A technique for growing human bone marrow cell colonies in agar-gel medium is described. "Feeder layers" containing 1 X 106 normal human peripheral white blood cells are used as the stimulus for colony growth. Human bone marrow aspirates are collected in heparinized syringes and plated as 2 x 1 0 cells on "feeder layers." Normal human bone marrow yields 32-102 colonies per 2 X 105 cells plated. Colonies are almost exclusively granulocytic. Growth rate of colonies is slower than with mouse bone marrow but colonies reach a comparable size (500-1500 cells) at days 12-16.

The purpose of these experiments was to determine the degree of reduction in the number of antigenbinding B lymphocytes in the spleens of mice that had been rendered tolerant in the perinatal period. Newborn or pregnant mice were injected with fluorescein (Flu) coupled onto human gamma globulin, and the spleen cells of the neonatally injected mice, or of the offspring of the pregnant mice, were analyzed 1-6 weeks later. Tolerogen doses were chosen so as to achieve either a two-thirds reduction (low dose) in the number of anti-Flu B cells capable of yielding anti-hapten plaque-forming cell clones after in vitro stimu ation, or as representing a supraoptimal tolerogenic stimulus (high dose). Antigen-binding B cells were studied by a two-cycle procedure, namely an initial cycle of binding to Flu-gelatin (about 30% of the Flu-gelatin-binding number) was then subtracted from the Flu-gelatin-binding number of the identical group to arrive at a net figure for Flu-binding cells.

Spleen cells from normal unimmunized adult mice were subjected to a 2-stage fractionation procedure designed to prepare a subpopulation highly enriched for the capacity to form antibody to the hapten fluorescein (FLU).First, the spleen cell suspension was rocked in a petri dish coated with a thin layer of FLU-gelatin (1 5 min at 4 "C), the nonadherent cells washed away and the adherent cells recovered by melting the gelatin and treated with colleenase. The cells were relabeled with a fluorescent protein, usually FLU-gelatin at 100 pg/ml, (25 "C, 15 min). After washing, the cells were sorted into cohorts of relatively homogeneous fluorescence intensity in the fluorescence-activated cell sorter (FACS). Finally, the sorted cells were again collagenase-treated and stimulated with FLU-coupled polymerized flagellin in a limit dilution microculture system capable of generating single clones of anti-hapten plaque-forming cells (PFC).It was found that the cohort of cells representing the 10 % most intensely fluorescent cells, yielded PFC clones with a frequency of 1 in 8, about 5 times higher than the hapten-gelatinenriched cells. Less fluorescent cohorts gave progressively lower frequencies of clonable PFC precursors. Some evidence was obtained, suggesting that the most highly fluorescent cells produced antibody of higher median avidity than the FLU-gelatin-fractionated cells, which in turn showed higher avidity than unfractionated spleen cells, findings consistent with the clonal selection hypothesis.Somewhat surprisingly, cells with high fluorescence intensity prepared by FACS sorting of FLU-gelatin-labeled unfractionated spleen cells, yielded much lower PFC precursor frequencies than cells of equivalent fluorescence intensity derived through the 2-stage fractionation procedure.Spleen cells from newborn mice were subjected"t0 2-stage fractionation. The method proved equally applicable to them, and preparations with a 1 in 11 PFC precursor frequency could be readily generated. However, the neonatal cells differed from the adult cells in responding significantly better to the polyclonal B cell activator, lipopolysaccharide, than to antigen.The theoretical implications of these findings and the potential uses of pure populations of hapten-binding cells are briefly discussed.

The clonal anergy theory of induction of immunological tolerance states that differentiating B lymphocytes that encounter multivalent antigen at the pre-B to B cell transition stage can receive and store a negative signal, which renders them anergic to later triggering stimuli. The theory was tested by using an anti-IL chain monoclonal antibody, E4, as a model tolerogen.The fluorescence-activated cell sorter was used to select B cellfree cell populations from adult murine bone marrow or newborn spleen, and later, to analyze B cell neogenesis in vitro. The presence of E4 at 21 jg/ml was required to impede the development of normal numbers of B cells with full receptor status. The subsequent capacity of these B cells to respond in vitro to mitogens was assessed in a filler-cell free microculture system that allows single B cells to proliferate and differentiate. Concentrations of E4 far below those required to affect B cell neogenesis had profound inhibitory effects on the subsequent functional capacity of the B cells. In fact, 10-3 jig/ml of E4 markedly impaired both proliferation and antibody formation, and 10' jig/ml, which had no effect on Ig receptor development, abrogated functional ca- It is now well established that immature cells of the B lymphocyte lineage can be rendered immunologically tolerant by certain multivalent antigens both in vitro and in vivo (1-8). Recently, we provided evidence that the stage at which the cell displayed its greatest sensitivity to negative signaling was that of first emergence of the surface membrane immunoglobulin (mIg) receptors, that is, during the pre-B to B cell transition (7). It is possible to introduce tolerogens into the tissues of the developing fetus by transplacental transfer (6) and thus ensure interaction between lymphoid cells and tolerogen at the first appearance of specific antigen binding receptors. When fluoresceinated human gamma globulin (FLU-HGG) was injected into mice at 14.5 days ofgestation, effective B cell tolerance was induced with doses far lower than those required to reduce the number of FLU-specific antigen-binding B cells in the offspring (8). In other words, B lymphocytes expressing a normal range of FLU-binding avidities appeared to emerge from the pre-B cell pool in normal numbers, but were incapable of giving rise to anti-FLU antibody-forming cell (AFC) clones when challenged with a B cell mitogen or a T lymphocyte-independent antigen. This suggested that interaction between the newly emerged anti-FLU receptors and antigen had neither impeded the subsequent development of a standard complement of mIg nor directly led to the death of the anti-FLU B cell. Rather, the cell, though still alive and capable ofbinding antigen, appeared to have received some signal rendering it incapable of responding to appropriate triggering stimuli. We termed this phenomenon clonal anergy.This conclusion was dependent on a technology that enumerated and characterized FLU-binding B cells, by counting cells adhering specifically to thin layers of hapten...