Endogenous lectins (including selectins) are expected to define carbohydrate(CHO)-dependent adhesion between a cell showing surface expression of lectin and a different cell expressing the target CHO epitope. This mechanism, however, is apparently not universal because surface expression of lectins is seen only in certain cell types, and the structures defined by known lectins and selectins are highly restricted. In contrast, surface expression of a great number of CHOs [including glycosphingolipids(GSLs)] is known to change dramatically during ontogenesis and oncogenesis, and cell adhesion occurring at defined stages of ontogenesis can be inhibited by specific CHOs or GSLs. Our early studies focused on the mechanism of tight cell adhesion(compaction) of morula-stage embryo cells, as mediated by Le^x, and led to the discovery of Le^x-Le^x interaction in the presence of bivalent cation. Later, more systematic studies on adhesion of various GSLs (incorporated in [¹⁴C]cholesterol-liposomes) to GSLs coated on plastic plates showed strong H-H, H-Le^y, GM₃-Gg₃Cer, GM₃-LacCer, and sulfatide-GalCer interaction. We also demonstrated (i) specific interaction between GM₃-expressing B16 mouse melanoma cells and Gg₃Cer-expressing L5178 mouse leukemia cells based on GM₃-Gg₃Cer interaction, and (ii) adhesion of B16 variants to endothelial cells(ECs) based on GM₃-LacCer interaction. Relative degree of GM₃ expression in four B16 variants, and degree of adhesion to ECs or LacCer-coated plates, was found to be in the same order as degree of metastatic potential of the B16 variants. Initial adhesion of B16 cells to non-activated ECs is mediated by GM₃-LacCer interaction. Experimental data from a dynamic flow adhesion system supported this hypothesis. The GSL-GSL interactions demonstrated in these experimental systems may reflect a generalized phenomenon, i.e., involvement of specific CHO-CHO interaction in the earliest stages of cell-cell recognition and adhesion, which triggers subsequent (and stronger) adhesion events mediated by integrin and immunoglobulin family receptors.

View full abstract

Many glycoproteins are critically involved in T-cell activation. The structure/function relationship of individual proteins has been extensively studied. However, it is not known whether glycans on these glycoproteins play any role at all in the activation process. We have identified a panel of T-cell hybridoma mutants that cannot synthesize the glycosylphosphatidylinositol(GPI) anchor due to a defect in dolichol-phosphate-mannose(Dol-P-Man) biosynthesis. These mutants also cannot make full-length N-glycan precursors. The truncated oligosaccharide can be further processed to become complex, but not typical high mannose glycans. Interestingly, all of the Dol-P-Man mutants are defective in activation by antigen, superantigen, and Con A, despite normal T-cell receptor expression. The observed activation defect could be due to abnormal glycosylation, or due to the absence of GPI-anchored proteins, or both. In order to distinguish between these possibilities, the yeast Dol-P-Man synthase gene was stably transfected into the mutants, which restored full expression of surface GPI-anchored proteins. However, N-linked glycosylation was either partially or completely corrected in different transfectants. Correction of activation defects correlates well with the status of N-linked glycosylation, but not with the expression of GPI anchored proteins. Thus, N-linked glycosylation plays an important role in T cell activation and may provide a novel target for the development of immunosuppressive drugs.

View full abstract

Fractionation of proteins from perinatal rat brain was monitored with a neurite outgrowth assay. These studies resulted in the isolation of the heparin-binding proteins amphoterin(p30) and HB-GAM(p18). Cloning of the proteins from cDNA libraries of rat brain revealed highly charged, lysine-rich sequences for both proteins. In amphoterin the lysine residues are clustered to the N-terminal region, which is followed by a stretch of anionic amino acids. In HB-GAM densely spaced lysine residues are found both at the N-terminal and Cterminal end of the sequence. Amphoterin is abundantly expressed in early, embryonic rat brain and in immature cells in general. It is functionally associated with process outgrowth in developing neural cells. It binds plasminogen and tissue plasminogen activator(t-PA), and thereby effectively enhances plasminogen activation. We suggest that amphoterin targets and enhances the generation of plasmin activity, which is required for the penetration of axonal and other forms of cytoplasmic processes in tissues during development or during regeneration after tissue injury. HB-GAM is expressed later in brain and has a clearly more limited tissue distribution as compared to amphoterin. It is strongly expressed in rat brain during the perinatal developmental stage that corresponds to rapid outgrowth of axonal processes and formation of synaptic connections. We suggest that HB-GAM is an extracellular matrix-associated protein that enhances the growth and pattern formation of differentiating axons.

View full abstract

An antigen conjugated with pullulan (α-1, 4'-; α-1, 6'-glucan) was found to be a strong tolerogen for IgE antibody response and cell-mediated immune response, but a good immunogen for IgM and IgG antibody responses. The ability of an allergen-pullulan conjugate, such as Sugi Basic Protein(SBP)-pullulan conjugate, to elicit PCA or Arthus reaction was markedly reduced. Moreover, when rat skin sites presensitized with SBP-specific IgE antibody were treated with SBP-pullulan conjugate, the treatment blocked the PCA reactions following challenge with allergic SBP. An immune complex formed by SBP-pullulan and SBP-specific IgG/IgM antibody was incapable of supporting the efficient activation of complement systems, in both classical and alternative pathways. A hapten-or an antigen-pullulan conjugate induced suppressor T cells which functioned in induction-phase or effectorphase of cell mediated immunity. The suppressor T cells were found to function in a common pathway of suppressor T cells involving idiotypic interactions.

View full abstract

New insights in the distribution of lipids over the various intracellular organelles have been gained by a newly developed method for the immuno-localization of lipids. The method involves the labeling of an antigenic lipid with an antibody followed by a protein A-gold probe. Labeling is performed on ultrathin sections after freeze-substitution and low temperature embedding of aldehyde-fixed and cryo-protected cells. In a first study, Forssman glycolipid was localized in epithelial MDCK cells. Abundant labeling of the plasma membrane, the Golgi complex, and the membranes of the endocytic pathway was in contrast with the complete absence of labeling from mitochondrial and peroxisomal membranes. This demonstrated the specificity of the method. Recently, the method has also been applied to the(glycophospho)lipid phosphatidylinositol 4, 5-bisphosphate(PIP₂), the central component of the phosphoinositide signal transduction pathway. In MDCK cells, two monoclonal antibodies labeled the plasma membrane, whereas membranes of the Golgi complex were virtually negative. One monoclonal in addition yielded abundant labeling over the heterochromatin in the nuclear matrix, whereas the other monoclonal antibody yielded only minor labeling dispersed over the nuclear matrix. This labeling in the nuclear matrix suggests the occurrence of a non membrane-associated form of PIP₂, most likely complexed to a nuclear protein. The general applicability of the method is discussed.

View full abstract