Endogenous and Exogenous Nitric Oxide Inhibits Norepinephrine Release From Rat Heart Sympathetic Nerves
Abstract This study was designed to elucidate whether nitric oxide (NO) controls norepinephrine (NE) release from sympathetic nerves of the rat heart. Hearts were perfused in the Langendorff mode with Tyrode’s solution. The right sympathetic nerve was stimulated with trains of 1 or 3 Hz and NE release was measured. The NO synthase (NOS) inhibitor NG-nitro-l-arginine (L-NNA) enhanced the evoked NE release in a concentration-dependent manner. This facilitation was independent of the increase in perfusion pressure and was stereospecifically reversed by l-arginine but not d-arginine. Another NOS inhibitor, NG-methyl-l-arginine, produced a similar increase in NE release. The NO-donor compound S-nitroso-N-acetyl-d,l-penicillamine, added in the presence of L-NNA, restored the suppression of NE release in a concentration-dependent fashion. A similar suppression was achieved with 3-morpholinosydnonimine. These results demonstrated that NE release is under the inhibitory control of endogenous NO. Western blots demonstrated the presence of neuronal NOS I and endothelial NOS III in the hearts. Perfusion of the hearts with a low concentration of the detergent CHAPS produced functional damage of the endothelium, as evidenced by an increase in perfusion pressure and a conversion of the acetylcholine-induced coronary vasodilation to a constriction. However, CHAPS treatment did not produce a facilitation of NE release (as did the NOS inhibitors), and L-NNA still increased NE release in CHAPS-treated hearts. Double-labeling immunofluorescence histochemistry showed NOS I immunoreactivity in stellate ganglion cells and in neurons of the heart, some of which also stained positive for tyrosine hydroxylase. These findings suggest that NO, predominantly generated in neurons, controls NE release in the rat heart.
Nitric oxide is synthesized by three different isoforms of NOS (for review see References 1 and 21 2 ). NOS I was originally isolated from rat and porcine cerebellum,3 4 5 but immunohistochemical studies have also localized this isoform to sympathetic and parasympathetic neurons.6 7 8 9 10 11 12
In parasympathetic neurotransmission, NOS inhibitors were shown to facilitate electrically evoked acetylcholine release from guinea pig myenteric plexus13 and canine myenteric plexus–free circular muscle14 and to enhance the cholinergic contractions of guinea pig ileum longitudinal muscle.15 16
NO is known to influence sympathetic neurotransmission in various vascular preparations; however, the implications of prejunctional and postjunctional effects are controversial. In the rat tail artery,17 the rat caudal artery,18 the rabbit pulmonary artery,19 and the dog temporal artery,20 NOS inhibitors enhanced the electrically stimulated vasoconstriction but failed to increase NE release. On the other hand, in the canine pulmonary arteries and veins21 and mesenteric artery,22 NO was shown to inhibit NE release via a prejunctional mechanism. Other investigators demonstrated both prejunctional and postjunctional involvement of NO in increased renal sympathetic nerve transmission.23
In the rat heart, NO released by degranulating mast cells was regarded as a permissive mediator that enabled coreleased histamine to increase the evoked NE overflow via activation of prejunctional, facilitatory H2 receptors on sympathetic nerve terminals.24
NOS I–containing neurons have been demonstrated in the rat heart.25 Similarly, NADPH-diaphorase–positive nerve fibers were also found.26 NOS III is present in all types of endothelial cells.27 In the rat heart, NADPH-diaphorase staining has been found in coronary arteries and arterioles.28
In this study on the isolated perfused rat heart, we demonstrate a prejunctional inhibition of NE release by endogenous and exogenous NO. Persistence of this NO inhibition after functional damage of the endothelium with the nondenaturing detergent CHAPS indicates that significant amounts of the NO are of nonendothelial origin. Immunohistochemistry localized NOS I and TH to adjacent neurons showing some overlap.
Materials and Methods
Perfusion of Rat Hearts According to Langendorff
Sprague-Dawley rats of either sex (160 to 220 g) were killed by a blow to the head and exsanguination. The thorax was opened and the ascending aorta was cannulated. The heart was perfused in a flow-through (nonrecirculating) manner at 5 mL/min via the coronary arteries according to the Langendorff technique. Tyrode’s solution (in mmol/L: NaCl 137, KCl 2.7, CaCl2 1.8, MgCl2 1.05, NaHCO3 11.9, NaH2PO4 0.42, d-glucose 5.6, and (+)ascorbic acid 0.057) at 35°C to 37°C oxygenated with 95% O2/5% CO2 was used as the perfusion medium. Under control conditions, l-arginine (0.1 mmol/L) was present in the solution to prevent substrate depletion of NOS. The right sympathetic nerve was dissected free from adjacent tissue, and the vagal nerves of both sides were excised. The heart with the attached sympathetic nerve was removed and connected to the Langendorff apparatus. The nerve was pulled through a platinum ring electrode (3-mm diameter) and superfused at a rate of 0.8 mL/min with Tyrode’s solution throughout the experiment. The right atrium and ventricle were connected to isometric transducers and subjected to a tension of 0.3 g (atrium) and 1 g (ventricle). CPP was monitored via an additional port in the aortic cannula. The viability of the sympathetic nerve was tested by applying square wave pulses of 1-ms duration. The optimal current strength was usually about 30 mA (as monitored on an oscilloscope). The contractility of the right atrium and ventricle, the beating frequency of the right ventricle, and the CPP were documented on a Gould 3000 recorder.
Electrical Stimulation Protocols
In one series of experiments (concentration-response curves), hearts were subjected to five periods of SNS (SNS1 through SNS5) at 25-minute intervals. The 3-minute stimulation periods consisted of trains of 3 Hz (one train of 6 pulses every 4 seconds, or 270 pulses). SNS1 and SNS2, applied in the presence of 0.1 mmol/L l-arginine, served as the control stimulation periods. Then the NOS inhibitor L-NNA was introduced at increasing concentrations, starting 24.5 minutes before the next stimulation period (SNS3) and continuing through SNS5. In other experiments the NO donor SNAP (dissolved in dimethylsulfoxide) was infused at 5 μL/min in the presence of 0.1 mmol/L L-NNA via an additional port in the aortic cannula, starting 5 minutes before SNS3, SNS4, and SNS5. Dimethylsulfoxide alone (0.1%, vol/vol) had no effect on NE release (n=2). Each treatment group had its own control experiments: five periods of SNS in the presence of l-arginine (0.1 mmol/L) as the control for the L-NNA concentration-response curve and five periods of SNS with L-NNA (0.1 mmol/L) present during SNS3 through SNS5 as the control for the SNAP concentration-response curve.
In a second series of experiments (single concentration studies), hearts were subjected to four periods of sympathetic nerve stimulation (SNS1 through SNS4). At SNS1 and SNS3, the nerve was stimulated with trains of 1 Hz (one train of 2 pulses every 4 seconds during 3 minutes, or 90 pulses); at SNS2 and SNS4, with trains of 3 Hz. SNS1 and SNS2, as well as SNS3 and SNS4, were separated by 15 minutes, whereas SNS2 and SNS3 were separated by 25 minutes. l-Arginine (0.1 mmol/L) was present throughout in control experiments, or replaced during SNS3 and SNS4 with the NOS inhibitors L-NNA or L-NMA (0.1 mmol/L) or the NO-donor compounds SNAP or SIN-1 (1 μmol/L). SNAP and SIN-1 (SIN-1 dissolved in 20 mmol/L potassium phosphate buffer, pH 5.5) were infused at 5 μL/min via an additional port in the aortic cannula, starting 5 minutes before SNS3 and SNS4, respectively. Phosphate buffer alone (pH 5.5; 0.1% vol/vol) had no effect on NE release (n=2). In one set of L-NNA experiments, desipramine (1 μmol/L) was present throughout to inhibit neuronal uptake of NE.
In order to mimic the vasoconstrictor effect of L-NNA and L-NMA, a set of experiments was performed with [Arg8]-vasopressin. The peptide was infused into the aortic cannula at a final concentration of 0.1 U/L before and during SNS3 and SNS4. The concentration of 0.1 U/L was selected to produce an increase in CPP slightly greater than L-NNA (0.1 mmol/L).
Functional Endothelial Damage
In order to elucidate whether endothelial NO was responsible for the effects on NE release, the coronary endothelium was damaged by perfusing the heart with the nondenaturing detergent CHAPS (1 mmol/L) for 10 minutes. ACh (0.1 μmol/L) was infused (for 2 minutes) before and after the CHAPS treatment. Conversion of the ACh-induced vasodilation to constriction was taken as evidence of a successful reduction of endothelial NO release. In some experiments L-NNA (0.1 mmol/L) was added to the perfusion medium after the CHAPS treatment, and its effect on NE overflow was compared with control experiments without CHAPS treatment.
Determination of Norepinephrine Overflow
NE was determined electrochemically following HPLC separation.29 Briefly, aliquots of the perfusates (3 mL) or the heart extracts obtained at the end of the experiments (0.1 mL, made up to 3 mL with distilled water)24 were mixed with DHBA as an internal standard. Two moles per liter Tris-HCl buffer, pH 8.68 (1 mL), Na2EDTA (10% wt/vol, 100 μL, to scavenge divalent cations), and Na2SO3 (12.5% wt/vol, 100 μL, to protect catecholamines from oxidation) were added. NE and DHBA were adsorbed to alumina and, after several washes with distilled water, eluted twice with perchloric acid (0.1 mol/L, 100 μL). Aliquots of the extracts were separated by HPLC on a reverse-phase column (Grom Spherisorb ODS-2), 3 μm, fitted with a precolumn of the same material. A citrate/phosphate buffer, pH 6.0 (in mmol/L: citric acid 30, Na2HPO4 15, Na2EDTA 1, 1-octanesulfonic acid 1.27, methanol 5% to 10% vol/vol), was used as the mobile phase. Catecholamines were detected electrochemically by a Waters 460 detector (potential set at +0.7 V). Calibration curves were generated with external NE standards. The peak height of the samples was corrected using the internal standard (DHBA).
Evoked NE overflow was determined by subtracting basal efflux during the 4 minutes before stimulation from the release during the following 4-minute SNS. The evoked overflow was expressed as a fraction of the heart NE content at the beginning of the respective SNS (estimated by determining the tissue content at the end of the experiment and adding the amounts of NE found in the perfusates).
For concentration-response curves, the fractional release during the first through fifth stimulation periods (S1-S5) was expressed as a percentage of S2. Changes of fractional NE release by L-NNA and SNAP were given as percentages of the control values.
For single concentration studies, the fractional releases at SNS3 (1 Hz) and SNS4 (3 Hz) in the presence of the test substances were given as percentages of the corresponding releases at SNS1 (1 Hz) and SNS2 (3 Hz).
Statistical differences were determined by Student’s unpaired t test or by factorial ANOVA followed by Fisher’s protected least-significant-difference test for comparison of multiple means.
Rat hearts were homogenized on ice as previously described for brain tissue and endothelial cells.4 30 CHAPS (20 mmol/L) was present during homogenization and thereafter to solubilize any particulate NOS.30 Homogenates were centrifuged at 100 000g for 1 hour, and the combined soluble and solubilized particulate material was partially purified on 2′,5′-ADP-Sepharose. Protein fractions (100 μg each) were separated by 7.5% polyacrylamide gel electrophoresis.31 The proteins were transferred to nitrocellulose membranes (Hybond-ECL, Amersham), and Western blots were performed with three NOS-isoform–selective antisera as previously described.32
The procedures for processing NOS I and TH immunoreactivity of the stellate ganglia were essentially the same as those described previously.33 Briefly, rats were deeply anesthetized with ketamine hydrochloride (70 mg/kg IP) and perfused intracardially with freshly prepared 4% paraformaldehyde in 0.1 mol/L PBS. Left and right stellate ganglia were prepared using a Vibratome.
Sections were first incubated in a rabbit polyclonal antiserum (No. 6761-9) to rat cerebellar NOS I (1:1000),25 with gentle agitation for 48 hours. After washing with PBS for 30 minutes, tissues were incubated with anti-rabbit IgG (1:50 with 0.4% Triton X-100 in PBS). Tissues were then washed and incubated with fluorescein avidin D (1:50 in PBS). After thorough washing with PBS for at least 2 hours, sections were incubated with the second primary antibody to TH (a mouse monoclonal, 1:500) for 48 hours. After washing with PBS for 30 minutes, sections were incubated with biotinylated horse anti-mouse IgG (1:50 in 0.4% Triton X-100 in PBS) for 2 hours. After several washes in PBS, sections were incubated in Texas Red avidin D (1:50 in PBS). Finally, tissues were washed for 30 minutes with PBS, mounted in Citifluor, and coverslipped. In control experiments, omission of the primary antibodies resulted in no staining in any of the sections.
To detect NOS I in postganglionic sympathetic neurons of the heart, hearts with the sympathetic nerves attached were rapidly frozen in −80°C 2-methylbutane. At −15°C serial sections of 30 μm were made with a Leica cryostat. The sections were fixed for 10 seconds in ice-cold acetone.
Double-labeling immunofluorescence was performed using the same antiserum to NOS I as described above (1:200) and a sheep polyclonal antibody against TH (1:100). After overnight incubation at 4°C an anti-rabbit IgG conjugated to indocarbocyanine (Cy 3; 1:150) and an anti-sheep IgG conjugated to FITC (1:64) were used to stain NOS I and TH simultaneously. In control experiments, the primary antibodies were replaced with normal rabbit or sheep sera, which resulted in a complete loss of immunofluorescence.
Sections were examined under an epifluorescent microscope equipped with appropriate filters. FITC-labeled cells and neurons and Texas Red– or Cy 3–labeled cells and neurons were identified by their respective color fluorescence when viewed with the appropriate filters.
Compounds and Antisera
Acetylcholine perchlorate, d-arginine hydrochloride, l-arginine hydrochloride, CHAPS, DHBA, L-NNA, (–)-norepinephrine bitartrate salt, and [Arg8]-vasopressin were obtained from Sigma Chemical Co, desipramine hydrochloride from Ciba Geigy, fluorescein avidin D and Texas Red avidin D from Vector Laboratories, L-NMA from Abbott Laboratories, SIN-1 from Hoechst-Cassella, and SNAP from Alexis.
Biotinylated horse anti-mouse IgG was purchased from Vector Laboratories, donkey anti-sheep immunoglobulin G conjugated to FITC and rabbit and sheep normal serum from Sigma, goat anti-rabbit immunoglobulin G conjugated to Cy 3 from Dianova, mouse monoclonal antibody to TH from Chemicon, sheep polyclonal antibody to TH from Pel-Freez Biologicals, and antisera to all three NOS isoforms were generated at Abbott Laboratories, as described.25 27 34
Effects of L-NNA and [Arg8]-Vasopressin on Coronary Perfusion Pressure
The NOS inhibitor L-NNA (0.1 mmol/L) enhanced CPP by 75±10 mm Hg (mean±SEM, n=4) over a time period of 40 minutes (see Table⇓). The endothelium-independent vasoconstrictor [Arg8]-vasopressin (0.1 U/L) increased CPP by 113±9 mm Hg during the same period of time (Table⇓). The contractility of the right atrium and ventricle and the beating frequency of the right ventricle were not changed by L-NNA or [Arg8]-vasopressin.
Effects of L-NNA, L-NMA, and [Arg8]-Vasopressin on Evoked NE Overflow
L-NNA, added in increasing concentrations 24.5 minutes before the stimulation periods SNS3, SNS4, and SNS5, enhanced the evoked NE overflow (Fig 1⇓). The basal outflow of NE was not affected by any concentration of L-NNA. Based on the results shown in Fig 1⇓, we used an L-NNA concentration of 0.1 mmol/L in subsequent experiments.
L-NNA, added 24.5 minutes before SNS3, significantly increased the evoked NE overflow at 1 Hz (90 pulses, S3/S1) and at 3 Hz (270 pulses, S4/S2) (Fig 2⇓). This effect was stereospecifically prevented by l-arginine (1 mmol/L) but not by d-arginine (1 mmol/L) (Fig 2⇓).
Desipramine (1 μmol/L), which inhibits neuronal uptake of NE, produced a marked increase in basal NE outflow (control: 59.5±9.0 pg/min, desipramine: 142.8±30.0 pg/min, mean±SEM, n=8). Desipramine did not attenuate the enhancement of stimulation-induced NE release by L-NNA (0.1 mmol/L). At 1 Hz, NE overflow in response to L-NNA (S3/S1) was 187.7±31.8% (n=4), which is slightly greater (P<.05) than in the absence of an uptake blocker. At 3 Hz, NE overflow in response to L-NNA (S4/S2) was 123.2±24.8% (n=4), which is not significantly different from data obtained without an uptake blocker.
Similar to L-NNA, the NOS inhibitor L-NMA (0.1 mmol/L) significantly enhanced stimulation-evoked NE overflow at both 1 Hz (S3/S1: 150.9±18.8%; n=4, P<.01) and 3 Hz (S4/S2: 117.9±11.1%; n=4, P<.05).
To mimic the vasoconstrictor effects of the two NOS inhibitors, 0.1 U/L [Arg8]-vasopressin was added 24.5 minutes before SNS3. Evoked NE overflow at both 1 Hz (S3/S1: 55.3±1.9%) and 3 Hz (S4/S2: 67.9±5.5%) did not differ from control values but did differ significantly (P<.001) from experiments using L-NNA (Table⇑).
Effects of SNAP and SIN-1 on Coronary Perfusion Pressure
Infusion of the NO-donor compound SNAP (1 μmol/L) in the presence of L-NNA (0.1 mmol/L) resulted in a rapid decrease in CPP by about 75 mm Hg. There was no effect of SNAP on contractility and beating frequency of the right atrium and ventricle. Similar effects were seen with SIN-1 (1 μmol/L).
Effects of SNAP and SIN-1 on the Evoked NE Overflow
SNAP, when infused into the coronary system starting 5 minutes before SNS3, SNS4, and SNS5 in concentrations from 1 nmol/L to 10 μmol/L, inhibited the evoked NE overflow (Fig 3⇓). SNAP (1 μmol/L) was used in subsequent experiments.
Infusion of SNAP (1 μmol/L), starting 5 minutes before SNS3 and SNS4, significantly inhibited the evoked NE overflow at 1 Hz (90 pulses, S3/S1) and at 3 Hz (270 pulses, S4/S2; Fig 4⇓). SIN-1, another NO-donor compound, produced a similar inhibition at both stimulation frequencies (Fig 4⇓).
Hemodynamic Effects of Endothelial Damage With the Detergent CHAPS
Perfusion of the hearts with CHAPS for 10 minutes increased CPP by about 30 mm Hg, which is compatible with a functional impairment of endothelial NO release (Fig 5⇓). A higher CHAPS concentration (5 mmol/L) was more effective in increasing CPP (up to 150 mm Hg during the 10-minute perfusion; not shown) but impaired the contractility of the heart and the viability of the sympathetic neurons (as evidenced by an increased basal outflow of NE). ACh (0.1 μmol/L) was infused for 2 minutes before and after perfusion of the heart with CHAPS (1 mmol/L). Before CHAPS treatment, ACh caused a decrease in CPP by up to 40 mm Hg; after CHAPS treatment an increase in CPP by up to 30 mm Hg was obtained (Fig 5⇓).
Impairment of Endothelial NO Release Does Not Alter NE Overflow
Treatment of the hearts with CHAPS (1 mmol/L) did not alter basal outflow of NE compared with untreated hearts (63.8±8.3 pg/min versus 59.5±9.0 pg/min; mean±SEM, n=8). Also, the release of NE in response to electrical stimulation did not vary from that of control experiments (Fig 6⇓). Also, in CHAPS-treated hearts, L-NNA (0.1 mmol/L) still facilitated significantly the fractional overflow of NE (Fig 6⇓).
Identification of NOS Isoforms by Western Blot
Western blotting of proteins from the whole rat heart (partially purified on 2′,5′-ADP-Sepharose) demonstrated weak bands for neuronal NOS I at 160 kD and endothelial NOS III at 135 kD. No immunoreactivity for the inducible NOS II was detected in these normal hearts (Fig 7⇓).
The stellate ganglion, the major sympathetic ganglion supplying fibers to the heart, showed NOS I immunoreactivity in preganglionic fibers and postganglionic neurons. Virtually all stellate ganglion cells exhibited light to moderate immunoreactivity to NOS I (Fig 8A⇓). In addition, there were numerous varicose NOS I immunoreactive fibers, some of which appeared to surround ganglionic cells (Fig 8A⇓). Like the NOS I immunoreactivity, all ganglion cells had moderate to intense TH immunoreactivity (Fig 8B⇓). Double-labeling experiments showed that the majority of ganglion cells exhibited both NOS I and TH immunoreactivity (Fig 8⇓).
Double-labeling immunofluorescence using antisera to NOS I and TH demonstrated NOS I and TH immunoreactivity in neurons innervating the right atrium of the rat heart (Fig 9A⇓ and 9B⇓). Combined NOS I and TH immunoreactivity was also seen in clusters of subepicardial neuronal cells (Fig 9C⇓ and 9D⇓).
The present study demonstrates that endogenous and exogenous NO inhibits the evoked NE release from rat heart sympathetic nerves. To our knowledge, this behavior has not been demonstrated for the heart so far. The influence of NO on sympathetic neurotransmission has been investigated in several vascular preparations, but only postjunctional effects were reported.17 18 19 20 However, prejunctional effects of NO on NE release have been shown in at least two tissues. In the canine kidney, endogenous NO seems to act as an inhibitory modulator of renal sympathetic nerve activity at both prejunctional and postjunctional sites.23 In the canine mesenteric artery, NO-donor compounds such as SIN-1, nitroglycerin, and sodium nitroprusside have been shown to inhibit the evoked NE release via stimulation of soluble guanylyl cyclase.22
Autoinhibition of NE release via prejunctional α2-autoreceptors may mask other inhibitory mechanisms, such as the NO mechanism described here. Therefore, appropriate stimulation frequencies that minimize autoinhibition had to be selected (for review see Reference 3535 ). In perfused rat hearts stimulated at frequencies ≤3 Hz, the biophase concentration of NE is too low to activate autoinhibition.36 Therefore, frequencies of 1 Hz and 3 Hz were used in the present study.
In the rat mesenteric vasculature, L-NNA decreased (rather than increased) the evoked NE overflow.37 38 However, this may have been caused by vascular effects of the inhibitor. The vasoconstriction produced by L-NNA may have modified tissue perfusion so that more time was available for neuronal reuptake of NE. Similarly, in a study on the isolated perfused rabbit heart, hemoglobin, by scavenging endothelial NO, decreased the efflux of NE without changing its release from the nerves.39
In our study, the rat hearts were perfused at a constant rate of 5 mL/min. Because the NOS inhibitor L-NNA still enhanced NE release in the presence of the uptake inhibitor desipramine, changes in neuronal uptake of NE are unlikely to explain the L-NNA effects in the rat heart.
In the current study, evidence for a negative control of endogenous NO over the stimulated NE release was obtained with the two independent NOS inhibitors L-NNA and L-NMA. NG-nitro-l-arginine methyl ester, another common NOS inhibitor, was not used, because alkyl esters of l-arginine have been described as muscarinic receptor antagonists40 and prejunctional M2 receptors on sympathetic nerve terminals inhibit NE release from rat heart.41
[Arg8]-vasopressin was used to mimic the vasoconstriction produced by the NOS inhibitors. Vasopressin was selected because it lacks prejunctional effects on NE release.42 Although [Arg8]-vasopressin (0.1 U/L) increased CPP more than did L-NNA or L-NMA, evoked NE overflow was not different from control experiments. These findings suggest that the observed facilitation of evoked NE release by the NOS inhibitors is not due to nonspecific effects such as ischemia.
The inhibitor experiments were corroborated by experiments with the NO donors SNAP and SIN-1. These compounds are chemically dissimilar, and their only common denominator is NO release.43 44 Both compounds were able to inhibit evoked NE release. NO had no significant effect on basal release of NE. This finding is at variance with a recent report suggesting a 26% increase in catecholamine content of heart neurons after a 3-minute treatment with NO (10 nmol/L).26 However, given the basal NE overflow within 3 minutes of <0.05% of the NE content (see “Results”), a measurable change in NE content seems to be unlikely in such a short period of time.
In the last part of this work, we attempted to characterize the predominant source of the endogenous NO controlling NE release. Western blots showed the presence of both neuronal NOS I and endothelial NOS III in the rat heart. A recent report suggested that cardiomyocytes can express NOS III.45 However, expression in these cells seems to be low, because our immunohistochemical staining for NOS III protein was negative in rat cardiomyocytes but positive in the coronary endothelium of the same hearts (not shown). In order to impair endothelial NO production, we perfused the hearts with the nondenaturing detergent CHAPS (1 mmol/L, 10 minutes). This treatment increased CPP and abolished the ACh-induced vasodilation that is mediated by endothelium-derived NO. At the same time, cardiac contractility and rhythm were not significantly changed. Also, the facilitatory effect of L-NNA on evoked NE overflow was still present, pointing to nonendothelial NO as being responsible for the repression of NE release.
Immunohistochemistry demonstrated numerous NOS I–positive nerve fibers in the rat heart, some of which costained for TH. The origin of NOS I–positive fibers in the rat heart is not entirely clear. Some may arise from intrinsic NOS I–positive neurons in the heart and others from sympathetic postganglionic neurons in the stellate ganglion. Sympathetic postganglionic neurons in this ganglion exhibited staining of NOS I. Furthermore, colocalization experiments showed that NOS I and TH immunoreactivity coexist in the majority of sympathetic postganglionic neurons of the rat stellate ganglion, as has been previously described for the bovine superior cervical ganglion.10
We conclude from the present experiments that NO of predominantly neuronal origin exerts a negative control over the evoked NE release from sympathetic nerves in the rat heart. This may occur in a paracrine fashion (with NO being generated in nitrergic nerves adjacent to the sympathetic nerves) or in an autocrine fashion (with NO being an atypical cotransmitter in sympathetic neurons themselves). A potential functional significance of this control mechanism for cardiac function in vivo remains to be determined.
Selected Abbreviations and Acronyms
|CPP||=||coronary perfusion pressure|
|NOS||=||nitric oxide synthase|
|NOS I||=||neuronal NOS|
|NOS II||=||inducible NOS|
|NOS III||=||endothelial NOS|
|SNS||=||sympathetic nerve stimulation|
This work was supported by grants Fo 144/3-1 and Fo 144/4-1 from the Deutsche Forschungsgemeinschaft, Germany (Dr Förstermann) and NIH grants NS-18710 and NS-24226 from the Department of Health and Human Services (Dr Dun). We thank Dr Ingolf Gath for helpful advice and Ute Gödtel-Armbrust for expert technical assistance with the immunohistochemistry. Suggestions from Professor Erich Muscholl, Mainz, are gratefully acknowledged.
- Received March 9, 1995.
- Accepted June 22, 1995.
- © 1995 American Heart Association, Inc.
Bredt DS, Snyder SH. Isolation of nitric oxide synthetase, a calmodulin-requiring enzyme. Proc Natl Acad Sci U S A. 1990;87:682-685.
Schmidt HH, Pollock JS, Nakane M, Gorsky LD, Förstermann U, Murad F. Purification of a soluble isoform of guanylyl cyclase-activating-factor synthase. Proc Natl Acad Sci U S A. 1991;88:365-369.
Hryhorenko LM, Woskowska Z, Foxthrelkeld JET. Nitric oxide (NO) inhibits release of acetylcholine from nerves of isolated circular muscle of the canine ileum: relationship to motility and release of nitric oxide. J Pharmacol Exp Ther. 1994;271:918-926.
Egi Y, Matsumura Y, Murata S, Umekawa T, Hisaki K, Takaoka M, Morimoto S. The effects of N-G-nitro-l-arginine, a nitric oxide synthase inhibitor, on norepinephrine overflow and antidiuresis induced by stimulation of renal nerves in anesthetized dogs. J Pharmacol Exp Ther. 1994;269:529-535.
Schmidt HHHW, Gagne GD, Nakane M, Pollock JS, Miller MF, Murad F. Mapping of neural nitric oxide synthase in the rat suggests frequent co-localization with NADPH diaphorase but not with soluble guanylyl cyclase, and novel paraneural functions for nitrinergic signal transduction. J Histochem Cytochem. 1992;40:1439-1456.
Pollock JS, Nakane M, Buttery LDK, Martinez A, Springall D, Polak JM, Förstermann U, Murad F. Characterization and localization of endothelial nitric oxide synthase using specific monoclonal antibodies. Am J Physiol. 1993;265:C1379-C1387.
Klimaschewski L, Kummer W, Mayer B, Couraud JY, Preissler U, Philippin B, Heym C. Nitric oxide synthase in cardiac nerve fibers and neurons of rat and guinea pig heart. Circ Res. 1992;71:1533-1537.
Förstermann U, Pollock JS, Schmidt HH, Heller M, Murad F. Calmodulin-dependent endothelium-derived relaxing factor/nitric oxide synthase activity is present in the particulate and cytosolic fractions of bovine aortic endothelial cells. Proc Natl Acad Sci U S A. 1991;88:1788-1792.
Tracey WR, Pollock JS, Murad F, Nakane M, Förstermann U. Identification of an endothelial-like type III NO synthase in LLC-Pk1 kidney epithelial cells. Am J Physiol. 1994;266:C22-C28.
Tracey WR, Xue C, Klinghofer V, Barlow J, Pollock JS, Förstermann U, Johns RA. Immunochemical detection of inducible NO synthase in human lung. Am J Physiol. 1994;266:L722-L727.
Buxton ILO, Cheek DJ, Eckman D, Westfall DP, Sanders KM, Keef KD. NG-nitro l-arginine methyl ester and other alkyl esters of arginine are muscarinic receptor antagonists. Circ Res. 1993;72:387-395.
Böhme E, Grossmann G, Herz J, Mülsch A, Spies C, Schultz G. Regulation of cyclic GMP formation by soluble guanylate cyclase: stimulation by NO-containing compounds. In: Greengard, ed. Advances in Cyclic Nucleotide and Protein Phosphorylation Research 17. New York, NY: Raven Press; 1984:259-266.
Feelisch M. The biochemical pathways of nitric oxide formation from nitrovasodilators: appropriate choice of exogenous NO donors and aspects of preparation and handling of aqueous NO solutions. J Cardiovasc Pharmacol. 1991;17(suppl 3):S25-S33.
Balligand JL, Kelly RA, Smith TW, Michel T. Identification of a constitutive isoform of NO synthase in adult rat cardiac myocytes. Circulation. 1994;90(suppl I):I-193. Abstract.