The biosynthesis of brain gangliosides. Ganglioside-glycosylating activity in rat brain neuronal perikarya fraction.

Rat brain homogenate and the synaptosmal and neuronal perikarya fractions from 17-day-old rats were compared for their activities in sialosylating endogenous gangliosides and transferring N-acetylneuraminic acid and galactose to several glycolipids in vitro. The sialosylation of endogenous gangliosides and the activities of sialosyltransferases acting either on lactosylceramide or haematoside as acceptors, as well as galactosyltransferase acting on Tay-Sachs ganglioside as acceptor, were between 3-and 12-fold higher in the neuronal perikarya fraction than in whole homgenate on a protein or ganglioside basis. The activities found in the synaptosomal fraction were negligible. No evidence was found to indicate that the low activities in this fraction were due to the presence of inhibitors of the transfer activities or to inacessibility of the substrates to their respective enzymes. These findings, and the time course of labelling of gangliosides of the neuronal perikarya and synaptosomes from rats that received an injection of N-[3H]acetylmannosamine, indicate that the main cellular site of glycosylation of neuronal gangliosides is in the neuronal perikarya.     

The biosynthesis of gangliosides. Labelling of rat brain gangliosides in vivo

1. After injection of [6-(3)H]glucosamine into 8-day-old rats it was found that all the major brain gangliosides and their sialyl groups were labelled at essentially the same rate, except the hematoside, which was the least labelled. In 18-day-old rats it was found that the two major gangliosides with the sialyl (2-->8)-sialyl linkage, and their sialyl groups were more labelled than the hematoside, the Tay-Sachs ganglioside, the other two major gangliosides and their respective sialyl groups. 2. No difference was found in any of the cases studied between the specific radioactivities of the neuraminidase-resistant and -labile sialyl groups belonging to the same ganglioside. The same was found for the specific radioactivities of the galactosyl groups proximal and distal to the ceramide moiety of total brain gangliosides from rats injected with [U-(14)C]glucose. From this it was concluded that partial turnover of the ganglioside molecule does not occur. 3. A model for the synthesis of gangliosides is presented that accounts for results from previous experiments in vitro and the lack of precursor-product relationships observed in experiments in vivo.     

Separation of the glycoprotein and ganglioside components of thyrotropin receptor activity in plasma membranes.

The two components of thyroid plasma membranes known to interact with thyrotropin, i.e., a glycoprotein with specific thyrotropin binding activity and the gangliosides of the thyroid membranes, are shown to segregate differently when membranes are solubilized with lithium diiodosalicylate. Individually examined, the interaction of each component with thyrotropin exhibits a different sensitivity to salts. The data suggest that the thyrotropin receptor on the thyroid membrane is a complex which is composed of both glycoprotein and ganglioside components and that its properties are derived from each component.     

Determination of brain gangliosides by determination of ganglioside stearic acid

A new method is described for the determination of brain gangliosides by measuring stearic acid, the chief acid of gangliosides, in an appropriately purified brain extract. The method includes extraction of tissue with chloroform-methanol, extraction of gangliosides from the extract with 0.1 M KCl, evaporation of the aqueous phase, methanolysis, and gas-liquid chromatography of the resultant methyl esters with a double internal standard. The method depends on the simple composition of ganglioside fatty acids (80% stearic acid) and allows determination of less than 0.05 micromole of gangliosides. Interfering lipids are removed from the ganglioside extract by washing with chloroform-methanol-water. The effects of contamination with nonlipid N-acetylneuraminic acid are avoided.     

Rat brain microsomal gangliosides. Accessibility to a neuraminidase preparation and the possible existence of different pools in relation to their biosynthesis

1. Treatment of rat brain microsomal membranes with a neuraminidase preparation from Clostridium perfringens resulted in an almost complete conversion of polysialogangliosides into monosialogangliosides. 2. Neuraminidase treatment of the membranes did not increase the incorporation of N-[(3)H]acetylneuraminic acid from CMP-N-[(3)H]acetylneuraminic acid into the gangliosidic fraction, indicating that a monosialoganglioside is an acceptor of N-acetylneuraminic acid in these membranes only if, in addition to having the right chemical structure, it is in a proper position, probably in relation to the endogenous sialyltransferases. 3. These experiments also indicated that no independent turnover of the neuraminidase-labile N-acetylneuraminyl groups of gangliosides occurred in vitro. 4. N-[(3)H]Acetylneuraminic acid from endogenous polysialogangliosides labelled in vitro was released by neuraminidase at a slower rate than N-acetylneuraminic acid from unlabelled gangliosides of the same membranes. From this it was concluded that recently synthesized polysialogangliosides (completed in vitro) are in the membranes in a position less accessible to neuraminidase than are those synthesized earlier which were present in the membranes at the start of the labelling experiment.     

The influence of ganglioside insertion into brain membranes on the rate of ganglioside degradation by membrane-bound sialidase

Microsomal membranes isolated from calf brain contain a sialidase which cleaves ganglioside substrates naturally occurring within these membranes as well as exogenously added [3H]ganglioside GD1a. Micelles of [3H]ganglioside GD1a bind to the microsomal membranes in two steps. The first step, called adsorption, is fast and reversible by treatment with trypsin; the second step, called uptake, is slower and not reversible. The product of the enzymic degradation, [3H]ganglioside GM1, is exclusively located in the ganglioside pool taken up by the sialidase-bearing membranes, and not in the trypsin-releasable pool. Electron spin resonance (ESR) studies using a spin-labelled analogue of ganglioside GD1a indicate that the ganglioside uptake by microsomal membranes is accompanied by the disappearance of the micellar structure and by the 'dilution' of the probe molecules with membrane lipids. These findings suggest that exogenously added ganglioside substrate inserts into the microsomal membrane before it is recognized as substrate by the membrane-bound sialidase. Therefore, the influence of pH, ionic strength and membrane-fluidizing agents on the degradation rate measured with exogenous ganglioside GD1a does not only reflect kinetic parameters of the enzymic reaction itself but also the velocity of ganglioside insertion. Increasing ionic strength reduces the degradation rate. The acceleration of insertion with falling pH values shifts the measured pH optimum of the ganglioside degradation to lower values (pH 3.6) and masks the substantial residual sialidase activity at pH 5-7. The membrane-fluidizing alcohol n-hexanol greatly accelerates ganglioside insertion as well as ganglioside degradation. The latter was clearly demonstrated by studying the hydrolysis of endogenous ganglioside substrates, and is due to a decrease of the apparent Km value and an increase in the Vmax value. The Vmax value was also enhanced by freezing and thawing of the microsomal membranes.     

The interaction of calcium with gangliosides in bilayer membranes

We studied the binding of calcium to bilayer membranes formed from mixtures of phosphatidylcholine and mono-, di-, or trisialoganglioside by measuring its effect on the electrophoretic mobility of multilamellar vesicles and the conductance of planar bilayers. In 0.001 M monovalent salt solutions the surface potential of the membranes is large and micromolar concentrations of calcium have a significant effect on the mobility and conductance. In 0.1 M monovalent salt solutions the surface potential is small and millimolar concentrations of calcium are required to affect these parameters. The strong apparent binding of calcium we observed at low ionic strength could be due to the nonspecific accumulation of calcium in the electrical diffuse double layer. To distinguish between this nonspecific effect and binding of calcium to the membrane, we substituted dimethonium for calcium. Dimethonium is a divalent cation that screens negative charges but does not bind to lipids. We also examined the effect of replacing phosphatidylcholine by monoolein: calcium binds to phosphatidylcholine but not to monoolein. We describe our electrophoretic mobility results by combining the Poisson-Boltzmann and Navier-Stokes equations with the Langmuir adsorption isotherm. We conclude that calcium binds weakly to gangliosides with an intrinsic association constant of less than 100 M-1, which is similar to the association constant of calcium with phospholipids.     

Blood-group-Ii-active gangliosides of human erythrocyte membranes.

More than ten new types of gangliosides, in addition to haematoside and sialosylparagloboside, were isolated from human erythrocyte membranes. These were separated by successive chromatographies on DEAE-Sephadex, on porous silica-gel columns and on thin-layer silica gel as acetylated compounds. Highly potent blood-group-Ii and moderate blood-group-H activities were demonstrated in some of the ganglioside fractions. The gangliosides incorporated into cholesterol/phosphatidylcholine liposomes stoicheiometrically inhibited binding of anti-(blood-group I and i) antibodies to a radioiodinated blood-group-Ii-active glycoprotein. The fraction with the highest blood-group-I-activity, I(g) fraction, behaved like sialosyl-deca- to -dodeca-glycosylceramides on t.l.c. Certain blood-group-I and most of the -i determinants were in partially or completely cryptic form and could be unmasked by sialidase treatment. Thus the I and i antigens, which are known to occur on internal structures of blood-group-ABH-active glycoproteins in secretions, also occur in the interior of the carbohydrate chains of erythrocyte gangliosides.     

Interactions of pig brain cytosolic sialidase with gangliosides. The formation of catalytically inactive enzyme-ganglioside complexes requires homogeneous ganglioside micelles and is a reversible phenomenon

Cytosolic sialidase A, obtained from pig brain and purified, interacts with ganglioside GT1b giving two catalytically inactive enzyme-ganglioside complexes. Treatment of these complexes with Triton X-100 under given conditions (1% detergent; 1 h at 37 degrees C; 0.1 M acetic acid-sodium acetate buffer, pH 4.8) leads to the liberation of part of the enzyme (about 47%) in a free and fully active form. Reversible inactivation of cytosolic sialidase requires the presence of homogeneous micelles of GT1b or of mixed micelles (for instance Triton X-100 and GT1b) with a high GT1b content. Triton X-100/ganglioside mixed micelles with a molar ratio above 50, as well as small unilamellar vesicles of egg yolk lecithin and GT1b (7-15 mol%), did not inactivate the enzyme at all; on the contrary these forms of ganglioside dispersion behaved as excellent substrates for the enzyme. It is to be concluded that under in vitro conditions the ability of ganglioside to interact with cytosolic sialidase, giving rise to catalytically inactive complexes or to Michaelis-Menten enzyme-substrate complexes, depends on the supramolecular organization of the ganglioside molecules. Arrangements of tightly packed molecules with strong side-side interactions facilitate the formation of complexes with the enzyme; arrangement with separated and loosely interacting molecules facilitates binding at the catalytically active site of the enzyme.     

Inhibition of adenylate cyclase activity of human thyroid membranes by gangliosides

Gangliosides inhibit basal, thyrotropin-induced and fluoride-induced adenylate cyclase activity of human thyroid membranes in physiological conditions. In contrast neutral glycolipids, phospholipids and neuraminic acid containing oligosaccharides show no effect. The efficacy of inhibition is more dependent upon the position of the sialic acid residues than upon their absolute number. In general gangliosides with disialyl groups are more inhibitory than those with single sialyl moieties. The inhibitory effects of the individual gangliosides on the two modes of stimulation are parallel. This parallelism suggests that the inhibitory effect is located at the postreceptor level and that the gangliosides interact directly with the adenylate cyclase system. A possible role of thyroid membrane gangliosides as suppressive cofactors of adenylate cyclase is discussed in relation to recent findings of stimulating anti-ganglioside antibodies in Graves' disease.     

Phospholipase B-like activity in rabbit brain membranes.

1. Phospholipase B-like activity was found in rabbit brain membranes. 2. The activity was greatly enhanced by 0.025% (w/v) Triton X-100 and was inhibited by both Ca2+ and Mg2+. 3. With increasing pH of the reaction mixture, the activity was augmented. 4. The characteristics of the enzyme activity possibly suggest that phospholipase B in rabbit brain may be distinct from those previously reported.     

Phase behavior of ganglioside-lecithin mixtures. Relation to dispersion of gangliosides in membranes.

Ganglioside GM1 and mixed brain gangliosides were mixed with 1-stearoyl-2-oleoyl lecithin (SOPC) and examined by differential scanning calorimetry as a function of ganglioside content and temperature. Low mole fractions of ganglioside GM1 and of mixed brain gangliosides are shown to be miscible with SOPC in the gel phase up to X = 0.3, with the possible exception of a small region of immiscibility for the mixed brain gangliosides system centered around X = 0.05. Above X = 0.3, the low-temperature phases demix into a (gel) phase of composition X = 0.3 and a (micellar) phase of composition X = 1.0. Above the endothermic phase transition temperature, no phase boundaries are discerned. It is pointed out that phase structures need to be determined in each domain delineated in the phase diagrams, and that cylindrical phases may exist at higher temperatures and intermediate compositions. The effects of addition of wheat germ agglutinin, which binds to ganglioside GM1, on a ganglioside GM1-SOPC mixture (X = 0.5), are described and interpreted in terms of partial demixing of ganglioside and lecithin. Behavior of the ganglioside-SOPC system is discussed with respect to the kinetics of cholera toxin action in lymphocytes, as well as to other physiological roles of gangliosides in membranes.