Axonogenesis in neuro-2a cells correlates with GM1 upregulation in the nuclear and plasma membranes.

GM1 ganglioside was previously shown to occur in the nuclear membrane, as well as the plasma membrane, of central nervous system (CNS) and peripheral nervous system (PNS) neurons undergoing morphological differentiation in culture. NG108-15 neuroblastoma cells showed the same phenomenon when induced to extend axon-like but not dendrite-like processes, although in both cases terminal differentiation was evidenced by failure of extended neurites to retract following washout of neuritogenic agent. The present study of Neuro-2a neuroblastoma cells subjected to similar treatments has revealed both similarities and differences compared to NG108-15 cells. Similar to the latter, Neuro-2a cells responded to neuraminidase and ionomycin with axon-like outgrowth together with upregulation of nuclear GM1, and to three other agents (retinoic acid, dibutyryl cyclic AMP, exogenous GM1) with dendrite-like outgrowth that was unaccompanied by nuclear GM1 increase. Although both cell types responded to low serum by extending neurites of mixed axonal-dendritic properties, Neuro-2a, in keeping with its greater tendency to extend some neurites of axonal character in low serum, showed elevated nuclear GM1 in a significant number of such differentiated cells. All three axonogenic agents induced parallel upregulation of GM1 in plasma-, nuclear-, and Golgi membranes, and these increases were stable to washout. Neurites generated in Neuro-2a cells by the three dendritogenic agents lacked stability, unlike those produced by the same agents in NG108-15 cells. This study also amplified the differences in response triggered by exogenous GM1 compared to that resulting from enzyme-mediated elevation of endogenous GM1.     

C6 cells express a sodium-calcium exchanger/GM1 complex in the nuclear envelope but have no exchanger in the plasma membrane: comparison to astrocytes

Previous work demonstrated the presence of an isoform of Na(+)/Ca(2+) exchanger in the nuclear envelope of neurons and NG108-15 cells that is tightly associated with GM1 ganglioside and potentiated by the latter. This contrasted with the Na(+)/Ca(2+) exchanger(s) in the plasma membrane, which were suggested to associate more loosely with GM1. To study these aspects of Na(+)/Ca(2+) exchanger expression in nonneuronal neural cells, we have examined nuclear and plasma membrane exchanger patterns in astrocytes and C6 cells, a glia-derived line. We find both cell types contain the tightly associated exchanger/GM1 complex in the nuclear envelope but, surprisingly, only astrocytes possess Na(+)/Ca(2+) exchanger activity in the plasma membrane. This is the first reported example of a cell (C6) with Na(+)/Ca(2+) exchangers in the nuclear envelope but not in the plasma membrane. RT-PCR established the presence of the NCX1 subtype in C6 cells and both NCX1 and NCX2 in astrocytes. Comparison was made with NG108-15 cells, which have Na(+)/Ca(2+) exchangers in both nuclear and plasma membranes, and Jurkat cells, which have no Na(+)/Ca(2+) exchanger in either membrane. Culturing of C6 cells in the presence dibutyryl-cAMP caused upregulation of a high molecular weight isoform of the exchanger together with GM1 in the nuclear envelope, resulting in significant elevation of Na(+)/Ca(2+) exchanger activity in the latter. Application of exogenous GM1 to nuclei from non-treated cells also potentiated exchanger activity, although to a lesser degree. The Na(+)/Ca(2+) exchanger/GM1 complex occurs in the inner membrane of the nuclear envelope, suggesting a functional role in transferring Ca(2+) between nucleoplasm and the envelope lumen.     

Correlation of gangliotetraose gangliosides with neurite forming potential of neuroblastoma cells.

Gangliosides of 11 different neuroblastoma cell lines, grown to confluence, were extracted and quantified with respect to: (a) total lipid-bound sialic acid, (b) total gangliotetraose family, and (c) GM1 content. The cultured cells were induced to grow neurites in 3 ways: (a) serum reduction, (b) exogenous ganglioside, and (c) retinoic acid. Neurite outgrowth was quantified in terms of % of cells bearing neurites and average number of neurites per cell. No correlation was observed between neurite outgrowth and total ganglioside concentration, but a reasonably good correlation was observed with respect to neuritogenesis and gangliotetraose content. When exogenous ganglioside was the stimulant the best correlation was with GM1, whereas retinoic acid-stimulated outgrowth was approximately proportional to GD1a content. The 'neurite minus' N1A-103 line, which had the lowest level of GM1, GD1a, and total gangliotetraose gangliosides, showed little if any response to any of the stimuli.     

Expression of gangliosides on glial and neuronal cells in normal and pathological adult human brain

Few studies have assessed the glycolipid phenotype of glial cells in the human central nervous system (CNS) in situ. We investigated by immunohistochemistry the expression and cellular distribution of a panel of gangliosides (GM1, GM2, acetyl-GM3, GD1a, GD1b, GD2, GD3, GT1b, GQ1b and the A2B5 antibody) in adult, human normal and pathological brain, namely multiple sclerosis (MS) and other neurological diseases (OND). In normal conditions, we found diffuse expression in the white matter of most gangliosides tested, with the exception of acetyl-GM3, GT1b and GQ1b. By double immunofluorescence with phenotypic markers, GM1 and GD1b were preferentially expressed on GFAP+ astrocytes, GD1a on NG2+ oligodendrocyte precursors, A2B5 immunostained both populations, while GD2 was selectively present on mature oligodendrocytes. In the gray matter, only GM1, GD2 and A2B5 were present on neuronal cells. Interestingly, those gangliosides present on astrocytes in normal conditions were preferentially expressed on NG2+ cells in chronic MS lesions and in OND. Selective expression of GT1b upon astrocytes and NG2+ cells was instead observed in MS lesions, but not in OND. The definition of the glycolipid phenotype of CNS glial cells may be useful to identify distinct biological glial subsets and provide insights on the potential autoantigenic role of gangliosides in CNS autoimmune diseases.     

Developmental appearance of nuclear GM1 in neurons of the central and peripheral nervous systems.

Previous studies demonstrated expression of GM1 ganglioside in the nuclear envelope of differentiating neuroblastoma cells and cultured cerebellar granule cells from neonatal rat brain. In the present study, relatively few of the latter cells were shown to possess a nucleus with appreciable GM1 during the first few days in culture, but increasing numbers of such cells possessed GM1-expressing nuclei as morphological differentiation progressed. This phenomenon reached a plateau by the 8th day in culture, approximately 90% of observed nuclei showing cytochemical evidence of GM1 at that time. Cerebral cortical neurons from embryonic rat brain in culture also gave clear evidence of GM1 in the nuclear membrane. Similar results were obtained with cultured neurons from the superior cervical ganglion from embryonic rats, demonstrating developmental appearance of GM1 in the nuclear envelope of PNS neurons. Cytochemical evidence for GM1 in purified nuclei from freshly isolated cortical neurons of neonatal rat brain indicated that expression of nuclear GM1 is not an artifact of cell culture. Study of NG108-15 neuroblastoma x glioma hybrid cells showed upregulation of nuclear GM1 to lag somewhat behind neurite outgrowth, suggesting nuclear GM1 to have a functional role subsequent to onset of morphological differentiation.     

Upregulation of nuclear GM1 accompanies axon‐like,but not dendrite‐like,outgrowth in NG108–15 cells

Recent work has demonstrated that induced neurite outgrowth in neuroblastoma cells and spontaneous differentiation of primary neurons in culture are accompanied by upregulation of GM1 ganglioside in the nuclear envelope. Previous reports have depicted morphological variations in the nature of stimulated neurites resulting from different neuritogenic agents, and a recent study by this laboratory demonstrated that such stimulants could be divided into two categories: those which induce axon‐like neurites (group I) as opposed to those that stimulate dendrite‐like outgrowths (group II). The former includes KCl, ionomycin, neuraminidase, and cholera toxin B subunit (all agents which elevate intracellular Ca²⁺), while the latter group is comprised of retinoic acid, dibutyryl cAMP, exogenous GM1, and low serum treatment. The present study was undertaken to determine whether differences in neuritic phenotype could be correlated with upregulation of nuclear GM1. The neuroblastoma × glioma NG108–15 cell line was employed because of its ability to respond robustly to a variety of neuritogenic stimuli. It was found that although both groups of stimulants are capable of inducing stable neurites (terminal differentiation) in this cell line, nuclear GM1 is elevated only in the presence of group I stimulants. Thus, a correlation is indicated between axonogenesis and upregulation of GM1 in the nuclear envelope. Additionally, these two events appear to coincide with elevation of intracellular Ca²⁺. Conversion of cells to the differentiated phenotype, with or without nuclear GM1 elevation, was found to depend in some cases on concentration of stimulant and duration of treatment. J. Neurosci. Res. 55:107–118, 1999. © 1999 Wiley‐Liss, Inc.     

Lipid composition of human malignant brain tumors

Malignant transformation is characterized by the uncontrolled proliferation of cells. And changes in the composition of glycolipids, cell surface component which may be involved in regulation of cell growth, were often observed in the malignant transformation. In this study, cholesterol, lipid-bound phosphorus, cerebroside, sulfatide and ganglioside were quantitated in the tissue of 20 human malignant brain tumors (malignant glioma, 8; low grade glioma, 4; metastatic tumor, 7; malignant meningioma, 1). As compared with normal brain, all tumor tissue contained lower cholesterol, sialic acid, cerebroside and sulfatide. Metastatic brain tumor or glioma showed characteristic patterns in the content of ganglioside, cerebroside and sulfatide respectively. The ganglioside patterns of metastatic tumor or glioma contained a greater proportion of structurally simpler gangliosides than normal brain. And in metastatic tumor, GM3 was a major ganglioside. On the contrary, glioma had increased proportion of GM3 and GD3 gangliosides. High grade glioma such as Grade 3-4 contained higher proportion of GM3 and GD3, whereas low grade glioma (Grade 1-2) contained less proportion of GM3 and GD3.     

Gangliosides of cultured astroglia

Cultured astrocytes prepared from newborn rat brain and 13-day-old chick embryonic brain were analyzed qualitatively and quantitatively for ganglioside content. All preparations contained approximately the same total level: 2.4-3.4 micrograms N-acetylneuraminic acid (NeuAc)/mg protein. In contrast, the value for primary cultures of neurons from chick embryonic brain was 5.9. The non-hexosamine-containing species, GM3 and GD3, comprised 75-85% of the total in astroglial cultures, the remainder consisting mainly of structural types other than the gangliotetraose series; choleragenoid assay revealed the latter to be virtually absent or to comprise at most a few percent. Deficiency of gangliotetraose synthesizing ability was indicated by the very low level of UDP-GalNac:GM3 N-acetylgalactosaminyltransferase detected in the cells. Treatment of cultured astrocytes with astroglial growth factor 2 or dibutyryl cyclic AMP caused little if any change in quantity or pattern of gangliosides. The large majority of cells stained in a manner characteristic of astrocytes: positive for glial fibrillary acidic protein, negative for galactosyl ceramides. Staining with cholera toxin and anti-GM1 antibody was essentially negative, as was that with tetanus toxin, A2B5 monoclonal antibody, and antibody to GD3. All evidence thus points to cultured astrocytes of rat and chick brain containing appreciable gangliosides, most of which are GM3 and GD3 with the majority of the remainder comprising structures other than the gangliotetraose type.     

Derivatives of ganglioside GM1 as neuronotrophic agents: comparison of in vivo and in vitro effects

Exogenously administered gangliosides have been shown to behave as neuronotrophic/neuritogenic agents in a variety of cell culture systems and animal models, but it is not known whether they operate by the same mechanism in vivo and in vitro. To probe this question we have employed two derivatives of GM1 lacking the negative charge: the methyl ester (GM1-CH3) and the NaBH4 reduction product of the latter (GM1-OH) in which the carboxyl group is replaced by a primary alcohol. Both derivatives proved to be as neuritogenic as GM1 in 3 cell culture systems: neuro-2A cels, PC12 cells and explanted dorsal root ganglia. However, GM1-OH proved ineffective when applied to two animal models involving reduction of cholinergic markers in: (a) hippocampus following lesion of the lateral fimbria and (b) nucleus basalis magnocellularis following cortical lesion; GM1-CH3 showed marginal activity in (a) but more in (b), possibly owing to slow hydrolysis to GM1 which was highly active in both animal models. These results indicate the necessity of a negative change on the ganglioside molecule for in vivo but not in vitro activity and point to different mechanisms for the trophic effects of exogenous gangliosides.     

Protection of neuro-2a cells against calcium ionophore cytotoxicity by gangliosides.

Gangliosides are known to assert both neuritogenic and neuroprotective effects when applied to a variety of neuroblastoma and primary neuronal cultures. We have developed a model employing Neuro-2a neuroblastoma cells with Ca2+ ionophore A23187 as neurotoxic agent causing neurite retraction and eventual cell death. Gangliosides attenuated the toxicity of this substance, increasing both cell survival and neurite stability. In one series of experiments, cells were exposed to A23187 for 24 hr and then incubated in fresh medium (washout) for 18 hr; gangliosides were present at varying times. The paradigm in which cells were only preincubated (2 hr) with ganglioside provided no benefit, nor did incubation of the cells in both ionophore and ganglioside during the 24-hr exposure period. Significant protection was achieved by exposing the cells to ganglioside after washout of A23187, or continuously throughout the whole period. Bovine brain ganglioside mixture and the four major components (GM1, GD1a, GD1b, GT1b) applied individually were all effective. By contrast, GM3 and GM1-alcohol, a neutral derivative of GM1, provided little or no protection. Dichlorobenzamil, an inhibitor of the Na(+)-Ca2+ exchanger, tended to block the neurite stabilizing effect of gangliosides, suggesting that the mechanism might involve potentiation of this antiporter.     

Gangliosides in cerebrospinal fluid in children with autism spectrum disorders

Gangliosides are sialic acid-containing glycolipids found in all cells, especially abundant in nerve cells and mainly situated on outer-membrane surfaces. The aim of this study was to provide data on the concentration of gangliosides in the CSF of children and adolescents with autism spectrum disorders (ASD) - 66 with autistic disorder, and 19 with other autism spectrum disorders. The comparison group consisted of 29 children and adolescents, whose CSF had been sampled to exclude acute infectious CNS disorder. The concentrations of the gangliosides GM1, GD1a, GD1b, and GT1b were determined using a microimmunoaffinity technique. The ASD group had a significantly higher concentration of ganglioside GM1 compared with the comparison group. The GM1 increase could not be explained as secondary to other clinical factors. Mean ganglioside levels did not differentiate subgroups with autistic disorder and those with a more atypical clinical picture, nor subgroups with known medical disorders and those with idiopathic autism. Altered patterns of gangliosides in the CNS might reflect important correlates of pathogenesis in autism.     

Neuroimmunology of gangliosides in human neurons and glial cells in culture

Gangliosides (sialic-acid-bearing glycolipids) have received attention in recent years because of their role in cell recognition phenomena, synaptic transmission, memory generation, and nerve regeneration in the fields of neurosciences. It is suggested that each brain region or each neural cell type may contain a specific and characteristic set of gangliosides. We have investigated the immunocytochemical localization of several classes of gangliosides that include GM1, GM4, GD3, and GQ gangliosides on the cell surface of various cell types found in human neural cell cultures with antibodies specific for these gangliosides. Cell cultures were obtained from adult human brains and fetal human dorsal root ganglia and spinal cord and cultured in vitro for the period up to 6 months and utilized for the ganglioside immunocytochemistry. It was demonstrated that GM1 ganglioside was present in all galactocerebroside-positive oligodendrocytes and most of glial fibrillary acid protein (GFAP)-positive astrocytes (80%), most of neurofilament-positive neurons (80%), 50-70% of Schwann cells, and 5-10% of fibronectin-positive fibroblasts; GM4 ganglioside could be detected in all oligodendrocytes, 80% of astrocytes, and 50% of Schwann cells, while no staining was found in neurons or fibroblasts; GD3 ganglioside was present in all oligodendrocytes and 5-10% of astrocytes but not in neurons, Schwann cells, or fibroblasts; and all of fetal CNS neurons and approximately 80-90% of fetal dorsal root ganglia (DRG) neurons and a small percentage of astrocytes (10-20% in fetal and less than 1% in adult astrocytes) was labeled by A2B5 antibody which is specific for GQ ganglioside, while this antibody did not stain cell surface of oligodendrocytes, Schwann cells, or fibroblasts. Three classes of gangliosides, GM1, GM4, and GD3 were found to be definite components of fetal and adult human oligodendroglial plasma membrane, while GM1 and GM4 gangliosides were detected on the surface of most astrocytes. Only a minor population of astrocytes from both fetal and adult human CNS contained GD3 and GQ gangliosides. Two classes of gangliosides, GM1 and GQ, were detected on the surface of fetal human neurons. More than half of fetal Schwann cells reacted to GM1 and GM4 antibodies but did not to GD3 or GQ antibodies. We recognized the presence of a specific and characteristic set of gangliosides on the cell surface of different human neural cell types and these findings should facilitate further investigation of the precise biological activity of these gangliosides.     

GD3 ganglioside is prevalent in fully differentiated neurons from rat retina

Adult mammalian retinas contain unusually high amounts of GD3, a ganglioside of the lactosylceramide series. In this respect, they differ from adult avian retina and other regions of the adult avian and mammalian brain, where GD3 is a minor ganglioside and gangliosides of the gangliotetraosylceramide series (GM1, GD1a, GD1b, GT1b) are the predominant ones. We compare here the ganglioside patterns of rat, human, horse, and guinea pig retinas, which are known to differ in the degree of vascularization and astrocytic cell content. All these retinas showed a prevalence of pathway "b" gangliosides over pathway "a" gangliosides but showed no correlation between GD3 content and the degree of vascularization and astrocytic cell content. Immunostaining of rat retina sections showed the presence of GD3 in the inner and outer plexiform layers and also in the ganglion cell and inner nuclear layers. About 60% of the cells dissociated from rat retina showed immuno-colocalization of GD3 and the neuronal marker class III beta tubulin isotype or cholera toxin binding. All morphologically identifiable glial Muller cells coexpress GD3 and gangliotetraosylgangliosides. GD3 was a minor ganglioside among these axonally transported by ganglion cells in rats and guinea pigs, suggesting that it is either not synthesized by ganglion cells or, if so, it is restricted to the cell soma and/or dendritic tree. Our results demonstrate that, unlike neurons from avian retina and other regions of avian and mammalian brain, neurons from mammalian retina not only contain gangliosides of the gangliotetraosylceramide series but also keep a prevalence of gangliosides of the lactosylceramide series (GD3) when they are fully differentiated.     

Ganglioside alterations in guinea pig brains at end stages of experimental Creutzfeldt-Jakob disease.

Gangliosides were isolated from guinea pig brains at the end stages of experimental Creutzfeldt-Jakob disease. Quantitative analyses revealed marked decreases of ganglioside levels in pathologically devastated tissues such as cerebral cortex (-21%), basal ganglia and thalamus (-18%), and brain stem (-23%). The cerebellum revealed only minor pathological abnormalities and its ganglioside level remained unchanged. Thin-layer chromatography of the Creutzfeldt-Jakob brain gangliosides showed aberrant ganglioside patterns in all regions studied, including the cerebellum. With some exceptions, a trend in ganglioside pattern changes was detected which consisted of proliferation of GM3, GD3, GD2 and loss of GM1, GD1a, GD1b and GT1b.     

Gangliosides have a bimodal effect on DNA synthesis in U-1242 MG human glioma cells

GM1, GD1a, and GT1b inhibit both PDGF-stimulated and serum-stimulated DNA synthesis in Swiss 3T3 cells and the human glioma cell line U-1242 MG in a dose-dependent manner. The ganglioside inhibitory effect is counteracted in a dose-responsive fashion by serum such that ganglioside-induced inhibition is essentially abolished in 10% serum. Because of the potentially important role that gangliosides play in growth regulation of human gliomas, this phenomenon was studied in detail using U-1242 MG cells. Stimulation of DNA synthesis by low doses of serum in U-1242 MG cells is inhibited in a dose-responsive fashion by ganglioside GM1. However, serum itself counteracts the inhibitory effect of ganglioside in a dose responsive way. Kinetic analyses demonstrate that GM-1 competes with some components of serum for sites on U-1242 MG cells (Kb of GM1 =12.5 μM). On the other hand, GM1, GD1a, and GT1b stimulate DNA synthesis in quiescent U-1242 MG cells in both sparse and confluent conditions, indicating that ganglioside-stimulated DNA synthesis is dependent on the phase of cellular growth rather than cellular density. This growth stimulatory effect of gangliosides is more potent on quiescent, confluent cells than quiescent, sparse cells. These results demonstrate that exogenously added gangliosides can have opposite (bimodal) effects on progression of human glioma cells through the cell cycle depending upon the growth phase of the cells. © 1995 Wiley-Liss, Inc.     

Anti-GM1 ganglioside antibodies with differing fine specificities in patients with multifocal motor neuropathy

Antibodies to gangliosides were detected in sera from three of 19 patients with chronic inflammatory polyneuropathy (CIP) by a thin-layer chromatogram overlay technique. All three of the patients fell into a clinical subset of the group that had multifocal motor neuropathy, and in all three patients the antibodies reacted with GM1 ganglioside. However, the fine specificities of the antibodies differed as demonstrated by cross-reactivity with different gangliosides in each of the three patients. The antibodies in patient 1 reacted with GM1, GD1b, and asialo-GM1 suggesting that the terminal Gal(beta 1-3)GalNAc moiety that is common to these three glycolipids is an important part of the epitope(s). This was confirmed by showing reactivity of the antibodies with Gal(beta 1-3)GalNAc conjugated to bovine serum albumin. Patient 2 had antibodies that did not react with GD1b, but cross-reacted with GM2 ganglioside suggesting that the epitope(s) involved the inner portion of the oligosaccharide moiety that is shared between GM1 and GM2. Patient 3 had antibodies that reacted with GM1 and asialo-GM1, but they did not cross-react with either GD1b or GM2. These results provide further evidence for a relationship between motor nerve syndromes and anti-GM1 antibodies and also suggest that GM1 could be a principal target antigen since other reactive gangliosides differed among the patients. However, the possible pathogenic effects of anti-GM1 antibodies on motor nerves remain to be established.     

Ganglioside GM1 and GM3 in early human brain development: An immunocytochemical study

The distribution of GM1 and GM3 gangliosides in human brain development between gestational week (g.w.) 6 and 15 was demonstrated by an immunocytochemical approach using polyclonal anti-GM1 and anti-GM3 antibodies. The first appearance of GM1- and GM3-positive cells was recorded as early as in g.w. 6. Both antibodies labeled the cells in the ventricular zone of the telencephalic wall, with radially oriented fibers toward the pial surface, which represent radial glia cells with glia fibers. The intensive GM3 immunoreactivity was also exhibited in proliferating cells in the ventricular zone between g.w. 6 and 12. During the period from g.w. 12 to 15, characterized by a rapid multiplication of neurons and glia cells, an increased number of GM1- and GM3-positive cells was observed. Prominent GM1 ganglioside staining was observed at the surface of the cell bodies in the ventricular zone. Besides surface labeling in migrating cells, GM1 immunoreactivity was identified inside the soma in the regions of cortical plate and subplate. GM1 immunoreactivity was more pronounced on the membrane of neuronal cells migrating along radial glia fibers, especially at the contact site between neuronal and glial cells. The GM3 ganglioside was localized mostly inside the soma, showing a granular immunoreactivity pattern.Our observations confirm the presence of GM1 and GM3 gangliosides in neuronal and glial cells in early human brain development. The involvement, especially of GM1 ganglioside in glia-neuronal contacts during migration of neuroblasts to their final destination, as well as the presence of GM3 ganglioside in proliferative cells in the ventricular zone of the telencephalic wall was also recorded.     

Cytosolic gangliosides of rat brain: their fractionation into protein-bound complexes of different ganglioside compositions

The sialic acid moiety of rat brain cytosolic gangliosides was radiolabeled by intracranial injection of N-(3H)acetylmannosamine. Upon ammonium sulphate fractionation, Sepharose 6B gel filtration, and hydroxylapatite-cellulose chromatography, ganglioside-bound radioactivity of brain cytosolic extract followed the behavior of protein and not that of purified gangliosides. This indicates that cytosolic gangliosides occur as ganglioside-protein complexes. By application of hydroxylapatite-cellulose column chromatography, fractions were obtained having different ganglioside composition. In particular, one fraction contained GM1, one GD1a, and one GT1b with a ganglioside homogeneity better than 95% in each fraction. This indicates the occurrence in brain cytosol of a GM1-protein complex, a GD1a-protein complex, and a GT1b-protein complex.     

Developmental changes of ganglioside expressions in postnatal rat cerebellar cortex

We previously described the differential distribution of gangliosides in adult rat brain as detected by specific antibodies. We report here the distribution of gangliosides during the development of postnatal rat cerebellum by an immunofluorescence technique with mouse monoclonal antibodies (mAbs). Eleven mAbs that specifically recognize each ganglioside were used. Our study revealed that the expression of each ganglioside changed dramatically during the development. GD3 and O-Ac-GD3 were expressed intensely in the external granular layer at 1, 5, and 10 days, whereas GD2 was firstly detected in the internal granular layer at 5 days and GD1b was diffusely detected throughout all layers of the cerebellar cortex at early postnatal days. GD2 and GD1b were more intensely expressed in the granular layer at 20, 30, and 80 days, suggesting that premature granule cells express GD3 and its derivative, O-Ac-GD3, whereas mature granule cells express GD2 and GD1b intensely. On the other hand, GM1 was exclusively detected in the external granular layer and the molecular layer at 1 and 5 days. The staining sites spread gradually from these outer layers into the internal granular layer and the white matter after 10 days. The positive cells in the external granular layer and the molecular layer appeared to be Bergmann glial cells and their radially ascending cytoplasmic processes. The intensity of the staining in these specialized astroglial cells decreased gradually during postnatal days. In contrast, the expression of GQ1b was very faint at birth, but gradually increased during the development and was detected intensely in the internal granular layer, particularly in the cerebellar glomeruli in adulthood, suggesting that GQ1b expression may be associated with synapse-related structures. The developmental changes of the expression of other gangliosides were also recognized in the postnatal rat cerebellum. These results suggest that specific gangliosides may play an important role in regulating the early events responsible for the orderly formation of the cerebellar cortex.