The delineation of distinct subsets committed to the production of antibodies with different antigen-binding activities supports the view of a compartmentalization and specialization of function in the B cell repertoire and is consistent with the hypothesis of a developmentally layered immune system; as originally proposed by Herzenberg and Herzenberg. On the basis of the data by Solvason and Kearney in the human fetus and our data in the adult, and in agreement with the findings of Herzenberg et al. and Hardy et al. in the mouse, we propose that the human B cell repertoire includes at least three distinct B cell subsets: B-1a cells, which develop from progenitors in the fetal splanchnic district, namely the omentum, and are maintained in adult life by virtue of their self-replenishing nature; B-1b cells, progenitors of which can be found in the splanchnic district and, perhaps, adult bone marrow; and, finally, B-2 cells, which arise in the fetal liver and are continuously replenished in adult life by progenitors in the bone marrow (Figure 5). The different B cells types are distinguished by their differential expression of surface CD5 and, perhaps, CD11b and CD14, their differential expression of CD5 mRNA, and the different classes and specificities of the Ig they produce (Figure 5). B-1 lymphocytes play a major role in autoimmunity and constitute the physiological equivalent of the neoplastic forms in various lymphoproliferative disorders, such as CLL and SLL, which are often associated with the production of monoclonal antibodies to self antigens. Human B-1a (CD5+ B) and B-1b (CD5- CD45RAlo B) cells are responsible for the production of natural (polyreactive and monoreactive) antibodies in the fetus, neonate, and adult, and can give rise to the autoantibody-producing cells characteristic of several autoimmune disease states. Our recent findings suggest that while in healthy subjects the majority of natural polyreactive antibodies is encoded in V genes in germline configuration, some polyreactive antibodies are encoded in somatically mutated V genes, in a fashion consistent with an antigen-driven process of selection of such mutations. The nature of the antigen(s) involved in these selection processes remains to be determined. Under possibly different circumstances, the application of an antigen-driven process of clonal selection to B-1a and/or B-1b cells, previously committed to natural antibody production, can result in the generation of monoreactive high affinity and possibly pathogenic autoantibodies (Figures 5A and 5B).(ABSTRACT TRUNCATED AT 400 WORDS)