| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Pharmacology, Shiga University of Medical Science, Ohtsu, Japan (N.T., T.O.); and Department of Toyama Institute for Cardiovascular Pharmacology Research, Chuo-ku, Osaka, Japan (N.T.)
Abstract I. Introduction II. Discovery of Nitrergic (Nitroxidergic) Nerve in Blood Vessels A. From Nonadrenergic Noncholinergic Nerve to Nitrergic Nerve B. Evidence for the Presence of Vasodilator Nerve-Releasing Nitric Oxide As a Neurotransmitter—Criteria for Transmitter C. Mechanism of Nitric Oxide Formation and Action 1. Formation of Nitrergic Neurotransmitter. a. L-Arginine. b. Ca2+ and Calmodulin. 2. Action of Nerve-Derived Nitric Oxide on Vascular Smooth Muscle. III. Nitrergic Innervation in Intra- and Extracranial Vasculature A. Cerebral Artery 1. In Vitro Studies in Various Mammals. 2. Nerve Stimulation by Electrical Pulses and by Nicotine and Related Compounds. 3. Is Protein Phosphorylation Involved in Neuronal Nitric-Oxide Synthase Activation? 4. In Vivo Studies. 5. Tracing the Origin of Nitrergic Nerve. 6. Hypercapnic and Hypoxic Cerebroarterial Dilation and Hypothermia. a. Hypercapnia. b. Hypoxia. c. Hypothermia. 7. Autoregulation. 8. Prejunctional Modulation of Nitrergic Nerve Function by Cholinergic and Adrenergic Neurotransmitters. 9. Histochemical Studies of Neurons Containing Nitric-Oxide Synthase. B. Ocular Vasculature 1. Retinal Artery and Arteriole. 2. Ciliary Artery. 3. Ophthalmic Artery. C. Lingual Artery D. Nasal Vasculature E. Temporal Vasculature IV. Nitrergic Innervation in Blood Vessels of Viscera A. Coronary Artery B. Pulmonary Vasculature C. Digestive Tract Vasculature D. Renal Vasculature E. Uterine Vasculature F. Penile Artery and Vein V. Nitrergic Innervation in Blood Vessels of Skin and Skeletal Muscle A. Cutaneous Small Artery B. Skeletal Muscle Vasculature VI. Interaction of Nitrergic, Cholinergic, and Adrenergic Nerves in Peripheral Vasculature VII. Nitrergic Innervation of Corpus Cavernosum and Penile Erection A. In Vitro Studies B. In Vivo Studies and Penile Erection C. Histochemical Studies of Neurons Containing Nitric-Oxide Synthase VIII. Blood Pressure Control by Neurogenic Nitric Oxide A. Involvement of Nitrergic Nerve Innervating Vasculature B. Effect of Centrally Applied Nitric Oxide Donors and Nitric-Oxide Synthase Inhibitors C. Nitric-Oxide Synthase Knockout Mice IX. Acupuncture, Axon Reflex, and Neurogenic Inflammation X. Pathological Implications of Neurogenic Nitric Oxide A. Cerebral Vasospasm after Subarachnoid Hemorrhage B. Migraine and Cluster Headache C. Impaired Ocular Circulation: Relation to Glaucoma D. Pre-Eclampsia (Pregnant Intoxication) E. Hypertension F. Erectile Dysfunction XI. Pharmacological Implications of Neurogenic Nitric Oxide A. Phosphodiesterase Type 5 Inhibitors B. Ginsenosides C. Free Radical Scavengers D. {alpha}2-Adrenoceptor Antagonists E. Antimuscarinic Agents F. Neuronal Nitric-Oxide Synthase Inhibitors G. Compounds That Suppress the Action of Endogenous Nitric-Oxide Synthase Inhibitors XII. A Possible Reason for Predominant Nitrergic Nerve Function in the Cerebral Artery and Corpus Cavernosum Compared with the Peripheral Vasculature XIII. A Proposal for a New Classification of Efferent Parasympathetic Innervation in Vascular and Nonvascular Smooth Muscle
 |
Abstract |
|---|
 Top Next |
|---|
|
I. Introduction |
|---|
|
TopPreviousNext |
|---|
; Lundberg, 1981; Owman, 1988). However, most of the NANC inhibitory neurotransmitters were not identified for many decades. The discovery of nitric oxide (NO) as a novel mediator of intercellular signal transmission (Ignarro et al., 1987; Furchgott, 1988; Moncada et al., 1988a), the determination of the NO-synthesizing process (Moncada et al., 1988b; Palmer et al., 1988a), discovery of NO synthase (NOS) inhibitors (Palmer et al., 1988b) and isolation of neuronal NOS (nNOS; Bredt and Snyder, 1990) that enabled the production of nNOS antiserum represented a major breakthrough toward understanding the mechanisms underlying inhibitory responses to NANC nerve activation. Putative neurotransmitters of peptides, amino acids, amines, acetylcholine, and nucleotides have been reevaluated, and some of them have been confirmed to be true, whereas others have required reconsideration. Despite the fact that there are still unidentified neurotransmitters, the discovery of nerves liberating NO that mediates the inhibitory response is an undoubtedly epoch-making discovery.
NANC vasodilator nerves were first discovered in dog cerebral arteries (Toda, 1975), and 15 years later, a hypothesis that NO acts as a neurotransmitter was raised (Toda and Okamura, 1990a,b). Similar conclusions were also reached by independent groups using the canine ileocolonic junction (Bult et al., 1990) and duodenum (Toda et al., 1990a), bovine retractor penis muscle (Gillepsie and Sheng, 1990), rat anococcygeus muscle (Gillepsie et al., 1989; Li and Rand, 1989) and guinea pig tracheal muscle (Tucker et al., 1990). These discoveries led us to consider ideas that defied conventional dogma that neurotransmitters are organic, relatively high molecular weight molecules and that the neurotransmission process involves neurotransmitters that are stored in nerve terminals, and upon their release would interact with membrane receptive sites. On the basis of long-term investigations in not only vasculature but also other smooth muscle on NO-mediated neurogenic response, the hypothesis of NO neurotransmission is now widely accepted as an important control mechanism in functions of autonomically innervated organs and tissues. The nerve whose transmitter function depends on the release of NO is called "nitroxidergic" (Toda et al., 1991b; Toda and Okamura, 1991a, 1992c) or "nitrergic" (Rand, 1992). However, the NO Nomenclature Committee of the International Union of Pharmacology (chairman: Paul M. Vanhoutte) has chosen "nitrergic" as the official name (Moncada et al., 1997).
Histological studies have revealed that vascular smooth muscle is innervated by neurons containing NOS immunoreactivity (Bredt et al., 1990), as well as by those containing norepinephrine/tyrosine hydroxylase and cholinesterase/choline acetyltransferase. Functionally, nitrergic nerves would be more important in vasculature than cholinergic nerves, which play only a role in modulating adrenergic and nitrergic nerve functions. The cerebral artery and corpus cavernosum have unique characteristics of an intense nitrergic dilatation with a minimal adrenergic contraction. Therefore, the role of an intercellular messenger NO in the peripheral nervous system in the vasculature is a crucial and current topic in Pharmacology, Toxicology, Clinical Pharmacology, Physiology, Pathophysiology, and Clinical Medicine.
This review article covers areas of research on vasodilatation mediated by NO from perivascular nerves in vitro and in vivo, on the pathophysiological implication of neurogenic NO, and possible development of new therapeutic strategies in reference to nerve-derived NO. A section on the corpus cavernosum and penile erection will be included in this review, because penile erection is controlled by blood inflow from the penile artery and by blood outflow to the penile vein, because endothelial cells covering cavernous smooth muscle have similar properties with those of arteries and veins, and because functional characteristics of the dilator nerve are quite similar to those in the cerebral artery.
|
II. Discovery of Nitrergic (Nitroxidergic) Nerve in Blood Vessels |
|---|
|
TopPreviousNext |
|---|
NANC vasodilatation due to nerve stimulation was discovered in dog cerebral arteries by the use of nicotine (Toda, 1975). Supportive evidence was obtained in isolated feline cerebral arteries applied transmural electrical stimulation (Lee et al., 1975). Transmural electrical stimulation and nicotine induced similar relaxations in human, monkey (Toda, 1981, 1982), and sheep cerebral arteries (Duckles et al., 1977). Pharmacological studies have suggested that substances, such as prostanoids, ATP and histamine, and electrogenic Na+ pump do not participate in the response (Toda, 1975, 1978). There is considerable literature suggesting the involvement of acetylcholine (Vasquez and Purves, 1977; Bevan et al., 1982a,b), VIP (Lundberg et al., 1979; Lee et al., 1984; Brayden and Bevan, 1986), substance P (Edvinsson et al., 1981, 1982), and CGRP (Goodman and Iversen, 1986; Saito et al., 1989) in neurogenic cerebral vasodilatation (Lundberg, 1996). This conclusion was drawn on the basis of histological evidence demonstrating the presence of neurons containing acetylcholinesterase/choline acetyltransferase and the peptides in the vascular wall and of the ability of these compounds when applied exogenously to elicit vasodilatation in cerebral arteries. However, no evidence was shown that the molecules liberated from perivascular nerves dilated the arteries. The possible contribution of the peptides and acetylcholine synthesized and released from the nerve as mediators or neurotransmitters was excluded in dog and monkey cerebral arteries by the fact that the response to nerve stimulation was not influenced by capsaicin (Okamura and Toda, 1994c), by muscarinic receptor antagonists (Toda, 1975, 1982), by removal of the endothelium (Toda and Okamura, 1990b,c), or in arteries made tachyphylactic to the peptides under consideration (Toda, 1982; Okamura et al., 1989; Toda and Okamura, 1996a). In some studies with feline cerebral arteries, atropine depressed the neurogenic relaxation, suggesting that acetylcholine is one of the neurotransmitters (Bevan et al., 1982a,b). Acetylcholine-induced relaxations are mediated by endothelium-derived relaxing factor (EDRF) (Furchgott and Zawadzki, 1980). Since the relaxation to nerve stimulation is not influenced in feline arteries by endothelium denudation (Saito et al., 1989; Ayajiki et al., 1994), acetylcholine appears not to mediate the response in feline cerebral arteries. Substance P, an EDRF-mediated cerebral vasodilator (Conner and Feniuk, 1987; Onoue et al., 1988), can therefore be excluded as a transmitter candidate.
Attention was directed to discover endogenous cerebroarterial vasodilators as neurotransmitter candidates. In addition to VIP, CGRP, and substance P, ATP, 
-aminobutyric acid, pentagastrin (Toda, 1982), and atrial natriuretic peptide (Okamura et al., 1989) produced dilatation of dog and monkey cerebral arteries, but they were excluded as candidates either by experiments with receptor antagonists or by tachyphylaxis studies. Many other peptides and amino acids failed to produce relaxation (Toda, 1982). Nitroglycerin, although exerting strong vasodilator activity on cerebral arteries, could not be considered as an endogenous substance. Therefore, it was an important task to find inhibitors that selectively depressed this neurogenic relaxation. Oxyhemoglobin (oxyHb), an NO scavenger (Martin et al., 1985b), was such an inhibitor (Toda, 1988b; Linnik and Lee, 1989). Methylene blue, an inhibitor of soluble guanylyl cyclase (Gruetter et al., 1981), also inhibited the response (Toda, 1988b). However, no one, including ourselves, had the insight to regard EDRF as a neurotransmitter in the late 1980s. Garthwaite and his coworkers (1988) theorized that EDRF was released from neurons in the brain. Therefore, the hypothesis that NO was a neurotransmitter in vasodilator nerve innervating the vascular wall was not raised until 1990 (Toda and Okamura, 1990a,b).
B. Evidence for the Presence of Vasodilator Nerve-Releasing Nitric Oxide As a Neurotransmitter—Criteria for Transmitter
Abolition by NG-monomethyl-L-arginine (L-NMMA), a NOS inhibitor (Palmer et al., 1988b), of vasodilatation caused by electrical nerve stimulation or by nicotine was obtained for the first time in isolated dog cerebral arteries (Toda and Okamura, 1990a, b, c). The inhibitory effect was reversed by L-arginine, a substrate of NOS, but not D-arginine. D-NMMA was without effect. Other NOS inhibitors, such as NG-nitro-L-arginine (L-NA) (Mulsch and Busse, 1990; Rees et al., 1990; Toda et al., 1990c) (Fig. 1), L-NA methylester (L-NAME) and N-iminoethyl-L-ornithine (Rees et al., 1990), were also effective. Removal of the endothelium did not inhibit the neurogenic relaxation (Toda and Okamura, 1990b, 1991b). On the other hand, nerve stimulation-induced relaxation of dog coronary arteries was not influenced at all by the NOS inhibitors (Toda et al.,1990c) but was abolished by 
-adrenoceptor antagonists (Toda and Hayashi, 1982). Exogenously applied NO (acidified NaNO2) (Furchgott 1988) and NO donors, including nitroglycerin and sodium nitroprusside, induced dose-dependent dilatation of cerebral arteries, which was not affected by treatment with NOS inhibitors but was abolished by oxyHb and methylene blue. As described in the previous section, the latter two compounds abolished the response to nerve stimulation. ODQ (1H[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one), a soluble guanylyl cyclase inhibitor (Garthwaite et al., 1995), was also effective in depressing the response to nerve stimulation (Toda et al., 1999).
|
NO is synthesized from L-arginine by nNOS, which is activated by Ca2+ in the presence of calmodulin and other cofactors (Bredt and Snyder, 1990). Responses to electrical stimulation were abolished by removal of extracellular Ca2+ and also by 
-conotoxin GVIA, an N-type Ca2+ channel blocker (Toda et al., 1995c), but not by blockers of the L-type Ca2+ channel (Toda et al., 1992). The Ca2+ entry into nerve terminals upon arrival of generated action potentials or via Ca2+ channel opening due to nicotinic receptor stimulation would be responsible for activation of nNOS and NO synthesis.
In superfused cerebral arterial strips without the endothelium, the release of NO, measured as nitrite and nitrate (NOx), into the superfusate was increased during nerve stimulation by electrical pulses and nicotine (Toda and Okamura, 1990b). Tetrodotoxin (TTX) abolished the effect of electrical nerve stimulation, and hexamethonium (C6) abolished the nicotine action. Increased release of NOx by electrical stimulation was also abolished by L-NA in superfused dog temporal arteries (Toda et al., 1991c). It may be possible in the cerebral artery to demonstrate the release of NO by bioassay as shown in the dog intestine (Bult et al., 1990), visualization by a reaction with luminol and hydrogen peroxide to generate photons as shown in the guinea pig ileum (Wiklund et al., 1997), or using a novel diaminofluorescein (Kojima et al., 1998) as shown in the porcine coronary artery (Itoh et al., 2000). The content of cyclic GMP in the tissue was also increased by electrical stimulation and nicotine, and L-NA abolished the stimulating effect. The nucleotide increment produced by electrical stimulation and nicotine was abolished by TTX and C6, respectively (Toda and Okamura, 1991b).
Histochemical studies have demonstrated that guanylyl cyclase and phosphodiesterase, an enzyme that degrades cyclic GMP, are located in cells surrounding the rat brain vasculature (Poeggel et al., 1992). Perivascular nerve fibers containing NOS immunoreactivity have been demonstrated in dog, monkey (Yoshida et al., 1993, 1994a), and rat cerebral arteries (Bredt et al., 1990) by the use of antiserum raised against NOS purified from the rat cerebellum (Bredt et al., 1990). NOS-immunoreactive neurons are reportedly identical to reduced NADPH-positive ones (Dawson et al., 1991); therefore, the presence of networks of NOS-containing nerve fibers and cells in the vascular wall and brain has been widely demonstrated by the NADPH-staining method (Vincent and Kimura, 1992; Minami et al., 1994).
Findings presented thus far support the hypothesis that NO or stable analog of NO (R-SNO) acts as a neurotransmitter of vasodilator nerves innervating the cerebral artery. The general lines of evidence to support the concept of neurotransmission as delineated in Goodman and Gilman's text book (Hoffman and Taylor, 2001) include 1) demonstration of the presence of a physiologically active compound and its biosynthetic enzymes at appropriate sites; 2) recovery of the compound from the perfusate of an innervated structure during periods of nerve stimulation, but not in the absence of stimulation; 3) demonstration that the compound is capable of producing responses identical with those to nerve stimulation; 4) demonstration that the responses to nerve stimulation and to the administered compound are modified in the same manner by various drugs, usually antagonists. Except for the first part of 1), NO or R-SNO, like S-nitroso thiol, fulfills the criteria for neurotransmitter. Verification of the presence of the unstable molecule NO or its storage in vesicles would not necessarily be required. Substrate L-arginine in sufficient amounts (Forstermann et al., 1994) (100–800 µM of L-arginine in endothelial cells) and nNOS activated by Ca2+ locally produce NO that easily crosses cell membrane and diffuses out of nerve terminals. This represents a new idea for this gaseous and labile transmitter substance with low molecular weight. Therefore, the criteria for defining a neurotransmitter may in the future need to be rewritten along these lines. Hypothetical scheme of the neurotransmission process from the nitrergic nerve to vascular smooth muscle is summarized in Fig. 2.
|
Many arguments against this quite unexpected hypothesis have arisen. For instance, it was suggested that NO may be liberated from extraneuronal tissues. Answers to this comment are: the presence of constitutive NOS is histochemically demonstrated in the neuron, nerve cell, and endothelium, but not in smooth muscle and other adjacent tissues (Bredt and Snyder, 1990); blockers of L-type Ca2+ channels that suppress Ca2+ entry in smooth muscle fail to interfere with the response to nerve stimulation (Toda and Okamura, 1992a); and NOS in tissues other than neurons and endothelium is inducible only when incubated for long hours in lipopolysaccharides and endotoxin. In addition, the fact that damage to the dog pterygopalatine ganglion for 1 week histologically causes the NOS-containing neurons to disappear in the cerebral arterial wall and abolishes the dilator response to nerve stimulation of the isolated artery (Toda et al., 1993d) excludes the possibility of NO release from extraneuronal sites. It has also been suggested that neurally induced vasodilatation in isolated sheep middle cerebral arteries may be mediated largely by VIP and that possibly VIP acts to stimulate the formation of NO within smooth muscle, macrophage, or other tissues in the vascular wall (Gaw et al., 1991). In accordance with the reasons presented above, this cannot be the case, at least in dog and monkey cerebral arteries.
The other comment raised is that NO may act as a modulator of a vasodilator neurotransmitter. However, in cerebral arteries in which the neurogenic relaxation was abolished by NOS inhibitors, the addition of NO or NO donors did not restore the response (Toda and Okamura, 1990c). In bovine cerebral arteries, it is suggested that NO produced by perivascular nerves fully accounts for the experimental neurogenic relaxation, and VIP, present in the same nerves, acts as a neuromodulator by acting on nNOS (Gonzalez et al., 1997).
Whether the transmitter released is NO or R-SNO is still debatable in the perivascular nerve as well as in the endothelium (Myers et al., 1990; Feelisch et al., 1994). The reasons are as follows: the response to nerve stimulation was not inhibited by antioxidants, such as pyrogallol and hydroquinone, or augmented by superoxide dismutase (SOD) (Gillespie and Sheng, 1990; Toda and Okamura, 1990b), whereas actions of free radical NO were abolished by antioxidants (Moncada et al., 1986) and enhanced by SOD (Gryglewski et al., 1986; Rubanyi and Vanhoutte, 1986). From studies showing that the response to NANC nerve stimulation in the rat anococcygeus muscle and gastric fundus was insensitive to carboxy PTIO [2-(4-carboxyphenyl)-4,4,5,5,-tetramethylimidazoline-1-oxyl 3-oxide], a scavenger of free radical NO (Akaike et al., 1993), in concentrations sufficient to depress the relaxation induced by endothelium-derived NO in the rat aorta, the authors concluded that the transmitter from nitrergic nerves does not appear to be identical to endothelium-derived NO and may not be the free radical NO (Rand and Li, 1995a). However, they did not examine the property of nitrergic nerve in vasculature. On the other hand, it was noted that treatment with diethyldithiocarbamate, a membrane-permeable inhibitor of Cu/Zn SOD, allowed pyrogallol, hydroquinone, and duroquinone to inhibit the NO-mediated neurogenic response in the bovine retractor penis muscle and mouse anococcygeus muscle (Martin et al., 1994; Lilley and Gibson, 1995). The inhibition was reversed by SOD. Similar results were also obtained in monkey (Okamura et al., 1998c) and porcine cerebral arteries (Tanaka et al., 1999), suggesting that endogenous SOD protects neurons from functional impairment by superoxide anions and that the free radical NO, rather than R-SNO, acts as a neurotransmitter in these vascular and nonvascular tissues.
C. Mechanism of Nitric Oxide Formation and Action
The peripheral neurotransmitter of the NANC vasodilator nerve is presumed to be NO in intracranial vasculature in mammals so far investigated and in peripheral arteries of those other than rodents. Neuronal NOS, now designated as nNOS
(Eliasson et al., 1997), was originally purified and cloned from the brain (Bredt and Snyder, 1990; Bredt et al., 1991). The distribution and functional role of nNOS have been elucidated in tissues other than the brain, such as the skeletal muscle (Brenman et al., 1995; Stamler and Meissner, 2001) and macula densa in the kidney (Wilcox et al., 1998). In contrast to endothelial NOS (eNOS), nNOS is regulated post-transcriptionally and has many alternative splice variants (Wang et al., 1999b; Alderton et al., 2001). However, whether or not the nNOSs in brain, peripheral nitrergic nerve, and extraneuronal tissues are identical is not yet clear. Neuronal NOSs in brain and skeletal muscle are located in and associated with an adapter protein PSD-95 in the postsynaptic cells (Brenman et al., 1996; Tochio et al., 2000), whereas nNOS in the peripheral nitrergic nerve is located mainly in presynaptic nerve terminals. Since biochemical and molecular biological information of nNOS in the nitrergic vasodilator nerve per se is largely incomplete, mechanisms of NO formation and action can only be speculated upon from an immunological analogy to nNOS in the brain and on the basis of functional and histological data obtained from physiological and pharmacological experiments with isolated vascular preparations, including susceptibility of biological responses to selective nNOS inhibitors.
1. Formation of Nitrergic Neurotransmitter.
a. L-Arginine.
L-Arginine serves as a substrate of NOS to yield NO and L-citrulline by two steps: from L-arginine and O2 to NG-hydroxy-L-arginine, which is then converted to NO and L-citrulline. L-Arginine alone does not potentiate the neurogenic response but completely prevents the inhibition induced by NOS inhibitors. The inability of this NOS substrate to increase the neurogenic relaxation of isolated blood vessels may be explained in terms of there being sufficient amounts of L-arginine in the nerve cells (approximately 100 µM in the brain) (Barbul, 1986). The Km values of the enzyme for L-arginine are 1.5 to 2.3 µM (Bredt and Snyder, 1990; Schmidt et al., 1991; Ohshima et al., 1992).
L-Arginine is supplied by uptake from the extracellular space through a cation amino acid transporter system (system y+) or by synthesis from L-citrulline intracellularly via argininosuccinate synthetase and argininosuccinate lyase in neuronal cells (Wiesinger, 2001) (Fig. 3). In isolated porcine cerebral arteries without the endothelium, the uptake of L-citrulline and the active conversion from L-citrulline to L-arginine were noted, but these were absent in denervated arteries (Chen and Lee, 1995a). Presence of argininosuccinate synthetase and argininosuccinate lyase has been demonstrated histologically in NADPH diaphorase-positive nerve fibers innervating the cerebral artery (Yu et al., 1997). Therefore, a recycling process to synthesize L-arginine from L-citrulline seems to function in the perivascular nitrergic nerve. In cultured macrophages under treatment with lipopolysaccharide (Baydoun and Mann, 1994; Closs et al., 1997), the cationic transporter system y+ mediates the uptake of L-arginine, L-NMMA, NG,NG-dimethyl-L-arginine (ADMA), and NG-iminoethyl-L-ornithine, whereas a neutral transporter mediates the uptake of L-citrulline, L-NA and L-NAME. On the other hand, two other L-arginine analogs, NG,N'G-dimethyl-L-arginine (SDMA) and -amino-
-isothioure-idovaleric acid that fail to inhibit NOS enzyme activity, compete for the L-arginine transport (Closs et al., 1997). At present, it is not known whether these differences in substrate recognition of the transporter systems are also operative in the nitrergic nerve.
|
Arginases I and II affect intracellular L-arginine concentrations by catalyzing L-arginine as a substrate. Km values for these arginases are 2 to 20 mM and approximately 1000 times higher than those for NOS (Grody et al., 1987; Griffith and Stuehr, 1995); therefore, arginases are not likely to compete with NOS in the enzymatic reaction. However, once arginase is induced by cytokines, arginases and NOS may compete for the utilization of L-arginine, since the Vmax of arginases is 1000 times greater than those for NOS (Griffith and Stuehr, 1995; Wu and Morris, 1998). NOS inhibitors of L-arginine analogs so far described are ineffective on the arginase activity (Hrabak et al., 1994), but a sufficient amount of NG-hydroxy-L-arginine, an intermediate product of L-arginine by NOS, inhibits the arginase activity in iNOS-expressing macrophages (Hecker et al., 1995). These complex interactions between arginase and NOS may be important under pathological conditions. However, no data are available concerning the functional role of arginases in the nitrergic nerve.
b. Ca2+ and Calmodulin.
Neuronal NOS is a constitutive type of NOS, and activity of the purified enzyme is dependent on Ca2+ (Mayer et al., 1992) and calmodulin (Bredt and Snyder, 1990) in the presence of cofactors such as FMN, FAD, NADPH, and tetrahydrobiopterin (BH4). Generated action potentials evoked by a rapid rise of Na+ influx in nerve terminals allow opening of slow channels, through which Ca2+ is introduced into the neuron along extra- and intracellular concentration gradients. Therefore, the increase in the intracellular Ca2+ mainly by increasing the transmembrane influx or by the release from intracellular storage sites facilitates its binding to calmodulin. The Ca2+-calmodulin complex directly activates the enzyme, thus initiating electron flow from the reductase domain to the oxygenase domain to produce NO (Alderton et al., 2001). The Ca2+-calmodulin complex may also activate Ca2+/calmodulin-dependent protein kinases. It has been reported that nNOS has a phosphorylation site and is phosphorylated at Ser847 by Ca2+/calmodulin-dependent protein kinase II (Ca/CaM kinase II; Hayashi et al., 1999). This phosphorylation leads to a decrease in NOS activity (Nakane et al., 1991; Komeima et al., 2000). Neuronal NOS is also phosphorylated by cyclic AMP-dependent protein kinase (protein kinase A) and protein kinase C, with each kinase phosphorylating a different serine site on NOS (Bredt et al., 1992). Phosphorylation of nNOS by protein kinase C reportedly decreases (Bredt et al., 1992) or increases (Nakane et al., 1991) the enzyme activity. No change (Brune and Lapetina, 1991) or decrease (Dinerman et al., 1994) in nNOS activity was observed upon phosphorylation by protein kinase A. Phosphorylation by cyclic GMP-dependent protein kinase (protein kinase G) is reported to reduce the activity (Dinerman et al., 1994). Such diverse actions may be due to differences in the phosphorylation site(s) of the enzyme. The data discussed above are based on that obtained from nNOS protein purified from different species. In intact cells or in vivo systems, there may be other complex mechanisms of nNOS activation and deactivation by protein phosphorylation.
2. Action of Nerve-Derived Nitric Oxide on Vascular Smooth Muscle.
The signal transduction system for smooth muscle relaxation produced by the nitrergic neurotransmitter involves the activation of soluble guanylyl cyclase and the production of the second messenger cyclic GMP (Schmidt et al., 1993) (Fig. 2). In isolated blood vessels, neurogenic relaxation is abolished by soluble guanylyl cyclase inhibitors, such as methylene blue and ODQ and is accompanied by an increase in intracellular cyclic GMP contents (Toda and Okamura, 1996b).
Soluble guanylyl cyclase expressed in vascular smooth muscle cells catalyzes the conversion of GTP to cyclic GMP (Papapetropoulos et al., 1996). The enzyme is a heterodimer of and subunits containing a heme prosthetic group (Lucas et al., 2000). Both subunits are required for basal and NO-stimulated catalytic activity (Harteneck et al., 1990; Nakane et al., 1990). NO activates soluble guanylyl cyclase by binding directly to heme to form a ferrous-nitrosyl-heme complex, breaking the bond between iron and His105, and displacing the iron from the plane of the porphyrin ring. Termination of activation probably results from dissociation of NO from the heme and return of the iron core to the plane of the porphyrin ring (Lucas et al., 2000). NO also decreases stability of mRNAs encoding soluble guanylyl cyclase subunits in rat pulmonary artery smooth muscle cells (Filippov et al., 1997). Gene transfer of soluble guanylyl cyclase subunits to balloon injured blood vessels increases the responsiveness to NO in rats (Sinnaeve et al., 2001).
Although important roles of cyclic GMP in vascular relaxation are well recognized (Waldman and Murad, 1987), intracellular mechanisms and signal transduction pathways may vary in cells, tissue types or animal species. The final common step is a reduction in the intracellular free Ca2+ concentration or Ca2+ sensitivity (Fig. 2), which prevents the Ca2+-dependent activation of myosin light-chain kinase and muscle contraction. A number of hypotheses on biosignaling processes from cyclic GMP production to decreased intracellular Ca2+ concentrations have been proposed. Cyclic GMP has been shown to inhibit Ca2+ entry into cells by inhibiting L-type Ca2+ channels (Lincoln, 1989). Cyclic GMP-mediated activation of protein kinase G may be involved (Lincoln and Cornwell, 1993), which is suggested to inhibit voltage-gated Ca2+ channels directly (Clapp and Gurney, 1991) or indirectly by activating Ca2+-sensitive K+ channels and hyperpolarizing the cell membrane (Archer et al., 1994). Protein kinase G has also been shown to activate the Na+/Ca2+ exchanger (Furukawa et al., 1991), the Ca2+-ATPase on the plasma membrane (Yoshida et al., 1991), or the Ca2+-ATPase on the endoplasmic reticulum indirectly by phosphorylation of phospholamban, an endoplasmic reticulum Ca2+-ATPase regulatory protein (Cornwell et al., 1991). Protein kinase G has been noted to inhibit the agonist-induced production of inositol trisphosphate (IP3) (Hirata et al., 1990) or the effect of IP3 by phosphorylation of IP3 receptors on the endoplasmic reticulum membrane (Komalavilas and Lincoln, 1994).
Neuronal NOS-derived NO produces several actions other than vascular smooth muscle relaxation. In the central nervous system, NO, produced in and released from the postsynaptic cells, controls the release of the neurotransmitter glutamate (Garthwaite and Boulton, 1995). The linking of NO to the release (Meffert et al., 1996) of other neurotransmitters and their effects has also been reported in the brain and nonvascular tissues; i.e., acetylcholine (Gustafsson et al., 1990), dopamine (Hanbauer et al., 1992), -aminobutyric acid (Kuriyama and Ohkuma, 1995), and bombesin (Beltran et al., 1999). The mechanisms responsible for these actions are not fully understood, but direct S-nitrosylation of receptors, activation of cyclic GMP-dependent protein phosphorylation cascades, regulation of neuronal energy dynamics, and modulating effect on transporters may be involved (Esplugeus, 2002).
S-Nitrosylation of cysteine residues in proteins by exogenous NO is thought to be involved in certain physiological and pathological events that are not readily explained by a mediation of cyclic GMP. Several proteins are endogenously S-nitrosylated by neuronally generated NO, since this modification is not observed in animals harboring a genomic deletion of nNOS (Jaffrey et al., 2001). Whether or not this pathway is involved in the events mediated by NO derived from perivascular nitrergic nerves remains to be clarified.
|
III. Nitrergic Innervation in Intra- and Extracranial Vasculature |
|---|
|
TopPreviousNext |
|---|
1. In Vitro Studies in Various Mammals.
Electrical (1–20 Hz) and chemical stimulation (10-6–10-4 M nicotine) elicited relaxations in a frequency- and concentration-dependent manner, respectively, in cerebral arteries previously contracted with vasoconstrictors (Fig. 1). Treatment with -adrenoceptor antagonists or adrenergic neuron blockers did not alter or augment the response in cerebral arteries of various mammals (Toda and Okamura, 1996b). However, in sheep cerebral arteries, the stimulation induced contraction, which was reversed to a relaxation response by -adrenoceptor blockers (Duckles et al., 1977). Therefore, to elucidate the mechanism underlying neurogenic relaxation, experiments have been conducted in vasculature with -adrenoceptor blockade.
Neurogenic vasodilatation mediated by NO in isolated cerebral arteries has been demonstrated in a variety of mammals, including the human (Toda, 1993), monkey (Toda and Okamura, 1990c; Toda et al., 1997b), dog (Toda and Okamura, 1990a,b, 1991b; Toda et al., 1993d), pig (Lee and Sarwinski, 1991; Tanaka et al., 1999), cow (Gonzalez and Estrada, 1991; Ayajiki et al., 1993), cat (Ayajiki et al., 1994), sheep (Matthew and Wadsworth, 1997), guinea pig (Jiang et al., 1997), and rat (Ignacio et al., 1997). In small mammals, only the basilar artery is used, although the intracranial, extracerebral arteries of different regions can be utilized in mammals of medium or large size. In dogs, nicotine (10-4 M)-induced relaxations average 45.4 ± 4.4% (n = 11) in anterior cerebral, 50.2 ± 3.1% (n = 18) in middle cerebral, 31.6 ± 5.8% (n = 12) in posterior cerebral, and 28.5 ± 3.7 (n = 10) in basilar arteries. Greater responses were obtained in the anterior and middle cerebral arteries than in the other vessels. According to the histochemical study by Suzuki et al. (1994), the density of NADPH diaphorase (NOS)-containing nerve fibers was higher in the anterior half of the circle of Willis than its posterior half in the rat. Main trunks of the dog middle cerebral artery responded to nicotine with a greater relaxation than did the third branches of the artery (56.2 versus 22.6%; Toda and Miyazaki, 1984). Nicotine-induced relaxations did not significantly differ in cerebral arteries isolated from beagles of different ages (30 days, 3 months, 1 year, and 3 years) (Toda et al., 1986). In contrast, relaxations induced by electrical nerve stimulation of cerebral arteries varied directly with age in Japanese monkeys ranging from infancy (3–4 weeks) to adulthood (over 7 years) (Toda, 1991). The values do not seem to reflect nitrergic nerve functioning directly, since nicotinic receptors are not always distributed evenly in these arteries. Studies with electrical nerve stimulation also have problems, such as positioning of the stimulating electrodes including clearance between the tissue and electrodes, thickness of the vascular wall, etc. Although responsiveness to vasoactive agents, including NO, of intracerebral microvessels were noted (Dacey et al., 1988; Takayasu et al., 1994), the functional role of nitrergic nerve was not evaluated. Histological evidence for neurons containing NOS immunoreactivity or NADPH diaphorase in the microvascular wall (Kummer et al., 1992; Estrada et al., 1993; Iadecola et al., 1993) suggests that NO derived from the nerve contributes to regulate vascular size.
In most but not all of the cerebral arteries from different species, relaxant responses to electrical nerve stimulation at frequencies up to 20 Hz are abolished by high concentrations of NOS inhibitors, oxyHb, and guanylyl cyclase inhibitors, suggesting that the response is exclusively or at least largely mediated by NO. However, in some cerebral arteries, other mechanisms are also involved.
In feline cerebral arteries, neurogenic relaxation was partially reduced but not abolished by high concentrations of L-NA, and the remaining response was abolished by capsaicin or in the arteries made tachyphylactic to CGRP but not to VIP (Ayajiki et al., 1994). Another study noted that the arterial relaxation was mediated by CGRP, although the involvement of neurogenic NO was not examined (Saito et al., 1989).
In bovine cerebral arteries, we concluded that NO appeared to be the sole transmitter (Ayajiki et al., 1993). In arteries that were made unresponsive to VIP and CGRP by their successive applications, the neurogenic relaxation was unaffected, suggesting that these peptides do not participate in the response. However, other authors have reported that the main mechanism mediated by nerve-derived NO is regulated by VIP liberated from perivascular nerves in the bovine cerebral artery (Gonzalez et al., 1997).
In sheep cerebral arteries, the major mediator is considered to be VIP, which releases NO from the vascular wall (Gaw et al., 1991). This idea is supported by studies with VIP antiserum in the same artery (Matthew et al., 1997). In the guinea pig uterine artery, relaxations to high frequency stimulation (>5 Hz) are mediated by peptides, possibly VIP (Morris, 1993). However, in the dog (Toda et al., 1990c), monkey (Toda et al., 1997b), and bovine cerebral arteries (Ayajiki et al., 1993), no evidence supporting the involvement of peptides was obtained even when the nerve was stimulated electrically at frequencies as high as 20 Hz.
Although 7-nitroindazole was introduced as a promising selective inhibitor of nNOS (Moore et al., 1993), controversial data have been reported (Reiner and Zagvazdin, 1998). In a study with monkey cerebral arteries, the selectivity on responses to nitrergic nerve stimulation by electrical pulses and nicotine and to histamine, an endothelium-derived NO-releasing substance (Toda, 1990), was examined (Ayajiki et al., 2001b). This inhibitor was about as potent as L-NMMA and 100 times less potent than L-NA in attenuating the neurogenic relaxation and was about 5 times more potent in inhibiting nNOS than eNOS. The conclusion is that 7-nitroindazole is a relatively selective nNOS inhibitor.
It has been reported that NO-mediated neurogenic dilatation in guinea pig basilar arteries is associated with increased formation of intracellular cyclic GMP and activation of large-conductance Ca2+-activated K+ channel (Jiang et al., 1998). Relaxation or hyperpolarization of smooth muscle cell membrane induced by endothelial NO or NO donors is reportedly ascribed to activation of ATP-sensitive K+ channels in cerebral arteries and arterioles of the pig and piglet (Armstead, 1996; Bari et al., 1996) or Ca2+-activated K+ channels in rabbit cerebral arteries and arterioles (Robertson et al., 1993; Taguchi et al., 1995; Dong et al., 1998).
2. Nerve Stimulation by Electrical Pulses and by Nicotine and Related Compounds.
As previously described, stimulation of the nerve with electrical square pulses and chemically with nicotine or other nicotinic agonists produce similar patterns of responses in cerebral and peripheral arteries (Toda, 1995), and the neurotransmitters involved are identical (Fig. 2). Isolated dog cerebral and coronary arteries respond to electrical and chemical stimulation with relaxation, and the other arteries, including mesenteric, renal, femoral, pulmonary, superficial temporal, etc., contract in response to these stimuli. Evidence supporting the idea that the response to either mode of stimulation is due to the release of neurotransmitters from the activated nerve terminals in blood vessels has been mounting (Su, 1982; Toda, 1982; Nedergaard, 1988). Not only C6 and pentolinium but also neosurgatoxin (Ayajiki et al., 1998) abolish the action of nicotine on nitrergic nerves innervating the dog cerebral artery and ganglionic cells of the dog duodenum. The inhibitory potency of the toxin is greater in the ganglion than in the nerve and is about 3000 times greater than C6 in abolishing the response of dog cerebral arteries.
There are some discrepancies in the susceptibility of these response to inhibitors. The response to electrical stimulation is abolished by TTX but is not influenced by C6. This response is potentiated by inhibitors of the amine transporter, such as cocaine, desipramine, etc., only when contractions are induced by adrenergic nerve stimulation (Toda, 1972). In contrast, the response to nicotine is abolished by C6 or other ganglionic blocking agents and also by amine transporter inhibitors (Su and Bevan, 1970; Toda, 1976) but is resistant to TTX. Studies on Ca2+ concentration changes in single nerve terminal varicosities of the mouse vas deferens demonstrated that the nicotine action was insensitive to TTX at a concentration that blocked action potential-evoked Ca2+ transients (Brain et al., 2001). These findings indicate that different mechanisms of action are involved in electrical and chemical stimulation responses. It is well known that the generation of nerve action potentials by electrical stimulation evokes the transmembrane influx of Na+ in nerve terminals, and the induced depolarization opens voltage-dependent Ca2+ channels, leading to Ca2+ influx. This influx triggers exocytosis of transmitter vesicles in adrenergic and cholinergic nerves or evokes activation of nNOS in nitrergic nerves (Fig. 2).
In contrast, the mechanism of nicotine actions is less well understood. For example, it is not clear whether nicotine can generate action potentials by acting on nicotinic receptors in nerve terminals. The variable efficacy of TTX on the vascular response to nicotine is puzzling. The action of nicotine is abolished (Bell, 1968a) or partially attenuated in the rabbit ear artery (Furchgott et al., 1975) by TTX, but it is not influenced in the rabbit pulmonary artery (Su and Bevan, 1970), rabbit aorta (Ikushima et al., 1981), and dog basilar artery (Muramatsu et al., 1978) by TTX. In dog mesenteric, renal, and femoral arterial strips, nicotine-induced contractions mediated by norepinephrine are partially inhibited by TTX in concentrations (1–3 x 10-7 M) sufficient to abolish the contraction induced by electrical nerve stimulation (Toda et al., 1976). However, the nicotine-induced relaxation of dog and monkey cerebral arteries mediated by NANC neurotransmitter (later proven to be NO) was not inhibited (Toda et al., 1976; Toda, 1982). These findings may lead us to speculate that action potentials are generated by nicotine in adrenergic nerve terminals in some mammals, including rabbits and dogs, but this is not the sole mechanism. In dog and monkey cerebral arteries, nicotine (10-4 M) appear not to mediate the generation of action potentials but may have acted on nicotinic receptors responsible for opening the Ca2+ channel to exert NO-mediated relaxation. Studies on brain neurons suggest that at least two populations of presynaptic nicotinic receptors (TTX-sensitive and -insensitive) participate in the nicotine-induced amine release (Marshall et al., 1996; Kaiser and Wonnacott, 1999).
In eNOS knockout mice, acetylcholine (10-5 M)-induced dilatation of pial arterioles is blunted by TTX and abolished by L-NA (Meng et al., 1998). Because the response was abolished by atropine, the authors suggested that muscarinic receptor activation evokes the release of NO from the endothelium in which nNOS is expressed. Another possibility is that acetylcholine stimulates nicotinic receptors in nitrergic nerve endings and releases NO, since atropine used (10-5 M) was sufficient to induce complete blockade of muscarinic receptors and partial inhibition of nicotinic receptors.
Another question regarding nicotine actions is how inhibitors of the amine transporter abolish vascular actions of nicotine, a problem that has not been answered over several decades. The inhibitors include cocaine, tricyclic antidepressants such as imipramine and desipramine, propranolol, phentolamine, etc.; therefore, interpretations based on the use of pharmacological blockers for the analysis of nicotine actions must be made with care. On the basis of accumulated data concerning nitrergic nerve stimulation by nicotine, we hypothesize that nicotine or other ganglionic stimulants are transported by a cocaine-sensitive active process to the inside of cells where they act on Ca2+ channels (Fig. 2). Acetylcholine at the high concentration of 10-4 M relaxes dog cerebral and retinal arteries and monkey ciliary arteries, and this response is resistant to atropine but blunted by cocaine (Toda, 1979) and abolished by C6 (Toda, 1979; Toda et al., 1995e, 1998). On the other hand, acetylcholine and nicotine act on nicotinic receptors of ganglionic cell membranes to increase ion permeability and to depolarize the cell membrane to a level of firing generated action potentials. This action is not influenced by amine transporter inhibitors; thus, nicotinic receptor stimulants do not necessarily cross the cell membrane but are generally recognized to act on receptors located on the outside of the membrane. Future refined microanalysis and molecular biology methods would be required to directly verify our hypothesis on nicotine's mechanism of action that is not based on the generation of action potentials.
The possibility that norepinephrine released by nicotine from adrenergic nerve stimulated the release of NO from nitrergic nerve was proposed in a study using porcine basilar arteries treated with guanethidine, an adrenergic neuron blocker, and 6-hyroxydopamine (Zhang et al., 1998), which chemically denervates adrenergic neurons. The validity of this idea is incomplete until it is shown that these compounds do not impair the uptake of nicotine by the amine transporter. In dog cerebral arterial strips, nicotine-induced relaxations are suppressed by treatment with propranolol, phentolamine, and bretylium but unaffected by sotalol, timolol, prazosin, and yohimbine (Toda, 1975; Leckstrom et al., 1993; Toda et al., 1995c). The inhibition by the former three inhibitors is considered to result from a blockade of the amine transporter. In contrast to the data presented by Zhang et al. (1998), we found no evidence supporting the view that norepinephrine applied exogenously stimulates the release of NO from vasodilator nerves in porcine cerebral arteries (Tanaka et al., 1999).
Susceptibility to pharmacological interventions of the NO-mediated relaxation induced by electrical and chemic
