Glycosphingolipid domains on cell plasma membrane

In this study we analyzed by immunofluorescence, laser confocal microscopy, immunoelectron microscopy and label fracture technique the ganglioside distribution on the plasma membrane of several different cell types: human peripheral blood lymphocytes (PBL), Molt-4 lymphoid cells, and NIH 3T3 fibroblasts, which mainly express monosialoganglioside GM3, and murine NS20Y neuroblastoma cells, which have been shown to express a high amount of monosialoganglioside GM2. Our observations showed an uneven distribution of both GM3 and GM2 on the plasma membrane of all cells, confirming the existence of ganglioside-enriched microdomains on the cell surface. Interestingly, in lymphoid cells the clustered immunolabeling appeared localized over both the microvillous and the nonvillous portions of the membrane. Similarly, in cells growing in monolayer, the clusters were distributed on both central and peripheral regions of the cell surface. Therefore, glycosphingolipid clusters do not appear confined to specific areas of the plasma membrane, implying general functions of these domains, which, as structural components of a cell membrane multimolecular signaling complex, may be involved in cell activation and adhesion, signal transduction and, when associated to caveolae, in endocytosis of specific molecules.

Antigenicity of the carbohydrate moiety of ganglioside GM3 having 3-O-acetyl ceramide

To elucidate the effect of a modification of ceramide on antigenicity of the carbohydrate of ganglioside, the reactivity of O-acetyl GM3 having 3-O-acetyl ceramide, which has been characterized as a glioma-related ganglioside, with monoclonal antibody M2590 was examined in comparison to that of non-acetylated GM3, by means of quantitative enzyme-linked immunosorbent assay, TLC-immunostaining and liposome immune lysis assay. In all these assay systems, O-acetyl GM3 showed less activity than GM3 as follows: GM3 was detected till 0.1 nmol in TLC-immunostaining, whereas O-acetyl GM3 could not be detected even at 0.25 nmol; the GM3 reaction was approximately twofold that of O-acetyl GM3 at each diluted point in the enzyme-linked immunosorbent assay; and 20% of the liposomes containing GM3 were lysed at 6 mol%, while liposomes containing O-acetyl GM3 did not lyse at that concentration. The lesser antigenicity of the sugar moiety of O-acetyl GM3 could be ascribed to the presence of an acetyl group in the ceramide at the 3-position of sphingosine.

Anti-GM3 antibodies activate calcium inflow and inhibit platelet-derived growth factor beta receptors (PDGFbetar) in T51B rat liver epithelial cells

Glycolipids expressed in the plasma membrane regulate a variety of cellular processes including intracellular calcium dynamics. We used flow cytometry to characterize the glycoconjugates on the plasma membrane of T51B liver epithelial cells. Antibodies against glycolipids found to be present were tested for their ability elevate intracellular calcium. An antibody against GM3 (DH2) nearly doubles intracellular calcium while an antibody against type II chains (1B2) increases calcium to nearly four times the baseline level, similar to levels obtained with epidermal growth factor (EGF). The antibodies stimulated calcium inflow but did not trigger calcium release from internal stores. In addition DH2 but not 1B2 inhibited platelet-derived growth factor beta receptor (PDGFbetar) function. This is the first demonstration of activation of calcium inflow by agents that bind GM3 and type II chains. The ganglioside-mediated calcium inflow is likely to stimulate secretion by these liver cells.

Increased circulating antiganglioside antibodies in primary and secondary progressive multiple sclerosis

Plasma samples from 70 patients with multiple sclerosis (MS), 41 patients with other neurological diseases (OND), and 38 healthy subjects were examined for antibodies against gangliosides GM1, GM3, GD1a, GD1b, and GD3 using enzyme-linked immunosorbent assays. The percentages of subjects with increased anti-GM3 responses were significantly higher in the primary progressive MS (56.3%) and secondary progressive MS (42.9%) groups than in the relapsing-remitting MS (2.9%), healthy subject (2.6%), and OND (14.6%) groups. Elevated antiganglioside antibodies may be secondary to axonal damage or may be a cause of axonal damage and accumulating disability in progressive MS. In either case, they may serve as a marker of axonal damage in MS.

Syngeneic anti-idiotypic monoclonal antibodies to an anti-NeuGc-containing ganglioside monoclonal antibody

An IgM monoclonal antibody (MAb), named P3, has the characteristic to react specifically with a broad battery of N-glycolyl containing-gangliosides and with antigens expressed on breast tumors. When this MAb was administered alone in syngeneic mice, an specific IgG anti-idiotypic antibody (Ab2) response was induced, this Ab2 response was increased when P3 MAb was injected coupled to a carrier protein and in the presence of Freund's adjuvant. Spleen cells from these mice were used in somatic-cell hybridization experiments, using the murine myeloma cell line P3-X63-Ag8.653 as fusion partner. Five Ab2 MAbs specific to P3 MAb were selected. These IgG1 Ab2 MAbs were able to block the binding of P3 MAb to GM3(NeuGc) ganglioside and to a human breast carcinoma cell line. Cross-blocking experiments demonstrated that these Ab2 MAbs are recognizing the same or very close sites on the Abl MAb. The five Ab2 MAbs were injected into syngeneic mice and four of them produced strong anti-anti-idiotypic antibody (Ab3) response. While these Ab2 MAbs were unable to generate Ab3 antibodies with the same antigenic specificity than P3 MAb, three of them induced antibodies bearing P3 MAb idiotopes (Ag-Id+ Ab3). These results demonstrated that these Ab2 MAbs are not "internal image" antibodies, but they could define "regulatory idiotopes."

Anti-tumor effect of internal image bearing anti-idiotypic monoclonal antibody in relation to GM3 ganglioside

Anti-idiotypic antibodies are a new type of useful tools for the possible treatment of cancer patients, since some act as antigen specific immunomodulators. Anti-idiotypic monoclonal antibody (anti-Id MAb) D704 (Ab2) was established which bore the internal image of the determinant defined by MAb M2590 (Ab1) against a sialic acid residue on GM3 ganglioside. In an in vivo syngeneic tumor system, anti-Id MAb D704 was more effective in preventing tumor progression, as compared with anti-GM3 MAb or no treatment. Significant suppression of tumor growth and prolongation of survival by administration of anti-Id MAb D704 in an animal group inoculated with 1 x 10(4)/mouse melanoma cells were seen, but not in a group inoculated with 5 x 10(4) cells/mouse. In an active specific immunotherapy protocol utilizing Ab2, the activity of anti-anti-Id antibodies (Ab3) specific for GM3 (antigen) which has a weak immunogenicity only, was maintained for more than 3 months. Ab2 generated cellular anti-tumor immune responses, including delayed type hypersensitivity (DTH) reaction. Immunohistological analysis indicated a marked infiltration of CD4 and CD8 positive cells into the DTH sites. Our results suggest that internal image bearing anti-Id MAbs have a therapeutic potential against tumors if the number of melanoma cells is relatively low or if hosts are at an early stage of melanoma progression.

Autoantibodies to gangliosides in sera of atherosclerotic patients

Using ELISA we studied the levels and clinical correlation of serum antibodies against gangliosides and 5-hydroxytryptamine (5-HT) in patients with atherosclerosis and clinical manifestations of cardiovascular disease. A range of 70-80% of the patients showed higher titers of anti-GM3(L) and anti-5HT as compared to normal serum. The anti-GM3(L) antibodies appeared to be directed mainly against GM3 present in platelets and were much less reactive against GM3 isolated from the aorta. We concluded that the antigens responsible for the elevated anti-GM3(L) and anti 5-HT levels in atherosclerotic sera are released by vessel-wall activated platelets. These results provide further evidence of on-going autoimmune processes in atherosclerosis. The content of total sialic (TS) and lipid-bound sialic acid (LBS) was measured in sera of patients with IHD and of similar numbers of healthy donors. In the patient groups the average TS and LBS concentration was about 25% higher than in the control group. These changes appeared to be associated with higher degrees of protein sialylation and larger amounts of LDL in the patient sera than in those of healthy controls.

Ratio of IgG:IgM antibodies to sialyl Lewis(x) and GM3 correlates with tumor growth after immunization with melanoma-cell vaccine with different adjuvants in mice

Human melanoma cells (from biopsies and culture) express sialyl-Lewis(x) and sialyl Lewis(a), the ligands for ECAM. These ligands may facilitate tumor progression and metastasis in human cancers. To test whether the antibodies to these ligands inhibit tumor progression, IgG and IgM responses to sLe(x) and sLe(a) were induced in C57BL/6j mice (n = 76) by immunization with human melanoma cells, with or without adjuvants (BCG, MPL, KLH). Control mice were treated with saline or BCG. Tumor growth and antigen expression were monitored after challenge with B16 mouse melanoma cells expressing sLe(x), sLe(a) and the ganglioside GM3. Tumor growth was reduced in mice immunized with BCG alone or cells with BCG or MPL, while tumors in mice receiving cells without adjuvants grew larger than in the control. Augmentation of IgM titers to sLe(x) and GM3 after immunization with BCG, or with cells with BCG or MPL correlated with retarded tumor growth, while increased IgG titers to sLe(x) significantly correlated with aggressive tumor growth in mice immunized with cells without adjuvants. SLe(x), sLe(a) and GM3 were expressed in tumors from mice treated with saline or BCG. SLe(x) expression, in particular, was lost in tumors growing in mice immunized with cells with or without adjuvants. Anti-sLe(x) antibodies may promote or prevent tumor growth by antigenic modulation or by cytotoxic killing of tumor cells. Since early anti-sLe(x) IgM correlated with tumor regression, in contrast to anti-sLe(x) IgG, it may serve as a potential early endpoint for the effectiveness of melanoma vaccines expressing the antigens.

Cell density regulates crypticity of GM3 ganglioside on human glioma cells

Human glioma cell line KG-1C contains GM3 ganglioside as its sole glycolipid. The degree of M2590 antibody binding to GM3 was found to be regulated by the cell density; the percentage of positive cells in FACS analysis decreased from approximately 20% to close to none as the cells increased their density from sparse to confluent. The contents of GM3 with different cell densities were consistent, being more than 0.4 micromol/g of the cellular weight, which was high enough to be recognized by the antibody. Trypsin treatment of the cells did not increase antibody reactivity. The extracted GM3 retained its antigenicity, being intensely stained with M2590 on a TLC plate; there was no change in chromatographic mobility either, indicating no modification of its chemical structure. The fluorescent microscope disclosed scattered dot-like staining of GM3, particularly at the periphery of the cells. We were able to expose cryptic GM3 fully within 12 h by dispersion of the cells to a sparse density. Surface labeling of GM3 with the use of limited sodium periodate oxidation of sialylated residue equally labeled GM3 either from the confluent cells or the sparse cells. Disassembly of actin filaments with cytochalasin B (10 microM) partially exposed cryptic GM3 of confluent cells, indicating reversibility of the crypticity. All together, the results indicate that cryptic GM3 actually exists on the cell surface, hidden from the surface not by other molecules but by other mechanisms associated with the cellular architecture. We are beginning to explore the possibility of selective localization of GM3 in small caves or folds of the cell membrane produced upon cell-to-cell contact.

Glycosphingolipid expression in human skeletal and heart muscle assessed by immunostaining thin-layer chromatography

In this study the comparative TLC immunostaining investigation of neutral GSLs and gangliosides from human skeletal and heart muscle is described. A panel of specific polyclonal and monoclonal antibodies as well as the GM1-specific choleragenoid were used for the overlay assays, combined with preceding neuraminidase treatment of gangliosides on TLC plates. This approach proved homologies but also quantitative and qualitative differences in the expression of ganglio-, globo- and neolacto-series neutral GSLs and gangliosides in these two types of striated muscle tissue within the same species. The main neutral GSL in skeletal muscle was LacCer, followed by GbOse3Cer, GbOse4Cer, nLcOse4Cer and monohexosylceramide, whereas in heart muscle GbOse3Cer and GbOse4Cer were the predominant neutral GSLs beside small quantities of LacCer, nLcOse4Cer and monohexosylceramide. No ganglio-series neutral GSLs and no Forssman GSL were found in either muscle tissue. GM3(Neu5Ac) was the major ganglioside, comprising almost 70% in skeletal and about 50% in cardiac muscle total gangliosides. GM2 was found in skeletal muscle only, while GD3 and GM1b-type gangliosides (GM1b and GD1 alpha) were undetectable in both tissues. GM1a-core gangliosides (GM1, GD1a, GD1b and GT1b) showed somewhat quantitative differences in each muscle; lactosamine-containing IV3Neu5Ac-nLcOse4Cer was detected in both specimens. Neutral GSLs were identified in TLC runs corresponding to e.g. 0.1 g muscle wet weight (GbOse3Cer, GbOse4Cer), and gangliosides GM3 and GM2 were elucidated in runs which corresponded to 0.2 g muscle tissue. Only 0.02 g and 0.004 g wet weight aliquots were necessary for unequivocal identification of neolacto-type and GM1-core gangliosides, respectively. Muscle is known for the lowest GSL concentration from all vertebrate tissues studied so far. Using the overlay technique, reliable GSL composition could be revealed, even from small muscle probes on a sub-orcinol and sub-resorcinol detection level.

Simultaneous expression by porcine aorta endothelial cells of glycosphingolipids bearing the major epitope for human xenoreactive antibodies (Gal alpha 1-3Gal), blood group H determinant and N-glycolylneuraminic acid

Glycosphingolipids were isolated from primary cultures of porcine endothelial cells labelled with 14C-galactose or 14C-glucosamine. They were characterized by their mobility on thin layer chromatogram, their sensitivity to exoglycosidases, and their labelling with antibodies. In addition to the major glycosphingolipids, globotetra- and globotriaosylceramide, minor ones were identified as penta- and heptaglycosylceramide of the neolactoseries terminated by either Gal alpha 1-3Gal- (xenoreactive epitope) or Fuc alpha 1-2Gal- (H determinant). Two gangliosides were found, GM3 and GD3, and N-glycolylneuraminic acid was their major sialic acid. Therefore, porcine endothelial cells differ from human endothelial cells by expression of glycosphingolipids that are absent in man: two Gal alpha 1-3Gal-terminated glycolipids recognized by human natural antibodies, and two N-glycolylneuraminic acid-terminated gangliosides which are potent immunogens.

Anti-GM3 (II3Neu5Ac-lactosylceramide) ganglioside antibody labels human fetal Purkinje neurons during the critical stage of cerebellar development

The ganglioside GM3 (II3Neu5Ac-lactosylceramide) represents a minor ganglioside in normal human brain compared to major gangliosides with gangliotetraose-backbone. In this study the presence of GM3 in three 23 and 24 weeks of gestation old human cerebella was demonstrated by immunostaining extracted gangliosides on thin-layer chromatography plate as well as by immunohistochemical analysis of cerebellar cryosections. During this stage of brain development GM3 was found to be dominantly expressed on cells corresponding to Purkinje neurons. Delipidation of histological sections with chloroform/methanol caused significant reduction of anti-GM3 immunostaining, thus confirming the prevalent ganglioside nature of this antigen. These results give evidence that (1) GM3 ganglioside is associated with a specific subset of human fetal cerebellar neurons during the critical development stage, and (2) a definite ganglioside in general is distributed to a specific subset of cells in normal human brain.

Overexpression of monosialoganglioside GM3 on lymphocyte plasma membrane in patients with HIV infection

SUMMARY

This study was undertaken to analyze both the GM3 expression on peripheral blood lymphocytes of HIV-infected patients and the relationship between ganglioside content and anti-GM3 reactivity. GM3 expression was determined as a percentage of lipid-bound sialic acid and by cytofluorimetric analysis in 25 AIDS patients, 20 anti-HIV+ asymptomatic subjects, 25 patients with different viral disease, and 25 healthy donors. GM3 distribution was analyzed by immunofluorescence and immunoelectron microscopy. A follow-up study to detect anti-lymphocytic GM3 antibodies was performed in progressive and nonprogressive anti-HIV+ subjects. Lymphocytes from HIV-infected patients showed a significant increase of plasma membrane GM3 content; no difference was found between CD4+ and CD8+ cells. Immunofluorescence and immunoelectron microscopic analysis showed that GM3 was distributed in large clusters over the cell plasma membrane. The follow-up study revealed that the occurrence of anti-lymphocytic GM3 antibodies was significantly higher in patients with progressive disease, compared with asymptomatic non-progressive subjects. These findings revealed that (1) the increased GM3 content in HIV-infected patients is detected at the plasma membrane level, (2) GM3 overexpression is able to induce an increased reactivity with anti-GM3 antibodies, and (3) the appearance of anti-lymphocytic GM3 antibodies in asymptomatic anti-HIV+ subjects could have prognostic relevance for the risk of developing AIDS.

Evidence for the existence of ganglioside molecules in the antigen of Entamoeba histolytica

Gangliosides were found to be present in Entamoeba histolytica. They were extracted from lyophilized trophozoites of the pathogenic strain HM-1:IMSS and purified by high performance thin-layer chromatography. Two resorcinol-positive bands, comigrating with GM2 and GD1a were demonstrated, revealing the existence of ganglioside molecules in Entamoeba histolytica. The GM2 content, determined as lipid-bound sialic acid, was 1.5 micrograms/10(8) amoebae, the content of the GD1a comigrating band was 0.32 microgram/10(8) amoebae. The identity of the GM2 comigrating band was confirmed by TLC immunostaining, using the monoclonal anti-GM2 antibody GMB28. Furthermore, six out of ten anti-amoeba positive sera selectively reacted with the GM2 comigrating band, as revealed by immunostaining on TLC plates. Absorption tests revealed that preincubation of anti-amoeba positive sera with standard GM2 was followed by a significant decrease in the reaction with amoeba trophozoites by indirect immunofluorescence. These results demonstrate that a GM2 comigrating component of Entamoeba histolytica may be one of the antigens responsible for the appearance of circulating antibodies in patients with amoebiasis.

Amplified ELISA to detect autoantibodies to N-glycolyl-GM3 ganglioside

Numerous immunochemical methods are now available for the detection of antibodies to gangliosides. An amplified ELISA method for detection of autoantibodies to NGcGM3 ganglioside in the sera of patients with various type of renal diseases was developed. IgM antibodies were found in 39 out of 53 sera of patients using 30 normal healthy blood donor as a negative control. For human IgG conjugate no reactivity to NGcGM3 was seen in the sera. Positive ELISA results were confirmed by TLC-immunostaining using GM3, NGcGM3, NGcGM2 and Standard bovine gangliosides (GM1, GD1a, GD1b and GT1b). All sera were also assayed for reactivity with GM3 in ELISA to determine the line specificity of these antibodies. Based on these results, a protocol for a sensitive and reproducible amplification ELISA system for serum anti-NGcGM3 antibodies in patients with renal or other diseases is presented. The ELISA method described here in appear to be useful adjunt to measure antiNGcGM3 antibodies in sera of patients with various type of renal or other diseases.

Generation of human monoclonal antibodies against ganglioside antigens and their applications in the diagnosis and therapy of cancer

Different approaches to generating human monoclonal antibodies (MAbs) against tumor-associated ganglioside antigens have been carried out in several laboratories. A specific goal addressed by our laboratory is to produce human MAbs to several ganglioside antigens of relevance as therapeutic targets, such as the GM2, GD2, GD3 and GM3 gangliosides in melanoma. In vitro immunization of human B lymphocytes from normal donors was performed using liposomes containing gangliosides as the immunizing antigen combined with either complete tetanus toxoid or a synthetic peptide corresponding to a T helper epitope to stimulate in vitro immunization. Specific human anti-ganglioside antibodies were obtained, indicating that the antibody response found in vitro was antigen-driven. To overcome the widely reported problems concerning stability of immunoglobulin production by the antibody-secreting cell lines, a method of positive selection using GM3-coated magnetic beads has been developed in order to rescue unstable clones. Development of new methods to reproducibly generate ganglioside-specific human MAbs will amplify the possibilities for diagnostic and therapeutic applications.

Major histocompatibility complex class I binding glycopeptides for the estimation of 'empty' class I molecules

Different forms of major histocompatibility complex (MHC) class I heavy chains are known to be expressed on the cell surface, including molecules which are functionally 'empty'. Direct peptide binding to cells is obvious during sensitization of target cells in vitro for cytotoxic T lymphocyte killing and 'empty' MHC-I molecules are comparatively abundant on TAP-1/2 peptide transporter mutant cells. In the present work we have estimated the fraction of 'empty' MHC class I molecules using glycosylated peptides and cellular staining with carbohydrate specific monoclonal antibodies. Synthetic Db and Kb binding peptides were coupled at different positions with different di- or trisaccharides, using different spacing between the carbohydrate and the peptide backbone. Binding of sugar specific mAbs was compared in ELISA and cellular assays. An optimal Db binding glycopeptide was used for comparative staining with anti-Db and anti-carbohydrate monoclonal antibodies to estimate fractions of 'empty' molecules on different T lymphoid cells. On activated normal T cells, a large fraction of Db molecules were found to be 'empty'. The functional role of such 'empty' MHC class I molecules on T cells is presently unclear. However, on antigen presenting cells they might participate in the antigen presentation process.

[Autoantibodies to ganglioside GM3 and serotonin in blood serum in atherosclerosis]

Using ELISA method, the sera from 17 patients with atherosclerosis and 13 normal controls were examined for ganglioside- and serotonin-reactive antibodies. Gangliosides from human liver (GM31) and human aorta (GM3a) as well as GM2, GM1, GT1b, human brain cerebrosides and the BSA-serotonin conjugate (5-HT) were used as antigens. A group of patients showed statistically significant higher levels of anti-GM31 (82%) and anti-5-HT (71%) as compared to the control group. Taking into account the identical fatty acid composition of GM3 from human liver and platelets, one can assume that antibodies are produced against blood clot gangliosides in atherosclerotic patients' sera. This conclusion is supported by a high correlation (r = 0.06, p < 0.01) between the level of antibodies to 5-HT and GM31 in the sera of all patients. The sera of three patients with the highest content of antibodies to GM31 were shown to contain antibodies to GM3a and GT1b which were absent in control sera. No reaction with brain cerebrosides or GM1 and GM2 was detected in the sera of the examined persons. The differences in sera reactions with GM31 and GM3a can be explained as resulting from differences in the chain length of fatty acid residues of the ceramide moiety of gangliosides. The data obtained confirm the fact that antibodies to aorta gangliosides appear in the sera of atherosclerotic patients. Thus, the formation of atherosclerotic plaques leading to platelet activation, clot formation and ganglioside accumulation in aortic cells increase the levels of anti-ganglioside GM3 antibodies.

Novel modification of ceramide: rat glioma ganglioside GM3 having 3-O-acetylated sphingenine

A novel O-acetylated GM3 containing 3-O-acetyl 4-sphingenine was isolated with one having a non-acetylated base from transplanted rat glioma tissue. The presence and position of the acetyl group were estimated by one- and two-dimensional proton nuclear magnetic resonance, and fast atom bombardment-mass spectrometries. In addition, the O-acetyl GM3 showed higher immunological activity toward anti-melanoma antibody in the presence of non-acetylated GM3 in complement-dependent liposome lysis than did non-acetylated or acetylated GM3 alone in the liposome, suggesting enhancement of immunological reactivity of the intact tumor cells by a small amount of O-acetyl GM3.

Clinical significance of anti-GM3 antibodies in recurrent pregnancy loss with elevated level of antiphospholipid antibodies

PROBLEM

The ganglioside-GM3 neutralizes the reactivity of antiphospholipid antibodies (APLs) to phospholipids in vitro. The question of whether anti-GM3 antibodies might exert influence in APLs-positive recurrent pregnancy loss patients who are undergoing prednisolone and aspirin (PSL/ASA) treatment was investigated.

METHOD

The anti-GM3 antibody assay of sera was accomplished by ELISA. Sera of 56 patients with recurrent pregnancy loss, including 30 APLs-positive cases given PSL/ASA treatment, were examined.

RESULTS

Patients positive for IgG or IgM type anti-GM3 antibodies constituted 13/30 (43%) of the APLs-positive group as compared with only 2/26 (8%) of those who were negative (P < 0.01). In pregnant women with APLs-positive treated with PSL/ASA, live births occurred in only 6/13 (46%) patients with detectable anti-GM3 antibodies, while in 16/17 (94%) who tested negative for anti-GM3 antibodies (P < 0.01).

CONCLUSIONS

This observation suggests the possibility that presence of anti-GM3 antibodies may be an indicator for determining the prognosis in recurrent pregnancy loss with elevated level of APLs.

Gangliosides from normal and osteoarthritic joints

A decrease in all major gangliosides, contrasting with a marked increase in the GM3 third band (3B-GM3), has been demonstrated in osteoarthritic fibrillated cartilage. We found the 3B-GM3 in osteoarthritic fibrillated, grossly intact and osteophytic human cartilage, but it was undetectable in pig and bovine cartilage, bone, synovium, synovial fluid white cells, erythrocytes, and various extracellular tissues. We made a similar finding with lactosyl ceramide, the GM3 neutral glycolipid precursor. The search for an antibody against the 3B-GM3, as a possible specific antigenic marker of osteoarthritis (OA), was unsuccessful. However, an antibody against a molecule very close to 3B-GM3 on high performance thin layer chromatography, which will probably be identified as a glycosylphosphatidylinositol, was found in the serum of some patients with OA.

Efficacy of tumor cell vaccine after incorporating monophosphoryl lipid A (MPL) in tumor cell membranes containing tumor-associated ganglioside

Murine B16 melanoma expresses the ganglioside GM3. GM3 shed from tumor cells is immunosuppressive and promotes tumor growth. Reduction or elimination of the shed GM3 could be therapeutic, and the anti-GM3 antibodies may reduce and clear the shed ganglioside. To test this hypothesis, mice were challenged with tumor cells, with or without inducing anti-GM3 antibody response. Since gangliosides are poor immunogens and T-cell independent antigens, an adjuvant (monophosphoryl lipid A (MPL), a non-toxic lipid A of Salmonella), directed against B-cells, was employed. MPL was incorporated onto liposomes and into the surface membrane of B16 mouse melanoma cells; both are rich in GM3. C57BL/6J mice immunized with MPL-liposomes or MPL-B16 cells responded with elevated levels of anti-GM3 IgM. Non-immunized mice or mice immunized with B16 cells alone or ganglioside GM3 alone (without MPL) elicited poor anti-GM3 IgM response, confirming the GM3's immunologic crypticity and MPL's immunopotentiating effect. MPL's immunopotentiating effect was improved by coupling it to melanoma cell membranes. C57BL/6J mice were immunized with irradiated B16 alone or MPL alone or MPL-conjugated irradiated B16. After three weekly immunizations, each mouse received a challenge dose of viable syngeneic B16. Neither MPL alone nor B16 alone had a significant effect on tumor growth or host survival; however, administration of MPL-conjugated B16 cells significantly prevented tumor growth and prolonged survival. Our results indicate that MPL-incorporated B16 cells augment the anti-GM3 IgM response, which may reverse GM3-induced immunosuppression by eliminating tumor-derived GM3, and restore immunocompetence.

Autoantibodies against ganglioside GM3 represent a portion of anti-lymphocyte antibodies in AIDS patients

In this study we analysed the relationship between anti-lymphocytic ganglioside antibodies and anti-lymphocyte antibodies in AIDS patients. Anti-lymphocytic ganglioside antibodies were detected by thin layer chromatography (TLC) immunostaining; three colour flow cytometry was used to analyse circulating antibodies against different lymphocyte subsets. Anti-lymphocytic ganglioside antibodies were detected in 23 out of 49 AIDS patients sera (46.9%). All positive sera reacted selectively with the GM3 comigrating band from AIDS lymphocytes. Twenty two out of the 23 anti-lymphocytic GM3 positive sera also had antibodies against CD4+T cells, versus 17/26 anti-GM3 negative. Furthermore, patients with lymphocytic GM3 antibodies showed a significantly higher antibody reactivity against CD4+ T cells than patients in which these antibodies were not detected. The absorption tests revealed that preincubation of positive sera with GM3 was followed by a decrease in the reaction with target lymphocytes. These findings suggest that anti-GM3 antibodies are a portion, but not the majority, of antibodies reacting with CD4+ T cells.

Glycosphingolipid antigens in cultured bovine brain microvascular endothelial cells: sulfoglucuronosyl paragloboside as a target of monoclonal IgM in demyelinative neuropathy [corrected]

Since a number of anti-glycosphingolipid (GSL) antibody activities have been demonstrated in patients with various neurological disorders, the presence of common antigens between brain microvascular endothelial cells (BMECs) and the nervous tissues presents a potential mechanism for the penetration of macromolecules from the circulation to the nervous system parenchyma. We first investigated GSL composition of cultured bovine BMECs. Bovine BMECs express GM3(NeuAc) and GM3(NeuGc) as the major gangliosides, and GM1, GD1a, GD1b, GT1b, as well as sialyl paragloboside and sialyl lactosaminylparagloboside as the minor species. Sulfoglucuronosyl paragloboside was also found to be a component of the BMEC acidic GSL fraction, but its concentration was lower in older cultures. On the other hand, the amounts of neutral GSLs were extremely low, consisting primarily of glucosylceramide. In addition, we analyzed the effect of anti-SGPG IgM antibody obtained from a patient of demyelinative polyneuropathy with macroglobulinemia against cultured BMECs. Permeability studies utilizing cocultured BMEC monolayers and rat astrocytes revealed that the antibody facilitated the leakage of [carboxy-14C]-inulin and 125I-labeled human IgM through BMEC monolayers. A direct cytotoxicity of this antibody against BMECs was also shown by a leakage study using [51Cr]-incorporated BMECs. This cytotoxicity depended on the concentration of the IgM antibody, and was almost completely blocked by preincubation with the pure antigen, sulfoglucuronosyl paragloboside. Our present study strongly supports the concept that immunological insults against BMECs induce the destruction or malfunction of the blood-nerve barrier, resulting in the penetration of the immunoglobulin molecule to attach peripheral nerve parenchyma.