Introduction

Top Summary Introduction Materials & Methods Results Discussion References

The GABA_A receptor is modulated positively by a wide variety of structurally diverse general anesthetics (Harris et al., 1995; Whiting et al., 1995). In particular, halogenated ethers such as enflurane and isoflurane (Nakahiro et al., 1989; Wakamori et al., 1991; Jones et al., 1992) along with intravenous anesthetic agents such as the barbiturates (Barker and Ransom, 1978), propofol (Hales and Lambert, 1991), etomidate (Uchida et al., 1995), and steroid anesthetics (Peters et al., 1988) all enhance the function of the GABA_A receptor at clinically relevant concentrations. In addition, intravenous and volatile anesthetics can activate the GABA_A receptor directly in the absence of GABA (Barker and Ransom, 1978; Robertson, 1989; Hales and Lambert, 1991; Yang et al., 1992).

The GABA_A receptor is a heteromeric complex formed by different glycoprotein subunits (1-6, 1-4, 1-4, , and ) that coassemble to form a chloride channel (Whiting et al., 1995). Most GABA_A receptors in vivo consist of pentameric complexes of , , and  subunits with a stoichiometry of (Chang et al., 1996), although receptors lacking the  subunit can be expressed in vitro and are fully sensitive to general anesthetics (Pritchett et al., 1989; Jones et al., 1995).

GABA_A receptors are members of a ligand-gated ion channel superfamily that also includes the glycine, serotonin₃, GABA  (GABA_C), and nicotinic acetylcholine receptors (Ortells and Lunt, 1995). GABA_A receptors share significant amino acid sequence homology with these receptors (Ortells and Lunt, 1995). Glycine receptor function is modulated positively by clinical concentrations of volatile anesthetics (Harrison et al., 1993; Downie et al., 1996) but is affected only weakly by barbiturates (Koltchine et al., 1996; Mascia et al., 1996) and etomidate (Mascia et al., 1996). In addition, some general anesthetics have potent actions on neuronal (but not muscle) nicotinic acetylcholine receptors (Flood et al., 1997; Violet et al., 1997) or serotonin₃ receptors (Jenkins et al., 1996). In contrast to other members of the ligand-gated channel superfamily, GABA  receptors are insensitive to nearly all anesthetic compounds, including propofol (Mihic and Harris, 1996), barbiturates (Shimada et al., 1992), volatile anesthetics (Harrison et al., 1993; Mihic and Harris, 1996), and steroid anesthetics (Mihic and Harris, 1996).

The dissimilar pharmacology of  receptors helped identify residues crucial for positive modulation by general anesthetics. Recently, Mihic et al. (1997) constructed glycine 1/GABA 1 receptor chimeras and characterized a 45-amino acid residue domain within the glycine 1 receptor that was necessary for positive modulation by enflurane. Mutagenesis within this domain led to the identification of specific amino acid residues in TM2 and TM3 that seem to be necessary for modulation or direct activation of GABA_A and glycine receptors by enflurane.

The current study demonstrates that mutation of these specific residues in TM2 and TM3 also abolishes the enhancement of submaximal GABA-activated currents by isoflurane. Additional experiments were performed to test whether these specific mutations in TM2, TM3, or both also affected modulation and direct activation by intravenous general anesthetics.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Site-directed mutagenesis. The S270I (i.e., serine at position 270 mutated to isoleucine), S270H, and A291W mutations of the human GABA_A 2 subunit (Hadingham et al., 1993) and the M286W mutation of the GABA_A 1 subunit (Hadingham et al., 1993) were introduced by the unique site elimination method (Deng and Nickoloff, 1992) with use of the USE kit (Pharmacia Biotech, Piscataway, NJ). The method uses a two-primer system in which one oligonucleotide primer encodes the desired mutation and the other alters a unique SspI restriction site on the pCIS2 plasmid to an MluI site. The mutagenic reaction mixture was digested with SspI restriction endonuclease to eliminate the parental template. Positive clones of transformed Escherichia coli DH5 (Pharmacia Biotech) then were screened for the appearance of the MluI site, and mutations were confirmed further by double-stranded sequencing (Sequenase 2.0; United States Biochemical, Cleveland, OH). Both MluI and SspI restriction enzymes were from New England Biolabs (Beverly, MA). The sequences and locations of the primers (Operon Technologies, Alameda, CA) used are 2(S270I): 5-GACAACTCTAATCATCAGTGCTCGGAATTC-3, corresponding to bases 879-908 of the 2 cDNA sequence; 2(S270H), 5-GACAACTCTACACATCAGTGCTCGGAATTC-3, corresponding to bases 879-908 of the 2 cDNA sequence; 2(A291W), 5-CATGGACTGGTTTATTTGGGTTTGTTATGCATTTG-3, corresponding to bases 936-970 of the 2 cDNA sequence; 1(M286); and 5-GATTGATATTTATCTGTGGGGTTGCTTTGTG-3, corresponding to bases 915-945 of the 1 cDNA sequence.

Cell culture and transfection. Wild-type or mutant receptor cDNAs were expressed via the pCIS2 vector, which contains the strong promoter from cytomegalovirus and a polyadenylation sequence from Simian virus 40. HEK 293 cells (American Type Culture Collection, Rockville, MD) were cultured in Eagle's minimal essential medium (Sigma Chemical, St. Louis, MO) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT), L-glutamine (0.292 µg/ml; GIBCO BRL, Grand Island, NY), penicillin G sodium (100 units/ml; GIBCO BRL), and streptomycin sulfate (100 µg/ml; GIBCO BRL). For electrophysiological experiments, cells were plated onto glass coverslips coated with poly-D-lysine (Sigma). Each coverslip of cells was transfected individually according to the calcium phosphate precipitation technique (Pritchett et al., 1989; Harrison et al., 1993). There are reports of endogenous GABA_A receptor subunit expression in HEK 293 cells (Ueno et al., 1996). Our own experience with HEK 293 cells to date does not concur with this finding. In fact, similar to other reports (Davies et al., 1997), we do not see significant GABA-induced currents in untransfected or sham-transfected HEK 293 cells or by transfection with either GABA_A 2 or 1 subunit cDNAs alone (Koltchine et al., 1996).

Electrophysiology. Electrophysiological recordings were performed at room temperature using the whole-cell patch-clamp technique as described previously (Harrison et al., 1993; Koltchine et al., 1996). The coverslips were transferred 24-72 hr after removal of the cDNA to a 70-ml chamber that was continuously perfused (2-3 ml/min) with extracellular medium containing 145 mM NaCl, 3 mM KCl, 1.5 mM CaCl₂, 1 mM MgCl₂, 5.5 mM D-glucose, and 10 mM HEPES, pH 7.4, osmolarity 320-330 mOsM. The intracellular solution contained 145 mM N-methyl-D-glucamine hydrochloride, 5 mM K₂ATP, 5 mM HEPES/KOH, 2 mM MgCl₂, 0.1 mM CaCl₂, and 1.1 mM EGTA, pH 7.2, osmolarity 315 mOsM. Pipette-to-bath resistance was 4-6 M. Cells were voltage-clamped at 60 mV. Because the intracellular and extracellular solutions contained equal concentrations of chloride, the chloride reversal potential was 0 mV.

Data analysis. Modulator-induced potentiation of an agonist-induced current was defined as the percentage increase of the peak control agonist response (0% indicates no difference from control response). Concentration-response data were fitted (KaleidaGraph, Reading, PA) with the logistic equation: I/I_max = 100 * [drug]ⁿ/([drug]ⁿ + (EC₅₀ )ⁿ), where I/I_max is the percentage of the maximum obtainable agonist response, EC₅₀ is the concentration producing a half-maximal response, and n is the Hill coefficient. The current elicited by direct activation of the receptor by an anesthetic compound was expressed relative to the maximal current that could be elicited by GABA. Any direct activation produced by a modulator during pre-equilibration was subtracted from the total current elicited by coapplication of GABA and modulator. The subtraction of direct anesthetic activation is appreciable only at high anesthetic concentrations (e.g., 10 µM propofol or etomidate). In general, for the screening of potentiation, concentrations of anesthetics were chosen carefully (see Results) so little or no direct activation contributed to the observed current. Pooled data are presented throughout as mean ± standard error. Statistical significance was determined by Student's two-tailed, unpaired t test.

Drugs. Stock solutions of GABA (Sigma) and anesthetic compounds were diluted into extracellular solution daily before use. The other drugs used in this study were propofol (2,6-diisopropylphenol; Aldrich, Milwaukee, WI), methohexital sodium (Brevital sodium; Eli Lilly and Co., Indianapolis, IN), alphaxalone, picrotoxin (both from Research Biochemicals, Natick, MA), and isoflurane (Forane; Ohmeda Caribe, Guayama, PR). Propofol, etomidate, picrotoxin, and alphaxalone were first prepared as stock solutions in dimethylsulfoxide (Sigma) before being dissolved in the extracellular medium. The maximum final concentration of dimethylsulfoxide was 0.05% (v/v), which was determined during carrier control experiments to have no significant effect on GABA-induced currents in the receptor constructs analyzed in this study. Preparation and measurement of isoflurane solutions have been described previously (Jones et al., 1992). The MAC equivalent for isoflurane at 25° used for this study was 0.5 mM (Jones et al., 1992). Research-grade etomidate was a generous gift from Prof. Alfred Doenicke (Institute of Anesthesiology, Ludwig Maximilians University of München, Germany).

	Results

Top Summary Introduction Materials & Methods Results Discussion References

Expression and characteristics of wild-type and mutant receptors. Six wild-type and mutant GABA_A receptors were expressed by transfection of HEK 293 cells: 21 wild-type, 2(S270I)1, 2(S270H)1, 2(A291W)1, 21(S265I), or 21(M286W). After transient expression in HEK 293 cells, all of the wild-type and mutant receptors tested in this study produced inward currents in response to application of GABA via the rapid solution changer. The EC₅₀ and Hill slope values estimated from GABA concentration-response curves for receptors that contain either mutant 2 or 1 subunits demonstrate that these receptors retain GABA concentration-response relationships that are similar to those of wild-type 21 receptors: 21 wild-type (EC₅₀ = 8.7 ± 0.4 µM, n_H = 1.9 ± 0.2, nine experiments), 2(S270I)1 (EC₅₀ = 14.6 ± 0.1 µM, n_H = 2.4 ± 0.1, six experiments), 2(S270H)1 (EC₅₀ = 3.5 ± 0.2 µM, n_H = 1.5 ± 0.1, five experiments), 2(A291W)1 (EC₅₀ = 2.4 ± 0.1 µM, n_H = 1.7 ± 0.2, five experiments), 21(S265I) (EC₅₀ = 37.5 ± 8.5 µM, n_H = 1.2 ± 0.2, seven experiments), and 21(M286W) (EC₅₀ = 8.7 ± 0.4 µM, n_H = 0.8 ± 0.2, seven experiments).

Mutations in TM2 and TM3 of the GABA_A 2 and 1 subunits abolish positive modulation by isoflurane. At the clinically relevant concentrations of 0.25 and 0.5 mM (0.5 and 1.0 MAC equivalents, respectively), isoflurane strongly enhanced responses to an EC₂₀ concentration of GABA in wild-type 21 receptors (Fig. 1A). In contrast, coapplication of isoflurane up to 1.0 mM (2 MAC) produced no enhancement of submaximal GABA currents at mutant 2(S270H)1 or 2(A291W)1 receptors (Fig. 1, B and C). In fact, all five mutant receptors studied in this report were insensitive to isoflurane concentrations up to 1.0 mM (Table 1). To further examine the mechanism by which these mutations altered potentiation by isoflurane, we examined whether the addition of a 2s subunit to one of the mutant receptors would lead to any potentiation by isoflurane. Submaximal GABA currents at the 2(A291W)12s mutant receptor, however, also were insensitive to potentiation by isoflurane: potentiation of EC₂₀ GABA currents by 0.5 and 1.0 mM isoflurane was 3.8 ± 2.7% and 8.3 ± 4.7%, respectively (five experiments for both concentrations; negative values indicate slight inhibition).

View larger version (21K): [in this window] [in a new window]   Fig. 1.   Mutations within TM2 and TM3 abolish positive modulation by isoflurane. A, Submaximal GABA currents in wild-type GABAA 21 receptors are strongly enhanced by coapplication of clinically relevant concentrations of isoflurane (0.25 and 0.5 mM). B and C, In contrast, submaximal GABA currents in 2(S270H)1 or 2(A291W)1 mutant receptors are not enhanced by coapplication of isoflurane concentrations up to 1 mM. Individual recordings are from HEK 293 cells transfected with cDNAs encoding the indicated subunit combination.

Propofol potentiation of GABA. A propofol concentration of 1 µM was chosen to assay GABA potentiation because it is within the estimated clinically relevant concentration range (Franks and Lieb, 1994) and produces no direct activation in most of the receptors tested (see below). Mutations in the 2 subunit had little or no effect on GABA potentiation by 1 µM propofol (Fig. 2, Table 1). A point mutation in TM3 of the 1 subunit (M286W), however, eliminated GABA potentiation by 1 µM propofol (Table 1). In fact, submaximal GABA currents at the 21(M286W) mutant receptor were not enhanced by propofol at concentrations up to 10 µM (Fig. 3). The addition of a 2s subunit did not alter the effect of this 1 mutation because submaximal currents at the 21(M286W)2s mutant receptor also were insensitive to potentiation by propofol at concentrations up to 10 µM (Fig. 3B).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Mutations in the GABAA 2 subunit that abolish potentiation by isoflurane do not alter potentiation by propofol (PRO), etomidate (ETO), and methohexital (MTX). Submaximal GABA currents in wild-type GABAA 21 receptors are strongly enhanced by anesthetic concentrations of propofol (1 µM), etomidate (10 µM), and methohexital (5 µM). Similarly, submaximal GABA currents at 2(S270I)1 or 2(S270H)1 mutant receptors are enhanced by propofol, etomidate, and methohexital. Individual recordings are from HEK 293 cells transfected with cDNAs encoding the indicated subunit combinations.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Potentiation by both propofol and isoflurane is abolished in the 21(M286W) mutant receptor. A, Submaximal GABA currents in the 21(M286W) receptor are not enhanced by coapplication of 1.0 mM isoflurane (ISO) or supra-anesthetic propofol (PRO) concentrations (5 and 10 µM). Recording is from a single HEK 293 cell transfected with cDNAs encoding the 21(M286W) receptor combination. B, Concentration-response relationships for potentiation of GABA by propofol. For wild-type (WT) GABAA 21 receptors, significant potentiation of an EC20 test concentration of GABA occurs at all concentrations of 0.2 µM (p < 0.05 for each concentration 0.2 µM; 5-12 experiments for each data point). From the curve fit, the EC50 value for potentiation of wild-type 21 receptors by propofol is 1.6 µM with a Hill slope of 1.1. In contrast, propofol (10 µM) does not potentiate submaximal GABA currents at 21(M286W) (, 6-9 experiments) or 21(M286W)2s (, 7 experiments) mutant receptors.

Etomidate potentiation of GABA. Etomidate (1-20 µM) potentiated submaximal GABA currents at wild-type 21 receptors with an estimated EC₅₀ value of 3.4 µM, Hill slope of 1.6, and predicted maximal potentiation (E_max) of 216% (six or seven experiments for all concentrations). Etomidate (10 µM) significantly potentiated submaximal GABA currents in all wild-type and mutant receptors tested (Table 1). The magnitude of potentiation by etomidate was markedly reduced in the 2(A291W)1 mutant receptor, whereas all other mutant receptors retained normal potentiation by 10 µM etomidate (Table 1).

Methohexital potentiation of GABA. Previous work from our laboratory has shown that the barbiturate methohexital induces GABA potentiation, direct activation, and blocking effects on wild-type GABA_A 21 receptors (Koltchine et al., 1996). A concentration of 5 µM methohexital was used for the potentiation experiments because it produces substantial GABA potentiation without any direct activation of GABA_A 21 receptors (Koltchine et al., 1996). Methohexital potentiated the action of GABA at all receptors, although two receptors harboring mutations in TM3, 2(A291W)1 and 21(M286W), showed significantly lesser degrees of potentiation by 5 µM methohexital (Table 1).

Alphaxalone potentiation of GABA. None of the mutant receptors studied showed altered potentiation of submaximal GABA currents by 1 µM alphaxalone relative to wild-type 21 receptors (Table 1).

Effects of mutations in TM2 and TM3 on direct receptor activation by intravenous and volatile anesthetics. Direct activation by isoflurane was not studied in detail because isoflurane produces only minimal direct activation (5-10% of maximal GABA currents) in wild-type GABA_A 21 receptors. The direct activation of wild-type 21 receptors by isoflurane (Fig. 1A) was not evident in mutant receptors (Fig. 1, B and C). In fact, concentrations of isoflurane up to 5 mM (10 MAC) failed to elicit any direct activation of the 2(A291W)1 mutant receptor (data not shown).

Direct GABA_A receptor activation by propofol. Direct activation of wild-type GABA_A 21 receptors by propofol also was described previously (Jones et al., 1995). Concentrations of propofol of >50 µM can produce a profound block of the effect of GABA responses (Hales and Lambert, 1991), which is often followed by a "rebound" or "surge" current during washout. In our experiments, blocking and rebound effects were most evident at propofol concentrations of 100 µM, although small rebound currents immediately after washout of applied propofol were occasionally evident at 50 µM (Fig. 4). A propofol concentration of 50 µM thus was suitable to compare near-maximal direct activation by propofol in different receptors, with minimal interference from block.

View larger version (10K): [in this window] [in a new window]   Fig. 4.   Direct receptor activation by supra-anesthetic concentrations of propofol (PRO), etomidate (ETO), and methohexital (MTX) is similar to that for wild-type GABAA 21 receptors in most mutant receptors, except for 21(S265I), in which direct activation by propofol is reduced markedly. Wild-type GABAA 21 receptors are activated directly by propofol (50 µM), etomidate (50 µM), and methohexital (200 µM). Propofol, etomidate, and methohexital also directly activate 2(S270I)1, 21(S265I), and 21(M286W) mutant receptors. Direct activation by propofol is reduced markedly in 21(S265I) but not 2(S270I)1 receptors. Horizontal scale bar, 10 sec for direct activation by propofol, etomidate, and methohexital and 2 sec for the maximal GABA applications. The "rebound" currents at the tail end of drug applications are evident in some of the propofol and methohexital traces. Individual recordings are from HEK 293 cells transfected with cDNAs encoding the indicated subunit combinations.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Concentration-response relationships for direct activation of GABAA 21 receptors by propofol and etomidate. A, For wild-type (WT) 21 receptors (), propofol produces significant direct activation at all concentrations of 5 µM (p < 0.05 for each concentration 5 µM, 4-8 experiments). The direct activation of wild-type 21 receptors by propofol has an EC50 value of 10.5 µM, Hill slope of 2.6, and Emax value of 52.0% of the maximal GABA current. In contrast, the 2(A291W)1 receptor is more sensitive than the wild-type 21 receptor to the direct actions of propofol (). Propofol produces significant direct activation at all concentrations of 0.5 µM at the 2(A291W)1 receptor (p < 0.01 for each concentration 0.5 µM, 4-13 experiments). The direct activation of 2(A291W)1 receptors by propofol has an EC50 value of 1.0 µM, Hill slope of 1.2, and Emax value of 39.8%. The 21(M286W) mutant receptor, although insensitive to potentiation by propofol, is still directly gated by propofol () at all concentrations of 2 µM (p < 0.05 for each concentration  2 µM, 6-15 experiments). The direct activation of 21(M286W) receptors by propofol has an EC50 value of 9.5 µM, Hill slope of 2.2, and Emax value of 39.5%. B, For wild-type (WT) 21 receptors (), etomidate produces significant direct activation at all concentrations of 5 µM (p < 0.05 for each concentration 5 µM, 4-7 experiments). The direct activation of wild-type 21 receptors by etomidate has an EC50 value of 10.7 µM, Hill slope of 1.9, and Emax value of 47.5%. Similar to propofol, the 2(A291W)1 receptor is more sensitive to the direct actions of etomidate (). Etomidate produces significant direct activation at all concentrations of 0.5 µM at the 2(A291W)1 receptor (p < 0.05 for each concentration 0.5 µM, 7-10 experiments). The direct activation of 2(A291W)1 receptors by etomidate has an EC50 value of 1.9 µM, Hill slope of 1.2, and Emax value of 31.4%. The 21(M286W) mutant receptor is directly gated by etomidate () at all concentrations of 10 µM (p < 0.01 for each concentration 10 µM, 7-19 experiments). The direct activation of 21(M286W) receptors by etomidate has an EC50 value of 15.1 µM, Hill slope of 2.5, and Emax value of 23.5%. Error bars, standard error.

Direct GABA_A receptor activation by etomidate. It has been reported in experiments with Xenopus laevis oocytes that the direct action of etomidate is nearly absent or that etomidate has substantially reduced efficacy and/or apparent affinity in 1 compared with 2 or 3 subunit-containing receptors (Hill-Venning et al., 1997; Sanna et al., 1997). In our experiments using HEK 293 cells, etomidate produced significant direct effects in wild-type 21 receptors in the absence of GABA (Figs. 4 and 5B, Table 2). The direct activation by etomidate at wild-type 21 receptors had an EC₅₀ value of 10.7 µM, a much higher apparent affinity than seen in experiments with 1-containing subunits in X. laevis oocytes (Hill-Venning et al., 1997; Sanna et al., 1997).

Direct GABA_A receptor activation by methohexital. At high concentrations, methohexital produces substantial direct activation of GABA_A 21 receptors (Koltchine et al., 1996); 200 µM methohexital was chosen for the direct activation experiments because at this concentration, methohexital elicits a large direct effect with minimal blocking effects. Direct activation of all the mutant receptors by methohexital was similar to that for the wild-type 21 receptor (Fig. 4, Table 2). Direct activation by alphaxalone was not studied because it is not prominent in GABA_A 21 receptors (C. E. Rick, unpublished observations).

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

Actions of general anesthetics on wild-type GABA_A 21 receptors. Submaximal GABA currents at wild-type human GABA_A 21 receptors were potentiated by all the volatile and intravenous anesthetics studied. Thus, in agreement with other published studies, the presence of the  subunit is not required for potentiation by propofol (Jones et al., 1995), methohexital (Koltchine et al., 1996), etomidate (Uchida et al., 1995; Sanna et al., 1997), alphaxalone (Horne et al., 1993), or isoflurane (Harrison et al., 1993; Mihic et al., 1994).

A mutation in TM3 of the 2 subunit alters the concentration-response relationship for GABA and for direct receptor activation by propofol and etomidate. Although the GABA concentration-effect relationships for the mutants analyzed in this study mostly were very similar to that in the 21 wild-type receptor, a few changes were evident. A receptor with a mutation in TM3 of the 2 subunit, 2(A291W)1, has a higher apparent affinity for GABA than the wild-type 21 receptor, as indicated by a leftward shift in the GABA concentration-response curve. In addition, the concentration-response curves for direct activation by etomidate and propofol were shifted to the left of those for the wild-type 21 receptor by 6- and 10-fold, respectively. Thus, for the 2(A291W)1 mutant receptor, the apparent affinities for activation by GABA, etomidate, and propofol are all increased relative to wild-type. Also, the receptors containing mutations in the 1, but not 2, subunit had significantly lower Hill slopes compared with the wild-type 21 receptor for their GABA concentration-response relationships. These changes suggest altered gating mechanisms in these mutant receptors.

A mutation in TM3 of the 1 subunit abolishes potentiation of GABA but not direct receptor activation by propofol. For the most part, the mutations analyzed in this study had little effect on GABA potentiation and/or direct activation by the intravenous anesthetics propofol, etomidate, methohexital, or alphaxalone. An exception is the 21(M286W) mutant receptor, at which propofol concentrations of 10 µM failed to enhance submaximal GABA currents. The addition of a 2s subunit did not alter the deleterious effect of the 1(M286W) mutation because the 21(M286W)2s mutant receptor also was insensitive to propofol potentiation of GABA. In contrast to the lack of potentiation, propofol still directly activates this mutant receptor, with a concentration-response relationship that overlaps that for the wild-type 21 receptors.

Residues within TM2 and TM3 of the GABA_A receptor may form part of a volatile ether anesthetic binding site. The results from this study demonstrate that specific residues within both TM2 and TM3 of GABA_A 2 and 1 subunits are necessary for positive receptor modulation by the halogenated methyl ethyl ether anesthetic isoflurane. This extends the findings from a previous study that these residues also are critical for positive modulation by the related volatile anesthetic enflurane (Mihic et al., 1997). These same specific residues within TM2 and TM3 seem to be critical for modulation by n-alkanols (Mihic et al., 1997). In this study, mutations in either the 2 or 1 subunits in TM2 or TM3 abolished modulation by isoflurane. Similar to the situation seen with propofol potentiation of GABA at the 21(M286W)2s mutant receptor, the 2(A291W)12s mutant receptor was as insensitive to isoflurane potentiation of GABA as the 2(A291W)1 mutant receptor. This is consistent with the hypothesis that the  subunit is not a major determinant of volatile anesthetic modulation.