Molecular and kinetic analysis of an epitope-specific shift in the B cell memory response to a multideterminant antigen.

Our previous studies showed that the primary and memory B cell responses to the multideterminant antigen poly-(L-Tyr, L-Glu)-poly-D,L-Ala-poly-L-Lys ((T,G)-A-L), differ. The primary response is dominated by antibodies binding side-chain epitopes; there is little antibody response to epitopes on the poly-D,L-Ala-poly-L-Lys backbone of (T,G)-A-L. In contrast, B cells producing A-L+ antibodies constitute approximately a third of the memory response to (T,G)-A-L. To determine the basis of this epitope-specific repertoire shift, we have examined the kinetics of expression of A-L+ B cells and antibodies after in vivo antigen priming and identified VH and V kappa genes used by A-L+ hybridoma antibodies derived from primary vs memory B cells. Kinetic studies, using the splenic focus assay, showed that the clonal frequency of A-L+ B cells remains low (<3% of (T,G)-A-L-specific B cells) 1 wk after Ag priming, increases (9%) by 2 wk, but does not reach the memory frequency (30%) until at least 3 wk after immunization. Molecular analyses showed that both the primary and memory A-L+ antibody responses are heterogeneous, using different VH and V kappa gene families as well as different germ-line genes within a VH gene family. Both H and L chain gene sequences showed somatic mutations in primary as well as memory antibodies. Analysis of antibody binding patterns and somatic mutations in a set of clonally related B cells that use a new germ-line VH gene in the VGAM3.8 family (VGK7, described here), showed a direct correlation between somatic mutation and change in antibody binding specificity. Our results demonstrate how somatic mutation and Ag selection play a role in the development of the memory response to a multideterminant Ag. The data are discussed in the context of the single vs dual lineage models for memory B cell generation.