Cardiac structure and function in humans: a new cardiovascular physiology laboratory

As the traditional cardiovascular control laboratory has disappeared from the first-year medical school curriculum, we have recognized the need to develop another "hands-on" experience as a vehicle for wide-ranging discussions of cardiovascular control mechanisms. Using an echocardiograph, an automatic blood pressure cuff, and a reclining bicycle, we developed protocols to illustrate the changes in cardiac and vascular function that occur with changes in posture, venous return, and graded exercise. We use medical student volunteers and a professional echocardiographer to generate and acquire data, respectively. In small-group sessions, we developed an interactive approach to discuss the data and to make a large number of calculations from a limited number of measurements. The sequence of cardiac events and cardiac structure in vivo were illustrated with the volunteers lying down, standing, and then with their legs raised passively above the heart to increase venous return. Volunteers were then asked to peddle the bicycle to achieve steady-state heart rates of 110 and 150 beats/min. Data were collected in all these states, and calculations were performed and used as the basis of a small-group discussion to illustrate physiological principles. Information related to a surprisingly large number of cardiovascular control mechanisms was derived, and its relevance to cardiovascular dysfunction was explored. This communication describes our experience in developing a new cardiovascular control laboratory to reinforce didactic material presented in lectures and small-group sessions.

Selective A2A adenosine receptor agonist as a coronary vasodilator in conscious dogs: potential for use in myocardial perfusion imaging

The authors sought to demonstrate the advantages of a selective, potent, short-acting A adenosine receptor agonist, CVT-3146 (2-(N-pyrazolyl)Ado derivative), for potential clinical use as a coronary vasodilator during myocardial perfusion imaging. The use of adenosine in a pharmacological stress test during myocardial imaging is limited by side effects mediated by A1 and A2B adenosine receptors and by its ultrashort duration of action. CVT-3146 (0.1-5 microg/kg) and adenosine (13-267 microg/kg) were given as peripheral intravenous injections in 10 awake dogs instrumented for measurement of coronary blood flow (CBF). CVT-3146 caused a dose-dependent increase of CBF (ED50 = 0.34 +/- 0.08 microg/kg, maximal increase = 221 +/- 18%, n = 6). Adenosine was less potent (ED = 51 +/- 15 microg/kg, p < 0.05) but equieffective (maximal increase in CBF = 227 +/- 11%). The increase in CBF caused by 2.5 microg/kg CVT-3146 reached 84 +/- 5% of the maximal reactive hyperemia following 20 s of coronary occlusion (n = 4). After a 10-s injection of CVT-3146 (2.5 microg/kg), the increase in CBF remained at least twofold above baseline for 97 +/- 14 s, whereas for adenosine (267 microg/kg), the twofold increase in CBF lasted only 24 +/- 2 s (p < 0.01, n = 6). A 30-s injection of 2.5 microg/kg CVT-3146 prolonged the twofold increase in CBF up to 221 +/- 20 s. No atrioventricular block was noted. At 2.5 microg/kg, the peak effect of CVT-3146 on CBF was associated with a short-lasting (20 +/- 6 s) increase in heart rate (78 +/- 9 bpm) and decrease in mean arterial blood pressure (13 +/- 6 mm Hg, p < 0.05, n = 6). CVT-3146 is a potent coronary vasodilator. Its short duration of action, minimal and transient systemic hemodynamic effects, and ease of administration may make this agonist suitable for pharmacological coronary vasodilation during myocardial perfusion imaging for noninvasive detection of subcritical arterial stenosis.

Regulation of renal A(1) adenosine receptors in vivo

We have compared renal A(1) adenosine receptor (AR) regulations in rats after chronic agonist and antagonist treatments. In one group, R-phenylisopropyladenosine (R-PIA), a selective A(1) AR agonist, was infused subcutaneously for 7 days. Another group was fed theophylline, a non-selective AR antagonist, for 2 weeks. Renal cortex membrane A(1) AR binding with 1,3-[(3)H]-dipropyl-8-cyclopentylxanthine demonstrated approximately 40% reduction in the B(max ) for the R-PIA group without any changes in the K(d) values. Neither the B(max) nor the K(d) changed following chronic theophylline treatment. Renal cortex G(i)alpha-proteins from the R-PIA treated rats decreased by approximately 30%. Renal G(i)alpha levels did not change in theophylline-treated rats. Consistent with the A(1) AR desensitization, R-PIA-treated rats had significantly higher basal renin release and showed attenuated A(1) AR-mediated inhibition of renin release. These data suggest that prolonged A(1) AR stimulation results in downregulation of renal A(1) ARs and G(i)alpha, accompanied by desensitization of A(1) AR-mediated inhibitory effects on renin release. Unlike cardiac and brain A(1) ARs, renal A(1) receptors are not subject to up-regulation following chronic antagonist treatment.

Teaching the principles of hemodynamics

Knowledge of hemodynamic principles is crucial to an understanding of cardiovascular physiology. This topic can be effectively taught by discussing simple physical principles and basic algebraic equations. A variety of examples from everyday observations can be used to illustrate the physical principles underlying the flow of blood through the circulation, thereby giving the student an experiential feel for the topic in addition to an understanding of theory. Moreover, opportunities abound for showing how each hemodynamic principle can explain one or another functional feature of the cardiovascular system or a cardiovascular pathophysiological state. Thus hemodynamics can be used as an organizational thread to tie together other aspects of cardiovascular physiology.

Refresher course for teaching cardiovascular physiology

This report presents highlights of a refresher course presented at Experimental Biology '99 on Saturday, April 17, 1999, in Washington, District of Columbia.

Preconditioning with ischemia or adenosine protects skeletal muscle from ischemic tissue reperfusion injury

Prolonged tissue ischemia and subsequent reperfusion results in significant tissue injury due to the ischemic-reperfusion (IR) syndrome. Ischemic preconditioning (IPC) or adenosine (ADO) pretreatment are known to protect IR injury in cardiac muscle. Our aim was to determine whether IPC or ADO pretreatment attenuates and protects against ischemic tissue reperfusion injury in skeletal muscle. Rats were anesthetized and global hindlimb ischemia was induced by 60 min of suprarenal aortic clamping followed by 30 min of reperfusion period. The degree of skeletal muscle dysfunction was determined by decreases in maximum contractile force, and adenosine triphosphate (ATP) and creatine phosphate (CP) levels of extensor digitorum longus (EDL) muscle. The distal tendon of the EDL was attached to a force transducer for maximum isometric force measurement. Samples were taken from the EDL for measurement of ATP and CP levels. The following were protective protocols prior to the IR challenge: (1) four consecutive 5-min periods of ischemia separated by 5-min reperfusion periods (PC/I) or (2) i.v. adenosine infusion (350 microg/kg/min x 10 min, PC/A). Our data suggest that pretreatment with brief periods of ischemia or systemic ADO infusion attenuates ischemic tissue reperfusion injury in skeletal muscle. [Table: see text]

Cardiac desensitization to adenosine analogues after prolonged R-PIA infusion in vivo

To determine the effects of chronic in vivo stimulation of adenosine receptors, R-(-)-N6-(2-phenylisopropyl)adenosine (R-PIA), a selective A1 receptor agonist, was administered to rats as a continuous 7-day infusion (200 nmol/h). Inotropic and chronotropic responses of isolated atria to adenosine receptor agonists were markedly desensitized compared with the responses of atria from age-matched control animals. Carbachol's negative chronotropic effect was also attenuated, indicating a heterologous mode of desensitization. Antagonist radioligand binding assays indicated a 52% reduction in A1 adenosine receptor maximum binding, and competition binding assays revealed a significant loss of G protein-coupled high-affinity A1 receptors in atria from R-PIA-treated rats. Inhibitory G proteins (Gi) were significantly reduced, as quantified by immunoblot analysis, with no change in the amount of stimulatory G proteins. Ventricular membranes from R-PIA rats showed loss of Gi and uncoupling of A1 receptors, without a significant change in A1 receptor density. Thus chronic R-PIA infusion desensitized rat atrial muscle to the effects of adenosine receptor agonists via several regulatory adaptations, including downregulation of A1 adenosine receptors, uncoupling of A1 receptors from their associated G proteins, and loss of Gi proteins.

Differential sensitization of cardiac actions of adenosine in rats after chronic theophylline treatment

To determine the effect of chronic adenosine receptor blockade on atrial responsiveness, we administered theophylline to rats in their drinking water (0.6 mg/ml) for 2 wk. Inotropic and chronotropic responses to the adenosine receptor agonists N6-cyclopentyladenosine (CPA) and 5'-(N-ethylcarboxamido)-adenosine (NECA) were then measured in isolated atria from treated and control animals. The indirect (antiadrenergic) actions of CPA and NECA on force and rate, measured during beta-adrenergic receptor stimulation by isoproterenol, were markedly sensitized (2- to 10-fold reductions in the agonist concentration needed to obtain a half-maximal response) after theophylline. The sensitization was homologous because inotropic and chronotropic responses to carbachol were not affected by theophylline. The direct negative inotropic and chronotropic actions of CPA and NECA, measured without concomitant beta-adrenergic stimulation, were not sensitized after theophylline. The number of atrial A1-receptors, measured by antagonist radioligand binding (maximum specific binding at saturation), was increased by 22% in theophylline-treated rats [66.2 +/- 3.4 vs. 54.3 +/- 1.9 (control) fmol/mg protein, P < 0.05]. Competition binding indicated that the fraction of coupled (high-affinity) receptors was unchanged. The number of ventricular A1-receptors was increased to a similar extent without any change in coupling. Thus chronic dietary theophylline upregulated cardiac A1-adenosine receptors without changing coupling state or affinity and sensitized rat atria to the indirect, antiadrenergic, inhibitory inotropic and chronotropic actions of adenosine receptor agonists.

Role of nitric oxide in hypoxic coronary vasodilatation in isolated perfused guinea pig heart

To test the hypothesis that nitric oxide (NO) mediates hypoxic coronary dilatation in situ, isolated guinea pig hearts were perfused at constant pressure (Langendorff technique) with physiological salt solution. Switching from a control perfusate (95% O2-5% CO2) to one equilibrated with a lower O2 tension (20% O2) induced a large, but submaximal and reproducible, coronary dilatation. The NO synthase inhibitor NG-nitro-L-arginine (L-NNA) diminished baseline flow (3.67 +/- 0.24 vs. control 5.11 +/- 0.42 ml.min-1 x g-1; P < 0.05) and selectively blocked the coronary flow response to acetylcholine without reducing the response to papaverine. L-NNA reduced the absolute increase in coronary flow during hypoxia by 27 +/- 2% (delta flow = 5.83 +/- 0.49 vs. control delta flow = 8.04 +/- 0.74 ml.min-1 x g-1; P < 0.05). Hypoxic coronary dilatation was unaffected by infusion of the thromboxane mimetic U-46619, which decreased baseline coronary flow to the same extent as L-NNA. Prior addition of indomethacin did not alter the attenuating effect of L-NNA. Hypoxic coronary dilatation during constant flow perfusion at 14.7 +/- 0.28 ml/min was reduced by 65 +/- 5% after L-NNA. Therefore, the NO component of the response was not a consequence of the reduced baseline flow observed in the presence of L-NNA, did not depend on prostaglandin synthesis, and was not secondary to increased flow or intravascular shear stress. We conclude that hypoxic coronary vasodilatation in isolated guinea pig hearts is partially mediated by NO.

Adenosine causes bradycardia in pacing-induced cardiac failure

BACKGROUND

In normal, conscious dogs, systemic injection of adenosine causes arterial hypotension and a baroreceptor reflex tachycardia mediated in part by withdrawal of vagal tone from the sinoatrial node. After vagal section or muscarinic receptor blockade, however, adenosine injection causes bradycardia via a direct sinoatrial node inhibition. Because cardiac failure is marked by a loss of vagal tone, we hypothesized that adenosine injection in dogs with failing hearts would reduce heart rate.

METHODS AND RESULTS

Mongrel dogs were instrumented with indwelling catheters, manometers, and ventricular pacing electrodes. After the dogs had recovered from the surgery, the ventricles were paced continuously at 210 beats per minute for 3 weeks, followed by pacing at 240 beats per minute for an additional week. This regimen caused mild ventricular and more striking atrial hypertrophy and a gradual onset of physiological and clinical signs of congestive heart failure. Adenosine injections that caused large tachycardias before the pacing regimen began caused progressively smaller increments in heart rate during the first 2 weeks of pacing. After 3 and 4 weeks, adenosine injections caused overt reductions in heart rate despite the concomitant arterial depressor response.

CONCLUSIONS

We conclude that the loss of vagal tone associated with the development of cardiac failure unmasks the direct negative chronotropic effect of exogenous adenosine on the sinoatrial node.

Vasodilative and anti-adrenergic effects of adenosine in diabetic rat hearts

To determine the vasodilative and negative inotropic effects of adenosine in hearts of diabetic rats, isolated hearts, perfused at constant perfusion pressure (Langendorff technique), were prepared from age-matched control Wistar rats and rats made diabetic 10 weeks prior to study by a single injection of streptozotocin (65 mg.kg-1, i.p.). Adenosine and nitroprusside each increased coronary inflow when administered either as bolus injections or as infusions. Coronary flow responses to nitroprusside were unchanged in diabetic hearts. Coronary flow responses of diabetic hearts to adenosine injections were unchanged, but responses to adenosine infusions tended to be larger than in normal hearts. Diabetes had no significant effect on the EC50 for either vasodilator. Adenosine inhibited the inotropic effect of isoproterenol (enhanced left ventricular (LV) pressure (P) and LV dP/dtmax) in normal hearts, independently of its vasodilative action. This negative inotropic action of adenosine appeared equally strong in diabetic hearts. We conclude that adenosine's coronary vasodilative and anti-beta-adrenergic, negative inotropic effects in the rat heart were not diminished after 10 weeks of streptozotocin-induced diabetes mellitus. Thus, earlier reports of diminished adenosine dilative efficacy in experimental diabetes may have been unique to those particular models.

Mechanisms of coronary vasodilatation produced by ATP in guinea-pig isolated perfused heart

1. Isolated hearts of guinea-pigs were perfused in vitro with a physiological salt solution via a retrograde aortic cannulation (Langendorff preparation) at constant perfusion pressure. Bolus intra-arterial injections of various vasodilator drugs were made and the coronary flow responses were measured with an electromagnetic flow probe placed in the arterial inflow circuit. Inhibitory drugs were infused intra-arterially. 2. Nitro-L-arginine (NLA; 500 microM), an NO synthesis inhibitor, decreased coronary baseline flow by 16 +/- 0.8%, converted acetylcholine-induced coronary vasodilatation to vasoconstriction and had no effect on coronary flow responses to adenosine or papaverine. Sodium nitroprusside-induced responses were enhanced during NLA infusion by 46 +/- 11%. 3. Adenosine 5'-triphosphate (ATP) increased coronary flow but coronary flow responses to ATP were not altered by infusion of NLA. 4. ATP-induced coronary dilatation was not significantly attenuated by infusion of the adenosine receptor antagonist XAC, (xanthine amine congener; 2 microM), whereas XAC decreased coronary flow responses to adenosine by 75% +/- 5%. 5. ATP-induced coronary flow responses were reduced by only 31 +/- 4% during indomethacin infusion (2.8 microM) whereas indomethacin completely eliminated the initial vasoconstriction phase and greatly attenuated the peak flow and duration of the later vasodilatation phase seen in response to arachidonic acid (0.75 nmol). Indomethacin had no effect on vasodilatations produced by adenosine or prostaglandin I2. 6. These results indicate that ATP-induced coronary dilatation in the isolated, perfused heart of the guinea-pig is not dependent upon NO production or upon degradation of ATP to adenosine. The coronary dilator action of ATP may be partially dependent (approximately 30%) upon the production of vasodilator prostaglandins.

Glibenclamide attenuates adenosine-induced bradycardia and coronary vasodilatation

The effects of the ATP-sensitive K(+)-channel blocker glibenclamide on the cardiovascular responses to adenosine in dogs were determined. Adenosine (0.01-20 mumol/kg iv) caused coronary vasodilatation, arterial hypotension, and bradycardia in dogs with either combined beta-adrenergic and muscarinic receptor blockade or with bilateral cervical vagotomy plus beta-adrenergic receptor blockade. The 50% effective dose for adenosine-induced coronary dilatation was increased from 0.13 +/- 0.04 mumol/kg in the control state to 1.1 +/- 0.5 mumol/kg after 2 mg/kg of glibenclamide (P less than 0.001). Adenosine at 5 mumol/kg reduced heart rate by 19 +/- 5% from a baseline of 158 +/- 6 beats/min in five anesthetized dogs. After glibenclamide (10 mg/kg), this dose of adenosine failed to cause a significant change in heart rate. The arterial hypotensive effects of adenosine were also attenuated by glibenclamide. Thus glibenclamide inhibited adenosine-induced bradycardia, hypotension, and coronary dilatation. On the other hand, glibenclamide did not affect the reductions in heart rate caused by vagus nerve stimulation. The mechanism of this adenosine antagonism is not known but, in the case of bradycardia, it does not appear to involve any of the steps shared in common by both adenosine-induced and vagal responses of the sinoatrial node.

An unusual receptor mediates adenosine-induced SA nodal bradycardia in dogs

To characterize the receptor mediating the negative chronotropic effect of adenosine in dogs, experiments were performed on conscious dogs with chronically implanted cardiovascular instrumentation. Autonomic blockade was used to eliminate any reflex influences on heart rate. Intravenous bolus injections of various adenosine analogues caused dose-dependent, aminophylline-blockable reductions in heart rate with a potency order of 5'-(N-ethylcarboxyamido)-adenosine (NECA)-78:2-chloroadenosine-17:adenosine-1. Dipyridamole enhanced the potency of adenosine to equal that of 2-chloroadenosine. Moderately selective A1-receptor agonists N6-(L-2-phenylisopropyl)-adenosine (R-PIA) and N6-cyclohexyladenosine and an A2-selective agonist 2-phenylaminoadenosine (200 nmol/kg) had no negative chronotropic effect in the conscious dog. Adenosine and its analogues, including R-PIA, caused coronary vasodilatation at smaller doses than were required to slow the heart rate. The selective A1-adenosine receptor blocker xanthine amine congener (XAC) antagonized the negative chronotropic action of adenosine but did so nonselectively, as the coronary vasodilative and negative chronotropic actions of adenosine were antagonized equally well. The spontaneous contraction rate of isolated perfused dog right atrial preparations, which included the sinoatrial node, was reduced by intrasinoatrial node artery infusions of adenosine analogues with a potency ratio of NECA-100:adenosine-15:N6-cyclopentyladenosine-2.3:R-PIA-1. We conclude that the adenosine receptor mediating the negative chronotropic action of adenosine in the dog does not display the pharmacological characteristics of either typical A1- or A2-adenosine receptors. Instead, either a novel adenosine receptor or an A1-receptor with unusual agonist and antagonist binding properties appears to exist in the dog's sinoatrial node.

Mechanism of the apparent parasympathetic inhibition of adenosine induced heart rate slowing in the dog

The inhibitory action of intravenously administered adenosine on the sinoatrial (SA) node is not expressed in the conscious dog in the presence of normal vagal tone. After pharmacological or surgical parasympathetic blockade, however, adenosine exerts a powerful negative chronotropic effect. In order to determine the reason why this action of adenosine is blocked by the intact parasympathetic nervous system, we measured the chronotropic effects of adenosine while applying a constant cholinergic stimulus to the SA node. In conscious dogs during a systemic infusion of acetylcholine at a rate sufficient by itself to inhibit the SA node, intravenous adenosine injections caused further dose dependent reductions in heart rate. In anaesthetised, vagotomised dogs, intravenous adenosine caused similar negative chronotropic effects with or without concomitant electrical stimulation of the vagus nerve. As an example, 5 mumol.kg-1 adenosine reduced heart rate by 22 (SEM 4)% from a baseline heart rate of 172(10) beats.min-1; when heart rate was lowered to 66(1) beats.min-1 by electrical vagal stimulation, this dose of adenosine reduced heart rate by 36(8)%. Propranolol had no effect on these responses. We conclude that there is no direct cholinergic inhibition of the negative chronotropic action of adenosine on the canine SA node but rather that the inhibitory action of systemically administered adenosine on the SA node is simply masked by the withdrawal of vagal tone in response to the arterial hypotension resulting from this mode of adenosine administration.

Effects of alkylxanthines and calcium antagonists on adenosine uptake by cultured rabbit coronary microvascular endothelium

Adenosine uptake by cultured rabbit coronary microvascular endothelial cells was studied. Radiolabeled [2-3H]-adenosine, present initially in the extracellular space at 10(-6) mol/l, was incorporated into the cell cultures at a steady rate during 30 s-3 h incubations. Incorporated 3H was found mostly (83%) in adenine nucleotides. Incorporation of [3H]-adenosine was attenuated by an adenosine deaminase inhibitor (EHNA) but only at adenosine concentrations of 10(-5) mol/l or higher. Adenosine transport inhibitors (dipyridamole, nitrobenzylthioinosine) attenuated 3H incorporation. Adenosine uptake was also diminished by certain structural analogues of adenosine (e.g., 2-chloroadenosine), by several alkylxanthine drugs (theophylline, isobutylmethylxanthine, enprofylline and 8-phenyltheophylline), and by certain calcium antagonists (verapamil, nifedipine and trifluoperazine). The mechanisms of actions of these agents on adenosine uptake do not appear to be related to phosphodiesterase inhibition, adenosine receptor antagonism or calcium antagonism. The effects of varying adenosine metabolism may contribute to the pharmacologic actions of these agents.

Uptake and release of adenosine by cultured rat aortic smooth muscle

We wanted to determine whether CO2, H+ and K+ affect the adenosine metabolism of vascular smooth muscle in a way that could account for the effects of these substances on vascular reactivity and their ability to modulate adenosine-induced vascular relaxation. Accordingly, 1-week-old cultures of rat aortic smooth muscle were incubated in phosphate-buffered saline with various [K+]'s and pH's and aerated in an incubation chamber with gases containing various proportions of CO2. Uptake was measured as 14C incorporation into cellular constituents during exposure to 2 microM [14C]adenosine. Release was measured as net extracellular adenosine accumulation. Uptake of adenosine was not significantly affected by any of the experimental maneuvers, except that it was greatly attenuated by dipyridamole (10(-5) and 10(-4) M) and transiently enhanced by the low CO2 levels. Adenosine release, however, was depressed by lowering atmospheric CO2 (0% vs 5%) and also by normocapnic acidosis (pH 6.8 vs pH 7.4). We conclude that vascular smooth muscle in culture releases adenosine at a rate that might have vasoactive significance in vivo. Furthermore, some of the vascular actions of CO2 and H+, but not those of K+, may be partially explained by their effects on vascular smooth muscle's adenosine metabolism.

Superoxide anion selectively attenuates catecholamine-induced contractile tension in isolated rabbit aorta

Xanthine oxidase-derived oxygen metabolites caused a selective loss of norepinephrine-induced contractile tension in rings and helical strips from rabbit aorta. Phenylephrine-induced tension was not affected. The relaxation was selectively and completely blocked by superoxide dismutase but not by catalase. Isoproterenol-induced relaxation was also reversed by xanthine oxidase-derived oxygen metabolites. These observations are consistent with the chemical reaction of superoxide anion with catecholamines and suggest that the reaction may have significance at physiological concentrations of norepinephrine. The time course of the effects of superoxide generation on contractile tension was consistent with the properties of the chemical reaction (measured spectrophotometrically) and with the dependence of tone on norepinephrine concentration. These results indicate that superoxide anion, in situations at which submicromolar concentrations of this reduced oxygen metabolite are present, will selectively oxidize catecholamines, which may attenuate local adrenergic regulation.

Apparent reduction in baroreflex sensitivity to adenosine in conscious dogs

Relative effects of equihypotensive doses (-35 mmHg) of adenosine (5.0 mumol/kg) and nitroglycerin (25 micrograms/kg) on heart rate and, therefore, baroreflex sensitivity were studied in conscious dogs. Nitroglycerin increased heart rate 133 +/- 24% from 78 +/- 5.5 beats/min, whereas adenosine increased heart rate only 79 +/- 16% from 78 +/- 5.2 beats/min (P less than 0.01). Injection of nitroglycerin during combined beta-adrenergic and muscarinic receptor blockades caused arterial pressure to fall 38 +/- 3.4% from 107 +/- 3.2 mmHg without any significant change in heart rate (3.8 +/- 3.8 from 162 +/- 9.2 beats/min). During combined beta-adrenergic and muscarinic receptor blockades adenosine also reduced arterial pressure 45 +/- 2.7% from 106 +/- 2.9 mmHg but unexpectedly reduced heart rate as well by 37 +/- 1.7% from 160 +/- 9.7 beats/min. This bradycardia reflected an effect on the sinoatrial (SA) node rather than an induction of heart block, since the R-R interval increased by 70 +/- 7.8% from 371 +/- 20 ms (P less than 0.01), while the P-R interval increased only 13 +/- 2.3% from 97 +/- 7.2 ms (P less than 0.05) with no electrocardiographic evidence of nonconducted beats. Arterial plasma adenosine levels were 43 +/- 5 nmol/ml at this time. Adenosine also caused bradycardia during muscarinic blockade alone (-43 +/- 3.4% from 201 +/- 6.4 beats/min) and following bilateral vagal section (-33 +/- 1.9% from 151 +/- 5.9 beats/min). In summary, adenosine appears to alter normal baroreflex function in the conscious dog by reducing the tachycardia that normally follows a fall in systemic arterial pressure.(ABSTRACT TRUNCATED AT 250 WORDS)

The mechanism of adenosine release from hypoxic rat liver cells

Uptake of [14C]-adenosine into freshly dispersed rat hepatocytes was inhibited 44% by dipyridamole (50 microM) and 60% by nitrobenzylthioinosine (NBTI, 20 microM). The results are consistent with the known ability of these drugs to inhibit adenosine transport in other cell types. The nucleotide analogue, alpha, beta-methylene adenosine diphosphate (AOPCP, 50 microM), inhibited by 84% the degradation of exogenous 5' AMP that occurred rapidly when this substrate alone was presented to isolated hepatocytes. This confirms the ecto-5'-nucleotidase inhibitory properties of this analogue in isolated hepatocytes. During hypoxic incubation, isolated hepatocytes released adenosine, which accumulated in the extracellular volume. Dipyridamole and NBTI each markedly attenuated this extracellular adenosine accumulation. In contrast, AOPCP had no inhibitory effect on net hypoxic adenosine release. It is concluded that hypoxic rat hepatocytes produce adenosine intracellularly and that this adenosine is released via facilitated diffusion to the extracellular space, based on the inhibition observed with the transport inhibitors. The plasma membrane enzyme ecto-5'-nucleotidase does not appear to participate in hypoxic adenosine release from these cells as indicated by the lack of effect of the nucleotidase inhibitor, AOPCP.

Absence of a Role for Superoxide Anion, Hydrogen Peroxide and Hydroxyl Radical in Endothelium-Mediated Relaxation of Rabbit Aorta

We tested the hypothesis that the endothelium-dependent relaxation of rabbit thoracic aorta in vitro is mediated by reduced metabolites of oxygen. Helical vascular strips were contracted with either norepinephrine or phenylephrine. Oxygen metabolites, generated by the xanthine oxidase reaction, completely relaxed norepinephrine-induced contractile tone but not tone induced by phenylephrine. A mixture of oxygen metabolite scavengers (superoxide dismutase, catalase and mannitol) eliminated the relaxation induced by the xanthine oxidase products. Acetylcholine caused a dose-dependent and endothelium-dependent relaxation of the strips; this was not inhibited by the presence of the scavengers. We conclude that reduced oxygen metabolites have little direct effect on rabbit aortic smooth muscle in vitro, although they indirectly but specifically relax norepinephrine-induced tone, presumably by oxidation of norepinephrine. Oxygen metabolites do not appear to mediate the endothelium-dependent relaxation response of this tissue to acetylcholine.

Intracellular adenosine in isolated rat liver cells

Our objective was to determine whether a non-extracellular pool of adenosine exists in mammalian cells. Rat liver cells were dispersed by a collagenase perfusion technique and suspended in buffered salt solution. The adenosine content of these suspensions rose during hypoxia. Exogenous adenosine deaminase prevented or reversed the hypoxic increment but failed to reduce suspension adenosine levels to zero. This residual adenosine pool (average size = 85 +/- 10 pmol/mg protein) was not located in the extracellular medium, on surface adenosine receptors or in solution in the cytoplasm. A likely locus is the adenine-analog binding protein which has been described for liver and other tissues. Thus, our study supports the existence of an intracellular adenosine pool in isolated rat liver cells which is a large fraction of the total tissue adenosine. This situation may exist in other cell types as well, based on the ubiquity of the adenosine binding protein. Tissue adenosine content may not, therefore, accurately reflect interstitial adenosine concentration; thus, such measurements must be interpreted cautiously. It is not clear what, if any, functional role this putative, intracellular, bound adenosine pool plays in local vasoregulation.

Atrioventricular conduction disturbances during hypoxia. Possible role of adenosine in rabbit and guinea pig heart

Adenosine and related compounds can produce atrioventricular (A-V) conduction block. Similar conduction disturbances are observed in myocardial hypoxia. To investigate the possibility that adenosine might be causally involved in hypoxic conduction disturbances, we measured A-V conduction times, subdivided into atrial-to-His bundle (A-H) and His bundle-to-ventricular (H-V) intervals, with extracellular electrodes in isolated rabbit and guinea pig hearts perfused with modified Krebs-Henseleit solution. Adenosine produced dose-dependent prolongation of A-V conduction time in both species, although guinea pig hearts responded to lower doses (10(-7) M) and showed a steeper dose-response relationship than rabbit hearts. Higher adenosine doses produced second-degree heart block in both species. Conduction delay was confined to the A-H interval, implicating action on A-V node cells. Further investigation of guinea pig hearts revealed a specific antagonism towards adenosine's effects by 10(-5) M aminophylline. Conduction disturbances produced by acetylcholine or MnCl2 were unaffected by aminophylline as were adenosine's effects by atropine. Perfusion with hypoxic perfusate caused A-V conduction delays and second-degree block in guinea pigs hearts. This effect was dramatically attenuated by aminophylline. We conclude that endogenously released adenosine may cause at least some of the A-V conduction disturbances associated with acute myocardial hypoxia. Furthermore, methylxanthines may prove to be of therapeutic value in combatting such disturbances in a clinical setting.

The role of potassium in the metabolic control of coronary vascular resistance of the dog

We tested the hypothesis that potassium ion (K+) is involved in the local control of the coronary circulation. The left coronary artery was perfused at constant flow in closed-chest, anesthetized dogs. Step increases in heart rate caused transient (six dogs) or sustained (three dogs) increases in coronary sinus plasma [K+] averaging 0.53 mEq/liter. When the effects of vascular transit delay were accounted for, we found that [K+] changes preceded the vasodilation seen with increased heart rate. We used a mathematical model to calculate changes in interstitial [K+] from arterial and venous [K+] and K+ release rate. The magnitude of the changes in interstitial [K+] appeared to be sufficient to account for a considerable portion but not all of the initial changes in coronary vascular resistance associated with increased heart rate. Thus potassium seems to be involved at least transiently, and, in three of nine dogs, for a more sustained period, in heart rate-induced coronary vasodilation. Cessations (for 15 seconds) of coronary blood flow resulted in transient postischemic increases of coronary sinus [K+] averaging 0.55 mEq/liter. In this case, correction for vascular transit disclosed that the recovery of [K+] preceded the return of vascular tone to baseline for only the second half of the recovery, implying only a limited role for potassium in this response. Potassium appears to play a significant but transient role in the local control of the coronary circulation.

The role of adenosine in prolonged vasodilation following flow-restricted exercise of canine skeletal muscle

A period of prolonged vasodilation follows flow-restricted exercise of skeletal muscle. We tested the hypothesis that adenosine participates in mediating this vascular response. Vascularly isolated, anterior calf muscles of anesthetized dogs were stimulated to contract at a rate of 4 twitches/sec. Blood flow was held constant at 12.5 +/- 1.3 ml/min per 100 g which was about 14% of the expected free flow for this exercise level. Skeletal muscle tissue adenosine was measured with the an enzymatic, spectophotometric assay of trichloroacetic acid extracts of congruent to 50 mg biopsy samples. Tissue adenosine rose from 2.30 +/- 0.90 nmol/g in resting muscle to 22.5 +/- 5.8 nmol/g by the end of the 22-minute exercise. Following exercise, tissue adenosine fell toward its baseline value with a time course very similar to the early portion of the return of skeletal muscle vascular resistance to its control level. Thus, skeletal muscle adenosine content (1) increases to a sufficient magnitude and (2) falls with an appropriate time course to be at least partly responsible for the early portion of prolonged vasodilation seen after flow-restricted exercise of skeletal muscle.

Effect of indomethacin on coronary vascular response to increased myocardial oxygen consumption

We tested the hypothesis that arachidonic acid metabolites mediate the coronary vascular response to changes in cardiac activity. Isoproterenol was administered intravenously to five chloralose-anesthetized, open-chest dogs. Left anterior descending coronary artery blood flow, systemic arterial blood pressure, and great cardiac vein O2 content were continuously measured, and blood gas determination (including O2 content) were made before and after infusions. From these data, coronary vascular conductance, coronary O2 delivery, and myocardial O2 consumption were calculated. Isoproterenol increased conductance, O2 delivery, and O2 consumption. Indomethacin, a blocker of prostaglandin synthesis, was administered, and the isoproterenol infusions were repeated. The changes in conductance, O2 delivery, and O2 consumption associated with isoproterenol were not different after indomethacin was administered than before indomethacin was administered. Neither were the relations between conductance or O2 delivery and O2 consumption affected by indomethacin. We conclude that, in this preparation and with this stimulus, prostaglandins do not appear to mediate or modulate the coronary vascular response to changes in cardiac activity.

Coronary vascular resistance and myocardial oxygen consumption dynamics in response to catecholamine infusion

These experiments were performed in order to ascertain whether activation of coronary vascular adrenergic receptors could change the time course of the coronary vasodilatation accompanying increases in myocardial metabolic activity. The left common coronary arteries of dog hearts were perfused in situ with blood at constant flow. Coronary perfusion pressure and coronary sinus blood 02 content were continuously monitored. Norepinephrine was infused into the coronary artery at 0.5 to 5 microgram.min-1. and isoprenaline at 0.5 to 2 microgram.min-1. The amplitude of the vascular resistance change was less with norepinephrine than with isoprenaline infusion for a similar change in oxygen consumption. The time course of coronary vascular resistance, after correction for the effects of vascular transit, lagged significantly behind the time course of coronary sinus 02 content in the case of norepinephrine infusion. On the average, no lag was observed with isoprenaline infusion. It is concluded that stimulation of coronary vascular adrenergic receptors can alter the time course and the magnitude of the coronary vascular response to increases in myocardial metabolic activity resulting from myocardial beta-receptor stimulation.

Dynamics of myocardial oxygen consumption and coronary vascular resistance

Coronary vascular resistance may be regulated in part by substances whose concentrations are determined by or reflect the rate of myocardial oxygen consumption (e.g., adenosine, vessel wall PO2). We tested this hypothesis by comparing the time course of changes in myocardial oxygen consumption and coronary vascular resistance following 20 beat/min changes in heart rate. Main left coronary arteries of in situ dog hearts were perfused with blood at constant flow. Coronary sinus O2 content was monitored continuously with a densitometer and reflected the time course of changes in oxygen consumption and also the effects of vascular transit between tissue and the coronary sinus. These transit effects were estimated from dye transit curves and added to the time course of changes in coronary perfusion pressure which was proportional to coronary vascular resistance at constant flow. Coronary sinus O2 content changes preceded the adjusted time course of vascular resistance. This supports the hypothesis that coronary vascular resistance is regulated in part by factors closely linked to oxidative metabolism.

Simultaneous counting of 85Kr in lung and myocardium during measurement of coronary blood flow

Coronary blood flow rate (ml-min-1-100 g-1) was estimated by a) measuring pump flow into the cannulated circumflex branch of the left coronary artery and dividing by the weight of perfused myocardium and b) measuring the clearance of 85Kr following intra-arterial injection (detection with a 2-in. crystal with cylindrical collimation). Although the correlation between the two measurements was relatively high (r equals 0.90), the line best fitting the data was 85Kr flow equals 0.55 pump flow + 25.6. We tested the possibility that the discrepancy between the two methods was primarily due to the counting of 85Kr removed from myocardium and delivered to lung. Relative efficiency of lung counting versus myocardial counting was determined as well as clearance pattern of 85Kr from lung in each dog. A simple mathematical model which assumes no recirculation of 85Kr to heart allowed correction of coronary clearance curves using this information. When corrected 85Kr flow equals 1.00 pump flow + 4.1 (r equals 0.90). Thus, the major systematic cause for the discrepancy between the two measurements under the conditions of this experiment appears to be simultaneous counting of 85Kr in lung and in myocardium.