Research Paper J Vase Res 1992:29:2-7
James Quillen Frank Sellke Peter Banitt David Harrison
The Effect of Norepinephrine on the Coronary Microcirculation
Cardiovascular Center and Veterans Administration Medical Center, Departments of Surgery and Internal Medicine, University of Iowa College of Medicine, Iowa City. Iowa. USA
K e y w o rd s
A b stra ct
Norepinephrine Catecholamines Coronary arteries Endothelium-derived relaxing factor Isoproterenol Coronary microcirculation
The role of the sympathetic nervous system in the regulation of large coronary artery tone has been well defined. Studies of adrenergic regulation of coro nary-resistance vessels have largely been limited to indirect inferences based on flow measurement obtained in vivo. The purpose of the present study was to determine the effects of norepinephrine (NE) on the coronary microcircula tion using direct in vitro approaches. Porcine coronary microvessels (80-200 pm in diameter) were pressurized in isolated organ chambers. Diameters were measured using a Halpern microvessel imaging apparatus. After preconstric tion with leukotriene D4, NE caused complete relaxation. Relaxations to NE were inhibited by propranolol. Relaxations to NE were also inhibited by LY83583 (which depletes cGMP) and hemoglobin (which binds endotheliumderived relaxing factor. EDRF). N E caused minimal or no constriction in both preconstricted and nonpreconstricted microvessels even in the presence of hemoglobin and propranolol. In conclusion, NE predominantly dilates por cine coronary microvessels, both by P-adrenoceptor activation and by stimu lating release of EDRF. There is minimal a-adrenoceptor-mediated constric tion of coronary microvessels.
Norepinephrine (NE) released from sympathetic nerves has a significant influence of coronary vascular tone and myocardial perfusion. Although previous stud ies on large coronary arteries have demonstrated direct a-adrenergic receptor-mediated vasoconstriction [Feigl 1967; Zuberbuhler and Bohr, 1965; Vatner et al., 1980; Johannsen et al.. 1982; Young and Vatner, 1986; Cohen et al., 1983; Kuramoto et al., 1967; Holtz et al., 1977], studies on small coronary vessels have generally found
Received: Ma> 15. 1990 Accepted by Blood Vessels: December 14, 1990
minimal a-adrenoceptor-mediated constriction [Zuberbuhler et al., 1965; Nakayama et al., 1988; Mekata and Niu. 1969; Anderson et al.. 1972; Chilian et al., 1989]. Most previous studies on the adrenergic control of the coronary microcirculation have been performed with in J.Q. is a recipient o fa n individual NRSA from the NIH. F.S. is a recipient of a NHLBI Institutional Research Fellowship from the Cardiovascular Center, University of Iowa. D.H. is an Established Investigator of the AHA. P.B. was the recipient of an AHA Medical Student Research Fellowship. Supported by NIH Grants HL327I7. HI.201104ft, Ischemic SCOR HL32295. and a Merit Review Grant from the Veterans Administration.
David G. Harrison. MD. Professor of Medicine Cardiology Division. Department o f Internal Medicine Emory University Atlanta. GA 30322 (USA)
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
Introduction
M eth o d s Vessel Preparation Porcine hearts were attained from a slaughter house immediately after sacrifice. Hearts were removed and placed in cold Krebs buffer (118.3 mA/NaCI. 4.7 m.l/KCl. 2.5 mA/CaCk 1.2 mA/MgS04. 1.2 mA/ K H :P04. 25 mA/ NaHCOj. 11.1 mA/ glucose, adjusted to pH 7.4) and transported to the laboratory. Epicardial microvessels from the left anterior descending artery were carefully dissected using a forty-power dissecting microscope. Microvessels had internal diame ters of 80-200 pm and were 0.5-1 mm in length. All vessels were removed from the anterior free wall of the left ventricle, midway between the atrioventricular ring and the apex. The vessels were usually obtained by simply dissecting into the myocardium in this region until a microvesscl of the size desired was located. The vessels were subsequently placed in an isolated plexiglas organ chamber, eannulated with dual-glass micropipettes measuring 50-70 pm in diameter and secured with 10-0 nylon monofilament suture. Blood products were washed from the lumen of the microvessels before study with Krebs buffer. Oxygenated (95% O:. 5% CO?) Krebs buffer warmed to 37 °C was continuously circulated through the organ chamber. A pressure transducer monitored distending pres sures. The optimal distending pressure for constriction to KCI was found to occur at 20 mm Hg. thus the microvessels were maintained at this pressure for the remainder of the experiment. Using a micro scope equipped with an image-splitting device and connected to a video camera, the vessel image was projected on to a television mon itor. A videoelectronic dimension analyzer was used to measure luminal diameter, as described by Halpern et al. [1984]. Measure ments were recorded with a strip chart recorder. After at least I h of
equilibration, the microvessels were preconstrictcd with leukotrienc D4 to 30-60% of the resting baseline diameter. Stable prcconstriction with leukotrienc D4 was verified before dose-response curves were performed. In some vessels, both a control response to NE was examined, followed by a response to NE in the presence of an inhibi tor. Between these two concentration-response curves, at least 15 min was allowed to ellapsc and the vessels washed at least three times. In preliminary experiments, relaxations to two repeated con centration-response curves in the same vessel were found to be simi lar. Following a concentration-response intervention in which an antagonist was used, the vessel was discarded. Drugs The following pharmacologic agents were used: /.-norepineph rine hydrochloride, /.-isoproterenol hydrochloride. //,/.-propranolol hydrochloride, yohimbine hydrochloride were obtained from Sigma (St. Louis. Mo.. USA). Leukotrienc D4 was a gift from Merck Frosst Canada. Inc. (Dorval. Quebec. Que., Canada). LY83583 was pro vided by Eli Lilly Inc. (Indianapolis. Ind.. USA). Drugs were pre pared daily and were dissolved in distilled water. 0 .1% ascorbic acid was added to solutions containing NE and isoproterenol to prevent oxidation. Dcsaturated bovine hemoglobin solution (Sigma) was pre pared by dissolving 3.22 g free hemoglobin and 0.87 g sodium diethionate to 50 ml of distilled water. This solution was dialyzed with 5 I of distilled water for 2 h bubbled with nitrogen gas. Hemoglobin solutions were not stored beyond 1 day. Oxygen saturation was con firmed to be less than 20% by co-oximetry. Data Analysis Relaxation responses of the coronary microvesscls were ex pressed as percent relaxation from their preconstricted diameters. ICso's are expressed as log of the molar concentration. Mean responses at each dose of each drug and their respective ICjo’s were compared using analysis of variance. Whenever significance was indicated, Sheffe's F test for multiple comparison was used to com pare between groups. Values were expressed as mean ± SE. Signifi cance was assumed if p < 0.05.
Results
The mean microvessel diameter was 129 ± 18 pm with a range of 80-200 jam at 20 mm Hg distending pres sure. NE-induced constriction of coronary microvessels was examined in the presence of propranolol to block (3adrenergic dilatation. The results of these studies are sum marized in table 1. In the absence of pre-existing tone, NE (1 nA/to 1 pM ) produced constriction in only one of six experiments. With the addition of hemoglobin, to inacti vate EDRF, and induction of pre-existing tone, NE caused a minimal amount of constriction (2-3% of maxi mal KCI constriction) in 43% of the vessels. In contrast to the relative lack of a-adrenergic constric tion observed in coronary microvessels. NE (1 nA/ to 100 pM) caused potent, concentration-dependent relaxation of coronary microvessels preconstricted with leukotriene
3
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
vivo experiments, in which autoregulatory and metabolic influences may play a role despite efforts to minimize these effects. In these studies, vasodilatory effects evident with intracoronary infusion of NE or sympathetic-nerve stimulation may at least in part be due to metabolic fac tors as a result of (51-adrenergic-receptor activation [Vatner et al., 1986: Vatner et al., 1982; Mohrman et al.. 1978; Moreland and Bohr, 1984: Baron et al.. 1972; Johannson. 1973: Drew and Levy. 1972: McRaven et al.. 1971; Hamilton and Feigl. 1976], In addition to the effects of NE on coronary microvessel adrenergic recep tors on smooth muscle, an independent effect on microvascular endothelium must be considered. NE has been demonstrated to release endothelium-derived relaxing factor (EDRF) in the large coronary arteries [Cocks and Angus, 1983]. However, a similar effect has not been shown in the coronary microcirculation. The goal of the study was to determine the actions of NE on porcine coronary microvessels in vitro, free from metabolic and autoregulatory influences. In addition, the role of the endothelium in NE induced adrenergic modu lation of coronary microvascular tone was investigated.
Table 1. Con striction in response to NE (1 p.U)
Nonconstrictcd Preconstricted
a
Table 2.
Relaxation in response to NE
Number of vessels constricted with NE
Mean constriction“
1/6(16.7%)
propranolol (10 -6 47) propranolol ( l(H' 47 ) and hemoglobin ( 10 -5 M )
0 / 6 (0 %)
1.3 ± 1.1 % 0±0%
propranolol (10 ~6 47) propranolol (10 6 47)and hemoglobin (10- 5 .47)
3/7(42.9%) 3/7(42.9%)
2.5 ±1.1 % 2 .0 ± 0 .8 %
Asa percent of maximal KCI constriction (100 m47 KCI).
Vessel
n
- log ED50 % LTD4 preconstriction
Maximum relaxation, %
Control Propranolol (1 p47) Hemoglobin (10 p47) LY83583 (1 p47) Propranolol (1 p.V/) + yohimbine (1 p47) Propranolol (1 p.47) + hemoglobin (10 p47)
9 7 5
52 ±2 41 ±4 39 ±4 53 ±4 43±4 53±4
100
5 5 6
7.33 + 0.16 5.54 ±0.16* 6.47 ±0.23* 6.64 ±0.24* 4.90 ±0.28* 4.86 ±0.23*
98+1 100 100 100
91 ± 6
*p < 0.05 vs. control.
4
The relaxations to NE seemed independent of the pre constricted tone developed in response to LTD4. There were no differences in preconstricted tone between all of the treatment groups (table 2). Further, there was no cor relation between the IC50 to NE and the preconstricted tone for the various subgroups (r2 = 0.002-0.4) or for all vessels as a group (r2 = 0.014).
D iscussion
This study demonstrated that NE elicits relaxation of porcine coronary microvessels (80-200 pm) in vitro by both a P-adrenergic mechanism and by release of EDRF. a-Adrenoceptor-mediated constriction of this size class coronary vessel was minimal. Coronary microvessels of the size used in the present study have been shown to be true resistance vessels in other species. Nellis et al. [ 1981 ] and Chilian et al. [ 1989] studied vessels of this size in the rabbit heart and cat heart, respectively. Both studies demonstrated a signifi cant loss of arterial pressure throughout arteries of this size. Thus, it is reasonable to assume that porcine coro nary arteries of this size also contribute to coronary rcsis-
Quillcn/Scllke/Banitt/Harrison
NE and the Coronary Microcirculation
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
D 4 (fig. 1). The peak relaxation was 100% of the preconstricted tension with an IC50 o f -7.33 ± 0.16 M (table 2). This relaxation appeared in part related to the activation of (3-adrenergic receptors, as preincubation with 1 \lM propranolol markedly shifted the relaxation response to NE rightward (fig. 1). The relaxation to NE also appeared related to the release of EDRF. The addition of hemoglo bin (10 \iM ) to inactivate EDRF moderately shifted the dose-response curve rightward (fig. 2. table 2). Similarly, the addition of LY83583 (1 11 M ), to deplete vascular cGMP. shifted the dose-response curve to the right. The inhibition of NE-induced relaxation was not a nonspecific effect of hemoglobin, as hemoglobin had no effect on the relaxation of coronarv microvessels to isoproterenol (fig. 3). Endothelium-dependent relaxation to NE has been suggested to occur in large vessels as a result of cb-adrenergic receptors. Incubation with both hemoglobin (10 \iM ) and propranolol (1 \iM ) or yohimbine (an uy-adrenergic antagonist. 1 p.W) and propranolol (1 pM) similarly inihibitcd the NE-induced vascular relaxation (fig. 4, ta ble 2). This inhibition approached but did not achieve sig nificance compared to the inhibition achieved by pro pranolol alone.
Log [Norepinephrine]
Fig. 3. Average relaxations from baseline of porcine coronary microvessels to isoproterenol without (•: n = 7) and with hemoglobin (o; 10 (iA/t n = 6). Hemoglobin does not alter relaxations to isoproter enol. Values are expressed as mean ± SE.
Lo9 [Norepinephrine]
Fig. 2. Average relaxations from baseline of porcine coronary microvcsscls to NE without (•: n = 9) and with LY83583 (□: l pA/: n = 5) and hemoglobin ( a ; 10 pA/; n = 5). LY83583 and hemoglobin moderately blunt relaxations to NE. Values are expressed as mean ± SE. p < 0.05 vs. control.
Fig. 4. Average relaxations from baseline of porcine coronary microvesseis to NE with propranolol (•: I p.W; n = 7) and with or without hemoglobin (□: 10pA7;n = 6)oryohimbine(A; l pA7; n = 5). Hemoglobin and yohimbine mildly blunt relaxations (nonsignifi cant) to NE. Values are expressed as mean ± SE.
lance lo a significant degree. Further, vessels of this size class demonstrate several important differences in re sponse to pharmacologic [Lamping et al., 1989; Kanatsuka ct al., 1988; Chilian et al., 1989] and physiologic [Kanatsuka et al., 1989] stimuli compared to larger ves sels.
In the present study, minimal a-adrenoaceptor-mediated vasoconstriction was found in both quiescent and preconstricted microvesseis treated with (i-blockade with or without inhibition of EDRF with hemoglobin (to elim inate basal EDRF relaxation). This is in contrast to in vivo studies using sympathetic nerve stimulation [Mark
5
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
Fig. 1. Average relaxations from baseline of porcine coronary microvcsscls to NE without (•: n = 9) and with propranolol (o; I p.l/: n = 7). Propranolol markedly blunts relaxations to NE. Values are expressed as mean ± SE. p < 0.05 vs. control.
et al.. 1972; McRaven et al., 1971; Moreland et al., 1984; Kelley and Feigl. 1978] or NE infusion [Mark et al.. 1972; Kelley and Feigl. 1978; Vainer et al.. 1980] in which a transient a-adrenoceptor-mcdiatcd increase in vascular tone of coronary resistance vessels and reduced coronary flow was observed. Other investigators using in vitro preparations of smaller epicardial coronary arteries have reported findings similar to those in the present study [Zuberbuhler and Bohr. 1965; Mekata and Niu, 1969; Anderson et al., 1972; Nakayama et al., 1988], The lack of a-adrenoceptor-mediated vasoconstriction in epicardial microvessels, in the present study, may be due to one of several factors. First, the distribution of a-adrenoceptors is likely heterogeneous throughout the coronary circula tion [Cohen et al., 1983. 1984]. Thus, vasoconstriction may occur in vessels substantially larger or smaller than those analyzed in the present study. Secondly, the trans mural distribution of a-adrenoceptors may be different in the various size class vessels, with a greater proportion of functional a-adrenoceptors in the endocardial vessels ver sus the epicardial vessels which are examined in this study. Thirdly, with in vivo experiments, metabolic and autoregulatory influences may complicate the results de spite attempts to minimize the effects. Finally, in the case of sympathetic nerve stimulation, other vasoactive sub stances such as neuropeptide Y may be coreleased with NE which may independently alter vascular tone. It is unlikely that the low distending pressure em ployed in these studies accounted for the lack of constric tor responses to NE. In a large number of preliminary experiments, we altered perfusion pressure over a wide range (10-60 mm Hg). without observing a difference in the response to NE. Further, we have repeatedly found
that the optimal distending pressure for the development of isotonic contraction (which is the parameter measured using this preparation) to be 20 mm Hg. This is perhaps due to the fact that at higher distending pressures, the ves sels begin to generate greater amounts of isometric ten sion (which would be measured using a through the lumen wire or stirrup approach) and smaller amounts of isotonic tension. Another likely explanation is that once the vessel is removed from the myocardium and its external sup porting structures, the external forces upon the vessel are removed, and thus the internal pressure of 20 mm Hg is unopposed by an outside counterpressure. Under these circumstances, the transmural pressure in this prepara tion may actually approximate the normal physiologic distending pressures. This is particularly likely because the in vivo pressure within these vessels has been docu mented to be approximately 50% of aortic pressure, or only about 40-50 mm Hg during diastole. NE may elicit release of EDRF from large isolated arteries [Cocks and Angus, 1983]. In the present study, the relaxation of microvessels in response to NE was inhibited by both hemoglobin and by LY83583. demon strating the contribution of EDRF release in NE-induced relaxation. Hemoglobin failed to inhibit the relaxation of microvessels by the nonselective p-adrenergic agonist iso proterenol, showing that in the coronary microcirculation EDRF release is not through a p-adrenergic mechanism. In summary, NE causes vasodilation of porcine coro nary microvessels in vitro, free from metabolic and autoregulatory influences. Minimal a-adrenoccptor-mediated vasoconstriction occurs in response to NE. NE, in addi tion to its P-receptor-mediated relaxation, induces release of EDRF in the porcine coronary microcirculation.
Anderson. R.: Holmberg. S.: Svedmyr. N.: Aberg, G.: Adrenergic a- and (5-receptors in coronary vessels in man. An in vitro study. Acta tried, scand. /9 /. 241-244 (1973). Baron. G.D.: Speden. R.N.: Bohr. D.F.: Betaadrenergic receptors in coronary and skeletal muscle arteries. Am. J. Physiol. 223: 878-881 (1972). Chilian, W.M.; Layne. S.M.; Eastham, C.L.; Mar cus, M.L.: Heterogeneous microvascular coro nary a-adrenergic vasoconstriction. Circula tion Res. 64: 376-388 (1989). Cocks. T.M.: Angus. J.A.: Endothelium-dependent relaxation of coronary arteries by noradrena line and serotonin. Nature 305: 627-630 (1983).
Cohen. R.A.;Shepherd. J.T.; Vanhoutte, P.M.: Pre junctional and postjunctional actions of endog enous norepinephrine at the sympathetic neu roeffector junction in canine coronary arteries. Circulation Res. 52: 16-25 (1983). Cohen. R.A.; Shepherd. J.T.: Vanhoutte. P.M.: Ef fects of the adrenergic transmitter on epicardial coronary arteries. Fed. Proc. 43: 2862-2866 (1984). Drew. G.M.: Levy. G.P.: Characterization of the coronary vascular (5-adrenoceptor in the pig. Br. J. Pharmacol. 46: 348-350 (1972). Feigl, E.O.: Sympathetic control of coronary circu lation. Circulation Res. 20: 262-271 (1967).
Halpern. W.; Osol. G.; Coy. G.: Mechanical behav ior of pressurized in vitro prcarteriolar vessels determined with a video system. Ann. biomed. Eng. 121:463-479 (1984). Hamilton. F.N.; Feigl. E.O.: Coronary vascular sympathetic beta-receptor innervation. Am. J. Physiol. 230: 1569-1576(1976). Holtz. J.: Mayer. E.: Bassenge. E.: Demonstration of alpha-adrenergic coronary control in differ ent layers of canine myocardium by regional myocardial sympathectomy. Piliigers Arch. 372: 187-194(1977). Johannson. B.: The (5-adrenoceptors in the smooth muscle of pig coronary arteries. Eur. J. Phar macol. 24: 2 18-224 (1973).
6
Quillcn/Sellke/Banitt/Harrison
NE and the Coronary Microcirculation
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
R eferences
Mark. A.L.; Abboud. F.M.: Schmid. P.G.: Heistad. D.D.: Mayer. H.E.: Differences in direct effects o f adrenergic stimuli on coronary', cutaneous and muscular vessels. J. clin. Invest. 51: 279287(1972). McRaven. D.R.: Mark. A.L.: Abboud. F.M.; Mayer, H.E.: Responses of coronary vessels to adrenergic stimuli. J. clin. Invest. 50:1T3-11Ü (1971). Mekata, LL: Niu, H.: Electrical and mechanical responses of coronary artery smooth muscle to catecholamines. Jap. J. Physiol. 19: 599-608 (1969). Mohrman. D.E.; Feigl. E.O.: Competition between sympathetic vasoconstriction and metabolic vasodilation in the canine coronary circulation. Circulation Res. 42: 79-86 (1978). Moreland, R.S.; Bohr, D.F.: Adrenergic control of coronary arteries. Fed. Proc. 43: 2857-2861 (1984).’ Nakayama, K.: Osol, G.: Halpern. W.: Reactivity of isolated porcine coronary' resistance arteries to cholinergic and adrenergic drugs and trans mural pressure changes. Circulation Res. 62: 741-748(1988).
Nellis. S.; Liedtke. J.: Whitesell. L.: Small coronary vessel pressure and diameter in an intact beat ing rabbit heart using fixed-position and treemotion techniques. Circulation Res. 49: 342353(1981). Vatncr, S.F.: Pagani, M.: Manders. W.T.: Pasipoularides. A.D.: Alpha adrenergic vasoconstric tion and nitroglycerin vasodilation of large cor onary arteries in the conscious dog. J. clin. Invest. 65: 5-14 (1980). Vatncr. S.F.: Hintze. T.H.: Macho, P.: Regulation of large coronary arteries by p-adrencrgic mechanisms in the conscious dog. Circulation Res. 51:56-66 (1982). Vatner, D.E.; Knight. D.R.; Homey, C.J.: Vatner. S.F.; Young, M.A.: Subtypes of (3-adrcnergic receptors in bovine coronary arteries. Circula tion Res. 59. 463-473 (1986). Young, M.A.: Vatner, S.F.: Regulation of large cor onary arteries. Circulation Res. 59: 579-596 (1986). Zuberbuhler. R.C.: Bohr D.F.: Responses of coro nary smooth muscle to catecholamines. Circu lation Res. 16:431 -440 (1965).
7
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
Johannson, U.J.; Mark, A.L.: Marcus. M.L.: Re sponsiveness lo cardiac sympathetic nerve stimulation during maximal coronary dilation produced by adenosine. Circulation Res. 50; 510-517(1982). Kanatsuka. H.; Lamping, K.G.; Eastham. C.L.: Marcus. M.L.: Non-uniform effect of nitroglyc erin on different size coronary arterial micro vessels. (Abstract) FASEB J. 2: A496 ( 1988). Kanatsuka, H.; Lamping. K.G.: Eastham. C.L.: Dellsperger. K.C.: Marcus. M.L.: Comparison of the effects of i nereased M VO? and adenosi ne on the coronary microvascular resistance. Cir culation Res. 65: 1296-1305(1989). Kelley. K.O.: Feigl, E.O.: Segmental a-receptormediated vasoconstriction in the canine coro nary circulation. Circulation Res. 43: 908-917 (1978). Kuramoto, K.; Murata. K.; Fujii. K.; Kurihara. H.; Kimata. S.: Matsushita, S.: Kuramochi. M.; lkeda. M.: Nakao. K.: Sympathetic coronary vasoconstriction after adrenergic beta block ade. Tohoku J. exp. Med. 91;95-101 (1967). Lamping, K.G.: Kanatsuka. E.; Eastham. C.L.: Chilian. W.M.: Marcus, M.L.: Nonuniform va somotor responses of the coronary microcircu lation to serotonin and vasopressin. Circula tion Res. 65: 343-351 (1989).
James Quillen Frank Sellke Peter Banitt David Harrison
The Effect of Norepinephrine on the Coronary Microcirculation
Cardiovascular Center and Veterans Administration Medical Center, Departments of Surgery and Internal Medicine, University of Iowa College of Medicine, Iowa City. Iowa. USA
K e y w o rd s
A b stra ct
Norepinephrine Catecholamines Coronary arteries Endothelium-derived relaxing factor Isoproterenol Coronary microcirculation
The role of the sympathetic nervous system in the regulation of large coronary artery tone has been well defined. Studies of adrenergic regulation of coro nary-resistance vessels have largely been limited to indirect inferences based on flow measurement obtained in vivo. The purpose of the present study was to determine the effects of norepinephrine (NE) on the coronary microcircula tion using direct in vitro approaches. Porcine coronary microvessels (80-200 pm in diameter) were pressurized in isolated organ chambers. Diameters were measured using a Halpern microvessel imaging apparatus. After preconstric tion with leukotriene D4, NE caused complete relaxation. Relaxations to NE were inhibited by propranolol. Relaxations to NE were also inhibited by LY83583 (which depletes cGMP) and hemoglobin (which binds endotheliumderived relaxing factor. EDRF). N E caused minimal or no constriction in both preconstricted and nonpreconstricted microvessels even in the presence of hemoglobin and propranolol. In conclusion, NE predominantly dilates por cine coronary microvessels, both by P-adrenoceptor activation and by stimu lating release of EDRF. There is minimal a-adrenoceptor-mediated constric tion of coronary microvessels.
Norepinephrine (NE) released from sympathetic nerves has a significant influence of coronary vascular tone and myocardial perfusion. Although previous stud ies on large coronary arteries have demonstrated direct a-adrenergic receptor-mediated vasoconstriction [Feigl 1967; Zuberbuhler and Bohr, 1965; Vatner et al., 1980; Johannsen et al.. 1982; Young and Vatner, 1986; Cohen et al., 1983; Kuramoto et al., 1967; Holtz et al., 1977], studies on small coronary vessels have generally found
Received: Ma> 15. 1990 Accepted by Blood Vessels: December 14, 1990
minimal a-adrenoceptor-mediated constriction [Zuberbuhler et al., 1965; Nakayama et al., 1988; Mekata and Niu. 1969; Anderson et al.. 1972; Chilian et al., 1989]. Most previous studies on the adrenergic control of the coronary microcirculation have been performed with in J.Q. is a recipient o fa n individual NRSA from the NIH. F.S. is a recipient of a NHLBI Institutional Research Fellowship from the Cardiovascular Center, University of Iowa. D.H. is an Established Investigator of the AHA. P.B. was the recipient of an AHA Medical Student Research Fellowship. Supported by NIH Grants HL327I7. HI.201104ft, Ischemic SCOR HL32295. and a Merit Review Grant from the Veterans Administration.
David G. Harrison. MD. Professor of Medicine Cardiology Division. Department o f Internal Medicine Emory University Atlanta. GA 30322 (USA)
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
Introduction
M eth o d s Vessel Preparation Porcine hearts were attained from a slaughter house immediately after sacrifice. Hearts were removed and placed in cold Krebs buffer (118.3 mA/NaCI. 4.7 m.l/KCl. 2.5 mA/CaCk 1.2 mA/MgS04. 1.2 mA/ K H :P04. 25 mA/ NaHCOj. 11.1 mA/ glucose, adjusted to pH 7.4) and transported to the laboratory. Epicardial microvessels from the left anterior descending artery were carefully dissected using a forty-power dissecting microscope. Microvessels had internal diame ters of 80-200 pm and were 0.5-1 mm in length. All vessels were removed from the anterior free wall of the left ventricle, midway between the atrioventricular ring and the apex. The vessels were usually obtained by simply dissecting into the myocardium in this region until a microvesscl of the size desired was located. The vessels were subsequently placed in an isolated plexiglas organ chamber, eannulated with dual-glass micropipettes measuring 50-70 pm in diameter and secured with 10-0 nylon monofilament suture. Blood products were washed from the lumen of the microvessels before study with Krebs buffer. Oxygenated (95% O:. 5% CO?) Krebs buffer warmed to 37 °C was continuously circulated through the organ chamber. A pressure transducer monitored distending pres sures. The optimal distending pressure for constriction to KCI was found to occur at 20 mm Hg. thus the microvessels were maintained at this pressure for the remainder of the experiment. Using a micro scope equipped with an image-splitting device and connected to a video camera, the vessel image was projected on to a television mon itor. A videoelectronic dimension analyzer was used to measure luminal diameter, as described by Halpern et al. [1984]. Measure ments were recorded with a strip chart recorder. After at least I h of
equilibration, the microvessels were preconstrictcd with leukotrienc D4 to 30-60% of the resting baseline diameter. Stable prcconstriction with leukotrienc D4 was verified before dose-response curves were performed. In some vessels, both a control response to NE was examined, followed by a response to NE in the presence of an inhibi tor. Between these two concentration-response curves, at least 15 min was allowed to ellapsc and the vessels washed at least three times. In preliminary experiments, relaxations to two repeated con centration-response curves in the same vessel were found to be simi lar. Following a concentration-response intervention in which an antagonist was used, the vessel was discarded. Drugs The following pharmacologic agents were used: /.-norepineph rine hydrochloride, /.-isoproterenol hydrochloride. //,/.-propranolol hydrochloride, yohimbine hydrochloride were obtained from Sigma (St. Louis. Mo.. USA). Leukotrienc D4 was a gift from Merck Frosst Canada. Inc. (Dorval. Quebec. Que., Canada). LY83583 was pro vided by Eli Lilly Inc. (Indianapolis. Ind.. USA). Drugs were pre pared daily and were dissolved in distilled water. 0 .1% ascorbic acid was added to solutions containing NE and isoproterenol to prevent oxidation. Dcsaturated bovine hemoglobin solution (Sigma) was pre pared by dissolving 3.22 g free hemoglobin and 0.87 g sodium diethionate to 50 ml of distilled water. This solution was dialyzed with 5 I of distilled water for 2 h bubbled with nitrogen gas. Hemoglobin solutions were not stored beyond 1 day. Oxygen saturation was con firmed to be less than 20% by co-oximetry. Data Analysis Relaxation responses of the coronary microvesscls were ex pressed as percent relaxation from their preconstricted diameters. ICso's are expressed as log of the molar concentration. Mean responses at each dose of each drug and their respective ICjo’s were compared using analysis of variance. Whenever significance was indicated, Sheffe's F test for multiple comparison was used to com pare between groups. Values were expressed as mean ± SE. Signifi cance was assumed if p < 0.05.
Results
The mean microvessel diameter was 129 ± 18 pm with a range of 80-200 jam at 20 mm Hg distending pres sure. NE-induced constriction of coronary microvessels was examined in the presence of propranolol to block (3adrenergic dilatation. The results of these studies are sum marized in table 1. In the absence of pre-existing tone, NE (1 nA/to 1 pM ) produced constriction in only one of six experiments. With the addition of hemoglobin, to inacti vate EDRF, and induction of pre-existing tone, NE caused a minimal amount of constriction (2-3% of maxi mal KCI constriction) in 43% of the vessels. In contrast to the relative lack of a-adrenergic constric tion observed in coronary microvessels. NE (1 nA/ to 100 pM) caused potent, concentration-dependent relaxation of coronary microvessels preconstricted with leukotriene
3
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
vivo experiments, in which autoregulatory and metabolic influences may play a role despite efforts to minimize these effects. In these studies, vasodilatory effects evident with intracoronary infusion of NE or sympathetic-nerve stimulation may at least in part be due to metabolic fac tors as a result of (51-adrenergic-receptor activation [Vatner et al., 1986: Vatner et al., 1982; Mohrman et al.. 1978; Moreland and Bohr, 1984: Baron et al.. 1972; Johannson. 1973: Drew and Levy. 1972: McRaven et al.. 1971; Hamilton and Feigl. 1976], In addition to the effects of NE on coronary microvessel adrenergic recep tors on smooth muscle, an independent effect on microvascular endothelium must be considered. NE has been demonstrated to release endothelium-derived relaxing factor (EDRF) in the large coronary arteries [Cocks and Angus, 1983]. However, a similar effect has not been shown in the coronary microcirculation. The goal of the study was to determine the actions of NE on porcine coronary microvessels in vitro, free from metabolic and autoregulatory influences. In addition, the role of the endothelium in NE induced adrenergic modu lation of coronary microvascular tone was investigated.
Table 1. Con striction in response to NE (1 p.U)
Nonconstrictcd Preconstricted
a
Table 2.
Relaxation in response to NE
Number of vessels constricted with NE
Mean constriction“
1/6(16.7%)
propranolol (10 -6 47) propranolol ( l(H' 47 ) and hemoglobin ( 10 -5 M )
0 / 6 (0 %)
1.3 ± 1.1 % 0±0%
propranolol (10 ~6 47) propranolol (10 6 47)and hemoglobin (10- 5 .47)
3/7(42.9%) 3/7(42.9%)
2.5 ±1.1 % 2 .0 ± 0 .8 %
Asa percent of maximal KCI constriction (100 m47 KCI).
Vessel
n
- log ED50 % LTD4 preconstriction
Maximum relaxation, %
Control Propranolol (1 p47) Hemoglobin (10 p47) LY83583 (1 p47) Propranolol (1 p.V/) + yohimbine (1 p47) Propranolol (1 p.47) + hemoglobin (10 p47)
9 7 5
52 ±2 41 ±4 39 ±4 53 ±4 43±4 53±4
100
5 5 6
7.33 + 0.16 5.54 ±0.16* 6.47 ±0.23* 6.64 ±0.24* 4.90 ±0.28* 4.86 ±0.23*
98+1 100 100 100
91 ± 6
*p < 0.05 vs. control.
4
The relaxations to NE seemed independent of the pre constricted tone developed in response to LTD4. There were no differences in preconstricted tone between all of the treatment groups (table 2). Further, there was no cor relation between the IC50 to NE and the preconstricted tone for the various subgroups (r2 = 0.002-0.4) or for all vessels as a group (r2 = 0.014).
D iscussion
This study demonstrated that NE elicits relaxation of porcine coronary microvessels (80-200 pm) in vitro by both a P-adrenergic mechanism and by release of EDRF. a-Adrenoceptor-mediated constriction of this size class coronary vessel was minimal. Coronary microvessels of the size used in the present study have been shown to be true resistance vessels in other species. Nellis et al. [ 1981 ] and Chilian et al. [ 1989] studied vessels of this size in the rabbit heart and cat heart, respectively. Both studies demonstrated a signifi cant loss of arterial pressure throughout arteries of this size. Thus, it is reasonable to assume that porcine coro nary arteries of this size also contribute to coronary rcsis-
Quillcn/Scllke/Banitt/Harrison
NE and the Coronary Microcirculation
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
D 4 (fig. 1). The peak relaxation was 100% of the preconstricted tension with an IC50 o f -7.33 ± 0.16 M (table 2). This relaxation appeared in part related to the activation of (3-adrenergic receptors, as preincubation with 1 \lM propranolol markedly shifted the relaxation response to NE rightward (fig. 1). The relaxation to NE also appeared related to the release of EDRF. The addition of hemoglo bin (10 \iM ) to inactivate EDRF moderately shifted the dose-response curve rightward (fig. 2. table 2). Similarly, the addition of LY83583 (1 11 M ), to deplete vascular cGMP. shifted the dose-response curve to the right. The inhibition of NE-induced relaxation was not a nonspecific effect of hemoglobin, as hemoglobin had no effect on the relaxation of coronarv microvessels to isoproterenol (fig. 3). Endothelium-dependent relaxation to NE has been suggested to occur in large vessels as a result of cb-adrenergic receptors. Incubation with both hemoglobin (10 \iM ) and propranolol (1 \iM ) or yohimbine (an uy-adrenergic antagonist. 1 p.W) and propranolol (1 pM) similarly inihibitcd the NE-induced vascular relaxation (fig. 4, ta ble 2). This inhibition approached but did not achieve sig nificance compared to the inhibition achieved by pro pranolol alone.
Log [Norepinephrine]
Fig. 3. Average relaxations from baseline of porcine coronary microvessels to isoproterenol without (•: n = 7) and with hemoglobin (o; 10 (iA/t n = 6). Hemoglobin does not alter relaxations to isoproter enol. Values are expressed as mean ± SE.
Lo9 [Norepinephrine]
Fig. 2. Average relaxations from baseline of porcine coronary microvcsscls to NE without (•: n = 9) and with LY83583 (□: l pA/: n = 5) and hemoglobin ( a ; 10 pA/; n = 5). LY83583 and hemoglobin moderately blunt relaxations to NE. Values are expressed as mean ± SE. p < 0.05 vs. control.
Fig. 4. Average relaxations from baseline of porcine coronary microvesseis to NE with propranolol (•: I p.W; n = 7) and with or without hemoglobin (□: 10pA7;n = 6)oryohimbine(A; l pA7; n = 5). Hemoglobin and yohimbine mildly blunt relaxations (nonsignifi cant) to NE. Values are expressed as mean ± SE.
lance lo a significant degree. Further, vessels of this size class demonstrate several important differences in re sponse to pharmacologic [Lamping et al., 1989; Kanatsuka ct al., 1988; Chilian et al., 1989] and physiologic [Kanatsuka et al., 1989] stimuli compared to larger ves sels.
In the present study, minimal a-adrenoaceptor-mediated vasoconstriction was found in both quiescent and preconstricted microvesseis treated with (i-blockade with or without inhibition of EDRF with hemoglobin (to elim inate basal EDRF relaxation). This is in contrast to in vivo studies using sympathetic nerve stimulation [Mark
5
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
Fig. 1. Average relaxations from baseline of porcine coronary microvcsscls to NE without (•: n = 9) and with propranolol (o; I p.l/: n = 7). Propranolol markedly blunts relaxations to NE. Values are expressed as mean ± SE. p < 0.05 vs. control.
et al.. 1972; McRaven et al., 1971; Moreland et al., 1984; Kelley and Feigl. 1978] or NE infusion [Mark et al.. 1972; Kelley and Feigl. 1978; Vainer et al.. 1980] in which a transient a-adrenoceptor-mcdiatcd increase in vascular tone of coronary resistance vessels and reduced coronary flow was observed. Other investigators using in vitro preparations of smaller epicardial coronary arteries have reported findings similar to those in the present study [Zuberbuhler and Bohr. 1965; Mekata and Niu, 1969; Anderson et al., 1972; Nakayama et al., 1988], The lack of a-adrenoceptor-mediated vasoconstriction in epicardial microvessels, in the present study, may be due to one of several factors. First, the distribution of a-adrenoceptors is likely heterogeneous throughout the coronary circula tion [Cohen et al., 1983. 1984]. Thus, vasoconstriction may occur in vessels substantially larger or smaller than those analyzed in the present study. Secondly, the trans mural distribution of a-adrenoceptors may be different in the various size class vessels, with a greater proportion of functional a-adrenoceptors in the endocardial vessels ver sus the epicardial vessels which are examined in this study. Thirdly, with in vivo experiments, metabolic and autoregulatory influences may complicate the results de spite attempts to minimize the effects. Finally, in the case of sympathetic nerve stimulation, other vasoactive sub stances such as neuropeptide Y may be coreleased with NE which may independently alter vascular tone. It is unlikely that the low distending pressure em ployed in these studies accounted for the lack of constric tor responses to NE. In a large number of preliminary experiments, we altered perfusion pressure over a wide range (10-60 mm Hg). without observing a difference in the response to NE. Further, we have repeatedly found
that the optimal distending pressure for the development of isotonic contraction (which is the parameter measured using this preparation) to be 20 mm Hg. This is perhaps due to the fact that at higher distending pressures, the ves sels begin to generate greater amounts of isometric ten sion (which would be measured using a through the lumen wire or stirrup approach) and smaller amounts of isotonic tension. Another likely explanation is that once the vessel is removed from the myocardium and its external sup porting structures, the external forces upon the vessel are removed, and thus the internal pressure of 20 mm Hg is unopposed by an outside counterpressure. Under these circumstances, the transmural pressure in this prepara tion may actually approximate the normal physiologic distending pressures. This is particularly likely because the in vivo pressure within these vessels has been docu mented to be approximately 50% of aortic pressure, or only about 40-50 mm Hg during diastole. NE may elicit release of EDRF from large isolated arteries [Cocks and Angus, 1983]. In the present study, the relaxation of microvessels in response to NE was inhibited by both hemoglobin and by LY83583. demon strating the contribution of EDRF release in NE-induced relaxation. Hemoglobin failed to inhibit the relaxation of microvessels by the nonselective p-adrenergic agonist iso proterenol, showing that in the coronary microcirculation EDRF release is not through a p-adrenergic mechanism. In summary, NE causes vasodilation of porcine coro nary microvessels in vitro, free from metabolic and autoregulatory influences. Minimal a-adrenoccptor-mediated vasoconstriction occurs in response to NE. NE, in addi tion to its P-receptor-mediated relaxation, induces release of EDRF in the porcine coronary microcirculation.
Anderson. R.: Holmberg. S.: Svedmyr. N.: Aberg, G.: Adrenergic a- and (5-receptors in coronary vessels in man. An in vitro study. Acta tried, scand. /9 /. 241-244 (1973). Baron. G.D.: Speden. R.N.: Bohr. D.F.: Betaadrenergic receptors in coronary and skeletal muscle arteries. Am. J. Physiol. 223: 878-881 (1972). Chilian, W.M.; Layne. S.M.; Eastham, C.L.; Mar cus, M.L.: Heterogeneous microvascular coro nary a-adrenergic vasoconstriction. Circula tion Res. 64: 376-388 (1989). Cocks. T.M.: Angus. J.A.: Endothelium-dependent relaxation of coronary arteries by noradrena line and serotonin. Nature 305: 627-630 (1983).
Cohen. R.A.;Shepherd. J.T.; Vanhoutte, P.M.: Pre junctional and postjunctional actions of endog enous norepinephrine at the sympathetic neu roeffector junction in canine coronary arteries. Circulation Res. 52: 16-25 (1983). Cohen. R.A.; Shepherd. J.T.: Vanhoutte. P.M.: Ef fects of the adrenergic transmitter on epicardial coronary arteries. Fed. Proc. 43: 2862-2866 (1984). Drew. G.M.: Levy. G.P.: Characterization of the coronary vascular (5-adrenoceptor in the pig. Br. J. Pharmacol. 46: 348-350 (1972). Feigl, E.O.: Sympathetic control of coronary circu lation. Circulation Res. 20: 262-271 (1967).
Halpern. W.; Osol. G.; Coy. G.: Mechanical behav ior of pressurized in vitro prcarteriolar vessels determined with a video system. Ann. biomed. Eng. 121:463-479 (1984). Hamilton. F.N.; Feigl. E.O.: Coronary vascular sympathetic beta-receptor innervation. Am. J. Physiol. 230: 1569-1576(1976). Holtz. J.: Mayer. E.: Bassenge. E.: Demonstration of alpha-adrenergic coronary control in differ ent layers of canine myocardium by regional myocardial sympathectomy. Piliigers Arch. 372: 187-194(1977). Johannson. B.: The (5-adrenoceptors in the smooth muscle of pig coronary arteries. Eur. J. Phar macol. 24: 2 18-224 (1973).
6
Quillcn/Sellke/Banitt/Harrison
NE and the Coronary Microcirculation
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
R eferences
Mark. A.L.; Abboud. F.M.: Schmid. P.G.: Heistad. D.D.: Mayer. H.E.: Differences in direct effects o f adrenergic stimuli on coronary', cutaneous and muscular vessels. J. clin. Invest. 51: 279287(1972). McRaven. D.R.: Mark. A.L.: Abboud. F.M.; Mayer, H.E.: Responses of coronary vessels to adrenergic stimuli. J. clin. Invest. 50:1T3-11Ü (1971). Mekata, LL: Niu, H.: Electrical and mechanical responses of coronary artery smooth muscle to catecholamines. Jap. J. Physiol. 19: 599-608 (1969). Mohrman. D.E.; Feigl. E.O.: Competition between sympathetic vasoconstriction and metabolic vasodilation in the canine coronary circulation. Circulation Res. 42: 79-86 (1978). Moreland, R.S.; Bohr, D.F.: Adrenergic control of coronary arteries. Fed. Proc. 43: 2857-2861 (1984).’ Nakayama, K.: Osol, G.: Halpern. W.: Reactivity of isolated porcine coronary' resistance arteries to cholinergic and adrenergic drugs and trans mural pressure changes. Circulation Res. 62: 741-748(1988).
Nellis. S.; Liedtke. J.: Whitesell. L.: Small coronary vessel pressure and diameter in an intact beat ing rabbit heart using fixed-position and treemotion techniques. Circulation Res. 49: 342353(1981). Vatncr, S.F.: Pagani, M.: Manders. W.T.: Pasipoularides. A.D.: Alpha adrenergic vasoconstric tion and nitroglycerin vasodilation of large cor onary arteries in the conscious dog. J. clin. Invest. 65: 5-14 (1980). Vatncr. S.F.: Hintze. T.H.: Macho, P.: Regulation of large coronary arteries by p-adrencrgic mechanisms in the conscious dog. Circulation Res. 51:56-66 (1982). Vatner, D.E.; Knight. D.R.; Homey, C.J.: Vatner. S.F.; Young, M.A.: Subtypes of (3-adrcnergic receptors in bovine coronary arteries. Circula tion Res. 59. 463-473 (1986). Young, M.A.: Vatner, S.F.: Regulation of large cor onary arteries. Circulation Res. 59: 579-596 (1986). Zuberbuhler. R.C.: Bohr D.F.: Responses of coro nary smooth muscle to catecholamines. Circu lation Res. 16:431 -440 (1965).
7
Downloaded by: King's College London 137.73.144.138 - 3/6/2018 10:53:38 PM
Johannson, U.J.; Mark, A.L.: Marcus. M.L.: Re sponsiveness lo cardiac sympathetic nerve stimulation during maximal coronary dilation produced by adenosine. Circulation Res. 50; 510-517(1982). Kanatsuka. H.; Lamping, K.G.; Eastham. C.L.: Marcus. M.L.: Non-uniform effect of nitroglyc erin on different size coronary arterial micro vessels. (Abstract) FASEB J. 2: A496 ( 1988). Kanatsuka, H.; Lamping. K.G.: Eastham. C.L.: Dellsperger. K.C.: Marcus. M.L.: Comparison of the effects of i nereased M VO? and adenosi ne on the coronary microvascular resistance. Cir culation Res. 65: 1296-1305(1989). Kelley. K.O.: Feigl, E.O.: Segmental a-receptormediated vasoconstriction in the canine coro nary circulation. Circulation Res. 43: 908-917 (1978). Kuramoto, K.; Murata. K.; Fujii. K.; Kurihara. H.; Kimata. S.: Matsushita, S.: Kuramochi. M.; lkeda. M.: Nakao. K.: Sympathetic coronary vasoconstriction after adrenergic beta block ade. Tohoku J. exp. Med. 91;95-101 (1967). Lamping, K.G.: Kanatsuka. E.; Eastham. C.L.: Chilian. W.M.: Marcus, M.L.: Nonuniform va somotor responses of the coronary microcircu lation to serotonin and vasopressin. Circula tion Res. 65: 343-351 (1989).
