Complete chemoenzymatic synthesis of the Forssman antigen using novel glycosyltransferases identified in Campylobacter jejuni and Pasteurella multocida.

We have identified an alpha1,4-galactosyltransferase (CgtD) and a beta1,3-N-acetylgalactosaminyltransferase (CgtE) in the lipooligosaccharide (LOS) locus of Campylobacter jejuni LIO87. Strains that carry these genes may have the capability of synthesizing mimics of the P blood group antigens of the globoseries glycolipids. We have also identified an alpha1,3-N-acetylgalactosaminyltransferase (Pm1138) from Pasteurella multocida Pm70, which is involved in the synthesis of an LOS-bound Forssman antigen mimic and represents the only known bacterial glycosyltransferase with this specificity. The genes encoding the three enzymes were cloned and expressed in Escherichia coli as soluble recombinant proteins that can be used to chemoenzymatically synthesize the Forssman antigen, and its biosynthetic precursors, in high yields.

Blood group ABO gene-encoded A transferase catalyzes the biosynthesis of FORS1 antigen of FORS system upon Met69Thr/Ser substitution.

Blood group A/B glycosyltransferases (AT/BTs) and Forssman glycolipid synthase (FS) are encoded by the evolutionarily related ABO (A/B alleles) and GBGT1 genes, respectively. AT/BT and FS catalyze the biosynthesis of A/B and Forssman (FORS1) oligosaccharide antigens that are responsible for the distinct blood group systems of ABO and FORS. Using genetic engineering, DNA transfection, and immunocytochemistry and immunocytometry, we have previously shown that the eukaryotic expression construct encoding human AT, whose LeuGlyGly tripeptide at codons 266 to 268 was replaced with FS-specific GlyGlyAla tripeptide, induced weak appearance of FORS1 antigen. Recently, we have shown that the human AT complementary DNA constructs deleting exons 3 or 4, but not exons 2 or 5, induced moderate expression of FORS1 antigen. The constructs containing both the GlyGlyAla substitution and the exon 3 or 4 deletion exhibited an increased FS activity. Here, we report another molecular mechanism in which an amino acid substitution at codon 69 from methionine to threonine or serine (Met69Thr/Ser) also modified enzymatic specificity and permitted FORS1 biosynthesis. Considering that codon 69 is the first amino acid of exon 5 and that the cointroduction of Met69Thr and GlyGlyAla substitutions also enhanced FS activity, the methionine substitutions may affect enzyme structure in a mode similar to the exon 3 or 4 deletion but distinct from the GlyGlyAla substitution.

Expression of Forssman antigen in human large intestine.

Forssman antigen is a commonly occurring heterophile antigen but is thought not to be present in most humans. Recent biochemical studies, however, have shown the presence of Forssman antigen in several forms of human cancer, including gastric, colon, and lung cancers. Immunohistochemical staining with both monoclonal and polyclonal antibodies has failed to demonstrate this antigen in human tissues. In this study we conclusively demonstrated the presence of Forssman antigen in cytoplasm of colon goblet cells, especially those in the so-called transitional mucosa adjacent to carcinoma. Specimens from 69 of 70 patients with colon cancer contained the antigen in goblet cells in transitional mucosae. The antigen was successfully demonstrated only after removal of sialic acid by alkaline hydrolysis-neuraminidase digestion. Localization of the antigen was quite different from that of Tn and human blood group A, B, and H antigens. This antigen was thought to be associated exclusively with globo-series glycolipids, and its existence in glycoproteins has not been conclusively demonstrated. However, the results presented here, based on proteolytic digestion and lipid extraction studies, strongly suggest that the antigen is also contained in glycoproteins.

Surface expression of Forssman glycosphingolipid antigen on murine bone marrow-derived macrophages is subject to both temporal and population-specific regulation and is modulated by IL-4 and IL-6.

The Forssman glycolipid antigen (Fo) has been shown to be a differentiation marker for mouse macrophages both in vivo and in vitro. In order to determine whether or not there is a relationship between stage of differentiation and Fo expression, we have analyzed the kinetics of Fo expression during the growth of cultured mouse bone marrow-derived macrophages (BMDM). BMDM were grown in serum free medium to avoid the possible influence of undefined serum factors. In this medium they could be maintained over a period of up to 20 days with cell yields comparable to those obtained with serum-supplemented media. Fo antigen was assayed with a specific antibody using both a whole cell ELISA and immunocytochemical staining of cells grown on slides. With increasing age in culture, BMDM showed a gradual quantitative increase in Fo expression and parallel increase in the Fo+ BMDM fraction from about 10% Fo+ cells on the 10th day of culture to a maximum of 50%-60% Fo+ cells between the 17th and 19th days. The temporal control over the development of the Fo+ cell fraction was intrinsic to BMDM maturation but was specific for Fo. During the same time period expression of MHC class II (Ia) remained consistently low, whereas expression of both Mac-1 (C3bR) and the macrophage-specific marker ER-BMDM-1 was always high. The interleukins IL-4 and especially IL-6 induced a premature expression of Fo at earlier stages of BMDM culture, but neither could promote further Fo expression once the intrinsically occurring maximum had been reached. No evidence in support of an autocrine regulation of Fo expression by IL-6 could be obtained, nor could a connection between cell cycle status and Fo expression be established. These data provide further evidence that Fo is a temporally regulated differentiation marker for a mouse macrophage subpopulation and for modulation of its expression by lymphokines.

Modulation of Forssman glycosphingolipid expression by murine macrophages: coinduction with class II MHC antigen by the lymphokines IL4 and IL6.

In contrast to murine spleen M phi, resident peritoneal M phi from health mice express very little Forssman glycolipid antigen (Fo). The following experiments suggest that Fo expression by peritoneal M phi may be associated with inflammation. Balb/c and CBA/J mice were given inflammatory stimuli by i.p. injection of live BCG, thioglycollate (TG), Corynebacterium parvum (CP), proteose peptone (PP), or LPS. Control animals received pyrogen-free saline. Expression of Fo and Ia antigen by peritoneal M phi was determined by immunofluorescence after 4 d. Application of TG or CP led to an up to 30-fold increase in Fo+, Ia+ double positive M phi over that in control animals. LPS caused mainly an increase in the percentage of double-positive M phi, whereas no effects were seen in BCG or PP treated animals. To clarify the possible involvement of cytokines in this process and to identify these, the effects of LPS and various cytokines on in vitro induction of Fo and Ia expression were studied in further experiments. LPS, IL6, and IL4 caused induction of up to 15% Fo+ and Ia+ M phi after a 4 d culture period. M phi colony stimulating factor (M-CSF) from lung-conditioned medium was also moderately active. IL1, TNF, and IL2 had no influence, whereas IFN-gamma only induced Ia. For a successful in vitro induction of Fo and Ia, a prior priming of the mice with PP appeared mandatory. This suggests that only M phi of a certain developmental stage can acquire Fo under the influence of the appropriate cytokines. The data may provide the first evidence for cytokine-mediated modulation of a glycolipid antigen of known chemical structure.

Decreased biosynthesis of Forssman glycolipid after retinoic acid-induced differentiation of mouse F9 teratocarcinoma cells. Lectin-affinity chromatography of the glycolipid-derived oligosaccharide.

Glycolipids synthesized by the mouse teratocarcinoma F9 cells and F9 cells (RA/F9 cells) induced to differentiate by a 3-day treatment with 0.1 microM all-trans-retinoic acid were analyzed. Both F9 cells and RA/F9 cells were incubated in media containing either D-[6-3H]galactose or D-[6-3H]glucosamine; the metabolically-radiolabeled glycolipids were isolated and the oligosaccharides were released from the glycolipids by ozonolysis and alkali fragmentation. From both cells, a single major pentasaccharide was isolated from the mixture of neutral [3H]oligosaccharides by affinity chromatography on a column of immobilized Helix pomatia agglutinin. The structure of this oligosaccharide was analyzed by methylation analysis and specific exoglycosidase treatments and identified as the Forssman pentasaccharide alpha-D-GalpNAc-(1----3)-beta-D-GalpNAc-(1----4)-alpha-D-Galp-(1----4)-b eta-D- Galp-(1----4)-D-Glc. There was a 3-4-fold decreased amount of the Forssman pentasaccharide from RA/F9 cells relative to F9 cells. In contrast, there were no major differences between these cells in the levels of globoside, the precursor to Forssman glycolipid. To investigate the basis for the decline in Forssman glycolipid synthesis upon differentiation, the activity of UDP-D-Gal-NAc:GbOse4Cer alpha-(1----3)-N-acetyl-D-galactosaminyltransferase (Forssman synthase) was determined in extracts of both the F9 and RA/F9 cells. The specific activity of Forssman synthase was approximately 70% lower in differentiated relative to the nondifferentiated cells. These data demonstrated that F9 cells synthesize authentic Forssman glycolipid, and that its expression and the activity of Forssman synthase were decreased following induced cellular differentiation.

Molecular modeling of glycosyltransferases involved in the biosynthesis of blood group A, blood group B, Forssman, and iGb3 antigens and their interaction with substrates.

A terminal alpha1-3 linked Gal or GalNAc sugar residue is the common structure found in several oligosaccharide antigens, such as blood groups A and B, the xeno-antigen, the Forssman antigen, and the isogloboside 3 (iGb3) glycolipid. The enzymes involved in the addition of this residue display strong amino acid sequence similarities, suggesting a common fold. From a recently solved crystal structure of the bovine alpha3-galactosyltransferase complexed with UDP, homology modeling methods were used to build the four other enzymes of this family in their locked conformation. Nucleotide-sugars, the Mn2+ ion, and oligosaccharide acceptors were docked in the models. Nine different amino acid regions are involved in the substrate binding sites. After geometry optimization of the complexes and analysis of the predicted structures, the basis of the specificities can be rationalized. In the nucleotide-sugar binding site, the specificity between Gal or GalNAc transferase activity is due to the relative size of two clue amino acids. In the acceptor site, the presence of up to three tryptophan residues define the complexity of the oligosaccharide that can be specifically recognized. The modeling study helps in rationalizing the crystallographic data obtained in this family and provides insights on the basis of substrate and donor recognition.