Activation of α^2B-Adrenoceptors Mediates the Cardiovascular Effects of Etomidate

The generation of the mouse lines lacking single α²-adrenoceptor subtypes has been described previously. 18,19Mice lacking α^2A- and α^2B-receptors (α^2AB-KO) were generated by crossing of single gene knockout lines to first obtain double heterozygous mice, which were further intercrossed until homozygous α^2AB-deficient mice were born. Genotypes were confirmed by subtype-specific polymerase chain reactions performed with genomic DNA isolated from small tail biopsies as described in detail. 20Mice were maintained in a specified pathogen-free facility. The University of Würzburg and the Government of Unterfranken (Würzburg, Germany) approved all animal procedures (protocol No. 621–2531.01–28/01).

Mice were given an intraperitoneal injection of etomidate (5–50 mg/kg body weight), thiopental (50 mg/kg body weight), or dexmedetomidine (500 μg/kg or 1 mg/kg), and the time after injection at which the righting reflex of mice was lost and the recovery time of this reflex were monitored. Because dexmedetomidine at these doses did not induce a loss of the righting reflex, its sedative effect was also assessed by placing mice on a rotating wheel (30 rpm) and determining the time for which mice stayed on the rod (Rotarod; Ugo Basile, Varese, Italy). For these series of experiments, male and female mice (3–4 months old, 5–10 mice per genotype) were used. The observer was blinded with respect to the genotype of the mice.

For hemodynamic measurements, two groups of male mice were used at the age of 3–5 months. In the first group, the cardiovascular effects of intravenous dexmedetomidine (5 μg/kg) were determined (eight mice per genotype). The second group was used to measure hemodynamic parameters after etomidate injection (6–10 mice per genotype). For aortic and left ventricular catheterization with a 1.4F pressure-volume catheter, 21mice were anesthetized with tribromoethanol (13 μl of 2.5% solution per g of body weight) and placed on a 37°C table. The microtip catheter was inserted into the right carotid artery and the pressure tip was advanced into the aorta or into the left ventricle. For injection of drugs, a PE-10 polyethylene tubing was inserted into the left jugular vein. Etomidate or dexmedetomidine were injected in 0.9% saline into the jugular vein catheter. Hemodynamic data were digitized via  a MacLab system (AD Instruments, Castle Hill, Australia) connected to an Apple G4 PowerPC computer (Apple Computer, Inc., Cupertino, CA). 21 

Stable cell lines of human embryonic kidney (HEK293) cells with high expression levels of either the cloned murine α^2A-adrenoceptor (14.6 pmol receptor/mg protein) or α^2C-adrenoceptor (16.8 pmol receptor/mg protein) subtype were grown to near confluency in cell culture at 37°C. 17In addition, because a stable cell line expressing the α^2B-adrenoceptor subtype was not available, native HEK293 cells were transiently transfected with the cloned mouse α^2B-adrenoceptor subtype by the calcium phosphate precipitation technique 22and used 48 h after transfection. Transient transfection of α^2B-adrenoceptors resulted in expression levels that were similar to the stable cell lines.

Initially, HEK293 cells were washed with 5 ml of phosphate-buffered saline. The cells were then scraped from the culture dishes into 5 ml of ice-cold hypotonic buffer (5 mm Tris, 2 mm EDTA, pH 7.4) and washed again with 5 ml of ice-cold hypotonic buffer. Cells were homogenized with an Ultra-Turrax (Janke & Kunkel, Staufen, Germany) and cell nuclei and whole cells were removed by centrifugation at 1,700 g  for 10 min at 4°C. Membrane vesicles were prepared by centrifugation of the supernatant at 145,000 g  for 30 min at 4°C. The resulting final membrane pellet was resuspended in binding buffer (75 mm N -methyl-d-glucamine, 25 mm glycine, 40 mm HEPES, 5 mm EGTA, 10 mm MgCl, pH 7.4). Protein concentrations were determined using a colorimetric assay based on the method of Bradford 23with bovine serum albumin as a reference standard.

Competition binding of etomidate to α²-adrenoceptors was determined in a 200-μl reaction containing 10 nm to 1 mm etomidate, the α²-adrenoceptor antagonist [³H]RX821002 (3 nm), and binding buffer. Nonspecific binding was determined by addition of the α²-receptor antagonist atipamezole (1 μm). The 200-μl reactions containing only the solvent propylene glycol without etomidate were used as controls. Membranes (10–20 μg protein) were incubated at room temperature for 1 h in the binding mixture assay. The binding reaction was stopped by filtration using a cell harvester and three washes of buffer. Membrane-bound [³H]RX821002 was determined by scintillation counting.

To determine the agonist activity of etomidate, untransfected or α^2B-receptor–transfected cells were stimulated for 20 min with 10 μm etomidate. Before stimulation, HEK cells were incubated overnight in serum-free Dulbecco's Modified Eagle Medium. Stimulated HEK cells were rapidly scraped off the plates and frozen in liquid nitrogen. Cell lysates were prepared and used for Western blotting as described previously. 20Antibodies against the extracellular signal-related kinases ERK1/2 and phospho-ERK1/2 were purchased from Cell Signaling (Beverly, MA).

The α²-antagonist [³H]RX821002 was used as a radioligand with a specific activity of 67 Ci/mmol (Amersham Pharmacia Biotech, Freiburg, Germany). Nonspecific binding was determined using the antagonist atipamezole (Orion Corp., Turku, Finland). Etomidate was either used as Hypnomidate® (2 mg etomidate/ml dissolved in 35% propylene glycol in water; Janssen-Cilag, Neuss, Germany) or as Etomidate®-Lipuro (2 mg etomidate/ml liposome suspension; Braun, Melsungen, Germany). Propylene glycol was used as the inactive control solution.

The results of the competition binding were analyzed using a nonlinear regression program. A binding curve was fitted according to a one-site competition model (Prism; GraphPad, San Diego, CA). Data are presented as geometric mean with 95% confidence intervals. From the competition curves, Kⁱvalues were calculated from EC⁵⁰values using the Cheng Prusoff equation. 24Hemodynamic data were analyzed using Student t  test for unpaired samples. A P  value of less than 0.05 was considered as statistically significant. Results are displayed as means ± SEM.

In vivo , the sedative effect of α²-receptor activation is almost exclusively mediated by the α^2A-receptor subtype. 25Thus, to test whether the sedative/hypnotic effect of etomidate is mediated via  the α^2A-receptor, the times for the loss and for recovery of the righting reflex were determined in wild-type mice and in mice lacking α^2A- or α^2B-receptors (α²-KO). For these experiments, intraperitoneal etomidate doses between 5 mg/kg and 50 mg/kg were tested in wild-type mice. Whereas 5 mg/kg etomidate did not result in loss of the righting reflex, 27% of the mice lost the reflex at 10 mg/kg etomidate and all mice transiently lost the righting reflex at 20-mg/kg and higher etomidate doses. Thus, we chose a dose of 30 mg/kg intraperitoneal etomidate to investigate sedative effects of etomidate in α²-receptor-KO mice. After injection of etomidate, the righting reflex disappeared in wild-type and α^2A-deficient mice at similar times after intraperitoneal injection (fig. 1a). The duration of etomidate-induced sleep was not affected by the α^2A-deletion and the righting reflex recovered at similar times in wild-type and α^2A-KO mice. Similarly, hypnotic properties of intraperitoneal thiopental (50 mg/kg) did not differ between wild-type and α^2A-KO mice (fig. 1b). Deletion of the α^2B-receptor gene did not affect the sedative effects of etomidate or thiopental in mice (fig. 1). Intraperitoneal injection of the non–subtype-selective α²-agonist dexmedetomidine up to 1 mg/kg did not result in a loss of the righting reflex. Thus, the sedative effect of dexmedetomidine was assessed on a rotating rod. Dexmedetomidine exerted a sedative effect in wild-type mice and in α^2B-KO mice, as evidenced by the observation that the mice were unable to stay on a rotating rod for longer than 10–20 s (wild-type, 9 ± 2 s; α^2B-KO, 14 ± 6 s; n = 8 mice per genotype, 30 min after 500 μg/kg intraperitoneal dexmedetomidine). In contrast, α^2A-KO mice did not show any sedative effects in response to dexmedetomidine (α^2A-KO stayed on the rod for more than 60 s, n = 10 mice).

To test potential effects of etomidate on cardiovascular α²-receptor subtypes, we first generated double-knockout mice deficient in α^2A- and α^2B-receptors (fig. 2 a and b). The α^2AB-deficient mice were generated by crossing single receptor knockouts to obtain double heterozygotes that were further crossed to yield mice with the α^2A−/−^B−/− genotype (fig. 2c). At weaning age, α^2AB-KO mice were detected at a frequency that was close to the expected Mendelian ratio, indicating that genetic deletion of these receptors did not interfere with embryonic development. After weaning, α^2AB-deficient mice developed normally and showed no obvious phenotypes.

First, we determined basal cardiovascular parameters in anesthetized α^2AB-KO mice. Mean arterial pressure did not differ between α^2AB-deficient animals (85 ± 4 mmHg, n = 9) and wild-type mice (89 ± 6 mmHg, n = 11). However, α^2AB-KO mice had a higher resting heart rate (493 ± 13 min⁻¹, n = 9) than wild-type control mice (446 ± 13 min⁻¹, n = 11). The cardiovascular response to the α²-receptor agonist was tested in anesthetized mice of different genotypes. As reported previously, rapid intravenous injection of dexmedetomidine resulted in a transient hypertension in wild-type mice that was followed by a long-lasting hypotension (fig. 3a). Deletion of the α^2A-receptor disrupted the hypotensive response, whereas the initial hypertension was absent in α^2B-deficient animals. Interestingly, α^2AB-KO mice showed a small but significant decrease in systolic blood pressure (−13 ± 4 mmHg). This pressure effect was accompanied by a decrease in heart rate (−25 ± 8 min⁻¹, fig. 3b).

On intravenous injection of etomidate, wild-type mice showed a rapid and transient hypertensive response that was completely absent in mice lacking α^2B-receptors (fig. 4). During the hypertensive phase, heart rate decreased by 18 ± 6 min⁻¹in wild-type mice. No significant alterations in heart rate were observed in α^2B-KO or α^2AB-KO mice. After intravenous etomidate, mean arterial blood pressure remained above the preinjection value for 2.4 ± 0.2 min. The hypertensive effect of etomidate in wild-type mice was dose dependent: 1 mg/kg etomidate caused a small increase in blood pressure (systolic pressure, +6.7 ± 2.6 mmHg; diastolic pressure, +5.7 ± 2.2 mmHg; n = 6), whereas a dose of 2 mg/kg etomidate elicited significantly larger pressure responses (fig. 4; systolic pressure, +15.7 ± 2.7 mmHg; diastolic pressure, +17.0 ± 2.6 mmHg).

Two different preparations of etomidate, one in propylene glycol and the other in liposome suspension, showed identical effects on blood pressure in wild-type mice. Etomidate did not alter cardiac contractility as assessed by direct catheterization of the left ventricle. Maximal left ventricular contraction rate after intravenous injection of 2 mg/kg etomidate was identical (dp/dt^max101.6 ± 2.1% of control) to the value before etomidate.

To confirm the in vivo  data, we next tested whether etomidate interacts with recombinant α²-receptor subtypes expressed in HEK293 cells. For these experiments, cells expressing 14–18 fmol/mg of the three α²-receptor subtypes were used. Etomidate inhibited binding of the specifically bound α²-receptor antagonist [³H]RX821002 in a concentration- and subtype-dependent manner with greater potency for α^2B- and α^2C-receptors than for α^2A-adrenoceptors (fig. 5).

To test whether etomidate acted as an agonist or antagonist at α^2B-receptors, the effect of etomidate on α²-receptor–mediated mitogen-activated protein kinase activation was assessed in HEK293 cells. Among other intracellular signaling pathways, α²-receptors can effectively activate the mitogen-activated protein kinase cascade. 20In untransfected or α^2B-receptor–expressing HEK293 cells, basal phosphorylation of the ERK1/2 kinases was hardly detectable (fig. 6). However, short-term exposure to etomidate increased ERK1/2 phosphorylation in HEK293 cells transfected with α^2B-receptors but not in untransfected cells (fig. 6).

The present results indicate that etomidate interacts with α²-adrenoceptors and displaces the α²-adrenoceptor antagonist [³H]RX821002 in HEK293 cells from all subtypes in a concentration-dependent manner. In addition, etomidate leads to a transient increase in blood pressure in wild-type mice, but not in α^2B-KO or α^2AB-KO mice, suggesting an interaction of etomidate with α^2B-adrenoceptors in the peripheral vasculature.

α²-Adrenoceptors have been implicated in a variety of physiologic functions. Pharmacologic studies and the development of genetic mouse models elucidated the physiologic effects mediated by the different α²-adrenoceptor subtypes. 16,18,26,27The α^2A-adrenoceptor subtype has been reported to be the predominant subtype involved in the antinociceptive, sedative, hypotensive, hypothermic, and behavioral actions of α²-adrenoceptor agonists. 25,26,28,29Stimulation of α^2B-adrenoceptors in vascular smooth muscle leads to vasoconstriction, which causes the initial hypertension after administration of α²-adrenoceptor agonists. 18In addition, the α^2B-adrenoceptor subtype is involved in mediation of the antinociceptive action of nitrous oxide. 30,31The α^2C-receptor has been shown to modulate dopaminergic neurotransmission, various behavioral responses, and to induce hypothermia. 32,33In addition, this subtype contributes to spinal antinociception of the imidazoline moxonidine in mice. 34 

The exact molecular targets by which etomidate mediates its anesthetic effects remain under discussion. Clinically relevant plasma concentrations of etomidate to induce hypnosis vary considerably. During induction of anesthesia in humans, initial plasma concentrations of etomidate may be as high as 12–24 μm. 35These concentrations are similar to the Kⁱvalue of etomidate for α^2B-receptors expressed in HEK293 cells in our study. Etomidate concentrations from 1.2–8.2 μm have been reported as the minimal hypnotic plasma levels for anesthesia in humans. 35It has been shown that etomidate at these concentrations exhibits γ-aminobutyric acid–modulatory effects, and at higher concentrations exhibits γ-aminobutyric acid–mimetic effects at γ-aminobutyric acid receptor type A receptors. 4,5,36,37Interestingly, mice with a point mutation in the β3 subunit of the γ-aminobutyric acid receptor type A receptor did not show any suppression of noxious-evoked movements in response to etomidate. 6In this study, 5 mg/kg etomidate was injected intravenously in wild-type mice to elicit a loss of the reflex, which lasted for approximately 10 min. 6In our study, we observed a significant transient hypertensive response to intravenous injection of 2 mg/kg etomidate. Thus, higher doses of etomidate may even cause a more pronounced hypertensive phase in mice. However, in the other study, the mice were already anesthetized with tribromoethanol for arterial catheterization and we did not use higher doses of etomidate to prevent toxicity. The low affinity of etomidate for the α^2A-adrenoceptor subtype determined in the current study challenges the assumption that the anesthetic effects of etomidate could be partially mediated by an action on cerebral α²-adrenoceptors. Indeed, we found no difference in the sedative/hypnotic effect of etomidate between normal mice and animals lacking functional α^2A-adrenoceptors.

Our results show a higher affinity of etomidate for α^2B- and α^2C-adrenoceptor subtypes expressed in HEK293 cells. The role of the α^2Cadrenoceptor subtype in anesthesia and cardiovascular control is currently still unknown. Previous investigations in mice have shown that disruption of the α^2C-adrenoceptor subtype does not lead to hemodynamic effects. 18However, α^2C-receptors operate together with α^2A-receptors as presynaptic inhibitory regulators of sympathetic norepinephrine release. 21To unequivocally distinguish the cardiovascular roles of α^2C- and α^2B-receptors in vivo , we generated mice deficient in α^2A- and α^2B-receptors. Interestingly, these mice showed a small but significant decrease of the heart rate in response to the α²-agonist dexmedetomidine, which was accompanied by a decrease in aortic blood pressure. Administration of dexmedetomidine failed to produce the initial hypertension, which can be attributed to the α^2B-adrenoceptor subtype 18and the central hypotensive effect related to the α^2A-adrenoceptor subtype. 26These findings are consistent with the observation that activation of presynaptic sympathetic α^2C-receptors can decrease norepinephrine release. 21 

Etomidate elicited a transient increase in aortic blood pressure in anesthetized wild-type mice that was not observed in α^2B-KO or in α^2AB-KO mice. The hypertensive response was accompanied by a mild bradycardia. In humans, the most distinctive property of etomidate as compared with other anesthetics is its minimal effect on cardiovascular parameters. 2,3However, several studies in patients have shown that etomidate may cause transient increases in blood pressure and/or systemic vascular resistance. 38,39Thus, during induction of anesthesia etomidate may interact with vascular α^2B-adrenoceptors to elicit a transient hypertensive effect or to counteract the hypotensive effects of other anesthetics and drugs used in anesthesia. In addition to a direct effect on vascular α^2B-receptors, etomidate may elicit the transient hypertension by an indirect mechanism. It is tempting to speculate that increased levels of catecholamines (caused by etomidate) may contribute to the transient hypertension. However, wild-type or α²-deficient mice did not show any typical signs of increased norepinephrine secretion; i.e. , tachycardia (fig. 4). Further studies on isolated microvessels are required to determine whether the etomidate-mediated hypertension is indeed mediated by a direct vasoconstrictory mechanism.

Interaction of etomidate with the α^2A-receptor subtype is unlikely to play an important role in mediating its anesthetic action because of the low affinity of etomidate for this receptor subtype. This result is also consistent with the clinical profile of etomidate in humans. Activation of α^2A-receptors elicits analgesia and inhibits sympathetic outflow. Etomidate does not have significant analgesic effects in humans and it does not block the sympathetic activation that is usually observed during intubation. 40 

In conclusion, etomidate can stimulate α^2B-adrenoceptors in mice, which results in a transient hypertensive response after rapid intravenous injection. Although this study was performed with transgenic mouse models and recombinant murine α²-adrenoceptors, it is tempting to speculate about the implications for the mechanism of action of etomidate in humans. The cardiovascular stability of etomidate during induction of anesthesia even in patients with cardiac disease and during bolus application might be related to an interaction of etomidate with peripheral α^2B-adrenoceptors. The α^2B-mediated vasoconstriction may oppose hypotensive effects of other coadministered agents during induction of anesthesia, and thus be responsible for the known beneficial hemodynamic profile of etomidate.