• chimera;
  • oocyte expression;
  • γ2 subunit;
  • γ-aminobutyric acid type A;
  • δ subunit


  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We have expressed the α4β3δ and α4β3γ2L subtypes of the rat GABAA receptor in Xenopus oocytes and have investigated their agonist activation properties. GABA was a more potent agonist of the α4β3δ receptor (EC50 ≈ 1.4 μmol/L) than of the α4β3γ2L subtype (EC50 ≈ 27.6 μmol/L). Other GABAA receptor agonists (muscimol, 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol, imidazole-4-amino acid) displayed similar subtype selectivity. The structural determinants underlying these differences have been investigated by co-expressing chimeric δ/γ2L subunits with α4 and β3 subunits. A stretch of amino acids in the δ subunit, S238-V264, is shown to play an important role in determining both agonist potency and the efficacies of full or partial agonists. This segment includes transmembrane domain 1 and the short intracellular loop that leads to the second transmembrane domain. The effects of the competitive antagonists, bicuculline and SR95531, and the channel blocker, picrotoxin, were not significantly affected by the incorporation of chimeric subunits. As the δ and γ2L subunits have not been previously implicated directly in agonist binding, we suggest that the effects are likely to arise from changes in the transduction mechanisms that link agonist binding to channel activation.

Abbreviations used

γ-aminobutyric acid type A


imidazole-4-amino acid


ligand-gated ion channel



TM1 and TM2

transmembrane domains 1 and 2

γ-Aminobutyric acid type A (GABAA) receptors are the major inhibitory neurotransmitter receptors in the mammalian brain. These receptors are members of the cys-loop family of ligand-gated ion channels (LGICs) that includes the nicotinic, serotonin type 3, and glycine receptors (Sieghart et al. 1999). Each member of the LGIC family is likely to be a pentamer in which homologous transmembrane (TM) subunits are arranged in a rosette conformation to form a central ion channel pore (Nayeem et al. 1994). Nineteen mammalian GABAA receptor subunits have been identified i.e. α1-6, β1-3, γ1-3, ρ1-3, δ, ε, θ, and π (McKernan and Whiting 1996; Barnard et al. 1998). The most common subtype of GABAA receptor in the mammalian CNS is the α1β2γ2 combination (reviewed by Sieghart et al. 1999; Whiting 2003), where the likely stoichiometry is 2α : 2β : 1γ (Farrar et al. 1999). However, the inclusion of other subunits within the pentamer results in differences in physiological function and sensitivity to the large number of pharmacological agents that target these receptors (Sieghart 1995).

Recently, it has been suggested that there are two inhibitory neuronal signaling pathways which are mediated by GABAA receptors, namely phasic inhibition and tonic inhibition (Brickley et al. 2001; Mody 2001). Phasic inhibition is evoked by the action potential-dependent release of high concentrations of GABA (0.5–1 mmol/L) at the synapse. These high local concentrations of GABA are transient and induce rapid inhibitory responses. In contrast, tonic inhibition is mediated by the persistent but low concentrations of GABA that either overspill from the synapse or are released from glial cells. The extrasynaptic receptors that are responsible for these functional differences are likely to include the α4, α5, α6, γ2/3 or δ subunits, with the most likely combinations in the brain being α4βxδ, α5β3γ2/3, and α6β2/3δ (see Farrant and Nusser 2005). Sur et al. (1999), using immunoprecipitation techniques, suggested that approximately one-third of the α4-containing GABAA receptor subtypes in the hippocampus and thalamus include the γ2 subunit while two-thirds co-associate with the δ subunit. Another study, using a similar approach and brain membrane extracts, suggested a higher ratio of the α4γ2 containing subtypes compared to α4δ (Bencsits et al. 1999). Overall, the available data suggest that a heterogeneous population of GABAA receptors containing the α4 subunit (mainly α4βxγ2 and α4βxδ subtypes) exists in the same regions of the brain.

Extrasynaptic receptor subtypes appear to have a higher sensitivity to GABA than those that mediate synaptic responses (Saxena and Macdonald 1994). This is not unexpected as these receptors respond to low ambient concentrations of GABA. The major aim of the present study was to identify the molecular basis for the higher potency of GABA at the α4δ-containing receptors. There is considerable evidence to demonstrate that binding sites within the LGIC family lie at subunit-subunit interfaces and agonist binding sites that are involved in GABAA receptor activation have been localized to the β–α subunit interfaces (see Amin and Weiss 1993). However, structural determinants lying outside the putative agonist binding domains have also been shown to affect agonist sensitivity. These include motifs in the α subunit that confer differential sensitivities between α6- or α1-containing receptors (Korpi and Luddens 1993) or between α3 and other α subunits (Bohme et al. 2004). The current structural and functional evidence suggests that, following formation of the receptor–agonist complex, a concerted movement of all subunits may be required to induce channel opening (see Unwin 2005). Thus, all subunits may contribute to agonist affinity and efficacy.

In this study, we explore the role of structural determinants within the δ and/or γ2L subunit that are important for conferring the differential potencies of agonists. We have generated δ/γ2L chimeric subunits and co-expressed these with α4 and β3 subunits in Xenopus oocytes for functional studies. The results demonstrate that a domain within the GABAA receptor δ subunit (S238-V264) confers high agonist sensitivity to the α4β3δ subtype.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References


All drugs were purchased from Sigma-Aldrich (St Louis, MO, USA) and were made as stock solutions (1 mmol/L to 1 mol/L) in sterile water. Picrotoxin suspensions were ultrasonicated to make soluble 1 mmol/L preparations. The stock solutions for GABA, 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP), imidazole-4-amino acid (I4AA), SR95531, and picrotoxin were aliquoted and stored at −80°C until use. Muscimol and bicuculline stocks were made freshly prior to each experiment.

Clones and construction of δ/γ2 chimeras

The cDNAs encoding the rat GABAA subunits were subcloned into the pcDNA3.1(+) expression vector (Invitrogen, San Diego, CA, USA). The original cDNAs encoding the α4 and β3 subunits were from Dr P. H. Seeburg’s laboratory and those encoding the γ2L and δ subunits were generously provided by Drs D. L. Weiss and R. L. Macdonald, respectively. Randomly derived δ/γ2L chimeras were created following the protocol of Moore and Blakely (1994). In brief, the δ and γ2L subunit cDNAs were engineered into multiple cloning sites in the pcDNA3.1(+) vector with the δ subunit being positioned upstream of the γ2L subunit. A BamHI and an EcoRI restriction site were left in the polylinker between the δ and γ2L sequences. The dual plasmid DNA was digested and linearized by BamHI/EcoRI, and was then transformed into Library Efficiency® competent DH5αEscherichia coli cells (Life Technologies, Gaithersburg, MD, USA). During the transformation, random crossover events occurred at regions of homology between the δ and γ2L sequences, creating a series of random chimeric subunit cDNAs. Figure 1 shows the three in-frame hybrid DNA chimeras (χ237, χ255, χ277) that were generated. In each case, the N-terminal domain was derived from the original δ subunit DNA with the remainder coming from γ2L. The chimeras were named by the point of crossover with the number representing the last residue of the δ subunit prior to the in-frame switch. Figure 1 also shows an alignment of the δ and γ2L sequences, the position of the crossover points and the location of the putative TM domains based on the structural model of Ernst et al. (2005).


Figure 1.  Schematic representation of chimeric subunit construction. (a) The wild-type δ subunit cDNA (black) was subcloned upstream of wild-type γ2L subunit cDNA (grey) into pcDNA3.1(+). This dual plasmid was doubly digested, using BamHI/EcoRI to linearize the polylinker between the δ and γ2L sequence and this segment was subsequently transformed into DH5αEscherichia coli cells. During the transformation, the linearized plasmid was recircularized by recombination events in homologous regions between the δ and γ2L sequences. (b) Three chimeric subunits (χ277, χ255, and χ237) were isolated and named according to the last residue of δ sequence before the crossover point. The wild-type δ subunit is shown in black and the γ2L subunit is shown in grey. The diagrams show the approximate positions of the crossover points within the predicted transmembrane domains of the subunit. (c) Partial amino acid sequence alignments of the δ and γ2L subunits including the transmembrane domains, TM1 and TM2, according to Ernst et al. (2005). The positions of the chimeric crossover points (237, 255, and 277) are shown. The shaded areas illustrate stretches of amino acids that are conserved between the δ and γ2 subunits showing that in the χ255 and χ277 chimeras, the switch to the unique γ2 sequence occurs after δ M248 and V264, respectively. The asterisks denote the non-identical residues within the 238–264 domains of the two subunits.

Download figure to PowerPoint

Expression of GABAA receptors in Xenopus oocytes

Capped cRNAs encoding rat GABAA receptor wild-type and chimeric subunits were synthesized from linearized cDNA following standard protocols and using T7 RNA polymerase (Invitrogen) for in vitro transcription. cRNA concentrations were calculated from their absorbance at 260 nm. Stage V–VI Xenopus laevis oocytes were isolated and prepared as described (Smith et al. 2004). Oocytes were injected with 50 nL of 1 μg/μL total subunit cRNAs in a 1 : 1 : 1 ratio (α4 : β3 : δ, α4 : β3 : γ2L or α4 : β3 : δ/γ2L chimeric receptors) or 1 : 1 (α4 : β3). Other ratios of subunit cRNAs were also used as described in the text. Injected oocytes were incubated individually in ND96 buffer (96 mmol/L NaCl, 2 mmol/L KCl, 2 mmol/L CaCl2, 1 mmol/L MgCl2 5 mmol/L HEPES, pH 7.4) in 96-well plates at 14°C for at least 48 h prior to functional analysis.

Two-electrode voltage clamp recordings

Oocytes were continuously bathed in frog Ringer’s solution (110 mmol/L NaCl, 2 mmol/L KCl, 1.8 mmol/L CaCl2, 5 mmol/L HEPES, pH 7.4) by gravity flow (∼5 mL/min) in a custom-made recording chamber. Agonist-induced currents were measured by standard two-electrode voltage clamp techniques using a GeneClamp 500B amplifier (Axon Instruments Inc., Foster City, CA, USA) at a holding potential of −60 mV. The voltage-sensing and current-passing electrodes were filled with 3 mol/L KCl and only electrodes with a resistance between 0.5 and 3.0 MΩ in frog Ringer’s solution were used.

To measure the effects of GABA and other agonists (muscimol, THIP, I4AA), the agonist was applied via gravity perfusion (30–40 s at 5 mL/min) followed by a 10–15 min washout to ensure complete recovery from desensitization. In studies of antagonist effects, oocytes were pre-perfused with these ligands for 2 min prior to initiation of the receptor response by perfusion with GABA (at its EC50 concentration) and the same concentration of the antagonist as used in the pre-perfusion.

Data analysis

Concentration-effect curves for agonist activation were analyzed by non-linear regression techniques using GraphPad Prism 4.0 software (GraphPad Software Inc., San Diego, CA, USA) and the following equation:

  • image

where I is the amplitude of agonist-evoked current for a given concentration [L], Imax is the maximum amplitude of current, EC50 is the agonist concentration that evokes half maximal receptor activation, and nH is the Hill coefficient.

The inhibitory effects of antagonists were analyzed using the equation (GraphPad Prism 4.0):

  • image

where IC50 is the concentration of antagonist, [A], that reduces the amplitude of the GABA-evoked current by 50% and nH is the Hill coefficient.

Data were analyzed by one-way anova and levels of significance were determined by the Dunnett’s post-test for multiple comparisons.


  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Expression of GABAA receptor subtypes

Many investigators have reported difficulties in co-expressing the δ subunit with other GABAA receptor subunits in recombinant systems (see Brown et al. 2002). We have, therefore, looked at the expression of various combinations of subunits i.e. β3 alone, α4β3, α4δ, β3δ, and α4β3δ. When injected into oocytes, only the α4β3 and α4β3δ combinations produced functional receptors as measured by detectable current responses elicited by a saturating (3 mmol/L) concentration of GABA. Using 1 : 1 (α4 : β3) or 1 : 1 : 1 (α4 : β3 : δ) ratios of the encoding cRNAs, these receptor subtypes showed significant differences in both their time courses of expression and expression levels attained. Using the same batches of oocytes, the expression levels of the α4β3δ subtype were robust 48 h after injection but the α4β3 subtype required incubation for 4–5 days before stable responses could be recorded. Six days after injection, the α4β3δ population displayed an average GABA-induced maximum current that was 4.7-fold higher than that of the α4β3 subtype (1419 ± 79 nA, n = 7 vs. 301 ± 76 nA, n = 6, respectively). The EC50 values for GABA activation of the two subtypes were not significantly different but, in agreement with previous results (Storustova and Ebert 2006), the α4β3δ subtype displayed lower sensitivity to inhibition by Zn2+ (IC50 = 5.40 ± 0.72 μmol/L, n = 3) compared to the α4β3 combination (0.18 ± 0.02 μmol/L, n = 3). The two receptor subtypes could also be distinguished by the effects of THIP. Although this agonist displayed similar ‘super-agonism’ at both receptors (see below), its EC50 for activation of the α4β3 receptor (85.1 ± 17.7 μmol/L, n = 3) was ninefold higher than for the α4β3δ subtype (see below).

Effects of varying cRNA ratios on receptor expression

It has been reported that efficient expression of trimeric αβγ receptors in Xenopus oocytes may require an increase in the relative amount of cRNA encoding the γ subunit in the injection mix (Boileau et al. 2002). We have not previously observed anomalous expression of receptors containing the α, β, and γ subunits when a cRNA ratio of 1 : 1 : 1 was used (unpublished results). However, to investigate the expression of the α4β3δ subtype, different cRNA ratios were studied. When the subunit ratios (α4 : β3 : δ) were changed from 1 : 1 : 1 to 1 : 1 : 5 or 1 : 1 : 10, the EC50 values for GABA activation were progressively increased (1.4 ± 0.13, 3.90 ± 0.84, and 10.4 ± 0.4 μmol/L, respectively) while the Hill slope decreased significantly (0.70 ± 0.07, 0.64 ± 0.03, and 0.38 ± 0.09). In control experiments using similar higher ratios of the δ subunit cRNA, we have not been able to force the functional expression of β3δ or α4δ receptors. Possible explanations for these results are provided in the Discussion. As the properties of the α4β3δ and α4β3γ2L receptors were highly consistent when a 1 : 1 : 1 ratio was used, we have used this stoichiometry in all experiments described below.

Effects of GABA and muscimol on functional responses of wild-type and chimeric receptors

Concentration–response curves (Fig. 2a and Table 1) revealed GABA to be approximately 20-fold more potent in mediating activation of the α4β3δ receptor (EC50 ≈ 1.4 μmol/L) compared to the α4β3γ2L subtype (EC50 ≈ 27.6 μmol/L).


Figure 2.  The effects of (a) GABA and (b) muscimol on activation of wild-type and chimeric receptors expressed in Xenopus oocytes. Concentration effect curves for α4β3δ (bsl00001), α4β3γ2L (•), and the chimeric receptors α4β3χ277 (□), α4β3χ237 (○) are shown. The effects of the agonists on the α4β3χ255 subtype are omitted for clarity. Data represent the mean ± SEM from at least three independent experiments and the parameters obtained from curve-fitting are summarized in Table 1.

Download figure to PowerPoint

Table 1.   Activation of the wild-type and chimeric GABAA receptor subtypes by GABA and muscimol
log EC50EC50 (μmol/L)nHlog EC50EC50 (μmol/L)nH
  1. Log EC50 values were obtained from pooled data from at least three independent concentration-effect curves using different batches of oocytes and represent the mean ± SEM. Hill coefficients (nH) from individual curves were averaged to give the final estimates indicated.

  2. ap < 0.01 compared to wild-type α4β3γ2 receptor. bp < 0.01 compared to wild-type α4β3δ recepor. cp < 0.05 compared to wild-type α4β3δ receptor.

α4β3δ−5.85 ± 0.05a1.410.70 ± 0.07−5.97 ± 0.02a1.080.93 ± 0.05
α4β3χ277−5.64 ± 0.10a2.290.69 ± 0.05−5.67 ± 0.06a,b2.150.71 ± 0.07
α4β3χ255−5.04 ± 0.02a,b9.050.71 ± 0.04−5.74 ± 0.07a,c1.820.72 ± 0.10
α4β3χ237−4.67 ± 0.07b21.50.77 ± 0.07−5.27 ± 0.04b5.350.72 ± 0.02
α4β3γ2−4.56 ± 0.11b27.60.91 ± 0.09−5.21 ± 0.06b6.170.93 ± 0.09

Chimeric δ/γ2L receptor subunits were constructed to investigate structural domains in these subunits that contribute to the above differences in agonist potency. These chimeras (χ277, χ255, χ237; see Fig. 1) incorporated the N-terminal sequence of the δ subunit with the remainder of the sequence corresponding to that of γ2L. When co-expressed with the α4 and β3 subunits, clear differences in GABA potency were observed. The EC50 value for GABA activation of the χ277-containing receptor (EC50 ≈ 2.3 μmol/L), where the crossover point lies towards the middle of the TM2 domain, was not significantly different from that of the wild-type α4β3δ receptor. In contrast, the EC50 for GABA activation of the χ237-containing subtype (EC50 ≈ 21.5 μmol/L) was not significantly different from that of the α4β3γ2L subtype. The EC50 value for GABA activation of the χ255-containing subtype (EC50 ≈ 9 μmol/L) was intermediate between the two wild-type receptors, and was significantly different from both (see Table 1).

A similar trend in potency was seen when another ‘full’ GABAA receptor agonist, muscimol, was investigated (Table 1). Incorporation of the χ277 chimeric subunit again conferred α4β3δ-like potency whereas the presence of the χ237 chimera led to characteristics similar to those of the α4β3γ2L receptor. The above results suggest that a domain of the δ subunit lying between residues 237 and 277 plays an important role in determining agonist sensitivity.

Effects of THIP and I4AA on functional responses of wild-type and chimeric receptors

THIP and I4AA are often described as partial agonists of the GABAA receptor as they elicit maximum current responses at the major GABAA receptor subtype, α1β2γ2, that are lower than those produced by ‘full’ agonists, such as GABA and muscimol. The oocyte expression system is not ideal for determining efficacies as their large size limits the rates of agonist perfusion and thus the resolution of rapid responses. However, in agreement with previous studies that used a stably expressing cell line (Brown et al. 2002), we found that THIP acts as ‘super-agonist’ of the α4β3δ receptor, eliciting a maximum current that was 135.6% of that induced by 1 mmol/L GABA (Figs 3a and 4a). THIP, however, is only a partial agonist of the α4β3γ2L combination, producing a maximum current that was 47% of that mediated by GABA (Figs 3c and 4a). A similar pattern was seen for I4AA i.e. it is a partial agonist of the α4β3γ2L subtype (Figs 3d and 4b) but, at the α4β3δ combination, it elicits a maximum response that is not significantly different from GABA (Figs 3b and 4b). In order to investigate the possible involvement of the δ subunit in influencing agonist efficacy as well as potency, we studied the responses of the chimeric receptors to THIP and I4AA.


Figure 3.  Representative currents for activation of different GABAA receptors by varying concentrations of THIP and I4AA as indicated. Control responses of the same oocytes elicited by 1 mmol/L GABA are shown for comparison.

Download figure to PowerPoint


Figure 4.  The effects of (a) THIP and (b) I4AA on activation of α4β3δ (bsl00001), α4β3γ2L (•), and the chimeric receptors α4β3χ277 (□), α4β3χ255 (bsl00066), α4β3χ237 (○). The data represent the mean ± SEM from at least three independent experiments and the results from data fitting are summarized in Table 2. The amplitudes are normalized to the magnitude of the response induced by a maximal concentration of GABA (100%, dashed line in figure).

Download figure to PowerPoint

The effects of THIP and I4AA on wild-type and chimeric receptors are shown in Fig. 4 and the results are summarized in Table 2. Inclusion of the δ/γ2L chimeric subunits had similar effects on the potency of these agonists to the results obtained for GABA i.e. the presence of the χ277 subunit gave rise to a receptor that had similar sensitivity to the α4β3δ subtype, while the chimeric receptor incorporating χ237 displayed similar potency characteristics to the α4β3γ2L receptor. In these experiments, the potencies of THIP and I4AA for the α4β3χ255 receptor were not significantly different from the α4β3δ wild-type (Table 2).

Table 2.   Activation of wild-type and chimeric GABAA receptor subtypes by THIP and I4AA
log EC50EC50 (μmol/L)nHEfficacylog EC50EC50 (μmol/L)nHEfficacy
  1. Log EC50 values were obtained from pooled data from at least three independent concentration-effect curves using different batches of oocytes and represent the mean ± SEM. Hill coefficients (nH) from individual curves were averaged to give the final estimates indicated. The efficacy values reported are the amplitudes of the maximum current induced by the agonist relative to the maximum current induced by GABA (100%) for that receptor subtype.

  2. ap < 0.01 compared to wild-type α4β3γ2 receptor. bp < 0.01 compared to wild-type α4β3δ recepor. cp < 0.05 compared to wild-type α4β3δ receptor.

α4β3δ−5.02 ± 0.08a9.620.92 ± 0.08135.6 ± 4.7a−4.00 ± 0.05a100.20.66 ± 0.06110.3 ± 8.7a
α4β3χ277−4.83 ± 0.17a14.70.86 ± 0.0596.5 ± 2.0a,b−4.22 ± 0.03a60.30.78 ± 0.0858.7 ± 5.2b
α4β3χ255−4.62 ± 0.07a24.20.95 ± 0.0796.0 ± 1.5a,b−3.96 ± 0.11a110.20.74 ± 0.0554.3 ± 4.0b
α4β3χ237−4.02 ± 0.11b94.61.00 ± 0.10110.5 ± 11.0a,c−3.46 ± 0.05b345.90.57 ± 0.0149.6 ± 2.9b
α4β3γ2−3.70 ± 0.10b199.11.30 ± 0.1946.9 ± 5.1b−3.28 ± 0.06b520.00.97 ± 0.1140.0 ± 3.5b

The effects of the chimeric subunits on agonist efficacy were more complex (see Figs 3 and 4). The efficacies reported in Table 2 were calculated from the percentage of the maximum current responses induced by these agonists compared to the maximum effect of GABA. In the case of THIP, inclusion of any one of the chimeric subunits resulted in maximum current amplitudes that were similar to those of GABA. Thus, all receptors displayed significantly higher efficacy when compared to the α4β3γ2L receptor but no receptor combination displayed a similar extent of THIP ‘superagonism’ as the α4β3δ wild-type. I4AA was a partial agonist at all chimeric receptors i.e. the maximum responses induced were not significantly different from those of the α4β3γ2L subtype.

Effects of competitive antagonists and picrotoxin on functional responses of wild-type and chimeric receptors

To investigate whether the above results obtained using the δ/γ2L chimeric subunits may be explained by changes in the overall structure of the binding sites for agonists/competitive antagonists or the properties of the ion channel itself, we have characterized the functional effects of two competitive antagonists (bicuculline and SR95531) and the channel blocker, picrotoxin. In these experiments, the oocytes were pre-perfused with the antagonist for 2 min prior to challenge with a concentration of GABA equal to its EC50 value. The results shown in Fig. 5 and summarized in Table 3 show that the apparent affinities for SR95531 and picrotoxin are not significantly different between the wild-type and chimeric receptors. Similarly, the effects of bicuculline were comparable for all subtypes, except in the case of α4β3χ237 receptor where its IC50 value was approximately threefold higher than that of either the α4β3δ or α4β3γ2L receptor (Table 3).


Figure 5.  The effects of (a) SR95531, (b) bicuculline, and (c) picrotoxin on GABA-evoked currents in the α4β3δ (bsl00001), α4β3γ2L (•), and chimeric α4β3χ277 (□), α4β3χ255 (bsl00066), and α4β3χ237 (○) receptors. In each experiment, the oocyte was pre-perfused with the antagonist for 2 min before challenged with GABA and the same concentration of antagonist as used in the pre-perfusate. The GABA concentration used was equivalent to its EC50 value for each receptor subtype and the data were normalized to the GABA response in the absence of antagonist. Data represent the mean ± SEM from at least three independent experiments. The data obtained from curve-fitting are summarized in Table 3.

Download figure to PowerPoint

Table 3.   Inhibition of GABA-gated currents by the competitive antagonists (SR95531 and bicuculline) and the channel blocker (picrotoxin)
log IC50IC50 (μmol/L)log IC50IC50 (μmol/L)log IC50IC50 (μmol/L)
  1. Log IC50 values were determined from at least three independent experiments and represent mean ± SEM. The concentration of GABA used was its EC50 value for each receptor subtype (see Experimental procedures).

  2. ap < 0.05 compared to wild-type α4β3γ2 receptor. bp < 0.01 compared to wild-type α4β3δ receptor.

α4β3δ−6.05 ± 0.080.90−5.61 ± 0.192.44−5.31 ± 0.254.85
α4β3χ277−6.29 ± 0.180.51−5.68 ± 0.022.10−5.32 ± 0.024.83
α4β3χ255−6.23 ± 0.040.59−5.48 ± 0.023.28−5.55 ± 0.022.82
α4β3χ237−6.10 ± 0.060.79−4.98 ± 0.0710.5a,b−5.47 ± 0.043.41
α4β3γ2−6.13 ± 0.060.74−5.49 ± 0.123.24−5.55 ± 0.012.83


  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

The importance of tonic inhibitory conductances mediated by extrasynaptic GABAA receptors has become clear over the last few years (see Farrant and Nusser 2005). At least in some cases, the extrasynaptic GABAA receptors have subunit compositions that are distinct from their synaptic counterparts and this confers their unique activation and pharmacological properties. Thus far, the δ subunit has been localized exclusively to extrasynaptic and perisynaptic membranes in a variety of cell types, including cerebellar granule cells (Nusser et al. 1998) and the dentate gyrus of the hippocampus (Wei et al. 2003). This subunit appears to be preferentially expressed with the α6 or α4 subunit. The α4 and δ subunits are co-localized in restricted regions of the brain, especially the thalamus and hippocampus (Bencsits et al. 1999; Sur et al. 1999). Furthermore, parallel changes in the expression levels of these subunits in response to various physiological and pharmacological challenges (see e.g. Sundstrom-Poromaa et al. 2002; Lovick et al. 2005) suggest that the α4βxδ subtype is likely to be a native receptor combination.

The presence of the γ2 subunit in the GABAA receptor has been implicated in clustering of receptor subtypes at the post-synaptic synapse (Essrich et al. 1998) suggesting that the α4β3γ2 receptor is located predominantly in this region. In this study, we have, therefore, used the rat α4β3δ and α4β3γ2 receptors as representatives of putative extrasynaptic and synaptic receptors, respectively.

The human forms of these GABAA receptors were characterized previously in stably transfected cell lines (Brown et al. 2002) and it was shown that the α4β3δ receptor was more sensitive to GABA than the α4β3γ2 subtype. In the present study, we found a similar difference in potency (approximately 20-fold) for the rat receptor subtypes expressed in Xenopus oocytes. It is thus clear that inclusion of either the δ or γ2 subunit in the complex has a significant effect on the potency of GABA activation. However, neither of these subunits has been implicated directly in the binding of GABA or other receptor agonists. Site-directed mutagenesis studies have been used to localize the agonist activation sites to the interfaces between the β and α subunits (Amin and Weiss 1993). These sites appear to be formed by distinct loops of amino acids contributed by the primary (β) and secondary (α) subunits, in homologous positions to neurotransmitter binding sites in other members of the receptor family.

The aim of the present study was to identify structural determinants within the GABAA receptor δ subunit that contributes to the higher potency of δ-containing receptors for GABAergic agonists. We used a random chimeragenesis approach to form δ/γ2L subunits and expressed these with native α4 and β3 subunits in Xenopus oocytes. Similar approaches have been used to investigate structural requirements for the binding of benzodiazepine site ligands to other GABAA receptor subtypes (e.g. Boileau et al. 1998; Derry et al. 2004).

Our major finding is that a domain in the GABAA receptor δ subunit (S238-V264) confers high agonist sensitivity to the α4β3δ subtype. For all agonists studied (GABA, muscimol, THIP, and I4AA) inclusion of the χ277 chimeric subunit (containing the first 277 residues of δ) imparted δ-like characteristics to the receptor activation properties. In contrast, inclusion of the χ237-subunit led to receptors whose activation characteristics were not significantly different from the wild-type α4β3γ2L receptor. These results suggest that a specific region of the δ subunit (residues 238–277) plays a major role in conferring higher agonist potency. Inspection of the sequence homologies of the rat δ and γ2 subunits (Fig. 1c) further restricts this domain. Based on the recent receptor model and alignments of Ernst et al. (2005), the sequences of these subunits are identical between residues 265 and 277 (δ subunit numbering); thus the domain that confers high agonist potency can be narrowed further to residues 238–264. In the case of activation by either THIP or I4AA, the receptor that included the χ255 chimeric subunit displayed similar potency to the α4β3δ subtype. Again inspection of the sequence homology shows that there is conservation between residues 249 and 255 (δ subunit numbering) suggesting that the sequence S238-M248 of the δ subunit, lying in the N-terminal segment of TM1 may be of particular importance in conferring high affinity for these agonists.

Of the 27 amino acid residues lying in the S238-V264 domain of the δ subunit, 16 are identical to those in the equivalent positions of γ2 (see Fig. 1c). Where divergences occur, homologous amino acids are found in five positions (S238T, M240I, L245I, M248L, Q257K) leaving only six semi- or non-conservative substitutions (S242C, V243T, A247V, S256N, A258D, V264T), five of which lie within the putative TM1 domain (see Fig. 1c). In terms of overall structure, this domain appears to be highly conserved among all subunits of the cys-loop LGIC family (see Ernst et al. 2005) but, despite this homology, there is increasing evidence for its influence on the binding–gating coupling properties of different receptor subtypes (see below).

The structural determinants within the δ and γ2L subunits that underlie differences in agonist efficacy are less clear. In the α4β3δ receptor, THIP has been suggested to act as a ‘superagonist’ (see above and Brown et al. 2002). No clear patterns for determining agonist efficacy have emerged from the present study. This is perhaps not surprising as many factors lying in the pathway from ligand recognition to channel opening may dictate differences in transduction efficiency.

In agreement with the earlier results of Brown et al. (2002) who studied the human GABAA receptor subtypes, we found that bicuculline, SR95531, and picrotoxin did not differentiate between the wild-type rat α4β3δ and α4β3γ2 receptors. Apart from a small effect (approximately threefold) of bicuculline on the χ237 containing receptor, none of the expressed receptors showed any significant differences in their responses to these ligands. These results suggest that the influence of the δ/γ2L subunits on agonist potency does not involve radical changes in the architecture of the binding sites. Similarly, the lack of influence of the chimeric subunit on inhibition by the channel blocker, picrotoxin, suggests that the open channel characteristics are not dramatically affected. These results are consistent with the involvement of the δ/γ2L subunits, in particular the TM1 domain, in the transduction mechanism leading from agonist binding to channel opening rather than agonist recognition per se.

As both the α4 and δ subunits have been reported to be difficult to express in recombinant systems, the actual subunit composition of the receptors investigated in different laboratories has been questioned (see Borghese et al. 2006). In the present study, we have compared the expression of different combinations of α4, β3, δ, and γ2L subunits. Only the α4β3, α4β3δ, and α4β3γ2L combinations resulted in the expression of functional GABAA receptors. The expressed α4β3 and α4β3δ receptors displayed distinct properties with respect to the times required for expression, block by Zn2+, sensitivity to THIP and desensitization characteristics (unpublished results). However, increasing the relative amount of cRNA encoding the δ subunit affected both the EC50 for GABA activation and the apparent Hill coefficient of the response. Control experiments revealed that using higher than stoichiometric amounts of the δ subunit cDNA does not force the aberrant expression of other functional receptors containing only one or two subunit isoforms. These results are difficult to rationalize in terms of expression of a heterogeneous population of receptors. It seems more likely that the effects of changing cRNA ratios reflect changes in subunit stoichiometry within the pentameric α4β3δ complex. Further analysis of receptor composition under different expression conditions using biophysical approaches (see e.g. Barrera et al. 2005) will be required to interpret these observations. In the present study, although we cannot exclude the possibility that a heterogeneous population of receptors existed, the results obtained for the native receptor subtypes combinations are similar to those reported by others (see Brown et al. 2002; Storustova and Ebert 2006). Furthermore, the differential responses obtained for receptors incorporating the chimeric δ/γ2L subunits suggest that these subunits were also efficiently incorporated into the receptors examined.

A major issue in studying GABAA receptor properties is how the binding of an agonist to its extracellular sites that are predicted to lie approximately 30 Å above the membrane surface is communicated to the channel gate lying deep within the TM domains (see Unwin 2005). A number of previous studies have used chimeric/mutagenesis strategies to investigate the roles of the δ/γ2 subunits in signal transduction. Using a chimeragenesis approach, Jones-Davis et al. (2005), demonstrated that the TM1 domain of the γ2 subunit was important for benzodiazepine potentiation of recombinant α1β2γ/δ receptors. Haas and Macdonald (1999) reported that the kinetics of desensitization of the α1β3δ receptor were slower than those of the α1β3γ2L subtype and Bianchi and Macdonald (2002) subsequently demonstrated that these results could be attributed to determinants within the extracellular N-terminal domain and two residues in the TM1 domain (V233, Y234). In studies of general anesthetic action, it has been shown that determinants for modulation of GABA currents by propofol and pentobarbital lie within the N-terminal and TM1 domains of these subunits (Feng and Macdonald 2004; Feng et al. 2004). Together with our current results, the available data suggest that the TM1 domain of the δ subunit plays a significant role in conferring the unique functional and pharmacological properties of δ-containing receptors.

In conclusion, we have identified a structural domain within the GABAA receptor δ subunit (S238-V264) that confers high agonist sensitivity to the α4β3δ subtype. Further, we have shown that these effects are agonist-dependent and are likely to involve changes in the transduction mechanism that links agonist binding to channel activation.


  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This work was supported by the Canadian Institutes of Health Research and by scholarship support (to H.Y.) from UCB Pharma, Braine d’Alleud, Belgium. We thank Isabelle M. Paulsen and Delilah Mroczko for expert technical assistance. We are grateful to Professor Ian L. Martin (Aston University, UK) for many valuable discussions and to Drs Roy Massingham and Pierre Chatelain for their interest in this project.


  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Amin J. and Weiss D. S. (1993) GABAA receptor needs two homologous domains of the beta-subunit for activation by GABA but not by pentobarbital. Nature 366, 565569.
  • Barnard E. A., Skolnick P., Olsen R. W., Mohler H., Sieghart W., Biggio G., Braestrup C., Bateson A. N. and Langer S. Z. (1998) International Union of Pharmacology. XV. Subtypes of gamma-aminobutyric acidA receptors: classification on the basis of subunit structure and receptor function. Pharmacol. Rev. 50, 291313.
  • Barrera N. P., Herbert P., Henderson R. M., Martin I. L. and Edwardson J. M. (2005) Atomic force microscopy reveals the stoichiometry and subunit arrangement of 5-HT3 receptors. Proc. Natl Acad. Sci. USA 102, 1259512600.
  • Bencsits E., Ebert V., Tretter V. and Sieghart W. (1999) A significant part of native gamma-aminobutyric acidA receptors containing alpha4 subunits do not contain gamma or delta subunits. J. Biol. Chem. 274, 1961319616.
  • Bianchi M. T. and Macdonald R. L. (2002) Slow phases of GABA(A) receptor desensitisation: structural determinants and possible relevance for synaptic function. J. Physiol. 544, 318.
  • Bohme I., Rabe H. and Luddens H. (2004) Four amino acids in the alpha subunits determine the gamma-aminobutyric acid sensitivities of GABAA receptor subtypes. J. Biol. Chem. 279, 3519335200.
  • Boileau A. J, Kucken A. M., Evers A. R. and Czajkowski C. (1998) Molecular dissection of benzodiazepine binding and allosteric coupling using chimeric gamma-aminobutyric acidA receptor subunits. Mol. Pharmacol. 53, 295303.
  • Boileau A. J., Baur R., Sharkey L. M., Sigel E. and Czajkowski C. (2002) The relative amount of cRNA coding for gamma2 subunits affects stimulation by benzodiazepines in GABAA receptors expressed in Xenopus oocytes. Neuropharmacology 43, 695700.
  • Borghese C. M., Storustovu S., Ebert B., Herd M. B., Belelli D., Lambert J. J., Marshall G., Wafford K. A. and Harris R. A. (2006) The delta subunit of gamma-aminobutyric acid type A receptors does not confer sensitivity to low concentrations of ethanol. J. Pharmacol. Exp. Ther. 316, 13601368.
  • Brickley S. G., Revilla V, Cull-Candy S. G., Wisden W. and Farrant M. (2001) Adaptive regulation of neuronal excitability by a voltage-independent potassium conductance. Nature 409, 8892.
  • Brown N., Kerby J., Bonnert T. P., Whiting P. J. and Wafford K. A. (2002) Pharmacological characterization of a novel cell line expressing human alpha4beta3delta GABAA receptors. Br. J. Pharmacol. 136, 965974.
  • Derry J. M., Dunn S. M. and Davies M. (2004) Identification of a residue in the gamma-aminobutyric acid type A receptor alpha subunit that differentially affects diazepam-sensitive and -insensitive benzodiazepine site binding. J. Neurochem. 88, 14311438.
  • Ernst M., Bruckner S., Boresch S. and Sieghart W. (2005) Comparative models of GABAA receptor extracellular and transmembrane domains: important insights in pharmacology and function. Mol. Pharmacol. 68, 12911300.
  • Essrich C., Lorez M., Benson J. A., Fritschy J. M. and Luscher B. (1998) Postsynaptic clustering of major GABAA receptor subtypes requires the gamma 2 subunit and gephyrin. Nat. Neurosci. 1, 563571.
  • Farrant M. and Nusser Z. (2005) Variations on an inhibitory theme: phasic and tonic activation of GABAA receptors. Nat. Rev. Neurosci. 6, 215229.
  • Farrar S. J., Whiting P. J., Bonnert T. P. and McKernan R. M. (1999) Stoichiometry of a ligand-gated ion channel determined by fluorescence energy transfer. J. Biol. Chem. 274, 1010010104.
  • Feng H. J. and Macdonald R. L. (2004) Multiple actions of propofol on alphabetagamma and alphabetadelta GABAA receptors. Mol. Pharmacol. 66, 15171524.
  • Feng H. J., Bianchi M. T. and Macdonald R. L. (2004) Pentobarbital differentially modulates alpha1beta3delta and alpha1beta3gamma2 L GABAA receptor currents. Mol. Pharmacol. 66, 9881003.
  • Haas K. F. and Macdonald R. L. (1999) GABAA receptor subunit gamma2 and delta subtypes confer unique kinetic properties on recombinant GABAA receptor currents in mouse fibroblasts. J. Physiol. 514, 2745.
  • Jones-Davis D. M., Song L., Gallagher M. J. and Macdonald R. L. (2005) Structural determinants of benzodiazepine allosteric regulation of GABA(A) receptor currents. J. Neurosci. 25, 80568065.
  • Korpi E. R. and Luddens H. (1993) Regional gamma-aminobutyric acid sensitivity of t-butylbicyclophosphoro[35S]thionate binding depends on γ-aminobutyric acidA receptor α subunit. Mol. Pharmacol. 44, 8792.
  • Lovick T. A., Griffiths J. L., Dunn S. M. and Martin I. L. (2005) Changes in GABA(A) receptor subunit expression in the midbrain during the oestrous cycle in Wistar rats. Neuroscience 131, 397405.
  • McKernan R. M. and Whiting P. J. (1996) Which GABAA-receptor subtypes really occur in the brain? Trends Neurosci. 19, 139143.
  • Mody I. (2001) Distinguishing between GABA(A) receptors responsible for tonic and phasic conductances. Neurochem. Res. 26, 907913.
  • Moore K. R. and Blakely R. D. (1994) Restriction site-independent formation of chimeras from homologous neurotransmitter-transporter cDNAs. Biotechniques 17, 130136.
  • Nayeem N., Green T. P., Martin I. L. and Barnard E. A. (1994) Quaternary structure of the native GABAA receptor determined by electron microscopic image analysis. J. Neurochem. 62, 815818.
  • Nusser Z., Sieghart W. and Somogyi P. (1998) Segregation of different GABAA receptors to synaptic and extrasynaptic membranes of cerebellar granule cells. J. Neurosci. 18, 16931703.
  • Saxena N. C. and Macdonald R. L. (1994) Assembly of GABAA receptor subunits: role of the delta subunit. J. Neurosci. 14, 70777086.
  • Sieghart W. (1995) Structure and pharmacology of gamma-aminobutyric acidA receptor subtypes. Pharmacol. Rev. 47, 181234.
  • Sieghart W., Fuchs K., Tretter V., Ebert V., Jechlinger M., Hoger H. and Adamiker D. (1999) Structure and subunit composition of GABA(A) receptors. Neurochem. Int. 34, 379385.
  • Smith K. M., Ng A. M., Yao S. Y. et al. (2004) Electrophysiological characterization of a recombinant human Na+-coupled nucleoside transporter (hCNT1) produced in Xenopus oocytes. J. Physiol. 558, 807823.
  • Storustova S. I. and Ebert B. (2006) Pharmacological characterization of agonists at delta-containing GABAA receptors: functional selectivity for extrasynaptic receptors is dependent on the absence of gamma2. J. Pharmacol. Exp. Ther. 316, 13511359.
  • Sundstrom-Poromaa I., Smith D. H., Gong Q. H., Sabado T. N., Li X., Light A., Wiedmann M., Williams K. and Smith S. S. (2002) Hormonally regulated alpha4beta2delta GABAA receptors are a target for alcohol. Nat. Neurosci. 5, 721722.
  • Sur C., Farrar S. J., Kerby J., Whiting P. J., Atack J. R. and McKernan R. M. (1999) Preferential coassembly of alpha4 and delta subunits of the gamma-aminobutyric acid A receptor in rat thalamus. Mol. Pharmacol. 56, 110115.
  • Unwin N. (2005) Refined structure of the nicotinic acetylcholine receptor at 4A resolution. J. Mol. Biol. 346, 967989.
  • Wei W., Zhang N., Peng Z., Houser C. R. and Mody I. (2003) Perisynaptic localization of delta subunit-containing GABA(A) receptors and their activation by GABA spillover in the mouse dentate gyrus. J. Neurosci. 23, 1065010661.
  • Whiting P. J. (2003) GABA-A receptor subtypes in the brain: a paradigm for CNS drug discovery? Drug Discov. Today 8, 445450.