Immunoisolation of GABA-specific synaptic vesicles defines a functionally distinct subset of synaptic vesicles

Abstract

Synaptic vesicles from mammalian brain are among the best characterized trafficking organelles. However, so far it has not been possible to characterize vesicle subpopulations that are specific for a given neurotransmitter. Taking advantage of the recent molecular characterization of vesicular neurotransmitter transporters, we have used an antibody specific for the vesicular GABA transporter (VGAT) to isolate GABA-specific synaptic vesicles. The isolated vesicles are of exceptional purity as judged by electron microscopy. Immunoblotting revealed that isolated vesicles contain most of the major synaptic vesicle proteins in addition to VGAT and are devoid of vesicular monoamine and acetylcholine transporters. The vesicles are 10-fold enriched in GABA uptake activity when compared with the starting vesicle fraction. Furthermore, glutamate uptake activity and glutamate-induced but not chloride-induced acidification are selectively lost during immunoisolation. We conclude that the population of GABA-containing synaptic vesicles is separable and distinct from vesicle populations transporting other neurotransmitters.

Fig. 1.

Characterization of antibodies specific for the VGAT. A, Antibodies specific for the N- (VGAT/1 and VGAT/3) or C-terminal (VGAT/4) domain recognize identical bands in synaptic vesicles and in cells expressing VGAT. tsA201 cells were transiently transfected with either a rVGAT plasmid (VGAT) or a plasmid without insert (pcDNA) and analyzed by immunoblotting. For comparison, purified synaptic vesicles (SV) were analyzed in parallel. All three antibodies recognized a doublet band (57 and 50 kDa) in both synaptic vesicles and transfected cells (filled arrowheads). Note that some degradation was observed in the heterologous expression system (open arrowheads). B, VGAT copurifies with other vesicular transporters and the vesicle protein synaptophysin during the isolation of synaptic vesicles. Synaptic vesicles were purified using established procedures, with the following fractions being analyzed:Homogenate; P1, crude nuclear pellet;P2, crude synaptosomes (10,000 × gpellet); S2, 10,000 × gsupernatant; LP1, 25,000 × g pellet obtained after synaptosomal lysis; LP2, crude synaptic vesicles; Peak 1 and SV, large membrane and purified synaptic vesicles as separated by controlled-pore glass (CPG) bead chromatography, respectively (for details, see Huttner et al., 1983; Hell and Jahn, 1994).

Fig. 2.

VGAT is present on a subset of synaptic vesicles. Immunogold labeling of synaptic vesicles purified through controlled-pore glass chromatography using anti-synaptophysin antibody (poly-clonal) (A), anti-VGAT antibody (B), and preimmune serum (C). D shows an overview of VGAT labeling at a lower magnification. Counting of labeled vesicles from several independent experiments revealed that ∼16% of all small vesicular profiles were labeled with anti-VGAT antibody, whereas synaptophysin labeling was observed on virtually all synaptic vesicles. Scale bars: A, D, 100 nm. Comparison of the staining patterns for VGAT (E,H) with those of synapto-brevin 2 (F) and GAD (I) in sections of rat cerebellum. The staining patterns of VGAT and GAD are identical. In contrast, synaptobrevin 2 antibody stains many more nerve terminals than VGAT; this is particularly obvious in the molecular cell layer (MO). High-magnification overlays of VGAT (red) and GAD (green) (J), and VGAT (red) and synaptobrevin 2 (green) (G) showing Purkinje cell bodies. GC, Granular cell layer;PC, Purkinje cell layer; MO, molecular cell layer. Scale bars: E, F,H, I, 50 μm; G,J, 10 μm.

Fig. 3.

VGAT immunobeads bind a highly homogenous organelle population with the size and shape of synaptic vesicles. A, VGAT beads; B, synaptophysin beads; C, control beads. D shows a representative field of the enriched vesicle fraction (LP2;arrows indicate small profiles of the size of synaptic vesicles) used as starting material for immunoisolation. See Materials and Methods for details. Scale bars, 500 nm.

Fig. 4.

Vesicles isolated on beads containing VGAT antibodies (VGAT Beads) are devoid of VAChT and VMAT2 but contain other synaptic vesicle proteins. The resuspended vesicle fraction (LP2) used as starting material (Input) was compared with the bead-bound material (Beads) and the material remaining in the supernatant after bead incubation (Sup). To allow for direct comparison, equal relative amounts of each fraction were separated in parallel.

Fig. 5.

VGAT immunoisolates are enriched for GABA uptake activity and impoverished for glutamate uptake activity. FCCP-sensitive GABA (A) and glutamate (B) uptake activities in LP2 (Input), unbound material (Sup), and bound material on VGAT beads (Beads) were measured using radiolabeled substrates and a standard filtration assay as described in Materials and Methods. All values were normalized to the relative amount of vesicle marker synaptophysin. The figures shows mean ± SD values of at least two independent experiments (in each experiment, n = 4).

Fig. 6.

Comparison of ATP-induced acidification using glutamate and chloride as counterion in VGAT and synaptophysin immunoisolates. Acidification was monitored by double-wavelength spectroscopy using acridine orange as indicator dye. The reaction was started by adding ATP. At the end of the reaction, 20 mm(NH₄)₂SO₄ was added to equalize the intravesicular pH with that of the medium (Hell et al., 1990). The bottom shows an immunoblot for the 116 kDa subunit of the vacuolar proton ATPase for each fraction. Note that, in this experiment, the recovery of V-ATPase was lower in the VGAT bead fraction than in the synaptophysin bead fraction, requiring higher amplification to obtain comparable signals.