Human Antibodies

During the past decade our knowledge of the cellular and molecular events associated with key immunological responses has been greatly advanced by the use of isolated subpopulations of immunocompetent cells, cloned cell lines and recombinant derived cytokines. Valuable as these studies have been they do not truly reflect the complex integrative events which take place in both primary and secondary lymphoid tissue both in vivo and in vitro. In order to address this problem we have developed a tissue culture procedure which is a modification of that previously used by others to study T cell maturation in the thymus. This involves culturing precision cut slices of human lymphoid tissue in a sponge culture system. Using this technique we have observed marked differences in both immunoglobulin and cytokine secretion between slices and suspensions of human spleen. In brief cultured slices (mitogen stimulated or otherwise) consistently secrete higher levels of immunoglobulin, IL-1β, IL-6, IL-8 and IL-11 and exhibit much lower proliferation than suspensions of the same tissue. Mitogen stimulated suspensions on the other hand secrete higher levels of IL-2, IL-4, IL-10 and TNFα than do slices. These differences are also observed at the intracellular cytokine level. Additional studies reveal that the immunoglobulin and cytokine secretion observed is largely due to the de novo synthesis of these molecules and not as a result of spontaneous secretion of preformed products. Furthermore immunoglobulin secretion in both slices and suspensions can be inhibited by the addition of specific antibodies to IL-1β, IL-6 and TNFα while IL-6 production can be differentially modulated by a variety of substances. Preliminary studies indicate that close interaction between B cells and stromal cells within ex plants accounts for some of the observed differences. This review article describes the basic technique, summarises the results we have obtained in this system and outlines the possible basis of the observed differences.

Anti-tyrosinase antibodies are found in the sera of patients with diffuse vitiligo, metastatic melanoma and in sera of patients with melanoma and hypopigmentation (MAH). The autoantigen is tyrosinase itself, the enzyme that participates in pigment (melanin) formation by both melanocytes and melanoma cells. The production of autoantibodies in both diseases is associated with the development of white patches on the patients' skin. The presence of these autoantibodies in patients with melanoma may suggest a better prognosis. Cross-antigenicity between melanoma cells and normal melanocytes is most probably the key mechanism leading to the appearance of MAN Anti-tyrosinase antibodies are absorbed by melanocytes and by melanoma cells in all the 3 situations (melanoma, vitiligo, MAH). However, since the production of antibodies in vitiligo exceeds that in melanoma or MAH, the antibodies are detected in significantly higher levels only in vitiligo. It is suggested here that anti-tyrosinase antibodies may be responsible, or at least participate in destruction of normal melanocytes during the immune response to melanoma antigens. This mechanism may be responsible for the phenomenon of MAH in patients with melanoma, and for the formation of the autoimmune vitiligo. Anti-tyrosinase antibodies may serve for two clinical applications. One is a marker for monitoring and follow up of patients with melanoma treated by immune therapy. The second is active (or passive) immunotherapy. We have recently shown that C57BL/6J mice immunized with tyrosinase generated a high titer of anti-tyrosinase antibodies, and following the inoculation of melanoma cells developed lower number of lung metastases, compared to the unvaccinated control group.

Immunoglobulin E plays a central role in mediating the pathology of allergic disease. Conversely, it is involved in the normal protective immune responses against parasite infection. Both these biological processes depend on interaction between the variable regions (V_H and V_L) of IgE antibodies and target antigen. It is now feasible to investigate the molecular nature of V_H regions used to encode IgE at the genetic level. Using this technology to analyze IgE in patients with asthma has revealed features which may have relevance for allergic disease. First, preferential choice of V_H genes, with dominance of the small V_H5 family, particularly the V_H32 gene, has been found. This may implicate a B cell superantigen (superallergen) selectively driving the use of these genes. Second, V_H5 genes in IgE are somatically mutated with clear hot spots of mutational activity. Mutational hotspots, which are a feature of the V_H5 gene, are supplemented in IgE by ongoing mutations which may be involved in affinity maturation. Third, a single B cell can switch to either IgE or IgG4, with both variants coexisting in blood. These findings may provide clues to the mechanism by which IgE is generated, and suggest options for therapeutic intervention.

Patients with primary biliary cirrhosis (PBC) frequently produce autoantibodies against gp210, an integral glycoprotein of the nuclear pores. This protein consists of three main domains: a large glycosylated lumenal domain, a single hydrophobic transmembrane segment and a short cytoplasmic tail. It has been previously shown that autoantibodies from PBC patients exclusively react with the cytoplasmic tail when recombinant rat gp210 expressed in Escherichia coli was used as antigen. Using human gp210 isolated from HeLa cells we found the lumenal domain as the major target. The aim of this study was to further characterize the dominant autoepitopes of gp210. Sera from 88 patients with autoimmune liver disease and 20 controls were used. Gp210 protein was digested with papain or endoglycosidase H and then subjected to immunoblotting. Autoantibodies against gp210 were detected in 12 of 43 (28%) PBC patients, but in none of the autoimmune hepatitis and control sera. Four of 12 (33%) anti-gp210 positive sera reacted with a fragment consisting of the cytoplasmic tail and 8 (66%) sera targeted an epitope located within the large lumenal domain. Furthermore, our data show that antigenic determinant is restricted to the 64 kD glycosylated amino-terminal fragment and that carbohydrate residues are an essential part of this novel epitope. We suggest that antigens possessing both epitopes namely: the glycosylated lumenal domain and the cytoplasmic tail should be used for screening tests in order to detect all sera with anti-gp210 specificity.

Interleukin-6 (IL-6) inhibitors are good potential therapeutic agents in human patients, and anti-IL-6 antibodies are among the best candidates. Here, we have successfully humanized mouse monoclonal antibody SK2, which specifically binds to IL-6 and strongly inhibits IL-6 functions. Since this antibody possesses N-linked carbohydrates on Asn-30 of V_H region, which seems to be very close to an antigen-binding site, influence of these carbohydrates on antigen-binding was investigated. A biosensor study showed that the mouse SK2 Fab and its deglycosylated fragments had almost equal Kd (K_on/K_off), 26.8 nM (1.05×10⁶12.81×10⁻²) and 24.7 nM (1.28×10⁶13.15×10⁻²), respectively. Furthermore, a mutant chimeric SK2 antibody, in which the N-glycosylation site was removed from the V_H region, showed a Kd of 11 nM, almost similar to that of the original chimeric SK2 antibody, determined by Scatchard analysis with ¹²⁵I-IL-6. These data indicate the carbohydrates of mouse SK2 V_H region do not significantly influence antigen-binding activity. In the next step, two versions of each humanized SK2 V_L and V_H regions were carefully designed based on the amino acid sequences of human REI and DAW, respectively. Only one alteration, Tyr to Phe, was made at position 71 in the two light chains, according to the canonical residue for L1. A N-glycosylation site was introduced on the two heavy chains, by changing Ser to Asn at position 30. All four combinations of humanized light and heavy chains could bind to IL-6 as well as the chimeric SK2 antibody. The light chain first version, however, could not efficiently inhibit IL-6 binding to its receptor, indicating the importance of the L1 loop conformation for the inhibitory activity of SK2 antibody. In contrast, both versions of the heavy chains were comparable, in yielding good humanized SK2 antibodies, suggesting that the glycosylation of the SK2 V_H region has no influence in recreating a functional antigen-binding site in this humanization.