In (6)(7)(8). This is not due to biases introduced by transformation or hybridization, since overexpression of the VH7183 genes early in ontogeny was confirmed by "in situ" hybridization methods comparing VH gene family representation in B cells from fetal liver and adult spleen (9) and "natural" plasma cells of neonatal and adult mice (10). A similar preferential usage of the VH7183 gene family occurs in differentiating B cells of the adult bone marrow (11,12). These studies support the notion of developmental control in VH gene family repertoires determined by a nonrandom position-dependent rearrangement preference (13,14). These repertoires evolve thereafter in a strain-specific manner during B-cell maturation (9, 10, 15), but it remains to be established what mechanisms dictate the selection of the VH repertoires expressed in the peripheral B-cell pools of an adult mouse. It is likely that the antigenic composition of the environment may play an important role in the final modulation of the expressed repertoires (10).We have now compared the VH gene family usage in the bone marrow (emergent) and splenic (available and actual) B-cell repertoires of germ-free and conventional (SPF) BALB/c adult mice. We found that VH gene utilization differs in these mice, suggesting that peripheral selection of VH gene repertoires is influenced by environmental antigens.Exogenous antigens could modify the expressed repertoires either by direct stimulation of antigen-specific clones or indirectly by idiotype interactions mediated by the immunoglobulins produced in those responses (16). If this was the case, administration of serum immunoglobulins could lead to changes in the expressed B-cell repertoires. Indeed, we found that the injection of purified isologous serum immunoglobulins into germ-free mice modifies VH gene family utilization. These results suggest that B lymphocytes and their products contribute to the selection of peripheral repertoires.MATERIALS AND METHODS Animals. Inbred BALB/c mice raised in germ-free conditions for 40 generations were from CIBA-Geigy. Control SPF BALB/c mice were from the same stock and breeding facilities or from the Pasteur Institute. As no differences were observed between both stocks of control SPF BALB/c mice, they were used indifferently thereafter. For each experiment, mice were matched for age (8-10 weeks) and sex.Immunoglobulin Treatment. Blood samples from 8-to 9-month-old female BALB/c mice were clotted in the cold and the serum was collected. Protein concentration was adjusted to 10-12 mg/ml with phosphate-buffered saline (PBS), and immunoglobulin was precipitated by addition of saturated ammonium sulfate. The precipitate was sedimented for 60 min at 4000 x g, dissolved in 20 mM Tris HCI/25 mM NaCl, pH 7.9, and dialyzed trice against the same buffer. A DE-52 ion-exchange column was prepared according to Whatman protocol (1 ml of DE-52 per 10 mg of protein). The column was equilibrated with the same buffer, the precipitated dialyzed protein was passed through the column, ...