Tryptophan and Cysteine Mutations in M1 Helices of α1β3γ2L γ-Aminobutyric Acid Type A Receptors Indicate Distinct Intersubunit Sites for Four Intravenous Anesthetics and One Orphan Site

Abstract

Background: γ-Aminobutyric acid type A (GABAA) receptors mediate important effects of intravenous general anesthetics. Photolabel derivatives of etomidate, propofol, barbiturates, and a neurosteroid get incorporated in GABAA receptor transmembrane helices M1 and M3 adjacent to intersubunit pockets. However, photolabels have not been consistently targeted at heteromeric αβγ receptors and do not form adducts with all contact residues. Complementary approaches may further define anesthetic sites in typical GABAA receptors.

Methods: Two mutation-based strategies, substituted tryptophan sensitivity and substituted cysteine modification-protection, combined with voltage-clamp electrophysiology in Xenopus oocytes, were used to evaluate interactions between four intravenous anesthetics and six amino acids in M1 helices of α1, β3, and γ2L GABAA receptor subunits: two photolabeled residues, α1M236 and β3M227, and their homologs.

Results: Tryptophan substitutions at α1M236 and positional homologs β3L231 and γ2L246 all caused spontaneous channel gating and reduced γ-aminobutyric acid EC50. Substituted cysteine modification experiments indicated etomidate protection at α1L232C and α1M236C, R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid protection at β3M227C and β3L231C, and propofol protection at α1M236C and β3M227C. No alphaxalone protection was evident at the residues the authors explored, and none of the tested anesthetics protected γ2I242C or γ2L246C.

Conclusions: All five intersubunit transmembrane pockets of GABAA receptors display similar allosteric linkage to ion channel gating. Substituted cysteine modification and protection results were fully concordant with anesthetic photolabeling at α1M236 and β3M227 and revealed overlapping noncongruent sites for etomidate and propofol in β-α interfaces and R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid and propofol in α-β and γ-β interfaces. The authors' results identify the α-γ transmembrane interface as a potentially unique orphan modulator site.

Figure 1. GABA_A receptor transmembrane inter-subunit anesthetic sites

The diagram depicts the arrangements of α1 (yellow), β3 (blue), and γ2L (green) subunits and each subunit’s transmembrane four-helix bundle (M1 to M4). The ‘+’ and ‘–’ interfacial surfaces of each subunit, corresponding respectively to M3 and M1 aspects, are identified. The approximate position of photolabeled residues, α1M236 and β3M227 are labeled in magenta, while their homologs on other subunits (see Table 1) are labeled in black. We also depict the hypothesized inter-subunit sites for etomidate (red rhombi), mTFD-MPAB (R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; green rectangles), and propofol (white hexagons), based on photolabeling data.

Figure 2. GABA concentration-responses and anesthetic-shifts in α1-M1 and β3-M1 tryptophan substituted GABA_A receptor mutants

The top panel shows wild-type peak current data (mean ± sem) for GABA alone (black circles), combined with 3.2 µM etomidate (red triangles), or combined with 8 µM mTFD-MPAB (R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; green diamonds). Normalization for control responses was to maximal GABA and to GABA plus anesthetic for anesthetic shift studies. The four other panels show similar data for two tryptophan-substituted α1-M1 mutants and two β3-M1 mutants (labeled in each panel). Fitted GABA EC₅₀s and EC₅₀ shifts in the presence of anesthetics are summarized in Table 2.

Figure 3. Anesthetic EC5 enhancement in wild-type vs α1-M1 and β3-M1 substituted tryptophan GABA_A receptor mutants

The bar-graph depicts GABA EC5 enhancement ratios (mean ± sem) for five receptor types (x-axis labels) and four equi-potent anesthetic solutions: 3.2 µM etomidate (ETO; red), 8 µM mTFD-MPAB (R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; green), 5 µM propofol (PRO; white), and 2.5 µM alphaxalone (ALFAX; purple). The number of oocytes studied for each interaction is indicated by the numbers in each bar. Each drug’s effect in mutants was compared to the same drug effect in wild-type (by two-way ANOVA). Differs from wild-type at ** p < 0.01, *** p < 0.001.

Figure 4. Substituted cysteine modification and anesthetic protection in α1M236Cβ3γ2L receptors

The bar graph summarizes pCMBS modification rate data in the presence of GABA alone and in the presence of four anesthetics (x-axis labels: ALX is alphaxalone; ETO is etomidate; MPAB is R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; PRO is propofol). We have previously published data showing the functional effect of pCMBS modification in this mutant and protection by etomidate and propofol ^,. Color coding by drug is the same as in Figure 3 and the number of cells studied for each condition is indicated by the numbers in each bar. This mutant is characterized by low GABA efficacy (Table 2) and pCMBS modification in the presence of alphaxalone represents a control modification condition with high channel open probability matching that in the presence of other anesthetics . ** Differs from the rate with pCMBS + GABA + alphaxalone at p < 0.01 (one way ANOVA).

Figure 5. Substituted cysteine modification and anesthetic protection in α1L232Cβ3γ2L receptors

Panel A shows a series of current sweeps recorded from a single oocyte expressing α1L232Cβ3γ2L receptors before and after a series of 10 s exposures to p-chloromercuribenzenesulfonate (pCMBS) + GABA (arrows). Red traces show current elicited by 3 µM GABA (~EC5) and the black traces show current elicited by 1 mM GABA (black bars above traces indicate GABA application). The final sweeps depict the effect of full modification by 10 µM pCMBS + GABA for 20 s (asterix and arrow). Panel B shows normalized low:high GABA response ratios and linear least squares fits for individual oocytes from all control modification studies (black symbols and lines) and in the presence of 3.2 µM etomidate (red symbols and lines). Panel C is a bar graph summarizing α1L232Cβ3γ2L modification rate data in control studies and in the presence of four anesthetics (x-axis labels: ALX is alphaxalone; ETO is etomidate; MPAB is R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; PRO is propofol). The number of cells studied for each condition is indicated by the numbers in each bar. *Differs from pCMBS + GABA at p < 0.05.

Figure 6. Substituted cysteine modification and anesthetic protection in α1β3M227Cγ2L and α1β3L231Cγ2L receptors

Panel A shows a series of traces recorded from a single oocyte expressing α1β3M227Cγ2L receptors. Currents were elicited with 2 mM GABA (black bars above traces indicate GABA applications) before and after a series of 10 s exposures to GABA + 100 µM p-chloromercuribenzenesulfonate (pCMBS; arrows). The final trace was recorded after modification with GABA + 1 mM pCMBS for 20 s. Panel B shows normalized peak current data and linear rate analyses for cells modified with pCMBS + GABA (black symbols and lines) and cells modified with pCMBS + GABA + 8 µM mTFD-MPAB (green symbols and lines). Panel C is a bar graph summarizing α1β3M227Cγ2L modification rate data in the absence and presence of four anesthetics (x-axis labels: ALX is alphaxalone; ETO is etomidate; MPAB is R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; PRO is propofol). Panel D shows a series of traces recorded from a single oocyte expressing α1β3L231Cγ2L receptors. Currents were elicited both 2 µM GABA (red traces) and 1 mM GABA (black traces) before and after a series of 10 s exposures to GABA + 25 µM pCMBS (arrows). Black bars above traces indicate GABA applications. The final trace was recorded after modification with GABA + 250 µM pCMBS for 20 s. Panel E shows normalized low:high GABA response ratios and linear least squares rate analyses for cells modified with pCMBS + GABA (black symbols and lines) and cells modified with pCMBS + GABA + 8 µM mTFD-MPAB (green symbols and lines). Panel F is a bar graph summarizing α1β3L231Cγ2L modification rate data in the absence and presence of four anesthetics (x-axis labels). In panels C and F, results statistically differing from pCMBS + GABA control are * p < 0.05 and ** p < 0.01.

Figure 7. GABA concentration-responses and anesthetic modulation of α1β3γ2I242W and α1β3γ2L246W receptors

Panel A shows peak current results (mean ± sem) for GABA-dependent activation of α1β3γ2I242W receptors with GABA alone (black circles) and in the presence of 3.2 µM etomidate (red triangles). Both data sets are normalized to 3 mM GABA responses, illustrating the low efficacy of GABA alone. Lines through data represent logistic fits (Eq. 1, methods). Fitted EC₅₀ and shift results are reported in Table 2. Panel B shows peak current results (mean ± sem) for GABA-dependent activation of α1β3γ2L246W receptors with GABA alone (black circles) and in the presence of 3.2 µM etomidate (red triangles). Both data sets are normalized to 1 mM GABA responses. Lines through data represent logistic fits. Fitted results are reported in Table 2. Panel C is a bar graph summarizing GABA EC5 enhancement results for both α1β3γ2I242W and α1β3γ2L246W receptors. At equipotent concentrations, none of the anesthetics (x-axis labels: ALX is alphaxalone; ETO is etomidate; MPAB is R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; PRO is propofol) differ in their modulation of these two mutant receptors.

Figure 8. Substituted cysteine modification and anesthetic protection in α1β3γ2I242C and α1β3γ2L246C receptors

Panel A shows a series of traces recorded from a single oocyte expressing α1β3γ2I242C receptors. Currents were elicited with 3.5 µM GABA (red traces) and 1 mM GABA (black traces) before and after a series of exposures to GABA + 10 µM p-chloromercuribenzenesulfonate (pCMBS; arrows). Black bars above traces indicate GABA applications. The final trace was recorded after modification with GABA + 100 µM pCMBS for 20 s. Panel B shows normalized low:high GABA response ratios and linear least squares rate analyses for cells modified with pCMBS + GABA. Panel C is a bar graph summarizing α1β3γ2I242C modification rate data in the absence and presence of four anesthetics (x-axis labels: ALX is alphaxalone; ETO is etomidate; MPAB is R-5-allyl-1-methyl-5-(m-trifluoromethyl-diazirinylphenyl) barbituric acid; PRO is propofol). Panel D shows a series of traces recorded from a single oocyte expressing α1β3γ2L246C receptors. Currents were elicited both 1.5 µM GABA (red traces) and 1 mM GABA (black traces) before and after a series of exposures to GABA + 500 µM pCMBS (arrows). Black bars above traces indicate GABA applications. The final trace was recorded after modification with GABA + 1 mM pCMBS for 60 s. Panel E shows normalized low:high GABA response ratios and linear least squares rate analyses for cells modified with pCMBS + GABA. Panel F is a bar graph summarizing α1β3γ2L246C modification rate data in the absence and presence of four anesthetics (x-axis labels). None of the rates with anesthetics differed significantly from control.