The alpha2delta ligand gabapentin inhibits the Rab11-dependent recycling of the calcium channel subunit alpha2delta-2

Abstract

The α2δ subunits of voltage-gated calcium channels are important modulatory subunits that enhance calcium currents and may also have other roles in synaptogenesis. The antiepileptic and antiallodynic drug gabapentin (GBP) binds to the α2δ-1 and α2δ-2 isoforms of this protein, and its binding may disrupt the binding of an endogenous ligand, required for their correct function. We have shown previously that GBP produces a chronic inhibitory effect on calcium currents by causing a reduction in the total number of α2δ and α1 subunits at the cell surface. This action of GBP is likely to be attributable to a disruption of the trafficking of α2δ subunits, either to or from the plasma membrane. We studied the effect of GBP on the internalization of, and insertion into the plasma membrane of α2δ-2 using an α-bungarotoxin binding site-tagged α2δ-2 subunit, and a fluorescent derivative of α-bungarotoxin. We found that GBP specifically disrupts the insertion of α2δ-2 from post-Golgi compartments to the plasma membrane, and this effect was prevented by a mutation of α2δ-2 that abolishes its binding to GBP. The coexpression of the GDP-bound Rab11 S25N mutant prevented the GBP-induced decrease in α2δ-2 cell surface levels, both in cell lines and in primary neurons, and the GBP-induced reduction in calcium channel currents. In contrast, the internalization of α2δ-2 was unaffected by GBP. We conclude that GBP acts by preventing the recycling of α2δ-2 from Rab11-positive recycling endosomes to the plasma membrane.

Figure 1.

Insertion of a bbs tag in the VGCC α₂δ-2 subunit. A, Schematic representation of a Ca_v2.1 and a α₂δ-2 subunit, showing the position of the bbs tag (gray diamond) in α₂δ-2. B, tsA-201 cells were transfected with Ca_v2.1/α₂δ-2/β4 or Ca_v2.1/α₂δ-2 bbs/β4, and labeled with an antibody directed against α₂-2 (102–117 peptide) (red fluorescence; top row) and with αBgTx-AF488 (green; middle row). The merged images also showing nuclear staining with DAPI are shown in the bottom row. The images are 1-μm-thick optical sections. Scale bar, 20 μm. C, I–V relationships (left) and representative currents traces resulting from step potentials from −90 mV to between −25 and +5 mV, in 5 mV increments (right) for cells transfected with Ca_v2.1/β4/α₂δ-2 (■) (n = 7), or Ca_v2.1/β4/α₂δ-2 bbs in the absence (○) (n = 10) or in the presence of αBgTx-AF488 (▴) (n = 8). Error bars indicate SEM. D, Binding isotherm for [³H]GBP to cholesterol-rich microdomains fractions prepared from tsA-201 cells expressing either α₂δ-2 (■) (fitted to an hyperbola with K_D = 28.5 nm; n = 3) or α₂δ-2 bbs (○) (K_D = 34.9 nm; n = 3). The binding curves were normalized to each mean B_max.

Figure 2.

Chronic GBP reduces the cell surface expression of α₂δ-2 bbs. A, tsA-201 cells transfected with Ca_v2.1/β4/α₂δ-2 bbs or Ca_v2.1/β4/α₂δ-2 R282A bbs were incubated for 40 h with 100 μm or 1 mm GBP, or with addition of the equivalent volume of H₂O (control cells). Plasma membrane-expressed α₂δ-2 bbs was labeled at 17°C with αBgTx-AF488. Scale bar, 10 μm. B, Quantification of the fluorescence at the plasma membrane, measured within a region of interest covering an identical area across the plasma membrane of each cell present in the field of view, for WT α₂δ-2 bbs or for α₂δ-2 R282A bbs for control (white bars), 100 μm GBP (light gray bars), or 1 mm GBP (dark gray bars). The results are expressed as the mean ± SEM from three independent experiments (20–42 cells per condition per experiment). **p < 0.005, one way ANOVA and Bonferroni's post test. C, Primary cortical neurons expressing eCFP and α₂δ-2 bbs were incubated for 72 h with 1 mm GBP or the equivalent volume of H₂O (control cells). α₂δ-2 bbs subunits were labeled using αBgTx-AF555, in the presence of 100 μm tubocurarine, to remove any background caused by the binding of αBgTx to α7 nicotinic receptors. All cells analyzed were positive for CFP expression. The panels corresponding to α₂δ-2 bbs labeling correspond to a 1.2-μm-thick optical section, whereas the CFP panels show an average intensity projection of a z-stack containing three images. Scale bar, 20 μm. D, Quantification of αBgTx-AF555 labeling of neurites. The fluorescence density was measured in regions of interests corresponding to the neurites of individual neurons, in control condition (white bar) (n = 19) or after chronic incubation with GBP (gray bar) (n = 16). *p < 0.05, Student's unpaired t test.

Figure 3.

GBP does not affect the internalization rate of α₂δ-2 bbs. A, B, tsA-201 cells transfected with Ca_v2.1/β4/α₂δ-2 bbs (A) or Cav2.1/β4/α2δ-2 R282A bbs (B) were incubated with αBgTx-AF488 at 17°C for 30 min, washed to remove free αBgTx-AF488, and then incubated at 37°C to allow internalization of the subunit, in the absence or in the presence of 1 mm GBP, and imaged at the indicated times (5–60 min), as summarized in the protocol above the images. Vesicles containing α₂δ-2 bbs were seen intracellularly (indicated by the white arrow) after incubation at 37°C. Scale bar, 20 μm. C, D, Time course of the internalization of α₂δ-2 bbs (C) or α₂δ-2 R282A bbs (D), in the presence of 1 mm GBP (○) or control (■). The results are shown as the mean ± SEM of three independent experiments. In each experiment, 16–36 cells were analyzed per time point and condition.

Figure 4.

The rate of insertion of α₂δ-2 is not altered after 4 h in the presence of GBP. A, Diagram indicating the possible pathways (right) involved using the insertion assay (left). B, tsA-201 cells expressing Ca_v2.1/β4/α₂δ-2 bbs were preincubated for 4 h without (control) or with 1 mm GBP. The cells were then incubated at 17°C with unlabeled αBgTx for 30 min, and then at 37°C with αBgTx-AF488 for the indicated times, as summarized in the protocol in A. Scale bar, 20 μm. C, Quantification of the fluorescence appearing at the plasma membrane over time, in the presence of 1 mm GBP (○) or control (■). The results are expressed as the mean ± SEM of three independent experiments. In each experiment, 14–36 cells were analyzed per time point and condition.

Figure 5.

Effect of GBP on the forward trafficking of α₂δ-2 subunits. A, Diagram (left) indicating the possible pathways involved using the assay described (right). B, C, tsA-201 cells expressing Ca_v2.1/β4/α₂δ-2 bbs (B) or Ca_v2.1/β4/α₂δ-2 R282A bbs (C) were incubated for 4 h with 50 ng/ml BFA. One millimolar GBP or the equivalent volume of H₂O was subsequently added to the cells for 4 h. The cells were then incubated at 17°C with unlabeled αBgTx for 30 min, and then at 37°C with αBgTx-AF488 for the indicated times, as summarized in the protocol in A. Scale bar, 20 μm. The imaging settings were modified compared with Figure 4 to obtain a better visualization of the cells but were kept constant within the experiment. D, E, Time courses for the insertion of α₂δ-2 bbs (D) or α₂δ-2 R282A bbs (E) subunits at the plasma membrane, in the presence of 1 mm GBP (○) or control (■). The results are expressed as the mean ± SEM of three independent experiments. In each experiment, 14–41 cells were analyzed per time point and condition. *p < 0.05, two-way ANOVA and Bonferroni's post test. The data in D and E were fit with single exponentials, to estimate the rate of insertion under the different conditions. The time constant (τ) for the insertion of α₂δ-2 bbs at the plasma membrane was 27.6 min under control conditions and 17.6 min in the presence of GBP (D). In contrast, the calculated τ for α₂δ-2 R282A bbs insertion were similar in the absence (19.1 min) and in the presence (23.4 min) of GBP (E).

Figure 6.

GBP inhibits the trafficking α₂δ-2 through Rab11-associated endosomes. A, Diagram indicating the possible pathways involved using Rab11 S25N. B, Cells transfected with Ca_v2.1/β4/α₂δ-2 bbs and either a control cDNA (empty vector), WT Rab11, or Rab11 S25N were incubated with 1 mm GBP for 40 h. The α₂δ-2 subunits present at the plasma membrane were labeled with αBgTx-AF488 at 17°C. Scale bar, 20 μm. C, Diagram indicating the possible pathways involved using Rab4 S22N. D, Cells transfected with Ca_v2.1/β4/α₂δ-2 bbs and either a control cDNA (empty vector), WT Rab4, or Rab4 S22N were incubated with 1 mm GBP for 40 h. The α₂δ-2 subunits present at the plasma membrane were labeled with αBgTx-AF555 at 17°C. Scale bar, 20 μm. E, F, Quantification of the fluorescence at the plasma membrane in the absence (white bars) and in the presence of GBP (gray bars) for the three conditions, as indicated, in the presence of Rab11 (E) or Rab4 (F) constructs. Results are expressed as mean ± SEM of three independent experiments (23–65 cells per condition and per experiment). *p < 0.05, two-way ANOVA and Bonferroni's post test.

Figure 7.

The Rab11 pathway is involved in the reduction of calcium currents by chronic GBP. A, Representative currents traces resulting from step potentials from −90 mV to between −25 mV and +5 mV, in 5 mV increments for cells transfected with Ca_v2.1/β4/α₂δ-2/WT Rab11 in the absence (■) or the presence of chronic GBP (○), or Ca_v2.1/β4/α₂δ-2/Rab11 S25N in the absence (▵) or in the presence of chronic GBP (▾). B, I–V relationships for cells transfected with Ca_v2.1/β4/α₂δ-2/WT Rab11 under control conditions (■) (n = 13) or after chronic incubation with 1 mm GBP (○) (n = 10), or Ca_v2.1/β4/α₂δ-2/Rab11 S25N in the absence (▵) (n = 13) or in the presence (▾) (n = 12) of chronic GBP. C, Mean peak I_Ba calculated at +5 mV and normalized to the value of the control condition, in the absence (white bars) and in the presence (gray bars) of GBP. *p < 0.05, two-way ANOVA and Bonferroni's post test.

Figure 8.

Effect of GBP and Rab11 S25N on α₂δ-2 expression in primary cortical neurons. A, Cortical neurons were cultured for 5–6 d, and then transfected with α₂δ-2 HA and either GFP-Rab11 WT or GFP-Rab11 S25N, and incubated for 3 d in the absence or in the presence of 1 mm GBP for 72 h. Cell surface α₂δ-2 was detected by live labeling using an antibody directed against the HA epitope (column 2) before fixation and permeabilization of the cells and immunodetection of GFP-Rab11 constructs using an antibody directed against GFP (column 1), and MAP2, detected using a Cy5-conjugated secondary antibody (column 3). A merged imaged is shown in column 4. GFP-Rab11 S25N displays a diffuse localization corresponding to its lack of association with cellular membranes. Scale bar, 20 μm. B, Enlarged images of straightened neurite segments, localized 200–300 μm from the cell body, stained for cell surface-expressed α₂δ-2 (red) and for GFP-Rab11 (green). Scale bar, 5 μm. C, Quantification of the fluorescence density corresponding to the cell surface immunoreactivity for α₂δ-2 in neurites segments, in the absence (white bars) or in the presence (gray bar) of 1 mm GBP. The results are expressed as the mean ± SEM of 11–22 cells from four independent experiments. **p < 0.001, two-way ANOVA and Bonferroni's post test.