Synaptic Membrane Glycoproteins Research Articles

Glycan analysis of the chicken synaptic plasma membrane glycoproteins - a major synaptic N-glycan carries the LewisX determinant

The majority of synaptic plasma membrane components are glycosylated. It is now widely accepted that this post-translational modification is crucial during the establishment, maintenance and function of the nervous system. Despite its significance, structural information about the glycosylation of nervous system specific glycoproteins is very limited. In the present study the major glycan structures of the chicken synaptic plasma membrane (SPM) associated glycoprotein glycans were determined. N-glycans were released by hydrazinolysis, labelled with 2-aminobenzamide, treated with neuraminidase and subsequently fractionated by size exclusion chromatography. Individual fractions were characterized by the combination of high-pressure liquid chromatography, exoglycosidase treatment or reagent array analysis method (RAAM). In addition to oligomannose-type glycans, core-fucosylated complex glycans with biantennary bisecting glycans carrying the LewisX epitope were most abundant. The overall chicken glycan profile was strikingly similar to the rat brain glycan profile. The presence of the LewisX determinant in relatively large proportions suggests a tissue-specific function for these glycans.

Neural cell adhesion molecules play a role in rat memory formation in appetitive as well as aversive tasks.

A polyclonal antibody (R1), raised against chick synaptic membrane glycoproteins and recognizing the neural cell adhesion molecule (NCAM) caused amnesia for avoidance tasks when injected into day-old chicks and adult rats 5.5 h post-training. We investigated the effects of R1 antibody on memory formation in a non-aversive task, where stress is minimal: a massed trial odour discrimination task in rats. Preimmune serum or R1 antibody was injected i.c.v. 5.5 h after the last training session. Forty-eight hours after the training session, control rats showed very good retention whereas R1 antibody injection significantly disrupted retention. The results suggest that glycoproteins recognized by R1 in the rat play a specific role in memory formation for appetitive events as well as in memory formation for aversive situations.

Developmental shift of synaptic vesicle protein 2 from axons to terminals in the primary visual projection of the hamster

Synaptic vesicle protein 2 is an integral synaptic vesicle membrane glycoprotein which is present in all synapses for which it has been examined. We used an anti-synaptic vesicle protein 2 monoclonal antibody to examine synaptic vesicle protein 2 localization in the developing hamster retinofugal pathway. From postnatal day 0 to day 1, a period of elongation of retinal ganglion cell axons to their central targets, fiber fascicles in the optic tract over the lateral geniculate nucleus were intensely synaptic vesicle protein 2-immunoreactive. Adjacent to the optic tract, single fibers could be seen. We also observed a marked immunostaining in growth cones and fiber fascicles in retinal explants in culture. By postnatal day 2, the staining of single fibers had ended, and by postnatal day 5, during the formation of terminal arbors, numerous fine puncta of synaptic vesicle protein 2 immunoreactivity were distributed within the neuropil of the lateral geniculate nucleus. In the adult, the optic tract was devoid of synaptic vesicle protein 2 staining, while the neuropil contained distinct immunoreactive profiles, particularly in the outer shell of the lateral geniculate. These synaptic vesicle protein 2-positive profiles closely resembled the grape-like clusters and large swellings of two known retinal axon terminal types. Eye removal resulted in the rapid disappearance of these synaptic vesicle protein 2-labelled terminal profiles contralateral to the enucleation. A similar pattern of synaptic vesicle protein 2 immunoreactivity was observed in the superior colliculus. From postnatal day 0 to day 2, retinal fiber fascicles in the stratum griseum superficiale/stratum opticum were darkly stained for synaptic vesicle protein 2. By postnatal day 5, the immunoreactivity shifted to the neuropil and from postnatal day 6 onwards, the synaptic vesicle protein 2 immunoreactivity was more intense in the stratum griseum superficiale than in the optic fiber layer. This study demonstrates dense synaptic vesicle protein 2-labelling of elongating axons both in vivo and in vitro. However, coincident with the transition from retinal ganglion cell axon elongation to terminal arborization, synaptic vesicle protein 2 is progressively restricted to synaptic terminals and becomes undetectable in axons. This study is the first to document an axonal localization of synaptic vesicle protein 2 during development and raises the question as to its role during axonal elongation.

Synaptic Membrane Glycoproteins gp65 and gp55 Are New Members of the Immunoglobulin Superfamily

Glycoproteins gp65 and gp55 are major components of synaptic membranes prepared from rat forebrain. Both are recognized by the monoclonal antibody SMgp65. We have used SMgp65 to screen a rat brain cDNA expression library. Two sets of overlapping cDNAs that contain open reading frames of 397 and 281 amino acids were isolated. The deduced proteins are members of the immunoglobulin (Ig) superfamily containing three and two Ig domains, respectively. The common part has approximately 40% sequence identity with neurothelin/basigin. The identity of the proteins as gp65 and gp55 was confirmed by production of new antisera against a common recombinant protein fragment. These antisera immunoprecipitate gp65 and gp55. Furthermore, expression of gp65 and gp55 cDNAs in human 293 cells treated with tunicamycin results in the production of unglycosylated core proteins of identical size to deglycosylated gp65 and gp55. Northern analysis revealed that gp65 transcripts are brain-specific, whereas gp55 is expressed in most tissues and cell lines examined. The tissue distribution was confirmed at the protein level though the pattern of glycosylation of gp55 varies between tissues. In situ hybridization experiments with a common and a gp65-specific probe suggest differential expression of gp65 and gp55 transcripts in the rat brain.

Antibody to Day-Old Chick Brain Glycoprotein Produces Amnesia in Adult Rats

Polyclonal antibody R-1, raised against a chick synaptic membrane glycoprotein fraction whose synthesis is enhanced following training on a passive avoidance task, produces amnesia when injected into chick forebrain 5.5 h posttraining. The amnestic IgG fraction specifically recognizes a low sialylated isoform of NCAM (Mileusnic, Rose, Lancashire, & Bullock, 1995). We have now investigated the effects of this antibody on memory formation in adult rats. R-1, preimmune serum, or saline was injected intracerebroventricularly 5.5 h posttraining through bilaterally implanted cannulae. Rats injected with R-1 and tested 48 h later showed a significant amnesia for avoidance compared with the controls. Amnesia was not apparent at 24 h posttraining. R-1 injections were without effect on spontaneous locomotor or exploratory activity in a hole-board test. The results contribute to the argument that the role of cell adhesion molecules in neuronal plasticity is not limited to the developing nervous system, but they play a more general role in the experience-dependent synaptic remodeling underlying long-term memory.

Synaptic Vesicle and Synaptic Membrane Glycoproteins during Pre- and Postnatal Development of Mouse Cerebral Cortex, Cerebellum and Spinal Cord

Glycoproteins of synaptic vesicles and synaptic membranes play an important role during the process of synaptogenesis. In order to study the temporal expression of specific carbohydrates and the expression of selected neural proteins, we used peroxidase-conjugated lectin overlays on Western blots and immunoblots of synaptic vesicles and synaptic membranes isolated from pre- and postnatal mouse cerebral cortex, cerebellum and spinal cord. Our lectin overlays on Western blots showed that (1) the synaptic vesicle glycoproteins, gp80-100, gp47 and gp44, and (2) the synaptic membrane glycoproteins, gp180, gp72, gp70 and gp34, show temporal regulation of expression of carbohydrate moieties. Quite significantly, gp47 showed a decrease in the vesicles coinciding with an increase in membranes suggesting a shift in localization. Anti neural cell adhesion molecule (N-CAM) antibody cross-reacted with gp180. The developmental expression of synaptotagmin 1, a well characterized glycoprotein of synaptic vesicle, was determined by immunoblots analysis. Anti synaptosomal-associated protein 25 (SNAP-25) antibody immunoblots were performed in order to compare our results with a developmentally regulated synaptosomal protein demonstrating expression coincident with synaptogenesis. Our immunoblot studies showed that (1) N-CAM (gp180) immunoreactivity decrease with development; (2) the expression of synaptotagmin 1 and SNAP-25 increases as development progresses, and (3) synaptotagmin 1 and SNAP-25 show a shift in subcellular localization (from synaptic vesicle to synaptic membrane) during development, thereby indicating that these proteins are first identified in a vesicular fraction. Thus, our data suggest that synaptic vesicle and synaptic membrane glycoproteins show temporal regulation of specific carbohydrates as well as protein expression during development, which may be a key factor to our understanding of the process of synaptogenesis.

Peptide-containing innervation of rat femoral lymphatic vessels

The lymphatic vessels conduct lymph fluid, proteins, and potentially antigenic material from the interstitium back to the bloodstream via lymph nodes, where solids are removed by phagocytic cells and recirculating lymphocytes and immunoglobulins are added. Immunostaining for two general neuronal markers, protein gene product 9.5 (PGP 9.5), a cytoplasmic ubiquitin C-terminal hydrolase, and synaptophysin, a calcium-binding four-span integral synaptic vesicle membrane glycoprotein, disclosed an abundant innervation of the large femoral lymphatic vessels in rats. This confirms and extends earlier findings based on nonspecific intravital methylene blue and silver impregnation staining methods. Nerves containing neuropeptide Y, C-flanking peptide of neuropeptide Y, and tyrosine hydroxylase, markers of noradrenergic postganglionic sympathetic fibers, were frequent whereas immunoreactivity to vasoactive intestinal peptide, a neuropeptide present in many cholinergic parasympathetic nerve fibers, was sparse suggesting possible sympathetic and parasympathetic influences. Futhermore, calcitonin gene-related peptide- and substance P-containing fibers were also present in the walls of lymphatic vessels suggesting a possible sensory influence in the coordinated myogenic responses. By comparison to normal light microscopy, confocal microscopy was found useful to trace the perihilar penetration of blood and afferent lymphatic vessels in lymph nodes. PGP 9.5-immunoreactive fibers were found in and around lymph nodes suggesting that there is a neural regulation of lymphoid node function. Because of their distribution, peptide-containing nerves may participate in regulating the capacity of the lymphatic pumping activity, and may possibly exert paracrine effects on lymphocytes.

Cerebral Glycoprotein Synthesis and Long-term Memory Formation in the Chick (Gallus domesticus) Following Passive Avoidance Training Depends on the Nature of the Aversive Stimulus.

Chicks that peck a small bright bead coated in a distateful substance can learn in a single trial to subsequently avoid a similar bead. The taste aversant commonly used is methyl anthranilate, which also has a strong pervasive odour. We have compared the efficacy of methyl anthranilate and the apparently odourless quinine as aversants. Methyl anthranilate-trained chicks learnt the task and the memory apparently persisted undiminished for at least 24 h. Quinine-trained chicks exhibited a memory for the task similar to that of methyl anthranilate-trained chicks 45 min after training, this thereafter declined until, at 24 h after training, they showed no recall. We investigated the incorporation of a radio-labelled synaptic membrane glycoprotein precursor, [3H]fucose, into three regions of the chick forebrain; two of these regions have previously been implicated in learning using methyl anthranilate as the aversant. There was a significant increase in [3H]fucose incorporation into the left lateral cerebral area and numerically similar, but non-significant, increases in the intermediate part of the medial hyperstriatum ventrale and lobus parolfactorius. There were no such increases in the right hemisphere of methyl anthranilate-trained chicks or any region of either hemisphere of quinine-trained chicks. Thus, the memory for methyl anthranilate is longer-lasting than that for quinine and is associated with increased fucosylation in the left cerebral hemisphere and although in the short-term, chicks can retain a memory of the one-trial passive avoidance task with quinine as the aversant, this does not result in a localized increase in cerebral [3H]fucose incorporation.

Svp25, a synaptic vesicle membrane glycoprotein from Torpedo electric organ that binds calcium and forms a homo-oligomeric complex.

Svp25 is a major glycoprotein of cholinergic synaptic vesicles isolated from the Torpedo electric organ. On SDS-PAGE svp25 migrates as a protein of Mr 25,000 and on two dimensional gel electrophoresis separates into several isoforms around a pI of 6.0. It binds concanavalin A and on phase separation with Triton X-114 behaves as an integral membrane protein. Svp25 represents a major vesicular 45Ca2(+)-binding protein. Under non-reducing conditions svp25 forms complexes of higher molecular weight which are multiples of 25,000. Svp25 is contained in the dense web of nerve terminal ramifications at the ventral side of the electroplaque cells. Colloidal gold labelling using a monospecific antibody confirms the selective association of the protein with synaptic vesicles. Although the function of the vesicular svp25 glycoprotein is not known, its ability to bind Ca2+ suggests that it is regulated by activation of the nerve terminal.

Characterization of a synaptic membrane glycoprotein, gp85, with a monoclonal antibody

Characterization of a synaptic membrane glycoprotein, gp85, with a monoclonal antibody

Characterization of gp 50, a major glycoprotein present in rat brain synaptic membranes, with a monoclonal antibody

Several cell lines secreting monoclonal antibodies (Mabs) against a major forebrain synaptic membrane (SM) glycoprotein, gp 50, have been raised. Western blots show that the Mabs react with a polypeptide doublet of M rs 49 and 45 kDa. These polypeptides exist solely in a concanavalin A (Con A) binding form. Removal of the Con A receptors by digestion with endo-β-N-acetylglucosaminidase H (endo H) lowers the M rs of the glycoprotein doublet to 36.5 and 34 kDa. Western blots of 2D polyacrylamide gels indicate that gp 50 exists in several isoforms. Solid phase radioimmunoassay (RIA) and Western blots of brain subcellular fractions show thet antigenic material to be concentrated in the SM fraction, but to be present in much lower amounts in synaptic junctions and postsynaptic densities. Gp 50 appears to be brain specific. Regional distribution studies show that it is present in all brain regions but is two-fold concentrated in cerebellum, brainstem and midbrain compared to forebrain. Immunocytochemical studies of several brain regions show that gp 50-like immunoreactivity is neuron specific and is concentrated in selected neuronal species, particularly granule cells. In both cerebellar and hippocampal granule cells gp 50-like immunoreactivity is localized in the perikarya and primary dendrites. Though immunocytochemistry did not show staining of synaptic regions this may be due to masking of the reactive epitope. The results are discussed in terms of the molecular properties of gp 50 and its subcellular localization in brain tissue.

Phosphorylation of synaptic membrane glycoproteins: the effects of Ca2+ and calmodulin.

Synaptic membranes were incubated with [gamma-32P]ATP, and glycoproteins were isolated by affinity chromatography on concanavalin A agarose. Glycoproteins accounted for 1.5-2.5% of the total 32P incorporated into synaptic membrane proteins. Ca2+ and calmodulin enhanced the phosphorylation of synaptic membrane glycoproteins approximately threefold. In the presence of Ca2+ and calmodulin, the rate of glycoprotein dephosphorylation was also increased three- to four-fold. Gel electrophoretic analysis identified several synaptic membrane glycoproteins that incorporated 32P, with the most highly labeled glycoprotein under basal phosphorylating conditions having an apparent Mr of 205,000 (gpiii). Ca2+ and calmodulin produced a marked increase in the phosphorylation of a glycoprotein with an apparent Mr of 180,000 (gpiv) and lesser increases in the labeling of three other glycoproteins. Membranes that had been labeled with [gamma-32P]ATP were extracted with Triton X-100 under conditions that yield a detergent-insoluble residue enriched in postsynaptic structures. The Triton X-100 insoluble residue accounted for 20-25% of the 32P associated with synaptic membrane glycoproteins. Gpiv and other glycoproteins, the phosphorylation of which was stimulated by calmodulin, were located exclusively in the Triton X-100 insoluble residue, whereas gpiii and other calmodulin-insensitive glycoproteins partitioned predominantly into the Triton X-100-soluble fraction. Phosphopeptide maps and phosphoamino acid analysis of gpiv isolated from synaptic membranes and a postsynaptic glycoprotein of apparent Mr of 180,000 (gp180) isolated from synaptic junctions indicated that the former protein was identical to the previously identified postsynaptic-specific gp180. In addition to phosphoserine and phosphothreonine, gpiv also contained phosphotyrosine, identifying it as a substrate for tyrosine-protein kinase as well as for Ca2+/calmodulin-dependent protein kinase.

Characterization of retrograde axonal transport of antibodies in central and peripheral neurons.

Retrograde axonal transport of antibodies against synaptic membrane glycoproteins was studied in the hypoglossal nerve and several CNS pathways of the rat. Injection into the tongue of polyclonal antibodies against synaptic membrane glycoproteins produced immunocytochemically labeled cells in the hypoglossal nucleus 4-5 hr later. Immunoreactive staining increased through 48 hr after injection and then declined. Injections of Fab preparations of the antibody gave labeling patterns indistinguishable from those of the whole antibody. The specificity of this method is shown by control studies in which antibodies against antigens that are not known to be present on the surface of presynaptic membranes were injected and gave no retrograde labeling. Retrograde labeling was also demonstrated in CNS pathways. However, labeling was never as intense as that seen in the hypoglossal nucleus, and some CNS pathways failed to show any retrograde labeling. Furthermore, retrograde labeling after control injections could be demonstrated in some cases. To determine if antibodies were also transported anterogradely, injections were made into the vitreous body of the eye, and the superior colliculus was processed for immunocytochemistry. Unlike wheat-germ agglutinin and several other tracers, antibodies were not found to be anterogradely transported in the optic nerve.

Low cerebrospinal fluid concentration of brain-specific protein D2 in patients with normal pressure hydrocephalus

The synaptic membrane glycoprotein D2 was measured in cerebrospinal fluid (CSF) and plasma in 13 patients with normal pressure hydrocephalus (NPH), in 14 patients with primary degenerative dementia of Alzheimer type (PDD), and in 24 patients without evidence of organic nervous disease (controls). Mean CSF D2 concentration was significantly lower in NPH patients: 299 ± 48 ng/ml (SEM) ( P < 0.001) than in PDD patients: 658 ± 50 ng/ml (SEM) and in controls 641 ± 45 ng/ml (SEM). Plasma D2 concentrations were higher in PDD patients compared with those found in controls. Determination of CSF D2 concentrations might be of diagnostic value in discrimination between patients with NPH and PDD patients with enlarged ventricles associated with diffuse brain atrophy.

Binding and uptake of concanavalin A into rat brain synaptosomes: evidence for synaptic vesicle recycling.

The specific binding of radioiodinated concanavalin A (125I-con A) to rat brain synaptosomes was shown to be saturable. In the presence of excess on A binding was rapid and was completed within 5 min (t1/2 was 25 s) at 37 degrees C, and at saturation the amount bound did not change over time. Under the electron microscope, concanavalin A-ferritin (con A-ft) bound to synaptosomes in two regions: in the extra-junctional plasma membrane and within the synaptic cleft of Gray type 1 and 2 synapses. Synaptosomes incubated with con A-ft at 37 degrees C internalized bound lectin by endocytosis through coated pits. Endocytosis took place in the extra-junctional membrane, because it can occur before con A-ft has penetrated into the synaptic cleft, and continued for a considerable time (more than 30 min) after saturation of the receptor(s). Synaptic vesicles, which have at least two con A receptors on the internal aspect of their membranes, and cisternae, become labelled. When exocytosis was induced in synaptosomes by K+ depolarizations, synaptic vesicle con A receptors became incorporated into the plasma membrane and were labelled with 125I-con A causing a 2.5-fold increase in con A binding that was Ca2+ dependent. These experiments thus provide evidence for the transient incorporation of synaptic vesicle membrane glycoproteins into the plasma membrane during transmitter release.

Developmental changes in the biosynthesis of sialoglycoprotein substrates for intrinsic synaptic sialidase.

Abstract: N′‐Acetyl‐d‐[6‐3H]mannosamine was administered to 13‐ and 28‐day‐old rats by intraventricular injection. At various time intervals following the injection, synaptic membranes were prepared and the incorporation of radiolabel into sialic acid residues released from endogenous glycoproteins and gangliosides by intrinsic sialidase determined. Radiolabel was incorporated into synaptic membrane gangliosides and glycoproteins, and at all times tested, &gt;90% of the label was associated with sialic acid. Sialic acid released from endogenous glycoproteins by intrinsic sialidase present in 28‐day membranes incorporated only 20–25% as much radiolabel per nmole as sialic acid released by mild acid hydrolysis or by exogenous neuraminidase. In contrast, sialic acid released from glycoproteins present in 13‐day‐old membranes by intrinsic sialidase, mild acid hydrolysis, or exogenous neuraminidase all were similarly labelled. At both ages the specific radioactivity (cpm/nmol) of sialic acid released from gangliosides by the intrinsic enzyme was similar to the total ganglioside sialic acid released by mild acid hydrolysis. The results identify glycoprotein substrates for intrinsic synaptic membrane sialidase as a distinct metabolic class in the mature brain and suggest the occurrence of a developmentally related change in the metabolism of these glycoproteins.

Effects of maternal nutritional stress on synaptic plasma membrane proteins and glycoproteins in offspring

An earlier study in this laboratory demonstrated a 33 to 50% deficiency of synaptic plasma membrane (SPM) proteins in the brains of developing offspring of rats that were fed a protein-deficient diet during lactation. We have demonstrated that an SPM deficit persisted if the nutritional stress (in this case postnatal protein-calorie deficiency) was continued on a long-term basis. If nutritional rehabilitation was instituted on the 21st postnatal day, however, the SPM deficit was reversible. We also demonstrated abnormalities in the synthesis of fucosylated SPM glycoproteins in the developing offspring rats subjected to protein or protein-calorie deficient diets and in rehabilitated offspring. These abnormalities were detected using sensitive double-label isotopic procedures. Few abnormalities were found in the relative distribution of the dye-binding capacity of SPM proteins from developing control, undernourished, or rehabilitated rats on polyacrylamide gels that had been stained with Coomassie blue. The potential cause and importance of abnormalities in synaptic glycoproteins is discussed.

Maternal ethanol consumption and synaptic membrane glycoproteins in offspring.

The present study was undertaken to assess the influences of chronic maternal ethanol consumption, prior to and during gestation, on the development of synaptic plasma membranes (SPMs) and on the synthesis of SPM glycoproteins in offspring. Comparisons were made between animals whose mothers were pair-fed a control or 6.6% (v/v) ethanol liquid diet in which protein accounted for either 18% (original) (C & E) or 21% (revised) (*C & *E) of the calories. In addition, groups of pups that were either cross-fostered (*C & *E) with chow-fed surrogate mothers or reverse cross-fostered (offspring of chow-fed mothers with *C & *E mothers) were examined. Ethanol and matched (same dietary group) control pups had comparable brain and body weights, brain protein content, and yield of SPM proteins during the 10-24 day age period examined. However, the yield of SPM proteins from ethanol and control offspring of and/or reared by the three groups of rat mothers that received the *E and *C liquid diets was greater than that of the offspring of rats that were fed the original diets. This suggests that the original diets were not nutritionally adequate for pregnant rats. Despite the fact that the content of SPM proteins was comparable in ethanol and matched control pups, the offspring of ethanol-treated rats had an abnormal distribution of [3H]- or [14C]-fucose-derived radioactivity among SPM glycoproteins. The SPM abnormalities were most severe in the non-cross-fostered offspring of E rats. No SPM glycoprotein abnormalities were found in the reverse cross-fostered group. The results of the present study demonstrate that chronic maternal ethanol consumption prior to parturition has a severe effect on the synthesis of SPM glycoproteins in developing offspring without affecting the content of SPMs per se. It also demonstrates the importance of optimizing the composition of liquid diets used to feed pregnant rats.

The effects of development on activity, specificity and endogenous substrates of synaptic membrane sialidase

Synaptic plasma membranes were prepared from cortices of rats varying in post-natal age between 4 and 30 days. Sialic acid associated with synaptic plasma membrane glycoproteins and gangliosides increased 75% and 50%, respectively, between 4 and 30 days. The amount of sialic acid released from these membrane constituents by intrinsic synaptic sialidase increased 2–4-fold over the same period. Incubation of synaptic plasma membranes with exogenous gangliosides or glycopeptides demonstrated a 2–3-fold incrase in sialidase activity during development. The major gangliosides present in synaptic plasma membranes at all ages were GT 1, GD 1a, GD 1b and GM 1. Intrinsic sialidase hydrolyzed 50–70% of endogenous GT 1 and GD 1a gangliosides at all ages. Endogenous GD 1b ganglioside was poorly hydrolyzed in young rats and its susceptibility to enzymic hydrolysis increased during development. When exogenous GD 1a and GD 1b were used as substrates a preferential increase in activity against GD 1b occured during development, the ratio of activity (GD 1a/GD 1b) decreasing from 3.6 to 1.6 between 7 and 30 days. 10- and 30-day-old synaptic plasma membranes contained complex mixtures of sialoglycoproteins, an increase in the relative concentrations of lower molecular weight sialoglycoproteins occuring development. Intrinsic sialidase present in 10- and 30-day-old synaptic plasma membranes acted upon all molecular weight classes of sialoglycoproteins.

Sulphate and fucose incorporation into synaptic membrane glycoproteins.

Sulphate and fucose incorporation into synaptic membrane glycoproteins.