Apr 17, 1992 - bone was left intact and removed with forceps under micro- scopic observation to ..... stimulation; this results in a decrease in oxygen extraction ... attenuate the hyperemia and pial vessel dilation associated with cortical ...
Proc. Natl. Acad. Sci. USA Vol. 89. pp. 6649-6652, July 1992
Competitive inhibition of nitric oxide synthase prevents the cortical hyperemia associated with peripheral nerve stimulation (cerebral blood flow/cortical activation/microdialysis)
FRANCES J. NORTHINGTON*, G. PAUL MATHERNE*t, AND ROBERT M. BERNEt University of Virginia Health Sciences Center, Departments of TPhysiology and *Pediatrics, Charlottesville, VA 22908
Contributed by Robert M. Berne, April 17, 1992
60 mg of sodium thiomylal per kg of body weight administered i.p. initially, trifluoroethane 0.5-1.5% administered during surgical and experimental periods, and 2% viscous lidocaine administered locally to incision sites. Animals were ventilated throughout the experiments to maintain pH = 7.35-7.40, PACO, 35 mmHg, and PA(, 150 mmHg with a mixture of oxygen and nitrogen. Femoral arterial cannulas were inserted to continuously monitor arterial blood pressure, heart rate, and arterial blood gases. Temperature was monitored with a rectal probe and maintained at 370C with a warming blanket. The animal's head was secured in a stereotaxic frame for bilateral placement of the microdialysis cannulas in the hindlimb sensory-motor cortex. The coordinates for this area (1.5-4 mm lateral and 0.3-3 mm posterior to the bregma) were obtained from a stereotaxic atlas (13). A 2 x 2 mm area of the skull was removed with a variable-speed drill over the posterior parietal cortex but away from the hindlimb area to allow for horizontal introduction of the cannulas into the superficial cortex with a micromanipulator. A thin layer of bone was left intact and removed with forceps under microscopic observation to minimize trauma to the cortex. Cannulas were mounted horizontally on the stereotaxic arm and advanced to predetermined coordinates that would place the hindlimb sensory-motor cortex between the tips of the inflow and outflow silica tubes within the microdialysis cannula. The distance between the tips constitutes the effective dialyzing area of the cannula (14). Once the cannulas were in place, the area of the craniotomy was sealed, and the cannulas were fixed in position with dental cement. The animals were then removed from the stereotaxic apparatus and allowed a 90-min postsurgical equilibration period in the anesthetized state. Placement of the cannula within the hindlimb sensory-motor cortex was confirmed at the end of each experiment by dissection. After baseline data were obtained, the sciatic nerves were isolated for stimulation. Briefly, the isolated nerve was severed proximal to its division into the tibial and peroneal nerves, and stimulating electrodes were placed on the proximal end. Parameters used for stimulation (0.2 V, 5 Hz) were those shown to produce minimal cardiovascular effects in other studies (15). Dialysis. The modified brain dialysis technique used in these studies as adapted from that of Johnson and Justice (16) has been described in detail (14, 17). The dialysis probe consists of a single, hollow dialysis fiber, one end of which is sealed with epoxy. The dialysis membrane diameter is 300 gm, and it has a molecular mass cutoff of 5 kDa. Two hollow silica perfusion tubes are inserted into the dialysis fiber so that their ends are 3 mm apart. To estimate cortical blood flow with the hydrogen-clearance technique, a platinum wire
With combined microdialysis and hydrogen ABSTRACT clearance techniques for simultaneous local delivery of drugs and blood-flow measurement in the rat hindlimb sensorymotor cortex, we examined the role of nitric oxide in cerebral blood-flow regulation during sciatic nerve stimulation. Infusion of 1 mM nitric oxide synthase antagonist, N'1-nitro-L-arginine methyl ester (L-NAME), blocked the cortical blood-flow response to sciatic nerve stimulation (152 ± 43 ml mint.100 g't of tissue in controls and 73 + 11 ml-min- l 100 go I in the presence of L-NAME; P < 0.05). Addition of 10 mM L-arginine to the dialysate containing L-NAME partially restored the hyperemic response to nerve stimulation (125 ml mind.100 go ). L-NAME also produced a decrease in baseline cerebral blood flow when compared with the control (66 ± 14 ml mind 100 g- vs. 93 + 25 ml-min-t.100 g-l). We conclude that nitric oxide from activated neurons participates in the local regulation of cortical blood flow in response to sciatic nerve stimulation and also in the maintenance of basal cortical blood flow.
The mechanisms regulating cerebral blood flow in response to peripheral nerve stimulation have not been clearly defined. The regional nature of the response strongly suggests that a locally produced vasoactive mediator is involved. Nitric oxide is produced by neurons and astroglia in response to stimulation by excitatory amino acids (1-6) and by nonadrenergic, noncholinergic vasodilator nerves (7, 8). It also participates in cell-signaling events in the brain (9-11). Computer modeling suggests that nitric oxide is a mediator in the local regulation of cerebral blood flow during cortical activation (12), but this has not been proven. The present study was designed to determine if nitric oxide participates in the regulation of cerebral blood flow in response to peripheral nerve stimulation. To accomplish this, the combined microdialysis and hydrogen-clearance techniques were used for local drug administration and simultaneous cerebral blood-flow measurement, respectively, in the hindlimb sensory-motor cortex in response to sciatic nerve stimulation. Experiments were performed in the presence and absence of the competitive nitric oxide synthase inhibitor N'-nitro-t,-arginine methyl ester (L-NAME) and in the presence of both L-NAME and the nitric oxide precursor, L-arginine. The effect of L-NAME on baseline cortical blood flow was also determined.
MATERIAL AND METHODS Preparation. Adult male Wistar rats (300-450 g) were used. The experimental protocol was approved by the Animal Research Committee of the University of Virginia and conforms to the National Institutes of Health guidelines for the care and use of animals in research. Anesthesia consisted of
Abbreviations: L-NAME, N7'-nitro-L-arginine methyl ester; CSF, cerebrospinal fluid. :To whom reprint requests should be addressed at: Department of Physiology, 1116 MR4 Health Sciences Center, Charlottesville, VA 22908.
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Proc. Natl. Acad. Sci. USA 89
Physiology: Northington et al.
(bare diameter, 0.01 mm) is included in the cannula. The coating is removed from -5 mm of either end and one end is inserted into the dialysis membrane. After insertion, the inflow silica tube was connected to an infusion pump, and the dialysis cannula was perfused at 0.5 ,ul/min. The electrolyte millimolar concentration of the artificial cerebrospinal fluid (CSF) was as follows: KCl, 3.0; MgCl2, 0.65; NaCl, 131.8; NaHCO3, 24.6; urea, 6.7; dextrose, 3.7; and CaCl2, 2.0. The CSF was filtered, warmed to 370C, and bubbled with 95% N2/5% CO2 until it was drawn into gas-tight syringes to be perfused through the microdialysis cannulas. This produced a solution with pH = 7.30 + 0.05, Pco2 = 40 4 mmHg, Po, = 27 7 mmHg, and HCO= 20 2 mM. This is similar to the oxygen tension of CSF and brain tissue (18). The use of the hydrogen-clearance technique for measurement of cerebral blood flow in the area surrounding the dialysis cannula has also been described in detail previously (17, 19). Local cerebral blood flow (CBF) is calculated from the time required for the levels of hydrogen introduced and then removed from the inspired air to decrease from 90% to 40% of maximum by using the following formula: ±
ln(40/90) time (min) -
CBF (mlmin-'100 g-)
250 x 100
Baseline Stimulation Recovery FIG. 1. Cortical blood flow (mean SD) measured with hydroclearance in the hindlimb sensory cortex of the rat is shown during baseline, stimulation, and recovery under three experimental conditions. (Top) Perfusion of the microdialysis/hydrogen-clearance cannula with artificial CSF. (Middle) Perfusion with CSF containing 1 mM L-NAME. (Bottom) Perfusion with CSF containing 1 mM L-NAME and 10 mM L-arginine. *, P < 0.05 vs. baseline; t, P < 0.05 vs. corresponding period in Top; §, P < 0.05 vs. stimulation in Middle. ±
taining L-NAME (Fig. 1 Bottom). Comparison of basal cerebral blood flow in the presence (Fig. 1 Middle) and absence (Fig. 1 Top) of L-NAME revealed that baseline blood flow was significantly depressed in the presence of L-NAME. In the presence of both L-NAME and L-arginine (Fig. 1 Bottom) blood flow was between that observed in the presence of L-NAME alone and in the absence of L-NAME.
RESULTS The animals were hemodynamically stable, with normal blood gas levels throughout the experiments. Stimulation produced no hemodynamic alterations (Table 1). Sciatic nerve stimulation increased blood flow in the contralateral hindlimb sensory-motor cortex from a baseline value of 93 25 ml min-1-100 g-1 to 152 ± 43 ml min-1 100 g-1 (Fig. 1 Top). During recovery, blood flow decreased to 113 43 ml min-1-100 g-1. The addition of L-NAME to the dialysate completely prevented this increase in flow (Fig. 1 Middle). The hyperemic response to stimulation was partially restored with the addition of L-arginine to the dialysate con-
DISCUSSION The significant findings of this study are that the blood flow response to cortical activation produced by peripheral nerve stimulation can be abolished by the local infusion of a nitric oxide synthase antagonist and that this blockade can be partially overcome by the addition of L-arginine. These
Table 1. Arterial blood
(refs. 17, 19, and 20). The local response of cortical blood flow during baseline, sciatic nerve stimulation, and recovery periods was tested under three conditions. In one group (n = 10), contralateral cortical blood flow was measured in the hindlimb sensorymotor cortex during perfusion of the dialysis cannula with artificial CSF. In a second group (n = 11), cerebral blood flow was measured during perfusion of the hindlimb sensorymotor cortex with CSF containing 1 mM L-NAME. In the third group (n = 5), the dialysis cannula was perfused with artificial CSF containing 1 mM L-NAME and 10 mM L-arginine. L-Arginine and L-NAME were obtained from Sigma. Cerebral blood flow measurements during baseline, stimulation, and recovery periods were compared among the groups with two-way ANOVA for repeated measurements, and individual differences were identified with the NeumannKeuls multiple comparison procedure. P < 0.05 was considered significant. Hemodynamic and arterial blood gas data were compared with one-way ANOVA. Data are presented 1 SD in all cases. as means
and hemodynamic data
PAO2, PAco, HCO°' mM mmHg mmHg Baseline 7.40 + 0.04 38 + 5 24 + 3 132 + 21 37 5 Stim 7.41 0.05 24 + 3 134 18 7.41 0.05 40 6 25 3 136 23 Recovery MAP, mean arterial pressure; HR, heart rate; Stim, stimulation.
mmHg 106 + 11 109 14 106 + 15
HR, beats per
342 350 362
min 45 39 57
Proc. Natl. Acad. Sci. USA 89 (1992)
Physiology: Northington et al. results suggest that release of nitric oxide or some closely related compound serves to regulate local cortical blood flow responses to peripheral nerve stimulation. In a three-dimensional model of cell signaling in the brain, it has been hypothesized that nitric oxide might serve as a local mediator of blood flow during cortical activation (12). Our findings provide direct evidence that this hypothesis is correct. The infusion of the competitive nitric oxide synthase antagonist, L-NAME, prevents any increase in cortical blood flow in the contralateral hindlimb sensory-motor cortex during sciatic nerve stimulation. By including L-arginine in the dialysate with the nitric oxide synthase inhibitor, the expected increase in local cortical blood flow is partially restored. This indicates that the L-arginine pathway is involved in the local regulation of cortical blood flow. It also proves that the nitric oxide synthase inhibitor does not prevent an increase in flow in response to stimulation by a toxic effect on the surrounding neurons, astroglia, or microvasculature because the reactivity of the vasculature to cortical activation is restored with an excess of substrate. Although not a major emphasis of this study, our findings of decreased basal cerebral blood flow in the presence of L-NAME suggest that nitric oxide production contributes to the maintenance of basal cerebral blood flow. These findings are consistent with those of others who have demonstrated a role for basal endogenous nitric oxide in the control of vasomotor tone throughout the body (21, 22). Finding an intermediate baseline cerebral blood flow in the experiments in which L-arginine was added to the CSF containing L-NAME also supports a role for nitric oxide in control of basal cerebral blood flow. There is a large body of evidence that the necessary nitric oxide synthetic mechanisms are present and that neurons respond to stimulation with excitatory amino acids with the production of nitric oxide. Enzymes necessary for the synthesis of nitric oxide have been found in the brain, most abundantly in the cerebellum, but also in the cortex and, in particular, in neurons (9, 23, 24). It has been demonstrated that nitric oxide is produced by both neurons and astroglia (1, 3, 4). Specifically, these two cell types release nitric oxide in response to activation by excitatory amino acids glutamate, kainate, and quisqualate (3, 4, 6). Blockade of the N-methylD-aspartic acid receptor prevents both the release of nitric oxide from neurons and the accumulation of cGMP (5, 11). Since the original description ofthe role of the endothelium in the response of vascular smooth muscle to some vasoactive drugs (25), it has been well established that endotheliumderived relaxing factor/nitric oxide is a potent vasodilator that acts by stimulation of cGMP production in a variety of vascular beds and isolated cerebral vessel preparations (1, 7, 8, 26-28). In some isolated cerebral vessels, nitric oxide also mediates the endothelium-independent vasodilation produced by nonadrenergic, noncholinergic nerve stimulation (7, 8). This study describes the role of nitric oxide in the regulation of cortical blood flow during peripheral nerve stimulation. The nature of the mechanism responsible for regulation of cortical blood flow during peripheral nerve stimulation has been a topic of interest and controversy since the demonstration that extremely well localized increases in cortical blood flow occur in response to various stimuli (29). It was originally assumed that the mediator was released as a result of a local metabolic deficit that occurred during peripheral nerve stimulation (29, 30). However, it has subsequently been shown that there is probably no local metabolic deficit. Leniger-Follert and Lubbers (18) have shown that local tissue Po2 actually increases during cortical activation, and Yaksh and colleagues (31) suggest that nutritive flow is enhanced prior to the onset of any metabolic deficit during cortical activation. Additionally, Fox and coworkers (32, 33)
have shown that cerebral oxygen consumption and cerebral blood flow are "uncoupled" during both visual and tactile stimulation; this results in a decrease in oxygen extraction during cortical activation. Most recently, vasoactive intestinal peptide (VIP) and adenosine have been implicated as potential local mediators of cerebral blood flow during cortical activation. However, antibodies to VIP and adenosine antagonists only partially attenuate the hyperemia and pial vessel dilation associated with cortical activation (30, 31). Additionally, we have recently shown that there is no local increase in interstitial adenosine in the hindlimb sensory-motor cortex and that the adenosine antagonist 8-sulfophenyltheophylline fails to block the increase in flow associated with sciatic nerve stimulation (34). Furthermore, the local increase in cortical blood flow in response to peripheral nerve stimulation is not mediated through adrenergic or cholinergic receptors (35). One of the difficulties encountered in establishing a regulatory role for a specific mediator in the control of cortical blood flow during peripheral nerve stimulation has been the lack of a technique with which to both precisely deliver specific agonists or antagonists and to measure local changes in cortical blood flow in the activated area. The combined microdialysis and hydrogen clearance technique has been used for this purpose while studying potential mediators of cerebral blood flow in response to other stimuli (14, 17, 34). A related injection technique has also been used to establish a role for nitric oxide in the modulation of baroreceptor reflexes in the nucleus tractus solitarius (1). Despite the fact that cannula introduction causes initial release of vasoactive mediators, evidence suggests that both levels of local metabolites and cerebrovascular reactivity return to normal within the 90-min postsurgical recovery period, allowing accurate estimation of cortical blood flow with this technique (14, 36). In summary, we have shown that blockade of nitric oxide synthase prevents the hyperemia associated with cortical activation produced by peripheral nerve stimulation and that an excess of L-arginine partially restores the response. Based on these findings, it is likely that the mechanism for the regulation of local cortical blood flow during cortical activation involves the paracrine vasodilator action of nitric oxide released from activated neurons in the local microvascular bed or from nonadrenergic, noncholinergic vasodilator nerves. The authors gratefully acknowledge the helpful discussions of this work with Douglas Spitz, Ph.D. This work was supported by National Institutes of Health Grant HL10384, Veteran's Administration Grant VA-90-F-16, and an institutional grant from the Children's Medical Center, University of Virginia. F.J.N. is a research fellow of the Virginia Affiliate of the American Heart Association. G.P.M. is the recipient of National Heart, Lung, and Blood Institute Clinical Investigator Award HL02457 and a March of Dimes Basel O'Connor Starter Scholar Award. 1. Di Paloa, E. D., Vidal, M. J. & Nistico, G. (1991) J. Cardiovasc. Pharmacol. 17, S269-S272. 2. Garthwaite, J., Garthwaite, G., Palmer, R. M. J. & Moncada, S. (1989) Eur. J. Pharmacol. 172, 413-416. 3. Murphy, S., Minor, R. L., Jr., Welk, G. & Harrison, D. G.
(1991) J. Cardiovasc. Pharmacol. 17, S265-S268.
4. Dawson, V. L., Dawson, T. M., London, E. D., Bredt, D. S. & Snyder, S. H. (1991) Proc. Natl. Acad. Sci. USA 88, 6368-6371. 5. Southam, E., East, S. J. & Garthwaite, J. (1991) J. Neurochem. 56, 2072-2081. 6. Garthwaite, J., Southam, E. & Anderton, M. (1989) J. Neurochem. 53, 1952-1954. 7. Lee, T. J. F. & Sarwinski, S. J. (1991) Blood Vessels 28, 407-412.
Physiology: Northington et al.
8. Toda, N. & Okamura, T. (1991) J. Pharmacol. Exp. Ther. 258, 1027-1032. 9. Knowles, R. G., Palacios, M., Palmer, R. M. J. & Moncada, S. (1989) Proc. Natl. Acad. Sci. USA 86, 5159-5162. 10. Shuman, E. M. & Madison, D. V. (1991) Science 254, 15031506. 11. Garthwaite, J., Charles, S. L. & Chess-Williams, R. (1988) Nature (London) 336, 385-388. 12. GaIly, J. A., Montague, P. R., Reeke, G. N., Jr., & Edelman, G. M. (1990) Proc. Nati. Acad. Sci. USA 87, 3547-3551. 13. Paxinos, G. & Watson, C. (1986) in The RatBrain in Stereotaxic Coordinates (Academic, Orlando, FL), 2nd Ed., plates 19-31. 14. Van Wylen, D. G. L., Park, T. S., Rubio, R. & Berne, R. M. (1986) J. Cereb. Blood Flow Metab. 6, 522-528. 15. Ngai, A. C., Ko, K. R., Morii, S. & Winn, H. R. (1988) Am. J. Physiol. 254, H133-H139. 16. Johnson, R. D. & Justice, J. B. (1983) Brain Res. Bull. 10, 567-571. 17. Van Wylen, D. G. L., Park, T. S., Rubio, R. & Berne, R. M. (1989) J. Cereb. Blood Flow Metab. 9, 556-562. 18. Leniger-Follert, E. & Lubbers, D. W. (1976) Pflugers Arch. 366, 39-44. 19. Pasztor, E., Symon, L., Dorsch, N. W. C. & Branston, N. M. (1973) Stroke 4, 556-567. 20. Young, W. (1980) Stroke 11, 552-564. 21. Moncada, S., Palmer, R. M. J. & Higgs, E. A. (1991) Pharmacol. Rev. 43, 109-142. 22. Bellan, J. A., Minkes, R. K., McNamara, D. B. & Kadowitz, P. J. (1991) Am. J. Physiol. 260, H1025-H1029. 23. Forstermann, U., Schmidt, H. H. H. W., Pollock, J. S., Hel-
Proc. Natl. Acad. Sci. USA 89 (1992)
26. 27. 28. 29. 30. 31. 32.
33. 34. 35. 36.
ler, M. & Murad, F. (1991) J. Cardiovasc. Pharmacol. 17, S57-S64. Forstermann, U., Gorsky, L. D., Pollock, J. S., Schmidt, H. H. H. W., Heller, M. & Murad, F. (1990) Biochem. Biophys. Res. Commun. 168, 727-732. Furchgott, R. F. & Zawadzki, J. V. (1980) Nature (London) 288, 373-376. Gonzalez, C. & Estrada, C. (1991) J. Cereb. Blood Flow Metab. 11, 366-370. Abman, S. H., Chatfield, B. A., Hall, S. L. & McMurtry, I. F. (1990) Am. J. Physiol. 259, H1921-H1927. Kelm, M. & Schrader, J. (1990) Circ. Res. 66, 1561-1575. Lassen, N. A., Ingvar, D. H. & Skinhoj, E. (1978) Sci. Am. 239, 62-71. Ko, K. R., Ngai, A. C. & Winn, H. R. (1990) Am. J. Physiol. 259, H1703-H1708. Yaksh, T. L., Wang, J. Y., Go, V. L. W. & Harty, G. J. (1987) J. Cereb. Blood Flow Metab. 7, 315-326. Fox, P. T. & Raichle, M. E. (1986) Proc. Natl. Acad. Sci. USA 83, 1140-1144. Fox, P. T., Raichle, M. E., Mintun, M. A. & Dence, C. (1988) Science 241, 462-464. Northington, F. J., Matherne, G. P., Coleman, S. D. & Berne, R. M. (1992) J. Cereb. Blood Flow Metab., in press. Ibayashi, S., Ngai, A. C., Howard, M. A., III, Meno, J. R., Mayberg, M. R. & Winn, H. R. (1991) J. Cereb. Blood Flow Metab. 11, 678-683. Van Wylen, D. G. L., Park, T. S., Rubio, R. & Berne, R. M. (1988) Am. J. Physiol. 255, H1211-H1218.