Gamma-aminobutyric acid type B (GABA(B)) receptor internalization is regulated by the R2 subunit

Abstract

γ-Aminobutyric acid type B (GABA(B)) receptors are important for slow synaptic inhibition in the CNS. The efficacy of inhibition is directly related to the stability of cell surface receptors. For GABA(B) receptors, heterodimerization between R1 and R2 subunits is critical for cell surface expression and signaling, but how this determines the rate and extent of receptor internalization is unknown. Here, we insert a high affinity α-bungarotoxin binding site into the N terminus of the R2 subunit and reveal its dominant role in regulating the internalization of GABA(B) receptors in live cells. To simultaneously study R1a and R2 trafficking, a new α-bungarotoxin binding site-labeling technique was used, allowing α-bungarotoxin conjugated to different fluorophores to selectively label R1a and R2 subunits. This approach demonstrated that R1a and R2 are internalized as dimers. In heterologous expression systems and neurons, the rates and extents of internalization for R1aR2 heteromers and R2 homomers are similar, suggesting a regulatory role for R2 in determining cell surface receptor stability. The fast internalization rate of R1a, which has been engineered to exit the endoplasmic reticulum, was slowed to that of R2 by truncating the R1a C-terminal tail or by removing a dileucine motif in its coiled-coil domain. Slowing the rate of internalization by co-assembly with R2 represents a novel role for GPCR heterodimerization whereby R2 subunits, via their C terminus coiled-coil domain, mask a dileucine motif on R1a subunits to determine the surface stability of the GABA(B) receptor.

FIGURE 1.

Constitutive internalization of R1a^BBSR2 receptors in live GIRK cells and neurons. A, GIRK cells expressing R1a^BBSR2 receptors were incubated in 3 μg/ml BTX-AF555 for 10 min at room temperature and then imaged over 0–60 min at 16–18, 22–24, and 30–32 °C. The arrowheads indicate internalized R1a subunits. Scale bar, 5 μm. B and C, rate (B) and extent (C) of internalization of BTX-AF555-tagged R1a^BBSR2 at 16–18 °C (red), 22–24 °C (blue), and 30–32 °C (black). n = 6–15; ***, p < 0.001. In this and following figures, all points and bars represent means ± S.E. (error bars). D, Hippocampal neurons (14–21 DIV) expressing R1a^BBSR2 and eGFP were incubated in 1 mm d-tubocurarine for 5 min followed by 3 μg/ml BTX-AF555 for 10 min at room temperature and imaged at different time points at 16–18 °C or 30–32 °C. The arrowheads indicate internalized R1a subunits. Scale bar, 10 μm. E and F, rate (E) and extent (F) of internalization of BTX-AF555-tagged R1a^BBSR2 receptors at 16–18 °C (red) and 30–32 °C (black) in live hippocampal neurons, n = 6–12. ***, p < 0.001.

FIGURE 2.

GABA_B receptors are internalized via clathrin and dynamin-dependent mechanisms and recruited to lysosomes. Shown are the rate (A) and extent (B) of constitutive internalization of BTX-AF555-tagged R1a^BBSR2 receptors in the absence (●) and presence of dynasore (▴) or chlorpromazine (□). GIRK cells expressing R1a^BBSR2 receptors were incubated in 3 μg/ml BTX-AF555 for 10 min at room temperature and imaged over 0–60 min at 30–32 °C in the presence of either 50 μg/ml CPZ or 80 μm DYN, n = 5–11. **, p < 0.05 (one-way ANOVA). C, hippocampal neurons (14–21 DIV) expressing R1a^BBSR2 and either eGFP-Rab5, eGFP-Rab11, or Rab7-eGFP were incubated in 1 mm d-tubocurarine for 5 min, followed by 3 μg/ml BTX-AF555 for 10 min at room temperature. Cells were incubated at 37 °C for 30–60 min and then fixed and imaged. The arrowheads depict co-localization in the soma (top) and a dendrite (bottom). Scale bars, 5 μm; error bars, S.E.

FIGURE 3.

Faster and more extensive internalization of R1a^BBS-ASA compared with R1a^BBSR2. A, GIRK cells expressing either R1a^BBS-ASA (top) or R1a^BBSΔCT (bottom) were incubated in 3 μg/ml BTX-AF555 for 10 min at room temperature to label surface GABA_B receptors and imaged over 0–60 min at room temperature. B, the rate of internalization of BTX-AF555-tagged R1a^BBSR2 heteromers and R1a^BBS-ASA and R1a^BBSΔCT homomers at room temperature (n = 6–10). The inset shows the relative positions of the ASA motif and C-terminal truncation (ΔCT). C, exponential decay time constants (min) for surface R1a^BBSR2, R1a^BBS-ASA, and R1a^BBSΔCT. D, extent of internalization for R1a^BBSR2, R1a^BBS-ASA, or R1a^BBSΔCT. One-way analysis of variance was used. *, p < 0.05; **, p < 0.01; ***, p < 0.001. Scale bar, 5 μm; NS, not significant; error bars, S.E.

FIGURE 4.

Silent incorporation of the BBS into R2 subunits. A, schematic diagram showing the relative locations for the BBS and the Myc and FLAG epitopes in GABA_B R1a and R2 subunits. For the R2 subunit, the BBS was inserted between Val⁶⁷ and Thr⁶⁸, as shown in the segment of the primary sequence. B, images of GIRK cells expressing R1aR2^BBS and eGFP were incubated with (+) or without (−) 3 μg/ml BTX-AF555 for 10 min at room temperature. C, images of rat cultured hippocampal neurons expressing R1aR2^BBS and eGFP, incubated with 1 mm d-tubocurarine for 5 min followed by incubation with or without 3 μg/ml BTX-AF555 for 10 min at room temperature. Scale bars, 5 μm. D, GABA concentration response curves for R1aR2, R1aR2^BBS, and R1a^BBSR2 receptors and BTX-bound R1aR2^BBS and R1a^BBSR2 receptors all expressed in GIRK cells (n = 5–13). E, whole-cell radioligand binding experiments with ¹²⁵I-BTX for the R1aR2^BBS receptor (n = 6). Error bars, S.E.

FIGURE 5.

R2 subunits slow the rate of GABA_B receptor internalization. A, rate and extent of internalization for BTX-AF555-tagged R2^BBS (gray) or R1aR2^BBS (black) at 16–18 °C (▴), 22–24 °C (■), and 30–32 °C (●) (n = 5–13). GIRK cells expressing R2^BBS or R1aR2^BBS were incubated in 3 μg/ml BTX-AF555 for 10 min at room temperature prior to imaging over 0–60 min. B, comparison of internalization of R1a^BBS-ASA with R1a^BBSR2, R1aR2^BBS, and R2^BBS (data taken from A and Fig. 1A) at 22–24 °C. C, rate of internalization of BTX-AF555-tagged R1a^BBSR2 (□) and R1aR2^BBS (○) receptors at 16–18 °C and 30–32 °C in live hippocampal neurons (n = 6 - 14). Neurons (14–21 DIV) expressing R1aR2^BBS and eGFP were incubated in 1 mm d-tubocurarine for 5 min followed by 3 μg/ml BTX-AF555 for 10 min at room temperature before imaging at 16–18 or 30–32 °C.

FIGURE 6.

Dileucine motif on R1a determines the rate of internalization. A, GIRK cells expressing R1a^{BBS-ASA,L889A,L890A} (top) or R1a^BBS-ASAR2ΔCT (bottom) were incubated in 3 μg/ml BTX-AF555 for 10 min at room temperature and imaged over 0–60 min at room temperature. Scale bar, 5 μm. B, rates of constitutive internalization for R2^BBS, R1a^BBSR2, R1a^{BBS-ASA,L889A,L890A}, R1a^BBS-ASA, and R1a^BBS-ASAR2ΔCT receptors (n = 6). C, decay time constants for the surface membrane fluorescence for the subunits indicated, where R1a^{BBS-ASA-L889A,L890A} = R1a^BBS-ASA-AA. D, extent of internalization for the subunits indicated. *, p < 0.05; **, p < 0.01; ***, p < 0.001, one-way analysis of variance. Error bars, S.E.

FIGURE 7.

GABA_B receptors are constitutively internalized as heterodimers. A, schematic for the dual labeling strategy for R2 (wild-type BBS, left) and R1a (mutant BBS, right) subunits. B, GABA concentration response curves for R1aR2, R1aR2^BBS, R1a^BBSR2, and R1a^BBS-CCR2^BBS expressed in GIRK cells (n = 7–13). C, hippocampal neurons (14–21 DIV) expressing R1a^BBS-CCR2^BBS were incubated in 1 mm d-tubocurarine for 5 min followed by 200 μm DTT for 30 min at room temperature, 20 μm MTSES for 5 min, 3 μg/ml BTX-AF488 for 10 min at 4 °C, 5 mm DTT for 5 min at room temperature, and 3 μg/ml BTX-AF555 for 10 min at 4 °C and imaged at different times at 30–32 °C. The arrowheads indicate some examples of co-localized and internalized R1a and R2 subunits. D, rate of constitutive internalization of BTX-AF488-tagged R2^BBS and BTX-AF555-tagged R1a^BBS-CC receptors (n = 5).

FIGURE 8.

Trafficking model for GABA_B receptors. A, schematic diagram that illustrates the trafficking of GABA_B receptors from the cell surface to early/late endosomes and onto lysosomes. The surface replenishment pathway involves the recycling of receptors and the synthetic pathway from the Golgi stack. The respective rate constants and key for the GABA_B receptors are indicated. B, rates of constitutive internalization for R2^BBS, R1a^BBSR2, R1a^{BBS-ASA,L889A,L890A}, R1a^BBS-ASA, and R1a^BBS-ASAR2ΔCT receptors, taken from Fig. 6B. The curve fits are generated using the model in A with k_endo = 0.067 min⁻¹, k_in = 0.128 min⁻¹, k_pb = 0.01 min⁻¹, k_recyc = 0.04 min⁻¹, k_degrad = 0.0083 min⁻¹ (for R1aR2; blue) with 45% of receptors recycling, and k_endo = 0.11 min⁻¹, k_in = 0.128 min⁻¹, k_pb = 0.01 min⁻¹, k_recyc = 0.04 min⁻¹, k_degrad = 0.0083 min⁻¹ (for R1a^BBS-ASA; red) with 30% of receptors recycling.