Advertisement
Research ArticleArticle

γ-Aminobutyric Acid Type A α4, β2, and δ Subunits Assemble to Produce More Than One Functionally Distinct Receptor Type

Megan M. Eaton, John Bracamontes, Hong-Jin Shu, Ping Li, Steven Mennerick, Joe Henry Steinbach and Gustav Akk
Molecular Pharmacology December 2014, 86 (6) 647-656; DOI: https://doi.org/10.1124/mol.114.094813

Abstract

Native γ-aminobutyric acid (GABA)A receptors consisting of α4, β1–3, and δ subunits mediate responses to the low, tonic concentration of GABA present in the extracellular milieu. Previous studies on heterologously expressed α4βδ receptors have shown a large degree of variability in functional properties, including sensitivity to the transmitter. We studied properties of α4β2δ receptors employing free subunits and concatemeric constructs, expressed in Xenopus oocytes, HEK 293 cells, and cultured hippocampal neurons. The expression system had a strong effect on the properties of receptors containing free subunits. The midpoint of GABA activation curve was 10 nM for receptors in oocytes versus 2300 nM in HEK cells. Receptors activated by the steroid alfaxalone had an estimated maximal open probability of 0.6 in oocytes and 0.01 in HEK cells. Irrespective of the expression system, receptors resulting from combining the tandem construct β2-δ and a free α4 subunit exhibited large steroid responses. We propose that free α4, β2, and δ subunits assemble in different configurations with distinct properties in oocytes and HEK cells, and that subunit linkage can overcome the expression system-dependent preferential assembly of free subunits. Hippocampal neurons transfected with α4 and the picrotoxin-resistant δ(T269Y) subunit showed large responses to alfaxalone in the presence of picrotoxin, suggesting that α4βδ receptors may assemble in a similar configuration in neurons and oocytes.

Introduction

γ-Aminobutyric acid (GABA)A receptors are pentameric channel proteins that respond to the binding of the transmitter GABA with an opening of a channel pore that is permeable to Cl− and other small anions. GABAA receptors containing α4, β1–3, and δ subunits are mainly located in the cell soma and mediate responses to tonic GABA (Farrant and Nusser, 2005; Belelli et al., 2009). These receptors have high affinity for GABA and, unlike synaptic GABAA receptors, demonstrate little desensitization even during persistent presence of agonist, in accordance with their predicted role as mediators of a steady inhibitory tone on a neuron (Haas and Macdonald, 1999; Brown et al., 2002; You and Dunn, 2007; Karim et al., 2012).

Functional studies have shown ambiguity regarding the properties of recombinant α4βδ receptors. For example, there are conflicting reports on the GABA concentration-response relationship. Several studies had observed GABA activation curves for α4βδ receptors in Xenopus oocytes with half-maximal concentrations at 1–3 µM (Storustovu and Ebert, 2006; You and Dunn, 2007; Hoestgaard-Jensen et al., 2010). In contrast, Karim et al. (2012) found that the GABA activation curve for α4β3δ receptors expressed in Xenopus oocytes contained two components: one with the concentration of GABA producing a half-maximal response (EC50) at ∼10 nM, the other with the EC50 at ∼1 µM. From inhibition by Zn2+ and comparison with properties of binary α4β3 receptors, it was proposed that the micromolar component reflected the response of receptors containing only α4 and β3 subunits, whereas the high-affinity component represented activity of receptors containing all three (α4, β3, and δ) subunits. A single-component GABA activation curve with EC50 at 24 nM was observed for α4β1δ receptors (Karim et al., 2012).

Although the lack or incomplete inclusion of the δ subunit in surface receptor complexes could account for conflicting data, another possibility is that the stoichiometry of α4βδ GABAA receptors is flexible, dependent on experimental conditions. Work on concatemeric receptors that demonstrate functionality of receptors formed by expression of constructs in which the δ subunit is in a different predicted position in the receptor (Kaur et al., 2009) supports this idea.

To attempt to resolve some of the ambiguities, we studied the activation properties of GABAA receptors consisting of free α4, β2, and δ subunits and various concatemeric constructs. Our data show a GABA activation curve with an EC50 at 10 nM in Xenopus oocytes. However, estimated EC50 value for activation by GABA was over 200-fold greater in HEK cells. Furthermore, receptors activated by the steroid alfaxalone (ALF) had an estimated maximal open probability of 0.6 in oocytes, but only 0.01 in HEK cells. The use of concatemeric receptors showed that subunit linkage could overcome the effect of the expression system, thereby suggesting that the preferred stoichiometry or subunit order of α4β2δ receptors is different in HEK cells compared with oocytes. Finally, α4βδ receptors overexpressed in hippocampal neurons displayed similarities to receptors expressed in oocytes rather than HEK cells.

Materials and Methods

Clones, Constructs, Mutagenesis, and Expression.

The experiments were conducted on combinations of free and concatemeric α4, β2, and δ subunits of the GABAA receptor. The cDNA for α4 was obtained from P. Whiting (Merck, Harlow, UK); β2 was obtained from D. Weiss (University of Texas Health Science Center, San Antonio, TX); and the cDNA for δ was provided by R. Macdonald (Vanderbilt University, Nashville, TN). The species of origin for the α4 subunit is human, whereas the origin of β2 and δ subunits is rat. In some experiments, mutated subunits were used. These were α4(F63C), δ(F74C) (both in loop D at the “-” interface of the extracellular domain), and δ(T269Y) (T6′Y in the second membrane-spanning domain). The mutations to loop D residues were made to test the contributions of α4 and δ subunits to receptor activation by GABA (Petrini et al., 2011). The δ(T269Y) mutation was used to confer resistance to picrotoxin (Gurley et al., 1995). The mutations were made using the QuikChange site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). The δ subunit contained the FLAG epitope in the amino terminus of the subunit (Ueno et al., 1996).

The generation and initial characterization of δ-β2-α4 and β2-α4 concatemeric constructs have been described in an earlier publication (Shu et al., 2012). Additional concatemers, β2-δ-β2 and β2-δ-α4, were made as follows. Using the δ-β2-α4 construct as a template, we generated a fragment containing part of δ and all of the β2 subunit with additional sequence at the 3′ end encoding the restriction sites XbaI and ApaI for cloning purposes. This fragment and the DNA construct β2-δ were digested with the endonucleases SacII (in δ) and ApaI. The δ sequence was substituted with the δ-β2 sequence through standard subcloning methods, generating the β2-δ-β2 concatemer. Generation of the β2-δ-α4 construct was done using β2-δ-β2 as the recipient vector and a polymerase chain reaction fragment containing part of δ and α4. Both of these were digested with SacII and XbaI, substituting the δ-β2 sequence in the parent concatemer with the δ-α4 fragment using standard subcloning methods.

The α4 and δ subunits within concatemeric constructs contained mutations to remove the putative dibasic retention signals in the M3-M4 intracellular loop. The wild-type α4 sequence was modified from MEKAKRKTS to MEAAAVATS. The wild-type δ sequence was modified from ISRRQG to ISAAQG. The mutations were introduced to the δ-β2-α4, β2-δ-β2, β2-δ-α4, β2-δ, and β2-α4 constructs to enhance expression levels. We have previously shown that these mutations do not affect activation or basic pharmacological properties of the receptor (Bracamontes et al., 2014).

All cDNAs were subcloned into the pcDNA3 expression vector in the T7 orientation. For expression in Xenopus laevis oocytes, the cDNAs were linearized by XbaI (NEB Laboratories, Ipswich, MA) digestion, and cRNAs were produced using mMessage mMachine (Ambion, Austin, TX). The oocytes were injected with 3–15 ng cRNA per construct in a final volume of 20–60 nl and incubated in ND96 (96 mM NaCl, 2 mM KCl, 1.8 mM CaCl2, 1 mM MgCl2, 2.5 mM Na pyruvate, 5 mM HEPES, pH 7.4) at 16°C for 1–6 days before recording. The ratio of cRNAs used for injection was 5:1:5 for free α4, β2, and δ subunits. It was 5:1 when only α4 and β2 subunits were used, and 1:1 in the case of combinations of concatemeric constructs. A relative excess of δ and α4 cRNAs was used to enhance expression of the δ subunit and to reduce formation of β2 homo-oligomers.

Transfection of HEK 293 cells was carried out using a calcium phosphate precipitation-based transient transfection technique (Akk, 2002). A total of 3 μg cDNA in the ratio of 1:1:1 (α:β:δ) was mixed with 12.5 μl 2.5 M CaCl2, and dH2O to a final volume of 125 μl. In some experiments, a cDNA ratio of 1:1:5 was used. This mixture was added to 125 µl 2× N,N-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid buffered solution. The combined mixture was incubated at room temperature for 10 minutes, followed by gentle mixing of the contents and an additional 15-minute incubation. The precipitate was added to the cells in a 35-mm dish for overnight incubation at 37°C, followed by replacement of medium in the dish the next day. The experiments were conducted in the course of the next 2 days after changing the medium.

Primary microcultures of hippocampal cells were prepared from 1- to 3-day postnatal Sprague Dawley rats, as described previously (Mennerick et al., 1995). The α4 and δ(T269Y) subunits were transiently expressed in hippocampal microculture neurons 8 days after plating. The cDNA amount was 0.6 µg α4 and 0.3 µg δ along with reporter plasmid (green fluorescence protein; Clontech Laboratories, Mountain View, CA) as a positive transfection marker. Lipofectamine2000 (Life Technologies, Carlsbad, CA) was used as transfection reagent, according to the manufacturer’s protocol. Electrophysiological experiments were performed 48–72 hours following transfection.

Electrophysiological Recordings and Data Analysis.

Electrophysiological experiments in Xenopus oocytes were conducted using standard two-electrode voltage clamp. Voltage and current electrodes were patch-clamp electrodes that when filled with 3 M KCl had resistances of 0.5–1.5 MΩ. The oocytes were clamped at −60 mV. The chamber (RC-1Z; Warner Instruments, Hamden, CT) was perfused continuously at approximately 5 ml min−1. Bath solution (92.5 mM NaCl, 2.5 mM KCl, 1 mM MgCl2, 10 mM HEPES, pH 7.4) was perfused between all test applications. Solutions were gravity applied from 30-ml glass syringes with glass Luer slips via Teflon tubing to reduce adsorption, and switched by pClamp using a Warner Instruments VC-8T valve controller. A typical drug application consisted of recording a 10-second baseline, followed by a 30-second drug application and a bath application (up to 10 minutes) until full recovery. The current responses were amplified with an Axoclamp 900A amplifier (Molecular Devices, Sunnyvale, CA), digitized with a Digidata 1320 series digitizer (Molecular Devices) at a 100 Hz sampling rate, and stored using pClamp (Molecular Devices). The traces were analyzed with Clampfit (Molecular Devices) to determine the peak amplitude of current response.

HEK cells expressing high levels of GABAA receptors were identified using a bead-binding technique. The amino terminus of the δ subunit had been tagged with the FLAG epitope (Ueno et al., 1996). Surface expression of the FLAG epitope was verified using a mouse monoclonal antibody to the FLAG epitope (M2; Sigma-Aldrich, St. Louis, MO), which had been adsorbed to immunobeads with a covalently attached goat anti-mouse IgG antibody (Life Technologies).

Experiments on HEK cells were conducted on lifted cells using the standard whole-cell voltage clamp technique. The bath solution contained (in mM) the following: 140 NaCl, 5 KCl, 1 MgCl2, 2 CaCl2, 10 d-glucose, and 10 HEPES, pH 7.4. The pipette solution contained (in mM) the following: 140 CsCl, 4 NaCl, 4 MgCl2, 0.5 CaCl2, 5 EGTA, and 10 HEPES, pH 7.4. The drugs were applied onto the cells using a SF-77B fast perfusion stepper system (Warner Instruments). The solution exchange time, estimated from switching the bath solution from 100 to 75% Ringer solution, was 10–20 ms with a typical HEK cell. A typical drug application consisted of recording a 1-second baseline, followed by a 4- to 8-second drug application and a bath application (up to 1 minute) until full recovery. The cells were clamped at −60 mV. The currents were recorded using an Axopatch 200B amplifier (Molecular Devices), low-pass filtered at 2 kHz, and digitized with a Digidata 1320 series interface (Molecular Devices) at 10 kHz. The analysis of current traces aimed at determining the peak current was conducted using pClamp 9.0 software (Molecular Devices).

Whenever feasible, the current responses are expressed in units of estimated open probability (Poest) of the channel. The latter is a useful parameter because it reflects a basic physiologic property of an ion channel and is unbiased by caveats inherent in normalization to the response to highest concentration of the activator. The theoretical range of Po spans from 0 to 1. The conversion of raw data to Poest involved determination of a drug combination whose response could no longer be enhanced by the δ-specific potentiator 4-chloro-N-[2-(2-thienyl)imidazo[1,2-a]pyridine-3-yl]benzamide (DS-2) or the steroid ALF, and then normalization to the response to that drug combination. Such an approach has been described previously (Rusch et al., 2004). For example, in the case of α4β2δ receptors, we observed that responses to saturating pentobarbital (PEB) were larger than responses to any other agonist employed, and that the responses to saturating (0.5 mM) PEB could not be enhanced by DS-2 or ALF. Accordingly, we considered the response to 0.5 mM PEB to have a Poest of ∼1. Peak responses to GABA or neurosteroids were then normalized to the response to 0.5 mM PEB for expression in units of Poest. The saturating concentration of PEB was determined for each receptor type individually. In oocytes expressing the combination of δ-β2-α4 and β2-α4 constructs, responses to PEB saturated at 2 mM. Furthermore, responses to saturating PEB could be further potentiated by 1 µM DS-2. Accordingly, the response to 2 mM PEB + DS-2 was considered to have a Poest of ∼1, to which responses to GABA and steroid were normalized.

This approach was adopted to remove some of the ambiguity inherent when responses are simply normalized to the maximal response to particular agonists, and so to allow more straightforward comparisons among constructs. However, we note that there is one essential caveat: the accuracy of this normalization requires that the maximal (potentiated) response have an open probability approaching 1. Two factors could invalidate this assumption. One is experimental. If the peak response is experimentally underestimated due to slow perfusion (for example, for the rebound tail resulting from unblock), then the maximal Poest for GABA would be artifactually increased. The second reason arises from the possibility that the mechanism of potentiation might be affected in some constructs so that potentiators were inefficacious. In this case, again, the potentiated response would not necessarily approach a Popen 1. In the case of α4β2δ receptors expressed in HEK cells, direct evidence from single-channel recordings demonstrates a low-maximal Popen for GABA (Akk et al., 2004). In the case of other constructs, there is no direct evidence. However, we believe that the use of Poest provides a more consistent parameter for comparisons among different constructs.

Pentobarbital blocks GABAA receptors at millimolar concentrations, resulting in a characteristic response with a small initial peak, followed by a large tail or rebound response upon the termination of application of PEB (Akaike et al., 1987). The tail response reflects transient repopulation of the conducting state(s) of the channel. In situations in which the amplitude of the tail response was greater than the initial peak response (e.g., Fig. 2B), we used the tail or rebound response that reflects recovery from block, rather than peak response for comparing current amplitudes.

For recordings from neurons, the coverslips with hippocampal microculture neurons were transferred from culture medium to an extracellular recording solution consisting of (in mM) the following: 138 NaCl, 4 KCl, 2 CaCl2, 1 MgCl2, 10 glucose, and 10 HEPES, pH 7.25. Transfected hippocampal microculture neurons were selected for electrophysiology based on green fluorescence protein. Patch pipettes were filled with an internal solution containing (in mM) the following: 130 CsCl, 4 NaCl, 5 EGTA, 0.5 CaCl2, and 10 HEPES, pH 7.25. When filled with this solution, pipette tip resistance was 3–6 MΩ. Cells were clamped at −70 mV. Drug applications were made with a multibarrel, gravity-driven local perfusion system. The estimated solution exchange times were 120 ± 14 ms (10–90% rise), measured by the change in junction currents at the tip of an open patch pipette. Currents were filtered at 5 kHz and recorded with an Axopatch 200B amplifier (Molecular Devices). All recordings were performed at room temperature.

Where applicable, the effects of drugs are presented in percentage of control, that is, 100% represents no effect. Parameters are shown as mean ± S.D. for responses. The Hill equation was fit to the mean data for a given situation (constructs, drugs), and the best-fitting parameter value is given ± the estimated S.E. of the fit provided by the fitting program.

Chemicals.

GABA, pentobarbital, and the inorganic salts were purchased from Sigma-Aldrich (St. Louis, MO). DS-2 was purchased from R&D Systems (Minneapolis, MN). The steroids were from Sigma-Aldrich, R&D Systems, or Steraloids (Newport, RI).

Results

Activation of α4β2δ Receptors by GABA in Xenopus oocytes.

We commenced by examining activation of α4β2δ GABAA receptors by the transmitter GABA. Oocytes injected with cRNAs for α4, β2, and δ subunits exhibited robust responses to GABA. In the presence of saturating GABA (300 nM), the peak response was 2265 ± 829 nA (mean ± S.D., 7 cells, range: 1427–3922 nA). The concentration-response relationship for GABA was examined by exposing oocytes to 1–300 nM GABA. Sample traces are shown in Fig. 1A. Curve fitting to the Hill equation yielded an EC50 of 10.3 ± 0.1 nM and a Hill coefficient of 0.8 ± 0.1 (best-fitting parameter ± uncertainty of the fit; 7 cells; Fig. 1B). In three cells, the GABA concentration range was extended to 30 µM. In these cells, the relative mean peak responses to 1, 3, 10, and 30 µM GABA were 92 ± 14, 94 ± 17, 93 ± 23, and 93 ± 23% (mean ± S.D.), respectively, of the mean peak response to 300 nM GABA. Incorporation of data from these cells had only a small effect on the concentration-response relationship (the revised EC50 was 7 ± 0.9 nM, and the revised Hill coefficient was 0.9 ± 0.1; dashed line in Fig. 1B). We infer that the GABA concentration-response relationship for α4β2δ receptors expressed in Xenopus oocytes contains a single, nanomolar component.

Fig. 1.
Fig. 1.

Activation of α4β2δ receptors expressed in Xenopus oocytes. (A) Representative current responses in the presence of 1, 3, 10, 100, and 300 nM GABA. All traces are from the same oocyte. (B) A GABA concentration-response relationship for receptors formed from free α4, β2, and δ subunits. The data points show mean ± S.E.M. for data from seven cells. The data are in units of Poest that was estimated by comparing responses to GABA to a response to saturating PEB in the same cells (see Materials and Methods for more details). The line through the points shows the predictions of the Hill equation fit to the means, yielding an EC50 of 10.3 ± 0.1 nM and a Hill coefficient of 0.8 ± 0.1 (best-fit parameter ± uncertainty of the fit). The dashed line shows a fit resulting from the inclusion of data (small circles) obtained from three cells for which the range of GABA concentrations was extended to 30 µM. Best-fit parameters ± uncertainty were EC50 = 7 ± 0.9 nM, nH = 0.9 ± 0.1. (C) Comparison of responses to saturating GABA (G; 300 nM) and saturating PEB (0.5 mM). Both traces are from the same oocyte. (D) Comparison of responses to 0.5 mM PEB in the absence and presence of 1 µM DS-2. Both traces are from the same oocyte.

The δ subunit is notoriously difficult to express in heterologous expression systems (Meera et al., 2010), and receptors containing just α4 and β subunits are functional (Meera et al., 2011; Karim et al., 2012). The α4β2 receptors may have contributed to, or even underlain, the observed current responses. To confirm the presence of δ subunit in surface receptors, we coapplied 1 µM DS-2, a δ-specific potentiator (Wafford et al., 2009), with 300 nM GABA. In three cells, from among those included in the analysis of the concentration-response relationship, the mean peak response in the presence of DS-2 was 166 ± 36% of control, that is, of the response to GABA alone. We used a saturating, rather than subsaturating, concentration of GABA to reduce variability originating from cell-to-cell differences in the GABA EC50.

Further confirmation of the presence of the δ subunit in receptor complexes became apparent from comparing the concentration-response relationships for oocytes injected with α4, β2, and δ subunits versus α4 and β2 subunits. In oocytes injected with cRNAs for only α4 and β2 subunits, the GABA concentration-response relationship was shifted by 70-fold to higher GABA concentrations (Table 1). We estimate that for α4β2 receptors the EC50 is 0.71 ± 0.32 µM (4 cells).

View this table:
TABLE 1

Activation properties of receptors expressed in Xenopus oocytes

The table gives the EC50 (best-fit parameter ± uncertainty of the fit) and Hill coefficients (nH) of GABA concentration-response curves fitted to averaged data from four to six cells using the Hill equation. IGABA (mean ± S.D.) was determined as the peak response to the highest GABA concentration tested in concentration-response measurements. Maximal Poest was calculated as ratio of IGABA to the response to saturating concentration of pentobarbital in the absence or presence of DS-2, whichever generated the largest response. In the case of α4β2 receptors that are not potentiated by DS-2, alfaxalone replaced DS-2. Maximal Poest provides an estimate of gating efficacy of GABA.

Previous single-channel work has suggested that the maximal open probability of α4β2δ receptors expressed in HEK cells in the presence of GABA is relatively low (0.1–0.2), whereas that in the presence of saturating pentobarbital (PEB) reaches ∼1 (Akk et al., 2004; Bracamontes et al., 2014). To provide an estimate for the relative open probabilities in oocytes, we calculated a parameter Poest, as described in Materials and Methods. To do this, we determined the maximal response we could elicit from a cell using high concentrations of an agonist in the presence of a strong potentiator. We assumed that this response reflected a probability of being open close to 1. By normalizing to this maximal response, it was then possible to compare the efficacy of GABA more accurately among the various constructs studied. However, we note that there are some caveats associated with this procedure, as discussed in Materials and Methods. To determine the Poest of α4β2δ expressed in oocytes, we compared peak responses elicited by 300 nM GABA or 0.5 mM PEB (a saturating concentration; data not shown). In three cells, the relative response to GABA was 53 ± 13% of the response to PEB. Sample traces are shown in Fig. 1C. To confirm the Poest for PEB, we probed the effect of DS-2 and the steroid ALF on the peak response to 0.5 mM PEB. We reasoned that receptors that open with a Po near 1 lack the dynamic range to be potentiated by drugs acting through changes in kinetics (Colquhoun, 1998; Rusch et al., 2004). Coapplication of 1 µM DS-2 with 0.5 mM PEB was without effect on the peak response (102 ± 8% of control, 5 cells; Fig. 1D). Coapplication of 10 µM ALF with PB was also without effect on peak current (99 ± 14%, 5 cells). We infer that the maximal Poest in the presence of PEB is ∼1, and that the maximal Poest of GABA-activated α4β2δ receptors in oocytes is ∼0.5.

Activation of α4β2δ Receptors by GABA in HEK 293 Cells.

Studies examining properties of α4βδ receptors in HEK cells have reported a GABA EC50 in the micromolar range (Bracamontes et al., 2014; Patel et al., 2014). To confirm these prior findings, we examined GABA activation properties of α4β2δ receptors transiently expressed in HEK 293 cells. In four cells, the estimated EC50 was 2.3 ± 0.4 µM and the Hill coefficient 1.3 ± 0.3 (Fig. 2, A and B). The presence of the δ subunit in surface receptors was demonstrated by potentiation of receptors activated by 50 µM GABA by 1 μM DS-2 (160 ± 23% of control, 5 cells).

Fig. 2.
Fig. 2.

Activation of α4β2δ receptors expressed in HEK 293 cells. (A) Representative current responses in the presence of 0.5, 1, 2, 5, 20, and 50 µM GABA. All traces are from the same cell. (B) A GABA concentration-response relationship for receptors formed from free α4, β2, and δ subunits. The data points show mean ± S.E.M. for data from four cells. The data are in units of Poest that was estimated by comparing responses to GABA to a response to saturating PEB in the same cells (see Materials and Methods for more details). The curve was fitted to the Hill equation yielding an EC50 of 2.3 ± 0.4 µM and a Hill coefficient of 1.3 ± 0.3 (best-fit parameter ± uncertainty of the fit). (C) Comparison of responses to saturating GABA (50 µM) and saturating PEB (1 mM). Both traces are from the same cell.

To determine the maximal Poest for GABA, we compared responses to 50 µM GABA and 1 mM PEB (Fig. 2C). The mean ratio of the tail current observed upon the termination of the application of PEB to the peak response in the presence of GABA was 5 ± 3 (7 cells). Coapplication of 100 µM PEB with 50 µM GABA potentiated the peak response by 5 ± 2 (7 cells). These data suggest that the Poest in the presence of saturating GABA is 0.2. This is consistent with our previous single-channel work on α4β2δ receptors in HEK cells (Akk et al., 2004).

We probed whether changes in transfection ratio affect activation properties of α4β2δ receptors in HEK cells. Specifically, we wished to test whether a fivefold increase in cDNA for δ shifts the GABA concentration-response relationship to lower agonist concentrations and enhances Poest in the presence of saturating GABA. In cells transfected in the ratio of 1:1:5 (α4:β2:δ), the estimated EC50 was 3.5 ± 0.3 µM and the Hill coefficient 1.1 ± 0.1 (4 cells). The mean ratio of the tail current upon termination of the application of PEB to the maximal response to GABA was 4 ± 3 (4 cells). We infer that HEK cells transfected with α4β2δ in the ratio of 1:1:1 and 1:1:5 have similar activation properties.

Activation of Concatemeric Receptors by GABA in Xenopus Oocytes.

In the next set of experiments, we determined whether activation properties of free subunit-containing receptors can be replicated in concatemers that impose a specific order and stoichiometry of subunits. To imitate a receptor containing two α4 subunits, two β2 subunits, and one δ subunit, arranged in the counterclockwise order δβ2α4β2α4, we generated δ-β2-α4 and β2-α4 concatemeric constructs (Shu et al., 2012). Oocytes injected with cRNAs for δ-β2-α4 and β2-α4 constructs demonstrated peak currents of 186 ± 91 nA (range: 102-324 nA; 6 cells) in the presence of 100 µM GABA. Responses to 100 µM GABA were potentiated by 1 µM DS-2 to 436 ± 143% of control (6 cells), indicating the presence of δ subunit in concatemeric receptor complexes and a low Po in the presence of saturating GABA. Curve fitting of concentration-response data from four cells yielded an EC50 of 3.5 ± 0.6 µM and a Hill coefficient of 1.3 ± 0.2 (Fig. 3; Table 1).

Fig. 3.
Fig. 3.

GABA concentration-response relationship of receptors expressed in Xenopus oocytes. GABA concentration-response properties for receptors formed from a combination of the concatemeric β2-δ construct and free α4 subunit, a combination of δ-β2-α4 and β2-α4 concatemeric constructs, a combination of the concatemeric β2-α4 construct and free δ subunit, a combination of β2-δ-β2 and β2-α4 concatemers, or a combination of β2-δ-α4 and β2-α4 concatemers. The data points show mean ± S.E.M. for data from four to six cells. The data are in units of Poest that was estimated by comparing the responses to GABA to a response to saturating PEB ± 1 µM DS-2, whichever generated the largest response (see Materials and Methods for more details). The curves were fitted to the Hill equation. Fitting results are summarized in Table 1. The dashed line replicates data for receptors formed from free α4, β2, and δ subunits (Fig. 1B).

We also tested receptors resulting from combining the β2-α4 construct with a free δ subunit. We presumed that such receptors assemble as two pairs of the tandem construct and a single free δ subunit per receptor, in a stoichiometry similar to the combination of δ-β2-α4 and β2-α4. For β2-α4 + free δ receptors, the GABA EC50 was 4.6 ± 0.2 µM and the Hill coefficient was 1.2 ± 0.1 (5 cells). Thus, GABA activation curves of concatemeric receptors expected to assemble as δβ2α4β2α4 are right-shifted, compared with free α4β2δ receptors, by over 300-fold.

In δ-β2-α4 + β2-α4 receptors, 1 µM DS-2 potentiated the maximal response to saturating PEB (2 mM) to 177 ± 45% of control (5 cells). From this, we calculate a Poest of 0.6 (i.e., 1.77−1) for δ-β2-α4 + β2-α4 receptors in the presence of PEB. The mean peak response to 100 µM GABA was 10 ± 4% (14 cells) of the mean response to 2 mM PEB. Accordingly, the Poest of δ-β2-α4 + β2-α4 receptors activated by saturating GABA is 0.06. Employing a similar approach, we determined that the maximal Poest of β2-α4 + δ receptors is 0.45 in the presence of PEB, and 0.08 in the presence of GABA. The Poest values in the presence of both PEB and GABA are lower for linked δβ2α4β2α4 than free α4β2δ receptors.

Concatemers in β3δβ3β3α1 and β3δα1β3α1 configurations produce functional receptors (Kaur et al., 2009). We created concatemeric constructs β2-δ-β2 and β2-δ-α4 that when combined with the β2-α4 tandem construct were expected to assemble in the analogous β2δβ2β2α4 and β2δα4β2α4 configurations, respectively. Both combinations produced receptors that were activated by GABA, although the peak responses were smaller than with other concatemers or with free subunits. The maximal peak currents in the presence of saturating GABA were 8 ± 3 nA (6 cells) and 66 ± 73 nA (6 cells) in β2-δ-β2 + β2-α4 and β2-δ-α4 + β2-α4, respectively. The GABA EC50 was 1.2 ± 0.7 µM and the Hill coefficient 0.9 ± 0.4 (from 6 cells) for β2-δ-β2 + β2-α4. Receptors consisting of β2-δ-α4 + β2-α4 had an EC50 of 55 ± 17 µM and a Hill coefficient of 1.3 ± 0.4 (6 cells) in the presence of GABA. The concentration-response curves for the α4β2δ and concatemeric receptors are shown in Fig. 3, and the data are summarized in Table 1.

Coapplication of 1 µM DS-2 with saturating PEB (5 mM) enhanced the maximal response to 174 ± 42% (4 cells) in β2-δ-β2 + β2-α4. Responses elicited by 5 mM PEB from β2-δ-α4 + β2-α4 were potentiated to 123 ± 18% of control (5 cells) by DS-2. In both cases, we focused on the rebound current rather than the small initial peak response due to the latter being inhibited by channel block at such a high PEB concentration. The Poest in the presence of saturating PEB was 0.6 for β2-δ-β2 + β2-α4 and 0.8 for β2-δ-α4 + β2-α4 receptors. By comparing the peak responses to saturating GABA and PEB, we calculated the Poest for GABA as 0.08 and 0.13 in β2-δ-β2 + β2-α4 and β2-δ-α4 + β2-α4, respectively. We infer that concatemeric receptors predicted to assemble in β2δβ2β2α4 and β2δα4β2α4 configurations do not functionally resemble receptors containing free α4β2δ subunits expressed in oocytes.

Lastly, we tested the combination of a tandem construct β2-δ and a free α4 subunit. An application of 3 µM GABA elicited responses with a mean peak current of 1131 ± 221 nA (6 cells). Curve fitting of current responses obtained at 10 nM to 10 µM GABA gave an EC50 of 0.17 ± 0.02 µM and a Hill coefficient of 1.0 ± 0.1. The estimate for gating efficacy was obtained by first determining the effect of DS-2 on receptors activated by saturating PEB (2 mM), and then by comparing peak responses to saturating GABA and PEB + DS-2. Coapplication of 1 µM DS-2 enhanced the response to 2 mM PEB to 157 ± 48% of control (4 cells). The peak response to 3 µM GABA was 46 ± 6% of the response to PEB + DS-2 (6 cells). The maximal Poest of β2-δ + free α4 combination is calculated as 0.64 or 0.46 in the presence of saturating PEB or GABA, respectively. With regard to apparent affinity and gating efficacy in the presence of GABA, the combination of the β2-δ tandem and free α4 subunit produced a receptor that most closely resembled the receptor consisting of free α4β2δ subunits in oocytes.

Right-shifted concentration-response curves have been observed for concatemeric GABAA receptors previously (Baumann et al., 2002; Akk et al., 2009; Kaur et al., 2009). To determine whether the presence of the linker sequence could account for the difference in GABA EC50 observed for free α4β2δ (10 nM) and β2-δ + free α4 (170 nM), we created a free (nonlinked) β2 subunit that had the linker sequence attached to its carboxy terminus. Expression of this subunit with wild-type α4 and δ subunits produced receptors with a GABA EC50 at 518 ± 74 nM (4 cells). We infer that the mere presence of a linker sequence on the β2 subunit can affect properties of activation by GABA.

Activation of α4β2δ Receptors by Neuroactive Steroids in Xenopus Oocytes.

Neuroactive steroids, although best known as potentiators of GABA-activated receptors, can also directly activate the GABAA receptor through allosteric binding site(s). Steroids are considered low-efficacy agonists of the GABAA receptor (Shu et al., 2004; Hosie et al., 2006).

To our surprise, several steroids elicited large currents from α4β2δ receptors in oocytes. Receptors exposed to 10 µM ALF generated a peak response that was 51 ± 18% (27 cells) of the response to saturating PEB. Activation by 10 µM 3α-hydroxy-5α-pregnan-20-one (3α5αP) or 3α-hydroxy-5β-pregnan-20-one (3α5βP) also resulted in prominent responses (42 ± 18%, 8 cells, and 45 ± 19%, 4 cells, respectively). Ten micromolar is a saturating concentration for these steroids in terms of potentiation and activation (Shu et al., 2004), and is near the limit of solubility in aqueous solutions (Brewster et al., 1989). Testosterone, 17β-estradiol, dehydroepiandrosterone (DHEA), pregnenolone sulfate, and DHEA sulfate did not elicit current responses, although in some cases (e.g., in the presence of 10 µM DHEA sulfate) rebound currents following termination of steroid application were observed. Initial outward currents in the presence of DHEA sulfate and pregnenolone sulfate suggest the presence of spontaneous channel activity that was blocked by the sulfated steroids. Contributions from spontaneously active receptors were not taken into consideration in data analysis. Sample traces are shown in Fig. 4A.

Fig. 4.
Fig. 4.

Activation by steroids of α4β2δ receptors expressed in Xenopus oocytes. Sample traces from oocytes expressing free α4, β2, and δ subunits. (A) The oocytes were exposed to 10 µM ALF, 3α-hydroxy-5α-pregnan-20-one (3α5αP), 3α-hydroxy-5β-pregnan-20-one (3α5βP), DHEA, DHEA sulfate (DHEAS), pregnenolone sulfate (PS), 17β-estradiol (17β-E), or testosterone (T). For comparison, a response to 0.5 mM PEB in the same oocyte is shown next to the steroid response. The initial outward currents during application of DHEAS and PS are most likely due to block of spontaneous activity. (B) Representative traces from an oocyte expressing free α4, β2, and δ subunits, and exposed to 0.01–10 µM alfaxalone. For comparison, a response to saturating (0.5 mM) pentobarbital from the same oocyte is also shown. (C) Alfaxalone concentration-response relationship for α4β2δ receptors. The peak currents were normalized to response to 0.5 mM PEB in the same cells. The data points give mean ± S.E.M. for data from six cells. The curve was fitted to the Hill equation yielding an EC50 of 0.9 ± 0.8 µM and a Hill coefficient of 0.7 ± 0.2 (best-fit parameter ± uncertainty of the fit). The maximal fitted response was 0.61 ± 0.13.

We examined the concentration-response relationship for ALF by exposing oocytes expressing α4β2δ receptors to 10 nM–10 µM ALF (Fig. 4, B and C). For normalization, each cell was also exposed to 0.5 mM PEB. Curve fitting of combined data from six cells to the Hill equation yielded an EC50 of 0.9 ± 0.8 µM and a maximal fitted relative response of 61 ± 13% (maximal Poest ∼0.6).

We considered the possibility that large steroid responses resulted from a population of receptors other than the intended α4β2δ. First, we tested omission of the δ subunit. Oocytes were injected with cRNAs for the α4 and β2 subunits. Exposure to 2 mM PEB produced large currents with a Poest of nearly 1 (no potentiation by ALF). Exposure to 10 µM ALF alone, however, was essentially without effect (0 ± 5% of the response to PEB, 8 cells). Sample currents are shown in Fig. 5. Homomeric β2 receptors expressed in oocytes are activated by PEB (Cestari et al., 1996). Exposure of oocytes expressing the β2 subunit to 10 µM ALF resulted in a small outward current, likely reflecting closure or blockade of spontaneous activity from β2 GABAA receptors. Pentobarbital elicited robust responses from the same set of cells (Fig. 5). We also tested the combination of α4 and δ subunits. In three oocytes injected with cRNAs for α4 and δ, no currents were observed in response to application of 0.5 mM PEB or 10 µM ALF (data not shown).

Fig. 5.
Fig. 5.

Alfaxalone selectively activates the α4β2δ and β2-δ + α4 receptors in Xenopus oocytes. Representative traces from oocytes expressing free α4, β2, and δ subunits and various combinations of concatemeric and free subunits in the presence of 10 µM ALF. For comparison, a response to a drug mix (PEB in the absence or presence of DS-2) eliciting a Poest of 1 in the same cell is shown next to the steroid response. The concentration of DS-2 was 1 µM. Comparison of the peak responses within each pair provides insight into gating efficacy of alfaxalone. The steroid activated free α4β2δ subunits and the combination of β2-δ + free α4 with high efficacy.

We further tested subunit specificity by replacing the α4 subunit with α1, and, in another set of experiments, by replacing δ with the γ2 subunit. In oocytes expressing α1β2δ receptors, 10 µM ALF elicited currents that were 31 ± 4% (4 cells) of the response to saturating (0.5 mM) PEB. In contrast, 10 µM ALF elicited a response that was 8 ± 2% (3 cells) of the response to saturating (2 mM) PEB in oocytes expressing α4β2γ2L receptors. We infer that steroids selectively activate δ-containing receptors.

To determine whether currents elicited by ALF are mediated by the classic steroid binding site, we probed the effect of the α4(Q246L) mutation. A previous study had shown that mutations to the homologous residue in the α1 subunit (Q241) reduce gating efficacy of neurosteroids on α1β2γ2 GABAA receptors, possibly by eliminating a critical hydrogen bond (Hosie et al., 2006). Our experiments indicate that the α4(Q246L) mutation reduces the gating efficacy of ALF in α4β2δ receptors. The relative peak response to 10 µM ALF was 1 ± 0.4% (5 cells) of the response to saturating (1 mM) PEB in receptors containing the mutation. Incidentally, this also demonstrates the presence of the α4 subunit in receptors producing steroid responses.

Activation of Concatemeric Receptors by the Steroid ALF in Xenopus Oocytes.

We next compared activation of concatemeric receptors by the steroid ALF. Oocytes expressing δ-β2-α4 + β2-α4 constructs showed small responses to ALF. The mean response to 10 µM ALF was only 0.4 ± 0.2% (6 cells) of the response to 2 mM PEB. The combination of β2-α4 tandem + a free δ subunit showed similar lack of sensitivity to ALF, having a relative ALF response of 1 ± 1% (4 cells). The β2-δ-β2 + β2-α4 receptors produced responses to 10 µM ALF that were 4 ± 1% (6 cells) of the response to saturating PEB. The relative steroid response from the β2-δ-α4 + β2-α4 combination was 1 ± 1% (6 cells). Sample traces are shown in Fig. 5.

Receptors generated from combining β2-δ tandems + free α4 subunits produced large currents in the presence of steroid (Fig. 5). The mean peak current in the presence of 10 µM ALF was 32 ± 22% (7 cells) of the response to 2 mM PEB + 1 µM DS-2, determined to elicit currents with Poest near 1. Thus, direct gating of free α4β2δ by ALF is best replicated by the combination of β2-δ + free α4.

Activation of α4β2δ Receptors by ALF in HEK Cells.

In HEK cells, direct activation by ALF was examined on two forms of the α4β2δ receptor. First, we studied free α4β2δ subunits. This combination gave rise to small currents in the presence of 10 µM ALF. The mean peak response was only 0.6 ± 0.9% (6 cells) of the response to saturating PEB.

We also tested receptors resulting from the combination of the β2-δ tandem and a free α4 subunit. This combination produced relatively large steroid currents. The mean peak response to 10 µM ALF was 23 ± 5% (5 cells) of the response to 2 mM PEB. We infer that the combination of β2-δ + free α4 has similar properties when expressed in oocytes and HEK cells.

Configuration of the α4β2δ Receptor.

A previous study of concatemeric GABAA receptors proposed that the δ subunit may substitute for α at the negative side of the GABA-binding interface (Kaur et al., 2009; Karim et al., 2012). To explore the configuration of the α4β2δ receptor, we mutated a conserved phenylalanine residue in loop D to cysteine. This mutation in the α1 subunit (α1F64C) has been shown to effectively abolish activation by GABA without affecting gating by PEB in α1β1γ2 receptors (Petrini et al., 2011).

HEK cells expressing α4β2δ receptors containing the α4(F63C) mutation demonstrated essentially no currents in the presence of GABA up to 10 mM. The ratio of the current amplitude in the presence of 10 mM GABA to that in the presence of 3 mM PEB was ∼0.5%. In contrast, the α4(F63C)β2δ receptors expressed in oocytes demonstrated robust currents with a midpoint of the GABA concentration-response curve at 198 ± 69 µM (5 cells). The maximal Poest in the presence of GABA (0.29 ± 0.15) was slightly reduced compared with that of the wild-type α4β2δ receptor.

To explore a potential contribution of the δ subunit to the “-” interface of the GABA binding site, we made the F-to-C mutation at the homologous site in the δ subunit. Expression of the δ(F74C) subunit with wild-type α4 and β2 in HEK cells produced receptors that showed increased sensitivity to GABA. The EC50 of the GABA concentration-response curve was 1.1 ± 0.2 µM, and the Poest in the presence of saturating GABA was 0.35 (5 cells). In contrast, α4β2δ(F74C) receptors expressed in Xenopus oocytes had a right-shifted GABA concentration-response curve, with the midpoint at 0.32 ± 0.04 µM (6 cells). The maximal Poest in the presence of GABA was not affected by the mutation (0.72 ± 0.16).

In sum, the F-to-C mutation in the α4 subunit eliminates activation by GABA in HEK cells while showing a drastic rightward shift in the GABA activation curve in oocytes. When the analogous F-to-C mutation is made in the δ subunit, it has a gain-of-function effect in HEK cells but a loss-of-function effect in oocytes.

Properties of Overexpressed α4βδ Receptors in Hippocampal Neurons.

The data above demonstrate that gating of α4β2δ receptors by ALF depends on the expression system. When free α4β2δ receptors are expressed in HEK cells, steroid-activated receptors have a maximal Poest of less than 0.01. In oocytes, saturating ALF elicits a response with a Poest of nearly 0.6. To gain initial insight into which type of behavior more closely resembles properties of α4βδ receptors in neurons, we transiently expressed α4 and δ(T269Y) subunits in cultured hippocampal neurons. Electrophysiological experiments were conducted in the presence of 100 µM picrotoxin to block GABAA receptors not containing δ(T269Y), including native receptors.

In four cells, the average peak response to 200 nM GABA (a concentration producing a half-maximal response when this receptor is expressed in oocytes) was 14 ± 10 pA. Coapplication of 1 µM DS-2 enhanced the peak response to 254 ± 102 pA, indicating the presence of the δ subunit in receptors generating the responses. Application of 10 µM ALF in the absence of GABA produced a mean peak response of 148 ± 69 pA, about 60% of the response to GABA + DS-2.

To verify that none of the responses resulted from receptors lacking the δ(T269Y) subunit, we conducted an identical set of experiments on untransfected neurons (in the presence of 100 µM picrotoxin). In nine cells, the average peak response to GABA was 4 ± 4 pA. Coapplication of DS-2 had no effect (4 ± 3 pA). The mean response to ALF was 1 ± 4 pA. Representative traces are shown in Fig. 6. We infer that picrotoxin effectively blocks receptors lacking the δ(T269Y) subunit, and that receptors containing the mutated δ subunit are efficaciously activated by ALF in hippocampal neurons.

Fig. 6.
Fig. 6.

Hippocampal neurons transfected with α4 and δ(T269Y) subunits demonstrate prominent responses to the steroid alfaxalone. (A) Representative traces in the presence of 200 nM GABA, 200 nM GABA + 1 µM DS-2, or 10 µM alfaxalone from a hippocampal neuron transfected with α4 and δ(T269Y) subunits. The currents were recorded in the presence of 100 µM picrotoxin to block GABAA receptors not containing the mutated δ subunit. (B) Representative traces in the presence of 200 nM GABA, 200 nM GABA + 1 µM DS-2, or 10 µM alfaxalone from an untransfected hippocampal neuron. The recordings were made in the presence of 100 µM picrotoxin. Lack of current responses indicates that 100 µM picrotoxin is sufficient to block GABAA receptors not containing the δ(T269Y) subunit.

Discussion

We have examined activation properties of GABAA receptors containing α4, β2, and δ subunits. Receptors were expressed as free subunits or concatemeric constructs in which two or three individual subunits were joined with short linkers to constrain subunit order and stoichiometry. We were motivated by earlier reports of ambiguity in the incorporation of the δ subunit in heterologously expressed receptors (Kaur et al., 2009; Meera et al., 2011) and a lack of clarity in basic activation properties of recombinant α4βδ receptors (Meera et al., 2011; Karim et al., 2012).

Our findings demonstrate that the properties of α4β2δ receptors are dependent on the expression system. In the Xenopus oocyte expression system, receptors assembled from free α4β2δ subunits demonstrated high, nanomolar sensitivity to GABA and high-efficacy activation by the steroid analog ALF. In contrast, in HEK cells, receptors assembled from free α4β2δ subunits exhibited low, micromolar sensitivity to GABA and low-efficacy activation by ALF.

It is unlikely that post-translational modifications underlie differences in receptor properties in oocytes and HEK cells, because high-efficacy activation by ALF was observed in both oocytes and HEK cells expressing the combination of β2-δ tandem + free α4. Furthermore, it was possible to express receptors in oocytes with properties similar to those assembled from free subunits in HEK cells by using several different combinations of concatemers (e.g., δ-β2-α4 + β2-α4). One simple explanation is that α4β2δ receptors assemble in different configurations with distinct activation properties in oocytes and HEK cells. However, both expression systems permit assembly into the alternative configuration if necessary constraints, for example, subunit linkage in concatemers, are applied. Data from hippocampal neurons transfected with α4 and a picrotoxin-resistant δ subunit suggest that neurons favor the assembly of, at least recombinant, α4βδ receptors in the form characterized by large steroid currents.

We used a human α4 and rat β2 and δ subunits in this study. Species differences were unlikely to mediate the differences in activation properties in oocytes and HEK cells. Nanomolar sensitivity to GABA has been observed previously in all human α4β1δ and α4β3δ receptors expressed in oocytes, whereas micromolar sensitivity to GABA has been observed in all rodent α4β2δ receptors expressed in HEK cells (Karim et al., 2012; Patel et al., 2014).

Our data are consistent with free α4β2δ subunits in HEK cells assembling in the subunit order δβ2α4β2α4 (counterclockwise when viewed from the extracellular side). The properties of α4β2δ (GABA EC50 in the micromolar range and low-efficacy activation by ALF) are observed with concatemers predicted to assemble as δβ2α4β2α4 (e.g., δ-β2-α4 + β2-α4) in oocytes. Furthermore, introduction of the F63C mutation to the α4 subunit effectively eliminates activation by GABA for receptors in HEK cells, in agreement with previous data showing that the presence of the mutation in both GABA binding sites renders the receptor inactive in the presence of GABA (Petrini et al., 2011). The δβ3α4β3α4 order of α4β3δ receptors was proposed previously in a study employing atomic force microscopy imaging complexes between epitope-tagged receptors and antibodies in HEK cells (Barrera et al., 2008).

In oocytes, the combination of the β2-δ tandem with free α4 subunit effectively captured the properties (high sensitivity to GABA and high efficacy of ALF) of free α4β2δ subunits. Our results do not directly address the subunit order of receptors formed by free subunits or by the combination of β2-δ and free α4. Some evidence, however, suggests that in such receptors the δ subunit may contribute to the negative interface of the transmitter binding site. The F-to-C mutations in loop D (F63C in α4 and F74C in δ) differently affected receptor activation in HEK cells and oocytes. In HEK cells, the α4(F63C) mutation abolished activation by GABA, whereas δ(F74C) resulted in receptors with increased sensitivity to GABA. In contrast, in oocytes, both mutations resulted in reduced sensitivity to GABA. We speculate that α4β2δ receptors in oocytes contain one β2-α4 and one β2-δ interface that are involved in binding GABA.

Involvement and contribution of the δ subunit to transmitter binding have been proposed previously. In concatemeric receptors containing α1β3δ subunits, introduction of the β3(Y205S) mutation shifted the GABA activation curves to higher concentrations regardless whether the β subunit was paired to an α or a δ subunit (Kaur et al., 2009). In the β3 subunit, a conserved arginine residue (R207) lines the transmitter-binding pocket and has been shown to affect receptor affinity to GABA (Wagner et al., 2004). Mutation of an arginine in the homologous position in the δ subunit (R218A) strongly reduced sensitivity of α4β3δ receptors to GABA (Karim et al., 2012). Both these examples, as well as our data on δ(F74C), suggest that the δ subunit may be involved in binding of GABA in some variants of the α4βδ receptor.

Large ALF-gated, DS-2–sensitive currents were observed in hippocampal neurons overexpressing α4 and δ subunits. The δ subunit contained the T269Y mutation, which rendered receptors containing the exogenous δ subunit resistant to picrotoxin. Prominent ALF-gated currents, in the presence of picrotoxin, from these receptors indicate that the behavior of recombinant α4/δ receptors in neurons may more closely resemble that of α4β2δ receptors in oocytes. It remains to be determined whether native α4/δ receptors show high level of activation in the presence of steroids. If so, physiologic levels of many neuroactive steroids (or combinations of neuroactive steroids) can be expected to activate native α4βδ receptors. Indeed, a recent study found GABA-independent tonic currents in dentate granule cells (Wlodarczyk et al., 2013) that, we speculate, may be elicited by endogenous neuroactive steroids.

A prior study showed that overexpression of the δ subunit increases incorporation of the δ subunit into surface α4β2δ receptors in HEK cells (Wagoner and Czajkowski, 2010). The increase in δ was accompanied by a reduction in β2, whose average number per receptor dropped to ∼1; however, the functional properties of receptors with different stoichiometries were not studied. Our data indicate that the GABA concentration-response relationship in HEK cells is not affected by a fivefold increase in the amount of cDNA for δ subunit used in transfections. The data from oocytes (present work; Kaur et al., 2009; Karim et al., 2012) suggest that the δ subunit can replace an α4 at the GABA binding site. Further work will be required to determine how many δ subunits can be incorporated in a functional receptor and at which positions in the pentamer.

Previous studies of other members of the pentameric ligand ion channel family have demonstrated flexibility in subunit assembly. Neuronal nicotinic α3 and β4 subunits assemble in different subunit stoichiometries, with distinct conductances and kinetic properties, in Xenopus oocytes and sympathetic neurons (Sivilotti et al., 1997). Receptors containing different stoichiometries of the nicotinic α4 and β2 subunits show different pharmacological properties, as a result of intersubunit binding sites in addition to the canonical acetylcholine-binding sites (Harpsoe et al., 2011; Mazzaferro et al., 2011). We have extended these studies to a new subtype of receptor and reconcile previous disparities in reported properties of tonic GABAA receptors containing δ subunits.

Authorship Contributions

Participated in research design: Eaton, Mennerick, Steinbach, Akk.

Conducted experiments: Eaton, Bracamontes, Shu, Li.

Contributed new reagents or analytic tools: Bracamontes.

Performed data analysis: Eaton, Bracamontes, Li, Mennerick, Akk.

Wrote or contributed to the writing of the manuscript: Eaton, Mennerick, Steinbach, Akk.

Footnotes

    • Received July 16, 2014.
    • Accepted September 19, 2014.

  • This work was supported by the National Institutes of Health National Institute of General Medical Sciences [Grants P01-GM047969 and R01-GM108580] and National Institute of Mental Health [Grants R21-MH099658 and R01-MH078823]. J.H.S. is the Russell and Mary Shelden Professor of Anesthesiology.

  • dx.doi.org/10.1124/mol.114.094813.

Abbreviations

ALF
alfaxalone
DHEA
dehydroepiandrosterone
DS-2
4-chloro-N-[2-(2-thienyl)imidazo[1,2-a]pyridine-3-yl]benzamide
GABA
γ-aminobutyric acid
PEB
pentobarbital

References

  1. ↵
    1. Akaike N,
    2. Maruyama T, and
    3. Tokutomi N
    (1987) Kinetic properties of the pentobarbitone-gated chloride current in frog sensory neurones. J Physiol 394:85–98.
    OpenUrlPubMed
  2. ↵
    1. Akk G
    (2002) Contributions of the non-α subunit residues (loop D) to agonist binding and channel gating in the muscle nicotinic acetylcholine receptor. J Physiol 544:695–705.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Akk G,
    2. Bracamontes J, and
    3. Steinbach JH
    (2004) Activation of GABAA receptors containing the α4 subunit by GABA and pentobarbital. J Physiol 556:387–399.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Akk G,
    2. Li P,
    3. Bracamontes J, and
    4. Steinbach JH
    (2009) Activation and modulation of concatemeric GABA-A receptors expressed in human embryonic kidney cells. Mol Pharmacol 75:1400–1411.
    OpenUrlAbstract/FREE Full Text
  5. ↵
    1. Barrera NP,
    2. Betts J,
    3. You H,
    4. Henderson RM,
    5. Martin IL,
    6. Dunn SM, and
    7. Edwardson JM
    (2008) Atomic force microscopy reveals the stoichiometry and subunit arrangement of the α4β3δ GABAA receptor. Mol Pharmacol 73:960–967.
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Baumann SW,
    2. Baur R, and
    3. Sigel E
    (2002) Forced subunit assembly in α1β2γ2 GABAA –receptors: insight into the absolute arrangement. J Biol Chem 277:46020–46025.
    OpenUrlAbstract/FREE Full Text
  7. ↵
    1. Belelli D,
    2. Harrison NL,
    3. Maguire J,
    4. Macdonald RL,
    5. Walker MC, and
    6. Cope DW
    (2009) Extrasynaptic GABAA receptors: form, pharmacology, and function. J Neurosci 29:12757–12763.
    OpenUrlAbstract/FREE Full Text
  8. ↵
    1. Bracamontes JR,
    2. Li P,
    3. Akk G, and
    4. Steinbach JH
    (2014) Mutations in the main cytoplasmic loop of the GABA(A) receptor α4 and δ subunits have opposite effects on surface expression. Mol Pharmacol 86:20–27.
    OpenUrlAbstract/FREE Full Text
  9. ↵
    1. Brewster ME,
    2. Estes KS, and
    3. Bodor N
    (1989) Development of a non-surfactant formulation for alfaxalone through the use of chemically-modified cyclodextrins. J Parenter Sci Technol 43:262–265.
    OpenUrlPubMed
  10. ↵
    1. Brown N,
    2. Kerby J,
    3. Bonnert TP,
    4. Whiting PJ, and
    5. Wafford KA
    (2002) Pharmacological characterization of a novel cell line expressing human α(4)β(3)δ GABAA receptors. Br J Pharmacol 136:965–974.
    OpenUrlCrossRefPubMed
  11. ↵
    1. Cestari IN,
    2. Uchida I,
    3. Li L,
    4. Burt D, and
    5. Yang J
    (1996) The agonistic action of pentobarbital on GABAA β-subunit homomeric receptors. Neuroreport 7:943–947.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Colquhoun D
    (1998) Binding, gating, affinity and efficacy: the interpretation of structure-activity relationships for agonists and of the effects of mutating receptors. Br J Pharmacol 125:924–947.
    OpenUrlCrossRefPubMed
  13. ↵
    1. Farrant M and
    2. Nusser Z
    (2005) Variations on an inhibitory theme: phasic and tonic activation of GABAA receptors. Nat Rev Neurosci 6:215–229.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Gurley D,
    2. Amin J,
    3. Ross PC,
    4. Weiss DS, and
    5. White G
    (1995) Point mutations in the M2 region of the α, β, or γ subunit of the GABAA channel that abolish block by picrotoxin. Receptors Channels 3:13–20.
    OpenUrlPubMed
  15. ↵
    1. Haas KF and
    2. Macdonald RL
    (1999) GABAA receptor subunit γ2 and δ subtypes confer unique kinetic properties on recombinant GABAA receptor currents in mouse fibroblasts. J Physiol 514:27–45.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Harpsøe K,
    2. Ahring PK,
    3. Christensen JK,
    4. Jensen ML,
    5. Peters D, and
    6. Balle T
    (2011) Unraveling the high- and low-sensitivity agonist responses of nicotinic acetylcholine receptors. J Neurosci 31:10759–10766.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Hoestgaard-Jensen K,
    2. Dalby NO,
    3. Wolinsky TD,
    4. Murphey C,
    5. Jones KA,
    6. Rottländer M,
    7. Frederiksen K,
    8. Watson WP,
    9. Jensen K, and
    10. Ebert B
    (2010) Pharmacological characterization of a novel positive modulator at α 4 β 3 δ-containing extrasynaptic GABAA receptors. Neuropharmacology 58:702–711.
    OpenUrlCrossRefPubMed
  18. ↵
    1. Hosie AM,
    2. Wilkins ME,
    3. da Silva HM, and
    4. Smart TG
    (2006) Endogenous neurosteroids regulate GABAA receptors through two discrete transmembrane sites. Nature 444:486–489.
    OpenUrlCrossRefPubMed
  19. ↵
    1. Karim N,
    2. Wellendorph P,
    3. Absalom N,
    4. Bang LH,
    5. Jensen ML,
    6. Hansen MM,
    7. Lee HJ,
    8. Johnston GA,
    9. Hanrahan JR, and
    10. Chebib M
    (2012) Low nanomolar GABA effects at extrasynaptic α4β1/β3δ GABAA receptor subtypes indicate a different binding mode for GABA at these receptors. Biochem Pharmacol 84:549–557.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Kaur KH,
    2. Baur R, and
    3. Sigel E
    (2009) Unanticipated structural and functional properties of δ-subunit-containing GABAA receptors. J Biol Chem 284:7889–7896.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    1. Mazzaferro S,
    2. Benallegue N,
    3. Carbone A,
    4. Gasparri F,
    5. Vijayan R,
    6. Biggin PC,
    7. Moroni M, and
    8. Bermudez I
    (2011) Additional acetylcholine (ACh) binding site at α4/α4 interface of (α4β2)2α4 nicotinic receptor influences agonist sensitivity. J Biol Chem 286:31043–31054.
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Meera P,
    2. Olsen RW,
    3. Otis TS, and
    4. Wallner M
    (2010) Alcohol- and alcohol antagonist-sensitive human GABAA receptors: tracking δ subunit incorporation into functional receptors. Mol Pharmacol 78:918–924.
    OpenUrlAbstract/FREE Full Text
  23. ↵
    1. Meera P,
    2. Wallner M, and
    3. Otis TS
    (2011) Molecular basis for the high THIP/gaboxadol sensitivity of extrasynaptic GABAA receptors. J Neurophysiol 106:2057–2064.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Mennerick S,
    2. Que J,
    3. Benz A, and
    4. Zorumski CF
    (1995) Passive and synaptic properties of hippocampal neurons grown in microcultures and in mass cultures. J Neurophysiol 73:320–332.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Patel B,
    2. Mortensen M, and
    3. Smart TG
    (2014) Stoichiometry of δ subunit containing GABAA receptors. Br J Pharmacol 171:985–994.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Petrini EM,
    2. Nieus T,
    3. Ravasenga T,
    4. Succol F,
    5. Guazzi S,
    6. Benfenati F, and
    7. Barberis A
    (2011) Influence of GABAAR monoliganded states on GABAergic responses. J Neurosci 31:1752–1761.
    OpenUrlAbstract/FREE Full Text
  27. ↵
    1. Rüsch D,
    2. Zhong H, and
    3. Forman SA
    (2004) Gating allosterism at a single class of etomidate sites on α1β2γ2L GABAA receptors accounts for both direct activation and agonist modulation. J Biol Chem 279:20982–20992.
    OpenUrlAbstract/FREE Full Text
  28. ↵
    1. Shu HJ,
    2. Bracamontes J,
    3. Taylor A,
    4. Wu K,
    5. Eaton MM,
    6. Akk G,
    7. Manion B,
    8. Evers AS,
    9. Krishnan K,
    10. Covey DF,
    11. et al.
    (2012) Characteristics of concatemeric GABAA receptors containing α4/δ subunits expressed in Xenopus oocytes. Br J Pharmacol 165:2228–2243.
    OpenUrlCrossRefPubMed
  29. ↵
    1. Shu HJ,
    2. Eisenman LN,
    3. Jinadasa D,
    4. Covey DF,
    5. Zorumski CF, and
    6. Mennerick S
    (2004) Slow actions of neuroactive steroids at GABAA receptors. J Neurosci 24:6667–6675.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Sivilotti LG,
    2. McNeil DK,
    3. Lewis TM,
    4. Nassar MA,
    5. Schoepfer R, and
    6. Colquhoun D
    (1997) Recombinant nicotinic receptors, expressed in Xenopus oocytes, do not resemble native rat sympathetic ganglion receptors in single-channel behaviour. J Physiol 500:123–138.
    OpenUrlCrossRefPubMed
  31. ↵
    1. Stórustovu SI and
    2. Ebert B
    (2006) Pharmacological characterization of agonists at δ-containing GABAA receptors: functional selectivity for extrasynaptic receptors is dependent on the absence of γ2. J Pharmacol Exp Ther 316:1351–1359.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Ueno S,
    2. Zorumski C,
    3. Bracamontes J, and
    4. Steinbach JH
    (1996) Endogenous subunits can cause ambiguities in the pharmacology of exogenous γ-aminobutyric acidA receptors expressed in human embryonic kidney 293 cells. Mol Pharmacol 50:931–938.
    OpenUrlAbstract
  33. ↵
    1. Wafford KA,
    2. van Niel MB,
    3. Ma QP,
    4. Horridge E,
    5. Herd MB,
    6. Peden DR,
    7. Belelli D, and
    8. Lambert JJ
    (2009) Novel compounds selectively enhance δ subunit containing GABAA receptors and increase tonic currents in thalamus. Neuropharmacology 56:182–189.
    OpenUrlCrossRefPubMed
  34. ↵
    1. Wagner DA,
    2. Czajkowski C, and
    3. Jones MV
    (2004) An arginine involved in GABA binding and unbinding but not gating of the GABAA receptor. J Neurosci 24:2733–2741.
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Wagoner KR and
    2. Czajkowski C
    (2010) Stoichiometry of expressed α(4)β(2)δ γ-aminobutyric acid type A receptors depends on the ratio of subunit cDNA transfected. J Biol Chem 285:14187–14194.
    OpenUrlAbstract/FREE Full Text
  36. ↵
    1. Wlodarczyk AI,
    2. Sylantyev S,
    3. Herd MB,
    4. Kersanté F,
    5. Lambert JJ,
    6. Rusakov DA,
    7. Linthorst AC,
    8. Semyanov A,
    9. Belelli D,
    10. Pavlov I,
    11. et al.
    (2013) GABA-independent GABAA receptor openings maintain tonic currents. J Neurosci 33:3905–3914.
    OpenUrlAbstract/FREE Full Text
  37. ↵
    1. You H and
    2. Dunn SM
    (2007) Identification of a domain in the δ subunit (S238-V264) of the α4β3δ GABAA receptor that confers high agonist sensitivity. J Neurochem 103:1092–1101.
    OpenUrlCrossRefPubMed
Back to top

In this issue

Download PDF
Article Alerts
Email Article
Citation Tools

Research ArticleArticle

Distinct Types of the α4β2δ GABAA Receptor

Megan M. Eaton, John Bracamontes, Hong-Jin Shu, Ping Li, Steven Mennerick, Joe Henry Steinbach and Gustav Akk
Molecular Pharmacology December 1, 2014, 86 (6) 647-656; DOI: https://doi.org/10.1124/mol.114.094813
Reddit logo Twitter logo Facebook logo Mendeley logo

Jump to section

Advertisement