藤橋 浩太郎

Past studies have shown that freshly isolated human B cells from peripheral blood and tonsils do not express IL-6 receptors (IL-6R); however, mitogen or antigen activation of these B cells induces IL-6R and responsiveness to IL-6. In this study, we have shown that a high proportion of B cells enzymatically dissociated from human appendix, a gut-associated lymphoreticular tissue (GALT), expresses the IL-6R, and that recombinant human IL-6 induces significant increases in the number of Ig-producing cells. The recombinant human IL-6-induced increase in Ig-producing cells is restricted to the IgA isotype. Further, IgA2 is the major subclass; however, significant numbers of IgA1 producing cells are also seen. In contrast, human tonsillar and peripheral blood B cells express low levels of IL-6R, and exogenous IL-6 does not increase numbers of Ig-producing cells. When PBMC or tonsillar cells are stimulated with PWM, the former display an equal distribution of IgA1 and IgA2 secreting cells, while tonsillar B cells are mainly of the IgA1 subclass. The distribution of surface Ig-positive (sIg+) B cells in the appendix B cell population is sIgA+ > sIgG+ > sIgM+, and the sIgA+ B cells express higher levels of IL-6R when compared with sIgG+ or sIgM+ B cells. These studies show that human appendix contains B cell subsets that constitutively express IL-6R, and that a high proportion of these cells are committed to the IgA isotype. Furthermore, higher numbers of IL-6 responsive IgA2 B cells are present in the human appendix as compared to tonsils or PBMC.

It is well established that increased numbers of plasma cells occur in the localized tissues of chronic inflammatory diseases such as adult periodontitis, and enzymatic isolation has shown that most B lineage cells produce IgG-subclass with some IgA-subclass responses. It would be of importance to determine if excess production of cytokines in the localized lesion account for these responses and in the present study we have assessed gingival mononuclear cell (GMC) supernatants for cytokines that activate B cells including IL-6R expression and for levels of IL-6 present. Inasmuch as limited numbers (approximately 1 to 3 x 10(6) cells) of GMC were obtained from surgically removed tissues (approximately 400 mg), we have focused on the analysis of IL-6 production by GMC in this study. Further, initial evidence of additional cytokines that are produced by GMC and induce expression of IL-6R on resting B cells has been obtained. The GMC and PBMC from individual patients were cultured in the presence (or absence) of Con A. Higher levels of IL-6 were produced spontaneously by GMC when compared with Con A-stimulated PBMC. When PBMC cultures were supplemented with GMC supernatants obtained from the same patient, high numbers of spot-forming cells (SFC), mainly of IgG followed by IgA isotype, were seen. The induction of SFC by GMC supernatants was inhibited by incubation with a goat anti-human IL-6 antibody. When the effect of GMC supernatants on subclasses of PBMC SFC was determined, the response was IgG1 > IgG2 > IgG3 = IgG4 and IgA1 > IgA2, a pattern remarkably similar to the distribution of plasma cells in the GMC itself. To assess for cytokines in GMC supernatants that mediated B cell activation, supernatants containing anti-IL-6 were cultured with PBMC or purified B cells for 72 h. This treatment induced small proliferative B cell responses and elevated expression of IL-6R on B cells, but did not induce SFC responses. Further, incubation of B cells with GMC supernatants induced resting B cells (G0/G1) to enter the cell cycle (S and G2/M). Addition of human rIL-6 to these cultures on day 3 restored IgG- and IgA-subclass SFC responses by day 7. Cytokine-induced IL-6R expression also occurred in vivo because freshly isolated GMC expressed high levels of this receptor. These results show that GMC produce cytokines that induce B cell activation including IL-6R expression and secrete IL-6 that regulates B cell terminal differentiation into plasma cells of IgG- and IgA-subclasses. We are currently assessing the cytokines produced by GMC that induce the expression of IL-6R on B cells.

Our studies have given direct evidence that CD3+, CD4-, CD8+ T cells in the IELs can be separated into at least two subsets that produce IFN-gamma and/or IL-5 and may exhibit Th1- and Th2-type functions in the epithelium and in the underlying lamina propria region. Further, it was also shown that TCR1+ T cells in IELs are capable of producing IFN-gamma and/or IL-5. In addition to production of these cytokines, IELs derived from mice orally primed with TD antigen contain a subset of T cells which posses antigen-specific immunoregulatory function. CD3+; CD4-, CD8+ T cells isolated from IEL of TD antigen-primed mice abrogated systemic unresponsiveness to secondary-type responses including those of the IgA isotype upon adoptive transfer to mice with OT. Our studies further suggested that these CD3+, CD4-, CD8+ IELs use the gamma-delta-form of TCR (TCR1) for their immunoregulatory function.

In summary, we have shown that human appendix and murine PP B cells, freshly isolated from normal tissue, respond to IL-6 with significant increases in IgA SFC. Further, sIgA+ B cells from appendix express more IL-6R than is seen with B cells isolated from PBMC and spleen. When IgA subclass responses were measured, rhIL-6 induced both IgA1 and IgA2 SFC responses; however, 60-70 % of the total response was represented by the IgA2 subclass. Our studies suggest that the human appendix as well as murine PP are enriched sources for sIgA+ B cells which are responsive to cytokines such as IL-6 for induction of IgA plasma cell responses.