Two Invariant Tryptophans on the α1 Subunit Define Domains Necessary for GABAA Receptor Assembly*

Two invariant tryptophan residues on the N-terminal extracellular region of the rat α1 subunit, Trp-69 and Trp-94, are critical for the assembly of the GABAA(γ-aminobutyric acid, type A) receptor into a pentamer. These tryptophans are common not only to all GABAA receptor subunits, but also to all ligand-gated ion channel subunits. Converting each Trp residue to Phe and Gly by site-directed mutagenesis allowed us to study the role of these invariant tryptophan residues. Mutant α1 subunits, coexpressed with β2 subunits in baculovirus-infected Sf9 cells, displayed high affinity binding to [3H]muscimol, a GABA site ligand, but no binding to [35S]t-butyl bicyclophosphorothionate, a ligand for the receptor-associated ion channel. Neither [3H]muscimol binding to intact cells nor immunostaining of nonpermeabilized cells gave evidence of surface expression of the receptor. When expressed with β2 and γ2 polypeptides, the mutant α1 polypeptides did not form [3H]flunitrazepam binding sites though wild-type α1 polypeptides did. The distribution of the mutant receptors on sucrose gradients suggests that the effects on ligand binding result from the inability of the mutant α1 subunits to form pentamers. We conclude that Trp-69 and Trp-94 participate in the formation of the interface between α and β subunits, but not of the GABA binding site.

The ␥-aminobutyric acid, type A receptor (GABA A receptor) 1 is the postsynaptic target for GABA, the major inhibitory neurotransmitter of the mammalian central nervous system. Along with the nicotinic acetylcholine, glycine, and serotonin (5HT3) receptors, the GABA A receptor belongs to the superfamily of ligand-gated ion channels, all of which share significant sequence and structural features (1)(2)(3). Molecular cloning has led to the identification of nine genes that encode distinct GABA A receptor polypeptides, including six ␣, four ␤, three ␥, one ␦, three , one ⑀, and one polypeptides. Several splice variants are also known, suggesting that a large number of heteropentameric isoforms could form (4 -8). Most studies so far suggest that functional receptors consist of two ␣, two ␤, and one ␥ subunits (9 -12).
The actual diversity of GABA A receptors in the brain depends on the precise processes of receptor assembly. The Nterminal segments of the receptor subunits contain signals for subunit oligomerization (13), but the specific domains responsible for receptor assembly are not yet known. Extensive mutagenesis studies performed on GABA A receptors have led to the identification of specific residues responsible for GABA binding on the interface of the ␣ and ␤ subunits and for benzodiazepine binding on the interface of the ␥ and ␣ subunits (14 -16). Two of these residues, Phe-64 and His-101 on the ␣1 subunit of the GABA A receptor, were also identified by photoaffinity labeling studies using [ 3 H]muscimol and [ 3 H]flunitrazepam to contribute to the GABA and benzodiazepine binding sites, respectively (15,17,18). Although these residues apparently contribute to separate binding pockets for the receptor agonist and allosteric modulator respectively, they are separated by only 37 amino acid residues, and as indicated in Fig.  1, four of these residues are identical not only among all GABA A receptor subunits, but also among all ligand-gated ion channel subunits. This constancy suggests selective value, and we reasoned that these residues are likely to play a critical role in GABA A receptor function. In particular, Smith and Olsen (15) have hypothesized that the two tryptophan residues, Trp-69 and Trp-94, because of their bulky and unique structure, may define a functional domain that is critical for channel opening. Galzi and Changeux (16) have proposed that both these invariant Trp residues behave as "structural canonical residues" that constrain the folding of the loops. If indeed this is the case, mutating the Trp residues could affect the folding of the loops and alter the conformational characteristics of the receptor. We have tested this hypothesis by mutating Trp-69 and Trp-94 into Phe and Gly and then studying the effects of each of these mutations on the binding characteristics and assembly of the GABA A receptor.

EXPERIMENTAL PROCEDURES
Site-directed Mutagenesis--The epitope tag and point mutations were introduced in the cDNA for the ␣1 subunit of the GABA A receptor using the Altered Sites II in vitro Mutagenesis Systems (Promega, Madison, WI). First, the entire coding region of the ␣1 subunit was subcloned into pALTER, and a 10-amino acid 9E10 epitope (EQKLI-SEEDL) from c-Myc was introduced between residues 4 and 5 of the mature peptide as described in the manual using the mutagenic oligonucleotide, GTCCTTAAGTTCATCTAGATCTTCTTCTGATATTAGC-TTT TGTTCTTGGGAGGGCTGTC. Specific point mutations were incorporated in a second round of mutagenesis performed on the epitopetagged ␣1 subunit construct using the following mutagenic oligonucleotides: W94F, AAATGTATCTGGAGTAAAGATTTTACTGGCCAT; W94G, AAATGTATCTGGAGTCCCGATTTTACTGGCCAT; W69F, TAATCTTTCATCCTTAAAGCTTTGGCGGAAAAA; W69G, TAATCTT-TCATCCTTCCCGCTTTGGCGGAAAA. Positive clones were identified by restriction digests and DNA sequencing.
Recombinant Protein Expression-Sf9 cells (Pharmingen, San Diego CA) were maintained in serum-free XL-400 medium (JRH Biosciences, Lenexa, KS) in shaking culture at 26°C as described previously (19). cDNAs for the ␤2 and either wild-type or mutant ␣1 subunits were subcloned into the baculovirus transfer plasmid, pAcAB3, under the control of the p10 and polh promoters, respectively. Recombinant baculoviruses were constructed by homologous recombination between the transfer vector and linearized BaculoGold viral DNA (Pharmingen) using a liposome-mediated transfection protocol. After 5 days, the supernatant was collected to harvest recombinant viruses, which were then amplified to produce a high titer stock (0.5-1 ϫ 10 8 plaque-forming unit/ml). GABA A receptors were produced by infecting Sf9 cells with recombinant baculovirus (containing cDNAs for both subunits) at a multiplicity of infection of 5. In experiments testing for [ 3 H]flunitrazepam binding, cells were coinfected with a recombinant ␥2 virus in addition to the virus encoding the wild-type or mutant ␣1 and ␤2 subunits. Infected cells were harvested at 60 h post-infection by pelleting the cells at 1000 ϫ g for 10 min. Cell pellets were stored at Ϫ20°C until ready to use.
Western Blots-Membranes of Sf9 cells expressing either wild-type or mutant ␣1 along with ␤2 subunits were analyzed by SDS-PAGE, and proteins were transferred on to nitrocellulose membranes. The blots were blocked in phosphate-buffered saline with 0.1% Tween 20 and 10% nonfat dry milk for 1 h at room temperature and incubated with the polyclonal anti-myc tag antibody (Upstate Biotechnology, Inc., Lake Placid, NY) at a dilution of 1:1000 for 1 h at room temperature. After extensive washing, the blots were incubated with anti-rabbit horseradish peroxidase-linked secondary antibodies (Amersham Pharmacia Biotech) at a dilution of 1:1000 for 1 h at room temperature. After extensive washing, immunoreactivity was detected by chemiluminescence (ECL, Amersham Pharmacia Biotech).
Radioligand Binding Assays-Radioligand binding assays were performed either on lysed cell membranes or on intact cells. Cell membranes were prepared by homogenizing the cell pellet for 10 s in membrane wash buffer (20 mM K 2 HPO 4 /KH 2 PO 4 buffer with 50 mM KCl, pH 7.5, containing 0.02% NaN 3 , 0.5 mM dithiothreitol, and the following protease inhibitors: 1 mM EDTA, 2 mM benzamidine chloride, 0.1 mM benzethonium chloride, 50 units/ml bacitracin, 0.3 mM phenylmethylsulfonyl fluoride, 10 mg/liter ovomucoid trypsin inhibitor, 10 mg/liter soybean trypsin inhibitor). The membranes were recovered by centrifugation at 150,000 ϫ g for 20 min, resuspended to a final concentration of 0.5 mg/ml, and used for the binding assays. For whole cell binding, the frozen cell pellet was thawed, resuspended in 10 mM phosphate buffer with 100 mM KCl, pH 7.5, to a concentration of 0.5 mg/ml and used for the binding assays. Binding assays were carried out using an appropriate concentration of the radioligand in 10 mM phosphate buffer with 100 mM KCl, pH 7.5, and 100 g of total membrane protein in a total volume of 250 l. S]TBPS). The reaction mixtures were diluted with 4 ml of assay buffer and filtered under vacuum over Whatman GF/B filters using a Brandel cell harvester. The filters were washed twice with 4 ml of assay buffer and counted for radioactivity. Undisplaceable background was estimated in the presence of 100 M GABA, 10 M diazepam, or 100 M picrotoxin, respectively, and was subtracted from total binding to compute specific binding. GABA displacement assays were conducted using 20 nM [ 3 H]muscimol and nonradioactive GABA concentrations ranging from 1 nM to 0.1 mM. Doseresponse profiles were generated by fitting the data to a threeparameter Hill equation.
Cell Surface Localization-Sf9 cells were seeded on each well of poly-D-lysine-treated eight-well chamber slides. The cells were infected with appropriate baculovirus constructs and incubated at 26°C for 60 h. The cells were fixed in 4% paraformaldehyde and washed three times with 0.1 M phosphate-buffered saline. Cells were permeabilized with 1% Triton X-100 in 0.1 M phosphate-buffered saline for 30 min and blocked using 3% fetal calf serum. Incubations with the polyclonal anti-myc tag antibodies at a dilution of 1:250 were carried out overnight at 4°C in the presence of 1% fetal calf serum. The cells were washed extensively and treated with goat anti-rabbit fluorescein isothiocyanate (Vector Laboratories, Burlingame, CA) at a dilution of 1:200 for 2 h at room temperature. The cells were then washed extensively and examined using a confocal microscope (Carl Zeiss LSM 310).
Sucrose Gradients-Membranes were prepared from infected cells as described for the binding assays. An equal volume of 2ϫ solubilization buffer (20 mM K 2 HPO 4 /KH 2 PO 4 , pH 7.5, 2% Triton X-100, 200 mM KCl, containing 0.02% NaN 3 , 0.5 mM dithiothreitol, and various protease inhibitors as described for the membrane preparation) was added, and the homogenate was solubilized during magnetic stirring at 4°C for 1 h. The supernatant was collected and concentrated to about one-fifth of its original volume in Centriprep-30 concentrators. The concentrated solubilized protein solution was layered on top of a 3-30% sucrose gradient made of solubilization buffer containing only 0.1% Triton X-100. Gradients were centrifuged in a Beckman SW40 rotor for 40 h at 34,500 rpm. Fractions of 500 l were collected from the top of the tube, precipitated following a chloroform-methanol protocol (20), and analyzed by Western blots using the anti-myc tag antibody as described earlier. The intensity of the bands was quantitated by densitometry. Cytochrome c (1.9 S), bovine serum albumin (4.3 S), alcohol dehydrogenase (7.4 S), and catalase (11.4 S) were run as standards, in parallel in a separate tube. The positions of the standards in the gradient were determined by SDS-PAGE. Fig. 1 shows a sequence alignment of several GABA A receptor subunits and some examples of ligand-gated ion channels in a region flanked by the [ 3 H]muscimol-photoaffinity-labeled peptide, Tyr-59 to Gln-67 (17), and the [ 3 H]flunitrazepam-labeled peptide, Trp-94 to Asn-102 (15,18). This region shows high sequence identity, not only among the GABA A receptor subunits, but also among all ligand-gated ion channels. To explore the role of Trp-69 and Trp-94 in ␣1 polypeptides, we mutated these residues to Phe (to retain the hydrophobicity) and Gly (to eliminate steric factors). After expressing the mutant ␣1 polypeptides either with ␤2 alone or with ␤2 and ␥2 polypeptides, we examined the binding characteristics and assembly of the mutant receptors.

RESULTS
Epitope Tagging of ␣1 Does Not Affect the Function or the Subcellular Location of GABA Receptors in Sf9 Cells-GABA A receptors expressed in Sf9 and in mammalian cells showed appropriate binding and functional characteristics. Co-expression of ␣1 and ␤2 polypeptides produced GABA A receptors that bound [ 3 H]muscimol and [ 35 S]TBPS and that also showed GABA-activated chloride currents. Individual ␣1 or ␤2 polypeptides did not bind either ligand and did not show any GABA-activated currents. Co-expression of ␣1, ␤2, and ␥2 subunits formed functional receptors, which also showed [ 3 H]flunitrazepam binding and GABA-activated currents that were enhanced by benzodiazepines 2 (21,22). To aid in the biochemical analyses and subcellular localization of the ␣1 subunits of the GABA A receptor, we introduced a 9E10 epitope tag at the N-terminal of the mature wild-type ␣1 peptide, and we ex- pressed this construct along with the ␤2 subunit in Sf9 cells. The epitope tag had no effect on either GABA affinity in Sf9 cells (data not shown) or on the functional characteristics of the receptor expressed in A293 cells (23).
Epitope-tagged Mutant Polypeptides Are Synthesized in Sf9 Cells-We incorporated all four mutations into the epitopetagged ␣1 polypeptide to facilitate detection by Western blots and subcellular localization. All mutant ␣1 subunits were expressed and could be detected by antibodies that recognized the myc epitope tag on Western blots (Fig. 2). The antibody did not recognize the untagged ␣1 subunit or the ␤2 subunit.
Trp-69 and Trp-94 Mutants Bind Muscimol but Not TBPS-Homogenates of cells expressing the four mutant ␣1 subunits (Trp-69 and Trp-94 each converted to Phe and Gly) along with ␤2 subunits showed binding to [ 3 H]muscimol, a GABA site ligand (Fig. 3A). This result was reproducible in many experiments, with different batches of viruses. The density of [ 3 H]muscimol binding sites varied between batches of cells and viruses and reflects differences in the B max values of the expressed receptor.
We determined the affinity of the mutant receptors for GABA by GABA displacement of [ 3 H]muscimol binding (Table  I). There was no significant change in the GABA affinity of the [ 3 H]muscimol binding sites formed by the Trp-94 mutants. The Trp-69 mutants, however, showed a slight increase in GABA affinity. Taken together, these data suggest the formation of a relatively normal GABA binding site despite the mutation of the ␣1 subunit near ␣1F64, a residue implicated in GABA binding.
Trp-69 and Trp-94 Mutations Affect Intracellular Targeting-When ␣1 and ␤2 subunits are expressed independently in cells, they are retained in the endoplasmic reticulum, but when they are expressed simultaneously, both subunits are targeted to the cell surface (23). We tested the ability of mutant ␣1 subunits to target to the cell surface with an antibody that recognizes the myc epitope tag, 9E10. Surface expression was detected by immunostaining nonpermeabilized cells (Fig. 4, panels A1-H1). We confirmed that the Sf9 cells were expressing the appropriate peptide by processing permeabilized cells in parallel (Fig. 4, panels A2-H2).
In cells that express only epitope-tagged ␣1 (␣1 9E10 ), we found epitope-tagged peptide only in permeabilized cells (Fig.   4, panel A1) and not in the nonpermeabilized sample (Fig. 4,  panel A2). This result confirmed that expression of the ␣1 polypeptide alone resulted in an intracellular localization of the ␣1 subunit. As expected, the anti-myc antibody did not recognize the ␤2 subunit (Fig. 4, panels B1 and B2) or the untagged ␣1 subunit (Fig. 4, panels C1 and C2).
Cells that simultaneously expressed wild-type ␣1 9E10 and ␤2 subunits showed antibody staining even in nonpermeabilized cells (Fig. 4, panels D1 and D2), confirming surface expression. Similar results were observed using bd17, a monoclonal antibody that recognizes the ␤2 subunit (data not shown). Fig. 4, panels E1 through H1 and E2 through H2, show the localization of the ␣1W69 and ␣1W94 mutants expressed with the ␤2 subunits. In all cases, no immunostaining was observed in nonpermeabilized cells indicating that the mutant ␣1 subunits were deficient in their targeting to the cell surface. Peptide expression, as well as antibody reaction, was confirmed by positive staining in permeabilized cells (Fig. 4, panels E1-H1). Similar results were observed with bd17, which recognizes the ␤2 subunit (data not shown).
␣1W69 and ␣1W94 Mutant Polypeptides Fail to Assemble Normally into Pentamers-To identify the oligomeric species formed by the mutant ␣1 polypeptides, we solubilized Sf9 cells that produced ␣1 9E10 ␤2, ␣1 9E10 (W69G)␤2 and ␣1 9E10 (W94G)␤2 receptors and loaded the solubilized extracts on to sucrose gradients. We analyzed sucrose gradient fractions collected by immunoblotting with anti-myc antibody (Fig. 5). The ␣1W94G and ␣1W69G polypeptides formed oligomers that were most concentrated in fraction 8, while wild-type ␣1 9E10 ␤2 receptor was most concentrated in fraction 13. Based on the sedimentation velocity of known markers, the sedimentation coefficients of polypeptides in these fractions were 5.3 and 8.7 S. Similar results were obtained by monitoring the migration of the ␤2 subunits using the monoclonal antibody, bd17 (data not shown). Knight et al. (24) have identified a pentameric species of GABA A receptors expressed in Sf9 cells that migrates at about 8.8 S. Thus, the ␣1 9E10 ␤2 receptors probably migrated as pentamers. The receptors that contain the ␣1W69 and ␣1W94 mutations did not have the same sedimentation coefficient as the pentameric ␣1 9E10 ␤2 receptor. Instead, they formed smaller oligomers, probably dimers, which may be the [ 3 H]muscimol binding species. [ 3 H]Muscimol binding was lost on solubilization of cells coexpressing the mutant ␣1 and ␤2 subunits, preventing us from monitoring the [ 3 H]muscimol binding species on the sucrose gradient. DISCUSSION The GABA A receptor gene family forms a subset of the ligand-gated ion channel superfamily. At least 19 genes encode distinct GABA A receptor subunits, which can form a large number of heteropentamers (4). Since the channel characteristics and pharmacology of the receptor depend on its subunit composition (25)(26)(27), a major challenge to understanding the physiological and pharmacological diversity of GABA A receptors is determining the control of receptor assembly. Our data suggest the identity of a domain on the ␣1 subunit that is necessary for the formation of an interface between the ␣1 and the ␤2 subunits.
Pentameric When expressed with ␤2 subunits, the Trp-69 and Trp-94 mutant ␣1 subunits formed a normal GABA binding site, which bound [ 3 H]muscimol with normal affinity (Table I). But the mutants were completely unable to bind [ 35 S]TBPS, a ligand that binds the chloride ionophore (Fig. 3A).
The mutant ␣1 polypeptides, when expressed with ␤2 and ␥2 polypeptides, were unable to bind the benzodiazepine ligand, [ 3 H]flunitrazepam (Fig. 3C). Analysis of the oligomer sizes by velocity sedimentation in sucrose gradients of detergent-solubilized membranes indicated that the mutant polypeptides did not form pentamers (Fig. 5).
Mutant ␣1 subunits associate with ␤2 subunits to form a species that is smaller than pentamers and binds [ 3  H]muscimol will bind to smaller intermediates. These results are consistent with the studies of Im et al. (12), which showed that a tandem construct of ␣6 and ␤2 subunits expressed in HEK293 cells could not by itself form pentamers (channels), although it retained the ability to bind muscimol. The Trp-69 and Trp-94 mutations compromised the ability of the ␣1 polypeptides to assemble with ␤2 or ␤2 and ␥2 polypeptides to form a pentamer.
For GABA A receptors, only the ␣1␤2 and ␣1␤2␥2 combinations were detected on the surface of cells. Monomers, or ␣1␥2 and ␤2␥2 combinations remained in the endoplasmic reticulum (23). Furthermore, these were the only combinations that could form pentamers (9), suggesting that only fully assembled pentameric GABA A receptors are capable of surface expression. The ␣1W69 and ␣1W94 mutants showed no cell surface binding (Fig. 3C) or immunostaining (Fig. 4) when expressed with ␤2 subunits, consistent with their inability to assemble to form pentamers as confirmed by sucrose gradients (Fig. 5).
How Can the Two Trp Mutations Interfere with the Assembly of the Receptor into a Pentamer?-Since the GABA binding site lies on the interface of the ␣ and ␤ subunits, and since the mutant ␣1 subunits expressed with the ␤2 subunits bind [ 3 H]muscimol, the GABA binding interface is unaffected by the Trp-69 and Trp-94 mutations. The failure of these subunits to assemble into pentamers suggests that the other interface involving ␤-␣ interactions must not have formed. If we call the FIG. 3. ␣1W69 and ␣1W94   interface on which the GABA binding site lies the "␣ 3 ␤ interface," the ␣1W69 and ␣1W94 mutations prevent the assembly of the subunits into pentamers by interfering with the formation of the "␤ 3 ␣ interface" (Fig. 6A). Fig. 6B shows one possible scheme for the assembly of a receptor composed of two ␣ and three ␤ subunits and the point at which assembly of the receptor into a pentamer can be blocked by the mutant ␣1 subunits. A receptor composed of three ␣ and two ␤ subunits would not assemble into pentamers by a similar mechanism. The ␥ subunit is unable to rescue the altered ␣1 subunits created by the Trp-94 mutations, since mutant ␣1 subunits are unable to form [ 3 H]flunitrazepam binding sites when expressed with ␤2 and ␥2 subunits. This can be attributed to one of two FIG. 4. Mutant ␣1 subunits when coexpressed with ␤2 subunits are not detected on the surface of Sf9 cells by immunofluorescence using the anti-myc tag antibody. The following GABA A receptor combinations were expressed in Sf9 cells: ␣1 9E10 (panels A), ␤2 (panels B), ␣1␤2 (panels C), ␣1 9E10 ␤2 (panels D), ␣1 9E10 (W94F)␤2 (panels E), ␣1 9E10 (W94G)␤2 (panels F), ␣1 9E10 (W69F)␤2 (panels G), ␣1 9E10 (W69G)␤2 (panels H). Cells were either permeabilized using 1% Triton (panels A1-H1) or left intact (panels A2-H2). The localization of the epitope-tagged ␣1 subunits was detected using a polyclonal antimyc tag antibody and a fluorescein isothiocyanate-tagged secondary antibody.
FIG. 5. ␣1W94G and ␣1W69G mutants coexpressed with ␤2 subunits do not form pentamers as shown by sucrose gradient fractionation. Sf9 cells expressing the subunit combination indicated were solubilized and subjected to sucrose density gradient fractionation. Gradient fractions were separated by SDS-PAGE and the ␣1 9E10 subunit was detected by Western blotting using the anti-myc tag antibody and enhanced chemiluminescence (ECL). A, the intensity of the bands was estimated by densitometry and is plotted as a function of the fraction number. The S values were determined by running molecular weight markers in separate tubes. B, the blots used for the densitometry are shown here.
FIG. 6. The ␣1W69 and ␣1W94 mutants interfere with the assembly of the GABA A receptor into pentamers by preventing the formation of the ␤ 3 ␣ interface. A, the ␣ 3 ␤ interface, which includes the GABA binding site, and the ␤ 3 ␣ interface are not equivalent. The Trp mutants do not affect the ␣ 3 ␤ interface, since they are able to form normal [ 3 H]muscimol binding sites; they prevent the formation of the ␤ 3 ␣ interface and, hence, the assembly of the receptor into pentamers. The impaired interface is indicated by an X. B, one possible scheme for the assembly of an ␣␤ pentamer made of three ␤ and two ␣ subunits is shown. The Trp-69 and Trp-94 mutants can block assembly of such a receptor at the position marked X.
reasons: (i) they are unable to form a ␥ 3 ␣ interface, or (ii) the ␣1 mutants are unable to form pentamers due to a compromised ␤ 3 ␣ interface. In either case, no [ 3 H]flunitrazepam binding can be detected due to the absence of a pentamer.
Why Should the Mutation of Trp-69 or Trp-94 Residues on the ␣1 Subunit of the GABA A Receptor Affect Subunit Interactions?-Several recent reports in the literature have identified critical Trp residues involved in protein-protein interactions (28). Our inference that mutations in ␣1W69 and ␣1W94 disrupt subunit interactions in the GABA A receptor is consistent with the known roles of tryptophan in protein-protein interactions. Trp-69 and Trp-94 may both lie at the ␤ 3 ␣ interface, although the Trp-69 residue more probably lies in close proximity to the ␣ 3 ␤ interface. Alternately, these residues may demarcate a region necessary for appropriate interactions of the ␣1 subunit at the ␤ 3 ␣ interface. These residues could play a structural role and the lack of assembly of the subunits could be the result of misfolding of the protein in this region flanked by Trp-69 and Trp-94. If the mutant ␣1 peptides are folded improperly, however, the disruption is likely to be local, since the ␣1 peptide is still able to form a normal [ 3 H]muscimol binding site.
The only detectable difference in the effect of the mutagenesis of the two W residues was in the affinity of the [ 3 H]muscimol binding species for GABA (Table I). The ␣1W69 mutants expressed with ␤2 had a higher affinity for GABA than the wild-type receptors, as well as the ␣1W94 mutants. This change in affinity may be evidence of the role of Trp-69 in conformational coupling, an effect that is masked by the more dramatic effect of preventing the assembly of the subunits into pentamers.
Are Trp-69 and Trp-94 Critical for the Assembly of all GABA A Receptor Subunits and all Ligand-gated Ion Channels?-Residues Trp-69 and Trp-94 are invariant, not only in all GABA A receptor subunits, but also among all ligand-gated ion channel subunits (Fig. 1). If the two Trp residues perform the same function in all these proteins, as is suggested by their constancy, then this region may contribute to the assembly of all members of the ligand-gated ion channel superfamily. This model is consistent with the proposal of Galzi and Changeux (16) that these two tryptophan residues are structural canonical residues, suggesting that they play the same role in all ligand-gated ion channel receptor subunits.
In conclusion, ␣1W69 and ␣1W94 either directly contribute to or demarcate a region of the ␣1 subunit forming the ␤ 3 ␣ interface and possibly also the ␥ 3 ␣ interface. Mutagenesis of these residues destroys the ␤ 3 ␣ or ␥ 3 ␣ interface and prevents the assembly of the mutant ␣1 subunit into pentamers. Homologous regions on other GABA A receptor subunits are also likely to perform the same function.