- Email: [email protected]
Effect of phosphodiesterase inhibitors on human arteries in vitro
British Journal of Anaesthesia 1996; 76: 122–129
Effect of phosphodiesterase inhibitors on human arteries in vitro M. B. VROOM, M. PFAFFENDORE, H. B. VAN WEZEL AND P. A. VAN ZWIETEN
Summary In the present study, we investigated if the relaxant effects of phosphodiesterase (PDE) III inhibitors on human vessels could be inhibited by a nitric oxide synthase blocker, L-NAME, or by a blocker of ATPsensitive potassium channels (KATP), glibenclamide. The experiments were performed using an isometric myograph in isolated human s.c. small arteries obtained from healthy donors. After a priming procedure consisting of exposure to high potassium (120 mmol litre1) solutions, phenylephrine 10 mol litre1 and an equilibrium period of 30 min, the preparations were contracted with a thromboxane A2 mimetic agent, U46619 1 mol litre1. Subsequently, cumulative concentration–response curves were constructed for the selective PDE III inhibitors amrinone, milrinone and enoximone, and for theophylline and dipyridamole, with and without the addition of L-NAME 100 mol litre1 or glibenclamide 1 mol litre1. Addition of L-NAME to the organ bath resulted in significantly higher pEC50 values (log of the concentration required for 50 % relaxation) for milrinone compared with the control: 2.77 (SEM 0.24) mol litre1 (n 5) vs 3.49 (0.17) mol litre1 (n 7) (P 0.05). There was no significant difference between any other group. From our data we conclude that the relaxant properties of amrinone, enoximone, theophylline and dipyridamole are not dependent on nitric oxide release or on interaction with KATP channels. However, the effect of milrinone may be partly endothelium-dependent in human vessels in vitro. (Br. J. Anaesth. 1996; 76: 122–129) Key words Pharmacology, phosphodiesterase inhibitors. Pharmacology, nitric oxide. Pharmacology, glibenclamide. Arteries, relaxation.
Over the past decade, interest has focused on the selective inhibitors of the cardiac PDE peak III isoenzyme for the treatment of congestive heart failure, but the effects of chronic oral administration of these inotropic agents in this condition have been disappointing [6–8]. However, they continue to be used in the treatment of acute heart failure [9, 10]. One of the disadvantages of the currently available PDE III inhibitors, such as amrinone, milrinone and enoximone, is marked vasodilator activity at higher doses, with concomitant reductions in arterial pressure. The precise mechanism of this vasodilator activity remains to be elucidated. There is little doubt that selective inhibition of a low Km cAMP PDE, documented in both human arteries and arteries of different animal species, plays a major role [11–15]. In addition, several PDE inhibitors have been shown to produce endothelium-dependent vasodilatation in animal studies [16–18]. Recently, there has been mounting interest in the role of the ATP-dependent potassium (KATP) channels in cardiovascular tissue. Opening of these KATP channels may result not only in vasodilatation, but also in marked cardioprotective anti-ischaemic effects [19, 20]. In the present study, we investigated if the vasodilator effects of the PDE III inhibitors amrinone and milrinone, both bipyridines, the imidazolone derivative, enoximone, and the nonspecific PDE inhibitors, theophylline and dipyridamole, could be modulated by a nitric oxide synthase blocker, N G -nitro-L-arginine methyl ester hydrochloride (L-NAME) or by the KATP channel blocker, glibenclamide. The experiments were performed in isolated human s.c. small arteries, obtained from healthy donors.
Materials and methods VESSEL DONORS
Since the discovery of cyclic nucleotide phosphodiesterase (PDE) activity 30 yr ago, there have been major advances in knowledge on this group of enzymes [1]. Five families have been described, each composed of several isoforms with different patterns of tissue distribution, including cardiac muscle, vascular smooth muscle and platelets [2, 3]. Subsequently, selective isoenzyme inhibitors have been developed [4]. Theoretically, these agents may prove useful in a variety of clinical disorders, including heart failure, asthma, depression or dementia, or arterial thrombosis. None the less, their current use in clinical practice is still limited [5].
S.c. arteries were obtained from 11 healthy white female patients (aged 18–49 yr) undergoing reconstructive breast surgery. The study was approved by the hospital’s medical Ethics Committee and all patients gave informed consent. The patients were free from cardiovascular disorders and were not M. B. VROOM, MD, H. B. VAN WEZEL, MD, PHD (Department of Anaesthesiology); M. PFAFFENDORF, PHD, P. A. VAN ZWIETEN, MD, PHD (Departments of Pharmacotherapy and Cardiology); Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands. Accepted for publication: August 19, 1995. Correspondence to M.B.V.
Vasorelaxation by PDE inhibitors receiving any chronic drug treatments, except for oral contraceptive agents. Surgery was performed under general anaesthesia with volatile anaesthetics, supplemented with fentanyl and vecuronium. PREPARATION AND MOUNTING OF THE VESSELS
Small arteries were dissected atraumatically, immediately after removal of a skin segment with s.c. fat from the caudal segment of the patient’s breast. After dissection, a 40-m diameter stainless steel wire was inserted into the lumen. The length of the vessels was determined using an OSM-4 10 /13 ocular lens (Olympus, Japan). Subsequently, the vessels (estimated length of 1–2 mm) were placed in an organ preservation solution (University of Wisconsin (UW)) [21] and stored at 4 ⬚C for a maximum of 7 days. Before the experiment, the preparations were transferred to the chamber of an isometric myograph [22]. The wire was fixed at both ends to microscrews and a second wire was inserted, which was subsequently connected to a force transducer (Kistler Morse, DSG 6, Redmond, WA, USA). This enabled continuous measurement of the isometric wall tension of the vessel. Reproducible measurements could be obtained within at least 7 h. The preparations were suspended in Tyrode’s solution for 10 min at 37 ⬚C. Throughout the experiment, the buffer solution was oxygenated with carbogen (95 % oxygen, 5 % carbon dioxide) to achieve a pH of 7.4. Subsequently, the ratio between passive wall tension and internal circumference was determined by a normalization procedure, as described by Mulvany and Halpern [23]. The internal circumference was adjusted to a value that equals 90 % of the circumference generated by a transmural pressure of 100 mm Hg (13.3 kPa). This value was then divided by to obtain the normalized lumen diameter. Mechanical responses were expressed as active tension, ␦T (N m1), that is the developed active force divided by twice the vessel segment length. At the beginning of each experiment, the preparations were contracted three times for 5 min with high potassium Tyrode’s solution (K 120 mmol litre1, see below) and once with phenylephrine 10 mol litre1, an ␣1 adrenoceptor agonist, at intervals of 15 min. Between these episodes the chambers were flushed at least three times with Tyrode’s solution. This episode was followed by an equilibrium period of 30 min for the L-NAME and control groups and 60 min for the glibenclamide group. L-NAME 100 mol litre1 or glibenclamide 1 mol litre1 was added at the beginning of this period and remained present in the bath thereafter. An equilibrium period of 60 min was chosen for glibenclamide as this drug has a slow onset of action. Subsequently, U46619, a thromboxane A2 mimetic agent, was administered to all preparations at a concentration of 1 mol litre1. After stabilization of the contraction, a cumulative concentration– response curve, using concentrations of 0.1 mol litre1 to 1 mmol litre1, was constructed for amrinone, milrinone, enoximinone, theophylline and dipyridamole.
123 Our experimental model included two separate force transducers connected to one single organ bath, which enabled simultaneous evaluation of the contractile responses of two vessels to one drug combination. Therefore, the experiments described were performed in duplicate if sufficient vessels could be harvested from the patient’s breast. The data of these duplicate experiments were averaged and listed in the results section for each individual patient. To evaluate the effects of the preservation technique on endothelial function, the response of contracted arteries to methacholine 105 mol litre1 was determined immediately after dissection of the blood vessel and after preservation in UW fluid. These vessels were not used in any other experiment. SOLUTIONS AND DRUGS USED
In all experiments, Tyrode’s solution of the following composition was used (mmol litre1): NaCl 136; NaHCO3 11.9; NaH2PO4 0.42; KCl 2.5; CaCl2 1.8; MgCl2 0.5; and glucose 5.5. In the high potassium Tyrode’s solution, NaCl 117.5 mmol litre1 was replaced by KCl, resulting in a total potassium concentration of 120 mmol litre1. These solutions were prepared just before each experiment. Aliquots of 20 ml were kept at 37 ⬚C and bubbled with carbogen before they were introduced into the bath. Amrinone and milrinone (donated by Sanofi Winthrop, Dev. Centre, Fawdon, UK) and enoximone (donated by Marion Merrell Dow Inc., Cincinatti, OH, USA) were dissolved to a concentration of 101 mol litre1. Milrinone was dissolved in distilled H2O, amrinone and enoximone were dissolved in 99 % dimethylsulphoxide (DMSO). Subsequently, these agents were diluted to 102–105 mol litre1 in distilled water (amrinone and milrinone) or 66.7 % DMSO (enoximone). Theophylline and dipyridamole (Sigma Chemical Co., USA) were dissolved in distilled water and 66.7 % DMSO, respectively, and subsequently diluted to 102–105 mol litre1 in distilled water. The final concentrations (1 %) of DMSO in the bathing medium had no significant effect on vascular smooth muscle contraction and relaxation. (See fig. 1 for the chemical structures of the various compounds studied.) Nitro-L-arginine methyl ester hydrochloride (LNAME) (Sigma), U46619 (9,11-dideoxy-11␣,9␣epoxy methanoprostaglandin F2␣) (Sigma) and Lphenylephrine hydrochloride (Sigma) were dissolved in distilled water. Glibenclamide (Hoechst) was dissolved in 66.7 % DMSO. DATA ANALYSIS
A log concentration–response curve was fitted to the data of individual experiments using GraphPad software (GraphPad, Institute for Scientific Information, San Diego, CA, USA). This computer program allows calculation of the EC50 by non-linear regression analysis, which is the concentration of the compound required to achieve the half maximal relaxant effect, based on the equation: E
124
British Journal of Anaesthesia
Figure 1
Chemical structures of the various compounds studied.
Emax AP/(AP EC50P ), where E the response obtained with a given concentration A; the exponent P describes the slope of the relationship (Hill coefficient); and Emax the maximally obtainable effect. All data are presented as mean (SEM). Differences between means were compared using a two-tailed Student’s t test for unpaired data or analysis of variance where appropriate (Graphpad Instat). P 0.05 was considered significant.
Results The resistance arteries dissected from the mammary tissues and used to evaluate the effects of the PDE III inhibitors (n 116) had an average diameter of 415 (20) m and a length of 1.8 (0.03) mm. There were no significant differences in main vessel characteristics between the groups (table 1). All arteries showed spontaneous rhythmic activity under the experimental conditions. The preparations remained stable and fully responsive to methacholine Table 1 Characteristics of the human vessels used in this study (mean (SEM)). L-NAME NG-nitro-L-arginine methyl ester Age (days)
Length (mm)
Circumference (µm)
Amrinone L-NAME Glibenclamide
3.7 (1.5) 3.8 (1.3) 2.1 (0.4)
1.8 (0.1) 2.0 (0.0) 1.9 (0.1)
482 (80) 338 (102) 419 (57)
Milrinone L-NAME Glibenclamide
4.0 (1.4) 3.8 (1.8) 2.5 (0.6)
1.8 (0.1) 1.9 (0.1) 2.0 (0.0)
403 (59) 385 (96) 383 (61)
Enoximone L-NAME Glibenclamide
3.1 (0.9) 4.0 (1.5) 2.1 (0.5)
1.9 (0.1) 2.0 (0.0) 1.8 (0.2)
434 (79) 479 (125) 336 (51)
Theophylline L-NAME Glibenclamide Dipyridamole L-NAME Glibenclamide
3.0 (0.9) 2.6 (1.2) 2.3 (1.0) 3.4 (1.6) 3.8 (1.5) 3.8 (1.5)
2.0 (0.0) 1.6 (0.2) 1.6 (0.2) 1.5 (0.1) 1.9 (0.1) 1.9 (0.1)
377 (58) 259 (50) 357 (65) 287 (43) 305 (68) 337 (57)
for at least 1 week after preparation when stored in the UW solution. Maximal relaxation to methacholine 105 mol litre1 was 97.5 (1.5) % (n 4) on day 0 vs 97.3 (1.8) % (n 4) on day 7 (ns). Preservation UW fluid had no influence on the contractile response to U46619 1 mol litre1, as linear regression analysis revealed no time-dependent change (correlation coefficient r 0.159, slope not significantly different from zero) (fig. 2). In addition, the contractile responses to phenylephrine remained stable with time: 3.92 (0.32) N m1 (n 22) on day 0 vs 5.05 (1.62) (n 4) (ns) on day 7. All PDE inhibitors induced concentrationdependent vasodilatation in the isolated arteries. Dipyridamole was the most potent vasodilator in our study (pEC50 4.67 (0.22) mol litre1) and theophylline the least active (pEC50 2.72 (0.07) mol litre1). The pEC50 values for amrinone, milrinone and enoximone were similar in the control groups: 3.28 (0.01), 3.49 (0.17) and 3.29 (0.08) mol litre1, respectively. Compared with potassium and phenylephrine, U46619 proved to be the most potent vasoconstrictor
Figure 2 Linear regression analysis of the contractile response to U46619 1 mol litre1 with respect to days of vessel preservation.
Vasorelaxation by PDE inhibitors
125
Table 2 Contractile response to U46619 1 mol litre1 (N m1) in small human isolated vessels (mean (SEM)); influence of NG-nitro-L-arginine methyl ester (L-NAME) and glibenclamide
Amrinone Milrinone Enoximone Theophylline Dipyridamole
Control
L-NAME
Glibenclamide
No. individuals studied Force
No. individuals studied Force
No. individuals studied
Force
6 7 6 5 5
5 5 5 5 5
6 6 6 8 5
3.68 (0.67) 5.64 (1.24) 4.66 (0.98) 4.20 (0.68) 3.95 (1.38)
5.44 (1.64) 6.06 (1.54) 5.65 (1.77) 5.71 (0.65) 4.03 (2.11)
4.76 (0.75) 4.49 (0.70) 6.84 (1.21) 4.07 (0.64) 4.31 (1.65)
Table 3 Contractile response (N m1) to the fourth potassium administration (K+) and to phenylephrine 10 mol litre1 (Phe) in small human isolated vessels (mean (SEM)). L-NAME N G-nitro-L-arginine methyl ester L-NAME
Control
Amrinone Milrinone Enoximone Theophylline Dipyridamole
Glibenclamide
K+
Phe
K+
Phe
K+
Phe
4.78 (1.55) 5.66 (1.34) 5.00 (1.67) 4.81 (0.69) 4.39 (0.43)
4.03 (1.46) 4.94 (1.31) 4.16 (1.49) 4.86 (0.50) 4.03 (0.38)
4.17 (0.71) 3.71 (0.60) 5.76 (0.92) 3.45 (0.68) 4.15 (1.33)
3.15 (0.58) 3.18 (0.51) 4.65 (0.65) 2.84 (0.67) 3.63 (1.24)
4.12 (0.94) 5.40 (1.11) 4.20 (0.87) 4.00 (0.66) 4.01 (1.01)
3.69 (0.82) 5.02 (1.03) 3.63 (0.75) 3.52 (0.58) 3.70 (0.90)
Table 4 pEC50 values of the phosphodiesterase inhibitors with respect to the vasodilatation induced by various inhibitors in precontracted small human arteries (mean (SEM)). pEC50 log of the concentration of the compound required to achieve the half maximal relaxation effect; L-NAME N G-nitro-L-arginine methyl ester. *P 0.05 compared with the respective control group
Amrinone Milrinone Enoximone Theophylline Dipyridamole
Control
L-NAME
Glibenclamide
No. individuals studied pEC50
No. individuals studied
pEC50
No. individuals studied pEC50
6 7 6 5 5
5 5 5 5 5
3.35 (0.09) 2.77 (0.24)* 3.41 (0.10) 2.69 (0.06) 4.76 (0.09)
6 6 6 8 5
3.28 (0.10) 3.49 (0.17) 3.29 (0.08) 2.72 (0.07) 4.67 (0.22)
in all experiments. The contractile response to U46619 was not influenced by the administration of L-NAME or glibenclamide. The basal values of the U46619-induced response were comparable in all preparations (table 2). Of the 85 experiments described in table 2, 31 experiments were performed in duplicate, accounting for the total of 116 experiments. Both the contractile response to U46619 and the vasodilator effects of amrinone, milrinone, enoximone, theophylline and dipyridamole were completely reversible on washout. The contractile responses to the fourth potassium administration and to phenylephrine are shown in table 3. Amrinone was administered in cumulative concentrations of 108 to 103 mol litre1. At the highest concentration, amrinone caused relaxation of U46619-induced contraction by 72.4 % of the initial value. Milrinone and enoximone were also administered in a concentration range of 108–103 mol litre1. The maximal relaxant effects were 71.4 % and 82.2 %, respectively. Theophylline proved less potent and effective, as higher doses were required (3.103 mol litre1) and a lower degree of maximal relaxation (57.8 %) was achieved. Dipyridamole
3.25 (0.17) 3.46 (0.15) 3.49 (0.06) 2.76 (0.19) 4.53 (0.02)
Figure 3 Log concentration-response curve for amrinone in the absence (control (n 6) ( )) or presence of L-NAME (n 5) ( ) or glibenclamide (n 6) ).
126
British Journal of Anaesthesia
Figure 4 Log concentration-response curve for enoximone in the absence (control (n 6) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 6) (●).
Figure 6 Log concentration-response curve for dipyridamole in the absence (control (n 5) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 5) (●).
Figure 5 Log concentration-response curve for theophylline in the absence (control (n 5) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 8) (●).
Figure 7 Log concentration-response curve for milrinone in the absence (control (n 7) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 6) (●). In the presence of L-NAME, a significant (P 0.05) rightward shift of the relaxation curve was observed.
concentrations ranged from 108 to 104 mol litre1, resulting in 83 % maximum relaxation of the basal value. Neither L-NAME nor glibenclamide influenced the basal value of the contractile response to U46619 when administered simultaneously with the PDE inhibitors. In addition, neither L-NAME nor glibenclamide influenced the relaxant responses to amrinone, enoximone, theophylline or dipyridamole (table 4 and figs 3–6). However, in the presence of LNAME a significant (P 0.05) rightward shift was observed in the relaxation curve of milrinone. Accordingly, the pEC50 values of milrinone were also influenced by L-NAME: 2.77 (0.24) mol litre1 with
compared with 3.49 (0.17) mol litre1 without L-NAME (P 0.05) (fig. 7). L-NAME
Discussion Using animal models, several investigators have demonstrated that the PDE III inhibitors amrinone, milrinone and enoximone induce vasodilatation as a result of inhibition of cGMP-inhibited (cGI) PDE, thus leading to increased intracellular levels of cAMP [11–14]. The effects of theophylline and dipyridamole on vascular smooth muscle are more complex. The mode of action of theophylline involves
Vasorelaxation by PDE inhibitors both non-specific phosphodiesterase inhibition and adenosine receptor antagonism [24]. In addition, unknown cellular mechanisms appear to play an additional role [25]. Dipyridamole is not only a PDE inhibitor, but also modulates adenosine metabolism, as it inhibits the cellular uptake of adenosine [26, 27]. In the present study, all agents caused concentration-dependent vasodilatation. The concentrations required were rather high compared with the results reported in the literature. However, these data may not be comparable with our results as various types of tissues were used in these studies and were obtained from different species. The use of human vessel preparations provides an obvious advantage when studying the action of agents with an influence on PDE activity, as the distribution of PDE isoenzymes is known to differ considerably between species [28]. Only Lindgren and colleagues used human vessels in their studies, but these tissues were obtained either post mortem or from patients undergoing laparotomy for various abdominal disorders, and not from healthy patients as in our study [11, 29]. Interestingly, these authors showed the importance of the mode of produced contraction on both the magnitude of the relaxant effect and the potency of PDE inhibitors on human vessels in vitro. Contractions of isolated arteries induced by high potassium concentrations (K 127 mmol litre1) were more difficult to relax by PDE III inhibitors compared with lower K concentrations (30 mmol litre1) or noradrenaline-induced contractions. Similarly, Harris and colleagues demonstrated a difference in efficacy of PDE III inhibitors in voltage-dependent (K 80 mmol litre1) as opposed to receptor-mediated (U46619) vasoconstriction [30]. The response to the PDE inhibitors in the present study may, therefore, be related partially to the method by which vasoconstriction was induced. Other important factors which may influence the efficacy of PDE inhibitors are the characteristics of the vascular beds from which the human vessels are taken. Marked variations in vascular responsiveness of isolated human coronary, renal and lung arteries have been demonstrated by Lindgren and Andersson [29]. In potassium (30 mmol litre1) contracted lung arteries, cGI-PDE inhibitors (including milrinone 3 105 mol litre1) were incapable of producing 50 % relaxation. In contrast, the agents investigated were significantly more potent and effective in renal and coronary arteries (pEC50 for milrinone: 3.2 106 and 1.7 106 mol litre1, respectively). This reduced efficacy in both lung arteries in the aforementioned investigation and mammary arteries in our study may indicate low cGI-PDE activities in these particular tissues. Although not all vessels were used immediately after dissection, but stored in UW organ preservation fluid, there were no differences between the vessel characteristics, including maximal force development and function of endothelium (response to methacholine), between vessels of several minutes to several days old. Therefore, it is unlikely that this preservation technique was responsible for these findings. Thus far there are no reports in the literature that document the preservation of isolated vessel preparations in UW solution. Successful
127 preservation in UW fluid for 72 h, however, has been demonstrated for entire organs taken from animals [31, 32]. In our study, only the vasodilator effects of milrinone were partially endothelium-dependent. The addition of L-NAME 100 mol litre1 resulted in a significant (P 0.05) increase in the pEC50 value of milrinone compared with controls: 2.77 (0.24) vs 3.49 (0.17) mol litre1, respectively. So far we cannot explain why the vasodilator profile of milrinone was distinctly different from that of its structurally related analogue, amrinone. Harris and colleagues also described differential relaxant effects of milrinone and amrinone in canine arteries [30]. In their study milrinone proved a more potent vasodilator against receptor-mediated (U46619) as opposed to K-induced depolarizations, whereas amrinone appeared equipotent in these conditions. The reason for these different effects could not be elucidated. Limited data are available in the literature on the interaction between the currently used PDE inhibitors and endothelium. Amrinone was reported to cause vasodilatation by both endothelium-dependent and -independent mechanisms in small pulmonary arteries of rabbits [16]. The vasodilator effects of enoximone in isolated porcine pulmonary arteries were endothelium-independent and unchanged by the addition of L-NAME 104 mol litre1 [33]. Finally, Kauffman and colleagues demonstrated that the vasodilator effects of milrinone in concentrations ranging from 107 to 105 mol litre1 occurred predominantly independent of the presence of an intact endothelium in aortic strips from rats [14]. In addition, these authors demonstrated that a high concentration of milrinone (104 mol litre1) resulted in increased concentrations of the second messenger cGMP next to enhanced levels of the intracellular mediator cAMP. Furthermore, in the presence of methylene blue, the dose–response curve for milrinone was shifted to the right. Methylene blue acts as an inhibitor of soluble guanyl cyclase and also nitric oxide synthase [34]. Clearly the mechanism by which PDE inhibitors induce vasodilatation is complex, as recent evidence has been obtained demonstrating that cross-interactions between cAMP and cGMP occur at different levels [35, 36]. At high concentrations milrinone acts on PDE IV and PDE V, next to PDE III inhibition [36], PDE IV hydrolyses cAMP, whereas PDE V acts as a specific cGMP hydrolyser. It has been demonstrated that vasodilatation induced by selective PDE IV inhibitors were abolished by both L-NAME and methylene blue, indicating that the effect of PDE IV inhibitors is dependent on nitric oxide production and subsequent activation of soluble guanylate cyclase [36]. Thus, the aforementioned results combined with our data may indicate that milrinone acts as a PDE III inhibitor, independent of endothelial function, at low concentrations, whereas at higher concentrations (105 mol litre1) the inhibitory activity of milrinone on PDE IV, which can be partially abolished by L-NAME becomes more prominent. The interaction of glibenclamide with milrinone
128 was reported by Fujimoto [37]. In his experiments, inhibition of milrinone-induced vasodilatation by glibenclamide in vascular preparations of rats and guineapigs, was demonstrable only at glibenclamide concentrations as high as 10–30 mol litre1 [37]. However, at these high concentrations glibenclamide may not act as a specific KATP channel blocker [38]. In the present investigation glibenclamide 1 mol litre1 did not influence the vasodilator effect of any of the PDE inhibitors studied, indicating that KATP channels are not involved in the process of vasodilatation caused by these agents. In conclusion, L-NAME sensitive nitric oxide release may contribute to the vasodilator properties of milrinone, but not to those of any of the other PDE inhibitors studied. Vasodilatation caused by PDE inhibitors does not appear to involve KATP channels.
Acknowledgements We acknowledge the support offered by Dr E. J. F. Timmenga and Dr C. M. A. M. van der Horst, plastic and reconstructive surgeons, in the harvesting of human vascular tissue.
References 1. Butcher RW, Sutherland EW. Adenosine 3⬘,5⬘-phosphate in biological materials. Journal of Biological Chemistry 1962: 237; 1244–1250. 2. Beavo JA. Multiple isoenzymes of cyclic nucleotide phosphodiesterase. Advances in Second Messenger and Phosphoprotein Research 1988; 22: 1–37. 3. Weishaar RE, Burrows SD, Kobylarz DC, Quade MM, Evans DB. Multiple molecular forms of cyclic nucleotide phosphodiesterase in cardiac and smooth muscle and in platelets. Isolation, characterization and effects of various reference phosphodiesterase inhibitors and cardiotonic agents. Biochemical Pharmacology 1986; 5: 787–800. 4. Thompson WJ. Cyclic nucleotide phosphodiesterases: Pharmacology, biochemistry and function. Pharmacology and Therapeutics 1991; 51: 13–33. 5. Nicholson CD, Challis RAJ, Shahid M. Differential modulation of tissue function and therapeutic potential of selective inhibitors of cyclic nucleotide phosphodiesterase isoenzymes. Trends in Pharmacological Sciences 1991; 12: 19–27. 6. Packer M, Carver JR, Rodeheffer RJ, Ivanoe RJ, DiBianco R, Zeldis SM, Hendrix GH, Bommer WJ, Elkayam U, Kukin ML, Mallis GI, Sollano JA, Shannon J, Tandon PK, De Mets DL. Effect of oral milrinone on mortality in severe chronic heart failure: Promise study. New England Journal of Medicine 1991; 325: 1468–1475. 7. DiBianco R, Shabetia R, Kostuk W, Moran J, Schlant RC, Wright R. A comparison of oral milrinone, digoxin and their combination in the treatment of chronic heart failure. New England Journal of Medicine 1989; 320: 677–683. 8. Uretsky BF, Jessup M, Konstam MA, Dec W, Leier CV, Benotti J, Murali S, Herrmann HC, Sandberg JA. For the enoximone multicenter trial group. Multicenter trial of oral enoximone in patients with moderate to moderately severe congestive heart failure. Circulation 1990; 82: 774–780. 9. Dupuis JY, Bondy R, Cattran C, Nathan HJ, Wynands JE. Amrinone and dobutamine as primary treatment of low cardiac output syndrome following coronary artery surgery: A comparison of their effects on hemodynamics and outcome. Journal of Cardiothoracic and Vascular Anesthesia 1992; 6: 542–553. 10. Caldicott LD, Hawley K, Heppel R, Woodmansey PA, Channer KS. Intravenous enoximone or dobutamine for severe heart failure after acute myocardial infarction: a randomized double blind trial. European Heart Journal 1993; 14: 696–700.
British Journal of Anaesthesia 11. Lindgren S, Andersson KE, Belfrage P, Degerman E, Manganiello VC. Relaxant effects of the selective phosphodiesterase inhibitors milrinone and OPC 3911 on isolated human mesenteric vessels. Pharmacology and Toxicology 1989; 64: 440–445. 12. Silver PJ, Lepore RE, O’Connor B, Lemp BM, Hamel LT, Bentley RG, Harris, AL. Inhibition of the low Km cyclic AMP system in vascular smooth muscle by milrinone. Journal of Pharmacology and Experimental Therapeutics 1988; 247: 34–42. 13. Itoh H, Kusagawa M, Shimomura A, Suga T, Ito M, Konishi T, Nakano T. Ca2 dependent and Ca2 independent vasorelaxation induced by cardiotonic phosphodiesterase inhibitors. European Journal of Pharmacology 1993; 240: 57–66. 14. Kauffman RF, Schenck KW, Utterbach BG, Crowe G, Cohen ML. In vitro vascular relaxation by new inotropic agents: Relationship to phosphodiesterase inhibition and cardiac nucleotides. Journal of Pharmacology and Experimental Therapeutics 1987; 242: 864–872. 15. Weishaar RE, Burrows SD, Kobylarz DC, Quade MD, Evans DB. Multiple molecular forms of cyclic nucleotide phosphodiesterase in cardiac and smooth muscle and in platelets: Isolation, characterization and effects of various reference phosphodiesterase inhibitors and cardiotonic agents. Biochemical Pharmacology 1986; 35: 787–800. 16. Clarke WR, Soltow LOG. Amrinone causes vasodilation by both endothelium-independent and endothelium dependent mechanisms in small pulmonary arteries. American Review of Respiratory Diseases 1991; (Suppl A773). 17. Rosenblum WI, Shimizu T, Nelson GH. Interaction of endothelium with dilation produced by inhibitors of cyclic nuclide diesterases in mouse brain arterioles in vivo. Stroke 1993; 24: 266–270. 18. Martin W, Furchgott RF, Villani GM, Jothianandan D. Phosphodiesterase inhibitors induce endothelium-dependent relaxation of rat and rabbit aorta by potentiating the effects of spontaneously released endothelium-derived relaxing factor. Journal of Pharmacology and Experimental Therapeutics 1986; 237: 539–547. 19. Grover GJ, Sleph PG, Dzwonczyk S. Role of myocardial ATP-sensitive potassium channels in mediating preconditioning in the dog heart and their possible interaction with adenosine A1 receptors. Circulation 1992; 86: 1310–1316. 20. Gross GJ, Auchampach JA. Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circulation Research 1992; 70: 223–233. 21. Southard JH, Van Gulik TM, Ametani MS, Vreugdenhil PK, Lindell SL, Pinaar BL, Belzer FO. Important components of the UW solution. Transplantation 1990; 49: 251–257. 22. Mulvany MJ, Halpern W. Mechanical properties of vascular smooth muscle cells in situ. Nature (London) 1976; 260: 617–618. 23. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circulation Research 1977; 41: 10–26. 24. Fredholm BB, Persson CGA. Xanthine derivatives as adenosine receptor antagonists. European Journal of Pharmacology 1982; 81: 673–676. 25. Persson CGA, Andersson KE, Kjellin G. Effects of enprofylline and theophylline may show the role of adenosine. Life Sciences 1986; 38: 1057–1072. 26. Kubler W, Spieckermann PA, Bretschneider HJ. Influence of dipyridamole (Persantin) on myocardial adenosine metabolism. Journal of Molecular and Cellular Cardiology 1970; 1: 23–27. 27. Klabunde RE. Effects of dipyridamol on postischemic vasodilation and extracellular adenosine. American Journal of Physiology 1983; 244: H273–H280. 28. Weishaar RE, Kobylarz-Singer DC, Steffen RP, Kaplan HR. Subclasses of cyclic AMP, specific phosphodiesterase in left ventricular muscle and their involvement in myocardial contractility. Circulation Research 1987; 61: 539–547. 29. Lindgren S, Andersson KE. Effects of selective phosphodiesterase inhibitors on isolated coronary, lung and renal arteries from man and rat. Acta Physiologica Scandinavica 1991; 142: 77–82.
Vasorelaxation by PDE inhibitors 30. Harris AL, Grant AM, Silver PJ, Evans DB, Alousi AA. Differential vasorelaxant effects of milrinone and amrinone on contractile responses of canine coronary, cerebral, and renal arteries. Journal of Cardiovascular Pharmacology 1989; 13: 238–244. 31. Ploeg RJ, Goossens D, McAnulty JF, Stouthard JH, Belzer FO. Successful 72 hour cold storage of dog kidneys with UW solution. Transplantation 1988; 46: 191–196. 32. Wahlberg JA, Love R, Landegaard L, Stouthard JH, Belzer FO. 72 hours preservation of the canine pancreas. Transplantation 1987; 43: 5–8. 33. Butt AY, Dinh-Xuan AT, Pepke-Zaba J, Cremona G, Clelland CA, Higenbottam TW. In vitro pulmonary vasorelaxant effect of the phosphodiesterase inhibitor enoximone. Angiology 1993; 12: 289–294.
129 34. Mayer B, Brunner F, Schmidt K. Novel actions of methylene blue. European Heart Journal 1993; 14: (Suppl. I) 22–26. 35. Jiang J, Colbran JL, Francis SH, Corbin JD. Direct evidence for cross-activation of cGMP-dependent protein kinase by cAMP in pig coronary arteries. Journal of Biochemical Chemistry 1992; 267: 1015–1019. 36. Lugnier C, Komas N. Modulation of vascular cyclic nucleotide phosphodiesterases by cyclic GMP : role in vasodilatation. European Heart Journal 1993; 14 (Suppl. I): 141–148. 37. Fujimoto S. Effects of pimobendan, its active metabolite UDCG 212 and milrinone on isolated blood vessels. European Journal of Pharmacology 1994; 265: 159–166. 38. Ashcroft FM. Adenosine 5⬘-triphosphate-sensitive potassium channels. Annual Review of Neuroscience 1988; 11: 97.
Effect of phosphodiesterase inhibitors on human arteries in vitro M. B. VROOM, M. PFAFFENDORE, H. B. VAN WEZEL AND P. A. VAN ZWIETEN
Summary In the present study, we investigated if the relaxant effects of phosphodiesterase (PDE) III inhibitors on human vessels could be inhibited by a nitric oxide synthase blocker, L-NAME, or by a blocker of ATPsensitive potassium channels (KATP), glibenclamide. The experiments were performed using an isometric myograph in isolated human s.c. small arteries obtained from healthy donors. After a priming procedure consisting of exposure to high potassium (120 mmol litre1) solutions, phenylephrine 10 mol litre1 and an equilibrium period of 30 min, the preparations were contracted with a thromboxane A2 mimetic agent, U46619 1 mol litre1. Subsequently, cumulative concentration–response curves were constructed for the selective PDE III inhibitors amrinone, milrinone and enoximone, and for theophylline and dipyridamole, with and without the addition of L-NAME 100 mol litre1 or glibenclamide 1 mol litre1. Addition of L-NAME to the organ bath resulted in significantly higher pEC50 values (log of the concentration required for 50 % relaxation) for milrinone compared with the control: 2.77 (SEM 0.24) mol litre1 (n 5) vs 3.49 (0.17) mol litre1 (n 7) (P 0.05). There was no significant difference between any other group. From our data we conclude that the relaxant properties of amrinone, enoximone, theophylline and dipyridamole are not dependent on nitric oxide release or on interaction with KATP channels. However, the effect of milrinone may be partly endothelium-dependent in human vessels in vitro. (Br. J. Anaesth. 1996; 76: 122–129) Key words Pharmacology, phosphodiesterase inhibitors. Pharmacology, nitric oxide. Pharmacology, glibenclamide. Arteries, relaxation.
Over the past decade, interest has focused on the selective inhibitors of the cardiac PDE peak III isoenzyme for the treatment of congestive heart failure, but the effects of chronic oral administration of these inotropic agents in this condition have been disappointing [6–8]. However, they continue to be used in the treatment of acute heart failure [9, 10]. One of the disadvantages of the currently available PDE III inhibitors, such as amrinone, milrinone and enoximone, is marked vasodilator activity at higher doses, with concomitant reductions in arterial pressure. The precise mechanism of this vasodilator activity remains to be elucidated. There is little doubt that selective inhibition of a low Km cAMP PDE, documented in both human arteries and arteries of different animal species, plays a major role [11–15]. In addition, several PDE inhibitors have been shown to produce endothelium-dependent vasodilatation in animal studies [16–18]. Recently, there has been mounting interest in the role of the ATP-dependent potassium (KATP) channels in cardiovascular tissue. Opening of these KATP channels may result not only in vasodilatation, but also in marked cardioprotective anti-ischaemic effects [19, 20]. In the present study, we investigated if the vasodilator effects of the PDE III inhibitors amrinone and milrinone, both bipyridines, the imidazolone derivative, enoximone, and the nonspecific PDE inhibitors, theophylline and dipyridamole, could be modulated by a nitric oxide synthase blocker, N G -nitro-L-arginine methyl ester hydrochloride (L-NAME) or by the KATP channel blocker, glibenclamide. The experiments were performed in isolated human s.c. small arteries, obtained from healthy donors.
Materials and methods VESSEL DONORS
Since the discovery of cyclic nucleotide phosphodiesterase (PDE) activity 30 yr ago, there have been major advances in knowledge on this group of enzymes [1]. Five families have been described, each composed of several isoforms with different patterns of tissue distribution, including cardiac muscle, vascular smooth muscle and platelets [2, 3]. Subsequently, selective isoenzyme inhibitors have been developed [4]. Theoretically, these agents may prove useful in a variety of clinical disorders, including heart failure, asthma, depression or dementia, or arterial thrombosis. None the less, their current use in clinical practice is still limited [5].
S.c. arteries were obtained from 11 healthy white female patients (aged 18–49 yr) undergoing reconstructive breast surgery. The study was approved by the hospital’s medical Ethics Committee and all patients gave informed consent. The patients were free from cardiovascular disorders and were not M. B. VROOM, MD, H. B. VAN WEZEL, MD, PHD (Department of Anaesthesiology); M. PFAFFENDORF, PHD, P. A. VAN ZWIETEN, MD, PHD (Departments of Pharmacotherapy and Cardiology); Academic Medical Centre, Meibergdreef 9, 1105 AZ Amsterdam, Netherlands. Accepted for publication: August 19, 1995. Correspondence to M.B.V.
Vasorelaxation by PDE inhibitors receiving any chronic drug treatments, except for oral contraceptive agents. Surgery was performed under general anaesthesia with volatile anaesthetics, supplemented with fentanyl and vecuronium. PREPARATION AND MOUNTING OF THE VESSELS
Small arteries were dissected atraumatically, immediately after removal of a skin segment with s.c. fat from the caudal segment of the patient’s breast. After dissection, a 40-m diameter stainless steel wire was inserted into the lumen. The length of the vessels was determined using an OSM-4 10 /13 ocular lens (Olympus, Japan). Subsequently, the vessels (estimated length of 1–2 mm) were placed in an organ preservation solution (University of Wisconsin (UW)) [21] and stored at 4 ⬚C for a maximum of 7 days. Before the experiment, the preparations were transferred to the chamber of an isometric myograph [22]. The wire was fixed at both ends to microscrews and a second wire was inserted, which was subsequently connected to a force transducer (Kistler Morse, DSG 6, Redmond, WA, USA). This enabled continuous measurement of the isometric wall tension of the vessel. Reproducible measurements could be obtained within at least 7 h. The preparations were suspended in Tyrode’s solution for 10 min at 37 ⬚C. Throughout the experiment, the buffer solution was oxygenated with carbogen (95 % oxygen, 5 % carbon dioxide) to achieve a pH of 7.4. Subsequently, the ratio between passive wall tension and internal circumference was determined by a normalization procedure, as described by Mulvany and Halpern [23]. The internal circumference was adjusted to a value that equals 90 % of the circumference generated by a transmural pressure of 100 mm Hg (13.3 kPa). This value was then divided by to obtain the normalized lumen diameter. Mechanical responses were expressed as active tension, ␦T (N m1), that is the developed active force divided by twice the vessel segment length. At the beginning of each experiment, the preparations were contracted three times for 5 min with high potassium Tyrode’s solution (K 120 mmol litre1, see below) and once with phenylephrine 10 mol litre1, an ␣1 adrenoceptor agonist, at intervals of 15 min. Between these episodes the chambers were flushed at least three times with Tyrode’s solution. This episode was followed by an equilibrium period of 30 min for the L-NAME and control groups and 60 min for the glibenclamide group. L-NAME 100 mol litre1 or glibenclamide 1 mol litre1 was added at the beginning of this period and remained present in the bath thereafter. An equilibrium period of 60 min was chosen for glibenclamide as this drug has a slow onset of action. Subsequently, U46619, a thromboxane A2 mimetic agent, was administered to all preparations at a concentration of 1 mol litre1. After stabilization of the contraction, a cumulative concentration– response curve, using concentrations of 0.1 mol litre1 to 1 mmol litre1, was constructed for amrinone, milrinone, enoximinone, theophylline and dipyridamole.
123 Our experimental model included two separate force transducers connected to one single organ bath, which enabled simultaneous evaluation of the contractile responses of two vessels to one drug combination. Therefore, the experiments described were performed in duplicate if sufficient vessels could be harvested from the patient’s breast. The data of these duplicate experiments were averaged and listed in the results section for each individual patient. To evaluate the effects of the preservation technique on endothelial function, the response of contracted arteries to methacholine 105 mol litre1 was determined immediately after dissection of the blood vessel and after preservation in UW fluid. These vessels were not used in any other experiment. SOLUTIONS AND DRUGS USED
In all experiments, Tyrode’s solution of the following composition was used (mmol litre1): NaCl 136; NaHCO3 11.9; NaH2PO4 0.42; KCl 2.5; CaCl2 1.8; MgCl2 0.5; and glucose 5.5. In the high potassium Tyrode’s solution, NaCl 117.5 mmol litre1 was replaced by KCl, resulting in a total potassium concentration of 120 mmol litre1. These solutions were prepared just before each experiment. Aliquots of 20 ml were kept at 37 ⬚C and bubbled with carbogen before they were introduced into the bath. Amrinone and milrinone (donated by Sanofi Winthrop, Dev. Centre, Fawdon, UK) and enoximone (donated by Marion Merrell Dow Inc., Cincinatti, OH, USA) were dissolved to a concentration of 101 mol litre1. Milrinone was dissolved in distilled H2O, amrinone and enoximone were dissolved in 99 % dimethylsulphoxide (DMSO). Subsequently, these agents were diluted to 102–105 mol litre1 in distilled water (amrinone and milrinone) or 66.7 % DMSO (enoximone). Theophylline and dipyridamole (Sigma Chemical Co., USA) were dissolved in distilled water and 66.7 % DMSO, respectively, and subsequently diluted to 102–105 mol litre1 in distilled water. The final concentrations (1 %) of DMSO in the bathing medium had no significant effect on vascular smooth muscle contraction and relaxation. (See fig. 1 for the chemical structures of the various compounds studied.) Nitro-L-arginine methyl ester hydrochloride (LNAME) (Sigma), U46619 (9,11-dideoxy-11␣,9␣epoxy methanoprostaglandin F2␣) (Sigma) and Lphenylephrine hydrochloride (Sigma) were dissolved in distilled water. Glibenclamide (Hoechst) was dissolved in 66.7 % DMSO. DATA ANALYSIS
A log concentration–response curve was fitted to the data of individual experiments using GraphPad software (GraphPad, Institute for Scientific Information, San Diego, CA, USA). This computer program allows calculation of the EC50 by non-linear regression analysis, which is the concentration of the compound required to achieve the half maximal relaxant effect, based on the equation: E
124
British Journal of Anaesthesia
Figure 1
Chemical structures of the various compounds studied.
Emax AP/(AP EC50P ), where E the response obtained with a given concentration A; the exponent P describes the slope of the relationship (Hill coefficient); and Emax the maximally obtainable effect. All data are presented as mean (SEM). Differences between means were compared using a two-tailed Student’s t test for unpaired data or analysis of variance where appropriate (Graphpad Instat). P 0.05 was considered significant.
Results The resistance arteries dissected from the mammary tissues and used to evaluate the effects of the PDE III inhibitors (n 116) had an average diameter of 415 (20) m and a length of 1.8 (0.03) mm. There were no significant differences in main vessel characteristics between the groups (table 1). All arteries showed spontaneous rhythmic activity under the experimental conditions. The preparations remained stable and fully responsive to methacholine Table 1 Characteristics of the human vessels used in this study (mean (SEM)). L-NAME NG-nitro-L-arginine methyl ester Age (days)
Length (mm)
Circumference (µm)
Amrinone L-NAME Glibenclamide
3.7 (1.5) 3.8 (1.3) 2.1 (0.4)
1.8 (0.1) 2.0 (0.0) 1.9 (0.1)
482 (80) 338 (102) 419 (57)
Milrinone L-NAME Glibenclamide
4.0 (1.4) 3.8 (1.8) 2.5 (0.6)
1.8 (0.1) 1.9 (0.1) 2.0 (0.0)
403 (59) 385 (96) 383 (61)
Enoximone L-NAME Glibenclamide
3.1 (0.9) 4.0 (1.5) 2.1 (0.5)
1.9 (0.1) 2.0 (0.0) 1.8 (0.2)
434 (79) 479 (125) 336 (51)
Theophylline L-NAME Glibenclamide Dipyridamole L-NAME Glibenclamide
3.0 (0.9) 2.6 (1.2) 2.3 (1.0) 3.4 (1.6) 3.8 (1.5) 3.8 (1.5)
2.0 (0.0) 1.6 (0.2) 1.6 (0.2) 1.5 (0.1) 1.9 (0.1) 1.9 (0.1)
377 (58) 259 (50) 357 (65) 287 (43) 305 (68) 337 (57)
for at least 1 week after preparation when stored in the UW solution. Maximal relaxation to methacholine 105 mol litre1 was 97.5 (1.5) % (n 4) on day 0 vs 97.3 (1.8) % (n 4) on day 7 (ns). Preservation UW fluid had no influence on the contractile response to U46619 1 mol litre1, as linear regression analysis revealed no time-dependent change (correlation coefficient r 0.159, slope not significantly different from zero) (fig. 2). In addition, the contractile responses to phenylephrine remained stable with time: 3.92 (0.32) N m1 (n 22) on day 0 vs 5.05 (1.62) (n 4) (ns) on day 7. All PDE inhibitors induced concentrationdependent vasodilatation in the isolated arteries. Dipyridamole was the most potent vasodilator in our study (pEC50 4.67 (0.22) mol litre1) and theophylline the least active (pEC50 2.72 (0.07) mol litre1). The pEC50 values for amrinone, milrinone and enoximone were similar in the control groups: 3.28 (0.01), 3.49 (0.17) and 3.29 (0.08) mol litre1, respectively. Compared with potassium and phenylephrine, U46619 proved to be the most potent vasoconstrictor
Figure 2 Linear regression analysis of the contractile response to U46619 1 mol litre1 with respect to days of vessel preservation.
Vasorelaxation by PDE inhibitors
125
Table 2 Contractile response to U46619 1 mol litre1 (N m1) in small human isolated vessels (mean (SEM)); influence of NG-nitro-L-arginine methyl ester (L-NAME) and glibenclamide
Amrinone Milrinone Enoximone Theophylline Dipyridamole
Control
L-NAME
Glibenclamide
No. individuals studied Force
No. individuals studied Force
No. individuals studied
Force
6 7 6 5 5
5 5 5 5 5
6 6 6 8 5
3.68 (0.67) 5.64 (1.24) 4.66 (0.98) 4.20 (0.68) 3.95 (1.38)
5.44 (1.64) 6.06 (1.54) 5.65 (1.77) 5.71 (0.65) 4.03 (2.11)
4.76 (0.75) 4.49 (0.70) 6.84 (1.21) 4.07 (0.64) 4.31 (1.65)
Table 3 Contractile response (N m1) to the fourth potassium administration (K+) and to phenylephrine 10 mol litre1 (Phe) in small human isolated vessels (mean (SEM)). L-NAME N G-nitro-L-arginine methyl ester L-NAME
Control
Amrinone Milrinone Enoximone Theophylline Dipyridamole
Glibenclamide
K+
Phe
K+
Phe
K+
Phe
4.78 (1.55) 5.66 (1.34) 5.00 (1.67) 4.81 (0.69) 4.39 (0.43)
4.03 (1.46) 4.94 (1.31) 4.16 (1.49) 4.86 (0.50) 4.03 (0.38)
4.17 (0.71) 3.71 (0.60) 5.76 (0.92) 3.45 (0.68) 4.15 (1.33)
3.15 (0.58) 3.18 (0.51) 4.65 (0.65) 2.84 (0.67) 3.63 (1.24)
4.12 (0.94) 5.40 (1.11) 4.20 (0.87) 4.00 (0.66) 4.01 (1.01)
3.69 (0.82) 5.02 (1.03) 3.63 (0.75) 3.52 (0.58) 3.70 (0.90)
Table 4 pEC50 values of the phosphodiesterase inhibitors with respect to the vasodilatation induced by various inhibitors in precontracted small human arteries (mean (SEM)). pEC50 log of the concentration of the compound required to achieve the half maximal relaxation effect; L-NAME N G-nitro-L-arginine methyl ester. *P 0.05 compared with the respective control group
Amrinone Milrinone Enoximone Theophylline Dipyridamole
Control
L-NAME
Glibenclamide
No. individuals studied pEC50
No. individuals studied
pEC50
No. individuals studied pEC50
6 7 6 5 5
5 5 5 5 5
3.35 (0.09) 2.77 (0.24)* 3.41 (0.10) 2.69 (0.06) 4.76 (0.09)
6 6 6 8 5
3.28 (0.10) 3.49 (0.17) 3.29 (0.08) 2.72 (0.07) 4.67 (0.22)
in all experiments. The contractile response to U46619 was not influenced by the administration of L-NAME or glibenclamide. The basal values of the U46619-induced response were comparable in all preparations (table 2). Of the 85 experiments described in table 2, 31 experiments were performed in duplicate, accounting for the total of 116 experiments. Both the contractile response to U46619 and the vasodilator effects of amrinone, milrinone, enoximone, theophylline and dipyridamole were completely reversible on washout. The contractile responses to the fourth potassium administration and to phenylephrine are shown in table 3. Amrinone was administered in cumulative concentrations of 108 to 103 mol litre1. At the highest concentration, amrinone caused relaxation of U46619-induced contraction by 72.4 % of the initial value. Milrinone and enoximone were also administered in a concentration range of 108–103 mol litre1. The maximal relaxant effects were 71.4 % and 82.2 %, respectively. Theophylline proved less potent and effective, as higher doses were required (3.103 mol litre1) and a lower degree of maximal relaxation (57.8 %) was achieved. Dipyridamole
3.25 (0.17) 3.46 (0.15) 3.49 (0.06) 2.76 (0.19) 4.53 (0.02)
Figure 3 Log concentration-response curve for amrinone in the absence (control (n 6) ( )) or presence of L-NAME (n 5) ( ) or glibenclamide (n 6) ).
126
British Journal of Anaesthesia
Figure 4 Log concentration-response curve for enoximone in the absence (control (n 6) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 6) (●).
Figure 6 Log concentration-response curve for dipyridamole in the absence (control (n 5) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 5) (●).
Figure 5 Log concentration-response curve for theophylline in the absence (control (n 5) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 8) (●).
Figure 7 Log concentration-response curve for milrinone in the absence (control (n 7) (■)) or presence of L-NAME (n 5) ( ) or glibenclamide (n 6) (●). In the presence of L-NAME, a significant (P 0.05) rightward shift of the relaxation curve was observed.
concentrations ranged from 108 to 104 mol litre1, resulting in 83 % maximum relaxation of the basal value. Neither L-NAME nor glibenclamide influenced the basal value of the contractile response to U46619 when administered simultaneously with the PDE inhibitors. In addition, neither L-NAME nor glibenclamide influenced the relaxant responses to amrinone, enoximone, theophylline or dipyridamole (table 4 and figs 3–6). However, in the presence of LNAME a significant (P 0.05) rightward shift was observed in the relaxation curve of milrinone. Accordingly, the pEC50 values of milrinone were also influenced by L-NAME: 2.77 (0.24) mol litre1 with
compared with 3.49 (0.17) mol litre1 without L-NAME (P 0.05) (fig. 7). L-NAME
Discussion Using animal models, several investigators have demonstrated that the PDE III inhibitors amrinone, milrinone and enoximone induce vasodilatation as a result of inhibition of cGMP-inhibited (cGI) PDE, thus leading to increased intracellular levels of cAMP [11–14]. The effects of theophylline and dipyridamole on vascular smooth muscle are more complex. The mode of action of theophylline involves
Vasorelaxation by PDE inhibitors both non-specific phosphodiesterase inhibition and adenosine receptor antagonism [24]. In addition, unknown cellular mechanisms appear to play an additional role [25]. Dipyridamole is not only a PDE inhibitor, but also modulates adenosine metabolism, as it inhibits the cellular uptake of adenosine [26, 27]. In the present study, all agents caused concentration-dependent vasodilatation. The concentrations required were rather high compared with the results reported in the literature. However, these data may not be comparable with our results as various types of tissues were used in these studies and were obtained from different species. The use of human vessel preparations provides an obvious advantage when studying the action of agents with an influence on PDE activity, as the distribution of PDE isoenzymes is known to differ considerably between species [28]. Only Lindgren and colleagues used human vessels in their studies, but these tissues were obtained either post mortem or from patients undergoing laparotomy for various abdominal disorders, and not from healthy patients as in our study [11, 29]. Interestingly, these authors showed the importance of the mode of produced contraction on both the magnitude of the relaxant effect and the potency of PDE inhibitors on human vessels in vitro. Contractions of isolated arteries induced by high potassium concentrations (K 127 mmol litre1) were more difficult to relax by PDE III inhibitors compared with lower K concentrations (30 mmol litre1) or noradrenaline-induced contractions. Similarly, Harris and colleagues demonstrated a difference in efficacy of PDE III inhibitors in voltage-dependent (K 80 mmol litre1) as opposed to receptor-mediated (U46619) vasoconstriction [30]. The response to the PDE inhibitors in the present study may, therefore, be related partially to the method by which vasoconstriction was induced. Other important factors which may influence the efficacy of PDE inhibitors are the characteristics of the vascular beds from which the human vessels are taken. Marked variations in vascular responsiveness of isolated human coronary, renal and lung arteries have been demonstrated by Lindgren and Andersson [29]. In potassium (30 mmol litre1) contracted lung arteries, cGI-PDE inhibitors (including milrinone 3 105 mol litre1) were incapable of producing 50 % relaxation. In contrast, the agents investigated were significantly more potent and effective in renal and coronary arteries (pEC50 for milrinone: 3.2 106 and 1.7 106 mol litre1, respectively). This reduced efficacy in both lung arteries in the aforementioned investigation and mammary arteries in our study may indicate low cGI-PDE activities in these particular tissues. Although not all vessels were used immediately after dissection, but stored in UW organ preservation fluid, there were no differences between the vessel characteristics, including maximal force development and function of endothelium (response to methacholine), between vessels of several minutes to several days old. Therefore, it is unlikely that this preservation technique was responsible for these findings. Thus far there are no reports in the literature that document the preservation of isolated vessel preparations in UW solution. Successful
127 preservation in UW fluid for 72 h, however, has been demonstrated for entire organs taken from animals [31, 32]. In our study, only the vasodilator effects of milrinone were partially endothelium-dependent. The addition of L-NAME 100 mol litre1 resulted in a significant (P 0.05) increase in the pEC50 value of milrinone compared with controls: 2.77 (0.24) vs 3.49 (0.17) mol litre1, respectively. So far we cannot explain why the vasodilator profile of milrinone was distinctly different from that of its structurally related analogue, amrinone. Harris and colleagues also described differential relaxant effects of milrinone and amrinone in canine arteries [30]. In their study milrinone proved a more potent vasodilator against receptor-mediated (U46619) as opposed to K-induced depolarizations, whereas amrinone appeared equipotent in these conditions. The reason for these different effects could not be elucidated. Limited data are available in the literature on the interaction between the currently used PDE inhibitors and endothelium. Amrinone was reported to cause vasodilatation by both endothelium-dependent and -independent mechanisms in small pulmonary arteries of rabbits [16]. The vasodilator effects of enoximone in isolated porcine pulmonary arteries were endothelium-independent and unchanged by the addition of L-NAME 104 mol litre1 [33]. Finally, Kauffman and colleagues demonstrated that the vasodilator effects of milrinone in concentrations ranging from 107 to 105 mol litre1 occurred predominantly independent of the presence of an intact endothelium in aortic strips from rats [14]. In addition, these authors demonstrated that a high concentration of milrinone (104 mol litre1) resulted in increased concentrations of the second messenger cGMP next to enhanced levels of the intracellular mediator cAMP. Furthermore, in the presence of methylene blue, the dose–response curve for milrinone was shifted to the right. Methylene blue acts as an inhibitor of soluble guanyl cyclase and also nitric oxide synthase [34]. Clearly the mechanism by which PDE inhibitors induce vasodilatation is complex, as recent evidence has been obtained demonstrating that cross-interactions between cAMP and cGMP occur at different levels [35, 36]. At high concentrations milrinone acts on PDE IV and PDE V, next to PDE III inhibition [36], PDE IV hydrolyses cAMP, whereas PDE V acts as a specific cGMP hydrolyser. It has been demonstrated that vasodilatation induced by selective PDE IV inhibitors were abolished by both L-NAME and methylene blue, indicating that the effect of PDE IV inhibitors is dependent on nitric oxide production and subsequent activation of soluble guanylate cyclase [36]. Thus, the aforementioned results combined with our data may indicate that milrinone acts as a PDE III inhibitor, independent of endothelial function, at low concentrations, whereas at higher concentrations (105 mol litre1) the inhibitory activity of milrinone on PDE IV, which can be partially abolished by L-NAME becomes more prominent. The interaction of glibenclamide with milrinone
128 was reported by Fujimoto [37]. In his experiments, inhibition of milrinone-induced vasodilatation by glibenclamide in vascular preparations of rats and guineapigs, was demonstrable only at glibenclamide concentrations as high as 10–30 mol litre1 [37]. However, at these high concentrations glibenclamide may not act as a specific KATP channel blocker [38]. In the present investigation glibenclamide 1 mol litre1 did not influence the vasodilator effect of any of the PDE inhibitors studied, indicating that KATP channels are not involved in the process of vasodilatation caused by these agents. In conclusion, L-NAME sensitive nitric oxide release may contribute to the vasodilator properties of milrinone, but not to those of any of the other PDE inhibitors studied. Vasodilatation caused by PDE inhibitors does not appear to involve KATP channels.
Acknowledgements We acknowledge the support offered by Dr E. J. F. Timmenga and Dr C. M. A. M. van der Horst, plastic and reconstructive surgeons, in the harvesting of human vascular tissue.
References 1. Butcher RW, Sutherland EW. Adenosine 3⬘,5⬘-phosphate in biological materials. Journal of Biological Chemistry 1962: 237; 1244–1250. 2. Beavo JA. Multiple isoenzymes of cyclic nucleotide phosphodiesterase. Advances in Second Messenger and Phosphoprotein Research 1988; 22: 1–37. 3. Weishaar RE, Burrows SD, Kobylarz DC, Quade MM, Evans DB. Multiple molecular forms of cyclic nucleotide phosphodiesterase in cardiac and smooth muscle and in platelets. Isolation, characterization and effects of various reference phosphodiesterase inhibitors and cardiotonic agents. Biochemical Pharmacology 1986; 5: 787–800. 4. Thompson WJ. Cyclic nucleotide phosphodiesterases: Pharmacology, biochemistry and function. Pharmacology and Therapeutics 1991; 51: 13–33. 5. Nicholson CD, Challis RAJ, Shahid M. Differential modulation of tissue function and therapeutic potential of selective inhibitors of cyclic nucleotide phosphodiesterase isoenzymes. Trends in Pharmacological Sciences 1991; 12: 19–27. 6. Packer M, Carver JR, Rodeheffer RJ, Ivanoe RJ, DiBianco R, Zeldis SM, Hendrix GH, Bommer WJ, Elkayam U, Kukin ML, Mallis GI, Sollano JA, Shannon J, Tandon PK, De Mets DL. Effect of oral milrinone on mortality in severe chronic heart failure: Promise study. New England Journal of Medicine 1991; 325: 1468–1475. 7. DiBianco R, Shabetia R, Kostuk W, Moran J, Schlant RC, Wright R. A comparison of oral milrinone, digoxin and their combination in the treatment of chronic heart failure. New England Journal of Medicine 1989; 320: 677–683. 8. Uretsky BF, Jessup M, Konstam MA, Dec W, Leier CV, Benotti J, Murali S, Herrmann HC, Sandberg JA. For the enoximone multicenter trial group. Multicenter trial of oral enoximone in patients with moderate to moderately severe congestive heart failure. Circulation 1990; 82: 774–780. 9. Dupuis JY, Bondy R, Cattran C, Nathan HJ, Wynands JE. Amrinone and dobutamine as primary treatment of low cardiac output syndrome following coronary artery surgery: A comparison of their effects on hemodynamics and outcome. Journal of Cardiothoracic and Vascular Anesthesia 1992; 6: 542–553. 10. Caldicott LD, Hawley K, Heppel R, Woodmansey PA, Channer KS. Intravenous enoximone or dobutamine for severe heart failure after acute myocardial infarction: a randomized double blind trial. European Heart Journal 1993; 14: 696–700.
British Journal of Anaesthesia 11. Lindgren S, Andersson KE, Belfrage P, Degerman E, Manganiello VC. Relaxant effects of the selective phosphodiesterase inhibitors milrinone and OPC 3911 on isolated human mesenteric vessels. Pharmacology and Toxicology 1989; 64: 440–445. 12. Silver PJ, Lepore RE, O’Connor B, Lemp BM, Hamel LT, Bentley RG, Harris, AL. Inhibition of the low Km cyclic AMP system in vascular smooth muscle by milrinone. Journal of Pharmacology and Experimental Therapeutics 1988; 247: 34–42. 13. Itoh H, Kusagawa M, Shimomura A, Suga T, Ito M, Konishi T, Nakano T. Ca2 dependent and Ca2 independent vasorelaxation induced by cardiotonic phosphodiesterase inhibitors. European Journal of Pharmacology 1993; 240: 57–66. 14. Kauffman RF, Schenck KW, Utterbach BG, Crowe G, Cohen ML. In vitro vascular relaxation by new inotropic agents: Relationship to phosphodiesterase inhibition and cardiac nucleotides. Journal of Pharmacology and Experimental Therapeutics 1987; 242: 864–872. 15. Weishaar RE, Burrows SD, Kobylarz DC, Quade MD, Evans DB. Multiple molecular forms of cyclic nucleotide phosphodiesterase in cardiac and smooth muscle and in platelets: Isolation, characterization and effects of various reference phosphodiesterase inhibitors and cardiotonic agents. Biochemical Pharmacology 1986; 35: 787–800. 16. Clarke WR, Soltow LOG. Amrinone causes vasodilation by both endothelium-independent and endothelium dependent mechanisms in small pulmonary arteries. American Review of Respiratory Diseases 1991; (Suppl A773). 17. Rosenblum WI, Shimizu T, Nelson GH. Interaction of endothelium with dilation produced by inhibitors of cyclic nuclide diesterases in mouse brain arterioles in vivo. Stroke 1993; 24: 266–270. 18. Martin W, Furchgott RF, Villani GM, Jothianandan D. Phosphodiesterase inhibitors induce endothelium-dependent relaxation of rat and rabbit aorta by potentiating the effects of spontaneously released endothelium-derived relaxing factor. Journal of Pharmacology and Experimental Therapeutics 1986; 237: 539–547. 19. Grover GJ, Sleph PG, Dzwonczyk S. Role of myocardial ATP-sensitive potassium channels in mediating preconditioning in the dog heart and their possible interaction with adenosine A1 receptors. Circulation 1992; 86: 1310–1316. 20. Gross GJ, Auchampach JA. Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circulation Research 1992; 70: 223–233. 21. Southard JH, Van Gulik TM, Ametani MS, Vreugdenhil PK, Lindell SL, Pinaar BL, Belzer FO. Important components of the UW solution. Transplantation 1990; 49: 251–257. 22. Mulvany MJ, Halpern W. Mechanical properties of vascular smooth muscle cells in situ. Nature (London) 1976; 260: 617–618. 23. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circulation Research 1977; 41: 10–26. 24. Fredholm BB, Persson CGA. Xanthine derivatives as adenosine receptor antagonists. European Journal of Pharmacology 1982; 81: 673–676. 25. Persson CGA, Andersson KE, Kjellin G. Effects of enprofylline and theophylline may show the role of adenosine. Life Sciences 1986; 38: 1057–1072. 26. Kubler W, Spieckermann PA, Bretschneider HJ. Influence of dipyridamole (Persantin) on myocardial adenosine metabolism. Journal of Molecular and Cellular Cardiology 1970; 1: 23–27. 27. Klabunde RE. Effects of dipyridamol on postischemic vasodilation and extracellular adenosine. American Journal of Physiology 1983; 244: H273–H280. 28. Weishaar RE, Kobylarz-Singer DC, Steffen RP, Kaplan HR. Subclasses of cyclic AMP, specific phosphodiesterase in left ventricular muscle and their involvement in myocardial contractility. Circulation Research 1987; 61: 539–547. 29. Lindgren S, Andersson KE. Effects of selective phosphodiesterase inhibitors on isolated coronary, lung and renal arteries from man and rat. Acta Physiologica Scandinavica 1991; 142: 77–82.
Vasorelaxation by PDE inhibitors 30. Harris AL, Grant AM, Silver PJ, Evans DB, Alousi AA. Differential vasorelaxant effects of milrinone and amrinone on contractile responses of canine coronary, cerebral, and renal arteries. Journal of Cardiovascular Pharmacology 1989; 13: 238–244. 31. Ploeg RJ, Goossens D, McAnulty JF, Stouthard JH, Belzer FO. Successful 72 hour cold storage of dog kidneys with UW solution. Transplantation 1988; 46: 191–196. 32. Wahlberg JA, Love R, Landegaard L, Stouthard JH, Belzer FO. 72 hours preservation of the canine pancreas. Transplantation 1987; 43: 5–8. 33. Butt AY, Dinh-Xuan AT, Pepke-Zaba J, Cremona G, Clelland CA, Higenbottam TW. In vitro pulmonary vasorelaxant effect of the phosphodiesterase inhibitor enoximone. Angiology 1993; 12: 289–294.
129 34. Mayer B, Brunner F, Schmidt K. Novel actions of methylene blue. European Heart Journal 1993; 14: (Suppl. I) 22–26. 35. Jiang J, Colbran JL, Francis SH, Corbin JD. Direct evidence for cross-activation of cGMP-dependent protein kinase by cAMP in pig coronary arteries. Journal of Biochemical Chemistry 1992; 267: 1015–1019. 36. Lugnier C, Komas N. Modulation of vascular cyclic nucleotide phosphodiesterases by cyclic GMP : role in vasodilatation. European Heart Journal 1993; 14 (Suppl. I): 141–148. 37. Fujimoto S. Effects of pimobendan, its active metabolite UDCG 212 and milrinone on isolated blood vessels. European Journal of Pharmacology 1994; 265: 159–166. 38. Ashcroft FM. Adenosine 5⬘-triphosphate-sensitive potassium channels. Annual Review of Neuroscience 1988; 11: 97.