- Email: [email protected]
Etomidate potentiation of GABAA receptor gated current depends on the subunit composition
ELSEVIER
NeuroseienceLetters 185 (1995) 203-206
NHHOSCIHC[ IHT[IIS
Etomidate potentiation of GABA A receptor gated current depends on the subunit composition I. U c h i d a a, G. K a m a t c h i b, D. B u r t b, J. Y a n g a,* aDepartment of Anesthesiology and Pain Management, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75235-9068, USA bDepartment of P,~armacologyand Experimental Therapeutics, University of Maryland School of Medicine, Baltimore, MD, USA
Received 17 October1994;revisedversionreceived19 December1994;accepted20 December1994
Abstract
The role of the )'2 subunit in etomidate potentiation of GABAA receptor-gated chloride current was studied by whole cell patch clamp experiments on H293 cells expressing GABAA receptors. The GABAA receptor subunits a l f l l with or without the )'2 subunit expressed well, with an overall peak current of 157 ± 42 pA/pE At a clinically relevant concentration, etomidate potentiates the peak current induced by GABA equally well in receptors with or without the 7'2 subunit. In contrast, the time course of current decay was prolonged only in receptors with the ),2 subunit. This )'2 subunit-dependent prolongation of the current time course was not blocked by the benzodiazepine receptor antagonist flumazenil. These results show that etomidate, an imidazole general anesthetic, interacts with the GABAA receptor in a ?2 subunit-dependent manner. Keywords: GABAA receptor; General anesthetic; Etomidate; H293 cells; Patch clamp
Recent evidence suggests that general anesthetic interaction with the inhibitory neurotransmitter ?-aminobutyric acid (GABA) receptor may be one important pharmacological mechanism of action of these drugs [6,18]. Etomidate, an imidazole intravenous general anesthetic of widespread clinical acceptance, has this ability to interact with GABA A receptors. Both electrophysiological and binding studies indicate that etomidate shows GABAmodulatory and GABA-mimetic properties [3,5,16,19]. Of particular interes! is the observation that the stereospecific, etomida~te-induced increase in [3H]diazepam binding to rat brain is seen in the forebrain crude membrane preparation but not the cerebellar synaptosomes [2]. Given the difference in GABA A receptor subunit expression in different areas of the brain, this raises the possibility of a receptor subunit-dependent action of etomidate on brain GABAA receptors. The presence of the ?2 subunit has previously been shown to confer specific pharmacological and physiological properties to GABA A receptors. The ?2 subunit is essential for benzodiazepine potentiation of GABA-gated current [15], lack of sensitivity to externally applied zinc * Correspondingauthor, Tel.; +1 214 6485498; Fax: +1 214 6486799.
[4], and a large 29 pS single channel conductance [1]. Therefore, we investigated the potential role of the ?2 subunit in etomidate potentiation of GABAA receptors expressed in combination with a 1 and fl 1 subunits. The a l , ill, and ?2s murine GABA A receptor subunits [13] were subcloned into the pcDNA I eukaryotic expression vector (Invitrogen) by standard methods and analyzed by restriction enzyme mapping. Transfectionquality plasmids were prepared by alkaline lysis and CsCI gradient purification. Human embryonic kidney (H293) cells were grown on Matrigel-coated coverslips (Collaborative Research) in DMEM (Gibco no. 320-1960) supplemented with 10% fetal calf serum, 400/tg of glutamine, 50 U/ml penicillin, and 50/zg/ml streptomycin. One microgram each of the desired receptor subunitcontaining plasmids and an expression plasmid containing the SV40 large T antigen (kindly provided by Dr. Paulo Kofuji) were added to each 35-mm tissue culture dish containing the H293 cells. Co-expression of the large T antigen allows episomal replication of the subunitcontaining plasmid and greater protein expression. Transfection was accomplished by the calcium phosphate precipitation method following the manufacturer's protocol (Stratagene, CA), except that the CO 2 concentration in the
0304-3940/95/$09.50 © 1995 ElsevierScienceIrelandLtd. All fightsreserved SSD! 0304-3940(95)11263-S
204
L Uchida et al. I Neuroscience Letters 185 (1995) 203-206
incubator was increased to 10% to acidify the media to pH 7.2. Control transfection experiments with a fl-galactosidase reporter gene following the above protocol resuited in a transfection efficiency approaching 40%. Electrophysiological recordings from randomly selected cells 48-72 h after transfection with the GABA subunits confirmed the high transfection efficiency. For recording purposes, transfected cells growing on a 10-mm glass coverslip were placed in a recording chamber which was continuously perfused with an external solution (in mM: 140 NaC1, 4 KCI, 1 CaC12 , 1 MgCI2, 10 HEPES, 20 dextrose, pH 7.4), and cells were viewed under Hoffman optics (Nikon TMS). Cells growing in isolation were randomly selected, and a whole cell recording mode was established using standard patch-clamp techniques. The patch pipette contained (in mM: 140 CsC1, 4 NaCI, 10 EGTA, 10 HEPES, pH 7.4). GABA (10/zM) with or without etomidatq (8.2/zM) dissolved in the external solution was applie4 by pressure ejection (20 psi x 100 ms) through puffer electrodes placed approximately 50/zm from the cell body. The etomidate concentration of 8.2/zM (2/zg/ml) was chosen to reflect the serum concentration of this compound observed in patients undergoing general anesthesia [7]. The shift in holding currents upon puffer application of isotonic sucrose was monoexponential with a time constant of 500 ms. Therefore, the current decay measured in the experiments reflects the steadystate current remaining from random closing and re-
opening of channels as the GABA concentration decays with an exponential time course of 500 ms. The average cell input capacitance of the H293 cells, directly read from the capacitance neutralization knob (AI200, Axon Instruments), was 19.2 pF (n = 43). Etomidate in 35% polypropylene glycol (Amidate) was obtained from Abbot Laboratories, and all other chemicals were purchased from Sigma. The drug vehicle at the concentration used in these experiments has no effect on the GABA response. Statistical significance was defined as P < 0.05 by the two-sided t-test. Application of GABA induced an inward-going current at negative holding potentials, and the current had a linear current-voltage relationship (Fig. 1A,B). The reversal potential of 1.5_ 0.4 mV (n = 8), measured from ramp voltage command-generated IV curves were near the expected reversal potential for chloride-permeable channels. Both alfll and alflly2 subunit combinations expressed well in transfected ceils, with an overall peak conductance of 3.09_ 0.77 nS/pF. For channels with a single channel conductance of approximately 30 pS [1], this corresponds to an estimated channel density of about 100 channels//zm2. Appropriate expression of the alfll subunits was deduced from the large currents induced by GABA applications. Control applications of GABA (10/tM) by pressure pulses of 100 ms at 20 psi gated currents with an Ipk of 155 _+38 pAJpF, and the current decay was well described
C
o~1131
Canh'ol
k_
d 4OOO
Control
20OO
-100-80-60-40~ !
,
|
i
*
V (u~V~ l/
|
!
|
i
20
40
§0
80
~:=1:_8ooo [ ](pAl
I00 J
/
Fig. 1. Current-voltage relationship and diazepam modulation of GABA-gated currents in transfected cells. (a) A ramp voltage command from - 8 0 to +80 mV over 400 ms was issued during the steady-state current induced by 10/~M GABA application in a cell transfected with alfl172 subunits. (1o) A current-voltage plot of the same data shows a linear IV curve with a reversal potential at 0 mV. (c) Current traces in response to 1 0 p M GABA application in an ctlfll subunit-transfected cell, with or without 1/~M diazepam. (d) Same as (c) except in a a l f l l y 2 transfected cell. The superimposed continuous lines are the best fitting monoexponential functions (i(t) = A x cxp(-t/z)) fit to the decay phase of the currents. For the traces shown, the amplitude (A) and the time constant (7) were A = -2288 pA, v = 950 ms, and A = --4782 pA, v = 1682 ms for control and with 1 p M diazepam, respectively. Calibrations are 4000 pA, 100 ms (a), 500 pA (c), 1000 pA (d), and 1000 ms for both (e) and (d).
!. Uchida et al. I Neuroscience Letters 185 (1995) 203-206
al~l'f2 ~[
CABA
GABA+ glomidate
205
¢z1131 GABA
C
L__ b
GABA GABA+ Etemidate GABA
L__ d
S
Fig. 2. Etomidate effects on GABA-gated currents in receptors composed of (a,b) a l f l l ? 2 subunits, and (c,d) alfll subunits. The top traces (a and c) are raw current traces showicLg reversible etomidate modulation of GABA-gated currents. The bottom traces (b and d) are superimposed traces of the control and the etomidate cttrrents normalized to the same peak. Prolongation of the time course of current decay for the etomidate trace (b) can be noted. The time constant of decay was 1170 ms for GABA only and 3240 ms for GABA + etomidate for an increase to 277% for this cell. Calibrations are 500 pA (a), 125 pA (b), and 2.4 s for the top traces and 700 ms for the bottom traces.
by a single exponential decay with a time constant of 1225 _ 75 ms. The corresponding values for the alfll)'2 subunits were 159 _+30 pA/pF and 1137 _ 77 ms. Thus both the magnitude and the decay properties of the control GABA responses were independent of the receptor subunit composition, in agreement with previous estimation of Bmax by binding studies [13]. Diazepam (1/zM) co-applied with GABA had no effect on alfll-transfected cells (Fig. 1C). In contrast, the same concentration of diazepam increased the peak response by 210% and increased the time constant of current decay by 160% (n = 3) (Fig. 1D) in alfll)'2-transfected cells, verifying the expression of the )'2 subunit. Etomidate at a clinically relevant concentration potentiated lpk to 162 _ 19% (n = 8) and 126 _+5% (n = 12) f~or alfll and alfll)'2, respectively. The decay time constant was significantly prolonged to 294_+58% only for alfll)'2 (Fig. 2). This apparent )'2 subunit-dependent IGABAmodulation by etomidate was further studied using a specific benzodiazepine receptor antagonist, flumazenil. Flumazenil (1/~M), while sufficient to completely block the effect of diazepam modulation of/GABA, did not block the etomidate action in all three cells examined. Thus, although etomidate shares with diazepam its dependence on the ),2 subunit for modulatory action, it appears not to act through the benzodiazepine site. One kinetic model of the spinal cord GABA channel indicates that the channels spend more time in the long burst state when the receptors are bound by two GABA molecules [I 1]. Accordingly, the receptor subunit combination which results in a higher affinity receptor should
show a slower current decay after a brief GABA application. Although the ED5o for the alfll)'2 and the alfll subunit combinations expressed in the L929 cells have been reported to be 7.4/zM and 1.0/zM, respectively [1], we detected no difference between the two subunit combinations in the decay rate of the current response to 10/zM GABA. This is because the decay of the current after a pressure application of GABA reflects the aggregate property of numerous channels opening and closing as the agonist concentration slowly decays back to zero. For a fixed time profile of agonist concentration, the time course of the current decay is more a reflection of the steepness of the dose-response curve rather than the burst duration of channels. Etomidate may increase the cooperativity of GABA action and increase the slope of the dose response in a)'2 subunit-dependent manner. It is well established that the benzodiazepine potentiation of GABAA receptor-gated current requires the presence of the )'2 subunit [15]. For the non-benzodiazepine general anesthetic GABAA modulators, the subunit dependence of their action is less clear. In expressed GABAA receptors containing alfll and a2fll subunits with or without the )'2 subunit, isoflurane potentiates the peak current due to exogenously applied GABA [8]. A comparison of pentobarbital effects on expressed a2fll subunit receptor [ 14] and native spinal cord receptors [ 12] suggests that the )'2 subunit has little effect on this drug's action either. In both the native receptors, presumed to contain the y2 subunit based on the benzodiazepine response, and the expressed receptors without the )'2 subunit, pentobarbital increased the amount of time the
206
L Uchida et al. / Neuroscience Letters 185 (1995) 203-206
channels reside in the longer burst states. However, a recent direct investigation of the role of y2L subunit in pentobarbital- and alphaxalone-induced shifts in the G A B A a dose-response demonstrate a subunit dependent difference [10]. Further evidence supporting the subunitdependent action of pentobarbital and isoflurane on the G A B A receptor is the observation that expressed homomeric receptors containing only the p l subunit, although opened by G A B A , are not modulated by either of these general anesthetics [8,17]. Our results support the idea that etomidate, an imidazole general anesthetic, potentiates G A B A ^ receptormediated current in a subunit-dependent manner. The basis for the segregation of etomidate effects on the peak current versus the current duration is probably a subunit dependent action of etomidate on specific rate constants governing the transition of channels from one state to another, as has been shown for the butyrolactone G A B A a receptor allosteric modulators [9]. The y2 subunit is widely distributed in the bt~ain including the forebrain and cerebellum [20]. Howevel, the exact cellular localization of the various subunits is largely unknown. Therefore, whether or not the earlier report of a difference between the effect of etomidate on [3H]diazepam binding in forebrain and cerebellar synaptosomes is due to the y2 subunit-dependent G A B A receptor modulation we describe here remains unknown. Further studies of subunit dependence o f general anesthetic modulation o f G A B A a channels, with particular attention to the similarities and differences in the amino acid sequences of the different subunits, will help define the molecular mechanisms and sites of action of these clinically essential drugs. This work was supported by the Merit Review Award from the Veterans Affairs (JY), the Young Investigator Award from the Foundation for Anesthesia Education and Research (JY), and NS25525 (DB). [1] Angelotti, T.P. and Macdonald, R.L., Assembly of GABAA receptor subunits: a2fll and alfl172s subnnits produce unique ion channels with dissimilar single-channel properties, J. Neurosci., 13 (1993) 1429-1440. [2] Ashton D., Geerts, R., Waterkeyn, C. and Leysen, J.E., Etomidate stereospecifically stimulates forebrain, but not cerebellar, H3diazepam binding, Life Sei., 29 (1981) 2631-2636. [3] Ashton, D. and Wauqnier, A., Modulation of a GABA-ergic inhibitory circuit in the in vitro hippocampus by etomidate isomers, Anesth. Analg., 64 (1985) 975-980. [4] Draguhn, A., Verdoro, T.A., Ewert, M., Seeburg, P.H. and Sakmann, B., Functional and molecular distinction between recombi-
nant rat GABAA receptor subtypes by Zn++, Neuron, 5 (1990) 781-788. [5] Evans, R.H. and Hill, R.G., GABA-mimetic action of etomidate, Experientia, 34 (1978) 1325-1326. [6] Franks,N.P. and Lieb, W.R., Molecular and cellular mechanisms of general anaesthesia, Nature, 367 (1994) 607-613. [7] Giese, J.L. and Stanley, T.H., Etomidate: a new intravenous anesthetic induction agent. Pharmacotherapy, 3 (1983) 251-258. [8] Harrison, N.L., Kubler, J.L., Jones, M.V., Greenblatt, E.P. and Pritchett, D.B., Positive modulation of human GABAA and glycine receptors by the inhalation anesthetic isoflurane, Mol. Pharroacol., 44 (1993) 628-632. [9] Holland, K.D., Yoon, K.W., Ferrendelli, J.A., Covey, D.F. and Rothman, S.M., y-Butyrolactone antagonism of the picrotoxin receptor: comparison of a pure antagonist and a mixed antagonist/inverse agonist, Mol. Pharmacol., 39 (1991) 79-84. [10] Home, A.L., Harkness, R.C., Hadingham, K.L., Whiting, P. and Kemp, J.A., The influence of the 72L subunit on the modulation of responses to GABAA receptor activation, Br. J. Pharmacol., 108 (1993) 711-716. [11] Macdonald, R.L., Rogers, C.J. and Twyman, R.E., Kinetic properties of the GABAA receptor main conductance state of mouse spinal cord neurones in culture, J. Physiol. (London), 410 (1989) 479-499. [12] Macdonald, R.L., Rogers, C.J. and Twyman, R.E., Barbiturate regulation of kinetic properties of the GABAA receptor channel of mouse spinal neurons in culture, J. Physiol. (London), 417 (1989) 483-500. [13] Moss, S.J., Ravindran, A., Mei, L., Wang, J.B., Kofuji, P., Huganir, R.L. and Burt, D.R., Characterization of recombinant GABAA receptors produced in transfected cells from routine a l , ill, and 72 subunit cDNAs, Neurosci. Lett., 123 (1991) 265-268. [14] Porter, N.M., Angelotti, T.P., Twyman, R.E. and Macdonald, R.L., Kinetic properties of a2fll GABAA receptor channels expressed in Chinese hamster ovary cells: regulation by pentobarbital and picrotoxin, Mol. Pharmacol., 42 (1993) 872-881. [15] Pritchett, D.B., Sontheimer, H., Shivers, D., Ymer, S., Kettenmann, H., Schofield, P.R. and Seeburg, P.H., Importance of novel GABAA receptor subunit for benzodiazepine pharmacology, Nature (London), 338 (1988) 582-585. [16] Proctor, W.R., Mynlieff, M. and Dunwiddie, T.V., Facilitatory action of etomidate and pentobarbital on recurrent inhibition in rat hippocampal pyramidal neurons, J. Neurosci., 6 (1986) 31613168. [17] Shimada, S., Cutting, G. and Uhl, G.R., GABA A or C receptor? GABA pl receptor RNA induces bicuculline-, barbiturate-, and benzodiazepine-insensitive GABA responses in Xenopus ooeytes, Mol. Pharrnacol., 41 (1992) 683-687. [18] Tanelian, D.L., Kosek, P., Mody, I. and Maciver, M.B., The role of the GABAA receptor/chloride channel complex in anesthesia, Anesthesiology, 78 (1993) 757-776. [19] Thyagarajan, R., Ramanjaneyulu, R. and Ticku, M.K., Enhancement of diazepam and GABA binding by (+)-etomidate and pentobarbital, J. Neurochem., 41 (1983) 578-585. [20] Wisden, W., Laurie, D.J., Monyer, H. and Seeburg, P.H., The distribution of 13 GABAA receptor subunit mRNAs in the rat brain. I. Telencephalon, diencephalon, rneseneephalon., J. Nenrosci., 12 (1992) 1040-1062.
NeuroseienceLetters 185 (1995) 203-206
NHHOSCIHC[ IHT[IIS
Etomidate potentiation of GABA A receptor gated current depends on the subunit composition I. U c h i d a a, G. K a m a t c h i b, D. B u r t b, J. Y a n g a,* aDepartment of Anesthesiology and Pain Management, UT Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75235-9068, USA bDepartment of P,~armacologyand Experimental Therapeutics, University of Maryland School of Medicine, Baltimore, MD, USA
Received 17 October1994;revisedversionreceived19 December1994;accepted20 December1994
Abstract
The role of the )'2 subunit in etomidate potentiation of GABAA receptor-gated chloride current was studied by whole cell patch clamp experiments on H293 cells expressing GABAA receptors. The GABAA receptor subunits a l f l l with or without the )'2 subunit expressed well, with an overall peak current of 157 ± 42 pA/pE At a clinically relevant concentration, etomidate potentiates the peak current induced by GABA equally well in receptors with or without the 7'2 subunit. In contrast, the time course of current decay was prolonged only in receptors with the ),2 subunit. This )'2 subunit-dependent prolongation of the current time course was not blocked by the benzodiazepine receptor antagonist flumazenil. These results show that etomidate, an imidazole general anesthetic, interacts with the GABAA receptor in a ?2 subunit-dependent manner. Keywords: GABAA receptor; General anesthetic; Etomidate; H293 cells; Patch clamp
Recent evidence suggests that general anesthetic interaction with the inhibitory neurotransmitter ?-aminobutyric acid (GABA) receptor may be one important pharmacological mechanism of action of these drugs [6,18]. Etomidate, an imidazole intravenous general anesthetic of widespread clinical acceptance, has this ability to interact with GABA A receptors. Both electrophysiological and binding studies indicate that etomidate shows GABAmodulatory and GABA-mimetic properties [3,5,16,19]. Of particular interes! is the observation that the stereospecific, etomida~te-induced increase in [3H]diazepam binding to rat brain is seen in the forebrain crude membrane preparation but not the cerebellar synaptosomes [2]. Given the difference in GABA A receptor subunit expression in different areas of the brain, this raises the possibility of a receptor subunit-dependent action of etomidate on brain GABAA receptors. The presence of the ?2 subunit has previously been shown to confer specific pharmacological and physiological properties to GABA A receptors. The ?2 subunit is essential for benzodiazepine potentiation of GABA-gated current [15], lack of sensitivity to externally applied zinc * Correspondingauthor, Tel.; +1 214 6485498; Fax: +1 214 6486799.
[4], and a large 29 pS single channel conductance [1]. Therefore, we investigated the potential role of the ?2 subunit in etomidate potentiation of GABAA receptors expressed in combination with a 1 and fl 1 subunits. The a l , ill, and ?2s murine GABA A receptor subunits [13] were subcloned into the pcDNA I eukaryotic expression vector (Invitrogen) by standard methods and analyzed by restriction enzyme mapping. Transfectionquality plasmids were prepared by alkaline lysis and CsCI gradient purification. Human embryonic kidney (H293) cells were grown on Matrigel-coated coverslips (Collaborative Research) in DMEM (Gibco no. 320-1960) supplemented with 10% fetal calf serum, 400/tg of glutamine, 50 U/ml penicillin, and 50/zg/ml streptomycin. One microgram each of the desired receptor subunitcontaining plasmids and an expression plasmid containing the SV40 large T antigen (kindly provided by Dr. Paulo Kofuji) were added to each 35-mm tissue culture dish containing the H293 cells. Co-expression of the large T antigen allows episomal replication of the subunitcontaining plasmid and greater protein expression. Transfection was accomplished by the calcium phosphate precipitation method following the manufacturer's protocol (Stratagene, CA), except that the CO 2 concentration in the
0304-3940/95/$09.50 © 1995 ElsevierScienceIrelandLtd. All fightsreserved SSD! 0304-3940(95)11263-S
204
L Uchida et al. I Neuroscience Letters 185 (1995) 203-206
incubator was increased to 10% to acidify the media to pH 7.2. Control transfection experiments with a fl-galactosidase reporter gene following the above protocol resuited in a transfection efficiency approaching 40%. Electrophysiological recordings from randomly selected cells 48-72 h after transfection with the GABA subunits confirmed the high transfection efficiency. For recording purposes, transfected cells growing on a 10-mm glass coverslip were placed in a recording chamber which was continuously perfused with an external solution (in mM: 140 NaC1, 4 KCI, 1 CaC12 , 1 MgCI2, 10 HEPES, 20 dextrose, pH 7.4), and cells were viewed under Hoffman optics (Nikon TMS). Cells growing in isolation were randomly selected, and a whole cell recording mode was established using standard patch-clamp techniques. The patch pipette contained (in mM: 140 CsC1, 4 NaCI, 10 EGTA, 10 HEPES, pH 7.4). GABA (10/zM) with or without etomidatq (8.2/zM) dissolved in the external solution was applie4 by pressure ejection (20 psi x 100 ms) through puffer electrodes placed approximately 50/zm from the cell body. The etomidate concentration of 8.2/zM (2/zg/ml) was chosen to reflect the serum concentration of this compound observed in patients undergoing general anesthesia [7]. The shift in holding currents upon puffer application of isotonic sucrose was monoexponential with a time constant of 500 ms. Therefore, the current decay measured in the experiments reflects the steadystate current remaining from random closing and re-
opening of channels as the GABA concentration decays with an exponential time course of 500 ms. The average cell input capacitance of the H293 cells, directly read from the capacitance neutralization knob (AI200, Axon Instruments), was 19.2 pF (n = 43). Etomidate in 35% polypropylene glycol (Amidate) was obtained from Abbot Laboratories, and all other chemicals were purchased from Sigma. The drug vehicle at the concentration used in these experiments has no effect on the GABA response. Statistical significance was defined as P < 0.05 by the two-sided t-test. Application of GABA induced an inward-going current at negative holding potentials, and the current had a linear current-voltage relationship (Fig. 1A,B). The reversal potential of 1.5_ 0.4 mV (n = 8), measured from ramp voltage command-generated IV curves were near the expected reversal potential for chloride-permeable channels. Both alfll and alflly2 subunit combinations expressed well in transfected ceils, with an overall peak conductance of 3.09_ 0.77 nS/pF. For channels with a single channel conductance of approximately 30 pS [1], this corresponds to an estimated channel density of about 100 channels//zm2. Appropriate expression of the alfll subunits was deduced from the large currents induced by GABA applications. Control applications of GABA (10/tM) by pressure pulses of 100 ms at 20 psi gated currents with an Ipk of 155 _+38 pAJpF, and the current decay was well described
C
o~1131
Canh'ol
k_
d 4OOO
Control
20OO
-100-80-60-40~ !
,
|
i
*
V (u~V~ l/
|
!
|
i
20
40
§0
80
~:=1:_8ooo [ ](pAl
I00 J
/
Fig. 1. Current-voltage relationship and diazepam modulation of GABA-gated currents in transfected cells. (a) A ramp voltage command from - 8 0 to +80 mV over 400 ms was issued during the steady-state current induced by 10/~M GABA application in a cell transfected with alfl172 subunits. (1o) A current-voltage plot of the same data shows a linear IV curve with a reversal potential at 0 mV. (c) Current traces in response to 1 0 p M GABA application in an ctlfll subunit-transfected cell, with or without 1/~M diazepam. (d) Same as (c) except in a a l f l l y 2 transfected cell. The superimposed continuous lines are the best fitting monoexponential functions (i(t) = A x cxp(-t/z)) fit to the decay phase of the currents. For the traces shown, the amplitude (A) and the time constant (7) were A = -2288 pA, v = 950 ms, and A = --4782 pA, v = 1682 ms for control and with 1 p M diazepam, respectively. Calibrations are 4000 pA, 100 ms (a), 500 pA (c), 1000 pA (d), and 1000 ms for both (e) and (d).
!. Uchida et al. I Neuroscience Letters 185 (1995) 203-206
al~l'f2 ~[
CABA
GABA+ glomidate
205
¢z1131 GABA
C
L__ b
GABA GABA+ Etemidate GABA
L__ d
S
Fig. 2. Etomidate effects on GABA-gated currents in receptors composed of (a,b) a l f l l ? 2 subunits, and (c,d) alfll subunits. The top traces (a and c) are raw current traces showicLg reversible etomidate modulation of GABA-gated currents. The bottom traces (b and d) are superimposed traces of the control and the etomidate cttrrents normalized to the same peak. Prolongation of the time course of current decay for the etomidate trace (b) can be noted. The time constant of decay was 1170 ms for GABA only and 3240 ms for GABA + etomidate for an increase to 277% for this cell. Calibrations are 500 pA (a), 125 pA (b), and 2.4 s for the top traces and 700 ms for the bottom traces.
by a single exponential decay with a time constant of 1225 _ 75 ms. The corresponding values for the alfll)'2 subunits were 159 _+30 pA/pF and 1137 _ 77 ms. Thus both the magnitude and the decay properties of the control GABA responses were independent of the receptor subunit composition, in agreement with previous estimation of Bmax by binding studies [13]. Diazepam (1/zM) co-applied with GABA had no effect on alfll-transfected cells (Fig. 1C). In contrast, the same concentration of diazepam increased the peak response by 210% and increased the time constant of current decay by 160% (n = 3) (Fig. 1D) in alfll)'2-transfected cells, verifying the expression of the )'2 subunit. Etomidate at a clinically relevant concentration potentiated lpk to 162 _ 19% (n = 8) and 126 _+5% (n = 12) f~or alfll and alfll)'2, respectively. The decay time constant was significantly prolonged to 294_+58% only for alfll)'2 (Fig. 2). This apparent )'2 subunit-dependent IGABAmodulation by etomidate was further studied using a specific benzodiazepine receptor antagonist, flumazenil. Flumazenil (1/~M), while sufficient to completely block the effect of diazepam modulation of/GABA, did not block the etomidate action in all three cells examined. Thus, although etomidate shares with diazepam its dependence on the ),2 subunit for modulatory action, it appears not to act through the benzodiazepine site. One kinetic model of the spinal cord GABA channel indicates that the channels spend more time in the long burst state when the receptors are bound by two GABA molecules [I 1]. Accordingly, the receptor subunit combination which results in a higher affinity receptor should
show a slower current decay after a brief GABA application. Although the ED5o for the alfll)'2 and the alfll subunit combinations expressed in the L929 cells have been reported to be 7.4/zM and 1.0/zM, respectively [1], we detected no difference between the two subunit combinations in the decay rate of the current response to 10/zM GABA. This is because the decay of the current after a pressure application of GABA reflects the aggregate property of numerous channels opening and closing as the agonist concentration slowly decays back to zero. For a fixed time profile of agonist concentration, the time course of the current decay is more a reflection of the steepness of the dose-response curve rather than the burst duration of channels. Etomidate may increase the cooperativity of GABA action and increase the slope of the dose response in a)'2 subunit-dependent manner. It is well established that the benzodiazepine potentiation of GABAA receptor-gated current requires the presence of the )'2 subunit [15]. For the non-benzodiazepine general anesthetic GABAA modulators, the subunit dependence of their action is less clear. In expressed GABAA receptors containing alfll and a2fll subunits with or without the )'2 subunit, isoflurane potentiates the peak current due to exogenously applied GABA [8]. A comparison of pentobarbital effects on expressed a2fll subunit receptor [ 14] and native spinal cord receptors [ 12] suggests that the )'2 subunit has little effect on this drug's action either. In both the native receptors, presumed to contain the y2 subunit based on the benzodiazepine response, and the expressed receptors without the )'2 subunit, pentobarbital increased the amount of time the
206
L Uchida et al. / Neuroscience Letters 185 (1995) 203-206
channels reside in the longer burst states. However, a recent direct investigation of the role of y2L subunit in pentobarbital- and alphaxalone-induced shifts in the G A B A a dose-response demonstrate a subunit dependent difference [10]. Further evidence supporting the subunitdependent action of pentobarbital and isoflurane on the G A B A receptor is the observation that expressed homomeric receptors containing only the p l subunit, although opened by G A B A , are not modulated by either of these general anesthetics [8,17]. Our results support the idea that etomidate, an imidazole general anesthetic, potentiates G A B A ^ receptormediated current in a subunit-dependent manner. The basis for the segregation of etomidate effects on the peak current versus the current duration is probably a subunit dependent action of etomidate on specific rate constants governing the transition of channels from one state to another, as has been shown for the butyrolactone G A B A a receptor allosteric modulators [9]. The y2 subunit is widely distributed in the bt~ain including the forebrain and cerebellum [20]. Howevel, the exact cellular localization of the various subunits is largely unknown. Therefore, whether or not the earlier report of a difference between the effect of etomidate on [3H]diazepam binding in forebrain and cerebellar synaptosomes is due to the y2 subunit-dependent G A B A receptor modulation we describe here remains unknown. Further studies of subunit dependence o f general anesthetic modulation o f G A B A a channels, with particular attention to the similarities and differences in the amino acid sequences of the different subunits, will help define the molecular mechanisms and sites of action of these clinically essential drugs. This work was supported by the Merit Review Award from the Veterans Affairs (JY), the Young Investigator Award from the Foundation for Anesthesia Education and Research (JY), and NS25525 (DB). [1] Angelotti, T.P. and Macdonald, R.L., Assembly of GABAA receptor subunits: a2fll and alfl172s subnnits produce unique ion channels with dissimilar single-channel properties, J. Neurosci., 13 (1993) 1429-1440. [2] Ashton D., Geerts, R., Waterkeyn, C. and Leysen, J.E., Etomidate stereospecifically stimulates forebrain, but not cerebellar, H3diazepam binding, Life Sei., 29 (1981) 2631-2636. [3] Ashton, D. and Wauqnier, A., Modulation of a GABA-ergic inhibitory circuit in the in vitro hippocampus by etomidate isomers, Anesth. Analg., 64 (1985) 975-980. [4] Draguhn, A., Verdoro, T.A., Ewert, M., Seeburg, P.H. and Sakmann, B., Functional and molecular distinction between recombi-
nant rat GABAA receptor subtypes by Zn++, Neuron, 5 (1990) 781-788. [5] Evans, R.H. and Hill, R.G., GABA-mimetic action of etomidate, Experientia, 34 (1978) 1325-1326. [6] Franks,N.P. and Lieb, W.R., Molecular and cellular mechanisms of general anaesthesia, Nature, 367 (1994) 607-613. [7] Giese, J.L. and Stanley, T.H., Etomidate: a new intravenous anesthetic induction agent. Pharmacotherapy, 3 (1983) 251-258. [8] Harrison, N.L., Kubler, J.L., Jones, M.V., Greenblatt, E.P. and Pritchett, D.B., Positive modulation of human GABAA and glycine receptors by the inhalation anesthetic isoflurane, Mol. Pharroacol., 44 (1993) 628-632. [9] Holland, K.D., Yoon, K.W., Ferrendelli, J.A., Covey, D.F. and Rothman, S.M., y-Butyrolactone antagonism of the picrotoxin receptor: comparison of a pure antagonist and a mixed antagonist/inverse agonist, Mol. Pharmacol., 39 (1991) 79-84. [10] Home, A.L., Harkness, R.C., Hadingham, K.L., Whiting, P. and Kemp, J.A., The influence of the 72L subunit on the modulation of responses to GABAA receptor activation, Br. J. Pharmacol., 108 (1993) 711-716. [11] Macdonald, R.L., Rogers, C.J. and Twyman, R.E., Kinetic properties of the GABAA receptor main conductance state of mouse spinal cord neurones in culture, J. Physiol. (London), 410 (1989) 479-499. [12] Macdonald, R.L., Rogers, C.J. and Twyman, R.E., Barbiturate regulation of kinetic properties of the GABAA receptor channel of mouse spinal neurons in culture, J. Physiol. (London), 417 (1989) 483-500. [13] Moss, S.J., Ravindran, A., Mei, L., Wang, J.B., Kofuji, P., Huganir, R.L. and Burt, D.R., Characterization of recombinant GABAA receptors produced in transfected cells from routine a l , ill, and 72 subunit cDNAs, Neurosci. Lett., 123 (1991) 265-268. [14] Porter, N.M., Angelotti, T.P., Twyman, R.E. and Macdonald, R.L., Kinetic properties of a2fll GABAA receptor channels expressed in Chinese hamster ovary cells: regulation by pentobarbital and picrotoxin, Mol. Pharmacol., 42 (1993) 872-881. [15] Pritchett, D.B., Sontheimer, H., Shivers, D., Ymer, S., Kettenmann, H., Schofield, P.R. and Seeburg, P.H., Importance of novel GABAA receptor subunit for benzodiazepine pharmacology, Nature (London), 338 (1988) 582-585. [16] Proctor, W.R., Mynlieff, M. and Dunwiddie, T.V., Facilitatory action of etomidate and pentobarbital on recurrent inhibition in rat hippocampal pyramidal neurons, J. Neurosci., 6 (1986) 31613168. [17] Shimada, S., Cutting, G. and Uhl, G.R., GABA A or C receptor? GABA pl receptor RNA induces bicuculline-, barbiturate-, and benzodiazepine-insensitive GABA responses in Xenopus ooeytes, Mol. Pharrnacol., 41 (1992) 683-687. [18] Tanelian, D.L., Kosek, P., Mody, I. and Maciver, M.B., The role of the GABAA receptor/chloride channel complex in anesthesia, Anesthesiology, 78 (1993) 757-776. [19] Thyagarajan, R., Ramanjaneyulu, R. and Ticku, M.K., Enhancement of diazepam and GABA binding by (+)-etomidate and pentobarbital, J. Neurochem., 41 (1983) 578-585. [20] Wisden, W., Laurie, D.J., Monyer, H. and Seeburg, P.H., The distribution of 13 GABAA receptor subunit mRNAs in the rat brain. I. Telencephalon, diencephalon, rneseneephalon., J. Nenrosci., 12 (1992) 1040-1062.