Neuropharmacology 45 (2003) 57–71 www.elsevier.com/locate/neuropharm
The in vitro and in vivo enantioselectivity of etomidate implicates the GABAA receptor in general anaesthesia Delia Belelli a, Anna-Lisa Muntoni a, Simon D. Merrywest b, Luc J. Gentet a, Anna Casula a, Helen Callachan a, Paola Madau a, David K. Gemmell c, Niall M. Hamilton c, Jeremy J. Lambert a, Keith T. Sillar b, John A. Peters a,∗ a
b
Department of Pharmacology and Neuroscience, Ninewells Hospital and Medical School, University of Dundee, Dundee DD1 9SY, Scotland, UK School of Biology, Division of Biomedical Sciences, Bute Medical Building, University of St. Andrews, St. Andrews KY16 9TS, Scotland, UK c Departments of Pharmacology and Chemistry, Organon Laboratories, Newhouse, Lanarkshire ML1 5SG, Scotland, UK Received 31 January 2003; received in revised form 21 March 2003; accepted 24 March 2003
Abstract General anaesthetics exhibiting enantioselectivity afford valuable tools to assess the fundamental mechanisms underlying anaesthesia. Here, we characterised the actions of the R-(+)- and S-(⫺)-enantiomers of etomidate. In mice and tadpoles, R-(+)-etomidate was more potent (~10-fold) than S-(⫺)-etomidate in producing loss of the righting reflex. In electrophysiological and radioligand binding assays, the enantiomers of etomidate positively regulated GABAA receptor function at anaesthetic concentrations and with an enantioselectivity paralleling their in vivo activity. GABA-evoked currents mediated by human recombinant GABAA receptors were potentiated by either R-(+)- or S-(⫺)-etomidate in a manner dependent upon receptor subunit composition. A direct, GABAmimetic, effect was similarly subunit dependent. Modulation of GABA receptor activity was selective; R-(+)-etomidate inhibited nicotinic acetylcholine, or 5-hydroxytryptamine3 receptor subtypes only at supra-clinical concentrations and ionotropic glutamate receptor isoforms were essentially unaffected. Acting upon reticulothalamic neurones in rat brain slices, R-(+)-etomidate prolonged the duration of miniature IPSCs and modestly enhanced their peak amplitude. S-(⫺)-etomidate exerted qualitatively similar, but weaker, actions. In a model of locomotor activity, fictive swimming in Xenopus laevis tadpoles, R-(+)- but not S-(⫺)-etomidate exerted a depressant influence via enhancement of GABAergic neurotransmission. Collectively, these observations strongly implicate the GABAA receptor as a molecular target relevant to the anaesthetic action of etomidate. 2003 Elsevier Science Ltd. All rights reserved. Keywords: General anaesthesia; GABAA receptor; Etomidate; GABAergic inhibitory synaptic transmission
1. Introduction Despite intensive investigation, the molecular mechanism(s) of general anaesthesia remain(s) elusive. General anaesthetics can modulate the function of many regulatory proteins mediating signal transduction within the nervous system, but actions at relevant concentrations involving anatomically and physiologically plausible targets are less extensive (Franks and Lieb, 1994). On such criteria, transmitter-gated ion channels
Corresponding author. Tel.: +44-1382-660111; fax: +44-1382667120. E-mail address:
[email protected] (J.A. Peters). ∗
emerge as likely targets of anaesthetic action. In particular, anion currents mediated by the GABAA receptor are reversibly potentiated by appropriate concentrations of many volatile and intravenous anaesthetic agents (Belelli et al., 1999; Krasowski and Harrison, 1999). Many general anaesthetics also exert a direct GABAmimetic effect, although this additional action is often most apparent at drug concentrations that are out with the range thought to be relevant to anaesthesia (Belelli et al., 1999). Congruence between stereoselectivity of action at the behavioural and molecular levels offers a powerful means of identifying molecular targets that are relevant to anaesthesia (Franks and Lieb, 1994). Studies utilising the enantiomers of anaesthetic barbiturates (Tomlin et
0028-3908/03/$ - see front matter 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0028-3908(03)00144-8
58
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
al., 1999), pregnane steroids (Wittmer et al., 1996), etomidate (Tomlin et al., 1998) and isoflurane (Jones and Harrison, 1993; Hall et al., 1994) have revealed a good correlation between potentiation of GABAA receptor activity and anaesthetic potency, strongly implicating modulation of GABAergic transmission in anaesthesia. The stereoselectivity of the imidazole, etomidate, is of particular interest because, in contrast to most clinical anaesthetics, it is utilised as the resolved R-(+)-enantiomer rather than a racemate (Doenicke and Ostwald, 1997). In mammals, the GABAA receptor is a hetero-pentameric protein assembled from a repertoire of α(1–6), β(1–3), γ(1–3), δ, ε and θ subunits (Hevers and Luddens, 1998). The most abundant GABAA receptor isoforms within the CNS are composed of α, β and γ subunits, with a proposed stoichiometry of 2α2β1γ, or 2α1β2γ. The precise subunit complement of the pentameric complex defines both the biophysical properties of the receptor and its pharmacological characteristics (Hevers and Luddens, 1998; Mohler et al., 2002). With particular pertinence to the present work, it has been shown that the interaction of R-(+)-etomidate with hetero-oligomeric GABAA receptors expressed in Xenopus oocytes is strongly influenced by the identity of the β subunit present within the receptor complex (Hill-Venning et al., 1997). For both the GABA-modulatory and GABAmimetic effects of R-(+)-etomidate, activity is favoured by the expression of β2- or β3- vs. β1-subunits within the receptor complex (Belelli et al., 1997; Hill-Venning et al., 1997). Although a previous study has examined enantioselective modulation of the bovine recombinant α1β1γ2L receptor GABAA receptor isoform by etomidate (Tomlin et al., 1998), we considered it important to extend these observations to human recombinant receptors containing the β3-subunit and also to GABAA receptors in their native environment. Despite receptors expressing the β3-subunit constituting only 19–25% of the total β-subunit-containing GABAA receptor population in the CNS (Benke et al., 1994), their relevance to anaesthesia is underscored by a very recent study employing mutant β3-subunit ‘knock-in’ mice wherein receptors harbouring the β3-subunit lack sensitivity to etomidate in in vitro assays. In such animals, the loss of the righting reflex (LRR) and suppression of nocifensive reflexes in response to etomidate were greatly attenuated (Jurd et al., 2003). In the present study, the potencies of the R-(+)- and S(–)-enantiomers of etomidate (Janssen et al., 1975) were quantified in vivo utilising loss of the righting reflex in Xenopus laevis tadpoles and mice as well characterised models of the obtunding effect of anaesthetics in man. The degree of enantioselectivity observed in vivo was compared with that found in vitro using assays of GABAA receptor function that included the binding of the selective radioligand [35S]t-butylbicyclophosphoro-
thionate ([35S]TBPS) to GABAA receptors native to rat brain and electrophysiological assays of human recombinant GABAA receptors expressed in Xenopus laevis oocytes. To evaluate the effect of the enantiomers of etomidate upon GABAergic synaptic transmission within the CNS, pharmacologically isolated GABAA receptormediated miniature IPSCs were recorded from neurones of the reticular nucleus of rodent thalamus in the absence and presence of anaesthetic. The latter, via their inhibitory intranuclear projections, are a crucial determinant of the patterning of activity in the thalamic relay nuclei (Cox et al., 1997). To facilitate a comparison of enantioselectivity at the behavioural and synaptic levels within the same model system, we determined the effect of etomidate upon fictive swimming in Xenopus tadpoles and established how the synaptic input to the spinal motoneurones that drive this response is modulated by the anaesthetic. Actions at the spinal level are likely to be of importance in mediating the immobilising effect of general anaesthetics (Collins et al., 1995). Overall, our primary goal was to provide a comprehensive description of the enantioselectivity of etomidate that would be useful in evaluating the hypothesis that modulation of GABAA receptor function by etomidate is an event important to general anaesthesia.
2. Methods 2.1. Determination of hypnotic potency of anaesthetics in mice and Xenopus laevis tadpoles All aspects of the work in this report were performed in accordance with the UK 1986 Animals (Scientific Procedures) Act. Male MF1 mice (Interfauna; 25–35 g), kept in a 12 h light/dark cycle, had access to food and water ad libitum. All tests were performed during the light period as previously described (Anderson et al., 2001). In brief, groups of eight mice were used to assess the effect of each dose of the enantiomers of etomidate tested. Compounds were administered i.v. (lateral tail vein; 10 ml kg–1) over a 10 s period. Mice were subsequently placed in separate enclosures and tested for loss of the righting reflex. Dosing was performed according to a protocol where the amount of anaesthetic injected (µmol kg–1) doubled until LRR was observed. Additional doses were introduced to provide data for LRR over a narrower range that permitted the calculation of anaesthetic potency. The dose of anaesthetic that caused LRR in 50% of animals for a period of not less than 30 s was quantified as the hypnotic dose50 (HD50). Hypnotic concentrations of the enantiomers of etomidate under equilibrium conditions were determined using pre-limb bud Xenopus laevis tadpoles. Groups of eight tadpoles (Blades Biological, Edenbridge, Kent,
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
UK) were placed in glass dishes each containing 100 ml of tap water at room temperature (~20 °C) and after a 1 h acclimatisation period, the etomidate enantiomers were added from concentrated stock solutions. Following a 180 min equilibration period, each tadpole was tested for LRR. If a tadpole did not right within 5 s following tipping over with a heat-polished rod, it was considered to have lost the righting reflex. The concentration of anaesthetic producing LRR in 50% of tadpoles was defined as the HC50. At the termination of the experiment, tadpoles were returned to tap water and return of the righting reflex was verified. Anaesthetic HD50 and HC50 values were determined by probit analysis (SAS Institute) and are reported with the 95% confidence limits. 2.2. Cell culture and transfection of HEK 293 cells Human embryonic kidney cells (HEK 293) were maintained in minimal essential medium supplemented with 10% (v/v) foetal calf serum, streptomycin (100 mg l–1), penicillin (1 × 105 i.u. l–1) at 37 °C in an atmosphere of 95% air/5% CO2 and 100% relative humidity. Transfection of cells with cDNA (15–20 µg per plate) encoding the human GABAA β3 subunit, subcloned into the pCDNAIAmp expression vector, was performed using the calcium phosphate precipitation technique following standard protocols (Chen and Okayama, 1988). After an incubation period of 24 h, cells were washed with phosphate buffered saline (PBS), collected by scraping, and pelleted by centrifugation at 6000 g for 10 min. Following two further cycles of centrifugation and resuspension into PBS, the cells were either frozen at ⫺70 °C, or used directly to prepare a crude membrane preparation as described below. 2.3. Analysis of [35S]TBPS binding to rat brain membranes and membranes of HEK 293 cells expressing the GABAA receptor b3 subunit. Male Sprague-Dawley rats (250–400 g) were killed by cervical dislocation, and synaptic membrane fractions prepared as previously described in detail (Hill-Venning et al., 1996). Radioligand binding assays ultilising [35S]TBPS were performed upon whole brain membranes subjected to freeze-thawing to remove endogenous GABA (Hill-Venning et al., 1996). Assay tubes, in triplicate, contained: 50 µl membrane homogenate (0.1 mg protein), 20 µl of either R-(+)- or S-(–)-etomidate, or an equivalent volume of assay buffer/vehicle and GABA (0.6 µM) in a final volume of 0.2 ml containing [35S]TBPS at a concentration of 1 nM. The assay buffer comprised: KH2PO4 (20 mM) and KCl (200 mM) adjusted to pH 7.4 with KOH. The binding of [35S]TBPS to the GABAA receptor complex equilibrated over a period of 150 min at 21 °C and was terminated by rapid
59
vacuum filtration over Whatman GF/B glass fibre filters. The filters were washed with 6 ml of ice-cold buffer and retained radioactivity was quantified by liquid scintillation spectroscopy. Non-specific binding, defined by the inclusion of picrotoxin (200 µM), typically amounted to ⬍5% of total binding. The protein concentration of the homogenate was estimated by the method of Lowry et al. (1951) using bovine serum albumin as standard. In assays examining the binding of [35S]TBPS to GABAA receptor β3 subunits expressed in HEK 293 cells, a crude membrane fraction was prepared in assay buffer from freshly harvested cells, or frozen stocks, using an Ultraturax (full speed, 15 s). Membrane aliquots (100 µl, approximately 0.5 mg protein) were equilibrated with [35S]TBPS (2 nM) in a total volume of 500 µl for 90 min at 25 °C. Free and bound radioligands were separated by the addition of 4 ml of ice-cold assay buffer followed by rapid filtration over Whatman GF/B filters. The latter were washed twice with 4 ml aliquots of icecold buffer and specific binding (defined by the inclusion of picrotoxin, 100 µM) was subsequently quantified by conventional liquid scintillation spectroscopy. The inhibition of [35S]TBPS binding by allosteric modulators of the GABAA receptor was described by an iteratively fitted four parameter equation of the form: [D]nH B ⫽ Bmax [D]nH ⫹ [IC50]nH where, Bmax is specifically bound [35S]TBPS in the absence of modulatory drug, B is binding in the presence of modulator at concentration [D], IC50 is the concentration of modulator causing half maximal inhibition of binding and nH is the interaction coefficient. All data are reported as the mean ± S.E.M. of values derived from the fitted curve. 2.4. Electrophysiological recordings from Xenopus laevis oocytes Complementary DNAs encoding the neurotransmitter receptor subunits employed in this study were kindly provided by the following sources: human GABAA α1, α6, β3 and γ2L, Dr. P. Whiting, Merck Sharpe and Dohme, Harlow, UK; rat NMDA NR1a and NR2A and rat AMPA GluR1 flop (GluR1o) and GluR2 flop (GluR2o), Professor P. Seeberg, Max-Planck Institute for Medical Research, Heidelberg, Germany; rat nicotinic ACh α4 and β2, Dr. J. Boulter, Department of Psychiatry, University of California at Los Angeles, CA, USA; chick (Gallus gallus) nicotinic ACh α7, Professor J.-P. Changeux, Institut Pasteur, Paris, France. The cDNA encoding the human 5-HT3A receptor subunit was isolated as described by Belelli et al. (1995). Where appropriate, cRNA transcripts were prepared from linearised cDNAs according to standard protocols (Belelli et al., 1996).
60
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
Xenopus laevis oocytes were isolated from toads anaesthetised by immersion in 0.2% w/v tricaine methanesulphonate in cool water and subsequently killed in accordance with Schedule 1 of the UK Government Animals (Scientific Procedures) Act 1986. Defolliculated oocytes were prepared as previously described (Belelli et al., 1995; Pistis et al., 1999) and incubated in Barth’s solution comprising (in mM) NaCl, 88; KCl, 1; NaHCO3, 2.4; MgCl2, 1; CaCl2, 1 and Hepes, 15 (pH 7.5). Stage VI oocytes were injected intranuclearly with 100–250 pg of the appropriate cDNAs (glutamate receptor subunits), or cytoplasmically with 30–50 ng of the appropriate cRNAs (GABAA, nicotinic ACh and 5-HT3A receptor subunits), both in 20 nl of double distilled deionised water using a Drummond Digital Microdispenser 510 (Drummond Scientific Co., Broomall, PA, USA). The injected oocytes were stored at 19 °C in individual wells of a 96 well microtiter plate in 200 µl of standard Barth’s supplemented with gentamycin (100 µg ml–1). Recordings were made from oocytes 2–12 days after cDNA or cRNA injection as appropriate. Agonist evoked currents were recorded at a holding potential of ⫺60 mV utilising either an Axoclamp 2A or a Gene Clamp 500 amplifier (Axon Instruments, Foster City, CA, USA) in the two-electrode voltage-clamp mode. Oocytes were held in a recording chamber (0.5 ml) and in the majority of recordings continually superfused (7–10 ml min–1) with frog Ringer solution with the composition (in mM): NaCl 120, KCl 2, CaCl2 1.8, Hepes 5, adjusted to pH 7.4 with NaOH). In experiments examining the effect of etomidate upon current responses mediated by NMDA and nicotinic ACh receptor subtypes, Ca2+ was replaced by Ba2+ to minimise the activation of the Ca2+-activated Cl-conductance endogenous to oocytes. In addition, NMDA receptor activity was recorded in nominally Mg2+-free Ringer solution supplemented with the coagonist glycine (10 µM). Current-passing and voltagesensing intracellular electrodes had resistances of 0.5– 1.3 M⍀ (when filled with 3 M KCl and measured in frog Ringer solution). In all studies, a maximally effective concentration of agonist was applied once every 15–20 min until the peak inward current response produced was stable. The magnitude of the GABA enhancing actions of anaesthetics is dependent upon the concentration of GABA utilised (Belelli et al., 1996). Hence, the concentration of GABA that evoked a response amounting to 10% of the current produced by a saturating concentration of agonist (i.e. EC10) was determined for each oocyte and used to evaluate the modulation of agonistevoked currents by etomidate. Potentiation by etomidate enantiomers was quantified by expressing the modulated response as a percentage of the maximal response evoked by GABA (denoted Emax; see Pistis et al., 1999). At receptor types where etomidate caused inhibition (i.e. nicotinic ACh, 5-HT3A, AMPA or NMDA receptors),
control responses were elicited by the agonist at the appropriate EC50. In those instances, the effect of etomidate was quantified as the percentage change in the peak amplitude of the control agonist-evoked current. Anaesthetics were pre-applied for 20–60 s before their coapplication with the appropriate concentration of agonist. Concentration–effect relationships for the potentiating, or inhibitory, effects of etomidate were fitted iteratively with the four parameter logistic equation: I Imax
⫽
[A]nH [A]nH ⫹ [EC50]nH
where, I is the amplitude of the agonist-evoked current in the presence of the anaesthetic at concentration [A], Imax is the amplitude of the response in the presence of a maximally effective concentration of etomidate, EC50 is the concentration of anaesthetic producing half-maximal enhancement, or inhibition, and nH is the Hill coefficient. Concentration–effect relationships for the GABA-mimetic effects of etomidate were similarly fitted where I now represents the amplitude of the current evoked by etomidate at concentration [A], Imax is the amplitude of the response in the presence of a maximally effective concentration of etomidate (expressed as a percentage of the maximal current evoked by a saturating concentration of GABA in the same population of oocytes) and EC50 is the concentration of etomidate producing a half-maximal response. Quantitative data are presented as the mean ± S.E.M. 2.5. Thalamic slice preparation and electrophysiology Thalamic slices were prepared from Sprague-Dawley rats, of either sex, aged 16–22 days as described previously (Huguenard and Prince, 1994). Animals were killed by cervical dislocation in accordance with Schedule 1 of the UK Government Animals (Scientific Procedures) Act 1986. Following decapitation, the brain was quickly removed and placed in high-Mg2+ ice-cold artificial cerebro-spinal fluid (aCSF) containing (in mM): NaCl, 124; KCl, 3; NaHCO3, 26; NaH2PO4, 1.25; CaCl2, 2, MgSO4, 10; d-glucose, 10, and was bubbled with 95% O2/5% CO2 to give a pH of 7.4. The tissue was maintained in ice-cold aCSF whilst horizontal 300 µm slices were cut using a Vibratome (Intracel, Royston, Hertfordshire, UK). The slices were incubated at 32 °C for 1 h in a chamber filled with circulating, oxygenated, standard aCSF containing 1 mM Mg2+ and subsequently allowed to cool to room temperature before use. Whole-cell patch-clamp recordings using an Axopatch 200B amplifier (Axon Instruments, Foster City, CA, USA) were made from reticulothalamic (nRT) neurones visualised under a Zeiss Axioskop FS (Carl Zeiss, Welwyn Garden City, UK) equipped with infrared differential interference contrast (IR-DIC). Patch pipettes
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
were prepared from thick-walled borosilicate glass (O.D. 1.55 mm × I.D. 1.15 mm, Garner Glass Company, Claremont, CA, USA) and had open tip resistances of 3–6 M⍀ when filled with intracellular solution comprising (in mM): CsCl, 140; Hepes, 10; EGTA, 10; CaCl2 2; MgATP, 5; QX-314 (lidocaine-N-ethyl bromide), 5, pH 7.3, 290–310 mOsm l–1. The series resistance (7–18 M⍀) was compensated up to 80% and monitored throughout the experiment. Miniature IPSCs were recorded at a holding potential of ⫺60 mV at 35 °C in the presence of 2 mM kynurenic acid to block ionotropic glutamate receptor-mediated synaptic transmission and 0.5 µM tetrodotoxin (TTX) to block action potentialdependent IPSCs. Data were stored to digital audio tape (Biologic digital tape recorder, DTR-120, Intracel, Royston, Hertfordshire, UK) and analysed using Axograph 4.0 (Axon Instruments, Foster City, CA). Currents were filtered and sampled at 2 and 10 KHz, respectively. Individual mIPSCs were identified using a template sliding along the current trace (Clements and Bekkers, 1997). All events were examined by eye and any contaminated by noise, or superimposed events, were excluded from analysis. Ensemble average traces of at least 50 mIPSCs were further analysed for extraction of peak amplitude, 10–90% rise time, and decay time constants. The decay phase of the mIPSCs was fitted with either one or two exponential functions given by the equations: y(t) = A·e(⫺t / t), or y(t) = A1·e(⫺t / t1) + A2·e(⫺t / t2), respectively, where A is amplitude, t is time and t is the decay time constant. An improvement in the fit by two exponentials compared to one was associated with a reduction in the standard deviation of the residuals i.e. the differences between the data points and the fitted curve. Goodness of fit was also assessed by eye and confirmed by an Ftest. Fitting of a single exponential to biexponentially decaying currents invariably resulted in an underestimation of the peak amplitude, in addition to an inadequate description of decay time course. Additionally, a weighted decay time constant (tw) was calculated according to the equation: tw = t1·P1 + t2·P2, where t1 and t2 are the decay time constants of the first and second exponential functions and P1 and P2 are the proportions of the current amplitude described by each component. Data are expressed as the mean ± S.E.M. Statistical tests were performed with repeated measures ANOVA with Student–Newman–Keuls multiple comparison comparisons test if appropriate. Statistical significance was accepted as P ⬍ 0.05. 2.6. Fictive swimming activity in Xenopus laevis embryos and larvae The methods employed have been previously described in detail (Reith and Sillar, 1997), so only a brief description is given here (see also preparation illus-
61
trated in Fig. 3A). All experiments were performed according to the UK 1986 Animals (Scientific Procedures) Act on pre-feeding embryos (stage 37/8) and larvae (stage 42; Nieuwkoop and Faber, 1956) of the South African clawed frog, Xenopus laevis. Immobilised (12.5 µM α-bungarotoxin) animals were pinned onto a Sylgard-coated rotatable Perspex platform within a preparation bath containing continuously re-circulating frog Ringer solution (composition in mM; 115 NaCl, 2.5 KCl, 1 MgCl2, 2.4 NaHCO3, 10 Hepes, 2 CaCl2 (extracellular recording) or 4 CaCl2 (intracellular recording) pH 7.4; 20–22 °C). Extracellular recordings of ventral root activity were performed via glass suction electrodes placed onto the exposed inter-myotomal clefts, wherein the motor axons are located. For intracellular recordings, a section of myotomes overlying the rostral spinal cord was removed and recordings were made with glass microelectrodes (filled with 2 M potassium chloride to reverse and enhance chloride dependent IPSPs) of 100–150 M⍀ resistance. Fictive swimming activity was initiated by applying a 1 ms current pulse to the tail skin via a glass stimulating electrode. Electrophysiological data were stored on videotape and data analysis was performed on 30 consecutive cycles of motor activity beginning 500 ms after the start of each of three different episodes (to minimise possible contamination by effects from the initiating sensory stimulus). Data are presented as means ± S.E.M. Statistics (Student’s t-test) were considered significant at P ⬍ 0.05. Three parameters of fictive swimming activity were analysed: (a) the episode duration (in s), measured from the first to the last motor burst observed during an episode; (b) the cycle period (in ms), measured as the time interval from the start of one burst to the start of the next; (c) the burst length (in ms), measured as the duration of each discrete motor burst.
2.7. Reagents
All cell culture reagents were obtained from Invitrogen Ltd (Paisley, UK). Drugs, with the exception of isomers of etomidate, were purchased from Sigma-Aldrich (Poole, UK). Drugs were dissolved as concentrated stock solutions in the appropriate medium, or diluted from stocks in double distilled deionised water. [35S]TBPS (2.6–7.4 TBq mmol–1) was purchased from NEN Life Science Products (Zaventem, Belgium). The hydrochloride salts of R-(+)- and S-(⫺)- etomidate were synthesised at Organon Laboratories (Newhouse, UK). Isomer ratios determined by chiral hplc were 98.83:0.71 for R-(+)-etomidate and 97.45:1.41 for S-(⫺)-etomidate. Stock solutions of the etomidate enantiomers were dissolved in double distilled deionised water and diluted into the appropriate medium.
62
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
3. Results 3.1. Hypnotic potency and duration of action in mice Intravenous injection of either of the enantiomers of etomidate induced hypnosis and LRR in mice. The HD50 determined for R-(+)-etomidate (6.0 µmol kg–1 (95% confidence limits = 5.4–6.5)) was approximately 10-fold less than that found for S-(⫺) etomidate (61.3 µmol kg–1 (95% confidence limits = 51.5–71.5). 3.2. Hypnotic potency in tadpoles Utilising a 180 min equilibration period, LRR in tadpoles occurred at HC50 values of 2.3 µM (95% confidence limits = 1.9–2.7) for R-(+)-etomidate and 24.5 µM (95% confidence limits = 20.2–30.7) for the S(⫺)-enantiomer. Thus, the degree of enantioselectivity observed at presumed equilibrium in tadpoles (approximately 10-fold) is identical to that found following administration as a bolus in mice. 3.3. Effects of etomidate enantiomers upon the binding of [35S]TBPS to rat brain membranes Both R-(+)-etomidate and S-(–)-etomidate produced a concentration-dependent and monophasic inhibition of [35]TBPS (1 nM) binding to rat brain membranes. The IC50 for R-(+)-etomidate (2.5 ± 0.5 µM, n = 4) was approximately 13-fold lower than that found for the S(⫺)-isomer (33 ± 2.6 µM, n = 4). Maximally effective concentrations of the etomidate enantiomers reduced the binding of [35S]TBPS to a level approximating to that of non-specific binding. The enantioselectivity and the IC50 values obtained for the etomidate enantiomers in this binding assay are remarkably similar to those determined for LLR in tadpoles. 3.4. Effects of etomidate enantiomers upon the binding of [35S]TBPS to GABAA receptors assembled from b3 subunits The expression of homomeric assemblies of the β3subunit is sufficient to constitute [35S]TBPS binding sites that are allosterically regulated by a number of intravenous anaesthetics, including R-(+)-etomidate (Slany et al., 1995; Davies et al., 1997). In the present study, approximately 95% of [35S]TBPS (2 nM) bound in a specific manner to crude membrane homogenates prepared from HEK 293 cells transfected with cDNA encoding the β3-subunit. In agreement with previous studies (Slany et al., 1995; Davies et al., 1997), such binding was insensitive to GABA (10⫺7–10⫺4 M, n = 3), but was completely inhibited in a concentration-dependent manner by R-(+)-etomidate (IC50 = 0.7 ± 0.05 µM, n = 3). The potency of the S-(⫺)-enantiomer was 10-fold less
(IC50 = 7.0 ± 0.9 µM, n = 3). Although affinity values for the enantiomers of etomidate cannot be inferred from measurements of their effects upon [35S]TBPS binding, it would appear that a major element of the etomidate binding site can be constituted by homo-oligomeric assemblies of the β3 subunit. This is because neither the potency, nor enantioselectivity, of the anaesthetic appear to be greatly different from that found for the isoforms of GABAA receptor expressed in rat brain (see above). 3.5. Enantioselectivity of action of etomidate at recombinant GABAA receptors In the present study, we evaluated the actions of the R-(+)- and S-(⫺)-enantiomers of etomidate at receptors assembled from α1β3γ2L, or α6β3γ2L subunits, the latter having been previously documented to be particularly sensitive to allosteric modulation by R-(+)-etomidate (Belelli et al., 1997). At both receptor isoforms, R-(+)etomidate (0.3–100 µM) potentiated the current evoked by GABA at EC10 with a bell-shaped concentration dependency (Fig. 1A). However, both the potency (expressed as the EC50 values derived from the ascending portion of the concentration–effect curve) and maximal effect of etomidate were greater for the α6 subunitcontaining receptor isoform (Table 1). High concentrations (ⱖ100 µM) of R-(+)-etomidate were associated with a reduced potentiation (Fig. 1A) and concomitantly, the development of an inwardly directed current that was blocked by picrotoxin (data not shown). Such direct activation of the GABAA receptor, which was concentration dependent, was most prominent for the α6 subunit-containing isoform (Fig. 1B, Table 1). S-(⫺)-etomidate exerted qualitatively similar actions to those described above at the α1β3γ2L or α6β3γ2L receptor isoforms, but was in all respects weaker than the R-(+)-enantiomer (Table 1). Thus, the potency of S-(⫺)-etomidate to potentiate GABA at the α1β3γ2L and α6β3γ2L receptors was, respectively, approximately 7.9- and 4.5-fold less than that found for the R-(+)-enantiomer (Fig. 1A, Table 1). Similarly, the maximal enhancement evoked by S(⫺)-etomidate was less than that produced by R-(+)-etomidate (by a factor of approximately 1.8 and 2.9 at α1β3γ2L and α6β3γ2L receptors, respectively, Table 1). The combination of these effects results in an enantioselectivity for etomidate that varies with the concentration of the anaesthetic if a simple comparison of the molar concentrations required to produce a comparable degree of potentiation is performed. Direct activation of the α1β3γ2L and α6β3γ2L receptors isoforms by S-(⫺)-etomidate was also far less pronounced than observed for the R-(+)-enantiomer (Fig. 1B). The limited solubility of the etomidate enantiomers precluded the determination of unequivocal maxima and hence of EC50 values, so it is not possible to state whether the isomers differ in potency, maximal effect, or both. The above results are
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
63
qualitatively consistent with those of Tomlin et al. (1998), who employed a relatively insensitive α1β1γ2L GABAA receptor isoform (Hill-Venning et al., 1997) that nonetheless displayed a preference for R-(+)-etomidate. 3.6. Effects of etomidate enantiomers upon membrane currents mediated by GABAA receptors assembled from b3 subunits Homo-oligomeric assemblies of β3-subunits expressed in Xenopus oocytes (Wooltorton et al., 1997), or HEK293 cells (Davies et al., 1997), are insensitive to GABA, but capable of mediating membrane currents in response to general anaesthetics that include propofol and pentobarbitone. In the present study, the enantiomers of etomidate evoked, in a concentration-dependent and reversible manner, inward current responses from Xenopus oocytes injected with cDNA encoding the human β3 GABAA receptor subunit (Fig. 1C). When expressed as a percentage of the peak current response produced by a maximally effective concentration of pentobarbitone (1 mM) in the same population of oocytes, the maximal inward current response evoked by R-(+)-etomidate (308 ± 77%; n = 4) was approximately threefold greater than that produced by the S-(-)-enantiomer (99 ± 7%; n = 3). 3.7. Selectivity of action of R-(+)-etomidate
Fig. 1. Modulation and activation of human recombinant GABAA receptor isoforms expressed in Xenopus laevis oocytes by the enantiomers of etomidate. (A) Concentration-dependent potentiation by R(+)-etomidate (circles) or S-(⫺)-etomidate (triangles) of membrane currents evoked by GABA at EC10 acting at α6β3γ2L (solid symbols) or α1β3γ2L (open symbols) GABAA receptor isoforms. Each data point is the mean ± S.E.M. of the results obtained from four to six oocytes. (B) Concentration-dependent activation by R-(+)-etomidate (circles) or S-(⫺)-etomidate (triangles) of α6β3γ2L (solid symbols) or α1β3γ2L (open symbols) GABAA receptor isoforms. Each data point is the mean ± S.E.M. of the results obtained from three to four oocytes. See Table 1 for fitted parameters. (C) Concentration-dependent activation by R(+)-etomidate (쎲) or S-(⫺)-etomidate (䊊) of the β3 GABAA receptor. GABA has no agonist activity at this receptor thus current responses to the etomidate enantiomers are alternatively expressed as a percentage of the maximal current response elicited by a saturating concentration (1 mM) of pentobarbitone. Each data point is the mean ± S.E.M. of the results obtained from three to four oocytes.
The selectivity of action of R-(+)-etomidate was examined across additional representatives of transmitter-gated ion channel families including nicotinic acetylcholine, 5-HT3 and ionotropic glutamate receptors, all of which were expressed in Xenopus laevis oocytes and examined under comparable conditions. It has previously been demonstrated that supra-anaesthetic concentrations of R-(+)-etomidate antagonise currents mediated by nicotinic acetylcholine receptors assembled from α4 and β4 subunits (Flood and Krasowski, 2000). Whether this particular subunit combination occurs within the mammalian CNS is uncertain, but the existence of hetero-oligomers composed of α4 and β2 subunits, or homo-pentamers of the α7 subunit is well documented (e.g. Cordero-Erausquin et al., 2000). Currents evoked by nicotine at EC50 from oocytes expressing the chick α7 subunit were antagonised by R-(+)-etomidate (1–200 µM) in a concentration-dependent manner with an IC50 of 25.3 ± 7.4 µM; n = 3; Table 1). At the rat α4β2 subunit combination, nicotine evoked currents were also blocked by R-(+)-etomidate, although less potently (IC50 = 48.8 ± 5.6 µM; n = 4; Table 1). The function of 5-HT3 receptors native to mouse N1E115 neuroblastoma cells is suppressed by high concentrations of etomidate (Barann et al., 1993). Here, we confirm a very weak blocking action of R-(+)-etomidate (1– 300 µM) upon currents evoked by 5-HT at EC50 acting
64
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
Table 1 Summary of the selectivity and enantioselectivity of action of etomidate. All receptor constructs were expressed in Xenopus laevis oocytes and the EC50 values for potentiation and direct activation of GABAA receptor isoforms, or IC50 values for inhibition of all other receptor types (except glycine), were determined under voltage-clamp at a holding potential of ⫺60 mV. Maximal potentiation, or activation, by etomidate enantiomers was quantified by expressing the modulated response as a percentage of the maximal response evoked by GABA (or glycine) (denoted Emax). For further details of experimental methods, see Methods. Abbreviations and symbols: ND—not determined due to the absence of a clear maximal effect or limited action Enantiomer
Receptor construct
Concentration range examined (µM)
EC50 or IC50∗ (µM)
Emax
R-(+)-etomidate R-(+)-etomidate R-(+)-etomidate R-(+)-etomidate
Chick nicotinic α7 Rat nicotinic α4β2 Human 5-HT3A Human glycine α1a Human GABA α1β3γ2L
S-(⫺)-etomidate
Human GABA α1β3γ2L
R-(+)-etomidate
Human GABA α6β3γ2L
S-(⫺)-etomidate
Human GABA α6β3γ2L
1–200 1–300 1–1000 1–300 Potentiation 0.1–100 Direct 3–300 Potentiation 1–300 Direct 3–300 Potentiation 0.03–100 Direct 1.0–300 Potentiation 0.1–300 Direct 1.0–300
25.3 ± 7.4∗ 48.8 ± 5.6∗ 147.7 ± 9.3∗ ND 4.00 ± 0.25 ND 31.6 ± 2.9 ND 2.96 ± 0.5 ND 13.2 ± 2.9 ND
– – – 29.2 ± 4.0 80.7 ± 4.8 苲28 (300 µM) 44.3 ± 0.2 苲13 (300 µM) 116.0 ± 6.9 苲86 (300 µM) 40.1 ± 1.7 苲18 (300 µM)
a
Data from Pistis et al. (1997).
upon homo-oligomeric receptors assembled from the human 5-HT3A receptor subunit (IC50 = 147 ± 9 µM; n = 4; Table 1). R-(+)-etomidate (100 µM) exerted little, or no, effect upon inward currents mediated by ionotropic glutamate receptors. Acting upon hetero-oligomeric AMPA receptors assembled from rat GluR1o and GluR2o subunits, non-desensitising currents evoked by kainate at EC50 were maintained at 96.1 ± 5% of control in the presence of etomidate (n = 3). Receptors incorporating the GluR2 subunit are thought to represent the majority of AMPA receptors expressed by principal neurones within the CNS (Seeburg et al., 1998). At rat recombinant NMDA receptors containing NR1a and NR2A subunits, an example of a receptor assembly likely to be targeted to synapses within the mature central nervous system (CullCandy et al., 2001), the anaesthetic, even at this high concentration, reduced responses evoked by NMDA at EC50 to only 79.3 ± 3.2% of control (n = 3). R-(+)-etomidate had no detectable direct effect at any of the nicotinic acetylcholine, 5-HT3, or ionotropic glutamate receptor isoforms examined.
and 10–90% rise time were 87.2 ± 7.4 pA and 383 ± 12 µs, respectively. Such events decayed biexponentially with a fast time constant (tf) of 6.37 ± 0.23 ms and a slow time constant (ts) of 42.6 ± 2.3 ms yielding a weighted time constant (tw) of 24.4 ± 1.5 ms. R-(+)etomidate (3 µM) had no effect upon rise time, but tended to increase peak mIPSC amplitude by 21.1 ± 6% and dramatically prolonged ts to 101.5 ± 10.4 ms whilst having little effect upon tf (5.1 ± 0.4 ms; n = 4; Fig. 2). tw increased to 58.0 ± 5.7 ms in the presence of R-(+)-etomidate. The actions of S-(⫺)-etomidate (3 µM) upon the mIPSC were qualitatively similar to those of the R-(+)-enantiomer, but were far less pronounced (Fig. 2). Peak current amplitude was increased by only 6.8 ± 1.6% (n = 7) along with a modest prolongation of ts (52.8 ± 4.0 ms) and tw (34.3 ± 2.0 ms; n = 7; Fig. 2). S-(⫺)-etomidate had no effect upon the rise time, or tf (6.96 ± 0.73 ms), of the mIPSC.
3.8. Effect of etomidate enantiomers on mIPSCs recorded from rat reticulothalamic (nRT) neurones
Experiments were performed on both embryonic (stage 37/8) and larval (stage 42) animals (Fig. 3A). Unless stated otherwise, the results presented are pooled from both stages. The effects of the etomidate enantiomers on ventral root motor bursts were examined initially. S-(⫺)-etomidate (2.5⫺20 µM) had no significant effects on any of the measured parameters of fictive swimming, even 20 min after application. However, R(+)-etomidate had dramatic effects on swimming. Fig. 3B illustrates an example of the increase in cycle period
Miniature IPSCs (mIPSCs) recorded from rat nRT neurones in the presence of TTX (0.5 µM) and kynurenic acid (2.0 mM) with symmetrical [Cl–] at a holding potential ⫺60 mV were abolished by bicuculline (10–30 µM), confirming their mediation by GABAA receptors (data not shown). In a sample of 17 neurones, prior to incubation with R-(+)-etomidate, the mean mIPSC amplitude
3.9. Effect of etomidate enantiomers on fictive swimming activity in Xenopus laevis embryos and larvae
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
Fig. 2. Modulation of mIPSCs recorded from rat thalamic reticular nucleus neurones by the enantiomers of etomidate. (A) Examples of continuous records of mIPSCs recorded from a rat thalamic reticular nucleus neurone in the absence and subsequently in the presence of R-(+)-etomidate (3 µM). Synaptic currents were recorded under voltage-clamp at a holding potential of –60 mV. (B) Superimposition of averaged mIPSCs recorded in the absence and presence of R-(+)- (left) or S-(⫺)-etomidate (right). Current traces are the ensemble average of a minimum of 50 mIPSCs recorded under each experimental condition. Note the marked prolongation of the decay phase of the mIPSC in the presence of R-(+)-etomidate which is due to an increase in the both slow decay time constant, ts, and the weighted decay time constant, tw. Similar, though less pronounced and not statistically significant trends (see C) are evident for S-(⫺)-etomidate. (C) Mean percentage change in mIPSC kinetics in the presence of R-(+)- or S-(⫺) etomidate (both 3 µM). Statistically significant increases (∗, P ⬍ 0.001) in ts and tw are evident for R-(+)- but not S-(⫺)-etomidate. Other parameters were not significantly different from control. The data are the mean of recordings from four and seven neurones for R-(+)- and S(⫺)-etomidate, respectively.
observed in the presence of R-(+)-etomidate (15 µM) and its reversal by co-applied bicuculline (40 µM). Cycle periods were significantly increased, in a concentrationdependent manner, after application of R-(+)-etomidate (2.5⫺20 µM) in all 14 animals tested (P ⬍ 0.05). On average, cycle periods (Fig. 3C) increased from a control value of 40.6 ± 0.5 ms to 52.4 ± 0.7, 64.2 ± 0.9, 79.0 ± 0.7 and 92.9 ± 1.2 ms in the presence of R-(+)-etomidate at 2.5, 5.0,10 and 20 µM, respectively. Subsequent addition of bicuculline to nine animals significantly reduced cycle periods (P ⬍ 0.05; 76.5 ± 1.8 ms; Fig. 3C). Episode durations were significantly reduced in the
65
Fig. 3. The influence of the enantiomers of etomidate upon fictive swimming in immobilised Xenopus laevis tadpoles. (A) Schematic illustration of the preparation (shown for a stage 42 larva). Extracellular recordings of ventral root activity were made via suction electrodes placed over the myotomal clefts following removal of the flank skin. Intracellular recordings, from presumed motor neurons, were made using sharp microelectrodes after removal of the overlying rostral muscle blocks. Swimming activity was elicited by applying a brief (0.5– 1 ms) current pulse to the tail skin. (B) Excerpts of ventral root activity in the absence and presence of R-(+)-etomidate (R-(+)), and R-(+)etomidate co-applied with bicuculline (Bicuc.). R-(+)-etomidate reduced the frequency of motor bursts and bicuculline reversed this effect. (C) Graphical depiction of the influence of S-(⫺)- and R-(+)etomidate upon mean cycle period, recorded from 14 preparations. The S-(⫺) enantiomer had no significant effect (ns) on swimming, whilst cycle periods between motor bursts increased significantly (∗) in comparison to control with incremental doses of R-(+)-etomidate. In the nine animals tested, the effect of etomidate (20 µM) was partially reversed with bicuculline. (D) An example of the reduction in the duration of fictive swimming activity in the presence of R-(+)-etomidate. Bicuculline partially reversed this effect. (E) Mean data obtained from seven preparations illustrating the depression of burst duration by R(+)-etomidate and its reversal by co-applied bicuculline.
presence of the R-(+)-enantiomer (2.5⫺20 µM) in 12/14 animals tested (P ⬍ 0.05). In the example of Fig. 3D, average episode durations fell by 77.9 ± 4.0% in the presence of 15 µM R-(+)-etomidate. Subsequent addition of bicuculline to nine animals significantly reversed this
66
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
reduction in seven animals (P ⬍ 0.05). R-(+)-etomidate reduced the duration of ventral root motor bursts in larval preparations (n = 7). Average burst durations fell significantly from 16.2 ± 0.4 to 7.9 ± 0.2 ms after addition of R-(+)-etomidate (10 µM; P ⬍ 0.01; Fig. 3E). Subsequent addition of bicuculline reversed these effects (18.5 ± 0.5 ms; n = 7; P ⬍ 0.001). In embryonic preparations, ventral roots typically consist of a single compound spike per cycle, the duration of which was unaffected by etomidate. Intracellular recordings from presumed motoneurons in the ventral spinal cord using KCl-filled electrodes showed that R-(+)-etomidate had no obvious effects on either the excitatory or mid-cycle inhibitory (glycinergic; asterisked) components of the synaptic drive during swimming (Fig. 4). Swimming episodes in embryos can be prematurely terminated following stimulation of the rostral cement gland, which activates mid-hindbrain reticulospinal (mhr) GABAergic neurons and terminates swimming (Boothby and Roberts, 1992). Thus, swimming terminates prematurely with a barrage of GABAergic IPSPs following a brief electrical stimulus to the cement gland (arrowed in Fig. 4). R-(+)-etomidate (10⫺15 µM; n = 5) enhanced the GABAergic barrage after cement gland stimulation, causing individual GABAergic potentials to summate so that swimming terminated with a single, presumably compound, IPSP following which the membrane potential took more than 1 s to return to rest, about twice the duration in control.
In three of these animals, to which bicuculline (40 µM) was subsequently added, stimulation of the cement gland no longer terminated swimming prematurely. During quiescent periods between episodes of swimming activity, two groups of spontaneous inhibitory potentials are revealed, which represent quantal release of GABA and glycine onto the recorded motor neuron. The two types of potential are readily distinguished on the basis of their duration and their pharmacological sensitivity (Reith and Sillar, 1997). The glycinergic population have durations of 20–80 ms and are strychninesensitive, whilst the second bicuculline-sensitive group have durations of 90–200 ms and are GABAergic. In the presence of R-(+)-etomidate (10–15 µM; n = 8), the duration of GABAergic potentials increased from 90–200 ms in control to 450 ms–1 s, while their frequency and amplitude remained unaffected. However, there was no change in the duration, amplitude or frequency of the glycinergic potentials (Fig. 5A). Subsequent addition of bicuculline (n = 8) abolished all of these GABAergic potentials (Fig. 5A). In the presence of tetrodotoxin (1 µM), which blocks spike-evoked transmitter release from pre-synaptic terminals, R-(+)-etomidate prolonged the duration of the GABAergic IPSPs (n = 3, Fig. 5B). There was no concomitant increase in the frequency or amplitude of either the GABAergic or glycinergic potentials, nor was there any detectable change in membrane conductance or resting potential (data not shown). Thus, in the relatively simple, but intact, central nervous system of the Xenopus laevis tadpole, etomidate acts in an enantioselective manner, and probably exclusively postsynaptically, to prolong the effects of GABA at the GABA receptor with no concomitant effect upon glycinergic inhibition.
4. Discussion
Fig. 4. Etomidate enhances the embryonic cement gland response. (A) Intracellular recordings (using KCl-filled electrodes allowing diffusion of KCl into the cell, reversing and enhancing the mid-cycle glycinergic chloride dependent IPSPs) of motor neuron (mn) and ventral root (vr12) activity illustrating that R-(+)-etomidate had no effect on either the excitatory or inhibitory (glycinergic, ∗) components of the synaptic drive underlying swimming. (B) Embryonic swim episodes terminate prematurely with a barrage of GABAergic potentials when a brief stimulus (↓) is applied to the cement gland (see Fig. 3A). R-(+)-etomidate caused individual GABAergic potentials to summate so that swimming terminates with a single potential of greater than 1 s duration. Recordings illustrated in panels A and B were made from the same motor neurone.
The HD50 for R-(+)-etomidate (6.0 µmol kg–1) in mice determined here corresponds well with the value of 4.2 µmol kg–1 calculated from the data of Janssen et al. (1975) and approximates to the median effective dose (2 µmol kg–1) in man (Doenicke and Ostwald, 1997). It has frequently been assumed that S-(⫺)-etomidate is not an anaesthetic (e.g. Heykants et al., 1975; Doenicke and Ostwald, 1997). However, the present results and those of Tomlin et al. (1998) indicate a clear enantioselectivity in favour of the R-(-)-isomer yet an appreciable anaesthetic activity of the S-(⫺)-isomer. Indeed, the HC50 value determined for S-(⫺)-etomidate in tadpoles (24 µM) and the HD50 calculated in mice (61.3 µmol kg–1) indicate a potency comparable to that of racemic thiopentone in the same model systems (Lee-Son et al., 1975; Glen, 1980). The degree of enantioselectivity observed for etomidate in tadpoles (approximately 10fold) in this study is slightly less than that
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
67
Fig. 5. Etomidate enhances GABAergic inhibition in the tadpole motor-generating network. (A) Upper panel. During quiescent periods between episodes of swimming, two types of spontaneous inhibitory potentials are recorded, representing quantal release of GABA and glycine onto motor neurons. These potentials are distinguished on the basis of their duration and their pharmacological sensitivity; the strychnine-sensitive glycinergic population having durations of 20–80 ms and the second bicuculline-sensitive GABAergic group (example asterisked) having durations of 90–200 ms (see text for further details). In the presence of R-(+)-etomidate (10 µM, middle panel), the duration of the GABAergic (example asterisked), but not the glycinergic, potentials is increased to 400 ms–1s, whilst their amplitude and frequency remains unchanged. The addition of bicuculline (40 µM) in the continued presence of etomidate (10 µM, lower panel) abolishes all GABAergic potentials and reveals the shorter duration glycinergic potentials in isolation. (B) Plots of the duration of synaptic potentials recorded in the presence of TTX (1 µM) in control (upper panel), the presence of R-(+)-etomidate (10 µM, middle panel) and etomidate (10 µM) co-applied with bicuculline (40 µM, lower panel). Note that etomidate enhances the duration of the bicuclline-sensitive GABAergic potentials whilst glycinergic potentials are unaffected.
68
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
(approximately 17-fold) reported by Tomlin et al. (1998). This discrepancy might be due to methodological differences in the assessment of the anaesthetic endpoint. The present study appears to be the first to quantify the enantioselectivity of etomidate in a mammal. In the original pharmacodynamic and pharmacokinetic studies of the enantiomers of etomidate in rats (Heykants et al., 1975), S-(⫺)-etomidate administered i.v. at 9 mg kg–1 achieved a maximal brain concentration of approximately 15 µg g⫺1 tissue, but did not produce hypnosis. The minimal brain concentration of R-(+)-etomidate required for sleep was about 1.5 µg g⫺1. Thus, if the 10fold enantioselectivity reported here for mice (and tadpoles) can be extrapolated to the rat, the dosing regimen of Heykants et al. (1975) would probably not have detected a robust effect of S-(⫺)-etomidate. Importantly, however, this early work did establish that the enantiomers of etomidate achieve an equal concentration within the brain following i.v. administration. Thus, potential differences in CNS penetration are unlikely to contribute to the observed enantioselectivity. The HC50 for R-(+)-etomidate in tadpoles (2.3 µM) provides an indication of the concentration likely to be relevant to clinical anaesthesia. At such a value, R-(+)etomidate displaced the binding of [35S]TBPS from rat brain homogenates and potentiated membrane currents evoked by GABA at the two human receptor GABAA receptor isoforms examined (α1β3γ2L and α6β3γ2L). S(⫺)-etomidate, at concentrations close to HC50 in tadpoles, was qualitatively similar. Concentrations of R(+)-, or S-(⫺)-etomidate, substantially higher than HC50 were required to directly activate these receptor isoforms suggesting that the GABA-mimetic effect is unlikely to be important in producing anaesthesia. Similarly, concentrations of R-(+)-etomidate that directly activate hetero-oligomeric receptors assembled from α2-, or α3-, in combination with β1/2- and γ2L- subunits are much higher than those likely to be encountered clinically (Hill-Venning et al., 1997). However, such observations do not exclude the possibility that GABAA receptor isoforms of different composition might be activated by clinically relevant concentrations of etomidate. In contrast to strongly potentiating GABA, R-(+)-etomidate, at HC50, did not affect responses mediated by the nicotinic, 5-HT3, or ionotropic glutamate receptor isoforms examined. Even the most sensitive of those, the chick α7 nicotinic receptor, required a 10-fold higher concentration of the anaesthetic for half-maximal inhibition. Thus, of the receptors examined, only the GABAA receptor satisfies the criterion of appropriate sensitivity to anaesthetic to be considered a plausible effector of etomidate action. Importantly, there is a remarkable quantitative similarity in the enantioselectivity of etomidate as an anaesthetic in tadpoles and mice and as a modulator of GABAA receptor function in binding and electrophysiological assays.
The interaction of R-(+)-etomidate with GABAA receptors is dependent upon the species of β subunit isoform within the hetero-oligomeric complex (Hill-Venning et al., 1997). For both the GABA-modulatory and GABA-mimetic effects of R-(+)-etomidate, activity is enhanced by the inclusion of β2-, or β3-, vs. β1-subunits within the receptor (Belelli et al., 1997; Hill-Venning et al., 1997). We have attributed this difference to the identity of a single amino acid residue located towards the extracellular aspect of the second transmembrane domain of the β subunits (asparagine in β2/3, serine in β1; Belelli et al., 1997). In addition, as demonstrated in this and previous studies (Hill-Venning et al., 1997), expression of the α6-, vs. α1-, subunit in combination with β3- and γ2L- subunits within the hetero-oligomeric complex is associated with an increased potency and maximal effect of R-(+)-etomidate as a positive modulator of GABA and as a GABA-mimetic. The influence of the α-subunit extends to the potency of S-(-)-etomidate as a potentiator, but the maximal effect of the compound and its GABA-mimetic effect were similar at receptors containing either α1- or α6- subunits. As reported by Zhang et al. (1997) and Huntsman and Huguenard (2000), GABAA receptor mediated mIPSCs recorded from rat nRT neurones decayed with biexponential kinetics. The mean amplitude and tw of the mIPSCs, at a temperature of 35 °C were comparable to previously published estimates (Huntsman and Huguenard, 2000). R-(+)-etomidate, at a concentration (3 µM) approximating to HC50 in tadpoles, increased the peak amplitude and dramatically prolonged the duration of mIPSCs. In agreement with previous observations made upon hippocampal neurones in culture, the prolongation was due exclusively to an increase in ts (Yang and Uchida, 1996). By contrast, S-(⫺)-etomidate, at an equivalent concentration, had a minimal effect upon mIPSC parameters. These results directly confirm the findings of Ashton and Wauquier (1982) who inferred, from extracellular recordings of paired pulse inhibition at the Schaffer collateral CA1 pyramidal cell synapse of the guinea-pig hippocampus, an enantioselective action of etomidate upon GABAergic neurotransmission. The ability of R-(+)-etomidate to dramatically prolong mIPSCs in nRT neurones is consistent with the predominant expression of β3-, together with α3-, β1- and γ2subunits in this brain region (Pirker et al., 2000). The intact, paralysed Xenopus tadpole preparation has been extensively utilised as a simple model of the neural control of behaviour at the cellular and synaptic levels (see Roberts et al., 1998). In addition, modulation of synaptic transmission within the spinal cord, producing depression of the sensory processing of noxious signals and or decreased motoneurone excitability, is regarded to be essential to the immobilising effect of general anaesthetics (Collins et al., 1995) The present results indicate that etomidate depresses fictive swimming in an
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
enantioselective manner. Whereas S-(⫺)-etomidate at concentrations ⱕ20 µM was ineffective, the R-(+)enantiomer had a general inhibitory action, reducing both swimming frequency and swim episode duration. Importantly, a significant effect of R-(+)-etomidate could be observed at a concentration (2.5 µM) approximating to the HC50 for Xenopus tadpoles (albeit of a different developmental stage), although a more pronounced action was apparent only at higher concentrations of the anaesthetic. We cannot exclude the possibility that comparisons across the two tadpole paradigms may be complicated by differences in the developmental stage of the animals employed. Mechanistically, changes in the frequency of swimming are often effected by alterations in the relative contribution of inhibition to the synaptic drive. The only inhibitory component phase-linked to each cycle of swimming is glycinergic and appears insensitive to R-(+)-etomidate. However, it is probable that prolongation of sporadic GABAergic IPSPs (Reith and Sillar, 1997) by R-(+)-etomidate allowed their summation across an episode of swimming, sufficient to reduce swimming frequency. Such an enhancement of inhibition across the course of a whole episode could also be responsible for the observed shortening in the duration of swimming. Similarly, R-(+)-etomidate enhanced the GABAergic barrage associated with the premature termination of swimming following stimulation of the cement gland. The ability of bicuculline to occlude all of the effects of R-(+)-etomidate implicates the GABAA receptor as its site of action within this system. Recordings from motoneurons during the quiescent periods between episodes of swimming revealed that R(+)-etomidate selectively enhanced GABAergic, but not glycinergic, inhibitory potentials confirming the selectivity of R-(+)-etomidate observed at the mammalian orthologues of these receptors expressed in Xenopus oocytes (Pistis et al., 1997). Furthermore, when neurones were synaptically isolated using TTX, the effect upon GABAergic IPSPs persisted, indicating a direct postsynaptic site of action. Since there was no detectable change in the frequency of spontaneous GABA IPSPs, or their amplitude distributions, it is unlikely that there is an additional presynaptic action of R-(+)-etomidate. In conclusion, our evidence indicates that R-(+)-etomidate, at clinically relevant concentrations, is a highly selective positive allosteric modulator of GABAA receptor activity in a variety of experimental systems and that this activity displays an enantioselectivity that is quantitatively similar to that observed in animal models of the obtunding effect of anaesthetics in man. The selective modulation of GABAA receptor activity in Xenopus tadpoles by R-(+)-etomidate is sufficient to impinge significantly upon motor behaviour. It is not unreasonable to suggest that a similar effect upon inhibitory synaptic transmission within the CNS of higher vertebrates, as
69
exemplified by the results we obtained with R-(+)-etomidate upon reticulothalamic neurones, underlies at least some aspects of the anaesthetic activity of this compound. Indeed, immediately prior to the submission of this work for publication, Jurd et al. (2003) obtained the first direct evidence that the anaesthetic activity of etomidate does indeed involve specific GABAA receptor populations. In ‘knock-in’ mice expressing GABAA receptor β3-subunits rendered insensitive to etomidate by substitution of the second transmembrane asparagine 265 residue by methionine (Pistis et al., 1999), suppression of the nocifensive reflex by etomidate was abolished. The duration of etomidate-induced LRR was additionally greatly reduced in the ‘knock-in’ animals. As would be predicted from our previous in vitro studies (Pistis et al., 1999), the effects of the mutation extended to the intravenous anaesthetic propofol, but were less pronounced, or absent, in the case of the volatile agents, halothane and enflurane. Together with a complete lack of effect of the mutation upon steroid anaesthesia, which is also consistent with in vitro data (Pistis et al., 1999), the work of Jurd et al. (2003) demonstrates that specific GABAA receptor populations contribute to the spectrum of anaesthetic actions in an agent-dependent manner. Moreover, it has also been shown that the modulation of GABAA receptors with discrete brain regions can have pronounced effects upon behaviour. The anaesthetics propofol, or pentobarbitone, when injected locally into the hypothalamic tuberomammilary nucleus (TMN), produce marked sedation in rodents (Nelson et al., 2002). Direct activation of GABAA receptor populations within the TMN by the locally applied muscimol elicited LRR in a manner that could be prevented by pre-treatment with the competitive antagonist, gabazine (Nelson et al., 2002). Such recent work, together with the present findings, leaves little doubt that the GABAA receptor is crucial to the anaesthetic activity of etomidate.
Acknowledgements DB is an MRC Senior Research Fellow. KTS is supported by the Wellcome Trust and SDM by a BBSRC studentship. We thank Tenovus Tayside for additional financial support (DB).
References Anderson, A., Belelli, D., Bennett, D.J., Buchanan, K.I., Casula, A., Cooke, A., Feilden, H., Gemmell, D.K., Hamilton, N.M., Hutchinson, E.J., Lambert, J.J., Maidment, M.S., McGuire, R., McPhail, P., Miller, S., Muntoni, A., Peters, J.A., Sansbury, F.H., Stevenson, D., Sundaram, H., 2001. α-Amino acid phenolic ester derivatives: novel water-soluble general anesthetic agents which allosterically modulate GABAA receptors. Journal of Medicinal Chemistry 44, 3582–3591.
70
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
Ashton, D., Wauquier, A., 1982. Modulation of a GABA-ergic inhibitory circuit in the in vitro hippocampus by etomidate isomers. Anesthesia and Analgesia 64, 975–980. Barann, M., Gothert, M., Fink, K., Bonisch, H., 1993. Inhibition by anaesthetics of 14C-guanidinium flux through the voltage-gated sodium channel and the cation channel of the 5-HT3 receptor of N1E-115 neuroblastoma cells. Naunyn Schmiedeberg’s Archives of Pharmacology 347, 125–132. Belelli, D., Balcarek, J.M., Hope, A.G., Peters, J.A., Lambert, J.J., Blackburn, T.P., 1995. Cloning and functional expression of a human 5-hydroxytryptamine type 3AS receptor subunit. Molecular Pharmacology 48, 1054–1062. Belelli, D., Callachan, H., Hill-Venning, C., Peters, J.A., Lambert, J.J., 1996. Interaction of positive allosteric modulators with human and Drosophila recombinant GABA receptors expressed in Xenopus laevis oocytes. British Journal of Pharmacology 118, 563–576. Belelli, D., Lambert, J.J., Peters, J.A., Wafford, K., Whiting, P.J., 1997. The interaction of the general anesthetic etomidate with the γ-aminobutyric acid type A receptor is influenced by a single amino acid. Proceedings of the National Academy of Sciences USA 94, 11031–11036. Belelli, D., Pistis, M., Peters, J.A., Lambert, J.J., 1999. General anaesthetic action at transmitter-gated inhibitory amino acid receptors. Trends in Pharmacological Sciences 20, 496–502. Benke, D., Fritschy, J.M., Trzeciak, A., Bannwarth, W., Mohler, H., 1994. Distribution, prevalence, and drug binding profile of γ-aminobutyric acid type A receptor subtypes differing in the β-subunit variant. Journal of Biological Chemistry 269, 27100–27107. Boothby, K.M., Roberts, A., 1992. The stopping response of Xenopus laevis embryos: behaviour development and physiology. Journal of Comparative Physiology [A] 170, 171–180. Chen, D., Okayama, H., 1988. Calcium phosphate-mediated gene transfer: a highly efficient transfection system for stably transforming cells with plasmid DNA. Biotechniques 6, 632–638. Clements, J.D., Bekkers, J.M., 1997. Detection of spontaneous synaptic events with an optimally scaled template. Biophysical Journal 73, 220–229. Collins, J.G., Kendig, J.J., Mason, P., 1995. Anesthetic actions within the spinal cord: contributions to the state of general anesthesia. Trends in Neurosciences 18, 549–553. Cordero-Erausquin, M., Marubio, L.M., Klink, R., Changeux, J.-P., 2000. Nicotinic receptor function: new perspectives from knockout mice. Trends in Pharmacological Sciences 21, 211–217. Cox, C.L., Huguenard, J.R., Prince, D.A., 1997. Nucleus reticularis neurons mediate diverse inhibitory effects in the thalamus. Proceedings of the National Academy of Sciences USA 94, 8854–8859. Cull-Candy, S., Brickley, S., Farrant, M., 2001. NMDA receptor subunits: diversity, development and disease. Current Opinion in Neurobiology 11, 327–335. Davies, P.A., Kirkness, E.F., Hales, T.G., 1997. Modulation by general anaesthetics of rat GABAA receptors comprised of α1 β3 and β3 subunits expressed in human embryonic kidney 293 cells. British Journal of Pharmacology 120, 899–909. Doenicke, A., Ostwald, P., 1997. Etomidate. In: White, P.F. (Ed.), Textbook of Anesthesia. Williams and Wilkins, Baltimore, pp. 93–109. Flood, P., Krasowski, M.D., 2000. Intravenous anesthetics differentially modulate ligand-gated ion channels. Anesthesiology 92, 1418–1425. Franks, N.P., Lieb, W.R., 1994. Molecular and cellular mechanisms of general anaesthesia. Nature 367, 607–614. Glen, J.B., 1980. Animal studies on the anaesthetic activity of ICI 35 868. British Journal of Anaesthesia 52, 731–742. Hall, A.C., Lieb, W.R., Franks, N.P., 1994. Stereoselective and nonstereoselective actions of isoflurane on the GABAA receptor. British Journal of Pharmacology 112, 906–910. Hevers, W., Luddens, H., 1998. The diversity of GABAA receptors.
Pharmacological and electrophysiological properties of GABAA channel subtypes. Molecular Neurobiology 18, 35–86. Heykants, J.J.P., Meuldermans, W.E.G., Michiels, I.J.M., Lewi, J.P., Janssen, P.A.J., 1975. Distribution, metabolism and excretion of etomidate, a short-acting hypnotic drug, in the rat. Comparative study of (R)-(+) and (S)-(⫺)-etomidate. Archives Internationales de Pharmacodynamie et de Therapie 216, 113–129. Hill-Venning, C., Peters, J.A., Callachan, H., Lambert, J.J., Gemmell, D.K., Anderson, A., Byford, A., Hamilton, N., Hill, D.R., Marshall, R.J., Campbell, A.C., 1996. The anaesthetic action and modulation of GABAA receptor activity by the novel water-soluble aminosteroid Org 20599. Neuropharmacology 35, 1209–1222. Hill-Venning, C., Belelli, D., Peters, J.A., Lambert, J.J., 1997. Subunitdependent interaction of the general anaesthetic etomidate with the gamma-aminobutyric acid type A receptor. British Journal of Pharmacology 120, 749–756. Huguenard, J.R., Prince, D.A., 1994. Clonazepam suppresses GABABmediated inhibition in thalamic relay neurons through effects in nucleus reticularis. Journal of Neurophysiology 71, 2576–2581. Huntsman, M.M., Huguenard, J.R., 2000. Nucleus-specific differences in GABAA-receptor-mediated inhibition are enhanced during thalamic development. Journal of Neurophysiology 83, 350–358. Janssen, P.A.J., Niemegeers, C.J.E., Marsboom, R.P.H., 1975. Etomidate, a potent non-barbiturate hypnotic. Intravenous etomidate in mice, rats, guinea-pigs rabbits and dogs. Archives Internationales de Pharmacodynamie et de Therapie 214, 92–132. Jones, M.V., Harrison, N.L., 1993. Effects of volatile anesthetics on the kinetics of inhibitory postsynaptic currents in cultured rat hippocampal neurons. Journal of Neurophysiology 70, 1339–1349. Jurd, R., Arras, M., Lambert, S., Drexler, B., Siegwart, R., Crestani, F., Zaugg, M., Vogt, K.E., Lederman, B., Antkowiak, B., Rudolph, U., 2003. General anesthetic actions in vivo attenuated by a point mutation in the GABAA receptor β3 subunit. FASEB Journal 17, 250–252. Krasowski, M.D., Harrison, N.L., 1999. General anaesthetic actions on ligand-gated ion channels. Cellular and Molecular Life Sciences 55, 1278–1303. Lee-Son, S., Waud, B.E., Waud, D.R., 1975. A comparison of the potencies of a series of barbiturates at the neuromuscular junction and on the central nervous system. Journal of Pharmacology and Experimental Therapeutics 195, 251–256. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J., 1951. Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry 193, 265–275. Mohler, H., Fritschy, J.M., Rudolph, U., 2002. A new benzodiazepine pharmacology. Journal of Pharmacology and Experimental Therapeutics 300, 2–8. Nelson, L.E., Guo, T.Z., Lu, J., Saper, C.B., Franks, N.P., Maze, M., 2002. The sedative component of anesthesia is mediated by GABAA receptors in an endogenous sleep pathway. Nature Neuroscience 5, 979–984. Nieuwkoop, P.D., Faber, J., 1956. Normal tables for Xenopus laevis (Daudin). Amsterdam, North Holland. Pirker, S., Schwarzer, C., Wieselthaler, A., Sieghart, W., Sperk, G., 2000. GABAA receptors: immunocytochemical distribution of 13 subunits in the adult rat brain. Neuroscience 101, 815–850. Pistis, M., Belelli, D., Peters, J.A., Lambert, J.J., 1997. The interaction of general anaesthetics with recombinant GABAA and glycine receptors expressed in Xenopus laevis oocytes: a comparative study. British Journal of Pharmacology 122, 1707–1719. Pistis, M., Belelli, D., McGurk, K., Peters, J.A., Lambert, J.J., 1999. Complementary regulation of anaesthetic activation of human α6β3γ2L and Drosophila (RDL) GABA receptors by a single amino acid residue. Journal of Physiology 515, 3–18. Reith, C.A., Sillar, K.T., 1997. Pre- and post-synaptic modulation of spinal GABAergic neurotransmission by the neurosteroid, 5β-pregnan-3α-ol-2O-one. Brain Research 770, 202–212.
D. Belelli et al. / Neuropharmacology 45 (2003) 57–71
Roberts, A., Soffe, S.R., Wolf, E.S., Yoshida, M., Zhao, F.-Y., 1998. Central circuits controlling locomotion in young frog tadpoles. Annals of the New York Academy of Sciences 860, 18–34. Seeburg, P.H., Higuchi, M., Sprengel, R., 1998. RNA editing of brain glutamate receptor channels: mechanism and physiology. Brain Research Reviews 26, 217–229. Slany, A., Zezula, J., Tretter, V., Sieghart, W., 1995. Rat β3 subunits expressed in human embryonic kidney 293 cells form high affinity [35S]t-butylbicyclophosphorothionate binding sites modulated by several allosteric ligands of gamma-aminobutyric acid type A receptors. Molecular Pharmacology 48, 385–391. Tomlin, S.L., Jenkins, A., Lieb, W.R., Franks, N.P., 1998. Stereoselective effects of etomidate optical isomers on gamma-aminobutyric acid type A receptors and animals. Anesthesiology 88, 708–717. Tomlin, S.L., Jenkins, A., Lieb, W.R., Franks, N.P., 1999. Preparation
71
of barbiturate optical isomers and their effects on GABAA receptors. Anesthesiology 90, 1714–1722. Wittmer, L.L., Hu, Y., Kalkbrenner, M., Evers, A.S., Zorumski, C.F., Covey, D.F., 1996. Enantioselectivity of steroid-induced gammaaminobutyric acidA receptor modulation and anesthesia. Molecular Pharmacology 50, 1581–1586. Wooltorton, J.R., Moss, S.J., Smart, T.G., 1997. Pharmacological and physiological characterization of murine homomeric β3 GABAA receptors. European Journal of Neuroscience 9, 2225–2235. Yang, J., Uchida, I., 1996. Mechanisms of etomidate potentiation of GABAA receptor-gated currents in cultured postnatal hippocampal neurons. Neuroscience 73, 69–78. Zhang, S.J., Huguenard, J.R., Prince, D.A., 1997. GABAA receptormediated Cl- currents in rat thalamic reticular and relay neurons. Journal of Neurophysiology 78, 2280–2286.
