Abstract
Barbiturate sensitivity was imparted to the human ρ1homooligomeric γ-aminobutyric acid (GABA) receptor channel by mutation of a tryptophan residue at position 328 (Trp328), which is located within the third transmembrane domain. Substitutions of Trp328 with a spectrum of amino acids revealed that nearly all hydrophobic residues produced receptor channels that were both directly activated and modulated by pentobarbital with similar sensitivities. Previous studies with ligand-gated ion channels (including GABA) have demonstrated that even conservative amino acid substitution within the agonist-dependent activation domain (N-terminal extracellular domain) can markedly impair agonist sensitivity. Thus, the lack of significant variation in pentobarbital sensitivity among the Trp328 mutants attests to an intrinsic difference between pentobarbital- and the GABA-dependent activation domain. Compared with the heterooligomeric αβγ receptor channel, the mode of modulation for homooligomeric Trp328 mutants by pentobarbital was more dependent on the GABA concentration, yielding potentiation only at low concentrations of GABA (fractions of their respective EC50 values), yet causing inhibition at higher concentrations. Agonist-related studies have also demonstrated that residue 328 plays an important role in agonist-dependent activation, suggesting a functional interconnection between the GABA and pentobarbital activation domains.
Molecular events leading to anesthesia have been attributed to the modulation of the excitatory or inhibitory ligand-gated ion channels (Nicoll, 1972;Nicoll et al., 1975; Barker and Ransom, 1978; Franks and Lieb, 1994). The primary target for a number of anesthetic compounds, including pentobarbital, are the γ-aminobutyric acid type A (GABAA) receptor channels (Schulz and Macdonald, 1981; Gage and Robertson, 1985; Parker et al., 1986; MacIver et al., 1991; Tanelian et al., 1993), the key components of synaptic inhibition in the central nervous system. GABA-gated chloride channels are heterooligomeric or homooligomeric pentamers composed of numerous combinations of homologous α, β, γ, δ, or ρ classes of subunits (Macdonald and Olsen, 1994). These subunits have the potential for creating a great number of GABA receptor channels with distinct pharmacology (Macdonald and Olsen, 1994). For example, GABA-evoked currents for heterooligomeric αβγ (GABAA;Schofield et al., 1987; Levitan et al., 1988; Macdonald and Olsen, 1994) and homooligomeric β receptor channels (Blair et al., 1988;Sanna et al., 1995) are potentiated in the presence of low concentrations of pentobarbital (modulatory action). Moreover, pentobarbital at higher concentrations can directly activate these channels (agonistic action; Mathers and Barker, 1980; Nicoll and Wojtowicz, 1980; Akaike et al., 1991; Sanna et al., 1995; Rho et al., 1996). In contrast, the homooligomeric ρ1receptor channel (GABAC; Cutting et al., 1991) is insensitive to pentobarbital (Shimada et al., 1992).
In this study, experiments were conducted to gain insight into the mechanisms of barbiturate modulation and activation of GABA-gated chloride channels. Results with ρ-β chimeras and site-directed mutagenesis of ρ1 indicate that hydrophobic amino acid substitution for Trp328 within the third transmembrane domain (TM3) imparts modulatory and agonistic properties of pentobarbital to ρ1 homooligomeric receptor channel. In addition, residue 328 plays an important role in agonist-dependent activation. Collectively, these results provide important clues concerning the mechanism of barbiturate action on GABA-gated ion channels.
Materials and Methods
All chimeras were constructed using either conserved restriction sites between the ρ1 and β2 subunits (e.g., HincII for ρ346/β305), or synthetic oligonucleotides containing designed restriction enzyme sites and polymerase chain reaction. Special care was taken not to alter the relative position of the conserved amino acids on both sides of the junction (with the exception of ρ405/β399). The DNA sequence of all chimeras was verified by DNA sequencing.
The cDNAs corresponding to ρ1 and β2 were cloned into the pSELECT vector (Promega, Madison, WI) and oligonucleotide-mediated site-directed mutagenesis was achieved according to the manufacturer’s protocol (Altered Sites; Promega). Successful mutagenesis was verified by DNA sequencing. The cDNAs were linearized with NheI leaving a several-hundred base pair tail (3′). These additional sequences at the 3′ end may increase cRNA stability in the oocyte. The cRNA was transcribed from the linearized cDNAs by standard in vitro transcription procedures (Megascript; Ambion, Austin, TX).
Xenopus laevis (Xenopus I; Ann Arbor MI) were anesthetized by hypothermia and oocytes were surgically removed from the frog and placed in Oocyte Ringer (OR2) that consisted of: 82.5 mM NaCl, 2.5 mM KCl, 10 mM HEPES, 1 mM CaCl2, 1 mM MgCl2, 1 mM Na2HPO4, 50 U/ml penicillin, and 50 μg/ml streptomycin, pH 7.5. Oocytes were dispersed and then incubated in OR2 minus Ca2+ plus 0.3% collagenase A (Boehringer Mannheim, Indianapolis, IN) for approximately 2 h. After isolation, the oocytes were thoroughly rinsed with OR2. Stage VI oocytes were separated and maintained overnight at 18°C.
Micropipettes for injecting cRNA were fabricated on a Narishige PP-83 puller (Narishige USA, Greenvale, NY) and the tips were cut off with microscissors. The cRNA in diethylpyrocarbonate-treated water was drawn up into the micropipette with negative pressure and then injected into the oocytes by applying positive pressure using a Picospritzer II (General Valve Corporation, Fairfield, NJ). The oocytes were incubated in OR2 at 18°C for 2 to 3 days before the experiment. To ensure that equal concentrations of cRNA for each construct were injected (especially important for comparison of maximum GABA-activated currents), set dilutions of cRNA from mutants were electrophoresed on a 1% formaldehyde-containing agarose gel. The amount of cRNA was judged and matched by interpolation of lanes containing different dilutions of the corresponding cRNA. In addition, for nearly all mutants, two independent isolates were characterized and tested.
Two to 3 days after injection, oocytes were placed on a nylon mesh suspended in a small volume chamber (∼75 μl). The chamber has an inlet in the top and an outlet in the bottom that allows continuous and rapid perfusion. Twenty separate reservoirs (100-ml glass containers) were connected to four six-way valves and the outlet of each of these six-way valves (the sixth position was connected to the reservoir containing the control solution) was connected to one four-way valve. The outlet of the four-way valve lead to the chamber. In this way, up to 20 different solutions could be introduced to an individual oocyte. Switching between the different solutions was controlled manually. The oocyte was continuously perfused with recording OR2 (OR2 without antibiotics and the 1 mM Na2HPO4 replaced with 1 mM NaCl) and briefly switched to the test solution containing drug.
Recording microelectrodes were fabricated with a Narishige PP-83 puller and filled with 3 M KCl. Electrodes with resistances of 0.6–1 MΩ were used. Standard two-electrode voltage-clamp techniques (Turbo TEC-05 npi; Adams and List, Westbury, NY) were used to record currents in response to application of drugs. In all cases, membrane potential was clamped to −70 mV. Data were played out on a Gould EasyGraf chart recorder (Gould Inc., Glen Burnie, MD) during the experiment and recorded on a VCR (Instrutech PA10b; Instrutech Labs., Plymouth Meeting, PA) for off-line analysis.
The EC50, and Hill numbers were estimated by fitting the concentration-response relationships to the following equation: [I = Imax/(1 [EC50/(A)]n )] using computer software provided by Dr. David S. Weiss, where I is the peak current at a given concentration of agonist A,Imax is the maximum current, EC50 is the concentration of agonist yielding a current half the maximum, and n is the Hill coefficient.
Results
TM3 of β2 Subunit Is Sufficient To Impart Pentobarbital Sensitivity to ρ1.
To determine the crucial domain(s) for the dual agonistic and modulatory action of pentobarbital, chimeric human ρ1 and rat β2 subunits were constructed. The cRNA from the different ρ-β chimeras were expressed in Xenopus oocytes and the responses of these receptor channels were electrophysiologically recorded in the presence of GABA, pentobarbital, and a combination of both. The summary of these experiments is depicted in Fig. 1A (see also Tables1 and 2). The most striking result was the role of the TM3 from the β2 subunit in conferring pentobarbital sensitivity (compare the ρ324/β283 and ρ346/β305). The ρ1 receptor channel containing both the TM3 and the TM4 from the β2subunit displayed marked sensitivity to pentobarbital. In contrast, deletion of the sequences corresponding to the TM3 of the β2 within the ρ324/β283 chimera and replacement with the TM3 of the ρ1 subunit abolished pentobarbital sensitivity in the resulting receptor channels (Fig. 1A, ρ346/β305 and ρ405/β399).
Determination of crucial residue in conferring pentobarbital-sensitivity to ρ1 receptor channel. A, chimeras between human ρ1 and rat β2 subunits were constructed. Special care was taken not to alter relative position of conserved amino acids on both sides of junction (with exception of ρ405/β399). cRNAs from different ρ-β chimeras were expressed in Xenopus oocytes and resulting receptor channels were examined using GABA (up to 20 mM), pentobarbital (up to 2.5 mM), or both; * indicates spontaneously open channels (see Results). In these channels (*), chloride leak (judged based on reversal potential for chloride) was directly proportional to amount of injected cRNA (data not shown). β2 receptor channels had a severe depression in maxima when tested with GABA alone. All receptor channels that responded to GABA or pentobarbital or both are scored with +. All numbers indicate amino acid position for ρ1 and β2 subunit, respectively. Many constructed chimeras did not yield functional channels (data not shown). It is important to note that rat β2 subunit is highly suited to form heterooligomeric receptor channels. This is corroborated by nearly three orders of magnitude greater maximal current when equivalent cRNA concentrations for β2 subunit is coexpressed with α and γ subunit. B, TM3 (but not TM1) of β2 subunit confers pentobarbital sensitivity to ρ1 receptor channel. C, alignment of amino acid sequences corresponding to TM3 of human ρ1, rat β2, α1, γ2, andDrosophila (DRC) GABA subunits. Boxed residues represent amino acids that were mutated in this study. D, pentobarbital-dependent modulation of GABA responses from ρ1 and ρW328M receptor channels. Mutation of Trp328 to Met, imparted pentobarbital sensitivity to ρ1 receptor channel. For ρW328M receptor channel, GABA (0.3 μM) responses were markedly potentiated in presence of 50 μM pentobarbital. In contrast, GABA currents (0.6 μM) for ρ1 receptor channel were not altered in presence of equivalent concentration of pentobarbital. Thick line above each current trace represents duration of GABA application or coapplication of GABA and pentobarbital.
Parameters determined from fitting Hill equation to GABA concentration-response relationships. Numbers in parentheses indicate number of oocytes tested. Concentration of GABA required for half-maximal activation (EC50) and Hill coefficient are mean ± S.D.
Parameters determined from fitting Hill equation to pentobarbital concentration-response relationships. Numbers in parentheses indicate number of oocytes tested. For all pentobarbital-sensitive mutants, 2.5 mM pentobarbital was used (with exception of ρY198S/W328M and ρY198S/WVS328–330MGC receptor channels, where 1 mM pentobarbital was used) to obtain pentobarbital Imax (PbImax). Concentration of pentobarbital required for half-maximal activation (EC50), Hill coefficient and PbImax/GABA Imax are mean ± S.D.
To determine whether both the TM3 and the TM4 of the β2 subunit are needed to confer pentobarbital sensitivity, or whether the TM3 of the β2subunit alone is sufficient to impart pentobarbital sensitivity to the ρ1 receptor channels, the TM3 of the ρ1 subunit was replaced with the equivalent domain of the β2 subunit (Fig. 1B, ρ324/β283–304/ρ347). The expression of the cRNA for this chimera yielded a receptor channel highly sensitive to pentobarbital (Table2; EC50 = 77.2 ± 0.5 μM). For comparison, TM1 from the ρ1subunit was replaced with the equivalent domain from the β2 subunit. The expression of this chimera, however, produced a receptor channel that was insensitive to pentobarbital (Fig. 1B and Table 2, ρ268/β227–242/ρ285). These results indicate the importance of the amino acid sequence within the TM3 of the β2 subunit in imparting pentobarbital sensitivity to the ρ1 receptor channel.
Methionine Substitution for Trp328 Within TM3 of ρ1Confers Pentobarbital Sensitivity.
Comparison of the amino acid sequences encoding the TM3 of ρ1 and β2 subunits revealed nonconserved differences with respect to size and hydrophobicity mainly at three positions, Trp328, Val329, and Ser330. The corresponding residues within the β2 subunit were Met286, Gly287, and Cys288 (Fig. 1C). Using site-directed mutagenesis, the Trp328, Val329, and Ser330, within the ρ1 subunit, singly or in combination, were mutated to Met, Gly, and Cys, respectively. All mutant receptor channels that included Trp328 to Met substitution (ρW328M, ρWV328,329 MG, and ρWVS328–330 MGC) were sensitive to pentobarbital. For these mutant receptor channels, pentobarbital at low concentrations mediated the potentiation of the GABA responses (see below, e.g., Fig. 1D) and displayed agonistic properties at higher concentrations (Table 2). In contrast, ρV329G, ρS330C, or ρVS329,330GC receptor channels were insensitive to both modulatory and agonistic action of pentobarbital (Table 2). The specificity of position 328 in conferring pentobarbital sensitivity is corroborated by the lack of response of the ρV329G, ρS330C, or ρVS329,330GC receptor channels to pentobarbital, given the proximity of Val and Ser residues to Trp328.
Hydrophobic Residues at Position 328 Impart Pentobarbital Sensitivity to ρ1 Receptor Channel.
The Trp328 within the ρ1 subunit was replaced with an array of diverse amino acids differing in hydropathy index (HI) (Kyte and Doolittle, 1982), charge, and size. Remarkably, as with Met (HI = 1.9), substitutions of Trp328 (HI = −0.9) with other hydrophobic residues such as Leu, Ile, and Val (HI of 3.8, 4.5, and 4.2, respectively) and Ala (Fig. 1C, found at the corresponding position on the α1 subunit, HI = 1.8), imparted both the agonistic and modulatory properties of pentobarbital to the mutated ρ1 receptor channels. GABA responses from the ρ1 receptor channel containing the Phe substitution (HI = 2.5), however, were only weakly enhanced by pentobarbital. The homooligomericDrosophila GABA receptor channel responds to the modulatory action of pentobarbital at high concentrations (Chen et al., 1994). The TM3 of the Drosophila subunit contains Gly (HI = −0.4), Thr, and Cys at the equivalent position with respect to the TM3 of the ρ1 (Fig. 1C). The corresponding residues (Trp, Val, and Ser) within ρ1 receptor channel were mutated to Gly, Thr, and Cys. Pentobarbital potentiated the GABA-evoked currents from the ρWVS328,330GTC receptor channel (data not shown) and was also found to be an agonist for this mutant receptor channel (depressed maximum, Table 2). In contrast, the Tyr (HI = −1.3) substitution (ρW328Y) yielded receptor channel with a pharmacological profile resembling that of the wild type. Pentobarbital neither directly activated nor modulated the GABA-evoked currents for the ρW328Y receptor channel (see below). Finally, substitutions of Trp328 by Glu or Ser (Fig. 1C, found at the corresponding position on the γ2 subunit) or Pro (HI of −3.5, −0.8, and −1.6, respectively) failed to yield functional channel when tested with GABA (Table 1, up to 30 mM) or pentobarbital (Table 2, 2.5 mM). Thus, the HI of the substituted amino acid at position 328 appears to be crucial not only in imparting pentobarbital sensitivity, but also in GABA-dependent activation.
Mutation of Trp328 Transforms GABA Sensitivity.
Figure2 shows the current traces elicited by different concentrations of GABA, as well as the GABA concentration-response relationships from oocytes expressing ρ1, ρW328L, ρW328I, ρW328V, ρW328M, ρW328A, ρW328F, and ρW328Y receptor channels. GABA activated these mutants with similar efficacy (with the exception of ρW328Y, with the maximum current of approximately 10% of the wild type), when matched cRNA concentrations (see Materials and Methods) for these individual mutants were injected into oocytes and their maximal GABA-evoked currents were compared. These mutations, however, induced marked transformation in the GABA potency (Table 1). The ρW328L and ρW328I receptor channels displayed approximately a 3-fold increase in the sensitivity. The GABA EC50s for ρW328L and ρW328I were 0.35 and 0.39 μM, respectively (in comparison with 1.03 μM for the wild type). In contrast, the Phe and Ala substitution produced a drastic 30-fold reduction in GABA sensitivity, when compared with ρ1 receptor channel. The EC50 values estimated form GABA concentration-response relationships for ρW328A and ρW328F receptor channels were 32 and 30 μM, respectively. The GABA concentration-response relationship for ρWVS328–330GTC receptor channel was also altered (Table 1). In comparison with wild type, the GABA EC50 for the ρWVS328–330GTC receptor channel was increased by approximately 8-fold. Finally, the Met, Val, and Tyr substitutions at position 328 produced receptor channels with similar GABA sensitivity. The EC50 values for these receptor channels deviated 20 to 30% from the EC50 value for wild type. These results suggest that position 328 is not only crucial in conferring pentobarbital sensitivity, but also plays a key role in GABA-dependent activation.
GABA-dependent activation of ρ1 Trp328 mutant receptor channels. A, current traces evoked by different concentrations of GABA for ρ1 and ρ1 328 mutants. Thick line above each current trace represents duration of GABA application. Symbol representing each mutant in B (concentration-response relationships) is shown on left of current traces corresponding to that mutant. B, GABA concentration-response relationship for ρ1 Trp328 mutants. Each plot represents average of normalized peak currents versus GABA concentrations from three oocytes expressing ρ1 or ρ1 Trp328 mutants. Lines are best fit of Hill equation to data points, and error bars represent S.D. (n = 3). Note that there are approximately two orders of magnitude variation in concentration of GABA required to elicit half-maximal currents (EC50) among these mutants.
Pentobarbital-Dependent Potentiation Versus Inhibition of GABA Responses Occurs Over a Narrow Range of GABA Concentration for ρ1 328 Mutants.
Enhancement of the GABA-evoked currents by pentobarbital from the homooligomeric Trp328 mutants was dependent on GABA concentration. Figure3A shows the pentobarbital-mediated (50 μM) modulation of GABA-evoked currents at different agonist concentrations for ρW328L, ρW328I, ρW328V, ρW328M, ρW328A, and ρW328F receptor channels. It is noteworthy that pentobarbital alone at a concentration (50 μM) applied in this experiment does not activate these Trp328 mutants.
Pentobarbital modulation of GABA-evoked currents for Trp328 mutant receptor channels. A, pentobarbital modulation of GABA-evoked currents (at different concentrations of GABA) for ρW328L, ρW328I, ρW328V, ρW328M, ρW328A, and ρW328Y receptor channels. Note that pentobarbital potentiation versus inhibition occurs over a narrow range of GABA concentrations. Magnitude of disrupted current traces for ρW328M and ρW328V at 4 μM GABA are 3.0 and 2.1 μA, respectively. Thick line above each current trace represents duration of GABA application or coapplication of GABA and pentobarbital. B, plot of pentobarbital potentiation for ρW328M receptor channel in presence of 0.2, 0.3, and 0.5 μM GABA versus ratio of GABA concentrations to EC50 for ρW328M. Note exponential relationship between relative potentiation by pentobarbital and GABA concentrations. Pentobarbital potantiation occurs at GABA concentrations below GABA EC50 value for ρW328M. Error bars are S.D.s (n = 3).
Pentobarbital elicited potentiation of GABA responses only at low concentrations of GABA (fractions of their respective EC50 values; e.g., EC5). However, at higher concentrations of GABA, pentobarbital appeared to act as a antagonist. For ρW328M receptor channels, the relationship between the fold potentiation by pentobarbital (50 μM) versus three different GABA concentrations is plotted in Fig. 3B. At 0.2 μM GABA (∼0.15 of the EC50 for ρW328M), pentobarbital (50 μM) increased the peak GABA responses by approximately 18-fold, whereas the potentiation by pentobarbital was reduced to 5-fold in the presence of 0.3 μM GABA. In comparison, pentobarbital failed to potentiate the currents evoked by 0.5 μM GABA (∼0.38 of the EC50 for ρW328M) and at higher concentrations of GABA, displayed antagonistic properties (Fig. 3A). For ρW328L and ρW328I receptor channels with higher sensitivity to GABA, pentobarbital (50 μM) was inhibitory at GABA concentrations as low as 0.2 μM (∼50% of EC50) but not at 0.1 μM (data not shown).
For ρW328F receptor channel, pentobarbital appeared to be less potent than other pentobarbital-sensitive ρ1 mutants. The GABA responses from ρW328F were only weakly enhanced by pentobarbital. Finally, the substitution of Trp328 to a Tyr residue (ρW328Y) yielded a receptor channel with a pharmacological profile resembling that of wild type. Pentobarbital neither directly activated nor modulated the GABA-evoked currents for ρW328Y receptor channel (Fig. 3A and Table 2).
Another characteristic of the pentobarbital modulation of these Trp328 mutants was the increase in the residual current following the removal of pentobarbital and GABA. This phenomenon was most prominent in mutant receptor channels with the highest sensitivity to GABA. For instance, the residual current for ρW328L and ρW328I receptor channels nearly quadrupled in amplitude, following the removal of GABA and pentobarbital (Fig. 3A). In comparison, for mutants with greater GABA EC50s (e.g., ρW328A), the rate and extent of current rise following the wash were less pronounced.
Pentobarbital also increased the rate of deactivation for these Trp328 mutants. For example, the time for the current amplitude to fall to 50% (T1/2) of maximum following removal of GABA (0.5 μM) was 7.8 ± 0.6 s (n = 5) for ρW328M receptor channels, whereas the deactivationT1/2 for the same mutant increased to 35.2 ± 2.4 s (n = 5) following removal of GABA (0.5 μM) and pentobarbital (50 μM).
The ρW328L and ρW328I receptor channels also exhibited higher sensitivity to isoguvacine (a less potent GABA agonist for ρ1 receptor channel) when compared with ρ1 receptor channel. The isoguvacine EC50 values for ρW328L and ρW328I were 23.69± 3.38 μM (n = 3) and 37.31± 0.96 μM (n = 3) in comparison to 104 ± 3.62 μM (n = 4) for the wild-type receptor channel. In the presence of 8 μM isoguvacine (<EC10), pentobarbital (50 μM) enhanced the isoguvacine-induced currents for ρW328L and ρW328I by 257 ± 48% (n = 4) and 292 ± 16% (n = 4), respectively (Fig.4A).
A, pentobarbital potentiation of isoguvacine-evoked currents for ρW328I and ρW328L. Pentobarbital (50 μM) markedly potentiated isoguvacine (8 μM) responses for ρW328I and ρW328L receptor channels. Concentration of isoguvacine (8 μM) used in this study is equivalent to fraction of isoguvacine EC50s for ρW328I and ρW328L receptor channels. B, pentobarbital modulation of heterooligomeric α1β2γ2receptor channel. GABA concentrations used were equivalent (with respect to their EC50s) to those used in Fig. 3A for ρW328M receptor channel. Thick line above each current trace represents duration of GABA application or coapplication of GABA and pentobarbital. Note contrast in pentobarbital action between α1β2γ2 and ρW328M, where potentiation for α1β2γ2receptor channel occurs over a wide range of GABA concentration (relative to EC50 value). C and D, a mutation within GABA activation domain of ρW328M markedly reduces GABA potency but does not alter GABA concentration-dependent (relative to EC50 value) modulation by pentobarbital. C, at GABA concentration (200 μM) equivalent to a fraction of EC50value (3331 ± 346.7 μM), pentobarbital (20 μM) markedly enhanced GABA-evoked currents for homooligomeric ρY198S/W328M receptor channel. D, pentobarbital depression of GABA responses occurred at concentrations of GABA below EC50 for ρY198S/W328M receptor channels. At GABA concentration equivalent to 0.84 of EC50 (2800 μM GABA), pentobarbital (50 μM) inhibited GABA responses and at higher concentration (30,000 μM GABA) caused an apparent block desensitization. Thick line above each current trace represents duration of GABA application, or coapplication of GABA and pentobarbital.
Pentobarbital Potentiation Occurs Over a Wide Range of GABA Concentrations for α1β2γ2Receptor Channel.
Figure 4B shows the pentobarbital modulation of the heterooligomeric α1β2γ2receptor channel in the presence of different concentrations of GABA. The GABA concentrations used were equivalent (with respect to their EC50s) to those used for the ρW328M receptor channel (see Fig. 3A). For the α1β2γ2receptor channel (Table 1, GABA EC50 = 46.5 ± 4.7 μM), pentobarbital at a concentration of 50 μM markedly potentiated the GABA responses evoked by 7, 10, 17, and 34 μM GABA (Fig. 4B). Moreover, at concentrations of GABA (135 μM) approximately three times the EC50, moderate enhancement of the GABA response was still present. This experiment demonstrates that for the heterooligomeric α1β2γ2receptor channel, the potentiation by pentobarbital occurs over a much wider range of GABA concentrations (relative to EC50) than pentobarbital-sensitive Trp328 mutants. The contrast in pentobarbital mode of modulation for the Trp328 ρ1 mutants and α1β2γ2receptor channels is intriguing, given that these two classes of receptor channels may be activated differently by GABA (Amin and Weiss, 1996; see Discussion).
Impairment of GABA Sensitivity Does Not Alter Unique Pentobarbital Modulation of ρ1 328 Mutants.
The ρ1 receptor channel is approximately 40 times more sensitive to GABA than α1β2γ2receptor channel. To examine whether this difference in GABA sensitivity could account for the contrast in pentobarbital modulation between the two classes of receptor channels, a Tyr at position 198 (presumably located with in the extracellular receptor domain) was mutated to Ser within both ρW328M and ρWVS328–330 MGC mutant subunits. Previous studies have demonstrated that Tyr198 to Ser substitution within the ρ1 receptor channel results in a 2500-fold decrease in GABA sensitivity (Table 1, ρY198S, also Amin and Weiss, 1994). Similar to ρY198S, both ρY198S/W328M and ρY198S/WVS328–330MGC receptor channels exhibited a three orders of magnitude reduction in GABA sensitivity (Table 1, EC50s of ∼3000 μM). Nevertheless, the mode of pentobarbital modulation for these receptor channels remained the same as other Trp328 ρ1 mutants. Pentobarbital at a concentration of 20 μM synergistically potentiated the GABA (200 μM) responses from oocytes expressing ρY198S/W328M (Fig. 4C) or ρY198S/WVS328–330MGC by 390 ± 14% (n = 3) and 780 ± 160% (n = 3), respectively. However, at concentrations of 750 μM and 2800 μM of GABA (below the EC50 values for ρY198S/W328M), pentobarbital displayed antagonistic properties. Moreover, in the presence of 30 mM GABA, the pentobarbital effect was consistent with a channel block (or desensitization; see Fig. 4D for ρY198S/W328M).
Thus, marked decrease in GABA potency in ρW328M (or ρWVS328–330MGC) did not alter the paradigm in pentobarbital modulation for the homooligomeric ρ1 mutants. Pentobarbital yielded potentiation (for these activation-impaired Trp328 mutants) only in the presence of GABA concentrations equivalent to fractions of their respective EC50s, yet caused inhibition at higher concentrations.
Pentobarbital at Higher Concentrations Is an Agonist for ρ1 Trp328 Mutants.
In addition to the modulatory effect, pentobarbital at higher concentrations is also an agonist for ρW328L, ρW328I, ρW328V, ρW328M, and ρW328A receptor channels. Figure 5A shows current traces evoked by different concentrations of pentobarbital for ρW328L and ρW328M receptor channels. In pentobarbital-direct activation studies with GABAA receptor channels, the current amplitude increases before returning to the baseline following removal of pentobarbital (at high concentrations, Rho et al., 1996; J. Amin, unpublished observations). This phenomenon was absent in pentobarbital-direct activation of the Trp328 mutants. Note that for ρW328L and ρW328M receptor channels, even at the highest concentration of pentobarbital (2.5 mM), the evoked currents did not increase in amplitude following pentobarbital wash.
A, current traces evoked by different concentrations of pentobarbital for ρW328M and ρW328L receptor channels. Thick line above each current trace represents duration of GABA application. B, pentobarbital concentration-response relationship for ρ1Trp328 mutants. Each plot represents average of normalized peak (to extrapolated maximum) currents versus pentobarbital concentrations from three oocytes (except for ρW328I, two oocytes) expressing ρW328L, ρW328I, ρW328V, ρW328M and ρW328A receptor channels. Lines are best fit of Hill equation to data points, and error bars represent S.D. C, comparison of ρW328L, ρW328I, ρW328V, ρW328M, and ρW328A receptor channels’ EC50values for pentobarbital and GABA. Note that for these mutants, there are marked alteration in EC50 values for GABA, while there are only moderate difference in EC50 values for pentobarbital.
Figure 5B depicts pentobarbital concentration-response relationships for ρW328L, ρW328I, ρW328V, ρW328M, and ρW328A receptor channels. Each plot represents the average of normalized peak (to the extrapolated maximum) currents versus pentobarbital concentrations from oocytes (n = 3 except for ρW328I, n = 2) expressing the above Trp328 mutant receptor channels. Comparison of the pentobarbital and the GABA EC50 values for the same set of mutants is plotted in Fig. 5C. Unlike the effect on GABA potency, the difference in the pentobarbital potency is subtle (Table 2, pentobarbital EC50s of 0.8 to 2.4 mM). For example, the difference between the ρW328A and ρW328L receptor channels in pentobarbital potency is less than 2-fold. However, for the same mutants, there is approximately a 100-fold contrast in the GABA potency.
Pentobarbital is also an agonist for the activation-impaired ρY198S/W328M and ρY198S/WVS328–330MGC receptor channels. Interestingly, pentobarbital exhibited approximately a 4-fold higher potency for these mutants (EC50s of 202.7 ± 23.5 and 176.5 ± 29.8 μM, respectively) than for ρW328M and ρWVS328–330MGC receptor channels.
Similar to heterooligomeric α1β2γ2receptor channels (Table 2) pentobarbital is a partial agonist for all pentobarbital-sensitive mutants. Table 2 lists the ratio of maximal current (Imax) evoked by 2.5 mM pentobarbital to the Imax for GABA. Among these pentobarbital-sensitive mutants, the apparentImax for pentobarbital varied from 10 to 30% of Imax for GABA (Table 2).
Thiopental and Phenobarbital Modulation of Trp328 Mutants.
Thiopental and phenobarbital were also effective at potentiating GABA responses for pentobarbital-sensitive mutants, albeit with different potencies. As shown in Fig. 6, in the presence of 4 μM GABA, thiopental (50 μM) was nearly as potent as pentobarbital (50 μM) for ρW328A homooligomeric receptor channel. Thiopental and pentobarbital increased the GABA responses for ρW328A receptor channel by 659 ± 72.8% (n = 4) and 836.8 ± 90.4% (n = 4), respectively. On the other hand, phenobarbital at twice the concentration (100 μM), was less effective at potentiating ρW328A’s responses to GABA (452.5 ± 16.1%, n = 4). Preliminary results indicate that ρW328L, ρW328I, ρW328V, and ρW328M receptor channels were also modulated by thiopental and phenobarbital with similar relative potencies (data not shown). The comparative potencies of these barbiturates for Trp328 mutants are consistent, in general, with their relative clinical potencies (Franks and Lieb, 1994).
Comparison of pentobarbital, phenobarbital, and thiopental in modulating GABA responses from ρW328A receptor channel. Among tested barbiturates, pentobarbital and thiopental appear to be most potent positive modulators of GABA responses for ρW328A receptor channels. Thick line above each current trace represents duration of GABA application or coapplication of GABA and barbiturates.
Tryptophan Substitution for Met286 within TM3 of β2Subunit Abolishes Pentobarbital Sensitivity.
Within the β2 subunit, the Met at position 286 alone or in combination with Gly287 and Cys288 were mutated to their corresponding amino acid counterparts found in the ρ1subunit. Figure 7 illustrates the currents elicited by bath application of GABA (5 μM) or both GABA (5 μM) and pentobarbital (30 μM) to oocytes expressing β2 or βM286W receptor channel. Similar to β2, the expression of the cRNA for βM286W (or βMGC286–288WVS) yielded spontaneously open channels. The magnitude of the chloride ion leak (judged by the reversal potential for chloride) in these ion channels was proportional to the amount of injected cRNA (data not shown). In addition, β2wild-type and mutant receptor channels displayed severe depression in the Imax when tested with GABA (see legend to Fig. 1). Nonetheless, in contrast to β2, coapplication of pentobarbital and GABA to oocytes expressing βM286W (or βMGC286–288WVS, data not shown) receptor channels failed to increase the GABA-evoked currents (Fig. 7).
Pentobarbital-dependent modulation of GABA responses from β2 or βM286W receptor channel. Mutation of Met 286 to Trp (βM286W), abolished pentobarbital sensitivity of β2receptor channel. On other hand, GABA currents for wild-type β2 receptor channel markedly increased by coapplication of pentobarbital. Thick lines above each current trace represent duration of GABA application or coapplication of GABA and pentobarbital.
The coexpression of cRNA for the rat α1 subunit with either βM286W or βMGC286–288WVS yielded receptor channels highly responsive to GABA (similar potency and efficacy as wild-type α1β2 receptor channel, data not shown). However, in contrast to βM286W (or βMGC286–288WVS), the α1βM286W or the α1βMGC286–288WVS receptor channel was pentobarbital sensitive (data not shown). Comparison of the amino acid sequence encoding the TM3 domain of β2 and α1 subunits revealed that the α subunit contains an Ala residue at the corresponding position (Fig. 1C). In the α1βM286W or the α1βMGC286-288WVS receptor channel, the lost pentobarbital function of the mutated β2 subunit may be reverted by the presence of the Ala residue within the TM3 of the α subunit.
Discussion
The data presented here indicate that replacing residue 328 with a spectrum of amino acid residues can confer barbiturate modulation as well as alter GABA-dependent activation of the mutated ρ1 receptor channel. The apparent major determinant for pentobarbital sensitivity of the mutated ρ1 receptor channel was, however, the hydrophobicity of the substituted amino acid at position 328. There were also key differences in the pentobarbital modulation between the homooligomeric ρ1 328 mutants and the heterooligomeric αβγ receptor channels.
Pentobarbital Versus GABA.
The lack of stringency for amino acid side chains (except for hydrophobicity) at position 328 to confer pentobarbital sensitivity is unique. For instance, the Met side chain is different from that of Ala in both size and the constituent elements, whereas the EC50 for pentobarbital between ρW328M and ρW328A receptor channels varied by less than 2-fold. Mutational analysis of different ligand-gated ion channels (including GABA) has shown that even conservative amino acid substitutions (such as Tyr to Phe) within the agonist-dependent activation domain can markedly impair the agonist sensitivity (Vandenberg et al., 1992; Amin and Weiss, 1993). The differences in amino acid side chain requirement between the agonist and the pentobarbital activation domains is perhaps best manifested by the nature of the bond they form. The interaction between pentobarbital and its site of action may be mediated through the butyl side chain of the amphipathic pentobarbital molecule via the relatively weak hydrophobic interaction, whereas the GABA agonist may interact with its activation domain by more specific and stronger hydrogen bonding. In the pentobarbital-mediated potentiation of the GABA responses, the apparent sequential release of pentobarbital followed by GABA may attest to this notion, because there appeared to be a direct correlation between the magnitude of current rise following the wash, and the EC50 of the Trp328 mutants (Fig. 3A, smallest rise in the current for ρW328A and largest for ρW328L and ρW328I).
Target-Specific or General Perturbation?
The view that general anesthetics indirectly affect membrane-embedded ion channels by altering the fluidity of lipid membranes is being gradually replaced by a target-specific model (Franks and Lieb, 1994). Two chief arguments, namely the identification of protein targets such as luciferase, as well as the discovery of the stereoselectivity of anesthetic agents, support the target-specific model for anesthetic action (Huang and Barker, 1980; MacIver and Roth, 1987; Franks and Lieb, 1991). Recently, several groups (Belelli et al., 1997; Mihic et al., 1997) have shown that the residues within the TM2 and the TM3 appear to be crucial for the action of the general anesthetics etomidate and enflurane. Interestingly, Mihic et al. (1997) have conversely mutated the corresponding 328 residue within the α subunit of the glycine receptor, or α and β subunit of the GABAA receptor channel to a Trp, to abolish the action of enflurane. Pentobarbital and enflurane are two structurally diverse anesthetics that appear to exert their action through the same site. This notion, together with the lack of amino acid side chain specificity for pentobarbital-dependent modulation, rekindles the debate over the mechanism of anesthetic action. It is tempting to speculate that substitution of the Trp328 to a hydrophobic residue may cause the TM2 (gate) to be readily accessible to the pentobarbital’s induced local lateral pressure within the membrane bilayer (Gaines, 1966; Gruner and Shyamsunder, 1991;Cantor, 1997). In this scenario, anesthetics may only need the exposure of the channel’s gating component to the membrane bilayer to shift equilibrium between the open and closed states.
Pentobarbital Modulation of ρ1 Versus α1β2γ2 Receptor Channels.
The contrast in the pentobarbital modulation between homooligomeric ρ1 and heterooligomeric α1β2γ2receptor channels is intriguing given that the ρ1 receptor channel shows approximately 40-fold greater sensitivity to GABA than α1β2γ2, and ρ1 displays unique activation and deactivation kinetics.
The aforementioned difference in GABA sensitivity, however, does not appear to play a key role in pentobarbital’s unique modulation of ρ1 Trp328 mutants. In experiments in which the GABA sensitivity of ρW328M was decreased by (mutation of Tyr to Ser at position 198) nearly three orders of magnitude, the dual modulatory action of pentobarbital for the resulting receptor channel persisted. Therefore, pentobarbital’s unique modulation of (pentobarbital-dependent potentiation versus inhibition) ρ1 Trp328 mutants is independent of GABA potency.
Could the difference in activation mechanism between these two classes of receptor channels account for the contrast in pentobarbital modulation? Experiments with coexpression of different ratios of wild-type ρ1 and activation-impaired ρ1 subunits (Y198S) have demonstrated previously that the agonist-dependent activation of homooligomeric ρ1 receptor channel appears to be preceded by three binding steps (Amin and Weiss, 1996) rather than two binding steps observed for heterooligomeric receptor channels (Blount and Merlie, 1989). The three-step activation scheme for ρ1 receptor channel was derived based on the assumption of one binding site per subunit in a pentameric configuration (five potential binding sites). A speculative view is that in the presence of pentobarbital the forward rates for GABA are increased for pentobarbital-sensitive ρ1receptor channel and at relatively higher concentrations of GABA, two additional binding sites could become occupied, leading the channel into a closed/desensitized state. Consistent with this, binding studies for the GABAA receptor channel (e.g., α1β2γ2) have shown that GABA binding is enhanced in the presence of pentobarbital (Olsen et al., 1991; Wakamori et al., 1991; Lin et al., 1993). Alternatively, pentobarbital antagonistic action could arise from pentobarbital binding to an inhibitory site within the ρ1Trp328 mutants. Rho et al. (1996), based on barbiturate studies on GABAA receptor channels, has proposed the presence of a low-affinity inhibitory site for pentobarbital. The depression in the pentobarbital-direct activation Imax (with respect to GABAImax) could also be due to occupation of this postulated inhibitory site by pentobarbital.
Tryptophan Residue.
What architectural features within the ρ1 receptor channel might be created by Trp328 substitutions? The Trp residue is unique not only with respect to size, but also because of the indole moiety on its side chain. This residue can potentially anchor the TM3 to the extracellular side of the membrane. In this scenario, mutation of Trp328 to hydrophobic amino acids such as Met, Leu, Ile, Ala, or Val may dislodge the N-terminal amino acids of the TM3 from the interface of the extracellular side of the membrane and subsequently allow the residue 328 to rest deep within the membrane. This structural perturbation in the TM3 may then expose the gate of the channels to the membrane components (see above). Alternatively, the TM3 in the new configuration along with other TMs may constitute a binding cavity for pentobarbital. This phenomenon in which membrane-spanning domains interact to create a binding site, is not unique among membrane-embedded proteins. For example, the interactions of different TMs in the rhodopsin molecule constitute a binding cavity for the retinal molecule (Unger et al., 1997). Finally, Trp328 may impede the interaction of the pentobarbital with its binding/sensor domain solely based on its size. Consistent with this hypothesis, ρ1 receptor channels containing the Tyr at position 328 did not respond to pentobarbital. Furthermore, in comparison with other hydrophobic amino acid substitutions, ρ1 receptor channel containing the Phe (contains an aromatic ring on its side chain) substitution (ρW328F) exhibited lower pentobarbital sensitivity.
The amino acid residues in the center of the TM2 (leucine, the presumed gate) and the TM3 (Phe and Val) are conserved among all GABA subunits. Hypothetically, pentobarbital binding may induce interaction of these conserved residues leading to an increase in agonist affinity for its receptor, given that the binding of the agonist to its receptor and the gating of the channel are closely coupled. Interestingly, mutation of conserved Phe residue (ρF333 M) within the center of the TM3 resulted in receptor channels that responded to neither GABA nor pentobarbital (Tables 1 and 2). Alternatively, in a situation in which the agonist binding cleft resides proximal to the extracellular side of the membrane, the polar moiety of the pentobarbital can alter the agonist binding cleft and thereby change the affinity of the agonist for its receptor. In either proposed mechanism, marked variation in GABA sensitivity among the Trp328 mutants as well as the increase in pentobarbital potency concomitant with impairment of GABA activation domain (ρY198S/W328M), may attest to the close coupling of the agonist and pentobarbital binding/sensor site.
Within the TM3, Trp328 is positioned 5 amino acids from the presumed extracellular interface and 14 amino acids from the intracellular compartment. This positioning of residue 328 within the membrane is intriguing, because anesthetics in general exhibit membrane asymmetry in exerting their effect. It is also interesting that the length of the hydrophobic side chain of pentobarbital (also thiopental), which is nearly 5 angstroms in length, closely matches the depth in which position 328 may penetrate within the lipid bilayer in a presumed α-helical structure.
Acknowledgments
I am indebted to Dr. N. P. Franks, Dr. W. R. Lieb, and Dr. E. Bennett for constructive suggestions; Dr. L. Carlacci for helpful discussion; Dr. D. Fitzpatrick and Dr. P. Gottschall for reading, and Dr. V. Pollock and K. Morris for help in the preparation of this manuscript.
Footnotes
-
Send reprint requests to: Dr. Jahanshah Amin, University of South Florida, College of Medicine, 12901 Bruce B. Downs Blvd. MDC Box B9, Tampa, FL 33612-4799. Email:Jamin{at}pharm.med.usf.edu
-
This work was supported by grants from National Eye Institute (EY11531–01A1) and Council for Tobacco Research (SA052).
- Abbreviations:
- GABA
- γ-aminobutyric acid
- TM
- transmembrane domain
- Received July 7, 1998.
- Accepted December 18, 1998.
- The American Society for Pharmacology and Experimental Therapeutics
