Role of Nitric Oxide Synthase Inhibition in Leukocyte-Endothelium Interaction in the Rat Pial Microvasculature

Abstract

We investigated the role of nitric oxide (NO) in leukocyte-endothelium interaction, blood-brain barrier (BBB) function and oxygen free-radical production in the rat pial microcirculation. In a closed cranial window preparation (dura removed) over the parietal cortex of pentobarbital-anesthetized Wistar rats, NO synthase (NOS) was inhibited by systemic and/or topical application of N^ω-nitro-l-arginine (l-NNA) under physiological conditions and during leukotriene B₄ (LTB₄) activation. Circulating leukocytes were labeled by intravenous injection of rhodamine 6G. We used a confocal laser scanning microscope (CLSM) and studied leukocyte rolling and sticking in pial veins and arteries before and after NOS inhibition. At the end of the experiments, sodium-fluorescein was injected intravenously to test BBB integrity. Brain cortex oxygen free-radical production was investigated in the cranial window preparation using lucigenin-enhanced chemiluminescence (CL). l-NNA application did not lead to significant changes in leukocyte-endothelium interaction, BBB function, and oxygen free-radical production under physiological conditions [leukocyte-endothelium interaction: control (n = 5), l-NNA systemically (n = 5), l-NNA topically (n = 5): at baseline rollers/100 μm: 0.76 ± 0.55, 0.64 ± 0.94, 0.44 ± 0.55 and stickers/100 μm: 0.90 ± 0.28, 0.76 ± 0.24, 0.84 ± 0.42; at 60 min rollers/100 μm: 1.49 ± 0.66, 1.21 ± 0.99, 0.67 ± 0.66 and stickers/100 μm: 1.04 ± 0.20, 1.19 ± 0.23, 1.21 ± 0.54; oxygen free-radical production (n = 4): CL count before l-NNA application 35 ± 17 cps, after 1 h of topical superfusion of l-NNA 38 ± 14 cps; p < 0.05]. In contrast to the results achieved under physiological conditions, a significant further increase of rolling leukocytes and BBB permeability occurred due to NOS inhibition under LTB₄-activated conditions [76 ± 47% significant (p ≤ 0.01, n = 7) further increase of rollers/100 μm due to 60 min l-NNA application following the activation period of 120 min LTB₄ superfusion]. Our results support a modulatory role for NO in leukocyte-endothelium interaction and BBB permeability in the pial microcirculation when this interaction is increased.

Nitric oxide (NO) or a related compound has been shown to play an important role in the blood flow regulation of various tissues, including the regulation of resting and stimulated cerebral blood flow (CBF) during global (i.e., with CO₂) and functional activation of the cerebral cortex (for review, see Iadecola et al., 1994a).

In addition to its vascular effects via modulation of smooth muscle function, NO may play a role in the microcirculation by affecting leukocyte-endothelium interaction. Impairment of endothelial NO production in mesenteric venules in vivo leads to a drastic increase in leukocyte adherence to the vascular endothelium (Kubes et al., 1991) possibly due to a superoxide anion radical mechanism (Gaboury et al., 1993). In addition, increases of neutrophil adhesion during NOS inhibition have also been observed in the heart (Ma et al., 1993), liver (Harbrecht et al., 1992), and skeletal muscle (Persson et al., 1990). So far, the role of NO in leukocyte-endothelium interaction has only been studied in detail in the circulation of noncerebral tissues. The brain microvascular endothelium, which forms the blood-brain barrier (BBB) differs functionally, antigenetically, and morphologically from extracerebral vascular endothelium. In addition, in the brain, NO is constitutively not only produced by the endothelium, but also by the surrounding parenchyma (neurons, astrocytes, etc.). Data from extracerebral tissues concerning the role of NO in leukocyte–endothelium interaction may, therefore, not be applicable to the brain.

The aim of this study was to investigate whether NO also plays a role in leukocyte-endothelium interaction during physiological as well as leukotriene B₄ (LTB₄)-activated conditions in the cerebral microcirculation. Leukocyte-endothelium interaction in pial vessels of the cerebral cortex of anesthetized rats were studied before and during systemic and/or topical application of the irreversible nitric oxide synthase (NOS) inhibitor N^ω-nitro-l-arginine (l-NNA) under physiological conditions and during topical application of the leukotactic compound LTB₄ using a cranial window preparation and confocal laser scanning microscopy (CLSM). Because the oxygen free-radical superoxide anion has been implicated to participate in the role that NO might play in leukocyte-endothelium interaction, additional experiments were performed to investigate oxygen free-radical production of the cerebral cortex under physiological conditions before and during topical NOS inhibition in vivo with lucigenin-enhanced chemiluminescence (CL) in the same experimental model.

MATERIALS AND METHODS

General preparation

Male Wistar rats (280–320 g), were anesthetized with pentobarbital (Trapanal, 100 mg/kg body wt, Byk Gulden, Konstanz, Germany), tracheotomized, and mechanically ventilated. The left femoral artery and vein were cannulated for physiological monitoring, intravenous saline infusion (infusion rate 1 ml/h), and application of drugs. Body temperature was monitored and maintained at 38 ± 0.5°C using a heating pad. Systemic arterial blood pressure (SAP) and endexpiratory Pco₂ (Heyer EGM I, Bad Ems, Germany) were monitored continuously and blood gases measured periodically (AVL LIST, Graz, Austria). Animals were placed in a stereotactic frame and a closed cranial window was implanted over the right parietal cortex (Dirnagl et al., 1991). The dura mater was removed and the exposed brain continuously superfused (superfusion rate, 1 ml/h) with artificial cerebrospinal fluid (aCSF) (composition in mM):K⁺, 3; Na⁺, 150; Ca²⁺, 1.25; Mg²⁺, 0.6; Cl⁻, 132; HCO₃⁻, 24.5; glucose 3.7; urea 6 (Levasseur et al., 1975), which was equilibrated with a gas mixture containing (in %) O₂, 6.6; CO₂, 5.9; and N₂, 87.5 leading to the following gas tensions and pH: Po₂, 119 ± 8 mm Hg; Pco₂, 34.6 ± 3.9 mm Hg; and pH 7.36 ± 0.03.

CLSM setup

For monitoring leukocytes within the pial circulation, we used a Bio-Rad MRC 600 CLSM system (Bio-Rad Microscience, Watford, U.K.) (for details, see Dirnagl et al., 1992). In brief, the CLSM was attached to a Nikon Optiphot microscope with a Zeiss water immersion objective corrected for a cover slip with a numerical aperture of 0.75, a working distance of 1.6 mm, and ×40 magnification. We used a Krypton laser (Ion Laser Technology, Salt Lake City, UT, U.S.A.) with principle lines at 488, 568, and 647 nm as light source. Confocal microscopy was performed through the cranial window. To visualize leukocytes, rhodamine 6G (Sigma Chemicals) was injected intravenously (200 μg/ml in 0.9% saline, 1 ml as bolus injection, followed by continuous infusion at an infusion rate of 1 ml/h). After intravenous injection of rhodamine 6G, circulating leukocytes (polymorphonuclear leukocytes, lymphocytes and monocytes) and platelets were labeled, whereas endothelial (ECs) and red blood cells remained unstained (Dirnagl et al., 1994; Villringer et al., 1991). Rhodamine 6G staining, according to this protocol, does not affect leukocyte function (Lorenzl et al., 1993; Dirnagl et al., 1994; Villringer et al., 1991). Appropriate filter sets for rhodamine 6G imaging were used. An average of seven different sites of exposed pial microvasculature per animal was recorded (area, 317 × 211 μm) at a frame rate of 16/s for 60 s and digital/analog converted for recording on a video recorder (Panasonic NV-V 8000 EG, Osaka, Japan). Images were analyzed off-line from the video recording by redigitalization. The CLSM was equipped with a computer-controlled x-y-z stage, permitting movement of the object with an accuracy of 0.1 μm in every dimension. It was, therefore, possible to return exactly to each of the seven recording sites during the course of the experiment.

In vivo CL recording setup

Reactive oxygen species were detected by lucigenin-enhanced CL in vivo. For a detailed description of this technique, see Dirnagl et al. (1995). In brief, the animal in the stereotactic frame was housed in a dark box and covered with aluminum foil to exclude photons from sources other than CL and to prevent spontaneous ultraweak CL. The unshielded cranial window was positioned under a reflector focused to the photon sensitive area of a cooled (–20°C) photomultiplier (Hamamatsu R943–02) with a dark count of 3 cps at −20°C. The animal within the dark box was housed in a second, larger dark box, equipped with metal shielding (Faraday cage) to exclude photons and other electromagnetic waves unrelated to brain CL from outside. To record oxygen free-radicals optically, brain topical superfusion of the CL enhancer lucigenin (10⁻³ M, 1 ml/h) was used. Lucigenin reacts with oxygen free-radicals producing photons in the blue/green wavelength range. Lucigenin-induced CL is particularly sensitive to the intracellular production of the superoxide anion radical (Faulkner and Fridovich, 1993), regardless of the cell type within the brain. Counts of the photomultiplier were amplified and counted by a Hewlett Packard Universal counter (HP 5316 B9) connected to a PC for data recording and storage.

NOS activity determination

Immediately after experiments, tissue specimens were removed from the cortex in situ to depth of 1 mm under the cranial window and frozen in liquid N₂ for later analysis. The NOS assay was based on assessing the conversion of l-[³H]-arginine to l-[³H]-citrulline (Bredt and Snyder, 1989). Tissue samples of 10 mg per 150 μl buffer were homogenized in a Potter-Elvehjem homogenizer with a Teflon pestle, four-five strokes, in ice-cold 0.05 M TRIS buffer (pH 7.4) containing 1.15% (wt/v) KCl, 1 mM EDTA, 5 mM glucose, 0.1 mM d,l-dithiothreitol (DTT), 200 U/ml superoxide dismutase, 2 mg/L pepstatin A, 10 mg/L trypsin inhibitor, 44 mg/1 phenylmethylsulfonyl fluoride. The homogenate was centrifuged for 10 min in a refrigerated bench microcentrifuge at 2,000 g (4°C). The protein concentration was measured at 224 and 236.5 nm using serum albumin as a standard and a detection limit of 5 μg/ml. Determination of NOS activity was performed in a reaction buffer: 50 mM HEPES buffer (pH 7.4), containing 1 mM EDTA, 1.25 mM CaCL₂, 1 mM DTT, 1 μM flavin adenine dinucleotide, 1 μM calmodulin, 15 μM 6R-tetrahydrobiopterin, 1 μM l-[³H]-arginine (1 μCi final concentration). The supernatant was added to obtain a final protein concentration of 15–30 μg/ml in a reaction volume of 150 μl. The reaction was initiated by adding the cofactor NADPH to a final concentration of 1 mM. Samples were incubated for 30 min at 37°C. The reaction was terminated by adding 1 ml of ice-cold 100 mM HEPES buffer (pH 5.5) containing 10 mM EGTA (stop buffer). The total volume (1 ml) was applied to a 0.5 ml DOWEX AG50 WX-8 column (NA⁺-form) that had been equilibrated with the stop buffer minus EGTA. l-[2,3-³H]-citrulline was eluted twice with 0.5 ml stop buffer minus EGTA and the radioactivity determined by liquid scintillation counting.

NOS activity in sham-operated animals (n = 2) was assayed and used as control. Percent inhibition was calculated by comparing enzyme activity in tissue exposed to the NOS inhibitor l-NNA [applied either systemically (n = 2) or topically (n = 2) and systemically and topically in combination (n = 2)] versus controls.

Experimental paradigm (Fig. 1)

In the first two studies (I and II), leukocyte-endothelium interaction was investigated before and 30 and 60 min after l-NNA application by counting leukocytes moving slowly along the wall of pial vessels at a velocity less than that of the other blood components (rollers) or adhering to the vascular wall for >10 s (stickers) per each 100 μm of vessel-length during the 60-s observation period. In study I, l-NNA was applied systemically by i.v. injection of 10 mg/kg body wt. For topical application in study II, l-NNA (10⁻³ M) was dissolved in aCSF and continuously superfused over the cranial window.

Fig. 1.

Experimental paradigm.

In studies I and II, a laser-Doppler flow (LDF) probe monitored rCBF through the closed cranial window in the area of confocal imaging. In addition to the ex vivo NOS activity determination assay, a decrease of resting rCBF and appearance of vasomotion marked the effectivity of the NOS blockade in the brain cortex underlying the cranial window in all animals (Dirnagl et al. 1993a,b).

In studies III and IV, the leukotactic compound LTB₄ (7.5 × 10⁻⁹ M) was superfused over the cranial window for 3 h. Experiments of 3 h of topical superfusion of LTB₄ without l-NNA application served as controls in the LTB₄-activation study (study III, LTB₄). In study IV, after a 2-h LTB₄ activation period, in the third hour of the experiments, l-NNA was applied systemically (10 mg/kg i.v.) and topically (10⁻³ M) during further LTB₄ superfusion (study IV, LTB₄ + l-NNA). The effect of the combination of systemical and topical l-NNA application without previous activation due to LTB₄ superfusion was tested in another two animals.

To study BBB integrity, at the end of each experiment in studies I-IV, sodium-fluorescein (MW 376, Sigma Chemicals) was injected intravenously (2 mg/100 g body wt) (Leakage of this low molecular weight dye is a sensitive marker for BBB disruption). Extravasation of sodium-fluorescein was investigated in vivo using CLSM (Dirnagl et al., 1994). At 10 min after the bolus injection, leakage of sodium-fluorescein into the extravascular space was graded qualitatively as no extravascular fluorescence (grade 0), slight appearance of fluorescence outside the pial vessels along the vessel wall (grade 1), slight level of fluorescence not only along pial vessels but also throughout the brain surface (grade 2), and high fluorescence on brain surface (grade 3).

Because the role of NO in leukocyte-endothelium interaction has been implicated as being closely related to the oxygen free-radical superoxide, in study V, lucigenin-enhanced CL was recorded at baseline and during topical application of l-NNA. Superfusion of lucigenin had already been performed 60 min before start of the baseline measurement of oxygen free-radicals to obtain a stable photon count; 60 min is the time needed for lucigenin to diffuse into the extracellular space and into the cells of the cortex. After 60 min of baseline recording, l-NNA was superfused topically for 1 h Lucigenin-enhanced CL was recorded continuously throughout the whole experiment.

Data analysis

Numbers of rollers or stickers per 100 μm vessel length before and after 30 and 60 min of l-NNA application, in control and study groups I and II; and before and after 2 h of LTB₄ superfusion, and after 3 h LTB₄- with and without l-NNA-superfusion in the third hour in study groups IV and III were compared using repeated-measures analysis of variance (ANOVA). Changes in numbers of rollers or stickers per 100 μm vessel length after 3 h in study groups III (LTB₄ alone) and IV (LTB₄ + l-NNA) (in %; values at 2 h of LTB₄ activation = 100% in each group) were compared using Student's unpaired t-test. Lucigenin-enhanced CL before and after 1 h of NOS blockade was compared using Student's paired t-test; p-values < 0.05 were considered statistically significant. All data are presented as means ± standard deviation (SD).

RESULTS

Physiological variables

During experiments, all physiological parameters were within normal ranges, except for SAP, which increased significantly after systemic l-NNA application in studies I and IV. Physiological variables are summarized in Table 1.

Table 1.

Physiological variables: systemic arterial pressures (SAP) and arterial gas tensions

Study group (n)		SAP (mm Hg)	Paco2 (mm Hg)	PaO2 (mm Hg)	pH
Control (5)	Baseline	114 ± 15	36 ± 2	105 ± 16	7.39 ± 0.01
	60 min	104 ± 10	32 ± 6	116 ± 13	7.41 ± 0.03
I (5)	Baseline	104 ± 9	37 ± 2	114 ± 15	7.38 ± 0.02
l-NNA systemically	60 min	141 ± 15	34 ± 3	129 ± 19	7.38 ± 0.03
II (5)	Baseline	108 ± 3	34 ± 3	107 ± 19	7.38 ± 0.02
l-NNA topically	60 min	100 ± 10	29 ± 3	124 ± 17	7.40 ± 0.03
III (6)	Baseline	117 ± 11	34 ± 4	127 ± 27	7.40 ± 0.02
LTB4	120 min	111 ± 10	30 ± 4	165 ± 35	7.40 ± 0.01
	180 min	93 ± 11	28 ± 3	132 ± 37	7.38 ± 0.04
IV (7)	Baseline	105 ± 12	35 ± 4	101 ± 11	7.40 ± 0.04
LTB4 + l-NNA	120 min	104 ± 7	32 ± 3	127 ± 19	7.37 ± 0.05
	180 min	133 ± 11	29 ± 5	137 ± 25	7.34 ± 0.05
V (4)	Baseline	100 ± 13	32 ± 4	120 ± 18	7.39 ± 0.03
CL	60 min	97 ± 9	34 ± 5	135 ± 20	7.41 ± 0.04

Values are means ± SD.

NOS activity determination

NOS activity in sham-operated animals was assayed and set at 100% as control. Whereas systemic application of the NOS inhibitor l-NNA reduced the NOS activity only to 54%, topically applied l-NNA and systemic and topical combined application led to reductions of NOS activity by 89 and 90%, respectively. These results are in accordance with other data from the literature; nitro-l-arginine—the active metabolite of other NOS inhibitors, e.g., nitro-l-arginine methyl ester (l-NAME)—application according to this protocol has been shown in numerous studies to reduce brain NOS activity by ˜50% when applied systemically, and by >90% during topical application (Traystman et al., 1995; Irikura et al., 1994; Iadecola et al., 1994b).

The NOS activity assay as described here does not distinguish between constitutive and inducible NOS. In the ischemic brain, expression of inducible NOS mRNA starts as early as 12 h after onset of ischemia (Iadecola et al., 1995), suggesting that the LTB₄-activation period in studies III and IV is too short to cause expression of inducible NOS during the course of the experiment. Therefore, the main NOS activity measured in our experiments is expected to originate from the constitutive NOS in neurons, astrocytes, and endothelial cells in the brain.

CLSM

The mean length of the pial vascular network investigated per site per animal was 277 ± 52 μm in study I (systemical l-NNA application), 291 ± 46 μm in study II (topical l-NNA application), 294 ± 33 μm in controls, 150 ± 11 (Jim in study III (LTB₄), and 192 ± 76 μm in study IV (LTB₄ + l-NNA). Diameters of the vessels investigated ranged from 20 to 100 μm in the venous system and from 15 to 90 μm in the arterial system.

There was only a very slight and nonsignificant increase in the numbers of rolling or sticking leukocytes during the 60-min observation times in studies I and II (Table 2) and during the combination of systemic and topical l-NNA application (n = 2) (data not shown) in the pial venous system. This slight, nonsignificant increase was also observed in the control group during the observation time. Therefore, no significant difference in leukocyte behavior was detected between animals of the control group and of studies I and II groups with either systemic or topical, or systemic and topical application of the NOS inhibitor l-NNA. Table 2 summarizes these data.

Table 2.

Leukocyte-endothelium interactions in pial veins at seven cortical sites per animal

Study group		Rollers/100 μm	Stickers/100μm
Control	Baseline (n = 5)	0.76 ± 0.55	0.90 ± 0.28
	30 min (n = 5)	1.13 ± 0.77	0.90 ± 0.20
	60 min (n = 5)	1.49 ± 0.66	1.04 ± 0.20
I	Baseline (n = 5)	0.64 ± 0.94	0.76 ± 0.24
l-NNA systemically	30 min (n = 4)	0.94 ± 0.48	0.90 ± 0.24
	60 min (n = 5)	1.21 ± 0.99	1.19 ± 0.23
II	Baseline (n = 5)	0.44 ± 0.55	0.84 ± 0.42
l-NNA topically	30 min (n = 3)	0.6 ± 0.51	1.34 ± 0.72
	60 min (n = 5)	0.67 ± 0.66	1.21 ± 0.54
III	Baseline (n = 6)	0.42 ± 0.25	0.73 ± 0.40
LTB4	120 min LTB4 (n = 6)	2.68 ± 1.98 a	1.01 ± 0.73
	180 min LTB4 (n = 6)	2.37 ± 1.76 a	1.02 ± 0.75
IV	Baseline (n = 7)	0.88 ± 0.43	0.9 ± 0.19
LTB4 + l-NNA	120 min LTB4 (n = 7)	2.18 ± 1.2 a	1.22 ± 0.46
	120 min LTB4 + 60 min
	LTB4 + l-NNA (n = 7)	3.7 ± 1.85b,c	1.39 ± 0.32 b

Values are means ± SD.

a

Significantly different compared to baseline within the study (p < 0.05).

b

Significantly different compared to baseline within the study (p ≤ 0.01).

c

Significantly differnt compared to 2 h within the study (p ≤0.01).

In studies III and IV, leukocyte rolling was significantly increased during continuous superfusion of a low concentration of the leukoattractant compound LTB₄. In contrast to the physiological state in studies I and II, NOS inhibition due to systemic and topical application of the NOS inhibitor following 2 h of LTB₄ activation (study IV) led to a significant increase in the numbers of rolling leukocytes in the pial venous system (Table 2, Figs. 2 and 3) compared to LTB₄ administration alone (3 h in study III). Due to high variations in the absolute numbers of increased rolling leukocytes during LTB₄ superfusion, there was only a slight, but not significant, difference in the absolute values seen after the third hour of the observation period between studies III and IV. Nevertheless, a significant difference (p ≤ 0.01) exists when comparing the % increase of rolling leukocytes in the third hour (second hour = 100%) between studies III and IV. This discrepancy in the significance between absolute and percentage values can be explained by high variations in the degree of activation of leukocyte-endothelium interaction due to LTB₄ application from animal to animal. Regardless of the absolute degree of activation of leukocyte-endothelium interaction in each individual animal, the level reached after the second hour of the observation period did not increase further during the third hour of superfusion of LTB₄ alone throughout all experiments in study III.

Fig. 2.

Effect of l-NNA on numbers of rolling leukocytes per 100 μm (in 60 s) of pial veins. In the physiological state, there is no significant difference between baseline (bl) conditions and the situation during NOS inhibition (control group and studies I and II, n = 5 in each group). After activation of the leukocyte-endothelium interaction with LTB₄ (7.5 × 10^9– M) for 120 min (study IV, n = 7), however, a significant increase in numbers of rolling leukocytes occurred during l-NNA application. *, significantly different from baseline conditions, p ≤ 0.01; **, significantly different from 120 min LTB₄ activation conditions, p ≤ 0.01.

Fig. 3.

Effect of NOS inhibition on numbers of leukocytes per 100 μm of pial veins rolling along the vessel wall (60 s) under LTB₄-activated conditions. Whereas after 120 min of LTB₄ superfusion (7.5 × 10^9– M) a plateau was reached and another 60 min of LTB₄ superfusion alone did not further increase the numbers of rolling leukocytes (study III, n = 6), LTB₄, + l-NNA (applied topically and systemically) significantly enhanced leukocyte-endothelium interaction (study IV, n = 7). *, significantly different from 120 min of LTB₄ superfusion, p ≤ 0.01.

In previous experiments in our laboratory (Dirnagl et al., 1994), it was shown that, under physiological conditions, there is no change in leukocyte-endothelium interaction during a 240-min observation period (a prolonged control group). This information is provided in addition to the 60-min control group for comparison with studies III and IV.

A typical example of the imaging of leukocytes within a pial vein with CLSM before and after 2 h of LTB₄ application is shown in Fig. 4.

Fig. 4.

Imaging of rhodamine 6G-labeled leukocytes within a pial vein before and after 2 h of LTB₄ superfusion using CLSM. A: xy-image showing a pial venule during baseline conditions. Note that there is no rolling or sticking leukocyte in this area. One leukocyte, seen at the upper rim of the vessel, is moving at blood stream velocity. Scale bar is 20 um. B: Same optical section as in A after 2 h of topical LTB₄ application. Leukocytes in this section, which have an elongated or almost round shape, are rolling along the vessel wall at a velocity less than that of the blood stream (for discrimination of rolling or sticking leukocytes, this section was continuously imaged for 1 min). Scale bar is 20 μm.

LTB₄ application alone (study III) did not lead to a significant change in the numbers of sticking leukocytes within the pial circulation of the cerebral cortex. LTB₄ superfusion followed by l-NNA application (study IV) led to a small but significant increase of sticking leukocytes compared to baseline recordings (see Table 2).

In pial arteries, no significant rolling or sticking of leukocytes was observed at any time point in any experimental group.

NOS inhibition alone did not affect BBB integrity during systemic and/or topical l-NNA application (no statistical significant difference to grade 0). In addition, no BBB dysfunction was observed in the control group.

LTB₄ application alone led to only slight leakage of sodium-fluorescein into the extravascular space in all animals of study III (grade 1). This slightly enhanced BBB permeability was significantly increased due to NOS inhibition following LTB₄ superfusion (study IV) (grade 2 in all animals).

Lucigenin-enhanced CL

Topical application of the NOS inhibitor l-NNA in study V did not change the photon count in the CL experiments. The photon count before l-NNA application was 35 ± 17 cps, and after 1 h of topical superfusion of l-NNA, it was 38 ± 14 cps. Inhalation of 100% O₂ increased the photon count by 20 ± 8 cps. After animals were killed (i.v. injection of KCl), the photon count decreased to 8 ± 7 cps. The dark count of the system was 3 ± 1 cps. A typical tracing is shown in Fig. 5.

Fig. 5.

Effect of topical application of l-NNA on free radical production, measured with lucigenin-enhanced CL (study V, n = 4). The figure shows a typical tracing of one experiment. NOS inhibition did not alter the photon counts during 60 min. Inhalation of 100% of O₂ significantly enhanced free radical production; killing the animal by i.v. injection of KCl led to an immediate decrease in the CL signal close to dark count.

DISCUSSION

In the present study, we found that neither systemic nor topical application of the NOS inhibitor l-NNA led to significant changes in leukocyte-endothelium interaction in the pial circulation of the brain cortex of barbiturate-anesthetized rats under physiological conditions, and that BBB integrity was not affected by NOS inhibition. Topical application of low concentrations of the leukoattractant compound LTB₄ induced slight, but significant, rolling of leukocytes along the endothelium. Following this mild activation of leukocyte-endothelium interaction, NOS inhibition further significantly activated leukocyte-endothelium interaction and increased BBB permeability. In the following discussion, we will compare our results to studies performed mainly in noncerebral tissues and try to explain the different results found in different microvascular beds.

Effect of NOS inhibition on leukocyte-endothelium interaction and vascular permeability

In several noncerebral tissues (e.g., heart, skeletal muscle, liver), an important role for NO in the regulation of leukocyte-endothelium interaction has been demonstrated (Persson et al., 1990; Ma et al., 1993; Harbrecht et al., 1992). Most of the studies investigating the increase in leukocyte-endothelium interaction during NOS inhibition were performed in intestinal and mesenteric tissue. In the gastrointestinal tract, systemic as well as topical NOS inhibition increased leukocyte-endothelium interaction in postcapillary venules, leading to adhesion and emigration of leukocytes (Kubes et al., 1993; Kubes et al., 1991).

The role of NO in the regulation of vascular permeability is discussed controversely. In some studies, inhibition of NO production induced albumin and horseradish peroxidase leakage in venules of mesenteric and intestinal tissue (Kanwar et al., 1994; Kurose et al., 1993; Kubes and Granger, 1992) as well as cerebral arterioles (Prado et al., 1992). In contrast, other investigators showed that NOS inhibition did not affect basal levels of vascular permeability but reduced augmented permeability in various tissues, including brain (Haberl et al., 1994; Hughes et al., 1990; Mayhan, 1993).

Mechanisms of NO effects on leukocyte-endothelium interaction

The mechanisms by which NO modulates microvascular endothelium function are not sufficiently resolved. Because superoxide anion (O₂⁻) can be inactivated by NO, it has been suggested that NO suppression may enhance O₂⁻ levels in the tissue, thus leading to functional changes in microvascular endothelium, such as leukocyte-endothelium interaction and vascular permeability (Gaboury et al., 1993). In addition, mast cell degranulation, which releases proinflammatory agents that cause leukocyte adhesion to endothelium, can be induced by superoxide. These findings suggest that basal levels of NO in vivo may continuously inactivate the proadhesive molecule superoxide.

Besides this scavenging effect of NO on superoxide, other mechanisms of enhanced leukocyte-endothelium interaction are discussed. Recently, it was shown that NO suppression-induced enhancement of endothelial adhesiveness may be mediated by a direct intracellular increase in oxidative stress in arteriolar and venular endothelial cells, and in mast cells in the mesenteric microcirculation, that is followed by venular leukocyte adhesion (Suematsu et al., 1994).

Another mechanism of NO-associated antiinflammatory properties was shown recently (Peng et al., 1995). Those authors found that NO inhibits the activation of the transcription factor NF-κB (which itself, when activated, leads to the expression of proinflammatory mediators) through induction and stabilization of the NF-κB inhibitor IκBα, even under basal conditions in a cell culture system.

NO and leukocyte-endothelium interaction in the brain: comparison to other tissues

In our cranial window preparation studies on the brain cortex, very low baseline levels of rolling and sticking leukocytes were detected and no increase in leukocyte-endothelium interaction occurred due to NOS inhibition. In addition, we found no increase in superoxide production of the cerebral cortex using lucigenin-enhanced CL during topical l-NNA application. It is unlikely that this negative finding was due to a sensitivity problem of CL. In every experiment, oxygen inhalation led to a significant increase in the photon count. This increase was caused by a slight increase in superoxide production due to hyperoxia. Much larger increases in the photon count have been shown during reoxygenation after cerebral hypoxia or ischemia (Dimagl et al., 1995).

Whereas, in the mesentery in the intra-abdominal situs under physiological conditions, neither rolling nor sticking of leukocytes was present, leukocyte-endothelium interaction increased rapidly after exteriorization in experimental studies and remained high over a long period (Fiebig et al., 1991). Therefore, Fiebig et al. suggested that inflammation is present in experimental mesentery preparations, and this results in relatively high levels of rolling or sticking leukocytes at the beginning of the experiments. In addition, Ma et al. (1993) have shown, that NOS inhibition in myocardial ischemia promotes neutrophil adherence only after reperfusion, in which free radical production is enhanced.

To mimic these experimental conditions in other tissues, in which leukocyte-endothelium interaction was increase before NOS inhibition (i.e., by the experimental preparation), we produced a slight activation of leukocyte-endothelium interaction by topical LTB₄ application. This induces a weak respiratory burst in neutrophils (Gaboury et al., 1993). In addition, leukocytes have been shown to produce superoxide anion radicals due to LTB₄ application in vitro (Maghni et al., 1991; Murohara et al., 1995). In a previous video microscopy study, it was shown that cerebral superfusion of LTB₄ induces an increase in leukocyte-endothelium interaction in the same range as in our study, without an increase in BBB permeability (Schürer et al., 1994).

Under these LTB₄-activated conditions, we found a slight but significant increase of leukocyte-endothelium interaction and BBB permeability after NOS inhibition, findings that are in accordance with those of other studies in noncerebral tissues.

Besides these methodological issues, other tissue-specific mechanisms may also be involved in the differences in the role of NO in leukocyte-endothelium interaction and vascular permeability in the brain compared to noncerebral tissues: (a) smaller lymphocyte binding capacities of cerebral endothelial cells compared to endothelial cells of other tissues (Male et al., 1990); (b) tight junctions of high electrical resistance between cerebral endothelial cells, presenting an effective barrier even to ions. Peripheral endothelial cells are either fenestrated or have tight junctions of low resistance between the cells (Rowland et al., 1991); (c) low mast cell content of the cerebral cortex; mast cells in the brain are found particularly in the leptomeninges, thalamus, and hypothalamus (Theoharides, 1990). In the brain, stationary interstitial immunocytes are mainly represented by microglial cells. During activation of interstitial immunocytes, proinflammatory agents such as superoxide anion are released. In the normal brain, microglial cells are not activated, and lack phagocytic and secretory activities (Giulian, 1992).

In conclusion, our data support a modulatory role for NO in leukocyte-endothelium interaction and BBB integrity in the pial microcirculation of the brain when this interaction is increased. This may be important for pathophysiological states of the brain, e.g., cerebral ischemia, during which induction of leukocyte-endothelium interaction occurs (Garcia et al., 1994) and NO levels in the brain tissue increase (Malinski et al., 1993). However, our data imply that basal NO production is not involved in modulating leukocyte-endothelium interaction. This latter finding is in contrast to results of a number of studies of the role of NO in leukocyte-endothelium interaction in noncerebral tissues. It remains open whether this discrepancy is due to methodological factors, particularly the degree of activation of leukocytes at the beginning of NOS inhibition, or to differences in the endothelium and/or surrounding tissues, resulting in different levels of superoxide in the resting state.