Synaptic vesicular localization and exocytosis of L-aspartate in excitatory nerve terminals: a quantitative immunogold analysis in rat hippocampus

Abstract

To elucidate the role of aspartate as a signal molecule in the brain, its localization and those of related amino acids were examined by light and electron microscopic quantitative immunocytochemistry using antibodies specifically recognizing the aldehyde-fixed amino acids. Rat hippocampal slices were incubated at physiological and depolarizing [K+] before glutaraldehyde fixation. At normal [K+], aspartate-like and glutamate-like immunoreactivities were colocalized in nerve terminals forming asymmetrical synapses on spines in stratum radiatum of CA1 and the inner molecular layer of fascia dentata (i.e., excitatory afferents from CA3 and hilus, respectively). During K+ depolarization there was a loss of aspartate and glutamate from these terminals. Simultaneously the immunoreactivities strongly increased in glial cells. These changes were Ca2+-dependent and tetanus toxin-sensitive and did not comprise taurine-like immunoreactivity. Adding glutamine at CSF concentration prevented the loss of aspartate and glutamate and revealed an enhancement of aspartate in the terminals at moderate depolarization. In hippocampi from animals perfused with glutaraldehyde during insulin-induced hypoglycemia (to combine a strong aspartate signal with good ultrastructure) aspartate was colocalized with glutamate in excitatory terminals in stratum radiatum of CA1. The synaptic vesicle-to-cytoplasmic matrix ratios of immunogold particle density were similar for aspartate and glutamate, significantly higher than those observed for glutamine or taurine. Similar results were obtained in normoglycemic animals, although the nerve terminal contents of aspartate were lower. The results indicate that aspartate can be concentrated in synaptic vesicles and subject to sustained exocytotic release from the same nerve endings that contain and release glutamate.

Fig. 1.

Test filters (processed together with the tissue sections presented in Fig. 5) illustrating the immunocytochemical specificity. The spots (0.1 μl) contained brain macromolecules conjugated by a glutaraldehyde/formaldehyde (g/p) mixture to amino acids l-Asp, l-Glu, l-Gln, and GABA, or only treated with glutaraldehyde/formaldehyde (“0”). The diluted 435 l-Asp and 607 l-Glu antisera were mixed with soluble g/p complexes of amino acids [l-asparagine (l-Asn), l-Glu, GABA, l-Asp, and l-Gln; 0.2 mmeach] as indicated before the sera were used on the test and tissue sections. Note that the l-Asp antiserum selectively stained the l-Asp spot, whereas the l-Glu antiserum selectively stained the l-Glu spot. The staining of test filters and tissue sections was abolished by preabsorption with g/p complexes (0.3 mm) of the amino acid used for immunization.

Fig. 2.

Test filters showing that the antisera do not react with synaptic vesicle proteins or other tissue macromolecules. The test spots (0.1 μl, 0.1 μg protein) contained macromolecules from synaptic vesicles, cytosol, and mitochondria, as well as amino acids conjugated to tissue macromolecules as in Figure 1. The filters with the spots were fixed with glutaraldehyde/formaldehyde, washed, and then exposed to the l-Asp, Glu, Gln, or Tau antibodies (final dilutions 1:300, 1:3000, 1:100, 1:3000, respectively). Note that for each antibody only the spot with the corresponding amino acid conjugate was labeled. The numbers indicate: 1, synaptic vesicles; 2, cytosol from synaptosomes; 3, nonsynaptosomal cytosol; 4, mitochondria from synaptosomes; 5, conjugated l-Asp; 6, conjugated Glu; 7, conjugated Gln; and 8, conjugated Tau.

Fig. 3.

Double-labeling quantitation ofl-Asp-LI and l-Glu-LI in different tissue compartments in slices incubated for 1 hr under control (3 mm K⁺) and depolarizing conditions (55 mm K⁺) at physiological (1.3 mm) or low Ca²⁺ concentrations. The low Ca²⁺ medium contained 0.1 mmCa²⁺ and 10 mm Mg²⁺or 1.0 mm EGTA. The values are mean numbers of gold particles per square micrometer ± SD in n tissue element profiles corrected for background labeling over empty resin (average 1.1 and 1.6 particles per square micrometer forl-Asp and l-Glu, respectively). The data shown here are from one animal; similar results were obtained in another animal in a separate experiment. Each profile was labeled with both thel-Asp and the l-Glu antiserum, respectively, on the same ultrathin section. The particle densities are arbitrary units (simultaneously processed test sections with known concentrations of Asp and Glu indicated a lower labeling efficiency for Glu, showing that the glutamate- and aspartate-labeling densities in excitatory nerve terminals at control conditions correspond to a concentration of ∼4 mm in the fixed tissue). At control conditions, the media with the low Ca²⁺-concentration did not alter the level of l-Asp or l-Glu immunoreactivity in any of the profiles (data not shown). Symbols used here and in Figure 6:asR, terminals forming asymmetrical synapses with spines in stratum radiatum of CA1 (spR) (see Fig. 5 for illustration of strata); asMi, terminals forming asymmetrical synapses with spines in the inner third of the molecular layer of fascia dentata (spMi); den, dendritic shafts in stratum radiatum of CA1 and the inner third of the molecular layer of fascia dentata; glia, glial cells (in CA1 and fascia dentata) identified by fine filaments or contact with capillaries. Asterisks and open triangles, Values in asR and asMi at 55 mmK⁺ (n = 25 and 20) were significantly different (p < 0.02) from the values at 55 mm K⁺ with low Ca²⁺–high Mg²⁺(n = 28 and 21) and low Ca²⁺–EGTA (n = 30 and 23) and from the values at 3 mm K⁺(n = 27 and 23). Filled circles andfilled triangles, Values in asR and asMi at 55 mm K⁺ when exocytosis was blocked were significantly different (p < 0.05)from the values at 3 mm K⁺.Open stars, Values in glia at 55 mmK⁺ (n = 13) were significantly different (p < 0.01) from the values at 55 mm K⁺ with low Ca²⁺–high Mg²⁺(n = 11) and low Ca²⁺–EGTA (n = 12) and from the values at 3 mmK⁺ (n = 10). Filled stars, Values at 55 mm K⁺ when exocytosis was blocked were significantly different (p < 0.02) from the values at control conditions. In spR, spMi, andden the numbers of profiles were between 12 and 30.

Fig. 4.

Electron micrographs showing gold particles signaling l-Asp in stratum radiatum CA1 of hippocampal slices incubated at 55 mm K⁺(A) and 55 mm K⁺with low Ca²⁺–high Mg²⁺(B) for 1 hr before fixation. In Anote that the terminal with asymmetrical junction (at) on a spine (s) is weakly labeled, whereas the similar type of terminal in B is strongly labeled. The glial profile (g) in A is strongly immunopositive for l-Asp. [The glial mitochondrion (m) is heavily labeled, other mitochondria are not]. In B also note the contrast in labeling between the terminals (at) and the spines (s). A, inset, Higher power photomicrograph of a part of the terminal (at) in A showing individual synaptic vesicles with diameters of ∼20–60 nm (arrowheads), which is in the same range as in this type of terminal in vivo (Harris and Sultan, 1995). Asterisks, Synaptic cleft. Scale bars: A, B 200 nm;inset 100 nm.

Fig. 5.

Light micrographs showing the effect of depolarization and TeTx on l-Asp-LI (A–C) and Glu-LI (D–F) in CA1 and fascia dentata of hippocampal slices. The slices were incubated as follows:A, D, 3 mmK⁺ for 3 hr; B, E, 55 mm K⁺ for 1 hr after preincubation for 2 hr at 3 mm K⁺; and C,F, 55 mm K⁺ (without TeTx) for 1 hr after preincubation for 2 hr at 3 mmK⁺ in the presence of TeTx. To stain preferentially the tissue accessible to TeTx, the slices were immunoreacted with the antibodies without resectioning and the surface photographed. Note that TeTx prevented the change from a nerve terminal-like staining pattern (fine dots) to a predominantly glial pattern (coarse processes) at depolarizing conditions.P, R, LM, Layers of hippocampus (pyramidale, radiatum, and lacunosum moleculare, respectively). Mo, Mm, Mi,G, Layers of area dentata (outer, middle moleculare, inner moleculare, and granulare, respectively).Arrowheads mark the obliterated fissura hippocampi.Open stars mark tears in the slice. Scale bar, 100 μm.

Fig. 6.

Double-labeling quantitation of amino acids in different tissue compartments in slices preincubated for 2 hr at 3 mm K⁺ with or without TeTx (Statens Seruminstitut) before incubation for 1 hr at 3 or 55 mmK⁺. The values are mean numbers of gold particles per square micrometer ± SD in n-tissue element profiles corrected for background density (particles per square micrometer) over empty resin (average 1.1, 1.6, and 1.8 for thel-Asp, the l-Glu, and the Tau antiserum, respectively). The data shown here are from one animal, similar results were obtained in two other animals in separate experiments. For symbols and other details, see Figure 3. The top two panels showl-Asp-LI andl-Glu-LI, respectively, determined by double labeling on the same ultrathin section (i.e., Asp and Glu were recorded in the same profiles). The bottom panel showsTau-LI determined by single labeling. Note that Tau-LI is unaffected by depolarization and TeTX. Asterisks andopen triangles, Values in asR andasMi at 55 mm K⁺(n = 22 and 19) were significantly different (p < 0.01) from the values at 3 mm K⁺ with (n = 23 and 20) and without (n = 29 and 21) TeTx and from the values at 55 mm K⁺ with TeTx (n = 25 and 19). Open circles andfilled triangles, Values in asR andasMi at 55 mm K⁺ with TeTx were significantly different (p < 0.01) from the values at 3 mm K⁺ ± TeTx. Open stars, Values in glia at 55 mmK⁺ (n = 9) were significantly different (p < 0.01) from the values at 3 mm K⁺ with (n = 8) and without (n = 12) TeTx and from the values at 55 mm K⁺ with TeTx (n = 10). Filled stars, Values in glia at 55 mmK⁺ with TeTx are significantly different (p < 0.05) from the values at 3 mm K⁺ ± TeTx.

Fig. 7.

l-Asp-LI (A,B) and l-Glu-LI (C,D) in neighboring ultrathin sections showing terminals (at1 and at2) forming synapses with asymmetrical specializations (white arrows) on spines (s) in stratum radiatum of CA1. The terminal (at1) in A and C was incubated at 55 mm K⁺ after preincubation in the absence of TeTx. The terminal (at2) in B and D was incubated at 55 mm K⁺ after preincubation in the presence of TeTx. Note that at1 is very weakly labeled with both antisera, whereas at2 is strongly l-Asp- andl-Glu-positive. Note that the spines have low levels ofl-Asp and l-Glu immunoreactivities in both experimental conditions. To allow access of TeTx, the terminals sampled here and in Figure 6 are from the superficial 30 μm of the slices. Scale bar, 0.2 μm.

Fig. 8.

Electron micrographs of l-Asp-LI andl-Glu-LI in hippocampus CA1 from a hypoglycemic rat subjected to perfusion fixation. A and Bshow accumulation of immunoreactivities over synaptic vesicle clusters (sv) versus over cytoplasmic matrix (cm) in terminals making asymmetrical synapses on spines (s). Broken lines mark the boundary between the vesicle-rich and vesicle-poor parts of the terminals. Scale bar, 0.2 μm.

Fig. 9.

l-Asp (Asp),l-Glu (Glu), glutamine (Gln), and taurine (Tau) labeling ratios between vesicle clusters (vesicles) and cytoplasmic matrix (cytopl. m.) in terminals forming asymmetrical synapses on spines in stratum radiatum CA1 of a rat made hypoglycemic before perfusion fixation (Fig. 8). The ratios were calculated by dividing the densities (particles per square micrometer immunogold particles over the vesicle clusters with the densities over cytoplasmic matrix and are presented as mean ± SD of n profiles) (n = 31, 21, 18, and 20 for Asp-LI, Glu-LI, Gln-LI, and Tau-LI, respectively). Asterisks, Ratios produced by the l-Asp and the l-Glu antisera were significantly different (p < 0.02) from the ratios produced by the Gln and Tau antisera. The ratios forl-Asp and l-Glu were not significantly different, neither were the ratios for Gln and Tau.

Fig. 10.

Neighboring ultrathin sections of the same tissue block as in Figure 8 showing colocalization of l-Asp (A) and l-Glu (B) immunogold particles in a terminal (at1) with asymmetrical synaptic specialization (arrows) on a spine (s). Note that the terminal has high levels of both immunoreactivities, whereas the spines and dendritic shafts (d) are very weakly labeled. Scale bar, 0.2 μm.

Fig. 11.

Scatter diagram of the densities ofl-Asp-LI and l-Glu-LI in the same nerve endings making asymmetrical junctions with spines in neighboring ultrathin sections from a hypoglycemic rat subjected to perfusion fixation (Fig.9). The values along the x- and y-axes are numbers of gold particles per square micromolar. Eachcircle represents the densities of l-Asp-LI and l-Glu-LI over an individual profile, corrected for background labeling over empty resin (1.0 and 1.7 gold particles per square micromolar, respectively). There was a significant positive correlation (r = 0.73, p < 0.01) between the density of l-Asp-LI and that ofl-Glu-LI.

Fig. 12.

Electron micrograph of normoglycemic perfusion-fixed tissue embedded in Lowicryl by freeze-substitution showing l-Asp-LI (A) andl-Glu-LI (B) in excitatory-type terminals (t) in st. radiatum of CA1. Immunogold particles with their centers inside the outer border of a synaptic vesicle profile, as analyzed in Figure 13, are marked byarrowheads. Broken lines indicate the outer borders of the terminals. s, Dendritic spines;m, mitochondrion. Scale bar, 0.2 μm.

Fig. 13.

l-Asp-LI and l-Glu-LI, but not Gln-LI are associated with synaptic vesicles in terminals forming asymmetrical synapses on spines in normoglycemic perfusion-fixed hippocampus CA1 st. radiatum. The columns show the ratio of l-Asp (n = 23),l-Glu (n = 12), and Gln (n = 12) labeling densities between synaptic vesicles (vesicles) and cytoplasmic matrix (cytopl. m.). Unlike in Figure 9, the immunodensities over synaptic vesicles are the number of gold particles with centers within individual synaptic vesicle profiles per total area of synaptic vesicle profile. The ratios were calculated for each terminal by dividing the l-Asp, l-Glu, and Gln densities over synaptic vesicles with the densities over cytoplasmic matrix. Asterisks, Similar l-Asp- andl-Glu-labeling ratios, statistically significantly higher than the Gln-labeling ratio (p < 0.0001).

Fig. 14.

l-Asp immunogold particles are more frequently located close to vesicle profiles than Gln immunogold particles. The distances of the centers of gold particles from the centers of synaptic vesicles were sorted into bins of 20 nm. Thecolumns show the frequencies of intercenter distances for each 20 nm bin. The total numbers of Asp and Gln immunogold particles were 139 and 271, respectively. Distances >140 nm (data not shown) make up altogether 4 and 7% for l-Asp and Gln immunogold particles, respectively. Asterisk, Within a distance of 20 nm from the center of the vesicle profile, the frequency of l-Asp immunogold particles is significantly higher than that of Gln immunogold particles (p < 0.0001, see Materials and Methods).