Both the B cell-surface trigger(s) and the intracellular molecular mechanism(s) of somatic hypermutation in immunoglobulin (Ig) variable region genes remain unknown, partly because of the lack of a simple and reproducible in vitro model. Here, we show that upon surface immunoglobulin cross-linking followed by co-culture with activated cloned T cells, the Burkitt's lymphoma cell line BL2 is induced to mutate its IgV(H) gene. Repeated activation of BL2 cells increased the frequency of mutation. The in vitro-induced mutations, which do not affect the IgM constant region, are point mutations distributed over the entire V(H)DJ(H) gene segment and do not show evidence of antigen-driven selection.

We report the molecular characterization of five human monoclonal antibodies, BAB1-5 (BAB1: IgG(1); BAB4: IgG(2); BAB2, 3, 5: IgG(4)), with specificity for the major birch pollen allergen, Bet v 1. BAB1-5 were obtained after immunotherapy and contained a high degree of somatic mutations indicative of an antigen-driven affinity maturation process. While BAB1 inhibited the binding of patients IgE to Bet v 1, BAB2 increased IgE recognition of Bet v 1, and, even as Escherichia coli-expressed Fab, augmented Bet v 1-induced immediate type skin reactions. The demonstration that IgG antibodies can enhance allergen-induced allergic reactions is likely to explain the unpredictability of specific immunotherapy.

The B cell surface trigger(s) and the molecular mechanism(s) of somatic hypermutation remain unknown, partly because of the lack of amendable in vitro models. Recently, however, we reported that upon B cell receptor cross-linking and coculture with activated T cells, the Burkitt’s lymphoma cell line BL2 introduces mutations in its IgVH gene in vitro. We now confirm the relevance of our culture model by establishing that the entire spectrum of somatic mutations observed in vivo, including insertions and deletions, could be found in the DNA of BL2 cells. Additionally, we show that among four human B cell lines, only two with a centroblast-like phenotype can be induced to mutate. Triggering of somatic mutations in BL2 cells requires intimate T-B cell contacts and is independent of CD40-CD40-ligand (CD40L) interactions as shown by 1) the lack of effect of anti-CD40 and/or anti-CD40L blocking Abs on somatic mutation and 2) the ability of a CD40L-deficient T cell clone (isolated from an X-linked hyper-IgM syndrome patient) to induce somatic mutation in B cell receptor-engaged BL2 cells. Thus, our in vitro model reveals that T-B cell membrane interactions through surface molecules different from CD40-CD40L can trigger somatic hypermutation.

Birch pollen allergy is a very frequent pathology in Europe and North America. More than 95% of the tree pollen allergic patients display IgE reactivity against Bet v 1, the major birch pollen allergen. Starting with PBL from a patient desensitized by immunotherapy, we have generated five B cell lines (BAB1 to BAB5) that secrete human IgG mAbs of high affinity for Bet v 1. Although competition studies indicated that these human IgG mAb recognized different epitopes, broad cross-reactivity was found with Bet v 1 homologous allergens present in tree pollens and plant-derived foods. When tested for interference with allergic patients' IgE, BAB1 inhibited (by 80-100%) the binding of IgE to nitrocellulose-blotted Bet v 1, while BAB2 enhanced it. The biologic significance of the ability of BAB1 to interfere with patients' IgE binding is indicated by the finding that BAB1 completely inhibited Bet v 1-induced histamine release from allergic patients' basophils. Allergen-specific human IgG mAbs such as BAB1, which presents high blocking activity in both immunochemical and cellular IgE competition experiments, might have therapeutical application.

Background: Antigen recognition by antibodies of different isotypes can result in completely different effects as exemplified by Type I allergy. While the IgE–antibody–mediated release of biological mediators constitutes the immunopathological basis for the immediate symptoms observed in allergic patients, allergen–specific IgG antibodies are thought to have protective effects. Methods: Cell lines secreting five human monoclonal IgG antibodies (BAB1–BAB5) with specificity for the major birch pollen allergen Bet v 1 were established from a birch–pollen–allergic patient who had received birch– pollen–specific immunotherapy. The influence of the Bet v 1–specific IgG antibodies on IgE binding to Bet v 1 was investigated. BAB2 was expressed in Escherichia coli as recombinant Fab, purified and tested for its ability to modulate Bet v 1–induced immediate–type skin reactions. Results: The BAB antibodies belonged to different IgG subclasses (BAB1: IgG1; BAB2, BAB3, BAB5: IgG4; and BAB4: IgG2) reflecting a tendency towards Th2. BAB1 represented the only antibody which strongly blocked IgE binding to Bet v 1, whereas BAB 3–BAB5 had little effect on IgE binding. Surprisingly, natural BAB2 antibodies as well as recombinant BAB2 Fabs strongly enhanced IgE binding to Bet v 1 and Bet v 1–induced immediate–type skin reactions and thus represent ‘enhancing antibodies’. Conclusion: The demonstration that anti–allergen IgG antibodies can also enhance IgE binding to a given allergen explains the unpredictability of specific immunotherapy as well as the controversy on the role of IgG in atopy.