Nitric Oxide Inhibits Microvascular ProteinLeakage Induced by Leukocyte

Adhesion-Independent andAdhesion-Dependent

Inflammatory Mediators

BRENT JOHNSTON,* JEFFREY P. GABOURY,* MAKOTO SUEMATSU,†

AND PAUL KUBES**Immunology Research Group, Department of Physiology and Biophysics,

University of Calgary, Calgary, Alberta, Canada T2N 4N1; and †Departmentof Biochemistry, Keio University School of Medicine, Tokyo, Japan 160-8582

ABSTRACT

Objective: The purpose of this study was to determine whether exogenous nitricoxide could block permeability alterations induced by neutrophil-independent(histamine) and neutrophil-dependent (CINC/gro) inflammatory mediators.Methods: Intravital microscopy was used in the rat mesentery to examine leu-kocyte adhesion, transvascular FITC-albumin leakage, and perivascular mastcell activation (ruthenium red uptake) in response to local superfusion withhistamine or the chemokine CINC/gro. The effects of the nitric oxide donorspermine-NO, or the cGMP analog 8-Br-cGMP were examined.Results: Histamine superfusion increased vascular protein leakage within min-utes, but did not increase firm adhesion above that seen in control preparations.The increase in albumin leakage could be prevented by co-administration ofspermine-NO, but was not affected by 8-Br-cGMP. CINC/gro elicited a linearincrease in vascular protein leakage and a profound increase in leukocyte adhe-sion. Treatment with spermine-NO or 8-Br-cGMP significantly attenuated in-creases in both adhesion and albumin leakage. The actions of spermine-NO and8-Br-cGMP were not due to mast cell stabilization as neither histamine norCINC/gro elicited mast cell activation.Conclusions: This study demonstrates that exogenous nitric oxide and 8-Br-cGMP could block adhesion dependent alterations in vascular permeability in-duced by CINC/gro, while adhesion-independent alterations in permeability in-duced by histamine could be blocked by exogenous NO but not 8-Br-cGMP. Thissuggests that different NO-dependent signalling pathways are important inmodulating these two types of vascular protein leakage.

KEY WORDS: intravital microscopy, venule, leukocyte adhesion, rolling, vascularpermeability, mast cell, nitric oxide, histamine, chemokine, cGMP

INTRODUCTION

Inhibiting nitric oxide (NO) synthesis with the com-petitive L-arginine analogue Nw-nitro-L-argininemethyl ester (L-NAME) produces rapid and persis-tent increases in microvascular protein and fluidleakage out of the vasculature in the cat intestineand rat mesentery (31,38). Associated with these in-creases in protein leakage are increases in neutrophil

Supported by grants from the Alberta Heritage Foundation forMedical Research (AHFMR) and the Medical Research Council(MRC) of Canada. P. Kubes is an AHFMR and MRC Scientist. B.Johnston is the recipient of an AHFMR studentship.For reprints of this article, contact Dr. Paul Kubes, ImmunologyResearch Group, Department of Physiology & Biophysics, Fac-ulty of Medicine, The University of Calgary, Calgary, Alberta,Canada T2N 4N1.Received 17 November 1997; accepted 19 November 1998.

Microcirculation (1999) 6, 153–162© 1999 Stockton Press All rights reserved 1073-9688/99 $12.00http://www.stockton-press.co.uk

adhesion to the venular endothelium (36,38,54).These two events appear to be partially related; inanimals in which adhesion was prevented withmonoclonal antibodies against CD18, the early in-crease in protein leakage (first 15 min) was stillnoted, but the transvascular protein flux was signifi-cantly reduced at later time points (31,38). Thesedata suggest that endogenous NO is a homeostaticregulator of microvascular permeability that maymanifest its effects both directly upon the endothe-lium and indirectly through modulation of adhesion-dependent events.

Based on these findings, one might predict that en-dogenous NO may reduce or dampen the increase inmicrovascular permeability associated with inflam-matory mediators. Surprisingly, Yuan et al. (60)demonstrated that histamine-induced increases inpermeability of isolated porcine coronary venulescould be blocked by an inhibitor of NO synthase. Inother reports, NO synthesis inhibition has also beenfound to prevent increases in vascular permeabilityinduced by histamine, bradykinin, substance P, andADP (22,45,51). In contrast, addition of exogenousNO via NO donors reduced platelet-activating factor(PAF)-induced albumin extravasation in the intes-tine (34), and three different NO donors reduced thevascular protein leakage associated with ischemia/reperfusion injury (40). In additional studies, in-haled NO was found to block oxidant-induced vas-cular permeability in the rabbit lung (28,52), andischemia/reperfusion injury in the cat mesentery(12).

There are a number of potential explanations forthese dichotomous observations. First, mediatorssuch as histamine, bradykinin, and serotonin inducea leukocyte-independent rise in microvascular per-meability (4,30,56), whereas PAF, CINC/gro, LTB4,fMLP and more complex models of inflammationsuch as ischemia/reperfusion injury have a neutro-phil-dependent component (35,37,55,56). It is pos-sible that NO may only block adhesion-dependentalterations in vascular permeability. Second, differ-ent tissues may respond differently to NO. Third,alterations in blood flow could impact upon vascularprotein leakage, a parameter that was not measuredin many of the NO inhibition studies. In some studieswhere NO inhibition was found to be protective, va-soconstrictors were equally effective in blocking vas-cular protein leakage alterations (22,51). This sug-gests that reduced blood flow associated with NOinhibition may account for the reductions in vascularprotein leakage. Finally, differences between in vitroand in vivo studies suggest that other cell types, such

as mast cells, may be involved in permeability alter-ations. For example, NO inhibition can cause mastcell degranulation (50), and the release of mediatorswhich increase vascular permeability.

Based on these variables, we designed a series ofexperiments to address the possibility that the effectsof NO on vascular protein leakage are dependentupon whether the eliciting stimulus also induces leu-kocyte adhesion and/or mast cell activation. Wechose two mediators: 1) histamine which evokes per-meability via endothelial activation (adhesion-independent) (4); and 2) CINC/gro, a rat C-X-Cchemokine, which induces microvascular permeabil-ity alterations entirely via neutrophil adhesion (55).The experiments were performed using intravitalmicroscopy so that blood kinetics, leukocyte adhe-sion, and microvascular protein leakage could bemeasured in single postcapillary venules. The samespecies and vascular bed were used in these experi-ments, and systemic alterations in hemodynamic pa-rameters were avoided by administering all reagentslocally. Mast cell activation was examined to excludethe possibility that vascular permeability alterationsinduced by histamine and/or CINC/gro were causedindirectly via mast cell degranulation. The effect of8-Br-cGMP on vascular protein leakage was also ex-amined to determine whether this secondary signal-ing mediator could mimic the effects of exogenous NO.

METHODS

Intravital Microscopy

Male Sprague-Dawley rats (200–300 g) were main-tained on a purified laboratory diet and fasted 18–24 h prior to each experiment. The animals wereinitially anesthetized with an i.p. injection of sodiumpentobarbital (65 mg/kg body weight), and the rightjugular vein was cannulated for the administrationof additional anesthetic as required. The right ca-rotid artery was cannulated to monitor systemic ar-terial blood pressure (model P23XL pressure trans-ducer; Viggo-Spectramed, Oxnard, CA; and model 7Grass physiologic recorder; Grass Instruments Co.,Quincy, MA). A midline abdominal incision wasmade and the rats were placed in a supine positionon an adjustable Plexiglas microscope stage. A seg-ment of the mid-jejunum was exteriorized over anoptically clear viewing pedestal for observation ofthe mesenteric microvasculature (15,16,27,32).

The mesenteric preparation was observed throughan intravital microscope (Optiphot-2, Nikon Inc.,Mississauga, ON, Canada) with a ×25 objective lens

NO inhibits vascular protein leakageB Johnston et al.

154

(Wetzlar L25/0.35; E. Leitz Inc., Munich, Germany)and a ×10 eyepiece. A video camera (model 5100HS; Panasonic, Osaka, Japan) mounted on the mi-croscope projected the image onto a color monitor(model PVM 2030; Sony, Tokyo, Japan), and theimages were recorded using a videocassette recorder(model AG-1790; Panasonic) for subsequent play-back analysis. The final magnification of the imageon the monitor was ×1800. Single unbranched mes-enteric venules (25–45 mm in diameter) were se-lected for study. Venular diameter was measured on-line using a video caliper (Microcirculation ResearchInstitute, Texas A & M University, College Station,TX). Centerline red blood cell velocity (Vrbc) wasalso measured on-line using an optical Doppler ve-locimeter (Microcirculation Research Institute,Texas A & M University). Venular wall shear rate(g) was calculated based on the Newtonian defini-tion: g 4 8(Vmean/Dv), where Dv is venular diam-eter and Vmean is Vrbc/1.6 (21).

The number of adherent leukocytes was determinedoff-line during video playback analysis. Leukocyteswere considered adherent to the venular endothe-lium if they remained stationary for a period of timeequal to or exceeding 30 s.

Vascular Permeability Measurements

The degree of vascular albumin leakage in rat mes-enteric venules was assessed as described previously(16,39). Fluorescein isothiocyanate (FITC)-labeledbovine albumin (25 mg/kg; Sigma Chemical Co., St.Louis, MO) was administered intravenously to ani-mals 15 min prior to the start of the experimentalprocedure. Fluorescence intensity (excitation wave-length, 420 to 490 nm; emission wavelength, 520nm) was detected using a silicon-intensified fluores-cent camera (model C-2400-08, Hamamatsu Pho-tonics, Hamamatsu, Japan) and images were re-corded for playback analysis using a videocassetterecorder. The fluorescent intensity of FITC-albuminwithin a defined area of the venule under study (10mm × 50 mm) and in the adjacent perivascular in-terstitium (20 mm from venule) was measured at 15min intervals following administration of FITC-albumin. This was accomplished using a video-capture board (Visionplus AT-OFG, Imaging Tech-nology Inc., Bedford, MA) and digital imaging soft-ware (Optimas, Bioscan Inc., Edmonds, WA). Theindex of vascular albumin leakage (permeability in-dex) was determined as a ratio: (interstitial inten-sity—background)/(venular intensity—background)(16,39).

Mast Cell-Activation Measurements

Mast cell activation was measured on-line by super-fusing preparations with buffer containing 0.001%ruthenium red (wt/vol; Sigma Chemical Co., St.Louis, MO), a dye taken up selectively by activatedmast cells (15,26,41). The intensity of dye uptakewas measured from video recordings throughout theexperiment using the video-capture board and digi-tal imaging software described above. Images wereconverted to digitized gray scales and phase-inverted. The relative light intensity of individualmast cells was measured using a 256-shade grayscale, and data are presented as degree of light in-tensity relative to background.

Experimental Protocol

The mesentery of untreated control animals was su-perfused with warmed bicarbonate-buffered salinecontaining 0.001% ruthenium red, and revealedlittle or no change in hemodynamic parameters, leu-kocyte kinetics, FITC-albumin leakage, or mast cellactivation throughout the experiment. Video record-ings and hemodynamic measurements were madeover 60 min at 15 min intervals. In some experi-ments, the inflammatory mediators CINC/gro (10nM) or histamine (100 mM) were added to the bufferafter a baseline recording (time 5 min) and super-fused onto the mesentery. These concentrations ofCINC/gro and histamine have previously beenshown to elicit optimal inflammatory responses invivo (32,55). In a second series of experiments,CINC/gro or histamine were superfused over themesentery with 100 mM spermine-NO (CaymanChemical Co., Ann Arbor, MI), an NO donor whichhas previously been shown to inhibit ischemia/reperfusion and H2O2-dependent inflammatoryevents (27,40). Spermine-NO has a 39-min half-lifeat 37 °C which allows for slow release of NO (43,57).Slow release of NO combined with very local admin-istration avoids complications due to alterations inhemodynamic parameters. As many of the actions ofNO are known to be due to the activation of solubleguanylate cyclase and the generation of cGMP(23,47), a final series of experiments was conductedto examine the role of cGMP in altering histamineand CINC/gro-induced vascular protein leakage.The cell permeable cGMP analog 8-Br-cGMP (100mM; Sigma) was superfused over the mesentery dur-ing superfusion with histamine or CINC/gro. 8-Br-cGMP has been shown previously to block L-NAME-induced increases in leukocyte transvascular albu-min leakage in vivo (38), and block albumin fluxacross monolayers of rat endothelium in vitro (19).

NO inhibits vascular protein leakageB Johnston et al.

155

Statistical Analysis

All values are reported as means ± SEM. Multiplecomparisons were performed by one-way analysis ofvariance (ANOVA) and individual differences be-tween groups were tested using an unpaired Stu-dent’s t test with Bonferroni corrections for multiplecomparisons where appropriate. Statistical signifi-cance was set at P < 0.05.

RESULTS

NO Blocks Leukocyte-Independent VascularProtein Leakage

Figure 1 summarizes microvascular protein leakageand leukocyte endothelial cell interactions in re-sponse to histamine superfusion. Superfusion withbuffered saline alone (control) caused no increase in

vascular protein leakage over a 60-min experiment.Addition of histamine to the superfusion buffercaused an increase in vascular protein leakage whichpeaked within 15 min and remained elevated over60 min (Fig. 1A). Others have demonstrated com-plete reversal of protein leakage within minutes ofapplying histamine (20,58), suggesting that extra-vascular FITC-albumin is cleared poorly from themesenteric interstitium. Superfusion with the NOdonor spermine-NO completely blocked histamine-induced microvascular permeability alterations (Fig.1A). Superfusion with spermine, the NO carrier, didnot affect microvascular protein leakage (data notshown). Surprisingly, the histamine-induced in-crease in vascular protein leakage was not affectedby the cell permeable cGMP analog 8-Br-cGMP (Fig.1A).

Figure 1B demonstrates that during the time thatmaximal alterations in protein leakage were ob-served (first 15 min), there was no increase in leu-kocyte adhesion. This is consistent with the obser-vation that histamine-induced increases in vascularprotein leakage are due to direct effects upon theendothelium and not dependent on alterations inleukocyte-adhesive interactions (4). A slight increasein adhesion at 30 and 60 min of histamine superfu-sion was noted. However, this increase was not sig-nificantly different from adhesion observed in con-trol preparations at these same time points. Therewas no increase in leukocyte adhesion in animalstreated with histamine plus spermine-NO or 8-Br-cGMP (Fig. 1B). There were no significant alter-ations in hemodynamic parameters including venu-lar shear rates when animals were treated locallywith histamine, histamine plus spermine-NO, or his-tamine plus 8-Br-cGMP (Table 1), excluding thepossibility that the blockade of FITC-albumin leak-age observed in the presence of spermine NO wasdue to reduced blood flow.

NO Blocks Leukocyte Dependent VascularProtein Leakage

Figure 2 summarizes microvascular protein leakageand leukocyte endothelial cell interactions in re-sponse to superfusion with the C-X-C chemokineCINC/gro. CINC/gro elicited significant increases invascular protein leakage (Fig. 2A) and leukocyte ad-hesion (Fig. 2B). Vascular protein leakage increasedlinearly over 30 min while adhesion increased within15 min and remained stable over the experiment at23–27 cells/100-mm venule length. In this model,transendothelial protein flux has been linked to leu-kocyte firm adhesion as CINC/gro-induced increases

Figure 1. Histamine-induced alterations in vascular pro-tein leakage (A), and leukocyte adhesion (B) in mesentericpostcapillary venules. Preparations were superfused con-tinuously with 100 mM histamine following an initial5-min control period (n 4 6), or control buffer through-out (n 4 6). Vascular permeability was measured as theratio of leakage of FITC-albumin from the vasculatureinto the extravascular space. Leukocyte adhesion wasmeasured from light recordings. Some animals weretreated with spermine-NO (100 mM superfusion; n 4 4)or 8-Br-cGMP (100 mM superfusion; n 4 4) in addition tohistamine. * P < 0.05 relative to control. † P < 0.05 rela-tive to histamine.

NO inhibits vascular protein leakageB Johnston et al.

156

in vascular dysfunction could be blocked by prevent-ing leukocyte adhesion with an anti-CD18 mAb(55). We examined whether NO could alter CINC/gro-induced vascular dysfunction. Superfusion withspermine-NO or 8-Br-cGMP significantly reduced

the microvascular protein leakage induced by CINC/gro (Fig. 2A). Spermine-NO and 8-Br-cGMP alsoattenuated CINC/gro-induced adhesion (Fig. 2B).The reduction in vascular permeability paralleledthe reduction in leukocyte adhesion. Treatments didnot alter hemodynamic parameters including shearrates in these groups of animals (Table 1), excludingthe possibility that protection was due to reducedblood flow.

Histamine and CINC/gro-Induced Alterations inVascular Protein Leakage Are MastCell Independent

As inflammatory mediators may induce permeabilityalterations indirectly through mast cell degranula-tion (8,15,59), mast cell activation was evaluated bymeasuring the uptake of ruthenium red dye from thesuperfusion buffer. As shown in Figure 3, the mastcell-activating agent CMP 48/80 caused a profoundincrease in ruthenium red uptake within 15 min.Dye staining persisted in these preparations over thenext 45 min. In contrast, neither histamine norCINC/gro elicited increases in mast cell dye uptake(Fig. 4). To ensure that the mast cells in these prepa-rations were responsive to activating stimulus, CMP48/80 was added to the superfusion buffer at the endof each experiment. CMP 48/80 caused significantruthenium red uptake by perivascular mast cells ineach of the preparations in which histamine andCINC/gro had no effect (Fig. 4).

DISCUSSION

In this study we examined the effects of exogenousNO on vascular protein leakage induced by an ad-hesion-independent mediator, histamine (4), and anadhesion-dependent mediator, CINC/gro (55). His-tamine-induced albumin leakage occurred rapidlyand prior to any changes in firm adhesion. However,vascular protein leakage induced by CINC/gro wasmore gradual and associated with a large increase in

Figure 2. CINC/gro-induced alterations in vascular pro-tein leakage (A), and leukocyte adhesion (B) in mesentericpostcapillary venules. Preparations were superfused con-tinuously with 10 nM CINC/gro following an initial 5-mincontrol period (n 4 5), or control buffer throughout (n 46). Vascular permeability was measured as the ratio ofleakage of FITC-albumin from the vasculature into theextravascular space. Leukocyte adhesion was measuredfrom light recordings. Some animals were treated withspermine-NO (100 mM superfusion; n 4 5) or 8-Br-cGMP(100mM superfusion; n 4 4) in addition to CINC/gro. * P< 0.05 relative to control. † P < 0.05 relative to CINC/gro.

Table 1. Venular shear rates (s−1) in control and experimental groupsa

Treatment 0 min 15 min 30 min 60 min

Control (n 4 6) 397 ± 46 401 ± 53 415 ± 71 396 ± 89Histamine (n 4 6) 395 ± 75 405 ± 63 391 ± 69 413 ± 82Histamine + Spermine-NO (n 4 4) 314 ± 42 383 ± 89 433 ± 113 428 ± 88Histamine + 8-bromo-cGMP (n 4 4) 319 ± 23 309 ± 24 314 ± 28 305 ± 27CINC/gro (n 4 5) 345 ± 10 362 ± 10 349 ± 14 NDCINC/gro + Spermine-NO (n 4 5) 299 ± 44 304 ± 53 303 ± 43 NDCINC/gro + 8-bromo-cGMP (n 4 4) 309 ± 32 319 ± 37 315 ± 34 ND

aDifferences within and between groups are not significant.

NO inhibits vascular protein leakageB Johnston et al.

157

leukocyte adhesion to the venular endothelium.While the exogenous NO donor spermine-NO wasable to inhibit histamine-induced increases in vas-cular protein leakage and CINC/gro-induced in-creases in leukocyte adhesion and vascular proteinleakage, 8-Br-cGMP was only effective in blockingthe effects of the adhesion dependent mediatorCINC/gro. The cGMP analog did not inhibit hista-mine-induced increases in vascular leakage. Thesedata suggest that exogenous NO can prevent bothadhesion-independent and adhesion-dependent in-creases in vascular protein leakage, but may actthrough different mechanisms with respect to signal-ing via cGMP.

In vivo, the reduction in histamine-induced vascularprotein leakage with spermine-NO is likely due to

effects on the venular endothelium as permeabilityincreases do not require increases in leukocyte roll-ing or adhesion (4). An interesting aspect of thisstudy is that the histamine-induced rise in microvas-cular permeability could be blocked with excess NO,whereas previous work from our laboratory hasshown that serotonin-induced permeability alter-ations, which are also neutrophil-independent, couldnot be altered by the same NO-donor regimen (16).This suggests that exogenous NO is not simply tar-geting the cytoskeleton in a non-specific manner toreduce vascular protein leakage, yet must be selec-tively altering histamine but not serotonin-dependent activation of the endothelium. It is pos-sible that NO blocks one or more of the downstreamsignaling components required for endothelial re-traction. For example, exogenous NO has beenshown to inhibit protein kinase C (48), a signalingenzyme implicated in permeability alterations(25,42,49). However, other factors must also be im-portant in permeability alterations as serotonin-induced vascular alterations can be blocked withprotein kinase C inhibitors (6), but not spermine-NO(16). This suggests that NO may target a specificsignaling pathway that mediates histamine-inducedvascular permeability, while serotonin may activatean alternate pathway that is insensitive to NO.

Nitric oxide also blocked increases in vascular pro-tein leakage elicited by CINC/gro, an event whichhas been shown to be entirely dependent on leuko-cyte adhesion. Suzuki et al. (55) were able to blockCINC/gro-induced monastral blue leakage from themesenteric vasculature by preventing leukocyte ad-

Figure 4. Mast cell activation after exposure to histamineor CINC/gro. Preparations were superfused with 0.001%ruthenium red alone (control; n 4 4), or in the presenceof 100 mM histamine (n 4 4) or 10 nM CINC/gro (n 4 4).Mast cell ruthenium red uptake was measured at 60 minfollowing exteriorization of the mesentery. The mast cell-activating ionophore CMP 48/80 (1 mg/mL) was super-fused onto the mesentery for 30 min at the end of eachexperiment as a positive control for ruthenium red uptake.* P < 0.05 relative to the control group.

Figure 3. Mesenteric mast cell activation 0 min (A), 15min (B), 30 min (C), and 60 min (D) after superfusing themesentery with 1 mg/mL CMP 48/80. The mesentery wassuperfused with 0.001% ruthenium red throughout theexperiment. Activated mast cells took up ruthenium reddye within 15 min of exposure to CMP 48/80. The scalebar 4 100 mm.

NO inhibits vascular protein leakageB Johnston et al.

158

hesion with an anti-CD18 antibody. The moregradual increase in microvascular permeability ob-served with CINC/gro is also consistent with adhe-sion-dependent increases in microvascular perme-ability seen with other mediators (56). Our data donot reveal whether adhesion or the physical migra-tion of leukocytes between endothelial cells was suf-ficient to cause increased permeability, or whetheradherent leukocytes release inflammatory mediatorswhich then increase permeability. As NO has beenshown to block both neutrophil adhesion(14,17,33,44) and the release of neutrophil-derivedmediators (10,46), both mechanisms could contrib-ute to inhibiting the vascular dysfunction. However,the fact that spermine-NO clearly inhibited adhesionin this study, suggests that the anti-adhesive prop-erties of NO may be the major reason for reducedmicrovascular protein leakage.

It is unclear how NO blocks leukocyte adhesion. Ex-posure to NO has been shown to reduce the expres-sion of ICAM-1, VCAM-1, and E-selectin on acti-vated endothelium through transcriptional regula-tion (1,11,53). However, this is unlikely to accountfor the reduced adhesion in this study due to theshort-time course of the experiments. Direct effectson neutrophils have also been reported. Clancy et al.(9) demonstrated that NO inhibits actin polymeriza-tion and clustering in human neutrophils stimulatedwith fMLP, preventing their subsequent adhesion tofibronectin coated plates. NO and cGMP appear todown-regulate the formation of pseudopods on neu-trophils (13), providing a mechanism which mayprevent adhesion and spreading of neutrophils onendothelium. A recent study also demonstrated thatNO donors given at certain concentrations can pre-vent firm adhesion for prolonged periods by block-ing functional upregulation of CD11/CD18 on neu-trophils (5). Although we have no direct evidence tosupport or refute this hypothesis, the fact that bothan anti-CD18 antibody and spermine-NO can effec-tively block neutrophil adhesion in response toCINC/gro may support this contention.

An unusual finding in this study was that spermine-NO blocked vascular protein leakage induced byboth histamine and CINC/gro, while 8-Br-cGMPblocked CINC/gro-induced alterations in vascularprotein leakage, but not histamine-induced increasesin albumin leakage. One possibility is that exogenousNO may modulate microvascular permeability alter-ations through signaling mechanisms that do not in-volve the guanylate cyclase/cGMP pathway. It isclear that NO does have effects that are independentof cGMP. For example, NO activates Ca2+-

dependent K+ channels in vascular smooth musclecells independent of cGMP (7). In another study, NOdonors but not cGMP analogs significantly decreasedIL-1-induced ICAM-1 surface expression on mesan-gial cells (24), suggesting that the effect of NO onICAM-1 expression is independent of cGMP forma-tion or requires activation of additional NO-dependent, cGMP-independent pathways. Othershave shown that exogenous NO inhibits cytokine-induced expression of VCAM-1, ICAM-1, and E-selectin, as well as endothelial secretion of IL-6 andIL-8 (11,53). This may be related to the ability ofNO to upregulate IkB and prevent NFkB-dependenttranscription of these adhesion molecules indepen-dent of cGMP activation (11,53).

We have previously demonstrated that NO blocksnumerous features of mast cell-induced inflamma-tion, likely through the inhibition of mediator release(16). For example, mast cell activation induced byCMP 48/80 caused increases in vascular proteinleakage and leukocyte adhesion which could be in-hibited by spermine-NO. Although it has beensuggested that some chemokines and other inflam-matory mediators may activate mast cells (2,3,8,18,29,59), CINC/gro and histamine did not activateconnective tissue mast cells in the mesentery. There-fore, our data would suggest that the suppressiveeffects of spermine-NO on histamine and CINC/gro-induced vascular protein leakage are unlikely due tothe ability of NO to stabilize mast cells (50,54).

In an earlier study, Paul et al. (51) found that NOinhibition with L-NAME reduced histamine-inducedvascular protein leakage in guinea pig skin. How-ever, the vasoconstrictor phenylephrine was equallyeffective, suggesting that L-NAME-induced reduc-tions in blood flow may mediate the protective re-sponse. In the present study, local NO administra-tion was able to block histamine and CINC/gro-induced permeability without altering vascularhemodynamics. This suggests that NO may functionin vivo to suppress protein leakage without alteringblood flow. In vitro however, Yuan et.al. (60) re-ported that L-NAME was able to block histamine-induced permeability alterations in isolated porcinecoronary vessels in which the flow rate was held con-stant. These discrepant results were initially ex-plained by suggesting that neutrophil and/or mastcell-dependent events were being studied in the invivo scenario, while neutrophil and/or mast cell-independent mechanisms were being examined invitro . The data in this study clearly refute this view;in the rat mesentery, spermine-NO was able to re-duce vascular protein leakage associated with a me-

NO inhibits vascular protein leakageB Johnston et al.

159

diator that increases microvascular protein leakageexclusively via neutrophils and a mediator that in-creases vascular dysfunction entirely independent ofneutrophils. Although it is possible that other bloodcomponents such as platelets or red blood cells maybe contributing factors, we found no evidence thatNO was working by affecting mast cell reactivity. Itis possible that there may be organ or tissue-specificdifferences in the effects of NO. There is also thepotential that NO may exert different effects in dif-ferent species. Many of these areas require carefulcharacterization in order to fully understand the ef-fects of NO.

ACKNOWLEDGMENTS

The authors wish to thank Dr. K. Watanabe at theCytosignal Research Institute, Tokyo, Japan, whoprovided CINC/gro for this study.

REFERENCES

1. Adams MR, Jessup W, Hailstones D, Celermajer DS.(1997). L-arginine reduces human monocyte adhe-sion to vascular endothelium and endothelial expres-sion of cell adhesion molecules. Circulation 95:662–668.

2. Alam R, Forsythe P, Lett-Brown MA, Grant JA.(1992). Macrophage inflammatory protein-1 a is anactivator of basophils and mast cells. J Exp Med 176:781–786.

3. Alam R, Grant JA. (1995). The chemokines and thehistamine releasing factors: Modulation of function ofbasophils, mast cells, and eosinophils. In: Human Ba-sophils and Mast Cells: Biological Aspects. (G Ma-rone, Ed.) S. Karger AG. Basel. 148–157.

4. Asako H, Kurose I, Wolf R, DeFrees S, Zheng Z,Phillips ML, Paulson JC, Granger DN. (1994). Role ofH1 receptors and P-selectin in histamine-inducedleukocyte rolling and adhesion in postcapillary ve-nules. J Clin Invest 93:1508–1515.

5. Banick PD, Chen Q, Xu YA, Thom SR. (1997). Nitricoxide inhibits neutrophil b2-integrin function by in-hibiting membrane-associated cyclic GMP synthesis.J Cell Physiol 172:12–24.

6. Barman SA, Pauly JR, Isales CM. (1997). Canine pul-monary vasoreactivity to serotonin: Role of proteinkinase C and tyrosine kinase. Am J Physiol 272:H740–H747.

7. Bolotina VM, Najibi S, Palacino JJ, Pagano PJ, CohenRA. (1994). Nitric oxide directly activates calcium-dependent potassium channels in vascular smoothmuscle. Nature 368:850–853.

8. Cirino G, Cicala C, Bucci MR, Sorrentino L, Mara-ganore JM, Stone SR. (1996). Thrombin functions asan inflammatory mediator through activation of itsreceptor. J Exp Med 183:821–827.

9. Clancy R, Leszczynska J, Amin A, Levartovsky D,

Abramson SB. (1995). Nitric oxide stimulates ADPribosylation of actin in association with the inhibitionof actin polymerization in human neutrophils. J Leu-koc Biol 58:196–202.

10. Clancy RM, Leszczynska-Piziak J, Abramson SB.(1992). Nitric oxide, and endothelial cell relaxationfactor, inhibits neutrophil superoxide production viaa direct action on the NADPH oxidase. J Clin Invest90:1116–1121.

11. De Caterina R, Libby P, Peng HB, Thannickal VJ,Rajavashisth TB, Gimbrone Jr. MA, Shin WS, LiaoJK. (1995). Nitric oxide decreases cytokine-inducedendothelial activation. Nitric oxide selectively reducesendothelial expression of adhesion molecules and pro-inflammatory cytokines. J Clin Invest 96:60–68.

12. Fox-Robichaud A, Payne D, Hasan SU, Ostrovsky L,Fairhead T, Reinhardt P. (1998). Inhaled NO as aviable antiadhesive therapy for ischemia/reperfusioninjury of distal microvascular beds. J Clin Invest 101:2497–2505.

13. Fukuda S, Yasu T, Schmid-Schonbein GW. (1998).The regulation of leukocyte response to shear stressby nitric oxide via cyclic GMP. FASEB J 12:A2(Abstr.)

14. Gaboury J, Woodman RC, Granger DN, Reinhardt P,Kubes P. (1993). Nitric oxide prevents leukocyte ad-herence: Role of superoxide. Am J Physiol 265:H862–H867.

15. Gaboury JP, Johnston B, Niu X-F, Kubes P. (1995).Mechanisms underlying acute mast cell-induced leu-kocyte rolling and adhesion in vivo. J Immunol 154:804–813.

16. Gaboury JP, Niu X-F, Kubes P. (1996). Nitric oxideinhibits numerous features of mast cell-induced in-flammation. Circulation 93:318–326.

17. Gauthier TW, Davenpeck KL, Lefer AM. (1994). Ni-tric oxide attenuates leukocyte-endothelial interac-tion via P-selectin in splanchnic ischemia-reper-fusion. Am J Physiol 267:G562–G568.

18. Harris JG, Flower RJ, Watanabe K, Tsurufuji S,Wolitzky BA, Perretti M. (1996). Relative contribu-tion of the selectins in the neutrophil recruitmentcaused by the chemokine cytokine-induced neutro-phil chemoattractant (CINC). Biochem Biophys ResCommun 221:692–696.

19. Hempel A, Noll T, Muhs A, Piper HM. (1996). Func-tional antagonism between cAMP and cGMP on per-meability of coronary endothelial monolayers. Am JPhysiol 270:H1264–H1271.

20. Horan KL, Adamski SW, Ayele W, Langone JJ, GregaGJ. (1986). Evidence that prolonged histamine suf-fusions produce transient increases in vascular per-meability subsequent to the formation of venularmacromolecular leakage sites. Proof of the Majno-Palade hypothesis. Am J Pathol 123:570–576.

21. House SD, Lipowsky HH. (1987). Leukocyte-endothelium adhesion: Microhemodynamics in mes-entery of the cat. Microvasc Res 34:363–379.

NO inhibits vascular protein leakageB Johnston et al.

160

22. Hughes SR, Williams TJ, Brain SD. (1990). Evidencethat endogenous nitric oxide modulates oedema for-mation induced by substance P. Eur J Pharmacol191:481–484.

23. Ignarro LJ. (1989). Biological actions and propertiesof endothelium-derived nitric oxide formed and re-leased from artery and vein. Circ Res 65:1–21.

24. Ikeda M, Ikeda U, Takahashi M, Shimada K, MinotaS, Kano S. (1996). Nitric oxide inhibits intracellularadhesion molecule-1 expression in rat mesangial cells.J Am Soc Nephrol 7:2213–2218.

25. Johnson A, Phillips P, Hocking D, Tsan M-F, Ferro T.(1989). Protein kinase inhibitor prevents pulmonaryedema in response to H2O2. Am J Physiol 256:H1012–H1022.

26. Johnston B, Burns AR, Kubes P. (1998). A role formast cells in the development of adjuvant-inducedvasculitis and arthritis. Am J Pathol 152:555–563.

27. Johnston B, Kanwar S, Kubes P. (1996). Hydrogenperoxide induces leukocyte rolling: modulation by en-dogenous antioxidant mechanisms including NO. AmJ Physiol 271:H614–H621.

28. Kavanagh BP, Mouchawar A, Goldsmith J, Pearl RG.(1994). Effects of inhaled NO and inhibition of en-dogenous NO synthesis in oxidant-induced acute lunginjury. J Appl Physiol 76:1324–1329.

29. Kemeny L, Ruzicka T, Dobozy A, Michel G. (1994).Role of the interleukin-8 receptor in skin. Int ArchAllergy Immunol 104:317–322.

30. Kubes P, Gaboury J. (1996). Rapid mast cell activa-tion causes polymorphonuclear leukocyte and inde-pendent permeability alterations in postcapillary ve-nules. Am J Physiol 271:H2438–H2446.

31. Kubes P, Granger DN. (1992). Nitric oxide modulatesmicrovascular permeability. Am J Physiol 262:H611–H615.

32. Kubes P, Kanwar S. (1994). Histamine induces leu-kocyte rolling in post-capillary venules: A P-selectin-mediated event. J Immunol 152:3570–3577.

33. Kubes P, Kurose I, Granger DN. (1994). NO donorsprevent integrin-induced leukocyte adhesion, but notP-selectin-dependent rolling in postischemic venules.Am J Physiol 267:H931–H937.

34. Kubes P, Reinhardt PH, Payne D, Woodman RC.(1995). Excess nitric oxide does not cause cellular,vascular or mucosal dysfunction in the cat small in-testine. Am J Physiol 269:G34–G41.

35. Kubes P, Suzuki M, Granger DN. (1990). Modulationof PAF-induced leukocyte adherence and increasedmicrovascular permeability. Am J Physiol 259:G859–G864.

36. Kubes P, Suzuki M, Granger DN. (1991). Nitric ox-ide: An endogenous modulator of leukocyte adhesion.Proc Natl Acad Sci USA 88:4651–4655.

37. Kurose I, Anderson DC, Miyasaka M, Tamatani T,Paulson JC, Todd RF, Rusche JR, Granger DN.(1994). Molecular determinants of reperfusion-

induced leukocyte adhesion and vascular proteinleakage. Circ Res 74:336–343.

38. Kurose I, Kubes P, Wolf R, Anderson DC, Paulson J,Miyasaka M, and Granger DN. (1993). Inhibition ofnitric oxide production: Mechanisms of vascular al-bumin leakage. Circ Res 73:164–171.

39. Kurose I, Wolf R, Grisham MB, Aw TY, Specian RD,Granger DN. (1995). Microvascular responses to in-hibition of nitric oxide production: role of active oxi-dants. Circ Res 76:30–39.

40. Kurose I, Wolf R, Grisham MB, Granger DN. (1994).Modulation of ischemia/reperfusion-induced micro-vascular dysfunction by nitric oxide. Circ Res 74:376–382.

41. Lagunoff D. (1972). Vital staining of mast cells withruthenium red. J Histochem Cytochem 20:938–944.

42. Lynch JM, Ferro TJ, Blumenstock FA, BrockenauerAM, Malik AB. (1990). Increased endothelial albuminpermeability mediated by protein kinase C activation.J Clin Invest 85:1991–1998.

43. Maragos CM, Morley D, Wink DA, Dunams TM,Saavedra JE, Hoffman A, Bove AA, Issac L, HrabieJA, Keefer LK. (1991). Complexes of NO with nu-cleophiles as agents for the controlled biological re-lease of nitric oxide. Vasorelaxant effects. J MedChem 34:3242–3247.

44. May GR, Crook P, Moore PK, Page CP. (1991). Therole of nitric oxide as an endogenous regulator ofplatelet and neutrophil activation within the pulmo-nary circulation of the rabbit. Br J Pharmacol 102:759–763.

45. Mayhan WG. (1992). Role of nitric oxide in modu-lating permeability of hamster cheek pouch in re-sponse to adenosine 5’-diphosphate and bradykinin.Inflammation 16:295–305.

46. Moilanen E, Vuorinen P, Kankaanranta H, Metsa-Ketela T, Vapaatalo H. (1993). Inhibition by nitricoxide-donors of human polymorphonuclear leucocytefunctions. Br J Pharmacol 109:852–858.

47. Moro MA, Russel RJ, Cellek S, Lizasoain I, Su Y,Darley-Usmar VM, Radomski MW, Moncada S.(1996). cGMP mediates the vascular and platelet ac-tions of nitric oxide: Confirmation using an inhibitorof the soluble guanylyl cyclase. Proc Natl Acad SciUSA 93:1480–1485.

48. Murohara T, Scalia R, Lefer AM. (1996). Lysophos-phatidlcholine promotes P-selectin expression inplatelets and endothelial cells. Possible involvementof protein kinase C activation and its inhibition bynitric oxide donors. Circ Res 78:780–789.

49. Murray MA, Heistad DD, Mayhan WG. (1991). Role ofprotein kinase C in bradykinin-induced increases in mi-crovascular permeability. Circ Res 68:1340–1348.

50. Niu X-F, Ibbotson G, Kubes P. (1996). A balancebetween nitric oxide and oxidants regulates mast cell-dependent neutrophil-endothelial cell interactions.Circ Res 79:992–999.

NO inhibits vascular protein leakageB Johnston et al.

161

51. Paul W, Douglas GJ, Lawrence L, Khawaja AM,Perez AC, Schachter M, Page CP. (1994). Cutaneouspermeability responses to bradykinin and histaminein the guinea-pig: possible differences in their mecha-nism of action. Br J Pharmacol 111:159–164.

52. Poss WB, Timmons OD, Farrukh IS, Hoidal JR, Mi-chael JR. (1995). Inhaled nitric oxide prevents theincrease in pulmonary vascular permeability in-creases caused by hydrogen peroxide. J Appl Physiol79:886–891.

53. Spiecker M, Darius H, Kaboth K, Hubner F, Liao JK.(1998). Differential regulation of endothelial cell ad-hesion molecule expression by nitric oxide donors andantioxidants. J Leukoc Biol 63:732–739.

54. Suematsu M, Tamatani T, Delano FA, Miyasaka M,Forrest M, Suzuki H, Schmid-Schonbein GW.(1994). Microvascular oxidative stress preceding leu-kocyte activation elicited by in vivo nitric oxide sup-pression. Am J Physiol 266:H2410–H2415.

55. Suzuki H, Suematsu M, Miura S, Liu YY, WatanabeK, Miyasaka M, Tsurufuji S, Tsuchiya M. (1994). RatCINC/gro: A novel mediator for locomotive and se-

cretagogue activation of neutrophils in vivo. J LeukocBiol 55:652–657.

56. Wedmore CV and Williams TJ. (1981). Control ofvascular permeability by polymorphonuclear leuko-cytes in inflammation. Nature 289:646–650.

57. Wink DA, Hanbauer I, Krishna MC, DeGraff W,Gamson J, Mitchell JB. (1993). Nitric oxide protectsagainst cellular damage and cytotoxicity from reac-tive oxygen species. Proc Natl Acad Sci USA 90:9813–9817.

58. Wu NZ, Baldwin AL. (1992). Transient venular per-meability increase and endothelial gap formation in-duced by histamine. Am J Physiol 262:H1238–H1247.

59. Yano H, Wershil BK, Arizono N, Galli SJ. (1989). Sub-stance P-induced augmentation of cutaneous vascularpermeability and granulocyte infiltration in mice is mastcell dependent. J Clin Invest 84:1276–1286.

60. Yuan Y, Granger HJ, Zawieja DC, DeFily DV, ChilianWM. (1993). Histamine increases venular permeabil-ity via a phospholipase C-NO synthase-guanylate cy-clase cascade. Am J Physiol 264:H1734–H1739.

NO inhibits vascular protein leakageB Johnston et al.

162