B lymphocyte physiology: the beginning and the end.

Whereas lymphatic tissues were implicated as the chief sites for antibody production almost 100 years ago, three findings cemented the role of the plasma cell as the actual producer, namely the histological and tissue culture studies of Fagraeus, the beautiful immunofluorescent approach of Coons and his group, and the micromanipulation approach to the study of antibody formation by single cells introduced by Lederberg and myself. Proof that antibody-forming cells derived from B, rather than T, cells had to await the studies of Miller and Mitchell, with some minor technical contributions from myself. The physiology of the germinal centre also has a long history, and recently much interest has surrounded the relative roles of germinal centres as sites of somatic immunoglobulin V gene hypermutation and selection of high affinity B cells, versus the roles of extra-follicular proliferative foci as sources of the primary antibody response. Tolerance and self-antigens within the primary B lymphocyte repertoire is secured by clonal deletion within the bone marrow of those cells which are operationally the most threatening to the body, and a second level of functional impairment of B cell activity and life-span, capable of being induced centrally or peripherally, termed clonal anergy. As these concepts became progressively more refined through transgenic models of immunological tolerance, we turned our attention more towards tolerance within the secondary B lymphocyte repertoire. This is generated primarily inside germinal centres, where it appears that two quite separate mechanisms act as a bulwark against the possible creation of hypermutated anti-self cells. The first is that germinal centre activity and memory cell generation are dependent on antigen-specific, germinal centre-seeking CD4+ T cells, and if a putative anti-self mutant B cell gets no help because of T cell tolerance, it will not expand further. The second is a very specific mechanism confined to the germinal centre whereby antigen-specific B cells are especially sensitive to antigen-induced apoptosis if soluble, deaggregated antigen is presented to them before they reach the 'rescue' signal of follicular dendritic-cell-bound antigen. While some of the death in germinal centres is clearly apoptotic in nature, a further phenomenon observed electron microscopically relates to the formation of type B dark cells. It is not yet clear whether the DNA in this type of dying cell is cleaved. Transgenic expression of bcl-2 in germinal centre B lymphocytes confers incomplete protection from apoptosis caused by soluble antigen. The suggestion that at least some of this apoptosis is mediated via Fas-Fas ligand interactions is prompted by the observation that the apoptotic phenomenon is markedly reduced in lpr mice.