Introduction

Top Abstract Introduction Methods Results Discussion References

The catecholamine-induced increase in contractile force in mammalian myocardium is mainly due to activation of beta adrenoceptors. The existence of myocardial alpha-1 adrenoceptors mediating positive inotropic effects is, however, now well established (Osnes et al., 1985; Fedida 1993; Terzic et al., 1993). In rat and rabbit heart the alpha-1 adrenoceptor-mediated effect is influencing and contributing to the final inotropic effect of combined receptor activation by the endogenous agonist norepinephrine (Skomedal et al., 1988, 1990a). As alpha-1 adrenoceptor-mediated and beta adrenoceptor-mediated inotropic effects can be differentially influenced (Endoh and Motomura, 1979; Christiansen et al., 1987), it has been speculated that alpha-1 adrenoceptor mediated effects may have increased importance during conditions where beta-adrenoceptor mediated effects are attenuated (Christiansen et al., 1987, Skomedal et al., 1988, 1990a).

In the failing human heart where the beta-1 adrenoceptor mediated inotropic response is attenuated, it has been difficult to demonstrate an alpha-1 adrenoceptor-mediated inotropic effect of functional significance (e.g., Böhm et al., 1988, Brodde et al., 1992; Steinfath et al., 1992; Bristow 1993). Mügge et al. (1983) and Brückner et al. (1984) reported alpha-1 adrenoceptor-mediated inotropic effects in human ventricular myocardium, but these authors did not compare the alpha-1 adrenoceptor-mediated inotropic effect and the beta adrenoceptor-mediated inotropic effect. The study by Böhm et al. (1988) indicated, however, an increasing importance of the alpha-1 adrenoceptor-mediated inotropic effect during severe heart failure due to the increasing attenuation of the beta adrenoceptor-mediated effect. In all these studies phenylephrine (in the presence of a beta receptor antagonist) has been used as the alpha adrenoceptor agonist. In addition, Landzberg et al. (1991) also observed an alpha adrenoceptor-mediated inotropic effect of phenylephrine in humans in vivo. In the presence of a beta receptor antagonist, Aass et al. (1986) found an alpha-1 adrenoceptor-mediated inotropic effect of norepinephrine in human ventricular myocardium obtained from patients undergoing mitral valve replacement. Although not compared directly to the beta adrenoceptor-mediated inotropic effect, the alpha-1 adrenoceptor-mediated inotropic effect observed in this study was of such a magnitude that a functional role might be expected. Neumann et al. (1993) and Scholz et al. (1995, 1996) reported data that indicated that naturally occurring norepinephrine was more effective as agonist compared to phenylephrine at the myocardial alpha-1 adrenoceptors mediating increase in contractility in failing human hearts. The functional importance of the alpha-1 adrenoceptors in failing human myocardium is thus still unsettled.

The purpose of our studies was to further elucidate the functional role of the myocardial alpha adrenoceptors in regulation of contractility also in failing human heart by investigating the relative magnitude of alpha-1 adrenoceptor-mediated and beta adrenoceptor-mediated inotropic responses in explanted human hearts. In our work we report that despite great variation in individual maximal responses to adrenergic stimulation by norepinephrine, the magnitude of the alpha-1 adrenoceptor-mediated inotropic component was comparable to that of the beta adrenoceptor-mediated inotropic component. In addition, the contribution from the two adrenergic receptor systems to the inotropic response may differ due to different etiology of heart failure.

	Methods

Top Abstract Introduction Methods Results Discussion References

Myocardial preparations. Ventricular myocardium was obtained from heart transplant recipients immediately after explantation of the failing heart. During preparation of the tissue, the surface was kept wet with physiological saline. Thin trabeculae were localized in the ventricular cavities, cut free and placed in a relaxing oxygenated (95% O₂/5% CO₂) physiological salt solution at room temperature. To prevent contracture during transportation and further preparation, we used a Ca⁺⁺/Mg⁺⁺ concentration ratio of 1:8 comparable to that of St. Thomas' Hospital cardioplegic solution. Experimentally, magnesium cardioplegia has also been shown to effectively protect the myocardium from calcium overload (Ataka et al., 1993). The solution contained (mmol/liter): NaCl 118.3; KCl 3.0; CaCl₂ 0.5; MgSO₄ 4.0; KH₂PO₄ 2.4; NaHCO₃ 24.9; glucose 10.0; mannitol 2.2. The tissue was transported to the laboratory and while kept immersed in this solution, muscle strips (about 1 mm in diameter, 8-10 mm long, endocardium as intact as possible) were prepared. The muscle strips were mounted in four organ baths containing the same oxygenated solution at 37°C except for Ca⁺⁺ being 2.5 mmol/liter and Mg⁺⁺ 1.2 mmol/liter. The muscles were driven electrically (field stimulation) at a frequency of 1 Hz with impulses of 5-msec duration and current about 20% above individual threshold (10-15 mA, determined in each experiment). The isometrically contracting muscles were stretched to the maximum of their length-tension curve. The developed tension was recorded by a Grass FT03C force-displacement transducer connected to a Grass RPS7C8B polygraph recorder equipped with 7DAG driver amplifiers, with 7P1F bridge amplifiers and with 7P20C derivators. The direct current signals were analog-to-digital converted by a National Instruments NB-MIO-16X board mounted in an Apple Macintosh Quadra 700 computer. The data acquisition was externally triggered synchronously to the contraction-relaxation cycles by the square wave pulses that triggered the muscles to contract. Each channel was scanned at 1 kHz yielding a time resolution of 1 msec. For the human heart the time window to be scanned was set to 500 msec, i.e., 500 posttrigger scans were sampled for each contraction-relaxation cycle. The logged data were stored in files as time stamped and event marked unfiltred binary clusters by software developed for the purpose in the visual programming language LabVIEW. The software could later open the files for analysis and compute appropriate low-pass filtered trend curves for developed tension vs. time for each contraction-relaxation cycle. Areas representative for the different experimental periods (control, stimulated) could be selected to calculate averaged contraction-relaxation cycles for these periods. These cycles were then used to determine values for typical descriptive parameters including T_max, TPT and TR20.

Experimental design. The muscles were allowed to equilibrate for 60 to 90 min before addition of agonist. The salt solution was changed in the middle of the equilibration period. Adrenoceptor agonists (phenylephrine, norepinephrine) were added directly to the organ bath in volumes of 25 to 75 µl to give the appropriate final concentrations and were completely mixed in the bath within 2 to 3 sec. Dose-response experiments were performed by cumulatively increasing the concentration of norepinephrine in the organ baths.

6

5

6

6

4

4

6

Explanted hearts. In our report data obtained from 20 explanted hearts are presented. The explanted hearts were failing either due to ischemic cardiomyopathy (postinfarction failure, n = 10) or to nonischemic cardiomyopathy (dilated cardiomyopathy (n = 4), congenital malformations (n = 4), myocarditits (n = 1) or toxic (adriamycin) cardiomyopathy (n = 1). In the group of patients undergoing heart transplantation due to ischemic cardiomyopathy, nine were men (44-61 yr, mean 54 yr) and one was a woman aged 58 yr. In the group of patients undergoing heart transplantation due to nonischemic cardiomyopathy, six were men (aged 14-57 yr, mean 36 yr) and four were women (aged 24-55 yr, mean 39 yr). All patients were severely symptomatic from their heart failure (New York Heart Association functional class III-IV). Three of the patients in the nonischemic group had relatively normal left ventricular function. Two of these had severe pulmonary hypertension secondary to congenital heart disease (Eisenmenger's syndrome) and the third had Uhl's syndrome with failing of the right ventricle. The other 17 patients all had severely depressed left ventricular function with ejection fraction of 24 ± 6% (mean ± S.E.) and pulmonary capillary wedge pressure of 21 ± 7 mm Hg (mean ± S.E.) with no difference between failure due to ischemic heart disease or not. The nonischemic patients had more severe right ventricular failure revealed as a higher right atrial pressure than the ischemic group, 15 ± 5 vs. 4 ± 2 mm Hg (mean ± S.E., P = .001), respectively. The nonischemic group also had a lower resting cardiac index of 1.5 ± 0.3 liter/min/m² compared to 2.2 ± 0.4 liter/min/m² (mean ± S.E., P = .006) in the ischemic failure group. All patients were receiving supportive medical treatment before transplantation by at least two of the following drugs: digitalis (mostly digitoxin), loop diuretics (furosemide, bumetanide), ACE-inhibitors (captopril, enalapril, lisinopril), spironolactone and nitrates (isosorbidmono/dinitrate). Some patients also received one or more of the following drugs: aspirin, warfarin, lovastatin, dipyridamol and potassium-chloride. In addition some patients used sedatives (oxazepam, levomepromazine, haloperidol), antiasthmatics (inhalation of budesonide, ipratropium, salbutamol), paroxetine, chlormezanone/acetaminophen, omeprazole or glipizide. Due to serious arrhythmias, three patients with ischemic and two patients with nonischemic cardiomyopathy also received amiodarone. All the patients, except one, receiving amiodarone were euthyroid as judged by TSH values. One patient received thyroxine due to hypothyroidism appearing before amiodarone treatment was started. None of the patients received adrenergic blocking drugs.

Definitions. T_max = maximal developed tension; TPT = time to peak tension; TR20 = time to relaxation to 20% level; RT = TR20  TPT = relaxation time; pD₂ = log EC₅₀ (EC₅₀ = concentration giving half maximal effect). Changes in contractile force was expressed as changes in T_max. Horizontal positioning of the dose-response curves were expressed by pD₂ values.

Calculation. The values after responses to agonist were generally calculated as percent of control values (100%). Response curves were constructed according to Ariëns and Simonis (1964), by estimating centiles, i.e., t₁₀ to t₁₀₀ for each single time course experiment and EC₁₀ to EC₁₀₀ for each single dose-response experiment, and calculating the corresponding means for each experimental group. This was done by a computer program based on linear interpolation between actual observed values. The response values were either expressed as fractional responses in per cent of maximum or were recalculated with control values as 100% and the maxima in percent thereof to also express differences in efficacy. The significance levels of differences were expressed by calculating P according to Wilcoxon one-sample or two-sample tests as appropriate. P  0.05 was considered to reflect significant differences.

Drugs. (-)-Norepinephrine bitartrate and timolol maleate were purchased through Norwegian Medical Depot. Ascorbic acid, atropine sulfate, carbamylcholine chloride (carbachol), (-)-phenylephrine hydrochloride and prazosin hydrochloride were purchased from Sigma Chemical Co. (St. Louis, MO). Stock solutions were prepared in purified water and kept at 20°C to avoid oxidation. Further dilutions of the drugs were made fresh daily and kept cool (0-4°C) and dark. Repetitive experiments showed that drug solutions treated in these ways, are stable.

	Results

Top Abstract Introduction Methods Results Discussion References

Characterization of the alpha-1 adrenergic inotropic effect of phenylephrine. In the presence of 6 × 10⁶ mol/liter timolol, activation of alpha-1 adrenoceptors by 1.2 × 10⁴ mol/liter phenylephrine evoked an inotropic response with the characteristics expected for an alpha-1 adrenoceptor-mediated inotropic response in mammalian myocardium. After a lag phase of about 30 sec, a monophasic inotropic response developed with a time to 50% relative response of 96 ± 9 sec (mean ± S.E., n = 6) (fig. 1, table 1). TPT, TR20 and RT were not significantly changed compared to control (table 1).

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Time course curves showing the development of the inotropic responses in failing human ventricular myocardium exposed to 1.2 × 104 mol/liter norepinephrine in the presence of 6 × 106 mol/liter prazosin (beta adrenoceptor stimulation, n = 8, ); in the presence of 6 × 106 mol/liter timolol (alpha-1 adrenoceptor stimulation, n = 14, ); alone (concomitant alpha-1 and beta-adrenoceptor stimulation, n = 14, ) and to 1.2 × 104 mol/liter phenylephrine in the presence of 6 × 106 mol/l timolol (alpha-1 adrenoceptor stimulation, n = 6, ). Ordinate: Inotropic response expressed as Tmax in percent of individual maximal responses. Abscissa, Time in seconds after addition of agonist. The receptor blockers were added 30 to 45 min before the agonists. Horizontal bars (given where they exceed the width of the symbol) represent ± S.E. of time to 50% relative response (see also table 1).

The influence of adrenergic blocking agents on the time course of the inotropic response to norepinephrine. In the presence of 6 × 10⁶ mol/liter of the alpha-1 adrenoceptor blocker prazosin, 1.2 × 10⁴ mol/liter norepinephrine elicited a monophasic positive inotropic response after a lag phase of about 5 sec. Time from addition of norepinephrine to 50% relative response was 22 ± 2 sec (mean ± S.E., n = 8) (fig. 1, table 1). This response was sensitive to beta adrenoceptor blockade as it was reversed by addition of 6 × 10⁶ mol/liter timolol (fig. 2a).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   a, Recording of contraction-relaxation cycles in failing human ventricular myocardium exposed to 1.2 × 104 mol/liter norepinephrine in the presence of 6 × 106 mol/liter prazosin (beta adrenoceptor stimulation). Control contraction before addition of agonist (); maximal inotropic steady state response (---); reversal of the inotropic response by 6 × 106 mol/liter timolol (·····). b, Recording of contraction-relaxation cycles in failing human ventricular myocardium exposed to 1.2 × 104 mol/liter norepinephrine in the presence of 6 × 106 mol/l timolol (alpha-1 adrenoceptor stimulation). Control contraction before addition of agonist (); maximal inotropic steady state response (---); reversal of the inotropic response by 6 × 106 mol/liter prazosin (·····). Ordinate, Developed tension in mN. Abscissa, Time in msec after the initiating stimulus.

6

4

6

Time course of the inotropic response to norepinephrine alone and the reversal of this response by sequential addition of adrenergic blockers. In the absence of adrenoceptor blocking agents, 1.2 × 10⁴ mol/liter norepinephrine evoked a monophasic positive inotropic response that developed after a lag phase of about 5 sec. In this situation, time from addition of norepinephrine to 50% relative response was 34 ± 4 sec (mean ± S.E., n = 14) (table 1, fig. 1). The inotropic response thus elicited demonstrated sensitivity both to the alpha-1 adrenoceptor blocker prazosin and to the beta adrenoceptor blocker timolol. This was demonstrated as the reversal responses to sequential addition of the respective adrenoceptor blockers at the steady state phase of the inotropic response (fig. 3) (see also Skomedal et al., 1988). The inotropic response to norepinephrine thus showed two components: one sensitive to the alpha-1 adrenoceptor blocker prazosin, amounting to 26.7 ± 3.5% of the total response (mean ± S.E., n = 12) and one sensitive to the beta-adrenoceptor blocker timolol, amounting to 73.3 ± 3.5% of total response (mean ± S.E., n = 12).

View larger version (16K): [in this window] [in a new window]   Fig. 3.   a) Reversal of the inotropic response to concomitant alpha-1 and beta adrenoceptor stimulation by norepinephrine: steady state response to 1.2 × 104 mol/liter norepinephrine (), reversal response to 6 × 106 mol/liter timolol (---) and reversal response to 6 × 106 mol/liter prazosin (·····). b, Reversal of the inotropic response to concomitant alpha-1 and beta adrenoceptor stimulation by norepinephrine: steady state response to 1.2 × 104 mol/liter norepinephrine (), reversal response to 6 × 106 mol/liter prazosin (---) and reversal response to 6 × 106 mol/liter timolol (·····). Ordinate, Developed tension in mN. Abscissa, Time in msec after the initiating stimulus.

Qualitative characteristics of the inotropic responses to norepinephrine in the presence and absence of adrenergic blocking agents. Time to peak tension in the control period was 183 ± 5 msec (mean ± S.E., n = 16). The inotropic responses elicited by 1.2 × 10⁴ mol/liter norepinephrine in the presence of prazosin and timolol, respectively, were accompanied by different qualitative changes in the contraction-relaxation cycles. In the presence of 6 × 10⁶ mol/liter prazosin, there was a shortening of the duration of the contraction-relaxation cycles (fig. 2a) with a significant reduction in TPT, TR20 and RT (table 1) reflecting a selective increase in relaxation compared to contraction. The shortening of TPT, TR20 and RT was sensitive to the beta adrenoceptor blocker timolol (fig. 2a).

6

4

Influence of cholinergic stimulation on the mechanical responses to norepinephrine in the presence of adrenergic blockers. Activation of muscarinic receptors by 1.2 × 10⁴ mol/liter carbachol influenced differently the steady state response to norepinephrine in the presence of prazosin and timolol, respectively. In the presence of 6 × 10⁶ mol/liter prazosin, both the inotropic response and the shortening of time to peak tension induced by 1.2 × 10⁴ mol/liter norepinephrine were attenuated by carbachol (table 2). In the presence of 6 × 10⁶ mol/liter timolol, however, carbachol increased the inotropic response to norepinephrine although time to peak tension was almost unchanged (table 2). The carbachol-induced changes were sensitive to the muscarinic receptor blocker atropine (data not shown).

Dose-response relationships of the inotropic responses to norepinephrine in the presence of adrenergic blockers. Norepinephrine evoked a concentration-dependent increase in contractile force both in the presence of prazosin and of timolol, respectively (fig. 4, table 3). In muscle tissue obtained from hearts failing due to ischemic cardiomyopathy, norepinephrine appeared to be equipotent in the presence of prazosin and timolol, respectively (fig. 4a, table 3). In muscle tissue obtained from hearts failing due to nonischemic cardiomyopathy, however, norepinephrine was apparently about 10 times more potent in the presence of prazosin compared to the presence of timolol (fig. 4b, table 3).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Inotropic responses expressed as developed tension (Tmax) to increasing concentrations of norepinephrine in human ventricular myocardium failing a) due to ischemic cardiomyopathy and b) due to nonischemic cardiomyopathy: alpha-1 adrenoceptor-mediated response in the presence of 6 × 106 mol/liter timolol (, n = 6); beta adrenoceptor mediated response in the presence of 6 × 106 mol/liter prazosin (, a, n = 6; b, n = 5). Ordinate, Inotropic response in percent of individual maxima. Abscissa, Log concentration of norepinephrine (mol/liter). Horizontal bars represent ± S.E. of pD2 values.

Maximal inotropic responses to norepinephrine in the presence of adrenergic blockers. The responses to maximal stimulation by norepinephrine varied considerably in preparations from different individuals. There was, however, an apparent covariation between the inotropic response in the presence of prazosin and in the presence of timolol in preparations obtained from the same heart and chamber (fig. 5) with a rank order correlation of r² = 0.4793 (P = .013). Thus, the mean values for the maximal effect of the two inotropic response components of norepinephrine were comparable both in time course (single concentration) experiments (fig. 5) and in dose-response (multiple concentrations) experiments (table 3).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Maximal inotropic responses in failing human ventricular myocardium to separate alpha-1 and beta adrenoceptor stimulation by 1.2 × 104 mol/liter norepinephrine in the presence of appropriate receptor blockers obtained in corresponding hearts and chambers (, n = 12).  and bars represent mean ± S.E. of the separate values. Ordinate, beta adrenoceptor-mediated inotropic response in percent of control period. Abscissa, alpha-1 adrenoceptor-mediated inotropic response in percent of control period. Rank order correlation, r2 = 0.4793, P = .013.

Influence of thyroid status on adrenergic inotropic responses to norepinephrine. In one heart failing due to ischemic cardiomyopathy from a patient judged to be hypothyroid as indicated by an elevated TSH value (TSH = 20 nmol/liter), the dose-reponse relationships to norepinephrine deviated from those found in other hearts failing both due to ischemic and due to nonischemic cardiomyopathy (fig. 6). In this situation norepinephrine was about 30 times more potent at the alpha-1 compared to the beta adrenoceptors. The maximal inotropic responses mediated by the two adrenoceptor systems were, however, still comparable (table 3).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Inotropic responses expressed as developed tension (Tmax) to increasing concentrations of norepinephrine in ventricular myocardium from a patient with hypothyroidism failing due to ischemic cardiomyopathy: alpha-1 adrenoceptor-mediated response in the presence of 6 × 106 mol/liter timolol (, n = 2); beta adrenoceptor-mediated response in the presence of 6 × 106 mol/liter prazosin (, n = 2). Ordinate, Inotropic response in per cent of individual maxima. Abscissa, Log concentration of norepinephrine (mol/liter). Horizontal bars represent ± S.E. of pD2 values.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Our observations revealed the existence of an alpha-1 adrenoceptor-mediated inotropic response that can be elicited by norepinephrine also in the terminally failing human heart ventricle. This is supported by 1) the appearance of a slowly developing inotropic effect of norepinephrine in the presence of 6 × 10⁶ mol/liter timolol; 2) the sensitivity of this inotropic response to the alpha-1 adrenergic receptor blocker prazosin; 3) the qualitative characteristics of the response and 4) the insensitivity of this inotropic response to be attenuated by concomitant activation of muscarinic cholinergic receptors.

Furthermore, quantitatively the mean inotropic response evoked by activation of alpha-1 adrenergic receptors was found to be comparable to that evoked through activation of beta adrenergic receptors when expressed as maximal developed tension. This is in contrast to several ex vivo studies on failing human myocardium where the alpha-1 adrenoceptor-mediated inotropic response to phenylephrine has been found to be small compared to the beta adrenoceptor-mediated inotropic response to isoprenaline (e.g., Böhm et al. 1988, Brodde et al., 1992; Steinfath et al., 1992; Bristow 1993). Although a great variation in inotropic responses between different hearts was observed in our study, this apparently applied to the alpha-1 adrenoceptor-mediated and the beta-adrenoceptor-mediated inotropic responses in parallel (fig. 5).

In the presence of 6 × 10⁶ mol/liter prazosin, there was a fast developing inotropic response to norepinephrine. Qualitively this response was characterized by a shortening of the duration of the contraction-relaxation cycle. In addition the response was attenuated by concomitant activation of muscarinic receptors. These characteristics are typical for a beta adrenoceptor/cyclic AMP-dependent mediated inotropic response in mammalian heart muscle, e.g., of isoprenaline or of norepinephrine in the presence of an alpha-adrenoceptor blocker (Ledda et al., 1975, Brückner et al., 1978, Skomedal et al., 1982, Skomedal and Osnes, 1983, Aass et al., 1983, Skomedal et al., 1985; Aass et al., 1986; reviews: Scholz 1980; Osnes et al., 1985). This response was sensitive to the beta-adrenoceptor blocker timolol (fig. 2a) and was taken as the typical beta adrenoceptor-mediated component of norepinephrine in failing human ventricle.

In contrast to the situation described above, the inotropic response evoked by norepinephrine in the presence of 6 × 10⁶ mol/liter timolol developed slowly. Qualitively this response was characterized by a "symmetrical" change of the contraction-relaxation cycle with unchanged values for TPT, TR20 and RT (=TR20-TPT). In addition, the response was reinforced and not attenuated by concomitant activation of muscarinic receptors. These characteristics are typical for an alpha-1 adrenoceptor/cyclic AMP-independent mediated inotropic response in mammalian heart muscle, e.g., of phenylephrine or norepinephrine in the presence of a beta adrenoceptor blocker (Ledda et al., 1975, Brückner et al., 1978, Skomedal et al., 1982, Skomedal and Osnes, 1983, Aass et al., 1983, Skomedal et al., 1985; Aass et al., 1986; reviews: Scholz 1980; Osnes et al., 1985). This response was sensitive to the alpha-1 adrenoceptor blocker prazosin (fig. 2b) and was thus taken as the typical alpha-1 adrenoceptor-mediated component of norepinephrine in failing human ventricle.

Activation of muscarinic receptors by carbachol differentially influenced the alpha-1 adrenoceptor and the beta adrenoceptor-mediated inotropic responses. This phenomenon is well known from other mammalian species (Endoh and Motomura, 1979) and also described in human ventriclar muscle (Neumann et al., 1993; Scholz et al., 1996). The explanations for these different effects are mainly related to 1) the bidirectional regulation of adenylyl cyclase by beta adrenoceptors and muscarinic cholinergic receptors and accordingly to cyclic AMP mediated effects on e.g., sarcoplasmic reticulum and myofilaments (review: Sulakhe and Vo, 1995) and to 2) a possible reinforcement of breakdown of phosphoinositides by stimulation of alpha-1 adrenoceptors and of muscarinic cholinergic receptors (e.g., Brown et al., 1985; Poggioli et al., 1986; Ransnäs et al., 1986). The response to cholinergic agonists in this situation can thus be used as a functional test whether the cyclic AMP system is activated.

In experiments with norepinephrine in the absence of adrenoceptor blockers (combined alpha-1 and beta adrenoceptor stimulation), the time course deviated both from the time course to alpha-1 and from the time course to beta adrenoceptor stimulation (fig. 1, table 1). Thus, the development of the inotropic response in human failing heart to unopposed stimulation by norepinephrine was influenced by both receptor populations demonstrating an interaction between the two adrenergic receptor systems. This is in accordance with earlier findings in normal rabbit myocardium (Skomedal et al., 1990b).

Reversal responses to adrenergic receptor blockers revealed a functional contribution of the alpha-1-adrenoceptor-mediated component to the final and unopposed response to norepinephrine also in failing human myocardium. The relative contribution of the alpha-1 adrenoceptor-mediated (prazosin sensitive) component and of the beta adrenoceptor-mediated (timolol sensitive) component to the unopposed inotropic response to norepinephrine found in our work is in general agreement with findings in normal myocardium from rat and rabbit (Skomedal et al., 1988; Osnes et al., 1989, Skomedal et al., 1990a) and in atria from children with congenital heart defects (Borthne et al., 1995). Quantitatively, there is thus a discrepancy between the expression of the alpha-1 adrenoceptor-mediated inotropic response during selective activation and activation of the alpha-1 adrenoceptor concomitantly with the beta adrenoceptors. An inhibitory interaction between the alpha-1 and the beta adrenoceptors is thus revealed in failing human myocardium and this observation is also in general agreement with findings in normal myocardium from rat and rabbit (Skomedal et al., 1988, 1990a). These response patterns are thus apparently general ones for the dual adrenoceptor regulation of the mammalian myocardium whether normal or failing.

The concentration of norepinephrine used in the time course experiments was chosen to give close to maximal inotropic responses. In dose-response experiments also graded responses were demonstrated and these experiments provided classical dose-response relationships for norepinephrine in the presence of the respective adrenergic blockers. The typical qualitative characteristics of the alpha-1 adrenoceptor-mediated and the beta adrenoceptor-mediated inotropic responses were observed also in these experiments. Especially, at maximal concentrations of norepinephrine in the presence of 6 × 10⁶ mol/liter timolol there was no shortening of the duration of the contraction-relaxation cycle demonstrating the sufficiency of the beta adrenoceptor blockade (table 1).

Dose-response experiments indicated differences that may be interesting with respect to etiology of the heart failure. In failure due to ischemic cardiomyopathy, the alpha-1 and beta adrenoceptor systems were apparently equally sensitive to activation by norepinephrine (fig. 4a). In failure due to nonischemic cardiomyopathy, the beta adrenoceptor system was the more sensitive one by about one log unit compared to the alpha-1 adrenoceptor system (fig. 4b). This difference in sensitivity at the beta adrenoceptors did not parallel the difference in cardiac index observed between the ischemic and nonischemic heart failure groups, respectively. There were no obvious differences in responsiveness (expressed as maximal developed tension) between the two adrenergic systems to activation by norepinephrine although the mean responses were nominally less in nonischemic compared to ischemic cardiomyopathy (table 3). This nominal difference in responsiveness to adrenergic stimulation paralleled, however, the difference in cardiac index between the ischemic and the nonischemic failing hearts.

It has been reported that beta adrenergic responsiveness decreases with age in human heart (White et al., 1994). The observed difference in sensitivity to alpha-1 and beta adrenoceptor stimulation in preparations used for dose-reponse experiments may at least partly be due to age differences as the patients transplanted due to ischemic cardiomyopathy were about 20 yr older (51-61 yr, mean 56 yr) than the patients transplanted due to nonischemic cardiomyopathy (18-53 yr, mean 37 yr). Because our study was not designed to investigate more specifically mechanisms related to beta-adrenergic effects, other possible explanations for the observed differences will be speculative only. However, differences in the beta adrenergic effector mechanisms both with respect to downregulation and to uncoupling of receptors in ischemic vs. nonischemic cardiomyopathy have been reported (Bristow et al., 1991, review: Bristow, 1993).

In several reports the alpha-1 adrenoceptor-mediated inotropic effect of phenylephrine (in the presence of beta adrenoceptor blocker) has repeatedly been reported to be small compared to the beta adrenoceptor-mediated inotropic effect of isoprenaline (e.g., Böhm et al., 1988, Brodde et al., 1992; Steinfath et al., 1992; Bristow 1993). Although Böhm et al. (1988) suggested an increased importance of the alpha-1 adrenoceptor-mediated inotropic effect in severe heart failure, the alpha-1 adrenoceptor-mediated effect of phenylephrine in failing human heart has been considered to be of little interest, e.g., as a compensating mechanism for the attenuated beta adrenoceptor-mediated effect that is regularly observed in failing heart muscle (e.g., Brodde et al., 1992). Aass et al. (1986) were the first to report an alpha-1 adrenoceptor-mediated inotropic effect of norepinephrine in human ventricular myocardium. During the collection period for our data, Neumann et al. (1993) and Scholz et al. (1995, 1996) presented data in accordance with ours: norepinephrine is apparently able to elicit an alpha-1 adrenoceptor-mediated inotropic response in failing human ventricular myocardium that is of a magnitude that makes this component interesting compared to the beta adrenoceptor-mediated component. In our work, the alpha-1 adrenoceptor-mediated inotropic component of norepinephrine is of a magnitude comparable to the beta adrenoceptor-mediated inotropic component and will make it appropriate to reconsider the functional importance of this regulatory mechanism during terminal heart failure.

Norepinephrine and phenylephrine have apparently different intrinsic activities at the alpha-1 adrenoceptors in failing human heart. This has not been evident from ex vivo experiments in animal myocardium (e.g., Aass et al., 1983) and has, to our knowledge, not been addressed with respect to heart muscle before Neumann et al. (1993) and Scholz et al. (1995, 1996) presented their data where phenylephrine was found to be a partial agonist with very low intrinsic activity at the human cardiac alpha-1 adrenoceptors compared to norepinephrine. In our work phenylephrine exerted an inotropic response of about 70% compared to norepinephrine in preparations obtained from the same hearts. The time course curves were, however, almost superimposed (fig. 1), indicating no major difference between norepinephrine and phenylephrine in changing the kinetics of the involved processes. At present we have no explanation for the apparent difference between norepinephrine and phenylephrine with respect to intrinsic activity at the alpha-1 adrenoceptors in failing human heart muscle. To our knowledge, there is for example, no data indicating different intrinsic activity of norepinephrine and phenylephrine at different subtypes of alpha-1 adrenoceptors. However, the newly discovered alpha-1L subtype (Muramatsu et al., 1995; Noguchi et al., 1995) might be of interest because prazosin apparently is less potent in blocking alpha-1 adrenoceptor-mediated inotropic effects in human heart (Mügge et al., 1983, Skomedal et al., 1985) compared to rat and rabbit heart (Skomedal et al. 1980, Davey 1986, Hiramoto et al. 1988).

One patient was judged to be hypothyroid at the time of transplantation as indicated by an elevated level of TSH. This patient had a known hypothyroidism that was substituted with thyroxine. Due to serious arrhythmia the patient also received amiodarone which may cause hypothyroidism (review: Figge and Figge, 1990). Amiodarone will thus probably reinforce a hypothyroid effect upon the sensitivity of the cardiac beta adrenoceptors (e.g., Bjørnerheim et al., 1991). The sensitivity of the beta adrenoceptors to norepinephrine in this situation appeared to be markedly reduced compared to the alpha-1 adrenoceptors as expressed by a striking preference in stimulation of the alpha-1 adrenoceptors. Although these observations are from one heart only, the data are in general agreement with experimental findings in normal animal heart: hypothyroidism is known to increase the sensitivity to alpha-1 adrenocepor stimulation and to reduce the sensitivity to beta adrenoceptor stimulation (e.g., Nakashima et al., 1971, Kunos et al., 1974, Osnes 1976, Williams and Lefkowitz, 1979, Fox et al., 1985). Thus the thyroid status should apparently be kept in mind when interfering with the adrenoceptors of a failing heart.

In failing heart the beta adrenergic response is blunted (e.g., Brodde et al., 1992, Steinfath et al., 1992, Bristow, 1993). Our observations indicated that the alpha-1 adrenoceptor-mediated response thus may be of functional importance in comparison with the beta adrenoceptor-mediated response. In situations where the beta adrenoceptor-mediated inotropic component is further attenuated (cholinergic activity, use of beta adrenoceptor blockers, hypothyroidism), the alpha-1 adrenoceptor-mediated inotropic component may be even more important also during heart failure. Although speculative, an intriguing possibility is that the alpha-1 adrenoceptor-mediated effects may at least support the contractile reserve and may even contribute to the "recovery" of heart function seen in terminal failing hearts that are treated with beta-adrenoceptor blockers (see e.g., Eichhorn and Hjalmarson, 1994).

In conclusion, in contrast to findings with phenylephrine as alpha-1 adrenergic agonist, the endogenous agonist norepinephrine apparently exerts an alpha-1 adrenoceptor-mediated inotropic effect in failing human myocardium that is comparable to the beta-adrenoceptor-mediated one when expressed as developed tension. During concomitant activation of adrenergic receptors the final inotropic response to norepinephrine is mediated through both receptor systems. The alpha-1 adrenoceptor-mediated inotropic component is thus, although varying, apparently of potential functional importance in failing human ventricular myocardium.