Anti-a-galactosyl immunoglobulin G (anti-Gal) is a natural antibody present in unusually high amounts in human sera. It constitutes as much as 1% of circulating immunoglobulin G in humans and displays a distinct specificity for the carbohydrate epitope galactosyl a(1-3) galactosyl (Gala1-3Gal). Recently, it has been suggested by various investigators that anti-Gal may be related to some autoimmune phenomena, since marked elevation of its titer was found in sera of patients with autoimmune thyroid disorders, rheumatoid arthritis, glomerulonephritis, and Chagas' disease. In view of the ubiquitous presence of anti-Gal in high titers in humans, throughout life, we hypothesized that, analogous with synthesis of anti-blood group antibodies against bacterial antigens, bacteria within normal intestinal flora may provide constant antigenic stimulation for the synthesis of anti-Gal. This hypothesis would imply that anti-Gal may bind to a variety of bacterial strains of human flora. In the present study, the interaction between affinity chromatography-purified anti-Gal and various bacterial strains was studied. By the use of a direct immunostaining assay and an enzyme-linked immunosorbent assay, anti-Gal was found to interact with a variety of Escherichia coli, Klebsiella, and Salmonella strains, some of which were isolates from normal stool. Furthermore, the anti-Gal-binding sites in some strains were found to be present on the carbohydrate portion of bacterial lipopolysaccharides. It is thus suggested that Gald-*>3Gal epitopes in the outer membranes of normal flora enterobacteria may provide a continuous source for antigenic stimulation. Since there is no immune tolerance to the Galal->3Gal carbohydrate structure in humans, anti-Gal seems to be constantly produced in response to these enterobacteria. In addition, bacteria which express Gala-*3Gal epitopes and which may adhere to various cells mediated binding of anti-Gal to human cell lines. These findings raise the possibility that anti-Gal may damage normal human tissues via inflammatory processes facilitated by bacterial Galo+l-*3Gal epitopes.

One percent of circulating IgG in humans recognizes galactose al,3 galactose residues (anti-Gal) and is synthesized in response to stimulation by enteric bacteria. In this study, we found that the prevalence of binding of anti-Gal to blood isolates is significantly higher than its binding to normal stool isolates. When anti-Gal bound onto the lipopolysaccharide of a representative blood isolate, Serratia marcescens #21, it blocked its alternative complement pathway (ACP) lysis and made the organism serum resistant. In contrast, when anti-Gal bound to the capsular polysaccharide of a serum sensitive Serratia, #7, it increased ACP killing ofthis strain. The mechanism of blockade of ACP lysis by anti-Gal did not involve a decrease in the number ofC3 molecules deposited onto Serratia #21 or an inhibition of the binding of C3b to its LPS, nor did it change the iC3b and C3d degradation products of bound C3b or prevent membrane attack complex formation on this organism. Our findings suggest that the effect ofanti-Gal on immune lysis is dependent on the bacterial outer membrane structure to which it binds. We postulate that anti-Gal may play a role in the survival of selected Enterobacteriace in Gram-negative sepsis by blocking ACP-mediated lysis of such bacteria by the nonimmune host, and that this effect depends on where antiGal finds its epitope on the bacterial outer membrane. (J. Clin.

No abstract

Anti-␣-galactosyl (anti-Gal) is a natural human serum antibody that binds to the carbohydrate Gal␣1,3Gal␤1,4GlcNAc-R (␣-galactosyl epitope) and is synthesized by 1% of circulating B lymphocytes in response to immune stimulation by enteric bacteria. We were able to purify secretory anti-Gal from human colostrum and bile by affinity chromatography on silica-linked Gal␣1,3Gal␤1,4GlcNAc. We found similar secretory anti-Gal antibodies in human milk, saliva, and vaginal washings. Secretory anti-Gal from milk and saliva was exclusively immunoglobulin A (IgA); that from colostrum and bile also contained IgG and IgM isotypes. Serum was also found to contain anti-Gal IgM and IgA in addition to the previously reported IgG. Anti-Gal IgA purified from colostrum and bile had both IgA1 and IgA2. Secretory anti-Gal from saliva, milk, colostrum, and bile agglutinated rabbit erythrocytes (RRBC) and bound to bovine thyroglobulin, both of which have abundant ␣-galactosyl epitopes. The RRBC-hemagglutinating capacity of human saliva, milk, bile, and serum was specifically adsorbed by immobilized Gal␣1,3Gal␤1,4GlcNAc but not by Gal␣1,4Gal␤1,4GlcNAc, Gal␤1,3GalNAc, Gal␤1,4GlcNAc, Gal␤1,4GlcNAc␣1,2Man, or Fuc␣1,2Gal␤1,4GlcNAc. No RRBC-hemagglutinating activity could be detected in rat milk, rat bile, cow milk, or rabbit bile, suggesting a restricted species distribution for secretory anti-Gal similar to that found for serum anti-Gal. Colostral anti-Gal IgA bound strongly to a sample of gram-negative bacteria isolated from the throats and stools of well children as well as to an Escherichia coli K-1 blood isolate. Colostral anti-Gal IgA inhibited the binding of a Neisseria meningitidis strain to human buccal epithelial cells, suggesting that this antibody may play a protective role at the mucosal surface.