Subcellular distribution of GABA(B) receptor homo- and hetero-dimers

Abstract

GBRs (GABA(B) receptors; where GABA stands for gamma-aminobutyric acid) are G-protein-coupled receptors that mediate slow synaptic inhibition in the brain and spinal cord. In vitro assays have previously demonstrated that these receptors are heterodimers assembled from two homologous subunits, GBR1 and GBR2, neither of which is capable of producing functional GBR on their own. We have used co-immunoprecipitation in combination with bioluminescence and fluorescence resonance energy transfer approaches in living cells to assess directly the interaction between GBR subunits and determine their subcellular localization. The results show that, in addition to forming heterodimers, GBR1 and GBR2 can associate as stable homodimers. Confocal microscopy indicates that, while GBR1/GBR1 homodimers are retained in the endoplasmic reticulum and endoplasmic reticulum-Golgi intermediate compartment, both GBR2/GBR2 homodimers and GBR1/GBR2 heterodimers are present at the plasma membrane. Although these observations shed new light on the assembly of GBR complexes, they raise questions about the potential functional roles of GBR1 and GBR2 homodimers.

Figure 1. Immunoprecipitation of GBR1b and GBR2

(A) Immunoprecipitation studies: HEK-293 cells expressing Myc–GBR1b (lanes 3 and 4), HA–GBR2 (lanes 9 and 10) or both receptors (lanes 5, 6, 11 and 12) were lysed before (T, total expression) or after (S, cell-surface expression) the addition of anti-Myc (lanes 1–6) or anti-HA (lanes 7–12) antisera. The receptors were then immunoprecipitated (IP) by the addition of Protein G–Sepharose and the immune complexes resolved by SDS/PAGE. The Myc- or HA-tagged receptors were identified by immunoblotting (WB) with either anti-Myc (lanes 1–6) or anti-HA (lanes 7–12) antisera. (B) Co-immunoprecipitation studies. HEK-293 cells expressing Myc–GBR1b and HA–GBR1b (lanes 3 and 4), or Myc–GBR1b and HA–GBR2 receptors (lanes 5, 6, 9 and 10) were lysed and the receptors immunoprecipitated as described above using anti-Myc (lanes 1–6) or anti-HA (lanes 7–10) antisera. The receptors were revealed by immunoblotting with either anti-HA (lanes 1–6) or anti-Myc (lanes 7–10) antisera. The results shown are representative of at least three independent experiments.

Figure 2. Immunofluorescence analysis of the Rluc- and GFP10-fused GBR

HEK-293 cells were transiently transfected with Myc–GBR1b, Myc–GBR1b–Rluc or Myc–GBR1b–GFP10 alone (A) or in combination with HA–GBR2 (C) or with HA–GBR2, HA–GBR2–Rluc or HA–GBR2–GFP10 alone (B) or in combination with Myc–GBR1b (D). Cells were permeabilized (+) or not (−) 48 h after transfection, and receptors were detected using anti-Myc and/or anti-HA antibodies. Immunoreactivity was revealed using appropriate secondary antibodies coupled with Texas Red to monitor Myc–GBR1 and either Oregon Green or Alexa633 to monitor HA–GBR2. Labels in boldface indicate the visualized receptors. Insets show bright fields (Myc–GBR1 and Myc–GBR1-Luc) or direct fluorescence of GFP10 (Myc–GBR1–GFP10) to confirm the presence of cells in the selected field. Experiments using confocal microscopy were performed using a Leica DM IRBE microscope with an oil immersion ×100 objective. The results shown are representative of at least three independent experiments.

Figure 3. Functional analysis of the Rluc- and GFP10-fused GBRs

The ability of the GBRs, fused or not fused to Rluc or GFP10, to stimulate ERK1 and ERK2 phosphorylation was assessed in HEK-293 cells. GBR1b/GBR2, GBR1b–Rluc/GBR2–GFP10 or GBR1b–GFP10/GBR2–Rluc were transfected in HEK-293 cells and stimulated with 10 mM baclofen or 10% FBS for 5 min. Cell lysates were then resolved by SDS/PAGE and the ERK activity detected using anti-phospho ERK1/ERK2 antibodies. The total ERK loaded in each lane was controlled by using an anti-ERK1/ERK2 antibody. These results are representative of three independent experiments.

Figure 4. Interaction between GBR1b and GBR2 assessed by BRET in living cells

HEK-293 cells were transiently co-transfected with GBR1b–Rluc (left panel) or GBR2–Rluc (right panel) and either GBR1b–GFP10, GBR2–GFP10, CCR5–GFP10, hTfR–GFP10 or soluble GFP10. Luminescence, fluorescence and energy transfer (BRET) were measured 48 h after transfection. BRET ratios are expressed as the means±S.E.M. for three to four independent experiments performed in duplicate. The relative expression level of the Rluc and GFP10 constructs obtained under each condition is expressed as an average GFP/Rluc ratio and presented at the bottom of each bar.

Figure 5. Homo- and hetero-dimerization assessed by BRET titration in living cells

HEK-293 cells were transiently co-transfected with a constant amount of receptor-Rluc construct (GBR1b in A and GBR2 in B) and an increasing amount of DNA of either GBR1b–GFP10 or GBR2–GFP10. Total luminescence and fluorescence as well as energy transfer (BRET) were measured 48 h after transfection. BRET ratios are represented as a function of the GFP10/Rluc ratios used as an index for the relative concentration of receptor-GFP10 expressed. The curves were fitted using a non-linear regression equation assuming a single binding site (GraphPad Prism).

Figure 6. Interaction of GBR1a and GBR2 in cells assessed by FRET

(A) HEK-293 cells were transiently transfected with GBR1a–CFP, GBR2–CFP, GBR1a–YFP or GBR2–YFP and the receptor detected by confocal fluorescence microscopy. CFP-tagged constructs were detected after excitation (ex) at 450 nm and emission (em) at 488 nm, whereas YFP-tagged constructs were detected after excitation at 490 nm and emission at 550 nm. As a negative control for FRET, the GBR2–CFP construct expressed alone was excited at 450 nm and the emission detected at 550 nm. Emission was not detected at 550 nm when a YFP-tagged receptor was excited at 450 nm (results not shown). (B) HEK-293 cells were transiently co-transfected with GBR1a–CFP or GBR2–CFP and either GBR1a–YFP or GBR2–YFP. Each of the CFP- and YFP-tagged receptors were detected as in (A) and the FRET measured by exciting the CFP-tagged receptors at 450 nm and detecting the emission of the YFP-tagged receptors at 550 nm. Arrows point to cells where FRET signals originating from the cell surface can be distinguished.

Figure 7. Subcellular localization of GBR1b and GBR2

HEK-293 cells were transiently transfected with Myc–GBR1b or HA–GBR2 individually (A, C, E) or co-transfected with the two receptors (B, D, F). The subcellular localization of each receptor was determined by assessing its co-localization with specific markers of the ER (Calnexin; A, B), ERGIC (p58; C, D) and TGN (γ-AP1; E, F). Detection was performed using anti-Myc or anti-HA antibodies for the receptors and the anti-Calnexin, anti-p58 and anti-γ-AP1 antibodies for the organelle markers. Immunoreactivity was revealed using appropriate secondary antibodies coupled with Texas Red (red), Oregon Green (green) and Alexa633 (blue). The co-localization of a receptor with a subcellular compartment, results in the apparition of a third colour in the overlay image: yellow for red and green co-localization, turquoise for blue and green co-localization and fushia for red and blue co-localization. Confocal microscopy was performed using a Leica DM IRBE microscope with an oil immersion ×100 objective. Results shown are representative of at least three independent experiments.