Structural characterization of gangliosides isolated from mullet milt using electrospray ionization-tandem mass spectrometry.

Electrospray ionization (ESI) coupled with tandem mass spectrometry has been used in conjunction with microwave-mediated saponification, periodate oxidation, and clostridial sialidase hydrolysis to enable detailed structural characterization of gangliosides and their derivatives present in mullet milt. The gangliosides extracted from mullet milt were determined to be GM3, GM3 lactone, GM3 methyl ester, and 9-O-acetyl GM3. For the major ganglioside GM3 and all GM3 derivatives, the ceramide composition was revealed to be C18:1/C16:0. GM3 with a C18:0/C16:0 ceramide was also found as a minor ganglioside. Both the ganglioside intramolecular ester and the ganglioside methyl ester (lacking carboxylic acid groups) showed dominant chloride attachment peaks (M + Cl)- in negative ion ESI-MS in addition to low intensity peaks corresponding to (M-H)-. GM3 and O-acetyl GM3 bearing carboxylic acid functions showed only (M-H)-. In positive ion ESI, GM3 and O-acetyl GM3 revealed (M + 2Na-H)+ peaks in addition to (M + Na)+, indicating free exchange of the carboxylic acid proton with a sodium cation, while the ganglioside intramolecular ester and ganglioside methyl ester with no acidic protons yielded only (M + Na)+. The strategy of employing ESI-MS to detect products of established wet chemical reactions represents a general approach for elucidation of ganglioside structural details.     

Alteration of ganglioside synthesis by GM3 synthase knockout in murine embryonic fibroblasts

To probe the functions of membrane gangliosides, the availability of ganglioside-depleted cells would be a valuable resource. To attempt to identify a useful genetic model of ganglioside depletion, we assessed ganglioside metabolism in murine GM3 synthase (GM3S)-/- knockout primary embryonic fibroblasts (MEF), because normal fibroblast gangliosides (GM3, GM2, GM1, and GD1a), all downstream products of GM3S, should be absent. We found that heterozygote MEF (GM3S+/-) did have a 36% reduced content of qualitatively normal gangliosides (7.0+/-0.8 nmol LBSA/mg cell protein; control: 11+/-1.6 nmol). However, two unexpected findings characterized the homozygous (GM3-/-) MEF. Despite complete knockout of GM3S, (i) GM3-/- MEF retained substantial ganglioside content (21% of normal or 2.3+/-1.1 nmol) and (ii) these gangliosides were entirely different from those of wild type MEF by HPTLC. Mass spectrometry identified them as GM1b, GalNAc-GM1b, and GD1alpha, containing both N-acetyl and N-glycolylneuraminic acid and diverse ceramide structures. All are products of the 0 pathway of ganglioside synthesis, not normally expressed in fibroblasts. The results suggest that complete, but not partial, inhibition of GM3 synthesis results in robust activation of an alternate pathway that may compensate for the complete absence of the products of GM3S.     

GM(1)-functionalized liposomes in a microtiter plate assay for cholera toxin in Vibrio cholerae culture samples

Vibrio cholerae, the causative agent for cholera, infects its host by expressing a protein consisting of two subunits: the pentameric cholera toxin B (CTB) and cholera toxin A (CTA). CTB frequently is used as an indicator of the presence of pathogenic V. cholerae and typically is detected using enzyme-linked immunosorbent assays (ELISAs). In lieu of an enzyme-linked detection method, we have developed GM(1) ganglioside-functionalized fluorescent dye-encapsulating liposomes for the detection of CTB produced by V. cholerae in a simple microtiter plate assay. Liposomes were compared with fluorescein-labeled antibodies and enzyme-linked secondary antibodies for quantification of purified CTB. A limit of detection for CTB using the liposomes was 340pg/ml, which was comparable to that using the ELISA but 18 times lower than that using the fluorescein-labeled anti-CTB antibodies for the same purpose. The sensitivity of the assay provided by the liposomes was substantial, and the working range improved when compared with that of the fluorescein-labeled antibodies and the ELISA. In addition, the liposomes required shorter assay times, exhibited greater precision, and were less expensive compared with the ELISA. The liposomes were optimized with respect to phospholipid and ganglioside concentrations. The optimized liposomes were then used to probe culture supernatants from V. cholerae El Tor C6706 grown in Dulbecco's modified Eagle's medium and AKI medium for the presence of CTB.     

Sialidase activities of cultured human fibroblasts and the metabolism of GM3 ganglioside

Free sialic acid has been found in the cell-conditioned medium of human foreskin fibroblasts. It is proposed that the accumulation of extracellular sialic acid may result from the hydrolysis of GM3 ganglioside on the cell surface of these fibroblasts. Sialidase activities with GM3 ganglioside and sialyllactitol as substrates were demonstrated in cell-conditioned medium, and the levels of their activities correlated positively with cell density. The GM3 sialidase activity at pH 4.5 was 4.1 and 38 pmol/h/ml of medium at sparse and confluent densities, respectively; the corresponding activities with sialyllactitol as the substrate were 12 and 75 pmol/h/ml of medium (pH 4.5). The pH versus activity profiles with GM3 as the substrate suggested the presence of a second sialidase with an optimal activity at pH 6.5 in the conditioned medium of preconfluent cells. This activity was virtually absent in the medium of contact-inhibited cells and could not be assayed with sialyllactitol as the substrate. The turnover of cell surface GM3 was assessed by pulse labeling human foreskin fibroblasts with a radioactive precursor of sialic acid ([1-14C]N-acetylmannosamine) and a radioactive precursor of ceramide ([3,3-3H2]serine). During a chase period of 24 h turnover of the doubly labeled cellular GM3 was observed; there was a loss of about 35% of the 14C-labeled sialic acid without any measureable loss of 3H-labeled ceramide from GM3. We have speculated that the enzyme-catalyzed removal of sialic acid from the GM3 ganglioside on the extracellular aspect of the plasma membrane may be a necessary event involved in the modulation of cell growth.     

Inhibitory effects of gangliosides on immune reactions of antibodies to neutral glycolipids in liposomes

Specific immune damage to liposomes containing Forssman or globoside glycolipid was inhibited when the liposomes also contained ganglioside. The activity of a human monoclonal Waldenstr?m macroglobulin antibody to Forssman glycolipid was inhibited by each of three gangliosides tested, GM3, GD1a and GD1b. Inhibition of the monoclonal antibody was dependent on the amount of ganglioside in the liposomes, and was diminished by reducing the relative amount of ganglioside. Inhibition also correlated positively with the number of ganglioside sialic acid groups, with inhibition by GT1b greater than GD1a greater than GM3. Naturally occurring human antibodies to globoside glycolipid were detected in 18% (9 out of 50) of normal human sera tested. Immune damage to liposomes induced by each of the three highest-reacting human anti-globoside sera was blocked by liposomal GM3. We conclude that gangliosides can strongly influence immune damage to membranes induced by antibody interactions with adjacent neutral glycolipids.     

Density-dependent recognition of cell surface GM3 by a certain anti-melanoma antibody, and GM3 lactone as a possible immunogen: requirements for tumor-associated antigen and immunogen

A mouse monoclonal antibody M2590, previously established after immunization of mice (C57BL/6) with syngeneic melanoma B16 cells and showing preferential reactivity with various types of melanoma over other tumor and normal cells or tissues, was shown to be directed to GM3 ganglioside. Since GM3 is widely distributed in essentially all types of animal cells, there is a conflict with the concept of a tumor-associated antigen and immunogen. Studies on the reactivity of M2590 antibody with various cells having different GM3 density at their cell surface, including cells treated with sialidase, liposomes, and solid-phase lipid layer containing different GM3 concentrations, have indicated that 1) reactivity of the antibody M2590 depends greatly on the density of GM3 exposed at the cell surface, on liposomes, or on solid phase; and 2) there is a threshold density that is recognized by the antibody in all-or-none fashion. In addition, the antibody M2590 reacts not only with GM3 but also with GM3 lactone, and the binding affinity of the antibody to GM3 lactone is strikingly higher than to GM3; however, the antibody does not react with GM3 ethyl ester. GM3 lactone was detected in melanoma as 3H-labeled GM3 gangliosidol after melanoma cells were directly treated with NaB[3H]4. A comparative immunization of BALB/c mice with GM3 and GM3 lactone showed that GM3 lactone is a much stronger immunogen than GM3, although the antibody elicited reacts with both GM3 and its lactone. Thus, the real immunogen could be GM3 lactone, although it is a minor membrane component.     

Differential expression of fucosyl GM1 and a disialoganglioside with a NeuAc alpha 2-6GalNAc linkage (GD1e) in various rat ascites hepatoma cells

A distinct difference in ganglioside composition among various rat ascites hepatomas and Yoshida sarcoma was observed on TLC-immunostaining with anti-fucosyl GM1 antibody, and chemical and enzymatic analyses. Yoshida sarcoma and ascites hepatomas, AH13, AH66F and AH66, but not the other 9 tumor cell lines investigated, specifically contained a disialoganglioside, NeuAc alpha 2-3Gal beta 1-3(NeuAc alpha 2-6)GalNAc beta 1-4Gal beta 1-4Glc beta 1-1ceramide (GD1e), whereas the 9 ascites hepatoma cells without GD1e contained fucosyl GM1. The differential expression of fucosyl GM1 and GD1e in various tumor cell lines indicates that different cell lineages express distinct metabolic pathways for gangliosides, and that the gangliosides are useful markers for distinguishing tumor cell lines.     

Aberrant fatty acyl alpha-hydroxylation in human neuroblastoma tumor gangliosides

Abnormalities of ganglioside structure characterize the neoplastic state, and aberrant glycosylation has been implicated as underlying many new tumor ganglioside structures. However, variations in ceramide structure can also result in novel tumor gangliosides. To address systematically this aspect of ganglioside metabolism, we have initiated a study of the structures of the ceramide species of an oligosaccharide-homogeneous human tumor-derived ganglioside, GM2. The ganglioside was isolated from neuroblastoma tissue and purified by normal-phase high pressure liquid chromatography. Marked ceramide heterogeneity was observed; 18 individual ceramide species of neuroblastoma GM2 were separated by reversed-phase high pressure liquid chromatography and collected. Their structures were determined by a combination of negative- and positive-ion fast atom bombardment mass spectrometry and collisionally activated dissociation tandem mass spectrometry of the underivatized gangliosides. The striking finding was the detection of alpha-hydroxylation of a significant fraction of each of the major fatty acid species (16:0, 18:0, 20:0, 22:0, and 24:1); alpha-hydroxylated species quantitatively represented almost one-fifth of the total tumor GM2 species. Fatty acyl hydroxylation was also detected in the ceramide of several other human tumor gangliosides. In contrast, as previously known, fatty acyl hydroxylation was not detected in the normal human brain gangliosides GM3, GM2, and GM1. We propose that aberrant fatty acid alpha-hydroxylation is a novel and sometimes quantitatively significant characteristic of human tumor ganglioside metabolism.     

Involvement of the acyl chain of ceramide in carbohydrate recognition by an anti-glycolipid monoclonal antibody: the case of an anti-melanoma antibody,M2590, to GM3-ganglioside

The effect of the chain length of the fatty acid residue of the ceramide moiety of ganglioside G_M3 on the binding ability of monoclonal antibody M2590, which is specific for the carbohydrate structure of G_M3-ganglioside, was examined by means of a direct binding assay on thin layer chromatography plates (TLC immunostaining) and a quantitative enzyme-linked immunosorbent assay (ELISA). Derivatives of G_M3 with a long fatty acid chain reacted with the M2590 antibody, but those with a short fatty acid chain showed no reaction in either assay system. These results suggested that the acyl fatty acid moiety of the ganglioside played an important role in the formation or maintenance of the antigenic structure of the carbohydrate moiety of the ganglioside.     

Incorporation of ganglioside analogues into fibroblast cell membranes. A spin-label study

The uptake of ganglioside analogues by a permanent mouse fibroblast cell line has been studied by radio-tracer techniques and ESR spectroscopy with 3H- and nitroxide-labeled compounds. Analogues of GM1, GM2, and GM3 monosialogangliosides and of GD1a and GD3 disialogangliosides were synthesized. The spin-label group was situated on the 5-, 9-, or 13-carbon atom of the C18 fatty acid chain, and the 3H label was in the carbohydrate moiety. Part of the ganglioside associated with the cells could be removed by trypsin treatment and was shown to consist of ganglioside micelles attached to the cell surface. The trypsin-resistant component displayed characteristic anisotropic ESR spectra which closely resembled those of the same spin-labeled analogues at low dilution in liposomes prepared from the extracted cell lipids. The flexibility gradient, polarity profile, and temperature dependence displayed by the spectra were similar to those found for fluid phospholipid bilayer model membranes, and the high effective order parameters suggested a location in the cell plasma membrane. Similar results were obtained for all the different ganglioside analogues, indicating a common anchoring region in the hydrophobic interior of the membrane. Under the incubation conditions used the amount of trypsin-resistant ganglioside analogue taken up by the cells was about 15 nmol/mg of cellular protein, irrespective of the nature of the oligosaccharide moiety. By use of the natural ganglioside [3H]GM3, the trypsin-resistant uptake was about 19 nmol/mg of cellular protein. Although these amounts are quite similar, the uptake kinetics differed between the true ganglioside GM3 and the ganglioside analogues.     

An interspecies comparison of gangliosides and neutral glycolipids in adrenal glands

Gangliosides and neutral glycolipids of adrenal glands of mouse, rat, guinea pig, rabbit, cat, pig, cow, monkey, and chicken were analyzed by thin layer chromatography (TLC). The major gangliosides common to all species had lactosylceramide in their core structure. GM3 containing N-acetylneuraminic acid (NeuAc) was the major ganglioside in rat, guinea pig, rabbit, and cat, whereas GM3 containing N-glycolylneuraminic acid (NeuGc) was the major one in mouse, cow, and monkey. GD3 was also detected in all species except mouse and GD3(NeuAc)2 was the major in pig adrenal gland. GM4(NeuAc) was detected in the adrenal glands of guinea pig and chicken but not in those of the other species. In the neutral glycolipid fractions, galactosylceramide, glucosylceramide, lactosylceramide, globotriaosylceramide and globotetraosylceramide were detected and the proportions of these glycolipids varied among the species. Guinea pig and chicken adrenal glands contained large amounts of galactosylceramide, this being consistent with the presence of GM4 in these two species. Globopentaosylceramide was detected in mouse, guinea pig, cat, and chicken by the TLC-immunostaining procedure.     

Absence of ganglioside GM1 in astroglial cells from newborn rat brain

An immunohistochemical method utilizing anti-ganglioside GM1 antiserum combined with the peroxidase-antiperoxidase technique was applied to a mixed cell population in primary cultures of newborn rat brain. Ganglioside GM1 was demonstrated to be present in neurons and oligodendroglia, but was absent in astroglia. This demonstration was confirmed using a newly developed biotinylated choleragen-avidin-peroxidase procedure. Primary cultures from newborn rat brain cells that had been subjected to a single treatment with trypsin (first passage) and then cultured for 14 days were predominately (95%) composed of astrocytes that stained positively for glial fibrillary acidic protein but were negative for GM1 ganglioside. This preparation contained only 0.34 nmol ganglioside NeuNAc per mg protein compared to 23.9 nmol gangliosidic NeuNAc/mg protein for a five day culture of newborn rat brain mixed cell culture that had not been subjected to passage. Prolongation of culture time from 5 to 21 days in the latter preparation reduced the ganglioside NeuNAc content to 4.9 nmol gangliosidic NeuNAc/mg protein as the proportion of astrocytes in the culture increased. Ganglioside GM1 could not be detected by TLC analysis of the lipid extract obtained from the “pure” astrocyte culture, although small amounts of GM3 and some polysialogangliosides were detected. About half of the label incorporated upon 24 h incubation of astrocytes in the presence of $\text{N-[}{}^3\text{H]acetylmannosammine}$ appeared in ganglioside GM3. It is concluded that astrocytes in mixed cell primary cultures from newborn rat brain, as well as astrocytes in astroglial preparations derived from such cultures, do not contain ganglioside GM1.     

Receptor-mediated interaction of ricin with the lipid bilayer of ganglioside GM1-liposomes

The interaction of ricin with ganglioside GM1 or glycoprotein containing liposomes was investigated. At neutral pH, ricin bound to galactose moieties on the surface of the liposomes to form ricin-liposomes complexes, but did not associate with their lipid bilayers. When these ricin-liposomes complexes were exposed to a pH below 5, ricin bound to GM1-liposomes became associated with the lipid bilayer, whereas ricin bound to glycoprotein-liposomes (containing human erythrocyte Band 3) was only rarely associated. Association of ricin with the lipid bilayer of GM1-liposomes did not occur in the presence of lactose, which inhibits the binding of ricin to ganglioside GM1. Using a hydrophobic probe, 8-amino-1-naphthalene sulfonic acid (ANS), it was revealed that an acidity below pH 5 resulted in exposure of hydrophobic regions on the ricin molecule. These results strongly suggest that association of ricin with the lipid bilayer of GM1-liposomes at acidic pH is mediated by the binding of ricin to ganglioside GM1 at neutral pH and occurs through interaction between the exposed hydrophobic region on the ricin molecule and the lipid bilayer of GM1-liposomes at low pH.     

A new ganglioside showing choleragenoid-binding activity in mouse spleen

A new ganglioside showing choleragenoid-binding activity was purified from mouse spleen and characterized. From the results of sugar-composition analysis, enzymatic hydrolysis, a permethylation study, 1H-NMR spectroscopy, and negative-ion fast atom bombardment mass spectrometry, the structure of the ganglioside was determined to be as follows: Gal beta 1-3GalNAc beta 1-4Gal beta 1-3GalNAc beta 1-4Gal beta 1-4Glc beta 1-1'ceramide 3----NeuGc alpha 2 This ganglioside contains a terminal tetrasaccharide structure identical with that of II3NeuGc alpha-Gg4Cer (GM1(NeuGc]. By means of a TLC-immunobinding assay and an enzyme-linked immunosorbent assay, the ganglioside was demonstrated to have almost the same choleragenoid-binding activity as GM1. Another ganglioside, that migrated faster than the new choleragenoid-binding ganglioside, was also purified from the same source material and identified as IV4GalNAc beta,IV3NeuGc alpha-Gg4Cer (GalNAc-GM1b(NeuGc]. Since, in the previous study, we demonstrated the existence of IV3NeuGc alpha-Gg4Cer (GM1b(NeuGc] in mouse spleen (Nakamura, K. et al. (1984) J. Biochem. 96, 949-957), the results of this study suggest that the new choleragenoid-binding ganglioside is synthesized from GM1b(NeuGc) through GalNAc-GM1b(NeuGc).     

Preparation and Specificity of 11 Monoclonal Antibodies to GM1 Ganglioside

Eleven monoclonal antibodies to GM1 ganglioside were prepared from hybridoma clones obtained by fusion of spleen cells from mice immunized with GM1 with mouse myeloma cells. When the reactivities of these 11 monoclonal antibodies were determined by enzyme-linked immunosorbent assay with six glycosphingolipids (GM1, GD1a, GD1b, GT1b, GM2, and asialo-GM1), they showed different degrees of specificity. From their reactivity patterns, they could be divided into three groups: Group 1, those that react only with GM1 (C3 and D3); Group 2, those that react predominantly with GM1 (C6, B6, D1, e1, g1, g9, and e12); and Group 3, those that show poor discrimination (h2 and A4). The clones differed in their biological activities.     

Protective effect of antiganglioside antibodies against experimental Trypanosoma brucei infection in mice

Liposome-associated ganglioside antigens (ganglioside GM1 or bovine brain gangliosides) were prepared to facilitate the potential protective efficacy for Trypanosoma brucei. Mice were immunized with liposome-associated ganglioside GM1 or bovine brain gangliosides intraperitoneally (i.p.). After immunization, significantly higher antigen-specific IgG and IgM antibodies were detected in sera than in the nonimmunized control group. When sera from immunized mice were analyzed for isotype distribution, antigen-specific IgG1, IgG2a, and IgG3 antibody responses were also noted. After immunization, mice were challenged i.p. with 1 x 10(2) cells of T. brucei. Sixty percentage of liposome-associated ganglioside GM1-immunized mice survived the infection, and all the mice immunized with bovine brain gangliosides-containing liposomes survived. However, all control mice died within 7 days after infection. These data demonstrate that liposomes containing ganglioside antigens have the potential usefulness for the induction of a protective immune response against T. brucei infection and suggest the possibility of developing vaccines that may ultimately be used for the prevention of trypanosomiasis.     

Ganglioside compositions of erythrocytes from various strains of inbred mice. Ocurrence of sialosylgalactosylceramide in red blood cells of inbred mice

The gangliosides from the erythrocytes of various strains of inbred mice were analyzed. The ganglioside compositions differed in different strains and the strains could be divided into 4 types on the basis of this difference. The main ganglioside is GM4 or sialosyl galactosyl ceramide in Type 1, and unidentified ganglioside in Type 2, GM4 and GM2 in almost equal amounts in Type 3, and GM2 in Type 4.     

Large liposomes containing ganglioside GM1 accumulate effectively in spleen

Large liposomes, with a composition of egg phosphatidylcholine, cholesterol and ganglioside GM1, prepared by an extrusion method, were injected intravenously into mice. After 24 h, up to 50% of injected dose was accumulated in spleen compared with about 15% in spleen for liposomes containing no GM1. The effect of GM1 on spleen accumulation of liposomes was liposome size dependent. Only relatively large liposomes (d greater than 300 nm) showed high accumulation; smaller liposomes were progressively less accumulated. The spleen accumulation increased with increasing injection dose of the liposomes. It was noted that the enhanced uptake by spleen was accompanied by a decrease in the liver uptake, but the total uptake of liposomes by liver and spleen was not dependent on the diameter of liposome or the presence of the ganglioside GM1. Autoradiographs of fixed and sectioned spleen using 125I-labeled tyraminylinulin as a content marker for the liposomes, showed that liposomes localized at the reticular meshwork of the red pulp. These results suggest that larger liposomes containing GM1 are filtered by the spleen during the circulation in blood. The smaller ones with a mean diameter of less than 100 nm are not retained by the filter. The function of GM1 is to prevent liposomes from a rapid uptake by the liver so that liposomes may circulate through the spleen and be filtered. These results, together with the observation that the liposome-entrapped proteins were degraded by the spleen, suggest the potential use of these liposomes for specific drug delivery to the spleen.     

GT1b in human metastatic brain tumors: GT1b as a brain metastasis-associated ganglioside1

We studied ganglioside expression in 12 human metastatic brain tumors metastasized from colon (4), renal (3), lung (2), esophagus (1), pancreas (1), and mammary (1) carcinomas. GM3 was the major common ganglioside expressed in brain metastatic tumor tissues, and GT1b was also present in all the metastatic brain tumor tissues. The latter was identified by TLC-immunostaining and characterized structurally by secondary ion mass spectrometry combined with ‘Far-Eastern blot’. The immunohistochemical analysis of frozen tissue sections confirmed localization of GT1b in the tumor cell membrane or cytosol. GT1b was shown to be expressed both in the primary colon carcinoma and the metastasis of a single patient by immunohistochemical procedure. In systemic carcinomas without brain metastasis, GM3 was a common major component, but no GT1b was detected. These findings indicate that GT1b is a brain metastasis-associated ganglioside. We speculate that the presence of GT1b would be a useful marker for estimating metastatic potentials to the brain.     

Expression of Gangliosides GD3 and 3'-isoLM1 in Autopsy Brains from Patients with Malignant Tumors

Abstract: Three autopsy brains from patients who succumbed to malignant gliomas have been analyzed in various regions with regard to their ganglioside content. The study focused on the gangliosides GD3 and 3'-isoLM1, which in a previous study of biopsies were found to be associated with these tumors. In particular, 3'-isoLM1, was suggested to be a marker for malignant gliomas. The highest concentrations (200–1,000 nmol of sialic acid/g wet weight) of GD3 was found in specimens of macroscopically pure tumor, where the proportion of GD3 was, at the most, 78% (range, 11–78%) of the total ganglioside sialic acid compared with <10% in normal brain tissue. The proportion of the total ganglioside sialic acid made up by GD3 was also elevated in the periphery of the tumor and in the same region in the opposite hemisphere, where no tumor cells were detected. In four of eight brain metastases of various carcinomas, GD3 was >10% of the total ganglioside sialic acid (range, 3–37%). The ganglioside 3'-isoLM1, as determined by TLC-enzyme-linked immunosorbent assay using a specific monoclonal antibody (SL-50), was not present at detectable levels in any of the macroscopically homogenous tumor areas. It was, however, found in the periphery of the tumor, in the corpus callosum, and at highest concentrations in the region of the opposite hemisphere corresponding to the tumor. The concentration varied between 0.1 and 6.0 nmol/g wet weight of tissue. The 3'-isoLM1 ganglioside was not detected in normal gray or white matter or in the normal corpus callosum, but in one of three breast cancer metastasis, one of two low differentiated cancer metastases, and one stomach cancer. The concentration was 1–4 nmol/g wet weight. These results indicate a unique distribution of the gangliosides GD3 and 3'-isoLM1 and suggest that they play distinct roles in interaction between tumor cells and brain.