Antigen-driven selection of virgin and memory B cells.

This review has summarized the evidence indicating that far more B cells are produced in adult bone marrow than are required to maintain B cell numbers in the periphery. It is shown that most if not all these newly-formed B cells have the potential to become mature peripheral B cells. However, to do this they need to receive an appropriate signal in secondary lymphoid organs. Cells failing to receive such a signal die after a brief period. Two separate situations have been identified which result in recruitment of newly-formed virgin B cells into the peripheral B-cell pool: Following activation by antigen. When the peripheral B-cell pool has been depleted. It is proposed that the first of these signals requires T help and is initiated by antigen presented on interdigitating cells in extrafollicular areas of secondary lymphoid organs. This process seems to be confined to periods immediately following administration of antigen and does not continue in established immune responses to thymus-dependent antigens. It seems probable that continued B cell activation, occurring during long term antibody responses, takes place in the follicles of secondary lymphoid organs and is driven by antigen presented on follicular dendritic cells. Indirect evidence is cited which suggests that somatic mutation in rearranged immunoglobulin V-region genes occurs mainly following B-cell activation in follicles and not during primary B lymphopoiesis. It is suggested that this may involve a hypermutation process which is switched on in activated B cells in germinal centers. Evidence is presented suggesting that plasma cells generated from B cells activated early in immune responses have an average life-span of less than 3 d. However, plasma cells generated in established responses appear to have an average life-span in excess of 20 d. Later sections in the review consider how B-cell recruitment in thymus-independent antibody responses differs markedly from recruitment during thymus-dependent responses. The possible role of splenic marginal zone B cells in some thymus-independent antibody responses is discussed and the evidence indicating that SIgM + ve, IgD-ve marginal zone B cells develop as a distinct population from recirculating SIgM + ve, IgD + ve B cells is summarized.