A TM2 Residue in the β1 Subunit Determines Spontaneous Opening of Homomeric and Heteromeric γ-Aminobutyric Acid-gated Ion Channels

γ-Aminobutyric acid type A (GABAA) receptors are major inhibitory neurotransmitter-gated ion channels in the central nervous system. GABAA receptors consist of multiple subunits and exhibit distinct pharmacological and channel properties. Of all GABAA receptor subunits, the β subunit is thought to be a key component for the functionality of the receptors. Certain types of GABAA receptors have been found to be constitutively active. However, the molecular basis for spontaneous opening of channels of these receptors is not totally understood. In this study, we showed that channels that contain the β1 but not β3 subunits opened spontaneously when these subunits were expressed homomerically or co-expressed with other types of GABAA receptor subunits in Xenopus oocytes. Using subunit chimeras and site-directed mutagenesis, we localized a key amino acid residue, a serine at position 265, that is critical in conferring an open state of the β1 subunit-containing GABAA receptors in the absence of agonist. Moreover, some point mutations of Ser-265 also produced constitutively active channels. The magnitude of spontaneous activity of these receptors was correlated with the molecular volume of the residue at 265 for both homomeric and heteromeric GABAA receptors, suggesting that the spontaneous activity of the β1 subunit-containing GABAA receptors may be mediated through a similar molecular mechanism that is dependent on the molecular volume of the residue at 265.

␥-Aminobutyric acid type A (GABA A ) receptors are major inhibitory neurotransmitter-gated ion channels in the central nervous system. GABA A receptors consist of multiple subunits and exhibit distinct pharmacological and channel properties. Of all GABA A receptor subunits, the ␤ subunit is thought to be a key component for the functionality of the receptors. Certain types of GABA A receptors have been found to be constitutively active. However, the molecular basis for spontaneous opening of channels of these receptors is not totally understood. In this study, we showed that channels that contain the ␤1 but not ␤3 subunits opened spontaneously when these subunits were expressed homomerically or co-expressed with other types of GABA A receptor subunits in Xenopus oocytes. Using subunit chimeras and site-directed mutagenesis, we localized a key amino acid residue, a serine at position 265, that is critical in conferring an open state of the ␤1 subunit-containing GABA A receptors in the absence of agonist. Moreover, some point mutations of Ser-265 also produced constitutively active channels. The magnitude of spontaneous activity of these receptors was correlated with the molecular volume of the residue at 265 for both homomeric and heteromeric GABA A receptors, suggesting that the spontaneous activity of the ␤1 subunit-containing GABA A receptors may be mediated through a similar molecular mechanism that is dependent on the molecular volume of the residue at 265.
The ␥-aminobutyric acid type A (GABA A ) 1 receptors are the major sites of fast synaptic inhibition and the targets of action of a variety of therapeutic agents such as barbiturates, steroids, anesthetics, and benzodiazepines in the brain. These receptors belong to a superfamily of the Cys-loop pentameric ligand-gated ion channels, which includes nicotinic acetylcholine (nACh), serotonin type 3 (5-HT 3 ), and glycine receptors (1). The topology of these receptors comprises a large extracellular N-terminal domain, a large intracellular loop, and four transmembrane (TM) domains (1). The N-terminal extracellular domain contains the specific binding sites for agonists and antag-onists (2). The TM2 domain is thought to be a key channellining component, which determines channel properties such as conductance, rectification, and desensitization (2).
Molecular cloning has identified a number of receptor subunits including six ␣, four ␤, four ␥, one ␦, one , and one subunit(s) (3). Among these subunits, the ␤ subunit is thought to be a key component to assemble heterooligomeric functional ion channels, to play a central role in determining the subcellular locations of GABA A receptors (4), and to bear binding sites for agonists (5,6) and some clinically important drugs such as general anesthetics (7)(8)(9). The ␤ subunits are also found to be capable of forming homomeric functional channels when expressed in Xenopus oocytes or mammalian cells (5,7,10,11). These homomeric GABA A receptor ion channels have been found to be a valuable approach for localizing molecular determinants of receptor assembly (12,13) and receptor sensitivity to general anesthetics (14,15).
Certain types of heteromeric and homomeric GABA A receptors can form channels that open spontaneously in the absence of agonist (5,7,10,11,13,(15)(16)(17)(18). For homomeric ␤ subunits, the constitutive activity appears to represent a major form of their functionality. Previous studies have reported that the spontaneous channel activity can vary substantially among GABA A receptor channels that contain different ␤ subunits (11,17,19). However, the precise molecular basis for the constitutive activity of GABA A receptors is not totally understood. Here, we investigated whether the difference in spontaneous activity among different ␤ subunits can be explained by localizing discrete sites on the receptor proteins using subunit chimeras and site-directed mutagenesis. Our data show that a single amino acid residue at position 265 in the second transmembrane domain of the ␤1 subunit is crucial for conferring increased opening probability of GABA A receptor ion channels in the absence of agonist. Further molecular analysis found that the magnitude of channel spontaneous opening of GABA A receptor channels is correlated with the volume of the amino acid residue at 265 of the ␤1 subunit.

EXPERIMENTAL PROCEDURES
Chimeric Receptor-DNA fragments encoding the indicated regions of ␤1 and ␤3 subunits were generated by polymerase chain reaction. PCR primers were designed to introduce unique restriction sites into the targeted cDNAs without changing the encoded amino acid sequences. The chimeric ␤1/␤3 cDNAs were constructed by cloning the PCR fragments into appropriate restriction enzyme sites of a pCM-Vscript vector (Stratagene). The chimeric C1 and C2 receptors were constructed by introducing an AflII site at position 213 of the ␤1 and ␤3 subunits. The chimeric C3 and C4 receptors were constructed by introducing a HindIII site at position 314 of the ␤1 and ␤3 subunits. The authenticity of the DNA fragments that flank the mutation site was confirmed by double strand DNA sequencing using an ABI Prism 377 automatic DNA sequencer (Applied Biosystems).
Site-directed Mutagenesis-Point mutations of a cloned rat GABA A receptor were introduced using a QuikChange site-directed mutagenesis kit (Stratagene). The authenticity of the DNA fragments that flank the mutation site was confirmed by double strand DNA sequencing using an ABI Prism 377 automatic DNA sequencer (Applied Biosystems).

Preparation of cRNA and Expression of Receptors-Complementary
RNAs were synthesized in vitro from linearized template cDNAs with mMACHINE RNA transcription kits (Ambion Inc.). The oocytes of mature Xenopus laevis frogs were isolated as described previously (20). Each oocyte was injected with a total of 20 ng of RNA in 20 nl of diethyl pyrocarbonate-treated water. The injected oocytes were incubated at 19°C in modified Barth's solution (88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO 3 , 2.0 mM CaCl 2 , 0.8 mM MgSO 4 , 10 mM HEPES, pH 7.4).
Two-electrode Voltage Clamp Recording-After 2-5 days incubation, the oocytes were studied at 20 -22°C in a 90-l chamber. The oocytes were superfused with modified Barth's solution at a rate of ϳ6 ml/min. Agonists and antagonists were diluted in the bathing solution and applied to oocytes for a specified time using solenoid valve-controlled superfusion. Membrane currents were recorded by a two-electrode voltage clamp technique at a holding potential of Ϫ70 mV using a Ge-neClamp 500 amplifier (Axon Instruments, Inc.). Data were routinely recorded on a chart recorder (Gould 2300S). Average values are expressed as mean Ϯ S.E.
Data Analysis-Statistical analysis of concentration-response curves was performed using the following form of the Hill equation where I is the peak current at a given concentration of agonist A, I max is the maximal response, EC 50 is the half-maximal concentration, and n is the slope factor (apparent Hill coefficient). Data were statistically compared by the unpaired t test or analysis of variance followed by Scheffe's test as noted. Correlation analysis was carried out using nonparametric regression (Statistica, StatSoft).

RESULTS
Homomeric ␤1 but Not ␤3 Subunits Can Form Channels That Open Spontaneously-In oocytes previously injected with cRNA of rat GABA A receptor ␤1 subunit (Fig. 1A, top), 3000 M GABA activated a fast inward current. Application of 100 M bicuculline (BIC), a selective inhibitor of GABA A receptors, did not induce any detectable current. However, picrotoxin (PTX), a chloride channel blocker, at a concentration of 100 M produced a reversible outward current. On the other hand, both BIC and PTX did not induce any response in Xenopus oocytes expressing homomeric ␤3 subunits (Fig. 1A, bottom). To ensure that we could study the extent of channel opening in the absence and presence of agonist at an equivalent basis, we normalized the magnitude of PTX-sensitive outward current as percentage of maximal response, which is the sum of the amplitude of PTX-sensitive current and that of GABA-activated current (Fig. 1B). The majority of the ␤1 subunits appeared to be in a spontaneously active state since the maximal amplitude of outward current produced by PTX represented 88% of the normalized maximal response, which is 8-fold higher than the amplitude of inward current activated by 3000 M GABA. PTX inhibited tonically opened ion channels formed by the ␤1 subunits in a concentration-dependent manner over a concentration range of 1 nM-300 M (Fig. 1C). The EC 50 value and Hill coefficient of the PTX concentration-response curve for homomeric ␤1 subunits were 0.3 Ϯ 0.02 M and 0.6 Ϯ 0.04, whereas PTX in concentrations up to 300 M did not trigger any detectable current in Xenopus oocytes expressing homomeric ␤3 subunits (Fig. 1C). The EC 50 value of PTX that we found for the ␤1 subunits is very similar to that of PTX for the ␤1 subunits reported previously (10).
Certain Types of Rat ␤1 Subunit-containing GABA A Receptors Are Spontaneously Active-The above results suggest that the homomeric ␤1 but not ␤3 subunits are constitutively active. Next, we examined whether a similar scenario could occur in heteromeric expression of the ␣2␤1 or ␤1␥2 subunit combinations. While the average amplitude of maximal GABA-activated current was 27 Ϯ 4 nA (n ϭ 17) for the homomeric ␤1 subunits, the average amplitude of maximal GABA-activated currents was 1370 Ϯ 65 nA (n ϭ 12) for ␣1␤1 and 868 Ϯ 45 nA (n ϭ 21) for ␤1␥2 subunit combinations. The differential sen- subunits. The bars are normalized as percentage of the maximal response (PTX current ϩ GABA current). C, concentration-response curve of PTX-induced outward current in the absence of agonist for homomeric ␤1 receptors (solid squares). The amplitude of current activated by PTX is normalized as percentage of the maximal response. It should be noted that PTX at a concentration range of 0.001-300 M did not induce any detectable response in Xenopus oocytes expressing the ␤3 subunits (solid triangles). sitivity of these receptors to GABA allowed us to clearly distinguish homomeric from heteromeric GABA A receptors. In cells expressing heteromeric subunits, we found that 100 M PTX induced outward current in oocytes co-expressing the ␤1 subunits with either ␣2 or ␥2 subunits ( Fig. 2A). In contrast, PTX did not induce any detectable current in cells previously injected with cRNAs of the ␣2␤3 or ␤3␥2 subunits (Fig. 2B). In addition, no apparent outward current was observed in oocytes previously injected with either H 2 O or cRNAs of the ␣2␤1␥2 subunits on application of 100 M PTX (data not shown). These observations indicate that certain types of GABA A receptors containing the ␤1 subunit can form ion channels that are capable of opening independently of GABA. In addition, we found that the magnitude of the GABA-activated current is inversely correlated with that of the PTX-sensitive outward current ( Fig. 2C; R ϭ Ϫ0.99, a linear regression, p Ͻ 0.001), suggesting that the extent of channel opening in response to GABA depends on a preexisting conformational state of these receptor channels.
Both Homomeric ␤1 and ␤3 Subunits Can Form GABA-gated Ion Channels-Whether or not the ␤1 and ␤3 homomers can form functional GABA-gated ion channels has been controversial. To address this question, we examined further the function of the ␤1 and ␤3 homomers. Oocytes exhibiting inward current in response to 3 mM GABA greater than 25 nA in amplitude were selected for this experiment. Phenobarbital, an allosteric modulator of GABA A receptors, at a concentration of 500 M directly activated inward current when applied alone (not shown) or increased the amplitude of current activated by 10 M GABA (Fig. 3A). In addition, GABA directly activated inward currents in a concentration-dependent manner over a concentration range of 0.1-3000 M (Fig. 3B). The EC 50 and Hill coefficient values of the GABA concentration-response curves were 7.6 Ϯ 4 M and 0.8 Ϯ 0.03, respectively, for the ␤1 subunits and 22 Ϯ 6 M and 0.8 Ϯ 0.06 for the ␤3 subunits. The EC 50 values for the GABA concentration-response curves of the ␤1 and ␤3 subunits are significantly different (p Ͻ 0.02, unpaired t test, n ϭ 10). Next, we determined the sensitivity of these homomeric receptors to BIC and PTX. In cells expressing ␤1 homomers, the application of 100 M BIC slightly reduced the amplitude of current activated by GABA at a concentration of EC 50 , whereas 3 mM GABA failed to induce inward current in the presence of PTX (Fig. 3C). This suggests that PTX can completely inhibit GABA-activated current in oocytes expressing the ␤1 homomers and that the outward current induced by PTX is mediated through channels formed by homomeric ␤1 subunits. In cells expressing ␤3 homomers, although neither 100 M BIC nor 100 M PTX induced detectable outward current, these concentrations of both BIC and PTX significantly inhibited GABA-activated current. The bar graphs in Fig. 3D show the average percentage inhibition of GABA-activated inward current by 100 M PTX and 100 M BIC in cells expressing homomeric ␤1 (solid bars) and ␤3 subunits (open bars). PTX (100 M) inhibited currents activated by an EC 50 concentration of GABA by nearly 100% in oocytes expressing either ␤1 or ␤3 homomers (Fig. 3D). On the other hand, 100 M BIC nearly completely inhibited current activated by an EC 50 concentration of GABA in cells expressing homomeric ␤3 subunits but had only a very small inhibitory effect in cells expressing homomeric ␤1 subunits.
The Constitutive Activity of the ␤1 Subunits Was Not Affected by the Y205F Mutation-To investigate the molecular mechanisms by which the ␤1 receptor channels open spontaneously, we first tested whether reduction of agonist binding affinity alters the spontaneous opening of the receptor channels. To do this, we substituted tyrosine at position 205, a previously described agonist-binding site in the extracellular N-terminal domain of the ␤1 subunit (21), with phenylalanine. Consistent with a previous study (21), the Y205F mutation shifted the GABA concentration-response curve to the right in a parallel manner (Fig. 4A) and increased the EC 50 value by ϳ10-fold ( Fig. 4B; 7.6 Ϯ 4 M for the wild type (WT) and 82 Ϯ 5 M for the Y205F receptors, p Ͻ 0.001, unpaired t test). However, the sensitivity of the Y205F mutant receptors to PTX-sensitive current was nearly identical to that of the wild type receptors (Fig. 4, C and D), suggesting that the agonist-binding site is unlikely to be involved in the mechanisms that underlie the constitutive activity of homomeric ␤1 subunits.
Chimeric Constructs: TMs (1-3) of the ␤1 Subunit Are Associated with the Channel Spontaneous Opening-In view of our observation that homomeric ␤1 and ␤3 subunits of GABA A receptors exhibit a difference in PTX-induced outward current, we thought that chimeras between the ␤1 and ␤3 subunits might be an ideal approach to localize molecular domains that may be involved in the spontaneous opening of GABA A receptor channels, and therefore we constructed chimeras between the ␤1 and ␤3 subunits. Four chimeric receptors were generated, and the amplitude of PTX-induced outward current was determined. As shown in Fig. 5A, the chimeras that replaced the N terminus of the ␤1 subunit (C1) and the C terminus of the ␤1 subunit (C4) with the corresponding segments of the ␤3 subunit exhibited PTX-induced outward current, suggesting that the extracellular N-terminal domain, the large intracellular loop between transmembrane domains 3 and 4, and the fourth transmembrane domain as well as the extracellular C-terminal domain of the ␤1 subunit are not essential for spontaneous opening of the channels in the absence of GABA. However, the chimeras that contained a region from TM1 to TM3 of the ␤3 subunit (C2 and C3) became insensitive to PTX inhibition of channel spontaneous opening (Fig. 5B), suggesting these transmembrane domains of the ␤1 subunit may be critical for the constitutive activity of the receptors in the absence of agonist. In addition, to determine whether there is a relationship between the amplitude of PTX-induced current in the WT and chimeric receptors and the sensitivity of these receptors to GABA, we first determined the EC 50 s of the GABA concentration-response curves for these receptors as shown in Fig. 5C. Next we compared the amplitude of PTX-sensitive current with the EC 50 values for the GABA concentration-response curves of the wild type and chimeric receptors (Fig. 5D). We found that there was no correlation between the EC 50 values and the spontaneous activity of these receptors (R ϭ 0.22, a linear regression, p Ͼ 0.5, n ϭ 5).
Point Mutations: A Residue (Ser-265) in the TM2 Confers Spontaneous Activity of the ␤1 Subunit-containing GABA A Receptors-To identify the site or sites responsible for the spontaneous opening of the homomeric ␤1 subunits, we aligned the amino acid sequences that flank a segment between TM1 and TM3 of the ␤1 and ␤3 subunits. Within this region, the ␤1 and ␤3 subunits differ by only four amino acid residues (Fig. 6A). We then substituted each of these residues of the ␤1 subunit with the corresponding residue of the ␤3 subunit. Fig. 6B shows the GABA concentration-response curves for the wild type and mutant ␤1 receptors. The EC 50 and Hill coefficient values were, respectively, 30 Ϯ 3 M and 1.0 Ϯ 0.1 for T255I/L256M receptors (solid circles), 7.7 Ϯ 6 M and 0.7 Ϯ 0.2 for S265N receptors (solid triangles), and 68 Ϯ 10 M and 0.8 Ϯ 0.09 for the I283M receptors (solid diamonds) (Fig. 6B). Except for the S265N mutation, which did not alter the sensitivity of the receptor to GABA, the T255I/L256M and I283M mutations significantly decreased the sensitivity of the receptors to GABA by 4-and 9-fold, respectively (p Ͻ 0.01, unpaired t test, n ϭ 5). However, of these point mutations, the S265N mutation was the only point mutation that abolished spontaneous activity of the ␤1 subunits (Fig. 6C), suggesting that the amino acid residue at position 265 in the ␤1 subunit is critical for the channel spontaneous opening. To determine whether amino acid Ser-265 of the ␤1 subunit is also important for the spontaneous opening of heteromeric GABA A receptor channels, we co-expressed ␤1 (S265N) mutant subunits with ␥2 or ␣2 subunits. The trace records in Fig. 7A show that 100 M PTX induced an outward current in oocytes expressing the ␤1␥2 subunits. However, 100 M PTX did not induce a detectable outward current in oocytes co-expressing ␤1 (S265N) mutant with ␥2 subunits. The S265N mutation of the ␤1 subunit also blocked the spontaneously opening channels produced by the ␣2␤1 subunits expressed in Xenopus oocytes (Fig. 7B). It should be noted that the N265S mutation of the ␤3 subunit did not produce spontaneously active channels in cells expressing homomeric ␤3(N265S) subunits or co-expressing ␤3(N265S)␥2 subunits (data not shown).
Point Mutations: The Molecular Volume of the Residue at Position 265 of the ␤1 Subunit Is Correlated with the Spontaneous Activity of GABA A Receptors-The observations above suggest that residue 265 in the ␤1 subunit is critical for the spontaneous opening state of these receptor channels in the absence of agonist. To gain molecular insight into the structural/functional role of the residue at position 265, we replaced Ser-265 with multiple amino acid residues and examined the spontaneous activity of mutant ␤1␥2 subunits. Among seven mutant receptors, the ␤1(S265W)␥2 and ␤1(S265G)␥2 mutant receptors were constitutively active. Next, we used correlation analysis to compare the magnitude of the PTX-induced outward current with the hydropathicity (22), polarity (23), hydrophilicity (24), and molecular volume (25) of the amino acid residues replaced at position 265. In general, there is no significant difference among these variables (Fig. 8, A, B, C, and  D). However, because these receptors clearly fall into two distinct groups based on whether they open or not in the absence of agonist, we classified the receptors into two groups, those that exhibit spontaneous activity, group 1, and those that do not, group 2. For the group 1 receptors, the strongest correlation was found between the side chain molecular volume of the residues at position 265 and the extent of spontaneous channel opening ( Fig. 8D; R ϭ 0.99, p Ͻ 0.0001, nonparametric analysis). A similar scenario was observed for other spontaneously opening ␤1-containing heteromeric GABA A receptors (Fig. 8, E  and F). DISCUSSION In this study we have demonstrated that homomeric and heteromeric expression of rat GABA A receptors that contain the ␤1 subunit were constitutively active, whereas GABA A receptors that contain the ␤3 subunit were inactive in the absence of GABA. These observations are consistent with previous studies showing that homomeric GABA A receptor channels formed by ␤1 but not ␤3 subunits are constitutively active (7,10,15). In addition, we found that the magnitude of spontaneous opening of homomeric ␤1 receptor channels was predominant as the channel spontaneous activity accounted for 88% of total extent of the opening probability. In line with previous studies (5, 26), we also observed spontaneously opening ␤1-containing heteromeric GABA A receptor channels. It is FIG. 5. Identification of molecular domains that mediate the spontaneous activity of the ␤1 homomers by chimeric constructs. A, schematic illustration of chimeric receptors constructed between the ␤1 and ␤3 subunits and the sensitivity of these receptors to PTX inhibition of spontaneous activity. B, bar graph showing the average outward current induced by PTX in the absence of GABA for each of the wild type and chimeric receptors. Each bar is the average from 4 -5 cells. C, the GABA EC 50 values for the wild type and chimeric receptors. D, the EC 50 values of GABA concentration-response curves are not correlated with the magnitude of PTX-induced response in the absence of GABA for the wild type and chimeric receptors.
unlikely that lack of constitutive activity from rat ␤3 subunits was due to insensitivity of the homomeric ␤3 subunits to GABA A receptor antagonists since both BIC and PTX were found to completely inhibit GABA-activated inward current in oocytes expressing the homomeric ␤3 subunits.
It has been well documented that homomeric ␤ subunits can form functional channels that either can open spontaneously or can be directly activated by some general anesthetics (5,7,11,19,27). However, whether or not homomeric ␤ subunits can form functional GABA-gated ion channels is still controversial and appears to depend on different species. For instance, while human and bovine ␤1 receptors were found to form channels that can be gated by GABA (27)(28)(29)(30), rat and mouse ␤1 and ␤3 subunits were insensitive to GABA (5,11,19). In the present study, we found that GABA activated an inward current in oocytes expressing either homomeric ␤1 or ␤3 subunits. The amplitude of inward current was increased by barbiturate and inhibited by a selective GABA A receptor antagonist, bicuculline, suggesting that the inward current activated by GABA was mediated through homomeric ␤1 and ␤3 subunits. Although the precise reason behind the different results from our laboratory and other laboratories is not totally understood, there are a number of possibilities that could contribute, at least in part, to this discrepancy. First, it could be due to different levels of homomeric expression of the ␤1 subunits. Second, it may depend on the level of posttranslational modulation of the subunits, which could vary pronouncedly among different batches of Xenopus oocytes. Consistent with this hypothesis, we found that the major differences in the amino acid sequences between human and rat ␤1 subunits occur within the large intracellular loop between TM3 and TM4 domains. Another possibility to reconcile this discrepancy could be due to different levels of spontaneous activity of the ␤1 subunits expressed in oocytes under different experimental conditions. We and others have found that the magnitude of spontaneous activity appeared to be so predominant that it nearly overshadowed the magnitude of GABA-activated current in cells expressing homomeric ␤1 subunits (27)(28)(29)(30). Overall, the magnitude of PTX-sensitive current was 5ϳ9-fold larger than that of GABA-activated current. We also observed that the magnitude of spontaneous opening was inversely correlated with magnitude of GABA-activated current in oocytes expressing different combinations of homomeric and heteromeric GABA A receptors that contain the ␤1 subunit. This indicates that the tendency of homomeric ␤1 subunits to open spontaneously increases with a decrease in the sensitivity of these receptors to GABA. It is therefore plausible to predict that when homomeric ␤1 receptor channels open spontaneously at the maximal probability, these receptors might no longer respond to activation by GABA.
We observed that the magnitude of channel spontaneous opening was not affected by the Tyr-205 mutation, a distinct agonist-binding site of the ␤1 subunits. This observation raises the possibility that molecular mechanisms by which the GABA A receptor channels open in the absence and presence of ligands may be different. This hypothesis is consistent with our finding that a TM2 residue is critical for channel spontaneous opening of the ␤1 subunits and is also in line with a previous study showing that the Y205F mutation did not alter spontaneous activity induced by substitution of a highly conserved leucine in the TM2 of GABA C receptors (21).
The most important finding of this study is that residue 265 of the ␤1 subunit is found to be critical for channel spontaneous opening of GABA A receptors. In addition, we have revealed that the magnitude of such a spontaneous activity is correlated with the molecular volume of the side chain of the residue 265 for different combinations of ␤1-containing GABA A receptors, indicating that the spontaneous activity of these receptor channels may be mediated through a molecular mechanism that depends on the molecular volume of the residue at position 265. The residue corresponding to Ser-265 of the ␤ subunits has been the focus of a large number of recent studies of glycine and GABA A receptors, particularly in the area of alcohol and general anesthetic research (31). This particular residue has been found to be a critical site that determines the sensitivity of GABA A receptors to ethanol, general anesthetics, and anticonvulsant agents in vitro (15,32,33) and in vivo (9,34). Moreover, the sensitivity of glycine and GABA A receptors to ethanol and general anesthetics is inversely correlated with molecular volume of a residue equivalent to Ser-265 (35,36). This particular residue has been thought to be a binding site for alcohol and general anesthetics of glycine and GABA A receptors (32,37). The results presented in this study have indicated that the residue at position 265 of the ␤1 subunit determines the pre-existing conformational state of GABA A receptor protein and therefore is critical for channel-gating dynamics. This conclusion is consistent with a recent kinetic analysis that point mutations of a residue equivalent to Ser-265 of the ␣2 subunit can modulate the gating efficacy of GABA A receptors (38). Several recent studies have shown that the sensitivity of certain types of Cys-loop pentameric ligand-gated ion channels to agonist and allosteric modulators such as alcohol and general anesthetics may depend, at least in part, on the preexisting conformational states of these receptor channels (39 -42). It appears that the receptor channels that open spontaneously could become less sensitive to potentiation by ethanol and general anesthetics (42). This hypothesis is favored by observations from this and other studies; with increase of molecular volume of the residue at position 265, the sensitivity of GABA A receptors to volatile anesthetics decreases, whereas the spontaneous activity of these receptor channels increases (15,39).
It should be noted that molecular basis for spontaneous activity of GABA A receptor channels is complicated. Although our results presented in this study suggest that Ser-265 of the ␤1 subunits confers channel spontaneous opening, previous studies also showed that such a spontaneous activity of the receptor channels could depend on receptor assembly and stoichiometry (12,13). It is also unclear which combinations of GABA A receptor subunits may form channels that can open spontaneously in vivo. Although there is evidence suggesting that other types of GABA A receptor subunits also could be involved in channel spontaneous activity of GABA A receptors (5,43), the results from this and previous studies suggest that such spontaneous activity of the wild type GABA A receptors may, at least in part, rely on the presence of distinct ␤ subunits (5,7,10,11,27). There is also evidence showing that spontaneous opening of GABA A receptor channels can be detected in spinal cord neurons (44) and in pituitary cells (45,46). However, the physiological significance of spontaneous activity of GABA A receptors remains unclear, given the fact that spontaneously opening GABA-gated ion channels are somehow difficult to identify in vivo because of background GABA release.
In summary, we have identified a particular amino acid residue, Ser-265, in the TM2 of the ␤1 subunit as a critical site that confers spontaneous opening of GABA A receptor channels. The magnitude of spontaneous activity of these receptors is dependent on the molecular volume of the residue at position 265. We have proposed that this particular residue in the ␤1 subunit may serve as a key structural element, which confers an open state of GABA A receptor channels in the absence of agonist by lowering the energy barrier that is required for channel opening. These observations should help to enhance our understanding of molecular mechanisms by which GABA A receptor channels can open spontaneously. The study reported here also provides some molecular details for the structural/ functional role of the residue at position 265 in determining the preexisting conformational state of GABA A receptor channels. Finally, our analysis together with others of the residue at position 265 of GABA A receptors should raise the possible argument against a proposed hypothetical "anesthetic binding pocket" that involves the residue Ser-265.