Barbiturates Require the N Terminus and First Transmembrane Domain of the δ Subunit for Enhancement of α1β3δ GABAA Receptor Currents*

GABAA receptors are composed predominantly of αβγ receptors, which mediate primarily synaptic inhibition, and αβδ receptors, which mediate primarily extrasynaptic inhibition. At saturating GABA concentrations, the barbiturate pentobarbital substantially increased the amplitude and desensitization of the α1β3δ receptor but not the α1β3γ2L receptor currents. To explore the structural domains of the δ subunit that are involved in pentobarbital potentiation and increased desensitization of α1β3δ currents, chimeric cDNAs were constructed by progressive replacement of γ2L subunit sequence with a δ subunit sequence or a δ subunit sequence with a γ2L subunit sequence, and HEK293T cells were co-transfected with α1 and β3 subunits or α1 and β3 subunits and a γ2L, δ, or chimeric subunit. Currents evoked by a saturating concentration of GABA or by co-application of GABA and pentobarbital were recorded using the patch clamp technique. By comparing the extent of enhancement and changes in kinetic properties produced by pentobarbital among chimeric and wild type receptors, we concluded that although potentiation of α1β3δ currents by pentobarbital required the δ subunit sequence from the N terminus to proline 241 in the first transmembrane domain (M1), increasing desensitization of α1β3δ currents required a δ subunit sequence from the N terminus to isoleucine 235 in M1. These findings suggest that the δ subunit N terminus and N-terminal portion of the M1 domain are, at least in part, involved in transduction of the allosteric effect of pentobarbital to enhance α1β3δ currents and that this effect involves a distinct but overlapping structural domain from that involved in altering desensitization.

loop receptor family, each GABA A receptor subunit is thought to be composed of a large extracellular N terminus, four transmembrane domains (M1-M4), one extracellular M2-3 loop, two intracellular loops (M1-2 and M3-4), and an extracellular C terminus. Theoretically, an enormous number of receptors could be formed with different GABA A receptor subunit/subtype combinations. However, it has been proposed that ␣␤␥ and ␣␤␦ receptor isoforms are the predominant GABA A receptors in the brain (3). There is increasing evidence suggesting that ␣␤␥ GABA A receptors mainly locate within the synapses and mediate GABAergic phasic inhibition, whereas ␣␤␦ receptors are preferentially targeted to extra-or perisynaptic membranes and mediate tonic inhibition (4 -6).
Barbiturates exert their effects in the brain by affecting GABA A receptor functions in a concentration-dependent manner. At lower concentrations, these compounds allosterically modulate GABA A receptors to potentiate GABAergic currents. At higher concentrations, they can directly activate GABA A receptors in the absence of GABA (2,7,8). It has been reported that the barbiturate pentobarbital substantially potentiated peak currents and increased desensitization of ␣1␤3␦ receptor currents evoked by saturating concentrations of GABA, but that these effects of pentobarbital were not observed with ␣1␤3␥2L receptor currents at saturating GABA concentrations, although pentobarbital enhanced ␣1␤3␥2L currents at subsaturating GABA concentrations (8), suggesting that the ␦ subunit rather than the ␥2L subunit confers unique "modulatory potential" for this modulator at saturating GABA concentrations. The structural domains of the ␦ subunit that contribute to the potentiation and desensitization alterations by pentobarbital of currents evoked by saturating concentrations of GABA are currently unknown.
To explore this issue, we took advantage of the differential modulatory effects of pentobarbital on ␣1␤3␦ and ␣1␤3␥2L receptors at a saturating GABA concentration to construct a series of chimeras by progressively replacing the rat ␥2L subunit sequence with the corresponding rat ␦ subunit sequence or ␦ subunit sequence with the ␥2L subunit sequence and co-transfected these chimeric subunits with wild type rat ␣1 and ␤3 subunits. Using an ultrafast drug delivery system and preapplication protocol, we performed whole cell patch clamp recordings of the currents evoked by a saturating concentration of GABA as well as by co-application of a saturating concentration of GABA and pentobarbital. By comparing the peak current amplitude enhancement and kinetic properties of chimeric receptors with those of the wild type ␣1␤3␦ and ␣1␤3␥2L receptors, we explored the structural basis for the differential effects of pentobarbital on ␣1␤3␦ and ␣1␤3␥2L receptor currents evoked by saturating concentrations of GABA.

Expression of Wild Type and Chimeric Recombinant GABA A
Receptors-Chimeric subunits were constructed by gradual replacement of the GABA A receptor ␥2L subunit sequence with the corresponding ␦ subunit sequence or vice versa from the N terminus toward the C terminus using the splice overhang extension method. The cDNAs encoding rat wild type ␣1, ␤3, ␥2L, and ␦ subunits as well as the chimeric subunits were cloned using the expression vector pCMVneo. The sequence of all of the wild type and chimeric subunits was verified by DNA sequencing (Vanderbilt DNA Sequencing Facility, Nashville, TN). Three ␦-␥2L chimeras were constructed by progressively replacing the ␥2L subunit sequence with the ␦ subunit sequence from the N toward the C terminus: ␦-␥2L(M1e) (␦G232-␥Y235), ␦-␥2L(M1pre-iso) (␦Y234-␥T237), and ␦-␥2L(M1p) (␦P241-␥C244). Three reverse ␥2L-␦ chimeras were constructed by progressively replacing the ␦ subunit sequence with the ␥2L subunit sequence from the N toward the C terminus: Human embryonic kidney (HEK293T) cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen), supplemented with 10% fetal bovine serum (Invitrogen), 100 international units/ml penicillin and 100 g/ml streptomycin (Invitrogen) in an incubator at 37°C with 5% CO 2 and 95% air. Cells were transfected with 2 g of each cDNA encoding rat ␣1 and ␤3; ␣1, ␤3, and ␥2L; ␣1, ␤3, and ␦; or ␣1, ␤3, and chimeric GABA A receptor subunits using a modified calcium phosphate precipitation method (8). Two g of pHOOK (Invitrogen) were co-transfected with the GABA A receptor subunits as a marker for the subsequent selection using an immunomagnetic bead separation method (9). Whole cell recordings were performed 24 h after the cells were selected.
Whole Cell Recordings-Whole cell currents were obtained using the patch clamp technique at room temperature. The external solution was composed of 142 mM NaCl, 1 mM CaCl 2 , 6 mM MgCl 2 , 8 mM KCl, 10 mM glucose, and 10 mM HEPES (pH 7.4, 323-329 mosmol). The recording electrodes were pulled from thin wall borosilicate glass tubes (World Precision Instruments, Sarasota, FL) on a P-87 Flaming Brown micropipette puller (Sutter Instruments, San Rafael, CA). The electrodes were fire-polished on an MF-9 microforge (Narishige, Tokyo, Japan). The resistances of the recording electrodes were 0.8 -1.8 megohms after being filled with an internal solution consisting of 153 mM KCl, 1 mM MgCl 2 , 10 mM HEPES, 2 mM MgATP, and 5 mM EGTA (pH 7.3, 301-309 mosmol). Combination of the external and internal solutions produced a chloride equilibrium potential near 0 mV and a potassium equilibrium potential at Ϫ75 mV. Electrophysiological recordings were performed on either an Axopatch-1D or a 200A patch clamp amplifier (Molecular Devices, Union City, CA) and a Digidata 1200 series interface (Molecular Devices). Data were recorded on chart paper using a WR7400 arraycorder (Graph-tec, Yokohama, Japan) as well as stored in a computer for offline analysis. All chemicals were purchased from Sigma-Aldrich. GABA and pentobarbital sodium salt were dissolved in water and prepared as stock solutions. Working solutions were made by diluting the stock solutions to desired concentrations with external solution on the day of the experiment. Drugs were applied by gravity via multibarrel tubes (two three-barrel square glass tubes glued together) connected to a Perfusion Fast-Step system (Warner Instruments, Hamden, CT). The 10 -90% rise times of liquid junction currents were consistently Ͻ2 ms estimated by stepping a dilute external solution across an open electrode tip. GABA was applied for 4 s. The intervals between consecutive drug applications were at least 45 s to minimize desensitization accumulation.
Data Analysis-Whole cell currents were analyzed offline using Clampfit 8.1 (Molecular Devices). The peak currents were measured manually from the base line to the transient peak. The extent of enhancement of GABA currents by pentobarbital (I PB /I CONTROL ) was calculated by dividing the peak current of co-application of GABA and pentobarbital by the peak current evoked by GABA alone. The extent of desensitization (% desensitization) was calculated by dividing the amount of current loss (peak current Ϫ current at the end of GABA/drug application) by peak current. Deactivation currents were fitted with one or two exponential components using the standard exponential Levenberg-Marquardt method in the form of ⌺a n n , where a denotes the relative amplitude of the exponential component, represents the time constant, and n is the number of exponential components. A weighted (a 1 ⅐ 1 ϩ a 2 ⅐ 2 )/(a 1 ϩ a 2 ) was used to compare the rate of deactivation, where a 1 and a 2 are the relative amplitudes of the exponential components at time 0.
Data were reported as mean Ϯ S.E. A paired Student's t test was used to compare the changes prior to and after pentobarbital treatment. One-way analysis of variance followed by Newman-Keuls multiple comparison test was utilized to analyze the differences among wild type and chimeric receptors. The difference was considered to be statistically significant if p Ͻ 0.05.
Pentobarbital Modulated the Current Amplitudes of ␥2L-␦ Chimeric GABA A Receptors-The results obtained using ␦-␥2L chimeras suggested that the N terminus as well as the N-terminal portions of the M1 domain of the ␦ subunit were critical for pentobarbital modulation. However, all of the constructs contained the ␦ subunit sequence in the N terminus, so it was not possible to isolate the role of the M1 domain. Is this domain sufficient to support ␦ subunit-like modulation by pentobarbital? One way to address this issue is to examine reverse ␥2L-␦ chimeras.
Several modulators, including pentobarbital, enhanced ␣␤␦ currents evoked by a saturating concentration of GABA mainly by increasing channel gating efficacy (8,15,18,20). The present study sought to determine the structural domains of the ␦ subunit involved in transduction of the allosteric effect of pentobarbital to current enhancement and alteration of desensitization. We observed that the enhancement by pentobarbital of currents evoked by a saturating concentration of GABA for ␣1␤3␦-␥2L(M1p) receptors was not different from that for ␣1␤3␦ receptors, suggesting that the ␦ subunit domain from the beginning of the N terminus to proline 241 in M1 is sufficient for maximal potentiation by pentobarbital. The enhancement of the reversed chimeric ␣1␤3␥2L-␦(M1pЈ) receptor currents by pentobarbital was not different from ␣1␤3␥2L receptors, supporting the idea that the ␦ subunit domain from the beginning of N terminus to the proline 241 in M1 is necessary and sufficient for maximal potentiation by pentobarbital. The enhancement by pentobarbital of ␣1␤3␥2L-␦(M1eЈ) currents was Ͻ25% of ␣1␤3␦ receptors, implying that the ␦ subunit N terminus plays a critical role in pentobarbital enhancement. However, ␣1␤3␦-␥2L(M1e) currents were only slightly potentiated by pentobarbital. These observations lead to the conclusion that both the N terminus and M1 domain from valine 233 to proline 241 of the ␦ subunit are necessary and sufficient for pentobarbital potentiation of ␣1␤3␦ currents evoked by saturating concentrations of GABA.
Both GABA A and nicotinic cholinergic receptors are members of the Cys-loop receptor family and are proposed to share similar topological structures. It was reported that some of the residues in the N-terminal portion of M1 domain of the nicotinic cholinergic receptor were exposed in the channel lining and might be involved in gating (25). An invariant proline residue in M1 domain of nicotinic cholinergic receptor subunits is critically coupled with ligand binding and channel gating (26). Mutation of the M1 proline (equivalent to proline 241 in the ␦ subunit) in ␣1 or ␤1 subunits reduced the enhancement by barbiturates of submaximal GABA-evoked currents in ␣1␤1␥2 receptors (27), suggesting that the M1 proline is involved in transduction of allosteric effect of barbiturates. An invariant glycine residue (equivalent to glycine 232 in the ␦ subunit) at the entrance to M1 of GABA A receptors also might be involved in transduction of allosteric effect of anesthetics including pentobarbital (28). The present finding that the M1 domain from valine 233 to proline 241 of the ␦ subunit contributes to the pentobarbital potentiation of ␣1␤3␦ currents implies that some of these residues may be directly involved in the transduction of the pentobarbital allosteric effect to channel gating. Furthermore, in addition to the M1 domain residues, we also demonstrated that the N terminus of the ␦ subunit contributed to the pentobarbital potentiation of ␣1␤3␦ currents. Multiple residues in the GABA A receptor N terminus were reported to couple with the M2-3 linker to affect channel gating (29). The finding that the structural domains of the ␦ subunit beyond M1 may not be critically involved in transduction of the pentobarbital allosteric effect suggests that the ␦ subunit N terminus may be able to interact with either the ␦ or ␥2L subunit M2-3 linker. We cannot, however, rule out an interaction between these two domains that occurs via residues that are conserved between the ␦ and ␥2L subunits.
Like the nicotinic cholinergic receptor (30 -34), the M2 domains form the ion conduction pathway and the channel gate of GABA A receptors. Pentobarbital at modulating concentrations did not seem to interfere with the gate via this domain because the M2 domain of ␦ subunit was not required for potentiation of ␣1␤3␦ currents evoked by a saturating concentration of GABA. Some residues in transmembrane regions (including M2 domain) of ␣ and ␤ subunits have been reported to be involved in anesthetic modulation (35)(36)(37)(38)(39). These studies suggested that allosteric modulation of GABA A receptors by anesthetics is subunit-dependent and that different anesthetics may have different actions on GABA A receptors.
It is possible that the domain from the beginning of the N terminus to M1 proline 241 might only confer ␣1␤3␦ receptors with a "partial agonist" property, such that these residues were not necessarily involved in transduction of pentobarbital allosteric effects. If this was the case, the similar structural domain is likely to be involved in the enhancement of ␣1␤3␦ receptor currents by different modulators. However, we found that the neurosteroid tetrahydrodeoxycorticosterone required structural domains beyond M2 for full potentiation of ␣1␤3␦ currents induced by a saturating concentration of GABA. 3 In addition, mutation of the M1 proline in ␣1 or ␤1 of the GABA A receptor only reduced the enhancement by barbiturates without interfering with that by neurosteroids (27). Therefore, this required domain from the beginning of the N terminus to M1 proline 241 is to some extent "specific" for pentobarbital modulation. These residues are unlikely to contribute to the formation of a barbiturate binding site as previous studies in ␦ subunit knock-out mice suggest that general anesthetics like barbiturates, etomidate, and propofol are not selective for ␣␤␦ receptors, although there may be some ␦ subunit selectivity for neurosteroids (40,41).
Interestingly, it has been reported recently that general anesthetics like propofol, etomidate, and tetrahydrodeoxycorticosterone enhanced both ␣4␤3 and ␣4␤3␦ receptors to the same extent when applied with saturating concentrations of GABA (42), indicating that it is the ␣4 instead of the ␦ subunit that confers current enhancement by general anesthetics. However, in the current study, we demonstrated that ␦ subunits play an important role in pentobarbital potentiation of ␣1␤3␦ currents, as pentobarbital greatly enhanced ␣1␤3␦ currents but only slightly enhanced ␣1␤3 currents. Therefore, ␣ subunits may play a role in modulation of ␣␤␦ receptors by anesthetics.
Pentobarbital Required the N Terminus and N-terminal Portion of the ␦ Subunit M1 Domain from Valine 233 to Isoleucine 235 to Alter ␣1␤3␦ Current Desensitization-Pentobarbital decreased the desensitization of ␣1␤3␥2L currents but increased that of ␣1␤3␦ currents induced by a saturating concentration of GABA (8, this study). Pentobarbital did not alter desensitization of ␣1␤3␦-␥2L(M1e) receptor currents and decreased desensitization of most ␣1␤3␥2L-␦ chimeric receptor currents, suggesting that the N terminus of the ␦ subunit may be involved in the pentobarbital-induced increase of desensitization. Some residues in the ␦ subunit M1 domain may also cause the desensitization increase since an increase in desensitization by pentobarbital was observed for ␣1␤3␦-␥2L(M1pre-iso) and ␣1␤3␦-␥2L(M1p) receptor currents. These data imply that ␦ subunit residues from the beginning of the N terminus to M1 isoleucine 235 are a requirement for pentobarbital to increase ␣1␤3␦ current desensitization. Our previous study (10) and the current study (Fig. 1D, left trace) showed that these ␦ subunit residues were required to abolish the fast desensitization of currents evoked by a saturating concentration of GABA. One possibility is that pentobarbital might modulate the rate constants into and out of the desensitized states to increase the occupancy of the desensitized states. It was reported that the desensitized state of nicotinic cholinergic receptors was conferred by some residues lining the inside of the ion conduction pathway and that the desensitized state was structurally different from other states (43). We demonstrated here that the structural domain of the ␦ subunit involved in increasing desensitization overlapped with but was distinct from that involved in transduction of the allosteric effect of pentobarbital.
Pentobarbital Prolonged Deactivation of All Wild Type and Chimeric Receptor Currents-Pentobarbital was reported to prolong the deactivation of ␣1␤3␦ and ␣1␤3␥2L receptor currents evoked by a saturating concentration of GABA (8). It was proposed that prolongation of deactivation is "coupled" with increased desensitization (23,44). Consistent with this idea, pentobarbital prolonged deactivation and enhanced desensitization of ␣1␤3␦ receptors as well as most of the ␦-␥2L chimeric receptors. However, "uncoupling" of desensitization and deactivation was also observed. Pentobarbital decreased the desensitization of ␣1␤3␥2L receptors as well as most of the ␣1␤3␥2L-␦ chimeric receptors and did not modify the desensitization of ␣1␤3␦-␥2L(M1e) receptors, although deactivation of these receptors was prolonged by pentobarbital. These data further support the idea that other factors in addition to desensitization affect deactivation. Deactivation has been demonstrated to be prolonged by increased gating efficacy and agonist affinity (45). Interestingly, we observed that pentobarbital dramatically prolonged deactivation of ␣1␤3␦-␥2L(M1e) receptor currents, but that peak currents and current desensitization were not modulated dramatically by pentobarbital. One possible explanation may be that pentobarbital slowed GABA unbinding, which has been reported for another general anesthetic drug, halothane (46).