Effects of ethanol on neuroblastoma cells in culture: role of gangliosides in neuritogenesis and substrate adhesion.

Murine Neuro-2A neuroblastoma cells were exposed to ethanol in culture under two experimental paradigms: (1) short-term (24 hr or less) and low concentrations (0.05 to 0.5%; 8.5 to 86 mM) and (2) long-term (48 hr at 0.5%; 86 mM). Long-term ethanol exposure did not affect Neuro-2A viability, determined by DNA synthesis or the ability to exclude Trypan Blue. Similarly, long-term ethanol treatment did not inhibit differentiation, exhibited by the extension of neurites, promoted by either dibutyryl-cyclic-AMP or by incubation with exogenous ganglioside GM1. The incorporation of exogenous ganglioside GM1 into plasma membranes was not influenced by varying concentrations of ethanol (up to 1.2%; 204 mM). In contrast, ethanol did influence Neuro-2A cell attachment to collagen in a dualistic manner. During short-term ethanol exposure, cell attachment was enhanced. However, when cells were initially exposed to ethanol for 48 hr a marked inhibition of subsequent attachment was observed. Long-term ethanol exposure also inhibited attachment to other substrata, including laminin, fibronectin and vitronectin. Incubation of Neuro-2A cells with either exogenous ganglioside GM1 or a mixture of brain gangliosides partially reversed the inhibition of attachment to collagen. This reversal did not appear to be due to any one particular ganglioside structure, however. Mixed brain gangliosides were fractionated into three fractions, according to the number of sialic acid residues. Each of the three fractions were equally effective in partially restoring Neuro-2A cell attachment to collagen after long-term ethanol treatment. The results suggest that the mechanism by which these effects occur is at the level of plasma membrane fluidity, because both ethanol and glycosphingolipid content are known to influence membrane lateral mobility, although other mechanisms, such as changes in headgroup hydration, are possible.

Cytotoxicity of the cocaine metabolite benzoylecgonine.

The NG108-15 and C6 cell lines were used in the present study as neuronal and glial models, respectively, to examine the cytotoxicity of the major metabolite of cocaine, benzoylecgonine (BE). Exposure of both cell types to varying concentrations of BE resulted in a loss of cells from the growth surface. Analysis of the unattached cells after such exposure, using a variety of techniques, revealed that these cells were not viable. Therefore, this effect could not be ascribed to BE interfering with cell-substratum interactions. The early events in BE cytotoxicity were examined by observing cells cultivated on the stage of an inverted microscope, using differential interference contrast (Nomarski) optics. Upon exposure of either cell type to 10 microM BE a retraction of cellular processes could be observed within 30 min. Within 6 h cell death was apparent. Similar analyses using 50 microM BE in the growth medium resulted in similar results, except that process retraction could be observed as early as 15 min after exposure. These results demonstrate that the major metabolite of cocaine, benzoylecgonine, is cytotoxic to in vitro models of neuronal and glial cells.

Glycosphingolipids during skeletal muscle cell differentiation: comparison of normal and fusion-defective myoblasts.

The regulation of glycosphingolipid (GSL) synthesis in culture by fusion-competent (E63) myoblasts and fusion-defective (fu-1) cells was examined. Upon reaching confluency E63 cells fused to form multinucleated myotubes and demonstrated many characteristics of developing skeletal muscle including induction of creatine kinase activity and a shift in creatine kinase isozymes to the MM isoform. The fu-1 cells displayed none of these characteristics, despite the fact that both cells were cloned from the same parental myoblast line (rat L8). There was a transient increase in the synthesis of total neutral GSLs by E63 cells at the time of membrane fusion. In contrast, neutral GSL synthesis by fu-1 cells gradually decreased with time in culture. The major GSLs synthesized by both cell types were lactosylceramide and ganglioside GM3, with more complex structures being observed with prolonged time in culture. Several glycosyltransferase activities were assayed at varying times in culture. Generally, the changes in activities fell into three groups. One group was maximally activated at the end of the culture period (GalT-3, GalNAcT-1 and GalT-6). Another group was maximally activated during the time of active membrane fusion (GlcT and SAT-1). A third group was maximally activated at the time of cell contact and the beginning of membrane fusion (GlcNAcT-1 and GalT-2). In terms of the times of maximal activation there were few differences between E63 and fu-1 cells, with one notable exception. The activity of GalT-2 (lactosylceramide synthase) in E63 cells increased dramatically upon contact and the beginning of membrane fusion, whereas there were no changes in GalT-2 activity in fu-1 cells during time in culture. These results support our hypothesis that membrane glycosphingolipids play an important role in the differentiation of skeletal muscle cells.

Cocaine exposure during pregnancy affects rat neonate and maternal brain glycosphingolipids.

Behavioral dysfunctions in offspring exposed in utero to cocaine have been observed, along with alterations in dopamine systems, but few studies of the underlying biochemistry have been conducted. Because of their documented roles in neuronal maturation, glycosphingolipids were analyzed in whole brains of offspring exposed gestationally to cocaine. Rat offspring exposed to cocaine in utero exhibited markedly elevated levels of both total gangliosides (p < 0.001) and neutral glycosphingolipids (p < 0.01) at postnatal day 1. However, by postnatal day 11 levels of gangliosides and neutral glycosphingolipids returned to control values. These effects were not restricted to chronic cocaine exposure early in life, in that ganglioside content of whole maternal brains was also elevated (p < 0.001), though less than that observed with the neonate brains. Qualitatively, no differences in ganglioside nor neutral glycolipid structure distribution were observed between cocaine-exposed and normal animals following separation by HPTLC and HPLC. These elevations are in contrast to those following alcohol exposure, where decreases in brain gangliosides have been observed. Neurochemical consequences of prenatal exposure to cocaine may be far-reaching and may not be restricted to the dopamine system.

Inhibition of CMP-N-acetylneuraminic acid:lactosylceramide sialyltransferase by nucleotides, nucleotide sugars and nucleotide dialdehydes.

The effects of nucleotides, nucleotide sugars and nucleotide dialdehydes on the activity and kinetics of cytidine 5'-monophospho-N-acetylneuraminic acid:lactosylceramide (alpha 2-->3) sialyltransferase (SAT-1) in microsomes derived from embryonic chick brain were investigated. Although under physiological conditions this enzyme utilizes a CMP-sugar as substrate, it was found that UDP-dialdehyde was an effective inhibitor of SAT-1 activity. CMP-dialdehyde was only slightly more efficient at inhibiting SAT-1 activity. Similar findings were found for the inhibitory effects of UDP versus CMP. In addition, two UDP-sugars (UDP-Gal and UDP-GalNAc) were also slightly inhibitory. Kinetic analyses demonstrate that both UDP- and CMP-dialdehydes are competitive inhibitors of SAT-1 activity. The data suggests that the substrate specificity of microsomal SAT-1 resides more in the sugar moiety, rather than in the nucleotide portion of the substrate.

Regulation of glycolipid synthesis during differentiation of clonal murine muscle cells.

The two clonal murine muscle cell lines G7 and G8, originally derived from the M114 line, represent unique models for comparative studies of myogenesis. Glycolipid synthesis was examined during differentiation using [3H]-galactose and [3H]-glucosamine as precursors. Upon G7 contact glucosylceramide labeling increased and nLcOse5Cer labeling stopped. During membrane fusion, glucosylceramide labeling stopped and lactosylceramide became the major synthetic product. G8 cells presented a different pattern, with increased labeling of GbOse3Cer during myogenesis. The major ganglioside synthesized by both myoblasts was GM3, and more complex structures were observed following completion of myotube formation. Total glycopeptide labeling increased when G8 myoblasts fused and remained elevated in myotubes, whereas no differences during fusion of G7 cells were noted. Upon comparison of the two clonal lines, the only consistent observation was a significant increase in the synthesis of total gangliosides and neutral glycolipid during cell contact and membrane fusion (p less than 0.02). The results suggest that changes in the synthesis of specific glycolipid structures during myogenesis are unique to each muscle cell line examined. However, transient increases in synthesis of total myoblast gangliosides and neutral glycolipids may be a more general phenomenon, possibly by curbing proliferation or by altering myoblast membrane fluidity characteristics during differentiation.

A sensitive method to quantitate gangliosides of the gangliotetraose series directly on chromatograms using peroxidase conjugated cholera toxin.

A method is described whereby ganglioside GM1 can be quantitated directly on thin-layer chromatograms using cholera toxin subunit B conjugated to horseradish peroxidase and visualized with chloronaphthol. Overlay and color development were performed after separating gangliosides on nano-TLC plates, and fixing with polyisobutylmethacrylate. Absolute quantitation was realized using a Shimadzu CS-9000 integrating spectrodensitometer, scanning at 580 nm. A correlation coefficient of 0.98 was obtained in a linear range of detection from 10(-11) to 10(-16) moles. Statistical analysis revealed good reproducibility and over 99% of the added gangliosides remained with the chromatogram during all overlay and washing procedures. By comparison, standard chemical visualization by resorcinol-HCl was linear in the nanomole range with a detection limit of only 10(-10) moles. Since the carbohydrate portion of gangliosides immobilized in this manner is susceptible to the action of enzymes including neuraminidase, this technique can be applied to all structures of the gangliotetraose series.

Post-mortem stability of membrane-associated sialidase activity in brain.

The post-mortem changes in membrane-associated sialidase (N-acetylneuraminosyl glycohydrolase, EC 3.2.1.18) were examined in rat brain obtained and stored in a manner which paralleled neuropathological handling of human brains. When whole brains were held at 4 degrees C in covered containers for varying periods of time (0.67 h), sialidase activity toward endogenous membrane substrate was elevated. This elevated activity was maximal at 8 h of storage and decreased thereafter. The apparent decrease in enzyme activation from 8 to 67 h of storage was not due to a reduction of activity, but was the result of depletion of endogenous membrane substrate, since activity toward exogenous ganglioside remained elevated. The changes were due to whole brain storage at 4 degrees C, and not a result of being stored at -80 degrees C. The post-mortem activation of sialidase was not due to the expression of a new form of the enzyme, since it displayed characteristics similar to those reported previously: (a) membrane gangliosides being the preferred native substrate, and (b) ganglioside GM1 and lactosylceramide being the major hydrolytic products. The results underscore the importance of post-mortem storage conditions when analyzing complex carbohydrates of brain.

Incorporation of exogenous ganglioside GM1 into neuroblastoma membranes: inhibition by calcium ion and dependence upon membrane protein.

Since exogenous gangliosides are known to promote neuritogenesis, the incorporation of exogenous GM1 into neuroblastoma membranes was examined. Neuro-2A cells, synchronized in the G1/G0 phase, were suspended in HEPES buffered saline containing 10(-4) M [3H]GM1, and membrane incorporation was measured as radioactivity remaining with the cell pellet following incubation with serum-containing medium and trypsin. Calcium ion (0.01 to 10 mM) reduced incorporation of exogenous GM1, due to its interaction with GM1 micelles in solution. When cells were treated with proteases prior to incubation with GM1, the inhibitory effect of Ca2+ was lost and total incorporation into membranes was lowered by approximately one order of magnitude. Pretreatment of cells with 0.05% trypsin resulted in an inhibition of GM1 incorporation within 5 minutes. When trypsinized cells were resuspended in complete growth medium, the cells recovered the ability to incorporate GM1 with time, and this paralleled labeling of cellular protein with [3H]leucine. The role of membrane protein in the incorporation of exogenous GM1 could not be explained by the lytic release of cytosolic transfer proteins nor the artifactual coating of the cell surface by serum proteins. These results suggest that the incorporation of exogenous gangliosides into cellular membrane lipid bilayers cannot be fully explained by considerations of lipophilicity alone, and leads us to propose that initial recognition by membrane protein(s) is necessary.

Glycosphingolipid biosynthesis during myogenesis of rat L6 cells in vitro.

Glycosphingolipid biosynthesis was examined using [3H]-galactose as a precursor as rat L6 myoblasts fused to form multinucleated myotubes. Incorporation of label into neutral glycolipids decreased steadily as the population of myotubes increased, so that final biosynthesis was one-half that observed with myoblasts (p less than 0.02). Conversely, ganglioside biosynthesis doubled during myoblast confluency (p less than 0.02) and then decreased as myotubes formed. Qualitatively, L6 cells synthesized large amounts of ganglioside GM3 during all myogenic phases. The major neutral glycosphingolipid products were lactosylceramide and paragloboside (nLcOse4Cer). Few changes in TLC autoradiographic patterns were noted during differentiation, with the exception of a slight decrease in ganglioside GM1. The results indicate that the biosynthesis of glycosphingolipids is tightly regulated during myogenesis in vitro and suggest a role for membrane gangliosides in muscle cell differentiation.

In situ visualization of glycosphingolipids on thin-layer chromatograms using lectin-peroxidase conjugates.

A technique is described for direct detection by lectins conjugated to horseradish peroxidase of the oligosaccharides of glycosphingolipids on thin-layer chromatograms. Final visualization is accomplished using a substrate medium consisting of hydrogen peroxide and 3,3'-diaminobenzidine. The procedure is rapid, specific, sensitive and highly reproducible, and staining patterns are stable for years. In addition to providing preliminary structural information, this technique can be employed in combination with other methods, such as autoradiographic detection of tritium-labeled glycosphingolipids in the same chromatogram.

Abnormal glycosylation of myelin-associated glycoprotein in quaking mouse brain.

Brain slices from actively myelinating (26-28 days) quaking and normal littermates were dual-labeled with radioactive mannose and fucose for 2 h. Following the incubation myelin was isolated by sucrose density gradient centrifugation and the incorporation of sugars into the major myelinassociated glycoprotein (MAG) determined. The incorporation of mannose (an internal monosaccharide) and fucose (a terminal monosaccharide) was impaired in quaking by approximately 70 and 83% respectively as compared to control. The mannose/fucose ratio in quaking myelin was approximately 70% higher than in control. The results indicate an abnormal processing of the N-linked oligosaccharide moiety of MAG in quaking oligodendrocytes.

Lipid content of swine influenza and other vaccines.

An analysis of the lipids in swine influenza vaccines was performed, comparing six different lots of swine influenza, other influenza and noninfluenza vaccines. Cholesterol content and phospholipid content varied greatly, but there were no major differences between the types of vaccines. Appreciable amounts of phosphatidylethanolamine were found in only one swine influenza vaccine. The major phospholipids of influenza vaccines were phosphatidylcholine, sphingomyelin and phosphatidic acid. A detectable amount of phosphatidylserine was not found in any swine influenza vaccine, but was present in two of three nonswine influenza vaccines. Only two of six swine influenza vaccines showed trace amounts (less than 0.5 microgram/ml) of ganglioside (GM3). However, larger quantities of galactocerebroside were found (2.24-6.43 micrograms/ml) in all influenza vaccines examined, including swine influenza vaccines.

Phospholipid biosynthesis during normal and dystrophic avian muscle cell differentiation in culture.

The biosynthesis of phospholipids during differentiation of normal and dystrophic muscle cells was studied using [32Pi] added to primary culture media. Labeled phosphatides were separated by two-dimensional thin-layer chromatography, exposing the plates to HCl vapors between solvents to cleave the alk-1-enyl bonds. Phosphatidylethanolamine labeling did not vary during myotube formation. Although lower labeling of phosphatidylcholine was observed in dystrophic myoblasts at the single cell stage, no differences were observed between normal and dystrophic cultures after cell contact and myotube formation. A sharp decrease in phosphoinositide content was observed as myoblasts first achieved cell contact and continued to decline slowly as myotubes developed. During myogenesis, phosphatidylcholine content gradually increased, from 18.4% of the total labeled phospholipid in dividing myoblasts to 48.6% of the labeled phospholipids in fully differentiated myotubes.

Quantitation of the in vitro neuroblastoma response to exogenous, purified gangliosides.

Individual ganglioside species (possessing the gangliotetrose oligosaccharide) were purified from bovine brain gray matter and applied in varying concentrations to the culture medium of mouse neuroblastoma cells (N2A) in vitro. After 48 hr of incubation, the cells were stained, and the neuritogenic response quantitated with a video analysis system, employing a program to measure three parameters of neuroblastoma differentiation: neurites per cell (sprouting), neurite length (extension), and degree of neurite branching (arborization). All the individual gangliosides tested promoted neurite extension in a dose-dependent fashion. Asialogangliosides ("neutral" glycosphingolipids) were without effect, which suggests that sialic acid (N-acetylneuraminic acid) is necessary to elicit this cellular response. With increasing concentrations of GM1 (5 to 500 micrograms/ml), the average cellular neurite length increased significantly, whereas the number of neurites per cell decreased. With the trisialoganglioside GT1b, neurite length did not increase to the extent seen with GM1, but an increase in the number of neurites per cell (sprouting) and branch points per neurite (arborization) was observed. These results suggest that the in vitro neuronal response to exogenous gangliosides may combine specific responses to individual species making up the total.

A simplified procedure for the preparation of tritiated GM1 ganglioside and other glycosphingolipids.

The ganglioside II3NeuAc-GgOse4Cer and other glycosphingolipids can be radiolabeled to high specific activity by the galactose oxidase-NaB3H4 procedure, by purifying the oxidized compounds prior to reductive labeling. The oxidized products are separated from nonoxidized compounds and detergents (Triton X-100 and sodium taurocholate) present during the enzymatic oxidation. Since the oxidized derivatives are separated, the final specific activity depends solely upon the specific activity of the NaB3H4 and the reduction conditions.

Glycosphingolipids of an alveolar rhabdomyosarcoma and normal human muscle. A case study.

The distribution of neutral, or asialosyl-, glycosphingolipids and gangliosides in a rhabdomyosarcoma of alveolar type have been studied. Histologically, this muscle tumor is composed primarily of two cell types: one with oval or round hyperchromatic nuclei and very little cytoplasm, and one a giant cell, with multiple, peripherally placed nuclei and weakly staining eosinophillic cytoplasm. In comparing glycolipids of the rhabdomyosarcoma with normal muscle from the same leg, the striking alteration in the tumor was a virtual disappearance of ganglioside GM2. There was also a slight increase in GM3 and a decrease in GD1a. The asialosyl derivative of GM2 (GalNac-Gal-Glc-Cer) was markedly increased in the tumor. A loss of glucosylceramide was also observed. The results are discussed in terms of glycolipid metabolic changes in muscle oncogenesis and their implications.

Properties of endogenous, membrane-associated sialidase activity (N-acetylneuraminidase) of the goldfish visual system.

The endogenous sialidase (N-acetylneuraminidase) activity of membranes prepared from goldfish retina and optic tectum displays characteristics similar to those reported for neural plasma membrane sialidases of other organisms. Endogenous membrane sialidase activity was found to be optimal at ph 4.0, and maximal release was obtained at 37-50 degrees C, above which temperature thermal instability of the preparations was observed. Optic nerve crush, which results in regeneration of retinal ganglion cell axons, did not result in significant changes in measured endogenous membrane sialidase activity in either the retina or the optic tectum. Enzymatic hydrolysis of membrane sialoglycolipid (ganglioside) accounted for about 70% of the total sialic acid released. Ganglioside GM1 accumulated as the major lipid product in both retina and tectum, indicating that the inner sialosylgalactosyl linkage in the ganglio oligosaccharide series was resistant to hydrolysis by the endogenous enzyme.