Abstract
Volatile anesthetics and alcohols enhance transmission mediated by γ-aminobutyric acid type A receptors (GABAARs) in the central nervous system, an effect that may underlie some of the behavioral actions of these agents. Substituting a critical serine residue within the GABAAR α1 subunit at position 270 with the larger residue histidine eliminated receptor modulation by isoflurane, but it also affected receptor gating (increased GABA sensitivity). To correct the shift in GABA sensitivity of this mutant, we mutated a second residue, leucine at position 277 to alanine. The double mutant α1(S270H,L277A)β2γ2S GABAAR was expressed in Xenopus laevis oocytes and human embryonic kidney (HEK)293 cells, and it had near-normal GABA sensitivity. However, rapid application of a brief GABA pulse to receptors expressed in HEK293 cells revealed that the deactivation was faster in double mutant than in wild-type receptors. In all heterologous systems, the enhancing effect of isoflurane and ethanol was greatly decreased in the double mutant receptor. Homozygous knockin mice harboring the double mutation were viable and presented no overt abnormality, except hyperactivity. This knockin mouse line should be useful in determining which behavioral actions of volatile anesthetics and ethanol are mediated by the GABAARs containing the α1 subunit.
The γ-aminobutyric acid type A receptors (GABAARs) are ligand-gated ion channels important for fast inhibitory synaptic transmission in the central nervous system. Clinically relevant concentrations of volatile anesthetics and alcohols potentiate GABAAR function.
Using chimeric constructs, an alcohol and volatile anesthetic binding site was identified in α2β1 GABAARs (Mihic et al., 1997). Additional experiments showed that mutations in α subunits at Ser270 decreased alcohol potentiation (Ueno et al., 1999; Findlay et al., 2000). Mutation of Ser270 to cysteine allowed irreversible labeling of mutant receptors with alcohol analogs (sulfhydryl-specific reagents) that blocked subsequent application of alcohols and volatile anesthetics (Mascia et al., 2000). Volatile anesthetics act in a common domain within the transmembrane region of GABAAR α subunits, and their action depends on their molecular size (Jenkins et al., 2001; Kash et al., 2003).
One α subunit mutation, α1 Ser270 to His, has been extensively tested using volatile anesthetics. Mutant GABAARs [α1(S270H)β2γ2S] expressed in human embryonic kidney (HEK)293 cells showed no potentiation by isoflurane, along with a significant increase of GABA sensitivity (Nishikawa et al., 2002). When expressed in Xenopus laevis oocytes, α1(S270H)β2γ2S GABAARs provided a reduced isoflurane potentiation and a significant increase of GABA sensitivity (Hall et al., 2004). Knockin mice were produced bearing the α1(S270H) mutation; however, the mice presented several synaptic and behavioral abnormalities that limited their usefulness in studies of volatile anesthetic effects, probably as a consequence of the increased sensitivity to GABA (Homanics et al., 2005; Elsen et al., 2006). The α1(S270H) knockin mice, however, presented an interesting feature: their miniature inhibitory postsynaptic currents in hippocampal slices showed a decreased sensitivity to isoflurane, but halothane produced the same effect as in wild-type mice (Elsen et al., 2006).
The present study aimed to create a knockin mouse line expressing mutant α1 GABAARs that lack volatile anesthetic and alcohol sensitivity but present near-normal response to GABA. Such a genetically engineered mouse would enable a definitive study of the relevance of α1-containing GABAARs to the cellular and behavioral effects of volatile anesthetics and alcohols. As mentioned above, replacement of Ser270 with His eliminated potentiation by isoflurane, but the GABA concentration-response curve (CRC) shifted to the left. To correct for the enhanced sensitivity to GABA, a second mutation was needed. A variety of mutations (e.g., Leu at position 277 to Ala) in the transmembrane 2-3 linker of the α subunit inhibit gating of the GABAAR, increasing GABA EC50 and reducing the maximal GABA response (O'Shea and Harrison, 2000; Topf et al., 2003). Therefore, we constructed a double mutant that incorporated both mutations, S270H and L277A, in the same α1 subunit [α1(S270H,L277A)] and compared them with wild-type receptors. We coexpressed these α1 subunits along with β and γ subunits in heterologous systems (HEK293 cells and Xenopus oocytes). We subsequently studied the pharmacology of the volatile anesthetics isoflurane and halothane at these receptors, along with other modulators of the GABAAR, including alcohols. Drugs, such as etomidate and pentobarbital, were included in the present experiments because they were also to be included in subsequent behavioral and electrophysiological studies of the knockin mice to determine whether the double mutation introduced in the α1 subunit also produced any changes in the sensitivity to these drugs (Werner et al., 2006). Based on these findings, we created a knockin mouse line bearing the double mutation and performed biochemical and behavioral analyses to assess the phenotype of these mice.
Materials and Methods
Site-Directed Mutagenesis
To create the mutant α1 subunit, we introduced two single point mutations into the cDNA encoding the human GABAAR α1 subunit (Koltchine et al., 1996). Mutations were performed with the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) with commercially produced primers of 24 to 30 bases (QIAGEN Operon, Alameda, CA). Mutant cDNAs were confirmed by automated fluorescent DNA sequencing (Cornell DNA Sequencing Service, Ithaca, NY). All restriction enzymes were obtained from New England Biolabs (Beverly, MA).
Electrophysiology in HEK293 Cells
Cell Culture and Transfection. Wild-type or mutant human α1 subunit cDNAs were coexpressed with the rat GABAA β2 and human γ2S subunits via the plasmid vector pCIS2 in HEK293 cells, as described previously (Koltchine et al., 1999). HEK293 cells (American Type Tissue Culture Collection, Manassas, VA) were cultured on poly-d-lysine-treated coverslips (Sigma-Aldrich, St. Louis, MO) in Eagle's minimum essential medium (Sigma-Aldrich) supplemented with 5% fetal bovine serum (Hyclone Laboratories, Logan, UT), 0.292 μg/ml l-glutamine (Invitrogen, Carlsbad, CA), 100 units/ml penicillin G sodium (Invitrogen), and 100 μg/ml streptomycin sulfate (Invitrogen). Cells were transfected using the calcium phosphate precipitation technique (Chen and Okayama, 1987) to achieve transient expression. Each coverslip of cells was transfected with approximately 6 μg of total DNA. The transfected cells were cultured for 24 h in an atmosphere containing 3% CO2 before being removed and replaced with fresh culture medium in an atmosphere of 5% CO2.
Electrophysiological Recording in HEK293 Cells Using Standard Application Methods. Coverslips with the transfected cells were transferred after 48 to 72 h to a bath that was continuously perfused with extracellular saline. The extracellular saline contained 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, at an osmolarity of 320 to 330 mOsm. Recordings were performed at room temperature using the whole-cell patch-clamp technique as described previously (Krasowski et al., 1997). The patch pipette solution contained 147 mM N-methyl-d-glucamine hydrochloride, 5 mM CsCl, 5 mM K2ATP, 5 mM HEPES, 1 mM MgCl2, 0.1 mM CaCl2, and 1.1 mM EGTA, pH 7.2, at an osmolarity of 315 mOsm. The chloride equilibrium potential was therefore approximately 0 mV. Pipette-to-bath resistance was typically 5 to 10 MΩ. Cells were voltage-clamped at -60 mV. All drugs were dissolved in extracellular medium and rapidly applied to the cell by local perfusion with laminar flow using a multichannel infusion pump (Stoelting, Wood Dale, IL) (Koltchine et al., 1996). The loss of isoflurane/halothane using this perfusion device has been measured using gas chromatography and represents only 5 to 10% of the total applied drug concentration (M. D. Krasowski, unpublished observations). The solution changer was driven by protocols in the acquisition program pCLAMP5 (Molecular Devices, Foster City, CA). Throughout the experiment, each cell was periodically challenged with a submaximal concentration of agonist to ensure that no cumulative desensitization or rundown of the GABAAR currents had occurred. Responses were digitized (TL1-125 interface; Axon Instruments) using pCLAMP5 (Molecular Devices).
GABA was dissolved directly into extracellular solution or stored overnight at 4°C in sealed Nalgene tubes (Nalge Nunc International, Rochester, NY). All reagents were purchased from Sigma-Aldrich, with the exception of isoflurane (Abbott Laboratories, Abbott Park, IL) and halothane (Halocarbon Laboratories, River Edge, NJ).
Electrophysiological Recording in HEK293 Cells Using Rapid Application Methods. The procedure was essentially as described previously (Li and Pearce, 2000). In brief, human α1 wild-type (α1) or double mutant [α1(S270H,L277A)] rat β2 and human γ2S GABAA subunits were transiently expressed in HEK293 cells (ratio 1:1:10) using Lipofectamine 2000 (Invitrogen). Electrophysiological recordings were performed with rapid applications using a two-barrel “theta” glass application pipette connected to a piezoelectric stacked translator (∼2.5-ms solution exchange). GABA CRCs were determined using the Cellectricon Dynaflow df-16 chip (∼25-ms whole-cell 10 to 90% solution exchange; S. Dai and R. A. Pearce, unpublished observations).
Data Analysis. Concentration-response data were fitted (Kaleidograph; Synergy Software, Reading, PA, or OriginPro 7.5; Origin-Lab Corp, Northampton, MA) using the following equation (Hill equation): where I/IMAX is the fraction of the maximally obtained GABA response, EC50 (effective concentration 50) is the concentration of agonist producing a half-maximal response, [drug] is drug concentration, and nH is the Hill coefficient. Agonist responses in each cell were normalized to the maximal current that could be elicited by GABA. Percentage of potentiation was then calculated as the percentage of change from the control (EC20) response to GABA in the presence of anesthetic. Pooled data were represented as mean ± S.E.M. Statistical significance was determined at the p < 0.05 level by a two-tailed unpaired Student's t test assuming unequal variance.
Electrophysiology in Xenopus Oocytes
Materials. Adult female X. laevis were obtained from Xenopus Express (Plant City, FL). GABA, ethanol, zinc chloride, flunitrazepam, and sodium pentobarbital were purchased from Sigma-Aldrich, etomidate was purchased from Tocris Cookson Inc. (Ellisville, MO), and isoflurane was purchased from Marsam Pharmaceuticals (Cherry Hill, NJ). All other reagents were of reagent grade. GABA, zinc chloride, and pentobarbital sodium stocks were prepared in water; flunitrazepam and etomidate were dissolved in dimethyl sulfoxide. The drug stocks were then dissolved in buffer; the final dimethyl sulfoxide concentration was 0.1% (v/v), which does not affect GABAA-mediated current. Isoflurane solutions were prepared in buffer immediately before application.
Isolation and Injection of Oocytes.X. laevis oocytes were manually isolated from a surgically removed portion of ovary. Oocytes were treated with collagenase (type IA; 0.5 mg/ml) for 10 min and then placed in sterile modified Barth's solution [88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO3, 10 mM HEPES, 0.82 mM MgSO4, 0.33 mM Ca(NO3)2, and 0.91 mM CaCl2, adjusted to pH 7.5], supplemented with 10,000 units of penicillin, 50 mg of gentamicin, 90 mg of theophylline, and 220 mg of sodium pyruvate per liter (incubation medium). Oocytes were then injected into the nucleus with 30 nl of a solution containing cDNA encoding GABAA subunits (α:β:γ at 0.5: 0.5:1 in nanograms per oocyte). The cDNAs were human α1 (wild-type and mutated), rat β2 and human γ2S in vector pCIS2, and human β3 in vector pcDNA1AMP (provided by Dr. P. J. Whiting, Merck Sharp & Dohme Research Laboratories, Harlow, UK). The injected oocytes were kept at 13°C in incubation medium.
Electrophysiological Recordings. Recordings were carried out 1 to 5 days after injection. The oocytes were placed in a rectangular chamber (approximately 100 μl) and continuously perfused with 2 ml/min ND96 buffer at room temperature (24°C). The perfusion buffer composition was 96 mM NaCl, 1 mM CaCl2, 2 mM KCl, 1 mM MgCl2, and 5 mM HEPES, pH 7.5. The whole-cell voltage clamp at -70 mV was achieved through two glass electrodes (1.5-10 MΩ) filled with 3 M KCl using an oocyte clamp (model OC-725C; Warner Instruments, Hamden, CT).
All drugs were applied by bath perfusion. All solutions were prepared the day of the experiment. The preapplication of volatile anesthetics and alcohol provided more reliable responses in previous studies and was also used here.
The CRCs were obtained with increasing concentrations of GABA, applied for 20 to 30 s at intervals ranging from 5 to 15 min. From these CRCs, the concentration evoking a half-maximal response (EC50) was calculated, along with the Hill coefficient (see Statistical Analysis). To study the 0.3 mM isoflurane, 10 to 200 mM ethanol, 11 mM butanol, 10 μM Zn2+, 1 μM flunitrazepam, 1 μM etomidate, and 50 μM pentobarbital modulation of GABA currents, the GABA concentration equivalent to EC5 was determined after 1 to 3 mM GABA produced a maximal current. A washout of 5 min was observed in between all GABA applications, except after 1 to 3 mM GABA (15 min). After two applications of EC5 GABA, each of the modulators was preapplied for 1 min and then coapplied with GABA for 30 s. EC5 GABA was reapplied after coapplication of GABA and modulator. To observe the direct effect of etomidate on GABAARs, 10 μM etomidate was applied for 1 min. All experiments shown include data obtained from oocytes taken from at least two frogs. All oocytes that presented a maximal current >20 μA were discarded.
Data Analysis. Nonlinear regression analysis was performed with Prism (GraphPad Software Inc., San Diego, CA). CRCs were fitted to the eq. 2: where I/IMAX is the fraction of the maximally obtained GABA response, EC50 is the concentration of agonist producing a half-maximal response, [GABA] is GABA concentration, and nH is the Hill coefficient. Agonist responses in each cell were normalized to the maximal current that could be elicited by GABA. Percentage of potentiation was then calculated as the percentage of change from the control response to EC5 GABA in the presence of anesthetic. Pooled data were represented as mean ± S.E.M. Statistical significance was determined using Student's t test or ANOVA, as indicated.
Knockin Mouse Production
A targeting construct to modify the α1 locus was created from a vector that was previously used to create gene targeted mice that harbored an α1 GABAA-R Ser270 to His knockin mutation (Homanics et al., 2005). In brief, QuikChange XL site-directed mutagenesis (Stratagene) was used to replace the Leu277 codon (CTC) with an A codon (GCT) in a plasmid that already harbored the Ser270 to His substitution, and a silent 1-base pair substitution that introduced a novel EcoRI site (Fig. 6A). Note that exon 9 in the present publication is identical to exon 8 in our previous publication. The recent identification of an untranslated exon at the 5′ end of the α1 gene resulted in the renumbering of exons. The gene-targeting construct was completely linearized with PvuI before introduction into mouse embryonic stem cells. The linearized construct was electroporated into Strain 129SvJ Go Germline embryonic stem cells (Genome Systems, Inc., St. Louis, MO) under previously described conditions (Homanics et al., 1997; Homanics, 2002). G418 (Geneticin, 265 μg/ml; Invitrogen)-resistant embryonic stem cell clones were screened for gene targeting by Southern blot analysis of EcoRI-digested DNA and hybridization with a 3′ external probe (Fig. 6B) as described previously (Homanics et al., 2005). Targeted clones were also analyzed with additional enzymes and probes, and all results were consistent with correct targeting at the α1 locus. The results presented here are from clone 107A5.
Correctly targeted embryonic stem cell clones were microinjected into C57BL/6J blastocysts to produce chimeric mice. Male chimeras were mated to C57BL/6J females to produce the F1 generation. Mice heterozygous for the neo-containing targeted locus were mated to FLPe deleter mice (Rodriguez et al., 2000) obtained from The Jackson Laboratory (stock 3800; Bar Harbor, ME) to remove the neo cassette by FLPe-mediated site-specific recombination. The FLPe transgene was subsequently bred out. Mice heterozygous for the knockin were interbred to produce controls (SL/SL), heterozygous (SL/HA), and homozygous (HA/HA) knockins. All mice were genotyped by Southern blot analysis of tail DNA as described previously (Homanics et al., 2005).
All mice were of a mixed C57BL/6J × strain 129SvJ background of the F3-F6 generations. All animals were maintained under specific pathogen free conditions in a photoperiod-controlled environment (lights on at 7:00 AM and off at 7:00 PM) with ad libitum access to standard rodent chow and water. All experiments were approved by the Institutional Animal Care and Use Committees and were conducted in accordance with National Institutes of Health guidelines on the use of animals in research.
Whole-brain RNA was extracted using TRIzol (Invitrogen), and α1 mRNA was amplified by RT-PCR as described previously (Homanics et al., 2005). Purified RT-PCR products from control and knockin brain were sequenced and compared.
Quantitative Immunoblot Analysis
Cerebral cortices and cerebella of adult mice were rapidly dissected over ice, flash-frozen on dry ice, and stored at -80°C. P2 membrane fractions from cortex were processed and analyzed as pooled samples (three pools per genotype, eight mice per pool), whereas cerebellar samples were analyzed individually (n = 8/genotype). Aliquots of 25 μg of protein from each pooled sample were separated by electrophoresis on precast SDS-10% polyacrylamide gels (Bio-Rad, Hercules, CA) and subsequently transferred to polyvinylidene difluoride membranes (Bio-Rad) for detection by subunit specific antibodies. GABAAR anti-α1, -α2, and -α3 antibodies (Fritschy and Mohler, 1995) were generously donated by Dr. Jean-Marc Fritschy (University of Zurich, Zurich, Switzerland). Anti-β2 (NB 300-198), anti-β3 (NB 300-119), and anti-γ2 (NB 300-151) antibodies were obtained commercially (Novus Biologicals, Littleton, CO). Primary antibodies were detected with either horseradish peroxidase-conjugated goat anti-rabbit (α1, β2, β3, γ2, and actin) or rabbit anti-guinea pig (α2, α3) (both from Abcam, Cambridge, MA) IgG polyclonal antibodies and visualized by enhanced chemiluminescence (Western Lightning; PerkinElmer Life and Analytical Sciences, Boston, MA). To ensure equal loading, blots were stripped using Re-blot (Chemicon International, Temecula, CA) and reprobed with an anti-β-actin polyclonal antibody (ab8227-50; Abcam) for normalization. Multiple exposures of each membrane were used to ensure that the measured signal was within the linear range of the film. Band intensity was measured densitometrically (Kodak 1-D software, version 3.6; Eastman Kodak, Rochester, NY). Each sample was analyzed on three to four different blots. Data were analyzed by Student's t test.
[3H]Flunitrazepam Binding
Cortical tissue was harvested from male and female knockin and wild-type mice, and GABAAR binding was performed using [3H]flunitrazepam (PerkinElmer Life and Analytical Sciences). Data from male and female mice were combined for statistical analyses. Tissue was homogenized in 25 ml of ice-cold assay buffer (50 mM Tris and 25 mM HEPES, pH 7.4) and centrifuged twice at 47,500g for 10 min (4°C). Final pellets were resuspended in ice-cold assay buffer. Binding was initiated by adding 200-μl aliquots of cortical tissue (100-200 μg of protein) to a reaction mixture containing 200 μl of ice-cold assay buffer, 50 μl of [3H]flunitrazepam (84.5 Ci/mmol; 1, 3, 10, 30, or 100 nM), and 50 μl of additional ice-cold assay buffer (nonspecific binding) or 100 μM diazepam (total binding). The reaction mixture incubated for 60 min at 4°C and was quenched with 2 ml of ice-cold assay buffer and was then rapidly filtrated through a GB100R filter (Advantec MFS, Dublin, CA) and washed with ice-cold assay buffer. Filters were incubated overnight in 4 ml of Biosafe II scintillation liquid (Research Products International, Mount Prospect, IL) before analysis in a Beckman LS 6500 scintillation counter (Beckman Coulter, Fullerton, CA). Specific binding was calculated by subtracting nonspecific binding from total binding. KD and BMAX values were calculated using Prism 3.0 (GraphPad Software Inc.).
Motor Activity
Spontaneous locomotor activity was measured in standard mouse cages by an Opto-microvarimex (Columbus Instruments, Columbus, OH). Activity was monitored by six infrared light beams placed along the width of the cage at 2.5-cm intervals, 1.5 cm above the floor. Each cage contained bedding and food and was covered by a heavy flat plastic lid equipped with ventilation holes and a bottle of water. The apparatus allowed the activity of 16 individual mice (eight of each genotype) to be monitored simultaneously. Each experimental cage contained one mouse. Motor response to novelty was monitored during a 3-h session that began immediately after placing the mouse into the cage. Data were analyzed in 10-min time bins. For long-term monitoring of locomotion, each mouse was habituated for 3 h. After adaptation, activity levels were monitored for 24 h, with data analyzed in 15-min bins.
Results
Pharmacology of Recombinant Receptors
Studies in HEK293 Cells Using Standard Application Methods. GABAARs harboring α1 subunit mutations of Ser270 to His and Leu277 to Ala [α1(S270H,L277A)β2γ2S] or only the Ser270 to His [α1(S270H)β2γ2S] or only the Leu277 to Ala [α1(L277A)β2γ2S] were studied by whole-cell patch clamp after transient expression in HEK293 cells. Brief applications of agonist produced concentration-dependent inward currents (Fig. 1, A-C). The maximal amplitudes and Hill coefficients were similar in wild-type and mutant receptors (Table 1).
The currents activated by at least seven concentrations of GABA were expressed as a fraction of the maximal GABA response, and these normalized data were pooled and fitted by a Hill equation (see Materials and Methods). Inspection of the data (Fig. 1E) revealed that the EC50 for the α1(S270H)β2γ2S mutant was decreased 5-fold compared with wild type, expressed as a significant left shift in the CRC. In contrast, the α1(L277A)β2γ2S mutant increased the EC50 17-fold, indicating that the receptor was significantly less sensitive to GABA than the wild-type receptor (Table 1).
As we hypothesized, introduction of both mutations into the same α1 subunit neutralized the opposing effects of both single mutants on the CRC. Figure 1, A and D, show typical current traces of the wild type and the α1(S270H, L277A)β2γ2S mutant. Maximally elicited currents and Hill coefficients were similar (Table 1). Although the single mutations change the EC50 significantly, the receptor including both mutations yielded an EC50 value that did not differ significantly from the wild-type GABAAR (Fig. 1F; Table 1).
To examine the effect of volatile anesthetics on the wild-type and the mutant receptor, HEK293 cells expressing α1β2γ2S or α1(S270H,L277A)β2γ2S GABAARs were exposed to the volatile anesthetics isoflurane and halothane. Clinically relevant concentrations of isoflurane (0.31 mM) increased the currents elicited by submaximal (EC20 concentration) GABA at the wild-type receptor by 100%. Another volatile anesthetic, halothane (0.21 mM), potentiated the GABA response to a lesser degree (36%) (Fig. 2, A and C). For the double mutant α1(S270H,L277A)β2γ2S receptor, isoflurane had no enhancing effect on the GABA response compared with the wild-type receptor. In contrast, the enhancing effect of halothane was not significantly decreased in the double mutant receptor compared with the wild-type receptor (23 and 36%, respectively; Fig. 2, B and C).
Studies in HEK293 Cells Using Rapid Application Methods. Application of 500-ms pulses of varying concentrations of GABA produced CRCs with EC50 values that substantially exceeded those obtained using standard drug application techniques (Table 1). Presumably, this was due to differences in exchange rates. The double mutation induced a significant 2-fold increase in the GABA EC50 of α1(S270H,L277A)β2γ2S compared with wild-type α1β2γ2S GABAARs (Table 1).
To analyze the deactivation kinetics, brief pulses of a high concentration of GABA (1 mM; 20 ms) were applied to wild-type and mutant GABAARs, in the absence or presence of 0.25 mM isoflurane (Fig. 3, A and B). The values for the decay time constants are shown in Fig. 3C. The double mutation caused receptors to deactivate approximately twice as fast as wild type under control conditions (p < 0.005, unpaired t test). Although isoflurane significantly increased the decay time constants in wild-type GABAARs (p < 0.005, paired t test), mutant receptors were no longer sensitive to modulation by isoflurane (p > 0.05, paired t test).
Studies in Xenopus Oocytes. The immobilizing and sedative effects of the intravenous anesthetic etomidate are mediated by β3- and β2-containing GABAARs, respectively (Jurd et al., 2003; Reynolds et al., 2003). Alcohols and volatile anesthetics seem to possess a larger range of targets than intravenous anesthetics, but the studies with etomidate suggest the possibility that GABAARs with different subunit composition may also underlie different effects by alcohols and volatile anesthetics. Therefore, in the present study, the mutated α1(S270H,L277A) was coexpressed in Xenopus oocytes with γ2S, and with either β2 or β3 subunits.
The simultaneous replacement of α1 Ser270 with His and of Leu277 with Ala produced no changes in the apparent GABA affinity compared with wild-type α1 when coexpressed with either β2γ2S (Fig. 4A; Table 1) or β3γ2S (Fig. 4C; Table 1). However, the maximal currents of both α1(S270H,L277A)β2γ2S (Fig. 4B) and α1(S270H,L277A)β3γ2S (Fig. 4D) were significantly decreased compared with the corresponding wild type.
Responses to the volatile anesthetic isoflurane were significantly decreased in α1(S270H,L277A)-containing receptors compared with wild-type controls. Isoflurane (0.3 mM) potentiation in α1(S270H,L277A)-containing receptors was only half of the enhancement observed in wild-type receptors (Table 2).
The responses to alcohols were eliminated or significantly decreased in double mutant α1-containing receptors. Analysis of the effect of ethanol on GABA responses in α1β2γ2S and α1(S270H,L277A)β2γ2S receptors (Fig. 5A) showed a significant effect of receptor (F1,52 = 11.54; p < 0.005), ethanol concentration (F4,52 = 15.44; p < 0.0001), and interaction (F4,52 = 16.36; p < 0.0001). A similar result was observed after the analysis of α1β3γ2S and α1(S270H,L277A)β3γ2S, (Fig. 5B) resulting in a significant effect of receptor (F1,40 = 251; p < 0.0001), ethanol concentration (F4,40 = 56.7; p < 0.0001), and interaction (F4,40 = 120; p < 0.0001). The effect of butanol on GABA responses in α1β2γ2S and α1(S270H,L277A)β2γ2S receptors was also tested. Butanol (11 mM) showed a robust potentiation in wild-type receptors and only a very small increase in α1(S270H,L277A)β2γ2S receptors (Table 2).
The modulation of GABA responses by 10 μM Zn2+ was slightly decreased in α1(S270H,L277A)β2γ2S compared with α1β2γ2S receptors, whereas the modulation by 1 μM flunitrazepam showed no differences. However, there were no changes in the Zn2+ or flunitrazepam effects in α1(S270H,L277A)β3γ2S compared with α1β3γ2S GABAARs (Table 2). Pentobarbital (50 μM) potentiation was decreased in α1(S270H,L277A)-containing receptors (by -32 and -34% in β2- and β3-containing receptors, respectively; Table 2). Etomidate can act either as a modulator or an agonist of GABAARs, depending on the concentration. At a low concentration (1 μM), etomidate effects were decreased in α1(S270H,L277A)β2γ2S and α1(S270H,L277A)β3γ2S compared with their respective wild-type receptors. At higher concentrations (10 μM), etomidate can activate the GABAAR in the absence of GABA. When 10 μM etomidate was applied to α1β2γ2S and α1(S270H,L277A)β2γ2S receptors, the responses observed were not different, but the current induced by 10 μM etomidate was significantly decreased in α1(S270H,L277A)β3γ2S compared with α1β3γ2S receptors (Table 2).
Characterization of Knockin Mice
We used the targeting strategy depicted in Fig. 6, A and B, to modify the α1 locus in embryonic stem cells. This mutant locus was transferred through the germline, the neomycin selection marker was deleted by site-specific recombination, and homozygous wild-type (SL/SL) and homozygous knockin (HA/HA) mice were produced and analyzed. To verify that the knockin mutation was expressed, and no other unintended mutations were present in the α1 gene, brain α1 mRNA was amplified by RT-PCR and subsequently analyzed by DNA sequence analysis. As expected, only the intended mutations were present in the α1 gene product of HA/HA mice (Fig. 6C). GABAAR number and binding affinity were unchanged by the mutation as assessed with [3H]flunitrazepam binding (BMAX = 1744 ± 66 versus 1749 ± 74 fmol/mg protein; KD = 4.62 ± 0.71 versus 5.90 ± 0.96 nM; SL/SL versus HA/HA, respectively). Immunohistochemistry demonstrated normal distribution of α1 protein (data not shown). Western blot of α1 protein revealed normal amounts of α1 in cerebellum of HA/HA mice (94 ± 9%) compared with SL/SL controls (100 ± 11%; Fig. 6D). However, Western blot analysis of cortex revealed an unexpected decrease in the amount of α1 present in HA/HA mice compared with SL/SL (Fig. 6E). This analysis also revealed significant changes in the abundance of other GABAAR subunits in cortex, including an increase in α3, β2, and γ2, and a decrease in β3 (Fig. 6E). Analysis of genotype distributions at weaning indicated the expected (Mendelian) production of HA/HA mice (n = 101) at approximately the same frequency as SL/SL controls (n = 106).
We noticed a few (<10%) premature deaths of HA/HA mice between 7 and 12 weeks of age, but otherwise HA/HA mutants had normal life spans of well over a year of lifetime, indistinguishable from SL/SL littermates. Otherwise, the knockin mice were normal in size, appearance, and overt behavior. HA/HA mutant mice did not show any obvious impairments compared with SL/SL littermates with respect to cage behavior, growth, weight, and fertility. HA/HA mice showed the typical pattern of light cycle-dependent activity, with wheel running and food consumption during the dark cycle (see data presented in next section). When observed in their home cages, HA/HA mice did not display any obvious behavioral abnormalities, such as spontaneous seizures; running in circles, or other repetitive motions; aggression toward cage mates; or any increases or decreases in general locomotion compared with SL/SL littermates. There were no differences with respect to growth rate, body length, or body weight between HA/HA and SL/SL animals. For example, average body weight of male mice at 8 weeks of age was 24.8 ± 1.0 g for SL/SL and 23.3 ± 1.0 g for HA/HA mice and at 12 weeks of age was 27.8 ± 0.8 for SL/SL and 29.0 ± 1.0 for HA/HA mice. HA/HA mice show the normal (hunched) posture and gait as well as normal neurological reflexes such as righting reflex, postural reflex, and whisker-orienting reflex. Breeder pairs of mutant males and females generated litters of normal sizes (58 pups) and at the expected frequency (every 34 weeks). Pups were cared for in a manner indistinguishable from SL/SL breeders with respect to nesting, grooming, and feeding. In addition, we found no differences in basal motor coordination, anxiety-like behavior, or thermal nociception (our unpublished observations).
The normalcy of the double knockin mice starkly contrasts with the abnormalities observed in mice that bear only the Ser270 to histidine mutation. Mice heterozygous for that single mutation are hyper-responsive, exhibit muscle tremors, have reduced body size, reduced motor coordination, are hypoactive in the home cage, hyperactive in the open field (Homanics et al., 2005), and have EEG abnormalities and anesthetic-induced seizure-like activity (Elsen et al., 2006). The double mutant mice reported here have none of these abnormalities. Thus, inclusion of the Leu277 to Ala mutation seems to have corrected the underlying cause of the behavioral abnormalities induced by the single mutation. The most likely explanation is the restoration of GABA sensitivity to near normal.
Motor Activity
Wild-type and mutant mice demonstrated typical transient motor responses to novel situations (F17,396 = 17.7; p < 0.0001, dependence on time). However, the motor activity levels of knockin HA/HA mutant mice were consistently higher than those of wild-type animals (F1,396 = 7.3; p < 0.01, dependence on genotype) (Fig. 7A).
After 24 h of monitoring activity, main effects of genotype (F1,2288 = 81.9; p < 0.0001) and time (F103,2288 = 8.9; p < 0.0001) with strong interactions (F103,2288 = 1.5; p < 0.001) were observed (Fig. 7B). HA/HA knockin mutant mice were more active, both in the light and dark phases.
Discussion
This study showed that a GABAAR harboring two specific mutations within the α1 subunit and expressed in heterologous systems exhibited a near-normal GABA sensitivity but had a differential response to two commonly used volatile anesthetics. Isoflurane enhanced submaximal GABA currents in wild-type α1β2γ2S GABAARs but failed to potentiate them in receptors containing the double mutant α1(S270H,L277A). In contrast, halothane enhanced the GABA responses in both the wild-type and mutant receptors. In addition, when GABAARs containing α1(S270H,L277A) were expressed in Xenopus oocytes, they failed to show any ethanol-induced potentiation of the GABA responses.
Ser270 in the second transmembrane region of the α subunit of the GABAAR seems to be important for the binding of volatile anesthetics (Koltchine et al., 1999; Mascia et al., 2000). Mutation of Ser270 to amino acids of various side chain volumes has been shown to affect the GABA EC50 and anesthetic action. The differences in size between isoflurane (144 Å3) and halothane (110 Å3) could explain the differences in enhancing effects on the mutant GABAAR: replacing Ser270 with His allowed potentiation by the smaller volatile anesthetic halothane but not by the larger anesthetic isoflurane. When Ser was replaced with even larger amino acids (e.g., tryptophan), the enhancing effects of halothane were abolished (Nishikawa et al., 2002). Substitution of Ser270 with a larger amino acid also decreased alcohol potentiation (Mihic et al., 1997; Ueno et al., 1999; Findlay et al., 2000). In α1(S270H,L277A)-containing GABAARs, ethanol potentiation was abolished, and butanol potentiation in α1(S270H,L277A)β2γ2S was greatly decreased. The absence or marked decrease of alcohol potentiation was most likely due to the Ser270 to His mutation in the alcohol and volatile anesthetic binding site.
Additional GABAAR modulators tested on the wild-type and mutant receptors expressed in Xenopus oocytes were Zn2+ (endogenous modulator), flunitrazepam (benzodiazepine), and pentobarbital and etomidate (intravenous anesthetics). The Zn2+ inhibition of currents through the GABAAR is greatly reduced if the γ subunit is present in the receptor (Draguhn et al., 1990; Smart et al., 1991). Zn2+ inhibition was minimal in all the expressed receptors, indicating expression of γ2S. A small decrease in Zn2+ inhibition was observed in α1(S270H,L277A)β2γ2S but not in α1(S270H,L277A)β3γ2S. The flunitrazepam-induced potentiation was not altered in α1(S270H,L277A)-containing receptors, suggesting that the level of expression of γ2S was not modified in either combination. Discrete Zn2+ binding sites have been described in the GABAARs (Hosie et al., 2003). The small difference in the Zn2+ sensitivity between wild-type and mutant receptors could be due to an allosteric effect of the α1(L277A) mutation, which is near to a Zn2+ binding site, or the α1(S270H) mutation, that could be introducing a new Zn2+ binding site by placing a H in a water-filled pocket. The pentobarbital-induced potentiation of GABA responses was reduced by approximately 30% in receptors containing the double mutation; however, there is no indication of any of these mutations being located in a pentobarbital binding site, and the observed change is likely due to an allosteric effect. The β subunit of the GABAAR is crucial for etomidate action: etomidate is approximately 10 times more potent at β2/3-containing receptors than at β1-containing receptors due to the variable residue at 286, which is either Asn or Ser (Belelli et al., 1997; Hill-Venning et al., 1997), and substitution of Asn in β2/3 can eliminate etomidate potentiation (Jurd et al., 2003; Reynolds et al., 2003). The etomidate-induced potentiation of GABA responses was decreased in α1(S270H,L277A)-containing receptors, whereas the etomidate-induced currents through the GABAAR were decreased only in α1(S270H,L277A)β3γ2S receptors. This suggests that the conformational changes necessary for etomidate action are impeded by the mutations in α1, even though they do not modify the action of other agonists (GABA) or modulators (flunitrazepam). In summary, the effect of most of the anesthetics tested was decreased (pentobarbital and etomidate, isoflurane) or abolished (ethanol) in the α1(S270H,L277A)-containing receptors expressed in Xenopus oocytes.
When GABAARs expressed in Xenopus oocytes and HEK293 cells were tested using standard solution exchange techniques, the double mutation α1(S270H,L277A) did not affect the GABA sensitivity. However, when tested in HEK293 cells using rapid application methods, α1(S270H, L277A)β2γ2S GABAARs showed a lower sensitivity to GABA and a faster rate of deactivation at the termination of the GABA pulse than wild type. Nevertheless, the effect of isoflurane was substantially reduced in the double mutant when tested by rapid application. We do not have a ready explanation for the influence of solution exchange rates on EC50, or why this differs in degree for the mutant versus wild-type receptors. Nevertheless, it is apparent that the characteristics of the GABA transient influence the impact of the mutation on receptor activation. One implication of this finding is that the effect of the double mutation on drug-free responses at receptors in situ might depend on the conditions of receptor activation, with little or no influence on tonic currents produced by low concentrations of GABA, but with changes in synaptic currents produced by transient receptor activation, and the elimination of isoflurane sensitivity under both conditions. This prediction can now be tested, since the mice that bear the double mutation are viable.
Mice bearing the double mutation Ser270 to His and Leu277 to Ala in the GABAAR α1 subunit were successfully generated, and with no overt abnormal characteristics. In biochemical analysis, wild-type and knockin mice presented no differences; for instance, flunitrazepam binding was similar between wild-type and knockin mice, and the α1 protein distribution was normal in brain of HA/HA mice. However, differences in receptor subunit levels in cortex, and locomotor activity were observed between SL/SL and HA/HA mice. The maximal currents of wild-type and double mutant receptors did not differ when the receptors were expressed in HEK293 cells, but there was a decrease in the maximal currents for the double mutant α1(S270H,L277A)β2/3γ2S expressed in Xenopus oocytes. Different expression systems may reflect more or less accurately the receptor characteristics in vivo. Even though the current amplitudes observed in oocytes better reflected the effects of the mutation on protein levels in cortex (the levels of mutant α1 protein were decreased by 48% in HA/HA mice compared with wild-type SL/SL mice), and these results are consistent with decreased protein levels in the oocyte expression system, they do not definitively address this issue. The flunitrazepam binding was not modified in HA/HA mice, probably because of compensatory changes in other subunits (the levels of α3 were increased by 44%, whereas α2 showed an increase of 34%).
We found that HA/HA knockin mutant mice demonstrated a hyperactive behavioral phenotype. These mice exhibited a slightly greater motor response to novelty; this response was transient, and after 150 to 180 min the activity of mutant and wild-type mice was similar. Over 24 h of continuous monitoring of spontaneous locomotion, HA/HA knockin mutant mice demonstrated normal circadian motor rhythm, but were more active than control mice. GABAARs are known to be important for regulation of exploratory activity (File, 1985), and mice lacking GABAAR α1 or α2 subunits demonstrated reduction of spontaneous locomotion (Blednov et al., 2003; Boehm et al., 2004). In contrast, mice lacking the GABAAR β2 subunit showed high levels of locomotor activity (Sur et al., 2001; Blednov et al., 2003), similar to the β3 knockout mice (DeLorey et al., 1998). Thus, the reduction of α1 protein found in the HA/HA mice is unlikely to account for the motor activity phenotype.
In conclusion, we were able to design a GABAAR containing a mutant α1 subunit that maintained a near-normal GABA response in vitro but showed a differential response to volatile anesthetics and alcohols. This receptor provides a promising candidate for an in vivo model of altered volatile anesthetic sensitivity, because it did not markedly affect the normal response to GABA but showed a decreased response to isoflurane in both HEK293 cells and Xenopus oocytes, and a normal potentiation by halothane in HEK293 cells. It will also be useful in the study of alcohol effects through α1-containing GABAARs, since the double mutant showed in-sensitivity to ethanol. The homozygous knockin mice engineered with this mutation presented no striking differences from the wild-type mice except for behavioral hyperactivity. These knockin mice are a unique resource that will allow for an unprecedented dissection of the role of the α1 subunit of the GABAAR in clinically relevant whole-animal behavioral responses to volatile anesthetics and ethanol. For example, if α1-containing GABAARs are important mediators of anesthetic action, then knockin mice bearing the double mutation should resist the anesthetizing effects of isoflurane but not halothane.
Acknowledgments
G.E.H. thanks Carolyn Ferguson and Edward Mallick for technical support. C.M.B. and R.A.H. thank Kathryn L. Carter and Rachel Phelan for the oocyte harvesting. We thank Drs. Edmond I. Eger II and James R. Trudell for helpful suggestions and fruitful discussions.
Footnotes
-
This study was supported by National Institutes of Health Grants T32-MH18273, GM47818, GM55719, AA10422, AA06399, and GM45129; Integrative Neuroscience Initiative on Alcoholism Consortium AA13520; and the Waggoner Center for Alcohol and Addiction Research.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.106.104406.
-
ABBREVIATIONS: GABAAR, γ-aminobutyric acid type A receptor; HEK, human embryonic kidney; CRC, concentration-response curve; HA/HA, homozygous knockin mice; SL/HA, heterozygous knockin mice; SL/SL, control wild-type mice; RT-PCR, reverse transcription-polymerase chain reaction; ANOVA, analysis of variance.
- Received March 13, 2006.
- Accepted June 26, 2006.
- The American Society for Pharmacology and Experimental Therapeutics