Random assembly of GABA rho1 and rho2 subunits in the formation of heteromeric GABA(C) receptors

Abstract

1. Various combinations of the rho subunits (rho(1A), rho(1B), rho(2A), rho(2B)) of GABA(C) receptors cloned from white perch retina were expressed in Xenopus oocytes, and electrophysiological and pharmacological methods were used to test their ability to co-assemble into heteromeric receptors. Simultaneous injection of the two subunits, irrespective of their relative proportions, led invariably to the formation of a preponderance of heteromeric receptors. 2. The GABA deactivation responses elicited from these cells could be described by a single exponential decay, and their pharmacological responses deviated significantly from those expected of a simple mixture of two homomeric rho(1) and rho(2) receptors. In contrast, a double exponential function comprising fast and slow components was required to fit the GABA deactivation responses elicited from oocytes sequentially expressing rho(1) and rho(2) subunits, a condition that favors the formation of a mixture of homomeric rho(1) and rho(2) receptors. 3. Both the GABA-response kinetics and the sensitivity to picrotoxin of the heteromeric perch rho(1B)rho(2A) receptor varied with the proportion of the subunit RNA injected, indicating there is no fixed stoichiometry for their co-assembly into heteromeric rho(1)rho(2) receptors. 4. If native GABA(C) receptors in retinal neurons behave in a similar manner as in the oocyte expression system, these finding suggest that the properties of their GABA(C) receptors are likely to be influenced by the transcription/translation efficiency of GABA rho subunit genes.

Fig. 1.

Single-cell PCR: (Upper) Photographs of solitary retinal bipolar cells isolated from perch retina. (Lower) PCR products for each of the GABA ρ subunits expressed by the bipolar cells shown above. Positive PCR controls (lane c) were carried out using the cDNA of the cloned GABA ρ subunits.

Fig. 2.

Current recordings from oocytes expressing each of the homomeric GABA ρ subunits. The τ values for the GABA deactivation responses were derived from the decay curves following the abrupt switch from 10 τ M GABA to the Ringer’s solution (arrow). Inset: bar graphs show the averaged values of τ (mean ± SEM) obtained from 5 to 7 oocytes expressing each of the perch ρ receptors.

Fig. 3.

Kinetics of GABA deactivation responses elicited from oocytes expressing ρ_1B and ρ_2A subunits. Each filled circle indicates the recorded current amplitude at a particular time after GABA was turned off, and the response was normalized to the current elicited at the moment the applied GABA (10 μM) was switched to Ringer’s solution. (A) Response obtained from an oocyte sequentially expressing ρ_1B and ρ_2A subunits 2 days apart. Data were fitted with a single exponential function (dash line) or with a double exponential (solid line). (B) Response obtained from an oocyte co-expressing ρ_1B and ρ_2A subunits. The GABA inactivation response could be well fit by a single exponential function, indicating that a heteromeric ρ_1Bρ_2A receptor was predominant on this oocyte. The dotted lines illustrate typical GABA inactivation responses for the homomeric ρ_1B and ρ_2A receptors. (C, D, and E) GABA deactivation responses elicited from a combination of perch (ρ_1A and ρ_2A), (ρ_1A and ρ_2B), and (ρ_1B and ρ_2B) subunits, respectively. Each filled circle represents the current amplitude (normalized to the maximum GABA response) at the indicated time after GABA was turned off. For every GABA ρ₁ and ρ₂ subunit combination, the deactivation response was well fit by a single exponential decay, indicating the formation of heteromeric perch ρ receptors. The τ values (sec) listed in the figure are averaged from 5 to 8 cells and shown as means ± S.E.M. The dotted lines illustrate typical GABA inactivation responses for the respective homomeric receptors. Note that except ρ_1Aρ_2B (shown in D) where the kinetic of the heteromeric receptor is slower than either of its components, all other combinations result in kinetics that are intermediate between the two homomeric components.

Fig. 4.

Dose–response curve of heteromeric ρ_1Bρ_2A receptors. The amplitudes of the GABA-elicited responses from oocytes co-expressing perch ρ_1B and ρ_2A subunits (filled circles and solid line) were normalized to those elicited by 100 μM GABA. Data were fitted with a Hill equation with an EC₅₀ of 0.68 μM and Hill co-efficient of 1.3. The dose–response relations for homomeric ρ_1B and ρ_2A receptors (dashed lines) are also plotted for comparison (Qian et al., 1998). Although the heteromeric receptors exhibited intermediate sensitivity to GABA, the slope of the curve (Hill coefficient) was not significantly different from those of the homomeric receptors (1.3 and 1.7 for ρ_1B and ρ_2A receptors, respectively). These results are consistent with the formation of heteromeric ρ_1Bρ_2A receptors on oocytes co-expressing perch ρ_1B and ρ_2A subunits.

Fig. 5.

Pharmacological properties of heteromeric perch-ρ receptors. The four bar graphs show the pharmacological profiles of heteromeric ρ₁ρ₂ receptors compared with their respective homomeric receptors in response to β alanine (3 mM), taurine (10 mM), glycine (10 mM) and I4AA (100 μM). The responses were normalized to the maximum GABA response, and the data were averaged from 4 to 5 cells. The responses obtained from oocytes co-expressing ρ₁ and ρ₂ subunits diverge from simple averaged responses of the two homomeric receptors, providing further evidence for the formation of heteromeric ρ₁ρ₂ receptors.

Fig. 6.

Kinetic analysis of GABA deactivation responses from oocytes expressing various proportions of ρ_1B and ρ_2A subunits. (A) The normalized current amplitudes plotted on a semi-logarithmic scale show that the data from each oocyte could be fit by a straight line, indicating that a single exponential is the dominant component of the deactivation response. The slope of each line varied with the different proportions of ρ_1B:ρ_2A expressed, indicating that no fixed stoichiometry existed for heteromeric ρ_1Bρ_2A receptors. Homomeric receptors are shown as filled symbols, where 0:1 represents the ρ_2A receptor and 1:0 is the ρ_1B receptor. (B) The bar graphs illustrate the averaged time constant (τ) of the deactivation responses elicited from 4 to 5 oocytes expressing each proportion of ρ_1B and ρ_2A subunits.

Fig. 7.

Picrotoxin inhibition curves obtained from heteromeric ρ_1Bρ_2A receptors. Three different ratios of ρ_1B and ρ_2A subunit RNAs were tested. Data are presented as the responses to GABA (10 μM) in the presence of various concentrations of picrotoxin; responses were normalized to the response elicited by GABA alone. The continuous curves are fits to Hill equation with the following parameters: ρ_1B:ρ_2A = 1:1 (IC₅₀ = 3.2 μM, Hill coefficient = 0.91); ρ _1B:ρ_2A = 3:1 (IC₅₀ = 5.2 μM, Hill coefficient = 0.83); ρ_1B:ρ_2A = 6:1 (IC₅₀ = 21.2 μM, Hill coefficient = 0.73). Data were collected from 4 to 5 oocytes for each combination, and for comparison, are shown together with the picrotoxin inhibition curves for homomeric ρ_2A and ρ_1B receptors (open symbols). In the case of the homomeric receptors, the fit to the Hill equation (dash lines) yielded values for IC₅₀ of 2.07 μM and 56.3μ M, with Hill coefficients of 0.89 and 0.73 for homomeric ρ_2A and ρ_1B receptors, respectively. Although the sensitivity to picrotoxin changes with the proportion of the subunits expressed, the slopes were similar for the various curves.