The Ontogeny of Antigen-Binding Cells

In order to analyze the development of antigen-specific cells, the binding of a variety of antigens by cells in the fetal, neonatal, and adult mouse was compared. The fiber-binding assay was used in many of these experiments, because it provides a simple and uniform method for studying the specific interactions of cells with any of a wide variety of antigens. To demonstrate the specificity of the assay, cells from the spleens of immune and nonimmune adult mice were isolated and characterized. Specifically, after removal from the fibers, these cells were assayed for their viability, their ability to rebind to fibers of the same specificity, and their in vivo response to antigen after transfer to irradiated syngeneic recipients. These experiments indicated that the fiber method yields highly enriched populations of specific antigen-binding cells that are viable and include antigen-sensitive bone marrow-derived cells capable of undergoing differentiation into antibody secreting cells. This assay was then used to characterize cells specific for each of eleven different hapten and protein antigens. In all cases, specific antigen-binding cells were first detected in the liver, on the 15th day of the 19-day gestation period. These cells disappeared from the liver within a day of birth, but continued to increase in number in the spleen until adulthood. The proportions of antigen-binding cells of different specificities were similar in fetal, neonatal, and adult tissues. The antigen-binding cell populations from fetal livers and spleens were similar to each other and to adult spleen cell populations in the distributions of their relative avidities for several antigens. These results indicate that antigen-binding cells of various specificities arise relatively rapidly and in parallel during development, and therefore that strong positive antigenic selection is not likely to operate during . ontogeny. This has several implications for theories on the origin of antibody diversity, and in particular suggests that positive selection may not be required for somatic diversification to occur. These results also suggest that the sharply restricted ability of the neonatal animal to respond to antigenic stimulation is not due to the lack of antigen specific cells, but rather to the absence of mature cells capable of the interactions necessary for a full immune response. While measurements of the numbers of antigen-binding cells in the spleens of individual outbred fetal mice failed to detect subpopulations of individuals differing systematically from the fetal population as a whole, significantly more variation among individuals was found than would be expected if the actual number of cells binding a specific antigen were constant, or nearly so, among fetuses. To determine the source of this variation more precisely, the numbers of cells specific for each of two antigens in the spleens of individual outbred (Swiss-L) and inbred (Balb/c and CBA/J) fetal mice were measured as a function of spleen size. For outbred Swiss-L fetuses, the ratio of antigen-binding cells to nucleated cells varied significantly more than could be accounted for by sampling fluctuation. For each inbred strain, however, the number of cells specific for a given antigen was a constant proportion of the number of nucleated cells, within sampling error. These proportions varied from antigen to antigen, and from strain to strain. The ratio of the proportions of cells specific for the two antigens, however, differed no more from CBA/J to Balb/c mice than would be expected in repeated samples of cells from the spleen of a single fetus. These results confirm at the level of the individual fetus the uniform pattern of development seen for populations of fetuses. They reveal a surprising precision in the proliferation of specific antigen-binding cell populations and suggest that the development of these cell populations may be subject to strong genetic controls