Tandem subunits effectively constrain GABAA receptor stoichiometry and recapitulate receptor kinetics but are insensitive to GABAA receptor-associated protein

Abstract

GABAergic synapses likely contain multiple GABAA receptor subtypes, making postsynaptic currents difficult to dissect. However, even in heterologous expression systems, analysis of receptors composed of alpha, beta, and gamma subunits can be confounded by receptors expressed from alpha and beta subunits alone. To produce recombinant GABAA receptors containing fixed subunit stoichiometry, we coexpressed individual subunits with a "tandem" alpha1 subunit linked to a beta2 subunit. Cotransfection of the gamma2 subunit with alphabeta-tandem subunits in human embryonic kidney 293 cells produced currents that were similar in their macroscopic kinetics, single-channel amplitudes, and pharmacology to overexpression of the gamma subunit with nonlinked alpha1 and beta2 subunits. Similarly, expression of alpha subunits together with alphabeta-tandem subunits produced receptors having physiological and pharmacological characteristics that closely matched cotransfection of alpha with beta subunits. In this first description of tandem GABAA subunits measured with patch-clamp and rapid agonist application techniques, we conclude that incorporation of alphabeta-tandem subunits can be used to fix stoichiometry and to establish the intrinsic kinetic properties of alpha1beta2 and alpha1beta2gamma2 receptors. We used this method to test whether the accessory protein GABAA receptor-associated protein (GABARAP) alters GABAA receptor properties directly or influences subunit composition. In recombinant receptors with fixed stoichiometry, coexpression of GABARAP-enhanced green fluorescent protein (EGFP) fusion protein had no effect on desensitization, deactivation, or diazepam potentiation of GABA-mediated currents. However, in alpha1beta2gamma2S transfections in which stoichiometry was not fixed, GABARAP-EGFP altered desensitization, deactivation, and diazepam potentiation of GABA-mediated currents. The data suggest that GABARAP does not alter receptor kinetics directly but by facilitating surface expression of alphabetagamma receptors.

Figure 1.

Tandems and Western blot. A, Tandem subunits (depicted with arc linkage) were expressed alone or with free α1, β2, or γ2S subunits in a 2:1 ratio. Only α1 and γ2 coexpression gave functional expression. B, The α_FLAG-β tandem subunit is stable. Representative Western blot from cells expressing α_FLAG and α_FLAG-β subunits probed with anti-FLAG antibody. The immunoreactive band at 54 kDa corresponds to the α_FLAG monomer, and the band at 110 kDa corresponds to α_FLAG-β concatamerized dimer. In cells expressing the α_FLAG-β tandem subunit, no smaller immunoreactive molecular weight bands were detected, indicating that the tandem subunit is not appreciably breaking down. Similar results were obtained in three experiments. C, Peptide map of the linker between the α1 and β2 subunits in the αβ tandem. Nine additional glutamine residues were added using a CAG repeat, and the β2 signal peptide sequence was retained in the construct.

Figure 2.

Current traces comparing receptors formed from different transfections: α1β2, αβtan+α1, α1β2γ2S 1:1:10, αβtan+γ2S, and αβtan+γ2L. GABA (10 mm) pulse durations were 20 ms (top row, A, D), 2000 ms (middle row, B, E) and 20,000 ms (bottom row, C, F). Differences in desensitization are more clearly seen in longer (2000 and 20,000 ms) pulses, and comparisons of deactivation are visible in short (20 ms) traces and in longer traces when normalized to the current at the end of the pulse (middle row, insets). One example of αβtan+γ2L is shown for comparison in C (gray).

Figure 3.

Desensitization and deactivation profiles for different transfections. A, Weighted desensitization time constants (τ_w) differ betweenα1β2 or αβtan+α1 transfections (open bars) and α1β2γ2S 1:1:10 or αβtan+γ2S or γ2L transfections (black bars). Transfections with αβγ in a 1:1:0.5 or 1:1:1 ratio (gray bars) result in intermediate τ_w. Statistical comparisons reveal that αβ and αβtan+α1 transfections differ significantly from αβγ 1:1:10 and αβtan+γ2S transfections (*p < 0.001) but not from αβγ 1:1:0.5 or 1:1:1.αβγ 1:1:10 and αβtan+γ2S weighted time constants are not significantly different from each other. Data are mean ± SEM. B, Comparison of the percentage of the peak current remaining at the ends of 10 mm GABA pulses of varying length. Percentage desensitization was calculated as follows: 1 – (current amplitude at pulse end/peak current) × 100%. Transfections with α1β2γ2S 1:1:1 (gray circles) result in values intermediate to α1β2 and αβtan+α1 (open symbols) versus α1β2γ2S 1:1:10, αβtan+γ2S, and αβtan+γ2L transfections (closed symbols). This difference is most clearly observed at the ends of 2 s pulses (*p < 0.001 compared with α1 β2 γ2S 1:1:1 transfections). Data are mean ± SEM. C, Weighted deactivation time constants for GABA_A subunit transfections. Deactivation time constants (τ_w) for α1β2 and αβtan+α1 transfections (open symbols) and α1β2γ2S 1:1:10, αβtan+γ2S, and αβtan+γ2L transfections (closed symbols). Transfections with α1β2γ2S 1:1:1 (gray symbols) result in intermediate time constants in pulses of 200 or 2000 ms (*p < 0.01 compared with α1β2γ2S 1:1:1 transfections). Data are mean ± SEM. Components of each weighted time constant are listed in Tables 1 and 2.

Figure 4.

Zinc block of currents from transfections with GABA_A receptor subunits and tandems. A, Current traces from 200 ms 1 mm GABA pulses (black) superimposed with currents from the same patch exposed to a coapplication of 1 mm GABA plus 30 μm ZnCl₂ (gray) after equilibration in control solution plus 30 μm ZnCl₂. B, Histograms for percentage of peak current blocked by Zn²⁺ for 200 ms GABA pulses. α1β2γ2S 1:1:10, αβtan+γ2S, αβtan+γ2L 1:1:10, and αβtan+γ2L transfections (black bars) are significantly different (p < 0.001) from α1β2 and αβtan+α transfections (open bars). Data are mean ± SD (to show error) for 4–17 patches for each condition.

Figure 5.

GABA concentration responses. A, Current responses to increasing concentrations of GABA for α1β2 transfections (with 0.3, 1.0, 3.0, 10, 30, 100, 300, and 1000 μm GABA) and α1β2γ2S 1:1:10 transfections (with 1, 3, 10, 30, 100, 300, 1000, and 3000 μm GABA). Pulse durations are described in Materials and Methods. Note that less desensitization occurs in α1β2γ2S 1:1:10 at any given concentration. B, C, Concentration–response curves and fits for α1β2 and αβtan+α1 (open symbols) versus α1β2γ2S 1:1:10 and αβtan+γ2S (closed symbols) transfections (B) or α1β2γ2S 1:1:1 transfections (C, gray circles). Currents were normalized to a maximal response at a GABA concentration 10-fold higher than the curve-fit maximal responses shown. Data shown are means ± SEM for four or more patches each. In some cases, the error bars were smaller than the symbol for the mean. Note that 1 mm GABA is near-maximal for all combinations shown.

Figure 6.

Single-channel openings and conductances. A–D, Representative current traces for α1β2 (A), αβtan+α1 (B), α1β2γ2S 1:1:10 (C), and αβtan+γ2S (D) transfections showing single-channel openings in tail currents after deactivation from a 200 ms, 10 mm GABA pulse or during the pulse in the case of D. All points amplitude histograms (center) of the closed and open levels for the openings indicated (insets) and Gaussian fits. Current–voltage plots were constructed for each of the transfections(right). Calculated chord conductances are listed beneath each plot for main and subconductances. Data are mean ± SD from three or more patches.

Figure 7.

Single-channel conductances in α1β2γ2S 1:1:1 transfections. A, Single-channel openings in tail currents after deactivation from a 200 ms, 10 mm GABA pulse with openings indicative of αβγ receptors (e.g., 29 pS) or αβ receptors (e.g., 10 pS, inset). B, Clusters of single openings from an excised patch held at ± 80 mV. Amplitude histograms (center) of the closed and open levels for the clusters indicated (arrows) and Gaussian fits yielding ∼15 pS chord conductance are shown, consistent with αβ receptors. C, Current–voltage plots were constructed for each of the transfections (right), culled from seven patches from six different transfections in which single-channel openings were detectable. Calculated chord conductances are listed beneath the plot for main and subconductances (sub). Data are mean ± SD.

Figure 8.

Coexpression of GABA_A subunits with a GABARAP-EGFP fusion protein. A, Summary of effects of GABARAP-EGFP on desensitization. Desensitization weighted time constants (τ_w) for each of the transfection types are shown in the left panel, compared with the extent of desensitization in the right panel. Data are mean ± SEM. The asterisk denotes significant difference from the non-GABARAP transfection at p < 0.01. B, Representative current traces (black traces) of α1β2γ2S 1:1:0.5 (top), α1β2 (second row), α1β2γ2S 1:1:10 (third row), and αβtan+γ2S transfections (bottom) overlaid with current traces from the same subunits cotransfected with GABARAP-EGFP (gray traces). GABA pulse durations shown are 2000 ms (left) to show changes in desensitization and 200 ms (right) to show changes in deactivation. C, Summary of effects of GABARAP-EGFP on deactivation. Deactivation weighted time constants (τ_w) following GABA pulses of 20, 200, and 2000 ms duration are shown for αβ (top left panel), α1β2γ2S 1:1:0.5 (top right), α1β2γ2S 1:1:10, and αβtan+γ2S transfections (bottom left) with (closed symbols) and without (open symbols) coexpression of GABARAP-EGFP. The bottom right panel summarizes data for the 2000 ms GABA pulse. Data are mean ± SEM. The asterisk denotes significant difference from the non-GABARAP transfection at p < 0.05. D, Effects of cotransfection of GABARAP-EGFP on diazepam-induced potentiation of subsaturating GABA responses. Each point represents data from one patch pulsed for 500 ms with 3 μm GABA for several sweeps and then equilibrated in 1 μm DZ and pulsed with 3 μm GABA plus 1 μm DZ. Peak currents were measured preincubation and postincubation with DZ, and potentiation was calculated as follows: I_GABA+DZ/I_GABA – 1. Horizontal lines represent the mean. The asterisk denotes significant difference from the non-GABARAP transfection at p < 0.001. G-RAP, GABARAP.

Figure 9.

Model of currents for αβ receptors compared with αβγ receptors. A, Kinetic scheme. Using a backbone kinetic model with one unbound (U) state, single-liganded (B1) and double-liganded (B2) states, and an open state (O₁,O₂) from each of the liganded states (Jones and Westbrook, 1995; Hinkle and Macdonald, 2003), currents for αβtan+α1 versus αβtan+γ2S were simulated (see Materials and Methods). Parameters optimized were desensitization and deactivation time constants, apparent affinity (EC₅₀), and open probability for macroscopic currents. A simplified model for αβtan+α1(αβ) receptors had the following rates: k_on, 5 × 10^6–1 s^–1; k_on, 100 s^–1; β1 (B1 to O1), 200 s^–1 m;α1 (O1 to B1), 1100 s^–1; β2, 1800 s^–1; α2, 280 s^–1. The model for αβtan+γ2S (αβγ) receptors used the following rates: k_on, 5 × 10^4–1 s^–1; k_on, 200 s^–1; β1 (B1 to O1), 50 s^–1 m; α1 (O1 to B1), 3100 s^–1; β2, 1800 s^–1; α2, 280 s^–1. Both models differ dramatically in their desensitization (d) and recovery (r) rates. For αβtan+α1 receptors, faster entry desensitized states (D1, D2, gray) were connected to B2 with the following rates: d1, 200 s^–1; r1, 80 s^–1; d2, 60 s^–1; r2, 12 s^–1. Slower entry D states (D3, D4) were modeled with the following rates: d3, 12 s^–1; r3, 0.03 s^–1; d4, 6 s^–1; r4, 0.3 s^–1. For the two relatively slow desensitizing components seen in αβtan+γ2S currents (also designated D3 and D4), the rates were as follows: d3,2s^–1; r3, 0.75 s^–1; d4,1s^–1; r4, 0.05 s^–1. B, Model of diazepam potentiation for varying amounts of αβ and αβγ receptors in a mixture. Experimentally determined values for maximal diazepam potentiation (I_GABA+DZ/I_GABA – 1), EC₅₀, open probability, and conductance for both αβ and αβγ receptors were used to generate the curve of expected values and 95% confidence intervals (see Materials and Methods). C, Simulated model traces for αβtan+α1 and αβtan+γ2S (bold), with averaged traces generated from mixtures of αβtan+α1 and αβtan+γ2S receptors in a 3:1, 2:1, and 1:1αβtan+α1:αβtan+γ2S ratio (dotted line). D, Traces from three excised patches from cells transected at a 1:1:1 α1:β2:γ2S ratio, showing variability in macroscopic kinetics.