Characterization of Acetylcholine-Induced Membrane Hyperpolarization in Endothelial Cells Guifa Chen and Donald W. Cheung
The characteristics of the hyperpolarization response to acetylcholine (ACh) in endothelial cells from the guinea pig coronary artery were studied by microelectrode recording technique. ACh (30 nM to 3 ,uM) induced membrane hyperpolarization in a dose-dependent manner. The sustenance of the response required the presence of external calcium. The hyperpolarization was not affected by nifedipine (1 ,uM) but was inhibited by the potassium channel blockers charybdotoxin (10 nM), tetraethylammonium (1 mM), and 4-aminopyridine (0.5 mM). Glibenclamide (10 ,M) and apamin (1 ,uM) were not effective. The inhibitors of endotheliumderived relaxing factor/nitric oxide synthesis Nw-nitro L-arginine (50 ,M) and NG-monomethyl L-arginine (30 ,uM) had no effect on the resting membrane potential or the ACh-induced responses. No hyperpolarization was observed with application of sodium nitroprusside (10 ixM) or 8-bromo-cGMP (0.1 ,uM). Ouabain (10 tM) depolarized the membrane significantly by 5 mV, but the ACh hyperpolarization was not affected. Indomethacin (10 JAM) was without effect on the resting membrane potential or the hyperpolarization to ACh. These results show that ACh-induced hyperpolarization is dependent on external calcium and can be inhibited by certain potassium channel blockers. The hyperpolarization response is not mediated by endothelium-derived relaxing factor/nitric oxide, cGMP, a cyclooxygenase product, or stimulation of the Na-K pump. (Circulation Research 1992;70:257-263)
T he relaxation response to acetylcholine (ACh) in vascular tissues is dependent on the presence of an endothelium. In many cases, the relaxation can be attributed to the release of an endothelium-derived relaxing factor (EDRF).' It is now generally acknowledged that nitric oxide (NO) or a nitrosothiol compound is the EDRF and L-arginine is the precursor.2 The heme moiety of soluble guanylate cyclase is the NO receptor, and vasorelaxation is associated with an increase in the cGMP level.2 Another endothelium-dependent response to ACh in vascular smooth muscle cells is membrane hyperpolarization.3 The cellular mechanism mediating endothelium-dependent hyperpolarization is distinct from that of EDRF/NO.3 For example, vasorelaxation and production of cGMP induced by EDRF/NO are inhibited by methylene blue and hemoglobin.2 However, the hyperpolarization in reFrom the University of Ottawa Heart Institute, Canada. Supported by a grant from the Heart & Stroke Foundation of Ontario. G.C. is a Research Fellow of the Heart & Stroke Foundation of Canada. Address for correspondence: Donald W. Cheung, University of Ottawa Heart Institute, 40 Ruskin Street, Ottawa K1Y 4E9, Canada. Received April 29, 1991; accepted October 4, 1991.
sponse to ACh is not affected by these same agents.4 The release of a separate endothelium-derived hyperpolarizing factor has been proposed.3,5 In aortic endothelial cells, ACh also induces membrane hyperpolarization by activating calciumdependent potassium channels.6,7 The physiological role of the endothelial hyperpolarization response remains to be clarified. However, it is known that similar electrical activities have important regulatory functions in secretory cells.8 The endothelial cells are also coupled to vascular smooth muscle cells via gap junctions.9"10 The membrane hyperpolarization response in the endothelium could therefore influence the activity of vascular smooth muscle cells directly. The electrical response of the endothelial cells to ACh has not been fully described. In the present study, we examined the characteristics of the ACh-induced hyperpolarization by an intracellular recording technique. Materials and Methods The circumflex coronary artery was dissected out from male guinea pigs weighing 300-400 g (Charles River Laboratories, Inc., Wilmington, Mass.). The artery was cut open, and the intimal layer was gently teased away from the rest of the artery with a pair of
Downloaded from http://circres.ahajournals.org/ at AUSTRALIA NATL UNIV on March 14, 2015
Circulation Research Vol 70, No 2 February 1992
fine forceps. The intima was pinned to the bottom of an organ bath with the lumen facing up. The preparation was constantly perfused with physiological solution bubbled with 95% 02-5% CO2 and containing (mM) NaCl 120, NaHCO3 25, glucose 11, KCl 5, CaCl2 2.5, NaH2P04 1, and MgSO4 1, at a rate of 2 ml/min. In nominally calcium-free solutions, CaCi2 was omitted. Glass micropipettes filled with 3 M KCl and with tip resistances of 40-50 MQl were used for intracellular recordings. Membrane potentials were measured using an intracellular preamplifier (model 8700, Dagan Corp., Minneapolis, Minn.). After penetration into the cell, the membrane potential remained stable for up to 4 hours of continuous recording, and the zero baseline potential was not changed on sudden withdrawal of the electrode at the end of the experiment. The endothelial cells were impaled from the luminal surface. The dose-response relation between membrane hyperpolarization and the concentration of ACh was studied by cumulative addition of the agent after a maximal response to each dose was reached. Control studies had shown that the level of membrane hyperpolarization to ACh was the same whether ACh was added as a single dose or cumulatively. There was no desensitization to ACh. After a control response to different concentrations of ACh was obtained, the preparations were washed in physiological solutions for 25 minutes. Drugs in appropriate concentrations were added to the perfusing solution, and recordings were made in the continuous presence of these drugs after an incubation period of 10 minutes. In some experiments with the L-arginine analogues, longer incubation periods of up to 25 minutes were also attempted with no noticeable change in the results. In all experiments, the effects of drugs as compared with control were studied by continuous recordings from the same cell. Stable and reproducible responses could be routinely recorded from a single cell for 2-3 hours. All experiments were performed at
360C. The following chemicals were used in the study:
NG-monomethyl L-arginine (L-NMMA, Calbiochem Corp., La Jolla, Calif.); charybdotoxin (CTX, Natural Product Sciences, Inc., Salt Lake City, Utah); ACh, ouabain, sodium nitroprusside, tetraethylammonium chloride (TEA), glibenclamide, 8-bromo-cGMP, indomethacin, nifedipine, 4-aminopyridine (4-AP), apamin, dimethyl sulfoxide (DMSO), and N'-nitro L-arginine (L-NARG) (all from Sigma Chemical Co., St. Louis, Mo.). Glibenclamide was prepared as 10-mM stock solutions in DMSO. The final diluted DMSO concentration in the perfusing medium was 1%. The same concentration of DMSO was present in the control, and the responses to ACh were not different from those recorded in normal physiological solutions alone. Stock solution of nifedipine (10 mM) was prepared in DMSO, and the final diluted concentration of DMSO was 0.01%. The remaining drugs were dissolved in distilled water.
All data were presented as mean+ SEM; n denotes the number of preparations. Analysis of variance was performed to test the statistical significance in the change in response to various treatments. A value of psO.05 was the specified level of significance. Results Calcium Dependence of ACh Response The endothelial cells were electrically quiescent and had an average resting membrane potential of -44.7±0.24 mV (n=39 preparations). ACh (30 nM to 3 gM) induced prolonged membrane hyperpolarization, and in the continual presence of ACh, the membrane repolarized slowly with time (Figure 1A). The prolonged hyperpolarization was dependent on the presence of external calcium. In nominally calcium-free solutions, ACh elicited only a transient hyperpolarization of decreased magnitude (Figure 1B). The transient hyperpolarization diminished in amplitude with each subsequent dose of ACh until it was totally abolished by the seventh to eighth application (Figures 1C and 1D). In calcium-free solutions, a small sustained depolarization response (2.3±0.2 mV, n=6) was unmasked after the initial transient hyperpolarization (Figures 1C and 1D). The resting membrane potential of the endothelial cells was significantly depolarized in calcium-free solution from -45.0±0.4 mV to -33.5±0.5 mV (n=7). Washing out of the ACh in calcium-free solution with normal calcium solution induced a large transient hyperpolarization followed by repolarization of the resting membrane potential (Figure 1D). Membrane hyperpolarization to ACh is dose dependent. A small response could usually be detected at -30 nM, reaching to a maximum hyperpolarization of 32.5±0.7 mV (n=39 preparations) at 3 ,uM ACh (Figure 2). In calcium-free solution, the hyperpolarization response to cumulative doses of ACh was virtually abolished (Figure 2A). The hyperpolarization response to ACh was not related to calcium influx through classical voltage-dependent calcium channels. Nifedipine (1 gM) had no significant effects on the resting membrane potential (control, -44.0±0.24 mV; nifedipine, -43.7± 1.8 mV; n=5) or the ACh response (Figure 2B).
Effects of Potassium Channel Blockers The effects of the potassium channel blockers CTX, apamin, 4-AP, glibenclamide, and TEA on the hyperpolarization response to ACh were examined. CTX, known to block the large- and medium-conductance calcium-activated potassium channels," caused a dose-dependent depolarization of the resting membrane potential (control, -42.8±0.7 mV; 10 nM CTX, -39.3±1.8 mV [p=NS]; 50 nM CTX, -37.2±2.0 mV [p25 minutes), in sharp contrast to the transient response to bradykinin (