INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

OCULAR CIRCULATION plays an important role in maintaining the functions of neural, retinal, and other ocular tissues. The tissues responsible for controlling ocular pressure seem to be particularly sensitive to circulatory disturbance (8). Vascular tone in the eye is regulated by autonomic neural and humoral factors, as that in other organs and tissues. Sympathetic nerves are involved in the ocular vasoconstriction and increased vascular resistance, mainly by mediation of neurogenic norepinephrine and also by neuropeptide Y (17) or ATP (33). However, the roles of parasympathetic nerves in dilatating vasculature in the eye were not determined until recently.

The discovery of vasodilator innervation in dog and monkey cerebral arteries, in which nitric oxide (NO) acts as a neurotransmitter (28, 29), prompted us to reinvestigate autonomic innervation in ocular arteries. Functional and histological evidence of NO-mediated vasodilator nerves (nitroxidergic; see Ref. 31) have been reported in retinal, posterior ciliary, and ophthalmic arteries from monkeys (35, 36), dogs (25-27, 33), pigs (34), cattle (37), cats (10), and humans (13). In an earlier study (25), we have demonstrated that denervation of the pterygopalatine ganglion by injections of ethanol abolishes the perivascular neurons containing NO synthase in the ipsilateral middle cerebral arteries and the vasodilator response to electrical nerve stimulation of the isolated canine arteries. This denervation also abolishes the nicotine-induced relaxation of central retinal arteries (25), which is mediated by nerve-derived NO (30). Therefore, it is hypothesized that nitroxidergic neurons innervating the retinal artery are delivered from the parasympathetic, pterygopalatine ganglion. However, it awaits the direct evidence for the origin of vasodilator innervation in ocular arteries in vivo.

The present study was aimed to elucidate whether electrical stimulation of the pterygopalatine and geniculate ganglia dilates the ophthalmic artery of ipsilateral side in anesthetized monkeys and whether damage of the pterygopalatine ganglion constricts the artery for the determination of tonic innervation of this nerve. In addition, it was investigated whether the vasodilator responses to nerve stimulation by electrical pulses of the arteries in vivo and of those isolated from monkeys used for the in vivo study were mediated by NO.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The Animal Care and Use Committee at Shiga University of Medical Science approved the use of monkeys in this study.

In vivo study. Male and female Japanese monkeys (Macaca fuscata) weighing 6-10 kg were premedicated with intramuscular injections of 15 mg/kg ketamine and anesthetized intravenously with 20 mg/kg thiopental sodium. Stable anesthetic conditions were maintained by additional injections of thiopental as needed. The monkeys were intubated when needed but usually permitted to breathe spontaneously. PO₂ and PCO₂ were stable during the experiment. Arterial systolic and diastolic pressures were monitored with a pressure transducer (NEC San-ei, Tokyo, Japan) via a catheter inserted into the left femoral artery. The monkeys' body temperature was maintained at 37°C on a heated operating table. To make the right pterygopalatine ganglion visible, the zygomatic arch and underlying muscles were excised during microsurgery. Special care was taken to minimize bleeding when we reached deep into the fossa pterygopalatina. To reach the geniculate ganglion, a postauricular incision was made and the external auditory canal was cut and anteriorly retracted. After the temporal bone was removed with a cutting burr, the facial nerve was cut at the stylomastoid foramen and exposed along its course from the foramen to the geniculate ganglion. With the aid of a surgical microscope, a fine bipolar concentric stimulating electrode was inserted into the pterygopalatine or geniculate ganglion, and the electrode was then fixed by dental cement. The ganglion was stimulated by electrical pulses (1 ms duration and frequencies of 2, 5, and 10 Hz to the geniculate and 10 Hz to the pterygopalatine ganglion with 10 V intensity for 15 s); 5 s after the start of stimulation, the contrast medium for angiography was injected. Transfemoral internal carotid angiography was performed with a digital subtraction angiography system (DFA-3-30, Hitachi Medical, Tokyo, Japan) at the same magnification throughout the experiment. The contrast medium iopamidol (Iopamiron, 2 ml, Schering, Germany) was injected by an autoinjector (Angiomat 6000, Liebel-Flarsheim) connected to the angiography system in each angiography. The data obtained were stored in a digital data recorder (Hitachi Medical). The ophthalmic artery diameter was measured by the use of image analyzer included in the angiography system at two selected points. The two values were averaged, and the results were expressed as a percentage of the control artery diameter obtained before the electrical stimulation or the application of test drugs. In the preliminary study, diameters were measured six times with 1-h intervals; the mean values of the six measurements were 0.795 ± 0.026, 0.792 ± 0.027, 0.792 ± 0.027, 0.795 ± 0.026, 0.801 ± 0.019, and 0.810 ± 0.020 mm, respectively (n = 5). In the control series, arterial diameters before, during, and 10 min after the ganglionic stimulation were measured by angiography. The effects of treatment for 20 min with atropine and then phentolamine on the response to pterygopalatine ganglionic stimulation were determined. Except where otherwise mentioned, the studies were performed on monkeys treated with atropine and phentolamine. Modifications by N^G-nitro-L-arginine (L-NNA), L-arginine, and hexamethonium were compared in vasodilator responses to stimulation of the pterygopalatine and geniculate ganglia. In the experimental series, the arterial diameter was measured 30 and 60 min after intravenous L-NNA, 20 min after L-arginine, or 10 min after the damage of the ganglion by electric cauterization.

In vitro study. The monkeys from the in vivo study were placed under deep thiopental-induced anesthesia and were then euthanized by bleeding from carotid arteries. The eyeballs attached with optic nerves and extraocular tissues were then removed and stored overnight at 4°C. The ophthalmic arteries (0.5-0.7 mm outside diameter) were isolated and cut into helical strips ~15 mm in length, from which the endothelium was removed by gently rubbing the intimal surface with a cotton ball. The specimens were vertically fixed between hooks in a muscle bath (20 ml capacity) containing the modified Ringer-Locke solution, which was maintained at 37 ± 0.3°C and aerated with a mixture of 95% O₂-5% CO₂. The hook anchoring the upper end of the strips was connected to the lever of a force-displacement transducer (Nihon-Kohden Kogyo, Tokyo, Japan). The strips were placed between stimulating electrodes, and electrical pulses of 0.2 ms at a frequency of 5 Hz for 40 s were transmurally applied to stimulate perivascular nerves. Under these stimulus conditions, submaximal and reproducible relaxant responses were induced in monkey retinal arteries (33). The resting tension was adjusted to 1.0 g, which was optimal for inducing the maximal contraction. The composition of the bathing solution was as follows (in mM): 120 NaCl, 5.5 KCl, 2.2 CaCl₂, 1.0 MgCl₂, 25.0 NaHCO₃, and 5.6 dextrose. The pH of the solution was 7.38-7.44.

2

6

4

7

6

4

Statistics and drugs used. The results shown in the text and figures are expressed as means ± SE. Statistical analyses were made using the Student's paired and unpaired t-tests and the Tukey's test after one-way analysis of variance. Drugs used were L-NNA, N^G-nitro-D-arginine (D-NNA) (Peptide Institute, Minoh, Japan), L-arginine, nicotine (base), hexamethonium bromide (Nacalai Tesque, Kyoto, Japan), acetylcholine chloride (Daiichi Pharmaceutical, Tokyo, Japan), atropine sulfate (Tanabe, Osaka, Japan), physostigmine (eserine) sulfate (Sigma, St. Louis, MO), phentolamine mesylate (Novartis Japan, Tsukuba, Japan), prazosin hydrochloride (Wako, Osaka, Japan), prostaglandin F₂ (Pharmacia-Upjohn, Tokyo, Japan), tetrodotoxin (Sankyo, Tokyo, Japan) and papaverine hydrochloride (Dainippon, Osaka, Japan). 1H[1,2,4]oxadiazolol[4,3-]quinoxalin-1-one (ODQ) was a generous gift from Prof. S. Moncada (University College London, London, UK). Responses to NO were obtained by adding NaNO₂ solution, adjusted at pH 2 (6), just before the application, and the concentrations of NaNO₂ in the bathing media were expressed as those of NO.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vivo study. In anesthetized monkeys, electrical stimulation of the right pterygopalatine ganglion at a frequency of 10 Hz for 15 s dilated the ipsilateral ophthalmic artery, as shown in Fig. 1. The diameter of the artery and the response to electrical stimulation were not influenced by treatment with phentolamine (1 mg/kg iv, n = 4). Atropine (1 mg/kg iv) did not alter the arterial diameter under control conditions (0.804 ± 0.044 vs. 0.814 ± 0.042 mm, n = 5) but potentiated the response to nerve stimulation from 11.0 ± 1.8 to 13.6 ± 2.2%, (24.6 ± 5.2% increase, n = 5, P < 0.01, paired t-test). Atropine did not affect mean blood pressure (82.1 ± 4.3 vs. 82.9 ± 4.6 mmHg, n = 5, P > 0.05, paired t-test) and heart rate (152.6 ± 17.8 vs. 164 ± 21.7 beats/min, n = 5). Additional treatment with phentolamine (1 mg/kg iv) failed to alter the arterial diameter under basal conditions (0.789 ± 0.030 vs. 0.801 ± 0.023 mm, n = 5) and failed to alter the responses to nerve stimulation (16.4 ± 1.0 vs. 15.0 ± 1.2%, n = 5, P > 0.05). Phentolamine significantly decreased blood pressure from 81.3 ± 5.5 to 68.4 ± 3.9 mmHg (n = 5, P < 0.05, paired t-test) and increased heart rate from 159.3 ± 27.2 to 181 ± 28.2 beats/min (n = 5, P < 0.05, paired t-test). The remainder of this study was undertaken in the monkeys treated with atropine and phentolamine, unless otherwise mentioned.

View larger version (116K): [in this window] [in a new window]   Fig. 1.   Angiographic recordings of the right ophthalmic artery before ( , control) and during 10 Hz of electrical stimulation of the right pterygopalatine ganglion in an anesthetized Japanese monkey. Arrow denotes arterial dilatation induced by nerve stimulation.

View larger version (64K): [in this window] [in a new window]   Fig. 2.   Recordings of vasoconstriction by NG-nitro-L-arginine (L-NNA, 5 mg/kg iv) and its reversal by L-arginine (L-Arg, 500 mg/kg iv) in the ophthalmic artery () of an anesthetized monkey. Arrows denote arterial diameter changes in response to nerve stimulation.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Modification of the ophthalmic arterial diameter by intravenous injections of L-NNA (5 mg/kg) and L-Arg (500 mg/kg) or by damage (Denerv) of the pterygopalatine ganglion in anesthetized monkeys (n = 8). Data shown were obtained 30 and 60 min after L-NNA (), 20 min after L-Arg (), and 10 min after Denerv (×). Significantly different from control, aP < 0.01; significantly different from the value with L-Arg, bP < 0.01; significantly different from the value with Denerv, cP < 0.01, and dP < 0.05 (Tukey's test). Vertical bars represent means ± SE. Basal absolute values are 0.852 ± 0.032 mm (n = 8).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Frequency-vasodilatation relationship of the ophthalmic artery in response to geniculate ganglionic stimulation (2, 5, and 10 Hz) before and after L-NNA (5 mg/kg iv) and L-Arg (500 mg/kg iv) in anesthetized monkeys (n = 4 for 2 and 5 Hz, n = 5 for 10 Hz). Ordinate represents stimulation-induced increments in the arterial diameter relative to that before the stimulation. Significantly different from control, aP < 0.01; significantly different from the value with L-Arg, bP < 0.01 (Tukey's test). Vertical bars represent means ± SE. Basal absolute values for control at 2, 5, and 10 Hz are 0.828 ± 0.029 (n = 4), 0.828 ± 0.029 (n = 4), and 0.788 ± 0.045 mm (n = 5), respectively. Basal absolute values under treatment with L-NNA at 2, 5, and 10 Hz are 0.765 ± 0.032 (n = 4), 0.765 ± 0.032 (n = 4), and 0.725 ± 0.047 (n = 5) mm, respectively. Basal absolute values under treatment with L-NNA +L-Arg at 2, 5, and 10 Hz are 0.808 ± 0.043 (n = 4), 0.808 ± 0.043 (n = 4), and 0.766 ± 0.053 (n = 5) mm, respectively.

In vitro study. The monkeys used for the in vivo study were given additional injections of thiopental and were then euthanized by bleeding after the experiment was finished. The eyeballs of nonoperated side, attached to the optic nerve, was stored overnight for this study. In helical strips of the ophthalmic artery denuded of the endothelium and partially contracted with PGF₂, transmural electrical stimulation at 5 Hz produced reproducible relaxations, which were abolished by tetrodotoxin (3 × 10⁷ M). Treatment with L-NNA (10⁶ M) abolished the relaxant response in five strips and reversed to a slight contraction in the remaining three, which was suppressed by treatment with prazosin (10⁶ M). Additional treatment with L-arginine (3 × 10⁴ M) restored the relaxation to nerve stimulation (Fig. 5). Typical responses as affected by L-NNA, D-NNA, L-arginine, and tetrodotoxin are illustrated in Fig. 6A. The neurogenic relaxation was abolished by treatment with ODQ, a soluble guanylate cyclase inhibitor (10⁶ M, n = 4). Effects of the inhibitors used are quantitatively compared in Fig. 5. D-NNA (10⁶ M) did not affect the relaxation induced by nerve stimulation; mean values before and after the treatment were 21.4 ± 4.9 and 22.8 ± 4.1% (n = 5), respectively.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Quantitative data concerning the effect of L-NNA (106 M), L-Arg (3 × 104 M), and 1H[1,2,4]oxadiazolol[4,3-] quinoxalin-1-one (ODQ, 106 M) on the relaxant response to 5 Hz transmural electrical stimulation (TES) of ophthalmic arterial strips denuded of the endothelium and partially contracted with PGF2. Ordinate represents the response to nerve stimulation relative to that induced by 104 M papaverine. Significantly different from control (C), aP < 0.01; significantly different from the value with L-Arg, bP < 0.01. n = 5 strips from separate monkeys. Vertical bars represent means ± SE. Basal absolute values for control at 2, 5, and 10 Hz are 0.83 ± 0.03 (n = 4), 0.83 ± 0.03 (n = 4), and 0.79 ± 0.05 mm (n = 5), respectively.

View larger version (10K): [in this window] [in a new window]   Fig. 6.   Relaxant responses to transmural electrical stimulation (5 Hz) of an ophthalmic arterial strip denuded of the endothelium before and after treatment with L-NNA (106 M), L-Arg (3 × 104 M), ODQ (106 M), and tetrodotoxin (TTX, 3 × 107 M). The strip was partially contacted with PGF2. After the first series of experiment was over (A), the preparation was repeatedly rinsed and equilibrated, and the second series (B) was performed. PA denotes 104 M papaverine that produced the maximal relaxation.

7

7

6

4

7

6

2

6

5

6

3

7

6

6

View larger version (9K): [in this window] [in a new window]   Fig. 7.   Typical responses to nicotine (104 M) and nitric oxide (NO) (107 and 106 M) of a monkey ophthalmic arterial strip denuded of the endothelium before and after treatment with L-NNA (106 M) and L-Arg (3 × 104 M). The strip was partially contracted with PGF2. PA denotes 104 M papaverine that produced the maximal relaxation.

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Modification by L-NNA (106 M) and then L-Arg (3 × 104 M) of the response to nicotine (104 M, A) and NO (107 M, B) in monkey ophthalmic arterial strips denuded of the endothelium and partially contacted with PGF2. The ordinate represents agonist-induced relaxations relative to those elicited by 104 M papaverine. Significantly different from control (C), aP < 0.01; significantly different from the value with L-Arg, bP < 0.01 (Tukey's test). n = 5 strips from separate monkeys. Vertical bars represent means ± SE.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study revealed that electrical stimulation of the pterygopalatine ganglion-induced vasodilatation of the ipsilateral ophthalmic artery, which was abolished by intravenous injections of L-NNA and restored by L-arginine in anesthetized monkeys. Our previous study (38) has demonstrated the presence of abundant nerve cells, bundles, and fibers containing NO synthase immunoreactivity in monkey pterygopalatine ganglion. The ophthalmic artery isolated from monkeys, in which the vasodilator response had been elicited in vivo by the ganglionic stimulation, responded to transmural electrical stimulation with frequency-related relaxations that were suppressed by L-NNA but not by D-NNA and abolished by ODQ, a guanylate cyclase inhibitor (7). The L-NNA-induced inhibition was reversed by L-arginine. Relaxations induced by nicotine that chemically stimulates perivascular nerve terminals to liberate neurotransmitters (12, 20, 28) were also abolished by L-NNA and ODQ. Similar results with the electrical and chemical stimulation have also been obtained in canine, monkey, porcine, and human cerebral arteries (9, 21, 28, 29) and canine, monkey, and porcine retinal or ciliary arteries (25, 34, 35). In addition, release of NO from isolated, superfused canine cerebral arteries into superfusate in response to transmural electrical stimulation or nicotine (28) and increase in cGMP content in the tissue by nicotine (30) are reportedly abolished by treatment with NO synthase inhibitors and tetrodotoxin (for electrical stimulation) or hexamethonium (for nicotine). These findings support the hypothesis that NO plays a crucial role as a neurotransmitter in the vasodilator nerve innervating monkey ophthalmic arteries as well as cerebral arteries from various mammals (32). In ipsilateral retinal arteries isolated from dogs in which the unilateral pterygopalatine ganglion is degenerated, the relaxant response to nicotine is abolished, whereas the response is unaffected in the contralateral arteries, suggesting the nitroxidergic vasodilator innervation of ocular arteries from the ipsilateral pterygopalatine ganglion in dogs (25). The present study with anesthetized monkeys provided a direct evidence that the same ganglion projects the nitroxidergic nerve to ophthalmic arteries. Influence of anesthetics could not be ruled out in the present study in vivo that was performed only under persistent anesthesia.

Unilateral denervation of the monkey pterygopalatine ganglion decreased the diameter of ipsilateral ophthalmic arteries. In anesthetized dogs and monkeys, intracisternal injections of L-NNA constrict the basilar artery, and the effect is reversed by L-arginine (14, 23). The vasoconstrictor action of L-NNA is significantly attenuated by treatment with hexamethonium, suggesting that neurogenic NO continuously released under resting conditions is involved in the basilar arterial dilatation (14, 23). These findings strongly suggest that nitroxidergic tonic discharges from the vasomotor center contribute to the maintenance of ocular and cerebral arterial dilatation and of decreased arterial resistance. In in vivo studies on rat, mouse, goat, and pig pial arteries and arterioles or cerebral vascular resistance, the intravenous injection or topical application of NO synthase inhibitors produces vasoconstriction or decreases blood flow (1, 3-5, 18). Suppression of basal release of NO from the endothelium due to the NO synthesis inhibition is considered to be involved in the response. In the present study, L-NNA intravenously injected also produced vasoconstriction (about 16% of the diameter of preinjection control, Fig. 3) of the ophthalmic artery, and L-arginine reversed the action. Cauterization of the pterygopalatine ganglion constricted the artery by an average value of 9%. Therefore, under the experimental conditions used, it is postulated that vasodilatation is partly mediated by NO synthesized from L-arginine in perivascular nerve terminals innervating the ophthalmic artery, and the remaining vasodilatation is associated with NO liberated from extraneuronal tissues including the endothelium. One may argue the influence of autoregulation in the change of the arterial diameter, because intravenous administration of L-NNA elevates the blood pressure (present study, 15). However, phentolamine did not change the arterial diameter despite a significant fall of systemic blood pressure, suggesting that autoregulation in ocular circulation is, if any, minimal under the experimental conditions used. Intracisternal injections of L-NNA in anesthetized monkeys constrict basilar arteries without any change in blood pressure, and the response was blunted under treatment with hexamethonium (14). Therefore, it is concluded that ophthalmic arterial constrictions induced by intravenous L-NNA are associated at least mainly with suppression of nitroxidergic function, although the autoregulatory influence cannot completely be excluded.

Stimulation of the geniculate ganglion also dilated the ipsilateral ophthalmic artery to a similar extent to that of the pterygopalatine ganglion. Responses to geniculate ganglionic stimulation, but not those to stimulation of the pterygopalatine ganglion, were abolished by hexamethonium, whereas those to stimulation of these ganglia were suppressed by L-NNA. These findings support the hypothesis that cholinergic preganglionic neurons from the brain stem via the geniculate ganglion and greater petrosal nerve innervate the pterygopalatine ganglion.

Atropine failed to alter the arterial tone under resting conditions but potentiated the response to vasodilator nerve stimulation. In isolated monkey, bovine, and porcine cerebral and ocular arteries (2, 22, 24, 34, 36), relaxations induced by electrical nerve stimulation are attenuated by treatment with acetylcholine and physostigmine but potentiated by atropine. We hypothesized that acetylcholine from cholinergic nerves acts on prejunctional muscarinic receptors, possibly the M₂ subtype (24), and inhibits the synthesis and release of NO from vasodilator nerves. This is also true in the case of isolated monkey ophthalmic arteries. In addition, the present study revealed evidences supporting the hypothesis that the prejunctional inhibition by neurogenic acetylcholine is operative in monkey ophthalmic arteries in vivo. The reason why the potentiation by atropine of the ophthalmic arterial response was seen only when the nerve was stimulated may be because the liberation of acetylcholine from cholinergic nerves in vivo under resting conditions is not sufficient to induce prejunctional inhibition.

In conclusion, the monkey ophthalmic artery is innervated in vivo by vasodilator nerve in which NO acts as a neurotransmitter; the nerve is originated from the ipsilateral pterygopalatine ganglion that receives the projection of cholinergic nerves from the brain stem. Our recent study has demonstrated that NO per se, but not stable analogs of NO such as S-nitrosothioles (11), is involved in the neurotransmission in monkey cerebral arteries (16, 19). NO derived from the nerve would participate importantly in the arterial dilatation under resting conditions and also when nitroxidergic nerves are activated by impulses from the vasomotor center. The nitroxidergic nerve function appears to be negatively controlled by cholinergic nerve activation.