The authors recently demonstrated that acetylcholine-induced pulmonary vasorelaxation had two primary components, nitric oxide (NO) and endothelium-derived hyperpolarizing factor (EDHF). The goal was to investigate the effects of etomidate and ketamine on the NO- and EDHF-mediated components of pulmonary vasorelaxation in response to acetylcholine, bradykinin, and the calcium ionophore, A23187.
Canine pulmonary arterial rings with an intact endothelium were suspended in organ chambers for isometric tension recording. The effects of etomidate and ketamine (10(-5) M and 10(-4) M) on vasorelaxation responses to acetylcholine, bradykinin, and A23187 were assessed in phenylephrine-contracted rings. The NO- and EDHF-mediated components of relaxation were assessed using a NO synthase inhibitor (N-nitro-L-arginine methylester [L-NAME]: 10(-4) M) and a Ca2+-activated potassium channel inhibitor (tetrabutylammonium hydrogen sulfate [TBA]: 10(-3) M) in rings pretreated with a cyclooxygenase inhibitor (ibuprofen: 10(-5) M). Intracellular calcium concentration ([Ca2+]i) was measured in cultured bovine pulmonary artery endothelial cells loaded with acetoxylmethyl ester of fura-2.
Etomidate and ketamine attenuated pulmonary vasorelaxation in response to acetylcholine and bradykinin, whereas they had no effect on the response to A23187. The relaxant responses to acetylcholine and bradykinin were attenuated by L-NAME or TBA alone and were abolished by combined inhibition in rings pretreated with ibuprofen. Etomidate and ketamine further attenuated both L-NAME-resistant and TBA-resistant relaxation. These anesthetics also inhibited increases in endothelial [Ca2+]i in response to bradykinin, but not A23187.
These results indicate that etomidate and ketamine attenuated vasorelaxant responses to acetylcholine and bradykinin by inhibiting both NO- and EDHF-mediated components. Moreover, our results suggest that these anesthetics do not directly suppress NO or EDHF activity, but rather inhibit the endothelial [Ca2+]i transient in response to receptor activation.
THE endothelium plays a crucial role in regulating the tone of vascular smooth muscle via the release of vasoactive mediators in response to various neurohumoral agonists, including acetylcholine and bradykinin. Nitric oxide (NO), synthesized by the l-arginine–NO synthase pathway, and prostacyclin, produced by the arachidonic acid–cyclooxygenase pathway, have been identified as endothelium-derived mediators. 1,2In addition, the endothelium can release endothelium-derived hyperpolarizing factors (EDHFs), 3–5although their identity has not been firmly established. The relative contribution of NO and EDHFs to total endothelium-dependent relaxation depends on the endothelial stimulus and the vascular bed investigated. 6,7
We recently demonstrated in isolated canine pulmonary artery that acetylcholine-induced relaxation has two primary components, NO and EDHF, and that the intravenous anesthetic, propofol, selectively attenuates acetylcholine-induced relaxation by inhibiting both of these endothelium-derived mediators. 8Etomidate and ketamine are intravenous anesthetics that are recognized to have minimal inhibitory effects on clinically measured end points of cardiovascular function. As a result, they are often used as induction agents in patients with hemodynamic instability. Although both etomidate and ketamine are thought to inhibit endothelium-dependent vasodilation in the systemic circulation, little is known about their effects or mechanism of action on endothelium-dependent relaxation in the pulmonary circulation.
The first goal of this in vitro study was to investigate the effects of etomidate and ketamine on the NO- and EDHF-mediated components of pulmonary vasorelaxation in response to receptor-dependent and -independent endothelial activators. A second goal was to test the hypothesis that these intravenous anesthetics attenuate endothelium-dependent vasorelaxation by reducing increases in intracellular Ca2+concentration [Ca2+]iin endothelial cells in response to receptor activation.
Materials and Methods
All experimental procedures and protocols were approved by the Institutional Animal Care and Use Committee at The Cleveland Clinic Foundation (Cleveland, OH).
Preparation of Pulmonary Arterial Rings
Twenty healthy male mongrel dogs weighing 24–32 kg were anesthetized with intravenous pentobarbital sodium (30 mg/kg) and fentanyl citrate (15 μg/kg). After tracheal intubation, the lungs were mechanically ventilated. A catheter was placed in the right femoral artery, and the dogs were exsanguinated by controlled hemorrhage. A left lateral thoracotomy was performed through the fifth intercostal space, and the heart was arrested with induced ventricular fibrillation (Grass Instrument Co., Quincy, MA). The heart and lungs were removed from the thorax en bloc , and the lower right and left lung lobes were dissected free. Intralobar pulmonary arteries (2–4-mm ID) were carefully dissected and immersed in cold modified Krebs-Ringer bicarbonate solution composed of 118.3 mm NaCl, 4.7 mm KCl, 1.2 mm MgSO4, 1.2 mm KH2PO4, 2.5 mm CaCl2, 25 mm NaHCO3, 0.016 mm Ca-EDTA, and 11.1 mm glucose. The arteries were cleaned of connective tissue and cut into ring segments 4–5 mm in length with special care taken not to damage the endothelium. In some rings, the endothelium was denuded by gently rubbing the intimal surface with a cotton swab. The integrity of the endothelium was verified by assessing the vasorelaxant response to acetylcholine (10−6m).
Isometric Tension Experiments
Pulmonary arterial rings were vertically mounted between two stainless steel hooks in organ baths filled with 25 ml Krebs-Ringer bicarbonate solution (37°C) gassed with 95% air and 5% CO2. One of the hooks was anchored and the other was connected to a strain gauge to measure isometric force. The rings were stretched at 10-min intervals in increments of 0.5 g to achieve optimal resting tension. Optimal resting tension was defined as the minimum amount of stretch necessary to achieve the largest contractile response to 40 mm KCl and was determined in preliminary experiments to be 5 g for the size of arteries used in these experiments. After the arterial rings had been stretched to their optimal resting tension, the contractile response to 60 mm KCl was measured. After washing out KCl from the organ bath and the return of isometric tension to prestimulation values, a cumulative concentration–response curve to phenylephrine was performed in each ring. This was achieved by increasing the concentration of phenylephrine in half-log increments (from 10−8m to 3 × 10−5m) after the response to each preceding concentration had reached a steady state. All rings were pretreated with the β-adrenoreceptor antagonist, propranolol (5 × 10−6m, incubated for 30 min), before phenylephrine administration in all protocols to avoid the β-agonist effect of phenylephrine.
After washout of phenylephrine from the organ bath and return of isometric tension to baseline values, the rings were again pretreated with propranolol and precontracted to 50% of the maximal response to phenylephrine (ED50level of tension). When the contractile response was stabilized, concentration–response curves to the endothelial cell activators, acetylcholine (10−9to 10−5m), bradykinin (10−10to 3 × 10−7m) and A23187 (10−9to 3 × 10−6m) were obtained. Each ring was exposed to only one endothelial activator. Responses to the endothelium-independent NO donor, S -nitroso-N -acetylpenicillamine (SNAP: 10−9to 3 × 10−6m), were also investigated in endothelium-denuded rings.
The effects of etomidate (10−5and 10−4m) and ketamine (10−4and 10−5m) on the concentration–response curves for these vasodilators were assessed by comparing vasorelaxant responses in the presence and absence of the anesthetics. The anesthetics were directly added to the organ bath 15 min before phenylephrine contraction.
To characterize the relative contribution of NO- and EDHF-mediated components to the total relaxation responses to these endothelial activators, arterial rings were pretreated with the combination of ibuprofen (10−5m), a cyclooxygenase inhibitor, and N -nitro-l-arginine methylester (l-NAME: 10−4m), an inhibitor of NO synthase, or tetrabutylammonium hydrogen sulfate (TBA: 10−3m), an inhibitor of calcium-activated potassium channels (KCa). Previous studies have shown that EDHF-mediated relaxation can be inhibited by either membrane depolarization (30 mm KCl) or by inhibitors of KCa. 4,5After incubation with these inhibitors for 20 min, the rings were contracted to the ED50level of tension with phenylephrine. The inhibitors remained in the organ bath throughout the experiment. The relaxant responses to the endothelial activators in inhibitor-treated rings were compared with responses in untreated rings. The effects of etomidate (10−4m) and ketamine (10−4m) on the NO-mediated component of the relaxant response to acetylcholine and bradykinin were examined in rings pretreated with the combination of ibuprofen and TBA. The effects of these anesthetics on the EDHF-mediated component of the relaxant response to acetylcholine and bradykinin were examined in rings pretreated with the combination of ibuprofen and l-NAME.
Cell Culture of Pulmonary Artery Endothelial Cells
Bovine pulmonary artery endothelial cells were purchased from the American Type Culture Collection (CRL-1733, Rockville, MD). The cells were cultured in Ham F-12K medium (American Type Culture Collection) supplemented with 1.5% NaHCO3, 10% horse serum, and 1% antibiotic–antimycotic mixture solution (10,000 units/ml penicillin, 10,000 μg/ml streptomycin, 25 μg/ml amphotericin B) in a humidified atmosphere of 5% CO2and 95% air at 37°C. Cells were allowed to grow until they formed a confluent monolayer. The cells were then subcultured nonenzymatically to 35-mm glass dishes designed for fluorescence microscopy (Bioptechs, Inc., Butler, PA). Twenty-four hours before the experiment, the culture medium was replaced with a serum-free medium to arrest cell growth and allow for establishment of steady state cellular events independent of cell division.
Bovine pulmonary artery endothelial cells were washed twice in Krebs-Ringer-HEPES buffer, which contained 125 mm NaCl, 5 mm KCl, 1.2 mm MgSO4, 1.0 mm NaH2PO4, 2.5 mm CaCl2, 11 mm glucose, and 25 mm HEPES at pH 7.4 adjusted with NaOH. The cells were then incubated in Krebs-Ringer-HEPES containing 2 μm acetoxylmethyl ester of fura 2 (fura-2—AM; TEF Labs, Austin, TX) at room temperature for 40 min. After the 40-min loading period, the cells were washed twice in Krebs-Ringer-HEPES buffer. Culture dishes containing fura-2–loaded endothelial cells were placed in a temperature-regulated (37°C) chamber (Bioptechs, Inc.) mounted on the stage of an Olympus IX-70 inverted fluorescence microscope (Lake Success, NY). The volume of the culture dish was 1.5 ml, and temperature was maintained at 37°C. Fluorescence measurements were performed using a dual-wavelength spectrofluorometer (Deltascan RFK6002, Photon Technology International, South Brunswick, NJ) at excitation wavelength of 340 and 380 nm and an emission wavelengths of 510 nm. The 340 to 380 fluorescence ratio was used as an indicator of [Ca2+]i. Solution changes were accomplished by rapidly aspirating the buffer in the dish and adding 1.5 ml of the new buffer. Before data acquisition, background fluorescence was measured and subtracted automatically from the subsequent experimental measurements. Fura-2 fluorescence signals (340, 380, and 340/380 ratio) originating from bovine pulmonary artery endothelial cells were continuously monitored at a sampling frequency of 5 Hz and were collected using a software package from Photon Technology International.
Drugs and Solutions
All drugs were of the highest purity commercially available: acetylcholine chloride, A23187, bradykinin, l-NAME, phenylephrine HCl, propranolol HCl, SNAP, TBA (Sigma Chemical, St. Louis, MO), etomidate (Bedford Laboratories, Bedford, OH), ketamine HCl (Parke-Davis, Morris Plains, NJ), ibuprofen (Cayman Chemical, Ann Arbor, MI). All drug concentrations are expressed as the final concentration in the organ bath or cell dish. A23187 and ibuprofen were dissolved in dimethylsulfoxide and diluted with distilled water. The final concentration of dimethylsulfoxide in the organ bath or culture dish was less than 0.1 % (vol/vol). The vehicles have no effect on isometric tension or the 340/380 ratio at the concentrations used in these studies.
All data are expressed as mean ± SEM. Vasorelaxant responses to the endothelium-dependent and -independent vasodilators are expressed as the percentage relaxation of the precontraction induced by phenylephrine. The logarithm of the drug concentration eliciting 50% of the maximum relaxation response (IC50) was calculated by linear regression analysis, and the maximum relaxant response (Rmax) was determined. Rmax= 100% indicates complete reversal of phenylephrine contraction. Statistical analysis was performed using the Student t test for unpaired comparisons. When more than two means were compared, the Scheffé F test after one-way analysis of variance was used. A P value < 0.05 was considered significant. Sample size (n values) equals the number of dogs from which pulmonary arterial rings were taken in the isometric tension experiments, and the number of experiments in the endothelial cell studies.
Effects of Etomidate and Ketamine on Pulmonary Vasorelaxation
Neither etomidate nor ketamine altered resting tension. Acetylcholine elicited endothelium-dependent relaxation of pulmonary arterial rings that was maximal at 10−5m (fig. 1). Etomidate and ketamine had little effect on acetylcholine-induced relaxation at a concentration of 10−5m, whereas a higher concentration of each anesthetic (10−4m) increased the IC50without changing the Rmaxfor acetylcholine (table 1and fig. 1). The vehicle for etomidate (35% propylene glycol) had no effect on the relaxant response to acetylcholine (table 1).
Bradykinin also induced receptor-mediated endothelium-dependent relaxation in pulmonary arterial rings precontracted with phenylephrine (figs. 2A and 2B). Etomidate and ketamine (10−4m) caused rightward shifts (P < 0.05) in the bradykinin concentration–effect curve without affecting the Rmaxvalues (table 1and figs. 2A and 2B). In contrast to these receptor-operated agonists, the vasorelaxant response to the non–receptor-mediated endothelium-dependent vasodilator, A23187, was not altered by etomidate or ketamine (table 1and figs. 2C and 2D). Neither etomidate nor ketamine (10−4m) altered SNAP-induced relaxation in endothelium-denuded rings. The log IC50and Rmaxvalues in control, etomidate-treated, and ketamine-treated rings were −7.08 ± 0.12 and 100 ± 3%, −7.02 ± 0.15 and 105 ± 3%, and −7.06 ± 0.10 and 103 ± 3%, respectively.
Characterization of Relaxant Responses to Acetylcholine, Bradykinin, and A23187
The effects of the various inhibitors on the relaxant responses to each endothelial activator are summarized in table 2and figure 3. None of the inhibitors had an effect on baseline tension. The combination of l-NAME plus ibuprofen, as well as the combination of TBA plus ibuprofen, attenuated (P < 0.01) the pulmonary vasorelaxant responses to acetylcholine, bradykinin, and A23187. Pretreatment with either TBA or 30 mm KCI (to depolarize pulmonary arterial smooth muscle) essentially abolished (P < 0.01) the l-NAME plus ibuprofen–resistant relaxation responses to these endothelial activators. These results indicate that in ibuprofen-pretreated rings, the pulmonary vasorelaxant responses to acetylcholine, bradykinin, and A23187 are mediated by both NO and an EDHF.
Effects of Anesthetics on Nitric Oxide– and Endothelium-derived Hyperpolarizing Factor–mediated Relaxation
The effects of the two intravenous anesthetics (10−4m) on each component of the relaxation response to acetylcholine and bradykinin were investigated. In rings treated with l-NAME plus ibuprofen, etomidate (P < 0.05) and ketamine (P < 0.01) induced a rightward shift of the concentration–effect curve for acetylcholine without changing the Rmaxvalues (table 3and figs. 4A and 4B), i.e. , both anesthetics inhibited the EDHF-mediated component of the vasorelaxant response to acetylcholine. These anesthetics also attenuated (P < 0.05) the TBA plus ibuprofen-resistant acetylcholine relaxation (table 3and figs. 4C and 4D), i.e. , both anesthetics inhibited the NO-mediated component of the vasorelaxant response to acetylcholine. Etomidate and ketamine also inhibited (P < 0.05) the NO- and EDHF-mediated components of bradykinin-induced relaxation (table 3and fig. 5).
Effects of Etomidate and Ketamine on Endothelial Intracellular Ca2+Concentration in Response to Bradykinin and A23187
Bradykinin induced a rapid increase in endothelial [Ca2+]ifollowed by a gradual decline to a sustained plateau phase above the resting level (fig. 6A, left side). After washing out with buffer for 15 min, a second bradykinin application produced a similar but slightly smaller response in the same cell (fig. 6A, right side). To account for this slight rundown effect, changes in the 340/380 ratio in response to the second exposure to bradykinin in the presence or absence (control) of anesthetics are expressed as a percentage of the first response to bradykinin. Pretreatment with ketamine (10−4m) attenuated both the peak (68 ± 5%;P < 0.05) and sustained (35 ± 3%;P < 0.01) phases when compared with the control group (92 ± 8% and 89 ± 5%, respectively) (figs. 6B and 6C). Etomidate (10−4m) also attenuated increases in [Ca2+]iat the peak (67 ± 7%;P < 0.05) and sustained (50 ± 11%;P < 0.01) phases in response to bradykinin.
In contrast to bradykinin, A23187 (10−7m) induced a monophasic increase in endothelial [Ca2+]ithat was sustained even after washing out with fresh medium (fig. 7). The increase in [Ca2+]iinduced by A23187 was 65 ± 5% of the sustained increase in [Ca2+]iin response to bradykinin (10−8m) in the same cell (n = 5). Neither etomidate nor ketamine (10−4M) altered the A23187-induced [Ca2+]itransient (etomidate = 63 ± 8%; ketamine = 67 ± 9%, n = 5 each).
The key findings of our study are as follows. Both etomidate and ketamine attenuated endothelium-dependent pulmonary vasorelaxation in response to the receptor-mediated agonists, acetylcholine and bradykinin, but not the receptor-independent agonist, A23187. The relaxation responses induced by all three of these endothelial activators consisted of NO- and EDHF-mediated components. Etomidate and ketamine attenuated the vasorelaxant responses to acetylcholine and bradykinin by inhibiting both the NO and EDHF components of the response. Finally, both of these anesthetics inhibited the increase in endothelial [Ca2+]iinduced by bradykinin, but not A23187.
The three primary endothelium-derived vasodilators are NO, prostacyclin, and EDHFs. We recently demonstrated that NO and EDHFs are the primary mediators of acetylcholine relaxation in canine pulmonary artery, with prostacyclin only playing a role at high concentrations of acetylcholine. 8Prostacyclin also mediates a small component of bradykinin and A23187-induced pulmonary vasorelaxation. 9To focus on the effects of the anesthetics on the NO- and EDHF-mediated components of these three endothelial activators, rings were pretreated with the cyclooxygenase inhibitor, ibuprofen, before administration of l-NAME or TBA. In ibuprofen-pretreated rings, l-NAME and TBA each attenuated the vasorelaxant responses to acetylcholine, bradykinin, and A23187, and combined inhibition abolished these vasorelaxant responses. Thus, in ibuprofen-pretreated rings, the vasorelaxant response to these endothelial activators that remains after TBA is mediated by NO, and the vasorelaxant response that remains after l-NAME is mediated by EDHFs.
Etomidate and ketamine inhibited receptor-mediated (acetylcholine and bradykinin) but not receptor-independent (A23187) endothelium-dependent vasorelaxation. These results indicate that the anesthetics do not exert an inhibitory effect on NO or EDHF activity, because NO and EDHF are common mediators for all three vasodilators. Moreover, the normal vasorelaxant response to SNAP indicates that these anesthetics do not inhibit NO-induced pulmonary vascular smooth muscle cyclic guanosine monophosphate production. Taken together, these results indicate that the inhibitory effects of etomidate and ketamine involve endothelial signaling pathways requiring receptor activation.
Etomidate and ketamine inhibited the TBA- and l-NAME–resistant vasorelaxant responses to acetylcholine and bradykinin, which indicates that the anesthetics inhibited both the NO- and EDHF-mediated components of the response, respectively. An increase in endothelial [Ca2+]iis an essential step in the production of NO and EDHF. 10,11Thus, we hypothesized that the anesthetics attenuated receptor-mediated relaxation by inhibiting increases in endothelial [Ca2+]i. To test this hypothesis, we assessed the effects of the anesthetics on changes in endothelial [Ca2+]iinduced by bradykinin and A23187. Surprisingly, we were not able to measure consistent changes in endothelial [Ca2+]iin response to acetylcholine, perhaps reflecting phenotypic changes in receptor activity in these cultured cells. We used concentrations of bradykinin and A23187 that were close to the IC50values calculated from the isometric tension measurements. Bradykinin activates endothelial BK2receptors, which stimulates the release of Ca2+from intracellular stores via inositol 1,4,5-triphosphate production, as well as transmembrane Ca2+influx from the extracellular space. 12,13The bradykinin-induced increase in [Ca2+]iconsisted of two phases, an initial rapid increase followed by a gradual decrease to a sustained plateau (fig. 6). The initial phase is likely caused by intracellular Ca2+release, and the sustained phase is caused by Ca2+influx. 14,15Although volatile anesthetics have been reported to inhibit Ca2+mobilization in response to bradykinin in aortic endothelial cells, 14,16the effects of intravenous anesthetics on [Ca2+]itransients in endothelial cells have not been previously investigated. Our observation that etomidate and ketamine inhibited both the initial peak and sustained increases in [Ca2+]iin response to bradykinin suggests that these anesthetics may alter inositol 1,4,5-triphosphate–mediated Ca2+release from the sarcoplasmic reticulum and receptor-activated Ca2+influx.
Receptor-activated Ca2+influx is not mediated via voltage-gated Ca2+channels, because vascular endothelial cells lack functional L-type Ca2+channels. 17It has been postulated that agonist-induced Ca2+influx in endothelial cells is mediated by capacitative Ca2+entry through nonselective cation channels, where the magnitude of Ca2+entry is regulated by the extent of filling of intracellular Ca2+stores. 17–20The anesthetics clearly attenuated the initial peak [Ca2+]itransient in response to bradykinin. This could result in less depletion of intracellular Ca2+stores, thereby causing an indirect anesthesia-induced inhibitory effect on capacitative Ca2+entry. Alternatively, the anesthetics could directly inhibit capacitative Ca2+entry via effects on nonselective Ca2+channels. In preliminary studies, we have observed that propofol attenuates capacitative Ca2+entry in pulmonary artery smooth muscle cells. 21In contrast to receptor activation, A23187 increases [Ca2+]iby increasing the Ca2+permeability of the cell membrane, as well as intracellular organelles containing Ca2+. Neither anesthetic had an effect on the A23187 vasorelaxation response or the A23187-induced increase in [Ca2+]i. Taken together, it seems likely that etomidate and ketamine inhibited the pulmonary vasorelaxant response to acetylcholine and bradykinin by reducing the increase in endothelial [Ca2+]iin response to receptor activation, and therefore reducing the synthesis of NO and EDHF.
Volatile anesthetics have been reported to inhibit both the NO and EDHF components of acetylcholine relaxation in rat 22and rabbit 23mesenteric artery and rabbit carotid artery. 24Etomidate and thiopental attenuated the EDHF component of acetylcholine- and bradykinin-induced relaxation in human renal artery 25and the rat coronary circulation, 26whereas propofol and ketamine inhibited the NO component of acetylcholine relaxation in rat aorta. 27In contrast to the systemic circulation, very little is known about the effects of intravenous anesthetics on endothelium-dependent relaxation in the pulmonary circulation. Our laboratory has reported that propofol attenuates the NO and EDHF components of acetylcholine relaxation, but not bradykinin relaxation, in canine pulmonary arterial rings. 8The effects of etomidate and ketamine on endothelium-dependent relaxation in the pulmonary circulation have not been previously investigated.
The plasma concentration of ketamine after the intravenous administration of 2 mg/kg has been reported to be 1.1 × 10−4m. 28Thus, the effects of ketamine on pulmonary vasorelaxation and the endothelial [Ca2+]itransient were apparent at clinically relevant concentrations. The peak plasma concentration of etomidate during induction of general anesthesia is approximately 10−5m. 29,30Because 75% of etomidate is bound to plasma protein, 31the free plasma concentration of etomidate is likely to be less than 10−5m. Thus, the effects of etomidate reported in this study appear to be significant only at supraclinical concentrations.
In summary, etomidate and ketamine selectively attenuated receptor-mediated endothelium-dependent pulmonary vasorelaxation by inhibiting the NO and EDHF components of the response. These anesthetic effects appear to be caused by inhibition of the endothelial [Ca2+]itransient in response to receptor activation.