Secondary Logo

Journal Logo

Effects of Nitric Oxide on Blood-Brain Barrier Disruption Caused by Intracarotid Injection of Hyperosmolar Mannitol in Rats

Chi, Oak Za MD; Chang, Qiang PhD; Wang, Guolin MD; Weiss, Harvey R. PhD

Neurosurgical Anesthesia

We performed this study to evaluate the effects of changing the level of nitric oxide (NO) on disruption of the blood-brain barrier (BBB) by hyperosmolar mannitol. Under isoflurane anesthesia, control rats (control group, n = 6) were given infusions with 25% mannitol into the internal carotid artery before measuring the transfer coefficient (Ki) of14 C-alpha-aminoisobutyric acid (14) C-AIB). In the CAS group (n = 6), [3-(cis-2,6-dimethyl piperidino)-sydnonimine] (CAS 754), a NO donor, was injected to decrease the mean arterial pressure (MAP) to 55 mm Hg and in the L-NAME group (n = 6), NG-nitro-L-arginine methyl ester (L-NAME), a NO synthase inhibitor, was injected before administering mannitol. In additional control animals (control + P group, n = 6) and additional CAS 754-treated animals (CAS + P group, n = 6), phenylephrine was infused to keep MAP at 130 mm Hg during the experimental period. In the control group, with mannitol injection, the Ki of the ipsilateral cortex (IC) where mannitol was injected increased to 4.3 times that of the contralateral cortex (CC) (17.2 +/- 2.9 vs 4.0 +/- 2.6 micro l [center dot] g-1 [centered dot] min-1). Without blood pressure control, the Ki of the IC of the CAS group (7.0 +/- 4.5) was lower and that of the L-NAME group (26.2 +/- 12.7) was higher than that of the control animals. At the same MAP, the Ki of the IC of the CAS + P group (9.6 +/- 3.1) was significantly lower than that of the control + P group (21.3 +/- 14.5) or that of the L-NAME group. There was no significant difference in the Ki of the IC between the control + P and the L-NAME groups. In conclusion, L-NAME worsened BBB disruption induced by hyperosmolar solution, which may be due to the pressure effect of L-NAME. CAS 754 was effective in attenuating disruption of the BBB caused by hyperosmolar mannitol. This effect is apparently not due to decreased MAP.

(Anesth Analg 1997;84:370-5)

Departments of (Chi, Chang, Wang) Anesthesia, and (Weiss) Physiology and Biophysics, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, New Brunswick, New Jersey.

Accepted for publication September 25, 1996.

Address correspondence and reprint requests to Oak Za Chi, MD, Department of Anesthesia, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, 125 Paterson St., Suite 3100, New Brunswick, NJ 08901-1977.

Recently, many studies have implicated nitric oxide (NO) in regulating vascular and blood-brain barrier (BBB) permeability [1-10]. The results of these studies, however, are conflicting, and the mechanism of the action of NO on the BBB is not clear. Inhibition of NO synthase causes microvascular stasis and produces focal areas of disruption of the BBB [10]. In the small intestine, inhibition of NO synthesis increased the basal permeability and augmented the increase of microvascular and mucosa permeability induced by ischemia-reperfusion and endotoxin [2-4]. On the contrary, NO synthase inhibitors also attenuate the increase of BBB permeability caused by some pathologic conditions such as meningitis, focal ischemia, ischemia-reperfusion, and acute hypertension [6-9].

Most studies with NO have been performed using inhibitors of NO synthase because NO itself has an extremely short half-life. Alternatively, donors of NO such as 3-(cis-2,6-dimethyl piperidino) sydnonimine (CAS 754) can be used to study the effects of NO. CAS 754 a metabolite of CAS 936 (3-[cis-2,6-dimethyl piperidino]-N-[4-methoxybenzoyl]-sydnonimine) [11], dilates blood vessels exclusively via NO- and cyclic guanidine monophosphate (cGMP)-dependent mechanisms [12]. In vitro, the rate of release of NO from CAS 754 is much slower than the rate of NO release from another sydnonimine, 3-morpholino sydnonimine [13]. Therefore, CAS 754 may have enough time to reach the brain and release a larger amount of NO [14]. We speculated that the effects of NO on the BBB could be observed more easily when the BBB was disrupted. In this study, the BBB was disrupted by an intracarotid injection of hyperosmolar mannitol.

We hypothesized that changing the level of NO by either a NO donor or a NO synthase inhibitor could modify the degree of BBB disruption by the hyperosmolar mannitol. In this study, CAS 754 was used as a NO donor and NG-nitro-L-arginine methyl ester (L-NAME) was used as a NO synthase inhibitor. BBB permeability was determined by measuring the blood-brain transfer coefficient (Ki) of14 C-alpha-aminoisobutyric acid (14) C-AIB).

Back to Top | Article Outline


This study was approved by our animal care and use committee. Thirty male Sprague-Dawley rats weighing 300-330 g were anesthetized with 1.4% isoflurane in an air and oxygen mixture (fraction of inspired oxygen 0.25-0.3), and were mechanically ventilated through a tracheal tube. A femoral artery and two femoral veins were catheterized and the femoral arterial catheter was connected to a Statham P23AA transducer. Blood pressure was continuously monitored and recorded. The right common, internal, and external carotid arteries were exposed. The external carotid artery was catheterized with a polyethylene tube (PE-50) after its branches were ligated. The tip of the catheter was placed about 1 mm distal to the carotid bifurcation. Femoral venous catheters were used to administer drugs and radioactive tracers. Body temperature was maintained at 37 degrees C with a heat lamp and a servocontrolled rectal thermister probe throughout the experimental period. Arterial blood pressure was recorded and 0.2-mL arterial blood samples were drawn anaerobically and analyzed for PaO2, PaCO2, and pHo during the experimental period. Rats were divided into five groups: In the control group (n = 6), no drugs were administered before opening the BBB with hyperosmolar mannitol. In the CAS group (n = 6), 10 min before injecting hyperosmolar mannitol, administering incremental doses of CAS 754 intravenously (IV) was begun (total dose 10-13 mg/kg) to decrease the mean arterial blood pressure (MAP) to about 55 mm Hg. In the L-NAME group (n = 6), L-NAME 10 mg/kg was administered IV 10 min before administering mannitol. In the control + P group (n = 6), phenylephrine was infused to maintain MAP at about 130 mm Hg beginning 5 min before mannitol was injected until the measurement of Ki was completed. In the CAS + P group (n = 6), after animals were treated with CAS in the same way as the CAS group, phenylephrine was infused to keep MAP at about 130 mm Hg beginning 5 min before injecting mannitol until the measurement of Ki was completed. We kept the MAP at about 130 mm Hg because this level of pressure was obtained after injecting L-NAME.

In order to open the BBB, a solution of 25% mannitol, filtered and warmed to 37 degrees C, was infused through the catheter in the carotid artery for 20 s at a rate of 0.25 mL [center dot] kg-1 [center dot] s-1. The rate of infusion was set so that it was well below the rate that causes a hypertensive opening of the BBB in anesthetized rats, but with an optimum disruption of the BBB without serious immediate or delayed neurotoxicity [14,15]. To determine Ki, 20 micro Ci of14 C-AIB (Amersham, Arlington Heights, IL) was rapidly injected IV and flushed with 0.5 mL of normal saline, 2 min after mannitol infusion. Blood samples were collected from the arterial catheter at 15-s intervals for the first 2 min and then, every minute for the next 8 min. Five minutes after injecting14 C-AIB, 20 micro Ci of3 H-dextran (70,000 daltons, Amersham) was injected IV and flushed with 0.5 mL of normal saline. After collecting the 10-min arterial blood sample, the animals were decapitated and their brains quickly frozen in liquid nitrogen. The following brain regions were dissected: ipsilateral cortex ([IC] where mannitol was injected), contralateral cortex (CC), cerebellum, and pons. They were solubilized in Soluene[R] (Packard, Downers Grove, IL) before counting the radioactivity. Arterial blood samples were centrifuged and the plasma was separated and processed for scintillation counting in the same way as the brain samples. Plasma and brain samples were counted on a liquid scintillation counter that was equipped for dual-label counting. Quench curves were prepared using carbon tetrachloride and all samples were automatically corrected for quenching. The blood-to-tissue transfer coefficient for14 C-AIB was determined, assuming a unidirectional transfer of14 C-AIB over a 10-min period of the experiment, using the following Equation asdescribed by Gross et al. [16]: Equation 1 where Am is the equal amount of14 C-AIB radioactivity in the tissue per gram, Vp is the volume of plasma retained in the tissue determined from the3 H-dextran data where Vp is obtained by dividing the amount of3 H-dextran radioactivity in the tissue per gram by the concentration of3 H-dextran in the plasma at the time of decapitation, Cp(t) is the arterial concentration of (14) C-AIB over time t, and CT is the arterial plasma concentration of14 C-AIB at the time of decapitation. In the Equation todetermine Ki, Vp x CT is a correction term which accounts for the label14 C retained in the vascular compartment of the tissue, Am.

(Figure 1) For analysis of data, a factorial analysis of variance was used to assess the differences between the groups, and between the various examined regions for the transfer coefficient and vital signs of those groups. The statistical significance of differences was determined using Duncan's procedure. For the significance of the effect of blood pressure on Ki and on the vital signs, two-tailed, unpaired t-tests were performed. All data were expressed as mean +/- SD, and significance was defined as P < 0.05.

Figure 1

Figure 1

Back to Top | Article Outline


The MAP of the CAS group was 41% lower and the MAP of the L-NAME group was 23% higher than that of the control group (Table 1). Phenylephrine infusion in the control and in the CAS groups increased the MAP by 18% and 124%, respectively. Heart rate and blood gas variables were not significantly different among the control, the CAS, and the L-NAME groups. Increasing MAP in the control and the CAS groups did not influence heart rate and blood gas variables.

Table 1

Table 1

The Ki of the CC of the control group was 4.0 +/- 2.6 micro L [centered dot] g-1 [center dot] min-1. In the IC where the hyperosmolar mannitol was injected, Ki was increased to 4.3 times that of the CC. Without blood pressure control, the Ki of the IC of the CAS group was 59% lower and the Ki of the IC of the L-NAME group was 52% higher than that of the control animals. There was no significant difference in the Ki of the IC between the control and the control + P groups and between the CAS and the CAS + P groups. At the same MAP, the Ki of the IC of the CAS + P group was 55% and 63% lower than that of the control + P and the L-NAME groups, respectively. There was no significant difference in the Ki of the IC between the control + P and the L-NAME groups. In all of the experimental groups of animals, the Ki of the IC was significantly higher than that of the corresponding CC, and the Ki of the CC was similar among all of the experimental groups.

In the noncortical brain regions such as the cerebellum and pons, the Ki was similar among all of the experimental groups. In the L-NAME and the control groups, the Ki of the cerebellum and pons was higher than that of the CC. The Ki of the cerebellum or pons was lower than that of the IC in all of the groups except in the CAS and in the CAS + P groups. Increasing MAP with phenylephrine produced no significant changes in the Ki of cerebellum or pons in the CAS-treated (CAS + P group) or in the control animals (control + P group) (Table 2).

Table 2

Table 2

Back to Top | Article Outline


Our study demonstrated that without controlling blood pressure, in the cortical area where hyperosmolar mannitol was injected (IC), the Ki was higher in the L-NAME-treated animals, and lower in the CAS-treated animals than in the control animals. At the same MAP, the Ki of the IC of the CAS + P group was lower than that of either the control + P or the L-NAME group. There was no significant difference in the Ki of the IC between the control + P and the L-NAME groups.

In our previous study, at the same MAP, the regional cerebral blood flow (rCBF) (was significantly higher in CAS 754-treated animals than in control and nitroprusside-treated animals [17]. We speculated that a relatively slower release of NO from CAS 754 allowed more CAS 754 to reach the brain before releasing NO, and that more NO was available in the brain when compared with nitroprusside, which releases NO in a very short period by a simple dissociation [18]. Since the time structure of this study was very similar to that of the previous one, CAS 754 should have produced NO and should have affected the BBB during the experimental period.

In this study, we speculated that the effect of NO on the BBB could be observed more easily if the BBB were disrupted. The rate and duration of infusion and the concentration of hyperosmolar mannitol that were used in this study were reported to produce a reversible and optimal opening of the BBB [15]. The exact mechanism of the BBB disruption by hyperosmolar mannitol is not clear. It could be biophysical or biochemical or both [15]. Mannitol may affect production of NO via the endothelial pathway and may change the amount of NO available in the brain tissue.

Previous studies showed that Ki was independent of rCBF when rCBF was within the range that occurred during administration of L-NAME, CAS 754, or mannitol under isoflurane anesthesia [8,17,19,20]. Therefore, determination of rCBF was not performed in this study.

NO not only influences microvasculature directly but also works as an important messenger for intercellular and intracellular communication [21,22]. During neuroexcitation, NO may be released and modify the BBB permeability, and it may play a significant role in controlling the movement of metabolites across the BBB. The major target for NO is soluble guanylate cyclase. Capillary endothelial cells of the brain respond to exogenous NO donors, such as 3-morpholino sydnonimine and sodium nitroprusside, by greatly increasing cGMP formation [23]. The functional consequence of an increased production of cGMP in the endothelial cells in the brain is not clear. NO as well as cGMP may play a role in the regulation of transport of ions, nutrients, and other substances across the BBB [1].

Extreme hypertension could cause BBB disruption [24,25]. Since BBB was disrupted by hyperosmolar mannitol in our study, we speculated that Ki could be dependent even on a moderate change of MAP. Therefore, two sets of experiments with or without MAP control were performed.

In noncortical areas, as well as in the CC in our study, Ki was similar and was within reported ranges in all of the experimental groups. This suggests that the pathological opening of the BBB did not occur, and that the changing level of NO did not significantly influence BBB permeability in the absence of hyperosmolar BBB opening under the conditions of our study. Our previous study also showed that L-NAME decreased Ki only in the ischemic cortex, not in most of the other brain regions [8]. Our data suggest that normal brain tissue may be more resistant to disruption of the BBB by changing the level of NO.

Our study demonstrated that CAS 754 attenuated the increase of Ki by hyperosmolar mannitol, and that this effect was apparently not due to decreased MAP, because the great increase in MAP (63 +/- 18 to 141 +/- 26 mm Hg) in the CAS-treated animals did not significantly influence the Ki of the IC. At the same MAP, the Ki of the IC of the CAS 754-treated animals was lower than that of the control + P or the L-NAME group. Boughton-Smith et al. [4] reported that S-nitro-N-acetyl penicillamine, which also generates NO, inhibited increases in vascular permeability in the intestine induced by platelet activating factor. They reported that NG-monomethyl-L-arginine potentiated gastrointestinal damage and plasma leakage induced by endotoxin, but that it had no effect on platelet activating factor-induced damage of the intestine. The investigators speculated that the protection afforded by endogenous NO may be dependent on the nature of the inflammatory stimulus [4]. In monolayers of bovine aortic endothelial cells, the permeability coefficient of14 C-sucrose was significantly decreased by an activator of guanylate cyclase, and depletion of L-arginine increased14 C-sucrose permeability [5]. Our results are similar to their findings.

However, Boje [6] reported that, in the brain, lipopolysaccharides which induce NO synthase increased NO production and blood-cerebrospinal fluid (CSF) barrier permeability, and she suggested that the pathological production of NO may contribute to the disruption of the blood-CSF barrier during meningitis. The reason for the difference between the results of our study and Boje's is not clear. The mechanism by which brain endothelial cells are disrupted may be different between the hyperosmolar insult and meningitis. It is also possible that a certain amount of NO may be required to maintain the integrity and basal permeability of the BBB. Too little or an excessive amount of NO may be harmful. A hyperosmolar solution could decrease the available amount of NO, and production of NO by CAS 754 may help to maintain the normal range of NO and the integrity of the BBB.

NO inhibits platelet and leukocyte aggregation and their adherence to the endothelial cells [26,27]. Thus, NO could help keep the microvasculature intact during a hyperosmolar condition. NO may work on the channels in the endothelial cells and could make endothelial cells relax or expand [5]. This will make the intercellular junctions tighter and render endothelial cells resistant to hyperosmolar dehydration. Whatever the mechanism may be, our data showed that CAS 754 was effective in attenuating BBB disruption caused by hyperosmolar mannitol. These data suggest that NO may be effective in protecting the BBB during its disruption by hyperosmolar mannitol.

Our study showed that L-NAME enhanced the increase of Ki of the IC by hyperosmolar mannitol when MAP was not controlled. However, when MAP was controlled, there was no significant difference between the L-NAME-treated and the control animals. These data suggest that the higher Ki of the IC in the L-NAME-treated animals may be due to the increase of MAP by L-NAME. Inhibition of the NO synthase has been reported to increase microvascular permeability without a change of capillary pressure in the intestine [2,3]. Another study showed that inhibition of NO synthesis increased the flow of labeled sucrose across endothelial monolayers [5].

In contrast, it has been reported that during meningitis, administering aminoguanidine, an inhibitor of NO synthase, blocked meningeal NO production and significantly attenuated changes in the permeability of the blood-CSF barrier [6]. L-NAME has been reported to be effective in reducing BBB permeability during cerebral ischemia-reperfusion, focal cerebral ischemia, and acute hypertension [7-9]. It is not clear why our data are different from that of others and different from that of our previous study on focal ischemia. It is possible that L-NAME could not reach the ischemic area in a sufficient amount to produce a maximum pharmacological effect. An optimal amount of NO may be needed to maintain BBB integrity and each pathological condition may produce different microvascular changes as suggested in a previous discussion. In most studies on BBB permeability, albumin or dextran was used as an indicator. In our study, however, a small neutral amino acid was used. It is possible that the permeability of albumin or dextran could be more affected by vasoconstrictors such as L-NAME. If the effect of a blood pressure increase by L-NAME on the BBB was excluded, our data suggest that the decrease of NO in our study produced no significant effect on the BBB disruption caused by hyperosmolar mannitol.

In conclusion, the L-NAME caused a worsening of BBB disruption produced by the hyperosmolar solution. This effect may be due to the increase of MAP by L-NAME. CAS 754, a NO donor, was effective in attenuating disruption of the BBB caused by hyperosmolar mannitol. This effect was apparently not due to the decreased MAP caused by CAS 754. These data suggest that an increasing amount of NO may be effective in preventing disruption of the BBB by hyperosmolar mannitol.

The authors wish to thank Patricia A. Sheffield, MA, for her expert editorial assistance.

Back to Top | Article Outline


1. Janigro D, West GA, Nguyen TS, Winn HR. Regulation of blood-brain barrier endothelial cells by nitric oxide. Circ Res 1994;75:528-38.
2. Kubes P, Granger DN. Nitric oxide modulates microvascular permeability. Am J Physiol 1992;262:H611-5.
3. Kubes P. Ischemia-reperfusion in feline small intestine: a role for nitric oxide. Am J Physiol 1993;264:G143-9.
4. Boughton-Smith NK, Deakin AM, Whittle BJ. Actions of nitric oxide on the acute gastrointestinal damage induced by PAF in the rat. Agents Actions 1992;Spec No:C3-9.
5. Oliver JA. Endothelium-derived relaxing factor contributes to the regulation of endothelial permeability. J Cell Physiol 1992;151:506-11.
6. Boje KM. Inhibition of nitric oxide synthase partially attenuates alterations in the blood-cerebrospinal fluid barrier during experimental meningitis in the rat. Eur J Pharmacol 1995;272:297-300.
7. Zhang J, Benveniste H, Klitzman B, Piantadosi CA. Nitric oxide synthase inhibition and extracellular glutamate concentration after cerebral ischemia/reperfusion. Stroke 1995;26:298-304.
8. Chi OZ, Wei HM, Sinha AK, Weiss HR. Effects of inhibition of nitric oxide synthase on blood-brain barrier transport in focal cerebral ischemia. Pharmacology 1994;48:367-73.
9. Mayhan WG. Role of nitric oxide in disruption of the blood-brain barrier during acute hypertension. Brain Res 1995;686:99-103.
10. Prado R, Watson BD, Kuluz J, Dietrich WD. Endothelium-derived nitric oxide synthase inhibition: effects on cerebral blood flow, pial artery diameter, and vascular morphology in rats. Stroke 1992;23:1118-24.
11. Siegfried MR, Erhardt J, Rider T, et al. Cardioprotection and attenuation of endothelial dysfunction by organic nitric oxide donors in myocardial ischemia-reperfusion. J Pharmacol Exp Ther 1992;260:668-75.
12. Bohn H, Beyerle R, Martorana PA, Schonafinger K. CAS 936, a novel sydnonimine with direct vasodilating and nitric oxide donating properties: effects on isolated blood vessels. J Cardiovasc Pharmacol 1991;18:522-7.
13. Mulsch A, Hecker M, Mordvintcev PI, et al. Enzymic and nonenzymic release of NO accounts for the vasodilator activity of the metabolites of CAS 936, a novel long acting sydnonimine derivative. Naunyn Schmiedebergs Arch Pharmacol 1993;347:92-100.
14. Rapoport SI. Opening of the blood-brain barrier by acute hypertension. Exp Neurol 1976;52:467-79.
15. Cosolo WC, Martinello P, Louis WJ, Chrisophidis N. Blood-brain barrier disruption using mannitol: time course and electron microscopy studies. Am J Physiol 1989;256:R443-7.
16. Gross PM, Blasberg RG, Fenstermacher JD, Patlak CS. The microcirculation of rat circumventricular organs and pituitary gland. Brain Res Bull 1987;18:73-85.
17. Chi OZ, Wei HM, Weiss HR. Effects of CAS 754, a new nitric oxide donor, on regional cerebral blood flow in focal cerebral ischemia. Anesth Analg 1995;80:703-8.
18. Rudolph W, Dirschinger J. Effectiveness of molsidomine in the long term treatment of exertional angina pectoris and chronic congestive heart failure. Am Heart J 1985;109:670-4.
19. Chi OZ, Wei HM, Lu X, Weiss HR. Increased blood-brain permeability with hyperosmolar mannitol increases cerebral O2 consumption and O2 supply/consumption heterogeneity. J Cereb Blood Flow Metab 1996;16:327-33.
20. Smith QR. Quantitation of blood-brain barrier permeability. In: Neuwelt EA, ed. Implications of the blood-brain barrier and its manipulation. Vol 1. Basic science aspects. New York: Plenum, 1989:85-118.
21. Snyder SH, Bredt DS. Nitric oxide as a neuronal messenger. Trends Pharmacol Sci 1991;12:125-8.
22. McCall T, Vallance P. Nitric oxide takes centre-stage with newly defined roles. Trends Pharmacol Sci 1992;13:1-6.
23. Marsault R, Frelin C. The activation by nitric oxide of guanylate cyclase in endothelial cells form brain capillaries. J Neurochem 1992;59:942-5.
24. Heistad DD, Marcus ML. Effect of sympathetic stimulation on permeability of the blood-brain barrier to albumin during acute hypertension in cats. Circ Res 1979;45:331-8.
25. Mayhan WG, Heistad DD. Role of veins and cerebral venous pressure in disruption of the blood-brain barrier. Circ Res 1986;59:216-20.
26. Kubes P, Suzuki M, Granger DN. Nitric oxide: an endogenous modulator of leukocyte adhesion. Proc Natl Acad Sci USA 1991;88:4651-5.
27. Radomski MW, Palmer RM, Moncada S. Endogenous nitric oxide inhibits human platelet adhesion to vascular endothelium. Lancet 1987;2:1057-8.
© 1997 International Anesthesia Research Society