP attenuated AP-CAV-induced leukocyte adhesion (n = 4), and replacing AP-CAV with scrambled AP-CAV (AP-CAV-X) showed no effect on leukocyte adhesion (n = 3). The blank bars represent the control values. The arrows indicate the procedures of 30 min of AP-CAV or AP-CAV-X perfusion followed by 10 min of resumed blood flow. The dashed line bars represent values measured after resumed blood flow in AP-CAV or AP-CAV-X perfused vessels. *Significant increases from the control values (P < 0.05). †Significant decreases from AP-CAV group (P < 0.05).

Our previous study showed that fMLP can stimulate ROS release from TNF-α-induced adherent leukocytes, resulting in increases in microvessel Lp (48). Here, we further investigated whether exposing CAV-induced adherent leukocytes to fMLP increases Lp. We added fMLP (10 μM) to the perfusate after leukocyte adhesion in AP-CAV perfused vessels. To maximize the local concentration of ROS released from the adherent leukocytes upon fMLP stimulation, the perfusate was kept at a balanced pressure (no flow) for 5 min before Lp measurement. Lp measured after 5 min of stationary flow transiently increased to mean peak values of 4.8 ± 0.6 and 16.3 ± 2.1 times that of the control in 1 and 10 μM AP-CAV perfused vessels, respectively. The 5-min stationary flow in the absence of adherent leukocytes does not affect Lp (48). Figure 3B shows a typical time course of the Lp changes in a single experiment. Figure 3C shows result summary (two left panels).

NO donor prevents AP-CAV-induced leukocyte adhesion.

To examine whether the AP-CAV-induced reduction of basal NO is responsible for increased leukocyte adhesion, we applied a NO donor, SNP, during AP-CAV perfusion in four vessels. The mean baseline Lp was 1.7 ± 0.1 ×10⁻⁷ cm·s⁻¹·cmH₂O⁻¹. Each vessel was first perfused with AP-CAV (10 μM) and SNP (10 μM) for 30 min, followed by 10 min of resumed blood flow with superfusate containing SNP (20 μM). When each vessel was recannulated with albumin-Ringer perfusate, the mean number of adherent leukocytes was only 3.7 ± 1.1 per 100 μm of vessel length, which was a significant reduction from AP-CAV perfusion alone and not significantly different from that of the baseline. Figure 3A, right, demonstrates the representative images before and after AP-CAV and SNP perfusion, followed by resumed blood flow from one single vessel. We also measured Lp when vessels were perfused with SNP and AP-CAV, which showed no significant difference from that of the control (Fig. 3C, third panel from left).

Scrambled AP-CAV peptide does not affect leukocyte adhesion and microvessel Lp.

AP-CAV-X (10 μM) was used to examine the specificity of AP-CAV in another three microvessels. Experiments were conducted under identical conditions to that of AP-CAV application. No significant leukocyte adhesion or changes in Lp were observed after blood flow was resumed in AP-CAV-X perfused vessels. The mean number of adherent leukocytes before and after resumption of blood flow was 0.8 ± 0.1 and 1.1 ± 0.2 per 100 μm of vessel length, and the mean Lp measured under each condition was 1.7 ± 0.2 and 1.8 ± 0.2 × 10⁻⁷ cm·s⁻¹·cmH₂O⁻¹, respectively (Fig. 3C, right).

Inhibition of NOS by l-NMMA shows similar results to the application of AP-CAV.

If AP-CAV-induced leukocyte adhesion is attributed to its inhibitory effect on basal NO, we would expect the similar effect from the application of N^G-methyl-l-arginine (l-NMMA), a NOS inhibitor. Experiments were conducted in three microvessels. Figure 4A shows a representative time course of the Lp changes in a single experiment. The mean control Lp of three vessels was 2.4 ± 0.3×10⁻⁷ cm·s⁻¹·cmH₂O⁻¹. After application of l-NMMA to both perfusate (100 μM) and superfusate (500 μM) for 30 min followed by 20 min of resumed blood flow, the mean adherent leukocytes to the microvessel wall increased from a mean baseline level of 0.8 ± 0.2 to 14.8 ± 2.5 per 100 μm of vessel length. Consistent with the results of AP-CAV-induced leukocyte adhesion, Lp measured in the presence of l-NMMA-induced adherent leukocytes was 2.7 ± 0.6 ×10⁻⁷ cm·s⁻¹·cmH₂O⁻¹, which is not significantly different from its control. When fMLP was applied to the perfusate, Lp measured after 5 min of stationary flow transiently increased to a mean peak value of 5.4 ± 0.9 times that of the control. Figure 4B summarizes the results.

Fig. 4.

l-NMMA, a NO synthase (NOS) inhibitor, showed similar effects on leukocyte adhesion and microvessel permeability to those of CAV. A: time course of Lp changes from a representative experiment. B: summary of the number of adherent leukocytes (per 100-μm vessel length) and the corresponding changes in microvessel Lp (n = 3). *Significant increases from control (P < 0.05).

AP-CAV-induced leukocyte adhesion does not change endothelial adherens junction.

Our results showed that AP-CAV-induced leukocyte adhesion did not change basal Lp. We then further examined the location of adherent leukocytes and whether the adherent leukocytes cause local structural changes of endothelial junctions. Immunofluorescence staining was used to simultaneously label adherent leukocytes and VE-cadherin in three vessels. The representative confocal images shown in Fig. 5 illustrate that AP-CAV-induced leukocyte adhesion did not change the distributions of VE-cadherin at endothelial junctions. Endothelial junctions were outlined by continuous VE-cadherin that had no apparent interruptions, even underneath the site of the attached leukocyte. In addition, the dual staining of both leukocyte and VE-cadherin demonstrated that the majority of adherent leukocytes selectively adhered at or near endothelial junctions.

Fig. 5.

Confocal images of the co-staining of VE-cadherin and adherent leukocytes illustrating that CAV-induced leukocyte adhesion did not change VE-cadherin distribution. A: VE-cadherin staining under control conditions and after CAV-induced leukocyte adhesion. The third image shows the double staining of VE-cadherin and adherent leukocytes (indicated by arrows). B: magnified images from three different vessels showing no changes in VE-cadherin at the adhesion sites.

Reduced basal NO by AP-CAV increases the adhesive capacity of ICAM-1 on the surface of endothelium lining microvessel walls.

Our results showed that leukocytes were firmly attached to the venular walls after resuming blood flow in vessels that have been exposed to AP-CAV. We then investigated whether AP-CAV-induced NO reduction affects the adhesive capacity of intercellular adhesion molecule-1 (ICAM-1), an important molecule mediating the firm adhesion of leukocytes to the vascular endothelium. A monoclonal blocking antibody directed against ICAM-1 (mAb1A29) (9, 34) was used to evaluate the changes in adhesive activity of ICAM-1 on endothelial cells lining the venular walls after the vessel was perfused with AP-CAV for 30 min. AP-CAV perfusion at 1 and 10 μM significantly increased the mean FI of mAb1A29 to 2.4 ± 0.2 and 9.5 ± 1.7 times that of the control, respectively (n = 4 per group). The application of 10 μM SNP abolished AP-CAV-induced increases in mAb1A29 binding (n = 3), and perfusion of AP-CAV-X (10 μM) showed no difference in mAb1A29 binding from that of the control (n = 3). Figure 6 shows representative confocal images and summarized FI quantification of mAb1A29 staining in five group of studies.

Fig. 6.

AP-CAV-induced increase in ICAM-1 adhesive capacity causes leukocyte adhesion. A: representative confocal images of anti-ICAM-1 blocking antibody mAb1A29 and vascular cell nuclei co-staining from five groups of studies and the secondary antibody control (ICAM-1 is shown in green, and nuclei are red). B: quantification of total fluorescence intensity (FI) of ICAM-1 per unit area of vessel wall under control conditions (n = 4), after AP-CAV (1 μM, n = 3; and 10 μM, n = 4) perfusion in the absence and presence of SNP (n = 3), and after AP-CAV-X perfusion (n = 3). C: perfusion of vessels with mAb1A29 significantly attenuated AP-CAV (10 μM)-induced leukocyte adhesion (n = 5 per group). *Significant increase from control (P < 0.05). †Significant decrease from AP-CAV group (P < 0.05).

Anti-ICAM-1 blocking antibody attenuates AP-CAV-induced leukocyte adhesion.

To confirm that the specific binding of mAb1A29 to ICAM-1 are directly associated with the adhesive activity of ICAM-1, we perfused microvessels with mAb1A29 before and during AP-CAV application. The initial number of adherent leukocytes under albumin-Ringer perfusion was 1.0 ± 0.3 per 100 μm of vessel length (n = 5). Each vessel was then perfused with mAb1A29 (0.1 mg/ml) for 30 min followed by perfusion of AP-CAV (10 μM) in the presence of mAb1A29 for another 30 min before blood flow was resumed. When each vessel was recannulated with albumin-Ringer perfusate after 10 min of blood flow, the mean number of adherent leukocytes was only 4.5 ± 0.8 per 100 μm of vessel length, which is a significant reduction from that of AP-CAV perfusion alone (26.3 ± 3.2 per 100 μm of vessel length). Figure 6C summarizes the results.

ICAM-1 phosphorylation at tyrosine 526 is required for AP-CAV-induced leukocyte adhesion.

Studies have demonstrated that TNF-α-induced early phase of leukocyte adhesion involves Src activation-mediated phosphorylation of ICAM-1 (18, 31). We then examined whether this mechanism also applies to AP-CAV-induced ICAM-1-dependent leukocyte adhesion by using fluorescence immunostaining and confocal imaging. We found that 30 min of perfusion of AP-CAV (10 μM) caused a marked increase in ICAM-1 phosphorylation at tyrosine 526 (Y526) on endothelial cells lining the venular walls compared with that of the control. The same period of AP-CAV-X perfusion showed no such effect. Preperfusing vessels with PP1 (10 μM), a src kinase inhibitor, abolished AP-CAV-induced ICAM-1 phosphorylation. Representative confocal images are shown in Fig. 7A, and the FI quantifications are summarized in Fig. 7B (n = 3 per group). The role of AP-CAV-induced ICAM-1 phosphorylation in AP-CAV-induced leukocyte adhesion was examined in another three vessels. Each vessel was first perfused with PP1 (10 μM) for 30 min before AP-CAV (10 μM) was added for another 30-min perfusion. After 10 min of resumed blood flow, the mean number of adherent leukocytes was only 4.6 ± 0.1 per 100 μm of vessel length, which was significantly lower than that of AP-CAV perfusion alone (26.3 ± 3.2 per 100 μm of vessel length). Figure 7C summarizes the results.

Fig. 7.

AP-CAV-induced leukocyte adhesion involves Src activation-mediated ICAM-1 phosphorylation at tyrosine 526 (Y526). A: representative confocal images demonstrate increases in ICAM-1 phosphorylation at Y526 following AP-CAV perfusion relative to that of the control. Such increased ICAM-1 phosphorylation was blocked by the application of a Src kinase inhibitor, PP1, and was absent in vessels perfused with AP-CAV-X or secondary antibody alone. Phosphorylated ICAM-1 at Y526 is shown in green, and nuclei are red. B: summary of changes in FI of phospho-Y526-ICAM-1 relative to control in each experimental group (n = 3 per group). C: perfusion of vessels with PP1 that prevented ICAM-1 phosphorylation also significantly attenuated AP-CAV induced leukocyte adhesion (n = 3). *Significant increase from control (P < 0.05). †Significant decrease from AP-CAV group (P < 0.05).

DISCUSSION

Our study demonstrated for the first time that exogenously applied AP-CAV, in addition to inhibition of stimulus-induced increases in NO production and microvessel permeability (44, 46, 50), also reduces basal NO production in endothelial cells lining microvessel walls. Most importantly, this study revealed how reduced basal NO affects the adhesiveness and barrier function of endothelial cells in intact microvessels. Although reduced basal NO by AP-CAV did not directly change basal microvessel permeability, it increased the adhesive capacity of endothelial cells and promotes leukocyte adhesion via Src-mediated phosphorylation of ICAM-1. Our results that NO donor abolished the effect of AP-CAV on increased ICAM-1 binding to a blocking antibody and leukocyte adhesion support an important anti-adhesive role of basal NO in maintaining the homeostasis of vascular walls. Although the experiments focused on the effect of AP-CAV on microvessels, the identified cellular mechanisms may represent the microvascular manifestations associated with reduced basal NO under various pathological conditions in vivo.

CAV reduces basal NO production and promotes leukocyte adhesion in intact venules.

The relationship between CAV and eNOS is one of the fundamental discoveries of caveolae involvement in cardiovascular functions. Cell culture and purified protein binding studies demonstrated that CAV inhibits eNOS activity by a direct interaction with eNOS through its NH₂-terminal “scaffolding domain,” located between amino acids 82 and 101 within endothelial plasmalemmal caveolae (11, 19, 30). Although the effect of the CAV scaffolding domain peptide on eNOS activity has been extensively studied for decades, most of the in vivo studies were conducted in vessels with acute exposure to inflammatory mediators (2, 3, 44, 50). We previously demonstrated that, when individually perfused venules were exposed to PAF, the excessive NO production triggered by increased endothelial [Ca²⁺]_i was a necessary signal for permeability increases (46, 49) and that the application of AP-CAV, which inhibits PAF-induced NO, prevents the increases in microvessel permeability (44, 50). The systemic application of AP-CAV in mice has been shown to suppress carrageenan and mustard oil-induced vascular leakage, as well as the hyperpermeability of tumor vasculature (2, 3, 12). This in vivo evidence suggests a promising anti-inflammatory therapeutic potential of AP-CAV in protecting endothelial barrier function through its inhibitory effect on inflammation-associated eNOS activation (1, 2, 44, 50). However, NO has also been known as an anti-adhesive molecule and plays anti-inflammatory roles in preventing platelet and leukocyte adhesion on microvessel walls (22, 42). How the inhibitory action of AP-CAV on eNOS affects the anti-inflammatory roles of basal NO has not been previously investigated. Detecting changes in basal NO requires a sensitive method, because the magnitude of basal NO production is much smaller than that induced by inflammatory mediators. Using continuous perfusion of NO indicator DAF-2 to measure basal NO is a new method recently developed in our laboratory (45). This method enables the endothelial cells lining the microvessel wall to maintain a relatively constant dye concentration, which enhances the sensitivity of NO detection. Using this method, we provided direct evidence that the exogenously applied AP-CAV, in addition to its role in suppression of stimulus-induced increases in NO production and microvessel permeability, also inhibits basal NO in intact venules and promotes leukocyte adhesion. The addition of NO donor that reversed the CAV effect on leukocyte adhesion supports a causal relationship between the reduced NO and increased leukocyte adhesion. Based on these findings, we further investigated the mechanisms involved in increased adhesiveness of microvessel walls by reduced basal NO and the impact of increased adhesive activity of endothelium on microvessel permeability.

Role of basal NO in the regulation of microvessel permeability.

NO has been recognized as an important endothelial factor that is not only a potent vasodilator regulating vascular tone and blood pressure but is also an important intrinsic modulator of microvessel permeability (4, 6, 21, 23, 27). Despite decades of studies, the exact roles of NO in the regulation of microvessel permeability remain controversial. Some studies showed that the application of NOS inhibitors increased vascular leakages, especially in the presence of blood (21, 24, 26, 38). Others showed that inhibition of NO prevented permeability increases in response to inflammatory mediators (1–3, 6, 32, 33, 44, 46, 49, 50). The inconsistent results were usually derived from different experimental approaches, organ vasculatures, tissue conditions, and animal species. Our present study was conducted in intact microvessels under basal conditions. Combined with our previous findings on stimuli-induced NO and microvessel permeability (44, 46, 49, 50), we revealed dual actions of NO on endothelial functions using the same experimental preparation, but under different conditions, which provides a unified perspective of the roles of NO in the regulation of endothelial function and microvessel permeability. Our results demonstrated that reducing basal NO production by AP-CAV perfusion increased leukocyte adhesion through an alteration of endothelial ICAM-1 adhesive capacity on the surface of endothelial cells lining intact microvessel walls. Neither the increased adhesiveness of ICAM-1 alone nor the ICAM-1-mediated leukocyte adhesion directly affects basal microvessel permeability. However, the presence of chemokines can trigger a respiratory burst from adherent leukocytes, resulting in ROS-mediated increases in microvessel permeability (Figs. 3 and 4). These results demonstrate that the reduced basal NO does not directly increase permeability but rather primes the microvascular endothelial cells into an adhesive state. These findings reveal the dual roles of NO in the regulation of endothelial functions in microvessels. Although basal NO plays anti-adhesive roles in maintaining the homeostasis of vascular walls, the inflammatory mediator-induced excessive NO mediates increases in microvessel permeability.

The relationship between reduced basal NO, ICAM-1 expression, leukocyte adhesion, and microvessel permeability.

Leukocyte adhesion to microvessel walls has been characterized as a series of coordinated interactions mediated by different adhesion molecules. During acute inflammation, the rapid recruitment of leukocytes (minutes) usually involves P-selectin translocation and the secretion of chemokines (29, 36), which mediate loose interaction between leukocyte and endothelium, such as rolling and tethering. Firm adhesion, usually occurring hours after cytokine-mediated activation, is mediated by the enhanced expressions of VCAM-1 or ICAM-1 that require transcription and translation of new proteins (27). In contrast to this classical pattern of leukocyte/endothelium interactions, we observed a large number of tightly adhered leukocytes in CAV- and l-NMMA-perfused vessels within 10 and 20 min of resumed blood flow, respectively. The adherent leukocytes were tightly attached to the microvessel walls and could not be washed away by re-perfusing the vessel with albumin-Ringer perfusate, indicating ICAM-1-mediated adhesion. Our results showed a 9.5-fold increase in ICAM-1 immunostaining in intact microvessels following a 30-min CAV perfusion (images in Fig. 6). In this study, the anti-ICAM-1 antibody staining was achieved by direct perfusion of intact vessels with the antibody without fixation. Therefore, the observed antibody binding with ICAM-1 should only represent the levels of extracellular domain of ICAM-1 on the surface of endothelial cells. The timing of increased ICAM-1 staining could rule out the possibility of new ICAM-1 synthesis. Studies have demonstrated that mAb1A29 is a specific blocking antibody for ICAM-1 that inhibits ICAM-1-mediated leukocyte adhesion (9, 34). Our results that perfusing vessels with 1A29 significantly inhibited AP-CAV-induced leukocyte adhesion are consistent with those reports. Then the increased binding affinity of ICAM-1 for this antibody upon AP-CAV perfusion could be the result of increased adhesive binding sites of ICAM-1 for leukocyte adhesion. Although the evidence derived from our present study could not rule out the possibility of membrane translocation of ICAM-1 in response to AP-CAV-induced NO reduction, protein analysis from cytoplasmic and membrane fractions in cultured cells reported that TNF-α-induced early phase leukocyte adhesion involves the activation of constitutive ICAM-1 in cell surface (18). We also explored the signaling pathways involved in AP-CAV-induced increases in the adhesive activity of ICAM-1. The significant increases in phospho-(Y526)-ICAM-1 staining following AP-CAV perfusion, not AP-CAP-X perfusion, showed a direct correlation between ICAM-1 phosporylation and increased endothelial surface ICAM-1 binding to mAb1A29. The application of a Src kinase inhibitor, PP1, that abolished AP-CAV-induced ICAM-1 phosphorylation and also inhibited AP-CAV-induced leukocyte adhesion supports that Src activation-mediated ICAM-1 phosphorylation resulted in increases in the adhesive activity of ICAM-1. Src activation mediated ICAM-1 phosphorylation that was independent of protein synthesis has also reported in TNF-α-induced early phase leukocyte adhesion in both cell culture and mouse lung studies (18, 31). Our results that NOS inhibitor (l-NMMA) showed similar results to those of AP-CAV-mediated leukocyte adhesion and that adding exogenous NO donor (SNP), or blocking ICAM-1, prevented AP-CAV-induced leukocyte adhesion support that reduced basal NO-mediated, Src activation-dependent increases in adhesive activity of ICAM-1 are the mechanisms for AP-CAV-induced leukocyte adhesion. Most importantly, AP-CAV-induced increases in adhesive activity of ICAM-1 on endothelial cells appears sufficient to mediate firm adhesion of leukocytes within 10–20 min without the need of activating blood cells.

Consistent with our findings, overexpression of CAV was shown to reduce NO production in vitro (7, 8, 10), and eNOS-deficient mice show increased levels of leukocyte adhesion (28). A rapid increase (in 30 min) in ICAM-1 expression in thrombin-stimulated umbilical vein endothelial cell was not inhibited by a protein biosynthesis inhibitor, cycloheximide (39). However, the mechanisms involved in the anti-inflammatory roles of NO have been attributed to its ability to neutralize superoxide (25, 27). It has been proposed that eNOS-derived NO does not directly modulate endothelial cell conversion to a pro-adhesive phenotype but rather interferes with the generation of ROS by NADPH oxidase and prevents ROS-mediated inflammatory response (27, 35). In contrast to this proposal, the timing and pattern of leukocyte adhesion observed in our studies strongly support a direct association between basal NO and adhesive states of endothelial cells. Although detailed molecular mechanisms remain to be discovered, our present studies indicate that the signaling pathways involved in reduced basal NO-mediated leukocyte adhesion are different from those induced by cytokines.

Our results also demonstrated that CAV-induced leukocyte adhesion, without fMLP stimulation, does not increase microvessel permeability. The continuous VE-cadherin outlined underneath the AP-CAV-induced adherent leukocytes (Fig. 5) provide further evidence that the adhesion process did not cause local changes at endothelial adherent junctions, which is consistent with the permeability measurements. These findings also support that the process of leukocyte recruitments, such as adhesion and migration, may not necessarily cause vascular leakage. Our results that the increases in microvessel permeability only occurred upon the application of fMLP to vessels with AP-CAV-induced adherent leukocytes further support that the adhesion process alone can be dissociated with increases in microvessel permeability, and the additional stimulus-induced leukocyte oxidative burst is the direct cause of leukocyte adhesion-mediated increases in microvessel permeability (13, 14, 17, 40, 43, 47, 48). This could also explain the effective anti-inflammatory actions of systemically applied AP-CAV without reported injury of unaffected organs by several investigators (2–4, 12).

Summary.

Our experimental approach allows us to investigate microvessel functions under both basal and stimulated conditions, and in the absence and presence of blood components, thus providing us a unique opportunity to dissect the distinct roles of AP-CAV and NO in the regulation of endothelial function in vivo under different vascular conditions. Our present study on CAV-induced basal NO reduction, in combination with previous studies, demonstrated that basal NO is essential in maintaining the nonadhesive state of endothelium through regulation of endothelial surface adhesion molecules, whereas the stimulated excessive NO causes increases in microvessel permeability. The reduced basal NO-induced increases in adhesive activity of ICAM-1 to leukocytes were mediated by Src activation-dependent ICAM-1 phosphorylation.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants HL-56237 and HL-084338 (P. He), and American Heart Association Great Rivers Affiliate 12PRE11470010 predoctoral fellowship (S. Xu).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: S.X. and P.H. conception and design of research; S.X., X.Z., D.Y., and Y.X. performed experiments; S.X., X.Z., D.Y., and Y.X. analyzed data; S.X., X.Z., and P.H. interpreted results of experiments; S.X., X.Z., D.Y., and Y.X. prepared figures; S.X. and P.H. drafted manuscript; S.X. and P.H. edited and revised manuscript; S.X. and P.H. approved final version of manuscript.