Acetylcholine, ATP, and Proteoglycan Are Common to Synaptic Vesicles Isolated from the Electric Organs of Electric Eel and Electric Catfish as Well as from Rat Diaphragm

Cholinergic synaptic vesicles were isolated from the electric organs of the electric eel (Electrophorus electricus) and the electric catfish (Malapterurus electricus) as well as from the diaphragm of the rat by density gradient centrifugation followed by column chromatography on Sephacryl-1000. This was verified by both biochemical and electron microscopic criteria. Differences in size between synaptic vesicles from the various tissue sources were reflected by their elution pattern from the Sephacryl column. Specific activities of acetylcholine (ACh; in nmol/mg of protein) of chromatography-purified vesicle fractions were 36 (electric eel), 2 (electric catfish), and 1 (rat diaphragm). Synaptic vesicles from all three sources contained ATP in addition to ACh (molar ratios of ACh/ATP, 9-12) as well as binding activity for an antibody raised against Torpedo cholinergic synaptic vesicle proteoglycan. Synaptic vesicles from rat diaphragm contained binding activity for the monoclonal antibody asv 48 raised against a rat brain 65-kilodalton synaptic vesicle protein. Antibody asv 48 binding was absent from electric eel and electric catfish synaptic vesicles. These antibody binding results, which were obtained by a dot blot assay on isolated vesicles, directly correspond to the immunocytochemical results demonstrating fluorescein isothiocyanate staining in the respective nerve terminals. Our results imply that ACh, ATP, and proteoglycan are common molecular constituents of motor nerve terminal-derived synaptic vesicles from Torpedo to rat. In addition to ACh, both ATP and proteoglycan may play a specific role in the process of cholinergic signal transmission.     

Characterization of a membrane protein from cholinergic synaptic vesicles isolated from the electric organ of Torpedo marmorata

Rabbits were immunized with cholinergic synaptic vesicles isolated from the electric organ of Torpedo marmorata. The resultant antiserum had one major antibody activity against an antigen called the Torpedo vesicle antigen. This antigen could not be demonstrated in muscle, liver or blood and is therefore, suggested to be nervous-tissue specific. The vesicle antigen was quantified in various parts of the nervous system and in subcellular fractions of the electric organ of Torpedo marmorata and was found to be highly enriched in synaptic vesicle membranes. The antigen bound to concanavalin A, thereby demonstrating the presence of a carbohydrate moiety. By means of charge-shift electrophoresis, amphiphilicity was demonstrated, indicating that the Torpedo vesicle antigen is an intrinsic membrane protein. The antigen was immunochemically unrelated to other brain specific proteins such as 14-3-2, S-100, the glial fibrillary acidic protein and synaptin. Furthermore, it was unrelated to two other membrane proteins, the nicotinic acetylcholine receptor and acetylcholinesterase, present in Torpedo electric organ. The antiserum against Torpedo synaptic vesicles did not react with preparations of rat brain synaptic vesicles or ox adrenal medullary chromaffin granules.     

Purification and Characterization of a Nonvesicular Vesamicol-Binding Protein from Electric Organ and Demonstration of a Related Protein in Mammalian Brain

A protein that binds vesamicol has been purified from a soluble fraction of the Torpedo electric organ homogenate that does not contain synaptic vesicles. The purified vesamicol-binding protein (VBP) has a molecular mass of 470 kDa composed of 30- and 24-kDa subunits. Chemical deglycosylation yielded a single, heterogeneous protein of 24 kDa. The 30-kDa subunit is also sensitive to endo-beta-galactosidase. The dissociation constant of the VBP.vesamicol complex is 0.9 microM, and the Bmax is 5,500 pmol/mg. Antiserum raised to the 30-kDa subunit cross-reacts with the 24-kDa subunit, but not with synaptic vesicles. Drug binding studies and Western blot analysis show that VBP is present in other Torpedo tissues as well as mammalian brain. Immunofluorescence microscopy demonstrates that VBP-like immunoreactivity is not localized exclusively to the nerve terminal regions of the electric organ. Thermal stability, the pH dependence of vesamicol binding, and pharmacological comparisons demonstrate that the VBP is not the cholinergic synaptic vesicle receptor for vesamicol. The implications of this finding for current efforts to develop in vivo diagnostics of cholinergic nerve terminal status based on vesamicol are discussed.     

Identification and characterization of SV31, a novel synaptic vesicle membrane protein and potential transporter

Synaptic vesicle proteins govern all relevant functions of the synaptic vesicle life cycle, including vesicle biogenesis, vesicle transport, uptake and storage of neurotransmitters, and regulated endocytosis and exocytosis. In spite of impressive progress made in the past years, not all known vesicular functions can be assigned to defined protein components, suggesting that the repertoire of synaptic vesicle proteins is still incomplete. We have identified and characterized a novel synaptic vesicle membrane protein of 31 kDa with six putative transmembrane helices that, according to its membrane topology and phylogenetic relation, may function as a vesicular transporter. The vesicular allocation is demonstrated by subcellular fractionation, heterologous expression, immunocytochemical analysis of brain sections and immunoelectron microscopy. The protein is expressed in select brain regions and contained in subpopulations of nerve terminals that immunostain for the vesicular glutamate transporter 1 and the vesicular GABA transporter VGaT (vesicular amino acid transporter) and may attribute specific and as yet undiscovered functions to subsets of glutamatergic and GABAergic synapses.     

Analysis of the synaptic vesicle proteome using three gel-based protein separation techniques

Synaptic vesicles are key organelles in neurotransmission. Their functions are governed by a unique set of integral and peripherally associated proteins. To obtain a complete protein inventory, we immunoisolated synaptic vesicles from rat brain to high purity and performed a gel-based analysis of the synaptic vesicle proteome. Since the high hydrophobicity of integral membrane proteins hampers their resolution by gel electrophoretic techniques, we applied in parallel three different gel electrophoretic methods for protein separation prior to MS. Synaptic vesicle proteins were subjected to either 1-D SDS-PAGE along with nano-LC ESI-MS/MS or to the 2-D gel electrophoretic techniques benzyldimethyl-n-hexadecylammonium chloride (BAC)/SDS-PAGE, and double SDS (dSDS)-PAGE in combination with MALDI-TOF-MS. We demonstrate that the combination of all three methods provides a comprehensive survey of the proteinaceous inventory of the synaptic vesicle membrane compartment. The identified synaptic vesicle proteins include transporters, soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs), synapsins, rab and rab-interacting proteins, additional guanine nucleotide triphosphate (GTP) binding proteins, cytoskeletal proteins, and proteins modulating synaptic vesicle exo- and endocytosis. In addition, we identified novel proteins of unknown function. Our results demonstrate that the parallel application of three different gel-based approaches in combination with mass spectrometry permits a comprehensive analysis of the synaptic vesicle proteome that is considerably more complex than previously anticipated.     

Cholinergic synaptic vesicles from Torpedo marmorata contain an atractyloside-binding protein related to the mitochondrial ADP/ATP carrier

Atractyloside is known to bind to the ADP/ATP translocase of the inner mitochondrial membrane, a complex formed by two basic protein subunits of relative molecular mass around 30 000. We found that synaptic vesicles from the electric organ of Torpedo marmorata, which store acetylcholine and ATP, bind atractyloside as well. Similarly to mitochondria, a protein-atractyloside complex could be solubilized from vesicle membranes with Triton X-100. Characterization of the complex by gel filtration, isoelectric focusing and gel electrophoresis revealed that atractyloside was bound to protein V11, earlier described as a major vesicle membrane component with a relative molecular mass around 34 000 and a basic isoelectric point. Since earlier experiments have already shown that uptake of ATP into isolated vesicles in vitro is inhibited by atractyloside, we can conclude now that V11 constitutes the nucleotide carrier of this secretory organelle. The structural and functional relationship of the mitochondrial and vesicular nucleotide translocases suggest a common evolutionary origin.     

Immunopurification of a Sarcomeric Junctional Protein Complex Containing GAPDH

We have isolated a monoclonal antibody, P4B2, which localizes to multiple anchorage junctions, namely, a subset of focal adhesions, the Z-disk of muscle, and neuromuscular junctions. Immunopurification of the antigen to this antibody from chicken brain tissue yielded a complex of three prominent proteins with mobilities of 36, 30, and 18 kDa. Amino acid sequencing of the purified proteins identified the 36-kDa protein as glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The other two protein bands were heterogeneous, containing proteins found in the synaptic vesicle fusion core complex. Immunolocalization of P4B2 antigen in developing cultured muscle cells showed that the antigen is incorporated into Z-lines soon after the sarcomeric architecture was positive for α-actinin. Together, the data indicate the P4B2 antigen is part of a unique GAPDH-containing protein complex that may be involved in reinforcement of established cytoskeletal structures.     

Biogenesis of synaptic vesicles in vitro

Synaptic vesicles are synthesized at a rapid rate in nerve terminals to compensate for their rapid loss during neurotransmitter release. Their biogenesis involves endocytosis of synaptic vesicle membrane proteins from the plasma membrane and requires two steps, the segregation of synaptic vesicle membrane proteins from other cellular proteins, and the packaging of those unique proteins into vesicles of the correct size. By labeling an epitope-tagged variant of a synaptic vesicle protein, VAMP (synaptobrevin), at the cell surface of the neuroendocrine cell line PC12, synaptic vesicle biogenesis could be followed with considerable precision, quantitatively and kinetically. Epitope-tagged VAMP was recovered in synaptic vesicles within a few minutes of leaving the cell surface. More efficient targeting was obtained by using the VAMP mutant, del 61-70. Synaptic vesicles did not form at 15 degrees C although endocytosis still occurred. Synaptic vesicles could be generated in vitro from a homogenate of cells labeled at 15 degrees C. The newly formed vesicles are identical to those formed in vivo in their sedimentation characteristics, the presence of the synaptic vesicle protein synaptophysin, and the absence of detectable transferrin receptor. Brain, but not fibroblast cytosol, allows vesicles of the correct size to form. Vesicle formation is time and temperature-dependent, requires ATP, is calcium independent, and is inhibited by GTP-gamma S. Thus, two key steps in synaptic vesicle biogenesis have been reconstituted in vitro, allowing direct analysis of the proteins involved.     

The Synaptic Vesicle-Associated G Protein o-rab3 Is Expressed in Subpopulations of Neurons

Abstract: The distribution of o-rab3—a synaptic vesicle-associated low-molecular-weight GTP-binding protein—was studied in various neural tissues of the electric ray Torpedo marmorata. o-rab3 was shown to be associated selectively with isolated cholinergic synaptic vesicles derived from the electric organ. Gel filtration of cholinergic synaptic vesicles using Sephacryl S-1000 column chromatography demonstrated a copurification of o-rab3 with the synaptic vesicle content marker ATP and with SV2—a synaptic vesicle transmembrane glycoprotein. Indirect immunofluorescence using antibodies against o-rab3 and SV2 and a double labeling protocol revealed an identical distribution of both antigens in the cholinergic nerve terminals within the electric organ and at neuromuscular junctions. An immunoelectron microscopic analysis demonstrated the presence of o-rab3 at the surface of the synaptic vesicle membrane. In the CNS immunofluorescence of o-rab3 and SV2 overlap only in small and distinct areas. Whereas SV2 has an overall distribution in nerve terminals of the entire CNS, o-rab3 is restricted to a subpopulation of nerve terminals in the dorsolateral neuropile of the rhombencephalon and in the dorsal horn of the spinal cord. Our results demonstrate that the synaptic vesicle-associated G protein o-rab3 is specifically expressed only in subpopulations of neurons in the Torpedo CNS.     

Phosphorylation of Brain Synaptic and Coated Vesicle Proteins by Endogenous Ca2+/Calmodulin- and cAMP-Dependent Protein Kinases

Phosphorylation of brain synaptic and coated vesicle proteins was stimulated by Ca2+ and calmodulin. As determined by 5-15% sodium dodecylsulfate (SDS) polyacrylamide gel electrophoresis (PAGE), molecular weights (Mr) of the major phosphorylated proteins were 55,000 and 53,000 in synaptic vesicles and 175,000 and 55,000 in coated vesicles. In synaptic vesicles, phosphorylation was inhibited by affinity-purified antibodies raised against a 30,000 Mr protein doublet endogenous to synaptic and coated vesicles. When this doublet, along with clathrin, was extracted from coated vesicles, phosphorylation did not take place, implying that the protein doublet may be closely associated with Ca2+/calmodulin-dependent protein kinase. Affinity-purified antibodies, raised against clathrin used as a control antibody, failed to inhibit Ca2+/calmodulin-dependent phosphorylation in either synaptic or coated vesicles. Immunoelectron cytochemistry revealed that this protein doublet was present in axon terminal synaptic and coated vesicles. Synaptic vesicles also displayed cAMP-dependent kinase activity; coated vesicles did not. The molecular weights of phosphorylated synaptic vesicle proteins in the presence of Mg2+ and cAMP were: 175,000, 100,000, 80,000, 57,000, 55,000, 53,000, 40,000, and 30,000. Based on the different phosphorylation patterns observed in synaptic and coated vesicles, we propose that brain vesicle protein kinase activities may be involved in the regulation of exocytosis and in retrieval of synaptic membrane in presynaptic axon terminals.     

Fusion of phospholipid vesicles induced by muscle glyceraldehyde-3-phosphate dehydrogenase in the absence of calcium

Ca2+-induced fusion of phospholipid vesicles (phosphatidylcholine/phosphatidic acid, 9:1 mol/mol) prepared by ethanolic injection was followed by five different procedures: resonance energy transfer, light scattering, electron microscopy, intermixing of aqueous content, and gel filtration through Sepharose 4-B. The five methods gave concordant results, showing that vesicles containing only 10% phosphatidic acid can be induced to fuse by millimolar concentrations of Ca2+. When the fusing capability of several soluble proteins was assayed, it was found that concanavalin A, bovine serum albumin, ribonuclease, and protease were inactive. On the other hand, lysozyme, L-lactic dehydrogenase, and muscle and yeast glyceraldehyde-3-phosphate dehydrogenase were capable of inducing vesicle fusion. Glyceraldehyde-3-phosphate dehydrogenase from rabbit muscle, the most extensively studied protein, proved to be very effective: 0.1 microM was enough to induce complete intermixing of bilayer phospholipid vesicles. Under conditions used in this work, fusion was accompanied by leakage of internal contents. The fusing capability of glyceraldehyde-3-phosphate dehydrogenase was not affected by 5 mM ethylenediaminetetraacetic acid. The Ca2+ concentration in the medium, as determined by atomic absorption spectroscopy, was 5 ppm. Heat-denatured enzyme was incapable of inducing fusion. We conclude that glyceraldehyde-3-phosphate dehydrogenase is a soluble protein inherently endowed with the capability of fusing phospholipid vesicles.     

Boar acrosin is a two-chain molecule. Isolation and primary structure of the light chain; homology with the pro-part of other serine proteinases

Cholinergic synaptic vesicles from the electric organ of Torpedo marmorata are associated with a Mg2+-ATPase insensitive to ouabain and oligomycin. Treatment of vesicle membranes with dichloromethane releases a Mg2+-ATPase with apparent molecular mass of around 250 kDa as determined by gel filtration. The vesicular ATPase resembles the mitochondrial F1-ATPase in these properties. Gel electrophoresis of the solubilized ATPase shows however that only a single 50-kDa band is present as compared to the alpha-subunit (52 kDa) and beta-subunit (50 kDa) of electric organ mitochondrial F1-ATPase present in this range of molecular mass range. In agreement, covalent photoaffinity labelling of isolated vesicles with azido-ATP shows a 50-kDa band. Vesicle ghosts were found to accumulate [14C]methylamine in an ATP-dependent manner indicating the presence of an inwardly directed proton pump. We conclude that cholinergic vesicles contain a proton pump probably driven by the Mg2+-ATPase here described, which generates an electrochemical gradient across the vesicle membrane and is necessary for uptake and storage of acetylcholine within the vesicles.     

Identification of the 37-kDa protein displaying a variable interaction with the erythroid cell membrane as glyceraldehyde-3-phosphate dehydrogenase

In previous studies from this laboratory we isolated and characterized a 37-kDa protein that was associated with the membrane of erythroid cells. The polypeptide appeared to undergo a lineage-specific alteration in its interaction with the membrane during erythroid development and migrated as a family of isoelectric focusing variants during analyses on two-dimensional gels. We report here that the 37-kDa protein is homologous to the enzyme glyceraldehyde-3-phosphate dehydrogenase (EC 1.2.1.12). This conclusion was reached from the results of several experimental approaches comparing the biochemical and genetic properties of the 37-kDa protein (p37) with authentic glyceraldehyde-3-phosphate dehydrogenase. Peptide maps of highly purified p37 and glyceraldehyde-3-phosphate dehydrogenase, generated with Staphylococcus V8 protease, were identical. The nucleotide sequence of a cDNA clone encoding p37 was nearly identical to the published sequence for genes encoding glyceraldehyde-3-phosphate dehydrogenase. These results suggest that the interaction of the enzyme with the red cell membrane is more complex than previously envisioned. The existence of subpopulations of glyceraldehyde-3-phosphate dehydrogenase molecules is envisioned that exhibit different levels of enzyme activity and bind to the red cell membrane with varying affinities.     

Uncoupling of Cholinergic Synaptic Vesicles by the Presynaptic Toxin β-Bungarotoxin

The effect of the presynaptic neurotoxin beta-bungarotoxin (beta-BuTx) on the acetylcholine (ACh) storage system of synaptic vesicles isolated from the electric organ of Torpedo californica was studied. The toxin can totally inhibit active transport of [3H]ACh by the vesicles in a Ca2+-, time-, and concentration-dependent manner. Correlated with these effects is a 50-60% stimulation of the vesicle proton-pumping ATPase activity. The beta-BuTx-mediated transport inhibition and ATPase stimulation are antagonized by delipidated bovine serum albumin, not reversed by excess EGTA, and not mimicked by other cationic proteins or soybean or pancreatic trypsin inhibitors. The behavior is consistent with phospholipase A2 (PLA2)-dependent damage to the vesicle membrane caused by beta-BuTx, which results in uncoupling of the ATPase and ACh transporter systems. The nonneurotoxic Naja naja venom PLA2 causes similar effects, except that it is slightly more potent on a molar basis. About 100-fold more beta-BuTx is required to effect lysis of synaptic vesicles than to uncouple them. ATP is a strong inhibitor of beta-BuTx- but not of N. naja PLA2-mediated uncoupling. The observations suggest that a component of beta-BuTx toxicity in the cholinergic terminal might involve attack on synaptic vesicles or vesicle-like structures and that a nucleotide-like factor might modulate the toxicity.     

G-proteins in Torpedo marmorata electric organ. Differential distribution in pre- and post-synaptic membranes and synaptic vesicles

The nature of the G-proteins present in the pre- and post-synaptic plasma membranes and in the synaptic vesicles of cholinergic nerve terminals purified from the Torpedo electric organ was investigated. In pre- and post-synaptic plasma membranes, Bordetella pertussis toxin, known to catalyze the ADP-ribosylation of the alpha-subunit of several G-proteins, labels two substrates at 41 and 39 kDa. The 39 kDa subunit detected by ADP-ribosylation in the synaptic plasma membrane fractions was immunologically similar to the Go alpha-subunit purified from calf brain. In contrast to bovine chromaffin cell granules, no G-protein could be detected in Torpedo synaptic vesicles either by ADP-ribosylation or by immunoblotting.     

Synapsin I Is Associated with Cholinergic Nerve Terminals in the Electric Organs of Torpedo, Electrophorus, and Malapterurus and Copurifies with Torpedo Synaptic Vesicles

Using an affinity-purified monospecific polyclonal antibody against bovine brain synapsin I, the distribution of antigenically related proteins was investigated in the electric organs of the three strongly electric fish Torpedo marmorata, Electrophorus electricus, Malapterurus electricus and in the rat diaphragm. On application of indirect fluorescein isothiocyanate-immunofluorescence and using alpha-bungarotoxin for identification of synaptic sites, intense and very selective staining of nerve terminals was found in all of these tissues. Immunotransfer blots of tissue homogenates revealed specific bands whose molecular weights are similar to those of synapsin Ia and synapsin Ib. Moreover, synapsin I-like proteins are still attached to the synaptic vesicles that were isolated in isotonic glycine solution from Torpedo electric organ by density gradient centrifugation and chromatography on Sephacryl-1000. Our results suggest that synapsin I-like proteins are also associated with cholinergic synaptic vesicles of electric organs and that the electric organ may be an ideal source for studying further the functional and molecular properties of synapsin.     

Identification of the ATP-binding heat-inducible protein of MR = 37,000 as glyceraldehyde-3-phosphate dehydrogenase

We previously found a novel ATP-binding heat-inducible protein of Mr = 37,000 in BALB/c 3T3 cells. Here, we found that the peptide mapping of this 37-kDa protein was similar to that of rabbit glyceraldehyde-3-phosphate dehydrogenase. Therefore, we biochemically compared the 37-kDa protein with a product translated from mRNA which was hybrid-selected using a cDNA for encoding chick glyceraldehyde-3-phosphate dehydrogenase and found that these two proteins were very similar. Northern blotting analysis using its cDNA as a probe revealed that glyceraldehyde-3-phosphate dehydrogenase was a heat-inducible protein in BALB/c 3T3 cells and that it was induced by stresses including treatment with alpha, alpha'-dipyridyl.     

Purification of synaptic vesicles from elasmobranch electric organ and the use of biophysical criteria to demonstrate purity.

We have purified cholinergic synaptic vesicles from the electric organs of two related marine elasmobranchs, Torpedo californica and Narcine brasiliensis, to a specific activity higher than had previously been obtained. We have demonstrated the homogeneity of the vesicles by biophysical criteria. The purification scheme consisted of differential centrifugation, flotation equilibrium in sucrose density gradients, and permeation chromatography on glass bead columns of average pore size 3000 A. Our criteria for purity were that bound acetylcholine, bound nucleotide triphosphate, protein, and lipid--phosphorus behave identically when vesicles were analyzed by procedures which depend on vesicle size, density, and charge. Contaminants were not detected when vesicles were fractionated by preparative and analytical sedimentation, by preparative equilibrium sedimentation using glycerol density gradients, or by electrophoresis in Ficoll density gradients. Pure synaptic vesicles, which have been purified 290-fold from the initial homogenate, contain per mg of protein: 8 mumol of acetylcholine, 3 mumol of ATP, and 7 mumol of lipid phosphorus. These procedures may be of general value in the purification of membrane vesicles.     

An ATP-modulated specific association of glyceraldehyde-3-phosphate dehydrogenase with human erythrocyte glucose transporter

Glyceraldehyde-3-phosphate dehydrogenase was found to bind in vitro to purified, human erythrocyte glucose transporter reconstituted into vesicles. Mild tryptic digestion of the glucose transporter totally inactivated the binding, suggesting that the cytoplasmic domain of the transporter is involved in the binding to glyceraldehyde-3-phosphate dehydrogenase. The binding was abolished in the presence of antisera raised against the purified glucose transporter, further supporting specificity of this interaction. The binding was reversible with a dissociation constant (Kd) of 3.3 x 10(-6) M and a total capacity (Bt) of approximately 30 nmol/mg of protein indicating a stoichiometry of one enzyme-tetramer per accessible transporter. The binding was sensitive to changes in pH showing an optimum at around pH 7.0. KCl and NaCl inhibited the binding in a simple dose-dependent manner with Ki of 40 and 20 mM, respectively. The binding was also inhibited by NAD+ with an estimated Ki of 3 mM. ATP, on the other hand, enhanced the binding by up to 3-fold in a dose-dependent manner with an apparent Ka of approximately 6 mM. The binding was not affected by D-glucose or cytochalasin B. The binding did not affect either the glucose or cytochalasin B in binding affinities or the transport activity of the transporter. However, the enzyme was inactivated totally upon binding to the transporter. Based on these findings, we suggest that a significant portion of glyceraldehyde-3-phosphate dehydrogenase in human erythrocytes exists as an inactive form via an ATP-dependent, reversible association with glucose transporter, and that this association may exert regulatory intervention on nucleotide metabolism in vitro.     

Adenosine triphosphate. A constituent of cholinergic synaptic vesicles

1. Synaptic vesicles separated by density-gradient centrifugation from extracts of the cholinergic nerve terminals of the electric organ of Torpedo marmorata were found to contain appreciable amounts of ATP as well as acetylcholine. 2. Vesicular ATP was stable in the presence of concentrations of apyrase and myokinase that rapidly destroyed equivalent amounts of endogenous or added free ATP; pre-treatment of cytoplasmic extracts of electric tissue with these enzymes destroyed endogenous free ATP, but did not affect the vesicular ATP. 3. When [U-(14)C]ATP was added to electric tissue at the time of comminution and extraction of the vesicles, all the radioactivity was associated with soluble components in the subsequent fractionation: none was associated with vesicles or membrane fragments; thus it is unlikely that vesicular ATP can be accounted for by the sequestration of endogenous free ATP within any vesicles formed during comminution and extraction of the tissue. 4. When synaptic vesicles were passed through iso-osmotic columns of Bio-Gel A-5m, which separates vesicles from soluble proteins and small molecules, all the recovered ATP and acetylcholine passed through together in the void volume. 5. Regression analysis showed that vesicular ATP content was highly correlated with vesicular acetylcholine content in different experiments, the molar ratio acetylcholine/ATP being 5.32+/-(s.e.m.) 0.45 (21 expts.) for the peak density-gradient fraction. The ratio varied, however, somewhat across the density-gradient peak suggesting some degree of chemical heterogeneity in the vesicle population.