Allosteric Inhibition of a Vesicular Glutamate Transporter by an Isoform-Specific Antibody

Abstract

The role of glutamate in excitatory neurotransmission depends on its transport into synaptic vesicles by the vesicular glutamate transporters (VGLUTs). The three VGLUT isoforms exhibit a complementary distribution in the nervous system, and the knockout of each produces severe, pleiotropic neurological effects. However, the available pharmacology lacks sensitivity and specificity, limiting the analysis of both transport mechanism and physiological role. To develop new molecular probes for the VGLUTs, we raised six mouse monoclonal antibodies to VGLUT2. All six bind to a structured region of VGLUT2, five to the luminal face, and one to the cytosolic. Two are specific to VGLUT2, whereas the other four bind to both VGLUT1 and 2; none detect VGLUT3. Antibody 8E11 recognizes an epitope spanning the three extracellular loops in the C-domain that explains the recognition of both VGLUT1 and 2 but not VGLUT3. 8E11 also inhibits both glutamate transport and the VGLUT-associated chloride conductance. Since the antibody binds outside the substrate recognition site, it acts allosterically to inhibit function, presumably by restricting conformational changes. The isoform specificity also shows that allosteric inhibition provides a mechanism to distinguish between closely related transporters.

Figure 1.

Novel monoclonal antibodies bind VGLUT2 from different sides of the membrane. HEK293T cells coexpressing pmVGLUT2 and EGFP were stained with VGLUT2 mAb hybridoma supernatants. (A) After fixation and permeabilization, all VGLUT2 mAbs label only transfected cells identified with EGFP. (B) Live HEK293T cells were incubated with hybridoma supernatants. 6D3, 8E11, 9C6, 12C1, and 14E1 but not 13H4 bind to intact cells expressing pmVGLUT2. 13H4 thus binds to the cytoplasmic face of VGLUT2. Scale bar, 20 μm. (C) Quantification of total VGLUT2 mAb fluorescence from EGFP+ cells for each of the conditions in parts A and B. n = 14–26 cells for each condition. Left panel: ***, p < 0.001; ****, p < 0.0001 compared to 14E1 by Kruskal–Wallis. Right panel: ****, p < 0.0001 relative to 13H4 by Kruskal–Wallis.

Figure 2.

VGLUT isoform specificity of novel VGLUT2 mAbs. Fixed, permeabilized HEK293T cells expressing pmVGLUT1 (A) or pmVGLUT3 (B) were incubated with VGLUT2 mAb hybridoma supernatants as in Figure 1A. (A) mAbs 6D3, 8E11, 9C6, and 12C1 bind to cells expressing pmVGLUT1, whereas 13H4 and 14E1 do not. (B) None of the mAbs bind to cells expressing pmVGLUT3. Scale bar = 20 μm. (C) Quantification of total VGLUT2 mAb fluorescence from EGFP+ cells for each of the conditions in panels A and B. n = 16–25 cells for each condition. ***, p < 0.001; ****, p < 0.0001 compared to either 13H4 or 14E1 by Kruskal–Wallis.

Figure 3.

Structure of VGLUT2–8E11 reveals the basis of isoform specificity. (A) Cryo-EM structure of rat VGLUT2 (rainbow) and Fab 8E11 (black) with the interface enlarged on the left and right (PDB: 6V4D). The sequence on the right is rotated 180° around an axis perpendicular to the membrane. (B) Alignment of ECL4–6 of VGLUT1–3 with conserved residues highlighted in yellow. VGLUT residues contacting 8E11 are indicated in bold and side-chain interactions in red. Glutamate 344 (boxed) is the only residue conserved in VGLUT1 and 2 but not VGLUT3. (C) Live cell feeding for pmVGLUT2, pmVGLUT2 E344A, pmVGLUT3, and pmVGLUT3 A348E confirms the requirement of 8E11 for the divergent ECL4 glutamate. Scale bar = 20 μm.

Figure 4.

8E11 inhibits glutamate transport by VGLUT2. (A) Assay for glutamate efflux. HEK293T cells are loaded with ³H-glutamate in the absence or presence of 8E11 and washed in cold Ringer’s solution before incubation in efflux buffer at either pH 7.4 or pH 5.5. (B) Efflux of ³H-glutamate from HEK293T cells expressing pmVGLUT2 + EGFP or EGFP alone. Efflux was measured with or without 8E11 at pH 7.4 and pH 5.5 (n = 5). Data indicate the mean ± SEM *** p < 0.001 compared to all other conditions, by two-way ANOVA.

Figure 5.

8E11 inhibits the chloride conductance associated with VGLUT2. (A) Representative traces of the currents from an oocyte expressing pmVGLUT2-HA, at either pH 7.4 or 5.0, without or with 8E11. (B) I–V curves of the currents from X. laevis oocytes expressing pmVGLUT2-HA (top) and bar graphs of the same currents at −120 mV (bottom). The currents were recorded at pH 7.4 and pH 5.5 before and after incubation with mAb 8E11 or an EAT-4 (IgG) antibody as a negative control. n = 7 oocytes for 8E11, n = 3 oocytes for EAT-4 mAb. Data indicate the mean ± SEM. Bottom, ***, p < 0.001 by two-way ANOVA compared to all other conditions (left) and **, p < 0.01 by two-way ANOVA compared to both pH 7.4 conditions (right).