(Stroke. 1995;26:277-281.)
© 1995 American Heart Association, Inc.
Articles |
From the Departments of Anesthesia (J.E.B.), Internal Medicine (D.D.H., F.M.F.), and Pharmacology (D.D.H., F.M.F.), Cardiovascular Center and Center on Aging, University of Iowa College of Medicine, Iowa City.
Correspondence to J.E. Brian, Jr, MD, Department of Anesthesia, Room 6530 JCP, University of Iowa College of Medicine, Iowa City, IA 52242.
| Abstract |
|---|
|
|
|---|
Methods Cranial windows were implanted in anesthetized rabbits. Windows were flushed with artificial cerebrospinal fluid, artificial cerebrospinal fluid with lipopolysaccharide, or artificial cerebrospinal fluid with lipopolysaccharide and NG-monomethyl-L-arginine (an inhibitor of NO synthase) for 4 hours. Other rabbits received either dexamethasone or indomethacin intravenously 1 hour before lipopolysaccharide treatment of cranial windows.
Results Application of lipopolysaccharide in cranial windows produced marked, progressive vasodilatation, with diameter increased by 58±7% (mean±SEM) after 4 hours. The cerebral vasodilator response was inhibited by NG-monomethyl-L-arginine, dexamethasone, or indomethacin. Excess L-arginine reversed the inhibitory effect of NG-monomethyl-L-arginine.
Conclusions Inhibition of lipopolysaccharide-induced dilatation of cerebral arterioles by NG-monomethyl-L-arginine and dexamethasone suggests that a portion of the vasodilatation was mediated by inducible NO synthase. Indomethacin also inhibited lipopolysaccharide-induced vasodilatation. These findings suggest an important role for both nitric oxide and cyclooxygenase products in lipopolysaccharide-induced cerebral arteriolar dilatation in vivo.
Key Words: lipopolysaccharides vasodilation nitric oxide
| Introduction |
|---|
|
|
|---|
Bacterial infections of the central nervous system dilate cerebral arterioles and increase cerebral blood flow.9 10 11 12 Prior investigations in animal models suggest that prostanoids and/or oxygen-derived radicals are responsible for a portion of the vasodilatation.9 11 12 Recent evidence indicates that NO may also contribute to increased cerebral blood flow in experimental meningitis.13 In vitro, stimulation with LPS or cytokines leads to increased production of NO by astrocytes, microglia, endothelium, and smooth muscle.3 14 15 16 Therefore, it seems likely that induction of NO synthase occurs during infections of the central nervous system and that in- creased NO production contributes to cerebral vasodilatation. Dexamethasone can suppress production of inducible NO synthase in noncerebral systems.6 17
In this study, we tested the hypothesis that topical application of LPS would cause dilatation of cerebral arterioles, which could be blocked with an inhibitor of NO synthase and by dexamethasone. We also tested the hypothesis that a portion of LPS-induced dilatation is mediated by cyclooxygenase products, which can be blocked with indomethacin.
| Materials and Methods |
|---|
|
|
|---|
A cranial window was prepared as described previously.18 The scalp, muscle, and periostium overlying the parietal area of the skull were reflected, and bleeding was controlled with ferric chloride solution. A craniotomy (approximately 6 mm in diameter) was made in the parietal bone using an air-cooled drill; bone bleeding was controlled with bone wax. The dura was excised with microscissors, and the arachnoid overlying an arteriole was incised. Two blunt 18-gauge needles were affixed to a dam of bone wax surrounding the craniotomy, and a circular glass coverslip (12-mm) was fused to the wax. The window was reinforced with dental acrylic. An outlet tube was affixed to one needle and set to control intracranial pressure to 10 cm H2O pressure. A stopcock was attached to the other needle, and the window was filled with artificial cerebrospinal fluid (CSF) warmed to 37°C and equilibrated with 90% N2/5% O2/5% CO2 (pH, 7.44±0.01; PO2, 35±1 mm Hg; PCO2, 36±1 mm Hg). Cerebral arterioles were observed using a microscope equipped with a video camera, and images were recorded on videotape. Arteriolar diameter was measured with an image-shearing device. The preparation was allowed to equilibrate for 1 hour, during which time the window was flushed with 2 mL of artificial CSF every 15 minutes. Flushing the window with artificial CSF did not alter the diameter of cerebral arterioles.
After the equilibration period, arteriolar diameter was measured under control conditions and in response to acetylcholine (10-6 and 10-5 mol/L). Responses to acetylcholine were examined to test responsiveness of the preparation. The window was then flushed with artificial CSF several times, and the preparation was allowed to recover for 30 minutes. After a second measurement of baseline vessel diameter, rabbits were randomly allocated to receive (1) artificial CSF alone (n=5); (2) artificial CSF containing LPS (Sigma Chemical Co; Escherichia coli 055:B5, 100 µg/mL; n=7); or (3) artificial CSF containing LPS (100 µg/mL) and an inhibitor of NO synthase, NG-monomethyl-L-arginine (L-NMMA; Calbiochem; 0.3 mmol/L; n=6). In a separate group of rabbits, the cranial window was treated with LPS (100 µg/mL), L-NMMA (0.3 mmol/L), and L-arginine (Sigma; 1 mmol/L; n=3) simultaneously to test reversibility of the effects of L-NMMA. To evaluate the effect of prolonged exposure to L-NMMA on arteriolar diameter, in another group of rabbits cranial windows were treated with L-NMMA alone (0.3 mmol/L; n=2). Two other groups received LPS in artificial CSF after pretreatment with either dexamethasone (Elkins-Sinn; 3 mg/kg IV; n=5) or indomethacin (Sigma; 5 mg/kg IV; n=5) 1 hour before administration of acetylcholine.
The absolute diameter of cerebral arterioles was recorded at baseline and compared among groups. After baseline measurements were made, the cranial windows were flushed every 30 minutes for 4 hours with 2 mL of vehicle (artificial CSF) or LPS with or without inhibitors. Changes in arteriolar diameter are expressed as percent change in diameter compared with baseline. Arterial blood gases were measured at baseline and at hourly intervals.
Statistics
Data are expressed as mean±SEM. Data were compared among groups
by ANOVA and Duncan's post hoc test. Data within groups were analyzed
by repeated-measures ANOVA and post hoc comparison by means contrast.
P<.05 was considered significant.
| Results |
|---|
|
|
|---|
|
When compared with artificial CSF, treatment with LPS significantly
increased diameter of the arterioles at all measurements. After 4
hours, diameter of the arterioles had increased by 58±7%
(P<.05; Figure
, A). Treatment of the cranial window with
LPS and L-NMMA resulted in marked reduction of LPS-induced dilation
after 0.5, 2, 3, and 4 hours (P<.05; Figure
, B). L-NMMA
inhibited LPS-induced cerebral arteriolar dilatation by approximately
70% at 4 hours. In rabbits treated with LPS and L-NMMA, increases in
diameter of the cerebral arterioles did not achieve statistical
significance compared with artificial CSF until 4 hours
(P<.05; Figure
, B).
In separate animals, treatment of the cranial window with LPS (100
µg/mL), L-NMMA (0.3 mmol/L), and L-arginine (1 mmol/L;
n=3) reversed L-NMMA inhibition of LPS-induced arteriolar dilatation
(Figure
, B). In addition, treatment of cranial windows with L-NMMA
alone (0.3 mmol/L; n=2) did not alter arteriolar diameter over 4 hours
(data not shown).
Dexamethasone (Figure
, C) and indomethacin (Figure
, D) both inhibited
LPS-induced dilatation at hours 3 and 4 (P<.05). LPS
increased diameter of vessels in rabbits that were treated with
dexamethasone and indomethacin at hours 2, 3, and 4 (P<.05;
Figure
, C and D). Dexamethasone and indomethacin each inhibited
LPS-induced dilatation by approximately 50% to 60%.
Mean arterial pressure was not different among groups or over time and averaged 73±1 mm Hg. Arterial PO2 and PCO2 did not vary over time or among groups and averaged 126±2 mm Hg and 35±1 mm Hg, respectively. Baseline arterial pH values were not different among groups and averaged 7.41±0.01. Arterial pH did not change over time in the artificial CSF group (P>.05). All other groups exhibited a modest decline in pH over time. When compared with baseline values, arterial pH was significantly reduced at hours 3 and 4 in the LPS group and at hours 2, 3, and 4 in the LPS groups treated with L-NMMA, indomethacin, and dexamethasone. In each group exhibiting change, the lowest pH values occurred at 4 hours and were as follows: L-NMMA, 7.29±0.03; LPS, 7.33±0.02; dexamethasone, 7.29±0.04; and indomethacin, 7.34±0.02. In the group treated with LPS, L-NMMA, and L-arginine, arterial pH declined to 7.31±0.02 at 4 hours. This change was not significant (P>.05). However, this likely reflected the small number of animals (n=3) in this group.
| Discussion |
|---|
|
|
|---|
Lipopolysaccharide and Nitric Oxide
Increased production of NO by inducible NO synthase appears to
play a key role in systemic vasodilation after administration of
LPS.7 8 We hypothesized that a similar expression of
inducible NO synthase occurs in brain and influences cerebral
arterioles in vivo. Studies using cultured cells indicate that a
variety of brain cells including astrocytes, microglia, and endothelium
can express inducible NO synthase after stimulation with LPS and/or
cytokines.3 14 15 A study of large cerebral arteries in
vitro suggested that LPS may cause expression of inducible NO synthase
in cerebral vascular muscle.16 Recent evidence indicates
that an inhibitor of NO synthase can prevent the increase in cerebral
blood flow occurring in experimental meningitis.13
In noncerebral systems, dexamethasone suppresses LPS-mediated expression of inducible NO synthase in cultured cells and in vivo.6 17 Dexamethasone did not alter vasodilatation in response to acetylcholine in the present study, which suggests that dexamethasone does not affect constitutive NO synthase. However, dexamethasone did prevent LPS-induced vasodilatation. Because both L-NMMA and dexamethasone prevented LPS-induced dilatation, the evidence suggests that the inhibitory effect of dexamethasone is due in part to suppression of inducible NO synthase. However, dexamethasone may also alter cyclooxygenase metabolism, as outlined below.
We did not test higher doses of dexamethasone or administer repeated doses, so it is possible that some dilatation that occurred in response to LPS after dexamethasone was due to residual activity of inducible NO synthase. However, systemic administration of 3 mg/kg dexamethasone, the dose used in the present study, achieves maximal suppression of LPS-mediated production of inducible NO synthase in lung, liver, and aorta in vivo.17
These data do not rule out a contribution of constitutive NO synthase to LPS-induced dilatation. Administration of intravenous LPS results in a rapid decrease in arterial pressure, which is attenuated by inhibitors of NO synthase.19 The hypotensive response to LPS is too rapid to be mediated by inducible NO synthase and has been attributed to activation of constitutive NO synthase.19 Factors that mediate acute activation of constitutive NO synthase by LPS are not understood but apparently involve an increase in activity, not amount, of constitutive NO synthase. In the present study, dilatation of cerebral arterioles occurring at 30 minutes could be mediated by activation of constitutive NO synthase.
The potency of blockers for isoforms of NO synthase differs. We used L-NMMA to inhibit NO synthase in the present study because it may be more potent than NG-nitro-L-arginine in inhibition of inducible NO synthase.20 The inhibitory effect of L-NMMA on LPS-induced dilatation was reversed with L-arginine, which indicates that inhibition by L-NMMA was specific for NO synthase. We did not administer higher doses of L-NMMA, and we have not ascertained the degree of NO synthase blockade obtained with the dose of L-NMMA used. Thus, dilatation occurring during LPS and L-NMMA exposure could be due to residual NO synthase activity.
Lipopolysaccharide and Prostanoids
Because indomethacin attenuated LPS-mediated vasodilatation,
cyclooxygenase products appear to account for a portion of dilatation
in the present study. In isolated cerebral microvessels,
application of LPS causes release of prostaglandin
E2,21 and CSF prostaglandin E2
levels are increased in a model of meningitis.22
Indomethacin also inhibits increases in cerebral blood flow during
acute meningitis.12 These findings suggest that cerebral
vessels respond to LPS or cytokines with an increase in cyclooxygenase
products.
In this study, indomethacin did not inhibit dilatation to acetylcholine. In other studies, indomethacin did not inhibit dilatation caused by N-methyl-D-aspartate or nitroprusside, which suggests that the inhibitory effect of indomethacin is not a nonspecific effect.18 23 The dose of indomethacin used in the present experiment inhibits dilatation of cerebral arterioles in response to topical arachidonate.24
Recently, an inducible form of cyclooxygenase produced in response to
LPS and/or cytokines has been reported.25 26 Expression of
the inducible form of cyclooxygenase can be suppressed with
dexamethasone.27 28 Application of interleukin-1
to
cerebral arterioles of piglets produces vasodilatation that can be
blocked with indomethacin or actinomycin D, which suggests that
inducible cyclooxygenase can be expressed in brain.29 It
is possible that the inducible form of cyclooxygenase mediates a
portion of the cyclooxygenase-dependent dilation in the present
study.
Prostaglandins may enhance LPS-mediated expression of inducible NO synthase, as do other agents that increase intracellular cyclic AMP levels.30 31 Thus, indomethacin may inhibit vasodilatation in response to LPS by indirectly reducing LPS-mediated NO production. Cyclooxygenase enzymes are heme-containing compounds that can bind NO. In some systems, it appears that cyclooxygenases may be directly activated by NO.25 26 Thus, in the present study, inhibition of LPS-induced vasodilation by L-NMMA may be due to blockade of both NO production and NO-dependent cyclooxygenase products.
Overall, it appears that the interaction between cyclooxygenase products and NO is complex and bidirectional. We did not examine the dose-response relationships of indomethacin, dexamethasone, and L-NMMA with regard to responses to LPS, because potential interactions of NO and cyclooxygenase systems do not allow us to establish the relative importance of these vasodilator mechanisms. Instead, we used indomethacin, dexamethasone, and L-NMMA to determine whether NO and cyclooxygenase products were important in LPS-mediated dilatation.
In all groups except that treated with vehicle (artificial CSF), there was a modest decline in arterial pH over several hours. LPS was the only common agent in these groups, which suggests that LPS in cranial windows may have exerted an effect on systemic pH even though other variables such as arterial pressure were not affected. Mechanisms that mediate this effect are not clear. However, the modest decline in arterial pH (approximately 0.10 pH units) is not sufficient to account for the observed changes in cerebral arteriolar diameter. Furthermore, arteriolar diameter did not change when cranial windows were flushed with artificial CSF with normal pH, as observed previously.32
In summary, we have demonstrated that application of LPS in vivo causes dilatation of cerebral arterioles and that the vasodilatation can be attenuated by L-NMMA, dexamethasone, or indomethacin. These findings indicate an important role for both NO and cyclooxygenase products in the observed vasodilatation.
| Acknowledgments |
|---|
Received May 17, 1994; revision received August 18, 1994; accepted October 5, 1994.
| References |
|---|
|
|
|---|
. J Neuroimmunol. 1993;46:19-24. [Medline]
[Order article via Infotrieve]
in combination with tumor
necrosis factor, interleukin-1, or endotoxin. J Natl Cancer
Inst. 1990;82:722-776.
-induced pial arteriolar dilation and increased
prostanoid production in newborn pigs. Brain Res Bull. 1994;33:403-409. [Medline]
[Order article via Infotrieve]
This article has been cited by other articles:
![]() |
C.A. Gunnett, D.D. Lund, M.A. Howard III, Y. Chu, F.M. Faraci, and D.D. Heistad Gene Transfer of Inducible Nitric Oxide Synthase Impairs Relaxation in Human and Rabbit Cerebral Arteries Stroke, September 1, 2002; 33(9): 2292 - 2296. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Brian Jr, S. A. Moore, F. M. Faraci, and H. A. Kontos Expression and Vascular Effects of Cyclooxygenase-2 in Brain • Editorial Comment Stroke, December 1, 1998; 29(12): 2600 - 2606. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. Okamoto, O. Ito, R. J. Roman, A. G. Hudetz, and R. M. Bryan Jr Role of Inducible Nitric Oxide Synthase and Cyclooxygenase-2 in Endotoxin-Induced Cerebral Hyperemia • Editorial Comment Stroke, June 1, 1998; 29(6): 1209 - 1218. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Brian Jr, F. M. Faraci, and G. Feuerstein Tumor Necrosis Factor-{alpha}–Induced Dilatation of Cerebral Arterioles • Editorial Comment Stroke, February 1, 1998; 29(2): 509 - 515. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. M. FARACI and D. D. HEISTAD Regulation of the Cerebral Circulation: Role of Endothelium and Potassium Channels Physiol Rev, January 1, 1998; 78(1): 53 - 97. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1995 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |