Excitation of cells in the rostral medial medulla of the rat by the nitric oxidecyclic guanosine monophosphate messenger system

ELSEVIER

Neuroscience Letters 195 (1995)155-158

HEUROSCltNCE LHTERS

Excitation of cells in the rostral medial medulla of the rat by the nitric oxidecyclic guanosine monophosphate messenger system Ian D. Hentall* Department of Biomedical Sciences, University of lllinois College of Medicine, Rockford, IL 61107-1897, USA Received 27 March 1995; revised version received 26 June 1995; accepted 28 June 1995

Abstract

Analgesia has been reported to be facilitated by supraspinal nitric oxide (NO) and cyclic guanosine monophosphate (cGMP). In the rostromedial medulla, an important pain-suppressing region, iontophoretically delivered 8-bromo-cGMP excited most single recorded cells (9/10), and methylene blue (a guanylyl cyclase inhibitor) inhibited all cells (7/7). Nitrite and ferrous ions together, shown voltammetrically ex vivo to yield nitric oxide (NO), excited some cells (14/28) and inhibited others (7/28). Methylene blue blocked excitation (3/3) but not inhibition (4/4) by the putative NO. Spontaneous or glutamate-evoked firing was gradually inhibited (23/32) or unaffected by N~°-nitro-L-arginine(a NO synthase inhibitor), but was mostly inhibited by L-arginine (the NO precursor) (23/26), although a rapid onset militated against elevated NO production. These substances, excepting L-arginine, produced changes consistent with an excitatory cGMP-NO cascade contributing to analgesia.

Keywords:L-Arginine;Guanylyl cyclase; Iron; Methylene blue; Nitrite; Raphe.

Nitric oxide (NO) has been suggested to have an antinociceptive influence at the supraspinal level, because Larginine (the precursor of NO) and various blockers of NO synthesis alter pain reflexes when administered by the intracerebroventricular route [1,7,11,13]. A supraspinal antinociceptive action has also been proposed for cyclic guanosine monophosphate (cGMP) [15], whose synthesis in the brain is stimulated by NO. A possible neurophysiological basis for these phenomena is examined here in one of the best characterized of pain-modulating brainstem regions, comprising the nucleus raphe magnus and adjacent areas of the rostral medial medulla (RMM). In the RMM, electrical stimulation or microinjection of glutamate produces analgesia [6,16], and the brain isoform of NO synthase is prominent [14]. Single RMM cells were recorded during iontophoresis of substances which can intervene in the NO-cGMP system. These included the NO synthase inhibitor N~°-nitro-Larginine (L-NNA), the natural substrate for NO synthase L-arginine, a combination of nitrite and ferrous ions (which were shown here to produce NO in situ), the * Corresponding author, Tel.: +1 815 3955679; Fax: +1 815 3955666; E-mail: [email protected].

guanylyl cyclase inhibitor methylene blue, and the membrane-soluble cGMP agonist analog 8-bromo-cGMP (8Br-cGMP). D-Arginine (not a NO precursor), and separately applied ferrous and nitrite ions, were also examined. Rats (female, Sprague-Dawley, 250-300g, n = 18) were anesthetized with pentobarbital, initially intramuscular (60 mg/kg), supplemented thereafter by continuous intravenous (jugular) infusion (15 mg kg -1 h-l). The trachea was intubated. The animals were mounted in a stereotaxic holder. Rectal temperature was kept at 37°C by a thermostatic blanket. A craniectomy allowed dorsal access to the medial medulla. For recording and drug delivery, three- or five-barrel micropipettes were used. The recording barrel (4-7 Mr2) was filled with 2 M sodium acetate in 5% fast green. Fast green was deposited electrophoretically (-2 to -4 pA for 20 min) to mark the end-point of penetrations. A barrel filled with 2 M NaCI functioned as a floating return for iontophoresis current. Other barrels contained various combinations of the following: NaNO2 (0.1 M, pH 7.4) always paired with FeSO4 (0.1 M, p H 3.5), methylene blue (0.1 M, pH 3.9), 8-Br-cGMP (0.5 M, pH 8), L-arginine (0.1 M, pH 6), L-NNA (0.01 M, pH4), D-arginine

0304-3940/95/$09.50 © 1995 Elsevier Science Ireland Ltd. All fights reserved SSD! 0 3 0 4 - 3 9 4 0 ( 9 5 ) 1 1802-Z

LD. Hentall / Neuroscience Letters 195 (1995) 155-158

156 0.02

=~ v

~

0.015

m

~

0.01

E

•

g

o.0o5

0

-0.005

750

~

~

I

I

I

800

850

900

950

1000

polanzation voltage (mV) Fig. 1. Ex vivo detection of NO. These curves show the steady-state current (positive values for oxidation) produced at different electrode potentials when ferrous ions and nitrite ions were ejected separatelyor together. A 20 nA iontophoresis current of appropriate polarity was used for ejection, and opposite, 20 nA holding currents were otherwise applied. The peak near +950 mV closely matches the standard oxidation potential at 25°C and 1 atmosphere for NO.

(0.1 M, pH 6), and L-glutamate (0.1 M, pH 8). Retaining currents (of appropriate polarity) were _+5 nA for all drugs except -e20 nA for FeSO4 and NaNO2. All substances were obtained from Sigma Chemical Corp. Currents delivered to the FeSO4 and NaNO2 barrels always summed to zero. Ex vivo qualitative tests of NO production by simultaneous ejection of ferrous and nitrite ions (Fe/NO2) were performed in 0.9% NaCl at 20°C. The expected reaction is Fe 2+ + NO2- + 2H + = NO + Fe 3+ + H20 Redox currents were measured by DC differential voltammetry, using an amperometric amplifier (A-M systems, model 1900) and Teflon-insulated platinum-wire electrodes (0.05 mm uninsulated diameter). The measurement electrode was located roughly 0.02 mm from the tip of the iontophoresis micropipette, separated by a thin Nation membrane. This membrane, which impedes ion movement while allowing passage of NO [10], was roughly I mm in diameter and 10/zm thick, and was made by evaporating a film of 5% Nation solution (Aldrich Chemical Co.) over a hole drilled in a plastic cover slip. Polarization voltages were applied to both the measuring electrode and an identically constructed reference electrode positioned 1 mm away. A chloridized silver wire provided a distal ground point. NO forms the half-cell NOINO3-1H ÷ at about +0.95 V with respect to the standard hydrogen electrode [10]. Ejection of nitrite ions (-20 nA, 1 min) increased the DC oxidation current measured between +0.8 and +1.0 V (Fig. 1), but did not establish a maximum, sug-

gesting that there was no single predominant electrochemical reaction. Ejection of ferrous ions (-20 nA, 1 min) caused a small decrease in currents measured in this voltage range, consistent with their reducing properties. Ejection of both ions (+20 nA, 1 min) led to currents which were more than additive and peaked near the oxidation potential for NO. Firing rates were tallied initially in consecutive 2-s bins. Two periods of 20 s were compared, one just prior to drug delivery and the other centered on the maximum or minimum in the first minute after the start of drug application. A given cell's response to each drug was evaluated by the first application of that drug, categorized into inhibition, excitation or no change by a two-sided ttest (P < 0.05). All neurons were located histologically by the fast green dye marks within 0.5 mm of the midline, in or near the nucleus raphe magnus. (The brain stem was fixed after experiments by intracardiac perfusion of 10% formaldehyde; coronal sections were cut at 50/zm width on a vibrating microtome and counter-stained with cresyl violet.) Cells with spontaneous activity > 5 Hz were selected for testing; slower cells were rejected because their inhibition is harder to show. Cells were either inhibited (n = 54), excited (n = 13), excited then inhibited (n = 3), or unchanged (n = 3) by a noxious mechanical cutaneous stimulus; these are also known respectively as off-cells, on-cells, on-off cells and neutral cells [5]. The criterion of higher spontaneous activity may have selected a disproportionate number of off-cells [5]. A particular qualitative response to each substance was seen in both on-cells and off-cells, but this did not rule out statistical differences between these categories in the probable direction or magnitude of responses. The response to Fe/NO2 (_+20nA) was excitation (n = 14; e.g. Fig. 2A), inhibition (n = 7; e.g. Fig. 2B) or no change (n = 7). The mean first response was 1.54 (_+0.28 standard error) of baseline. Nitrite ions given alone (20 nA, n = 8) slightly depressed activity to 0.92 (-+ 0.05 standard error) of baseline. Ferrous ions given alone (20nA, n = 9 ) slightly increased activity to 1.18 (-+ 0.07 standard error) of baseline. Changes typically disappeared within 1 min of restoring the holding currents. By itself, methylene blue produced a slow reduction of spontaneous activity (0.82 _+0.08 after 1 min with 20 nA, n = 7; e.g. Fig. 2A,B). Methylene blue at a lower dose (2 nA) was tested on three neurons that were significantly excited by in situ synthesis of NO (-+20 nA); in all three cases, a partial blocking of the excitation was observed (e.g. Fig. 2A). In four other neurons, a fall in activity produced by Fe/NO2, inspected by averaging repeated trials, was not seen to be blocked by methylene blue at currents as high as 20 nA; the absolute fall from pre-treatment activity remained about the same and the relative fall increased (e.g. Fig. 2B). This frequency-distribution of

I.D. HentaU I Neuroscience Letters 195 (1995) 155-158

NO, 20 nA

NO, 20 nA

NO, 20 nA

8-Br-cGMP, 80 nA

NO, 20 nA

30

C

30.

A

25

157

25.

m

8-B~-~MP 60 nA

8-Br-cGMP 20 nA

8-Br-cGMP 40 nA

20.

~" 20 ~15

~15

~ lO

~ lO 5

6

D

B

7¸

40nA

¸

t,~i

•~. 4 •

~3.

, . -

1

0.

20hA

10 hA, L-NNA

10nAl~lnA, L-Arg I (

,

.

;iii~V'l'~i.:l

._>

[1

I" :'iii ~'"

- ....

nA

~, 40 ]

after

~30t

)

time scale: 100 s

2°1

'

o. oo,.,os

'°oI

Fig. 2. Iontophoresis of substances affecting the NO-cGMP pathway. (A) Partial block of the excitatory effect of exogenous NO by methylene blue (MB). Continuous firing rate is displayed in 4-s bins; heavy bars indicate treatment periods. (B) Inhibition with NO, showing a lack of blocking by MB. Each trace connects points representing the average activity in 2-s bins from five trials out of 15. NO was applied at 2-rain intervals and methylene blue was added during the 6th to 10th trials. (C) Effect of 8-Br-cGMP. The bars indicate 10-s periods of iontophoresis; activity is expressed in 2-s bins. (D) Effects of L-arginine (L-Arg) and L-NNA on glutamate-evoked activity. The top three traces indicate (by upward deflections) iontophoretic application of the test drugs and glutamate; activity is expressed in 2-s bins.

methylene blue's action (reversals or non-reversals) on the excitatory or inhibitory influences of Fe/NO 2 (in a 2 x 2 contingency table) was significant (P=0.029, Fisher exact test). Iontophoresis of 8-Br-cGMP led to significant excitation of nine out of ten cells (e.g. Fig. 2C). The effect was seen with repeated trials of currents in the range of -20 to -80 nA; the remaining cell was unaffected at these at currents. L-Arginine (40 nA) inhibited 13 neurons out of 16 tested, and reduced glutamate-evoked activity in all ten neurons tested (Fig. 2D). D-Arginine also inhibited spontaneous activity, but with a considerably weaker effect. For example, at a 20 nA ejection current, D-arginine reduced activity to 82% _+3.5 (mean + standard error, n = 6), whereas L-arginine reduced activity to 67% _+5.0 (n = 16). L-NNA (40 nA) inhibited 16 of 25 cells tested and excited one cell; it also produced a consistent partial reversal of glutamate-induced excitation (n = 7; e.g. Fig. 2D). The time-course of the inhibitory effect of L-NNA was slow in onset and recovery (>1 min), consistent with the time needed for intracellular uptake and disappearance of previously synthesized NO. The rapidity of L-arginine's inhibitory action suggests that it did not involve increased NO synthesis following intracellular uptake, as also sug-

gested by the inhibitory influence of the enzymatically inactive D-arginine. Joint iontophoresis of nitrite and ferrous ions was employed to circumvent problems with application of NO to deep brain tissue. NO itself has a half-life in tissue of 35 s, and is presumably unstable when held for injection in a microneedle. NO donors have been developed for clinical slow release of NO, and are likely to disappear from a microinjection site before significant NO production. Furthermore, the commonly used inorganic NO donor, sodium nitroferricyanide, may influence some types of neuronal activity via mechanisms unrelated to NO production [2,8], and this is an untested possibility for other NO donors. The new technique shares with all iontophoretic procedures the difficulty of reproducible quantification due to variability between pipettes and to the inaccessibility of the time-dependent spatial gradient of ejected substances. An added complication is that the product of interest (NO) must reach peak concentration at some distance from the micropipette tip. Thus ex vivo DC polarographic measurement of NO synthesis was useful only as a qualitative validation of the method. Anodal current applied to the ferrous sulfate (pH 3.5) barrel alone served as a control for both hydrogen ions and ferrous ions under the experimental conditions. Hydrogen ions cannot explain both the weak excitation with

158

I.D. Hentall I Neuroscience Letters 195 (1995) 155-158

ferrous ions and the inhibitory influences o f methylene blue (pH 3.9) and L-NNA (pH 4). The principal pathophysiological effect of ferrous ions and ferric ions is lipid peroxidation in the presence of ascorbate, which is thought to occur following reduction to chelatable molecular iron [4]. But excitation with Fe/NO2 cannot be accounted for by lipid peroxidation, since it was produced in a milieu which favored conversion to ferric ions, stopped soon after cessation o f iontophoresis (indicating no chronic damage), and was blocked by methylene blue. Furthermore, lipid peroxidation is associated with depression of firing [12]. The CNS effects of NO are incompletely understood. Some facilitatory and inhibitory effects are mediated through increased synthesis o f cytoplasmic c G M P [3]. NO may also act independently of cGMP, possibly via Snitrosothiol intermediates, to depress the sensitivity of N M D A receptors at a redox-modulatory site [9]. The latter mechanism might account for the minority of neurons in which Fe/NO 2 caused an inhibition not blocked by methylene blue. Nitrergic inhibition could be more frequent in R M M cells of a particular somatosensory class, but this preliminary survey did not include enough neurons from all classes to decide this issue. Alternatively, Fe/NO2 may have excited unobserved interneurons, in a process not reversible by methylene blue, which then inhibited the recorded cell. The R M M thus appears to possess NO-induced excitation mediated by guanylyl cyclase. This is consistent with the excitatory effect of 8-Br-cGMP and with the inhibitory effects of L-NNA and methylene blue, but not with the inhibitory effect o f L-arginine. L-Arginine's dominant action could occur through a cellular mechanism other than stimulated NO production, which might explain the dual effect of L-arginine on antinociceptive processing [7] and the present results with D-arginine. In conclusion, the supraspinal antinociceptive action ascribed to NO and c G M P by previous studies in behavioral pharmacology [1,7,11,13] could at least partly be based on an excitatory N O - c G M P cascade in the RMM, but the role of other pain-modulating regions also requires clarification.

This work was supported by PHS grant NS26116. The technical assistance of Marcia Andresen is gratefully acknowledged.

[1] Duarte, I.D.G. and Ferreira, S.H., The molecular mechanism of central analgesia induced by morphine or carbachol and the Larginine-nitric oxide-cGMP pathway, Eur. J. Pharmacol., 221 (1992) 171-174. [2] East, S.J., Batchelor, A.M. and Garthwaite, J., Selective blockade of N-methyl-D-aspartatereceptor function by the nitric oxide donor, nitroprusside, Eur. J. Pharmacol., 209 (1991) 119-121. [3] Garthwaite, J., Glutamate, nitric oxide and cell-cell signaling in the nervous system, Trends Neurosci., 14 (1991) 60--67. [4] Gutteridge, J.M.C., Iron and oxygen radicals in brain, Ann. Neurol., 32 (1992) S16-$21. 15] Hentall, I.D., Andresen, M.J. and Taguchi, K., Serotonergic, cholinergic and nociceptive inhibition or excitation of raphe magnus neurons in barbiturate-anesthetized rats, Neuroscience, 52 (1993) 303-310. [6] Hentall, I.D., Barbaro, N.M. and Fields, H.L., Spatial and temporal variation of microstimulation thresholds for inhibiting the tailflick reflex from the rat's rostral medial medulla, Brain Res., 548 (1991) 156-162. [7] Kawabata, A., Umeda, N. and Takagi, H., L-Arginine exerts a dual role in nociceptive processing in the brain: involvement of the kyotorphin-metenkephalin pathway and NO-cyclic GMP pathway, Br. J. Pharmacol., 109 (1993) 73-79. [8] Kiedrowski, L., Costa, E. and Wroblewski, J.T., Sodium nitroprusside inhibits N-methyl-D-aspartate-evokedcalcium influx via a nitric oxide- and cGMP-independent mechanism, Mol. Pharmacot., 41 (1992) 779-784. [9] Lei, S.Z., Pan, Z.-H., Aggarwal, S.K., Chen, H.-S.V., Hartman, J., Sucher, N.J. and Lipton, S.A., Effect of nitric oxide production on the redox modulatory site of the NMDA receptor-channel complex, Neuron, 8 (1992) 1087-1099. [10] Malinski, T. and Taha, Z., Nitric oxide release from a single cell measured in situ by a porphyrinic-based microsensor, Nature, 358 (1992) 676-677. [11] McDonald, C.E., Gagnon, M.C., Ellenberger, E.A., Hodges, B.L., Ream, J.K., Tousman, S.A. and Quock, R.M., Inhibitors of nitric oxide synthesis antagonize nitrous oxide antinociception in mice and rats, J. Pharmacol. Exp. Ther., 269 (1994) 601-608. [12] Sharma, D., Maurya, A.K. and Singh, R., Age-related decline in multiple unit action potentials of CA3 region of rat hippocampus correlation with lipid peroxidation and tipofuscin concentration and the effect of centrophenoxine, Neurobiol. Aging, 14 (1993) 319-330. [13] Tseng, L.F., Xu, J.Y. and Pieper, G.M., Increase of nitric oxide production by L-arginine potentiates i.c.v, administered fl-endorphin-induced antinociception in the mouse, Eur. J. Pharmacol., 212 (1992) 301-303. [14] Vincent, S.R. and Kimura, H., Histochemical mapping of nitric oxide synthase in the rat brain, Neuroscience, 46 (1992) 755-784. [15] Vocci, F.J., Petty, S.K. and Dewey, W.L., Antinociceptive action of the butyryl derivatives of cyclic guanosine 3':5'-monophosphate, J. Pharmacol. Exp. Ther., 207 (1978) 892-898. [16] Zhuo, M. and Gebhart, G.F., Characterization of descending facilitation and inhibition of spinal nociceptive transmission from the nuclei reticularis gigantocellularis pars alpha in the rat, J. Neurophysiol., 67 (1992) 1599-1614. -