Differences in the osmotic fragility of recycling and reserve synaptic vesicles from the cholinergic electromotor nerve terminals of Torpedo and their possible significance for vesicle recycling.

In this study we demonstrate differences in the osmotic fragility of two metabolically and physically heterogeneous synaptic vesicle populations from stimulated electromotor nerve terminals. When synaptic vesicles isolated on sucrose density gradients are submitted to solutions of decreasing osmolarity 50% of VP2-type vesicles lysed at (mean + S.E. (number of experiments] 332 +/- 14 (4) mosM and 50% of VP1-type vesicles lysed at 573 +/- 8 (3) mosM. These results indicate that recycling vesicles are more resistant to hypo-osmotic lysis and they are consistent with our earlier conclusion that changes in water content on recycling are secondary to changes in the content of the osmotically active small-molecular-mass constituents acetylcholine and ATP.

High-affinity, sodium-gradient-dependent transport of choline into vesiculated presynaptic plasma membrane fragments from the electric organ of Torpedo marmorata and reconstitution of the solubilized transporter into liposomes.

Vesiculated fragments of presynaptic plasma membranes have been isolated from the purely cholinergic electromotor nerve terminals of Torpedo marmorata. Synaptosomes, generated from the terminals by homogenization, were separated on a discontinuous Ficoll gradient and then lysed by osmotic shock at 2 degrees C, pH 8.5 in the presence of 0.1 mM MgCl2. These conditions for lysis were optimal for choline transport. Electron micrographs of lysed synaptosomes showed vesiculated membranes with diameters smaller than those of synaptosomes; occasionally, synaptic vesicles were observed attached to them. Intact mitochondria or synaptosomes and basal laminae were not present. High-affinity (KT = 1.7 microM) uptake of choline into these vesiculated membrane fragments showed: an absolute dependence on the Na+ gradient (outside greater than inside), a transient Na+-gradient-dependent accumulation of choline over the equilibrium concentration (over-shoot), electrogenicity and rheogenicity, since the uptake was further stimulated in the presence of a Na+ gradient by valinomycin, dependence on the presence of external Cl-, and partial dependence on a Cl- gradient (outside greater than inside), high-affinity (Ki = 25 nM) inhibition by hemicholinium-3 and temperature sensitivity. The plasma membranes were further purified by centrifugal density gradient fractionation on a 4-12% Ficoll gradient. Several enzymes and polypeptides copurified with the specific binding sites for choline present in the membranes. The fraction with the most binding sites was one denser than 12% Ficoll. This was also the fraction richest in acetylcholinesterase, 5'-nucleotidase and polypeptides of relative molecular mass, Mr (X 10(-3)) of greater than 200, 140, 68 (doublet), 57, 54 and 28. Acetylcholinesterase was positively identified as a Mr 68 000 component by immune blot. By contrast the ouabain-sensitive ATPase showed a negative correlation with choline binding sites. When the solubilized proteins of the vesiculated membranes were transferred to liposomes, they conferred on the latter the capacity to take up choline in a manner closely resembling its transport in natural membranes but with an initial (one minute) rate of uptake approximately 10-times greater per mg of protein. Several proteins were selectively transferred to the liposomes including ones of Mr (X 10(-3)) 34, 42, 47, 54, 60, 68, 92, 160 and greater than 200. The polypeptides of Mr (X 10(-3)) 140, 57 and 28 were lost in the transfer.(ABSTRACT TRUNCATED AT 400 WORDS)

The density and free water of cholinergic synaptic vesicles as a function of osmotic pressure.

Synaptic vesicles from the cholinergic electromotor nerve terminals of Torpedo marmorata are among the most uniform subcellular organelles known and are osmotically sensitive. Changes in density accompanying osmotic perturbation have enabled changes in water content to be calculated; when referred to a standard state of known volume and water content, fractional and absolute water contents could be calculated for the perturbed states and compared with the fractional free water content as measured by the glycerol space. Under hyperosmotic conditions, discrepancies were found between these two estimates, the glycerol space falling more rapidly than the water space predicted from the density change. This is attributed to a failure of glycerol to displace water imbibed by the membrane as it collapses round an aqueous core of decreasing volume. 'Reserve' vesicles obeyed a relationship between density, osmotic load and osmolality derived for a perfect osmometer, and independent estimates of fractional free water content under standard conditions and osmotic load were made. The former of these agreed well with the glycerol space under standard conditions and the latter agreed with previous estimates of the osmotic load using morphological and analytical data and an assumed activity coefficient of 0.65. Finally, it was possible to model the interconversion of reserve and recycling vesicles more accurately than in previous work.

Vesamicol blocks the recovery, by recycling cholinergic electromotor synaptic vesicles, of the biophysical characteristics of the reserve population.

The effect of vesamicol on the ability of recycling cholinergic synaptic vesicles to recover, during a period of post-stimulation rest, the biophysical properties of the reserve pool was studied in prestimulated perfused blocks of the electric organ of the electric ray, Torpedo marmorata, a tissue rich in cholinergic synapses. The effect of the drug was analysed by high-resolution centrifugal density-gradient fractionation in a zonal rotor of the extracted vesicles. The two vesicle fractions were identified by their ATP and acetylcholine content and the recycled vesicles by their acquisition of [3H]acetylcholine derived from [3H]acetate in the perfusate. Vesamicol (10 microM) blocked the uptake of tritiated acetylcholine by recycled vesicles and also prevented them from rejoining the reserve pool. This is consistent with a previously formulated model of the recovery process, whereby the increase in the acetylcholine and ATP content of the recycled vesicles which takes place during a post-stimulus period of rest increases their osmotic load and thus their content of free water. Vesamicol, by blocking acetylcholine uptake, also blocks rehydration of the recycled vesicles and thus the accompanying decrease in their density to the value characteristic of fully charged vesicles.

The separation and characterization of two forms of Torpedo electric organ calelectrin.

Two methods for extracting calelectrin, a Ca2+-regulated membrane-binding protein from the electric organ of Torpedo marmorata, have been compared and the more promising one was modified to increase the yield to 7-8 mg.kg-1 wet weight of tissue, that is 4-5 times greater than the original method. The calelectrin so obtain could be resoloved into a minor component (designated L-calelectrin) eluted from an anion-exchange column at relatively low ionic strength (100 mM NaCl) and a major component (H-calelectrin) eluted at higher ionic strength (300 mM NaCl). The two forms were also separated by chromatography on a hydrophobic resin. Electrophoresis on cellulose acetate indicated that L-calelectrin had a lower mean isoelectric point that the H-form and polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulphate showed that under reducing conditions (presence of 5% beta-mercaptoethanol) both forms migrated as single species, the L-form having a lower apparent relative molecular mass (Mr 32,000) that the H-form (34,000). Under non-reducing conditions, there was no change in the migration of L-calelectrin but the H-form was resolved into two components of Mr 34,000 and 32,000. The addition of 2 mM Ca2+ had no effect on the migration of either form. Both forms were equally recognized by an anti-calelectrin antiserum and were microheterogeneous with respect to their isoelectric points (pH 4.3-5.5) in two-dimensional gel electrophoresis. Physical measurements were carried out on the major H-form. The Stokes radius was estimated to be 3nm, corresponding to an apparent Mr of 44,000. It was unaffected by changes in ionic strength, pH or Ca2+ concentration. Analytical ultracentrifugation gave a sedimentation constant of 2.9 S and an apparent Mr of 36,000. Measurements of circular dichroism indicated that 78% of the molecule was in the alpha-helix configuration and 22% in random coil. Ca2+ had no significant effect on the conformation.

Ca2+- and phospholipid-binding properties of Torpedo electric organ calelectrin.

The Ca2+-regulated lipid-binding properties of the H- and L-forms of calelectrin present in the electric organ of Torpedo marmorata have been compared. Binding of H-calelectrin required Ca2+ in millimolar concentrations, whereas that of L-calelectrin occurred in the micromolar range. Dissociation of H-calelectrin previously bound to lipids in the presence of 2 mM Ca2+ took place only when the Ca2+ concentration was reduced to micromolar concentrations. Binding was most effective to acidic phospholipids such as phosphatidylserine. Both forms of calelectrin promoted the aggregation of membrane vesicles in the presence of Ca2+.Mg2+, Na+ and K+ inhibited the Ca2+-induced binding to phospholipid, decreasing in effectiveness in that order. Binding was also inhibited by high pH. The surface activity and hydrophobicity index showed that H-calelectrin is a hydrophilic molecule. It may represent a less active, more highly phosphorylated "down-regulated" form of L-calelectrin. The role of calcium in H-calelectrin binding to lipid appeared to be consistent with the formation of a ternary complex of the protein, an acidic lipid and Ca2+, rather than with a direct interaction of lipid with hydrophobic sequences in H-calelectrin whose accessibility is Ca2+-regulated.

The pharmacological properties of the cholinergic false transmitter, N-2-acetoxyethyl-N-methylpyrrolidinium, and its precursor, N-2-hydroxyethyl-N-methylpyrrolidinium.

1 The pharmacological properties of N-2-hydroxyethyl-N-methyl pyrrolidinium (pyrrolcholine) and its acetate ester, recently shown to be a false transmitter at the cholinergic electromotor synapses in Torpedo marmorata, also those of the corresponding morpholinium compounds (morpholinecholine, acetylmorpholinecholine) have been studied on the guinea-pig ileum, frog heart, rectus abdominis muscle, rat blood pressure, rat gastrocnemius muscle and dorsal muscle of the leech. 2 Acetylpyrrolcholine and acetylmorpholinecholine are full cholinoceptor agonists with dose-response curves parallel to that of acetylcholine. They are, however, less potent. Acetylpyrrolcholine is relatively more potent as a muscarinic drug (molar potency about 30% of that of acetylcholine in the ileum but only 4% on the leech) whereas acetylmorpholinecholine is more strongly nicotinic. The unacetylated compounds are very weak agonists with potencies comparable to that of choline. 3 Pyrrolcholine in high concentration showed a distinct neuromuscular blocking effect in the rat gastrocnemius muscle preparation. It is likely that this is a direct effect and not due to uptake by the presynaptic nerve terminals followed by conversion to a false transmitter since it was not reduced by hemicholinium-3, which is known to block uptake of choline and choline analogues by the presynaptic high affinity choline uptake system.

Chol-1: a cholinergic-specific ganglioside of possible significance in central nervous system neurochemistry and neuropathology.

By use of an antiserum raised against presynaptic plasma membrane purified from the purely cholinergic electromotor system of Torpedo marmorata we have been able to identify a group of antigenically-related minor gangliosides (collectively designated Chol-1) that appear to be exclusively localized on cholinergic neurons. The cholinergic-specificity of these antigens has been shown by the following findings: a) The anti-Chol-1 antiserum induces a selective complement-mediated lysis of the cholinergic subpopulation of mammalian brain synaptosomes; b) Section of the fimbria, which causes a massive degeneration of cholinergic terminals in the hippocampus, leads to a concomitant depletion of the level of the Chol-1 gangliosides in the hippocampus; c) The anti-Chol-1 serum can be used to immunostain cholinergic elements in the central and peripheral nervous systems of the rat. The discovery of a cell surface cholinergic-specific antigen has provided a new and effective tool with which to study the cholinergic neuron. For instance, we have immuno-isolated cholinergic synaptosomes from rat cortex and used this preparation to study transmitter coexistence. Our results indicate that approximately 75% of the cortical cholinergic neurons also express the neuropeptide VIP. Furthermore, we are investigating the expression of Chol-1 in patients affected by diseases such as ALS which primarily involve central cholinergic neurons.

A trisialoganglioside containing a sialyl alpha 2-6 N-acetylgalactosamine residue is a cholinergic-specific antigen, Chol-1 alpha.

Cholinergic-specific antigens termed the Chol-1 family have been suggested to be of a ganglioside nature by Richardson et al. (J. Neurochem. 38, 1605-1614, 1982). Two molecular species of polysialogangliosides among bovine brain gangliosides were found to react with anti-Chol-1 alpha antiserum. One of them, Chol-1 alpha-a, was isolated and characterized as a trisialoganglioside containing the gangliotetraose backbone in which 1 mol of sialic acid was attached to each of the reducing end galactose, N-acetylgalactosamine and internal galactose residues, respectively. The chemical structure of Chol-1 alpha-a was determined for the first time, being as follows: IV3NeuAc III6NeuAc II3NeuAc-GgOse4 Cer.

The historical significance of work with electric organs for the study of cholinergic transmission.

The historical significance of work with electric organs for the development of electrobiology and our understanding of the cholinergic synapse at the cell and molecular biological level is traced from its earliest beginning in folk medicine, through the controversy on bioelectricity between Galvani and Volta to the present day, the last decades of which have seen the sequencing of the nicotinic acetylcholine receptor, the isolation and biochemical characterization of the cholinergic vesicle and much else. In the concluding section of the review the continued relevance and usefulness of the electromotor system as a model for future neurobiological research is emphasized.

The cholinergic marker Chol-1 is selectively localized in nerve terminals.

The subcellular distribution of the cholinergic-specific gangliosidic Chol-1 antigens in rat brain has been compared with that of total ganglioside as a marker for neural plasma membranes, particle-bound choline acetyltransferase (ChAT) as a synaptosomal marker, and supernatant ChAT as a marker for neuronal perikaryal cytoplasm. Over 50% of brain Chol-1 was recovered in a synaptosome fraction prepared by a standard density-gradient method. The specific concentration of Chol-1 relative to protein or major gangliosides was also highest in the synaptosome fraction and low in a supernatant fraction which contained most of the neuronal plasma membrane fragments. ChAT was likewise mainly recovered in the synaptosome fraction but its relative concentration there was significantly less than that of Chol-1 (p < 0.01).

Chol-1 is a cholinergic marker in the human central nervous system.

The cholinergic-specific gangliosides Chol-1 alpha and beta were detected in human brain and spinal cord by immune staining of thin-layer chromatography (TLC) plates on which ganglioside extracts had been separated. The colocalization of choline acetyltransferase (ChAT) and the Chol-1 antigens in ventral horn motoneurons was demonstrated immunocytochemically. In analytical studies, Chol-1 was found to be more concentrated in dorsal cord than ventral but the reverse was true of ChAT. This difference was explained by differences in the subcellular location of the two markers. Within each region of the thoracic cord the levels of ChAT and Chol-1 in different cords showed covariance. The expected fall of ChAT and Chol-1 in amyotrophic lateral sclerosis (ALS) cords was not seen and possible reasons for this are discussed.

Characterization of gastrin-releasing peptide immunoreactivity in distinct storage particles in guinea pig myenteric and Torpedo electromotor neurones.

Using high resolution centrifugal density-gradient separation of cytoplasmic extracts of guinea pig myenteric plexus and Torpedo electric tissue, we have succeeded in isolating fractions of storage particles rich in gastrin-releasing peptide (GRP). In extracts of myenteric plexus and gradients derived therefrom, the 10-amino acid GRP peptide (GRP-10) was the sole form present; this was bimodally distributed in the gradients, one peak copurifying with Golgi membranes and apparently consisting of immature storage particles, the other with other synaptophysin-rich neuropeptide-containing particles. In extracts of electric organ, a tissue rich in cholinergic electromotor nerve terminals, and gradients derived therefrom, GRP-like immunoreactivity behaved in gel permeation and reversed phase high performance liquid chromatography like the 27-amino acid peptide (GRP-27). About half of the immunoreactivity sedimented in the centrifugal gradient to a region rich in particles containing vasoactive intestinal polypeptide-like immunoreactivity; the remainder was recovered in a very dense region of the gradient containing larger membrane fragments, including synaptosomes. The electromotor nerves and cell bodies also contained GRP-27-like immunoreactivity in relatively high concentration as did the Torpedo gut. It is concluded that this GRP-like peptide is packaged in dense storage particles in the electromotor neurones.

A peptide with N-terminal histidine and C-terminal isoleucine amide (PHI) and vasoactive intestinal peptide (VIP) are copackaged in myenteric neurones of the guinea pig ileum.

When cytoplasmic extracts of the myenteric plexus of guinea pig ileum are submitted to centrifugal density gradient separation in a zonal rotor, conditions which separate storage particles containing substance P, somatostatin and VIP from each other, PHI copurifies with VIP. The two immunoreactivities cannot be separated by particle exclusion chromatography, which depends on size rather than density. It is concluded that the posttranslational cleavage of the propeptide or precursor to PHI and VIP occurs after packaging into these storage particles.

Arcachon and cholinergic transmission.

The cholinergic nature of transmission at the electromotor synapse of Torpedo marmorata was established at Arcachon in 1939 by Feldberg, Fessard and Nachmansohn (J. Physiol. (Lond.) 97 (1939/1940) 3P-4P) soon after transmission at the neuromuscular junction had been shown to be cholinergic. In 1964, after a quarter of a century of neglect, workers in Cambridge, then in Paris, Göttingen and elsewhere, began to use this system, 500-1000 times richer in cholinergic synapses than muscle, for intensive studies of cholinergic transmission at the cellular and molecular level.

Cholinergic synaptic vesicles are metabolically and biophysically heterogeneous even in resting terminals.

The metabolic heterogeneity of synaptic vesicles in the cholinergic nerve terminals of the electromotor neurons of Torpedo marmorata has been studied in resting tissue by evaluating the molecular acetylcholine content (MAC) of synaptic vesicles after extraction from frozen and crushed tissue and high-resolution centrifugal density gradient separation in a zonal rotor. Although vesicular acetylcholine was distributed in the gradient as a single, more or less symmetrical peak, 3 subpopulations of synaptic vesicles could be identified: a small, relatively light subpopulation of low MAC on the ascending limb of the acetylcholine peak, designated V0, a main population of fully charged vesicles designated V1, and a small, denser subpopulation also of low MAC on the descending limb of the acetylcholine peak, designated V2. The mean proportions and MACs of the 3 pools were: V0, 13%, 58,000; V1, 53%, 246,000; V2, 34%, 79,000. When tritiated acetate was perfused through excised blocks of electric organ for 1-2 h before vesicle isolation, the specific radioactivity of the acetylcholine in the V0 and V2 pools was 10-30 times higher than in the V1 pool. This suggests that both the V0 and V2 pools are not generated by the isolation procedure but are present in the intact endings and are functionally active. On the basis of their density and uptake of newly synthesized acetylcholine, the V0 and V2 pools were identified with the previously described VP0 pool of axonal vesicles and the VP2 pool of recycling vesicles in stimulated nerve terminals respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

A novel cholinergic-specific antigen (Chol-2) in mammalian brain.

Three new antisera have been raised in sheep against cholinergic electromotor presynaptic plasma membranes prepared from the electric organs of the electric ray, Torpedo marmorata. They all recognized one or more cholinergic-specific antigens in the mammalian nervous system by the following criteria: they sensitized the cholinergic subpopulation of rat-brain synaptosomes--and only this subpopulation--to lysis by the complement system and, in an immunocytochemical study, selectively stained choline acetyltransferase-positive cholinergic neurons in the rat spinal cord. However, two of the three antisera failed to recognize Chol-1 alpha and -beta, two closely related minor gangliosides already identified as the cholinergic-specific antigens recognized by previous anti-Torpedo presynaptic plasma membrane antisera or indeed any other ganglioside and the third recognized only Chol-1 alpha. A further investigation of the antigen(s) recognized by the most antigenic of the new antisera indicated that it is proteinaceous in nature, but has epitopes in common with electric organ gangliosides.

The lipid and protein content of cholinergic synaptic vesicles from the electric organ of Torpedo marmorata purified to constant composition: implications for vesicle structure.

The lipid, protein, acetylcholine and ATP content of cholinergic synaptic vesicles isolated from the richly innervated electric organ of Torpedo marmorata and purified to constant composition has been determined. The number of vesicles present in the preparations has been estimated by quantitative electron microscopy and the mean composition of the vesicle deduced. The acetylcholine content of the purest preparations was considerably greater than that previously attained and reached a mean of 6.10 mmole/g of protein and 2.6 X 10(5) molecules/vesicle; the mean values, for all determinations, were 4.1 +/- S.E.M. 0.6 and 2.6 X 10(5) +/- S.E.M. 0.6 X 10(5) respectively. The lipid and protein content of the vesicle (about 140 and 80 ag/vesicle respectively) is relatively low, indicating a thin, lipid-rich membrane and a highly hydrated core of which not more than 1-2% can be occupied by protein. These findings are consistent with conclusions drawn from recent density determinations made at different osmotic pressures using penetrating and non-penetrating gradients.

The involvement of lysophosphoglycerides in neurotransmitter release; the composition and turnover of phospholipids of synaptic vesicles of guinea-pig cerebral cortex and Torpedo electric organ and the effect of stimulation.

(1) Crude synaptosomal fractions (P2) derived from guinea-pig cerebral cortex were incubated in the presence of 50 mM KCl in a Krebs-glucose medium. Torpedo marmorata electric organs were stimulated electrically in vivo at 5 pulses/sec for 30 min by electrodes placed on the electric lobe. Synaptic vesicles were isolated from each source and the phospholipid compositions analysed and compared with vesicles from unstimulated controls. (2) Lysophosphatidylcholine was the only lysophosphoglyceride demonstrable in the synaptic vesicles from either source and its low levels did not increase as a result of chemical or electircal stimulation. In each case there was a close similarity of the phospholipid distributions in the vesicles taken from control and stimulated samples. (3) Control experiments indicated extensive decreases in the acetylcholine content of the vesicles from the stimulated electric organ and smaller decreases in the acetylcholine content of the synaptic vesicles from stimulated crude synaptosomal fractions. These fractions were found to respire linearly in the presence of 10 mM glucose and the vesicle fractions were shown to have low levels of contaiminating membranes as judged by marker enzyme analyses. (4) Crude synaptosomal fractions from guinea-pig cerebral cortex were incubated in a Krebs-glucose medium with labelled fatty acids and [3H]glucose in the presence or absence of 50 mM KCl. Subsynaptosomal fractionation was carried out and specific radioactivities of phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine and phosphatidylinositol were determined in fractions D (synaptic vesicles), E (microsomes) and H (disrupted synaptosomes). The release of neurotransmitter did not significantly enhance the labelling of phospholipids in any of the fractions studied as compared with phospholipids from unstimulated fractions. This was found after two incubation times and using [14C]oleate, [14C]arachidonate, [3H]palmitate and [3H]glucose.

The preparation and characterization of synaptic vesicles of high purity.

Very pure preparations of synaptic vesicles have been obtained from guinea pig cerebral cortex and from the electromotor synapses of Torpedo marmorata by density gradient centrifugation in a zonal rotor followed by chromatography on columns of glass beads of controlled pore size. Markers for soluble cytoplasm (lactate dehydrogenase), plasma and endoplasmic membranes membranes (Na-K-ATPase; acetylcholinesterase, NADPH-cytochrome c reductase], mitochondrial membranes [cytochrome oxidase] and lysosomes [acid phosphatase] were used to assess contamination and were undetectable. The only enzymes detected in the highly purified preparations from guinea pig cerebral cortex were Mg- and Ca-activated ATPases, but their content relative to acetylcholine fell on chromatography suggesting that they may be constituents of non-cholinergic vesicles. Lipids analyses of the highly purified vesicles confirmed earlier results and showed that glycolipids and lysolecithin are present in negligible amounts; this suggests that lysolecithin is not required for exocytosis of synaptic vesicles. A discussion of the probable limiting concentration of acetycholine in cerebral cortical vesicles derived solely from cholinergic terminals suggests that from 13 to 56% of the vesicles isolated are cholinergic, depending on the assumptions made.