Of many dinitrophenylated (DNP) protein conjugates tested, only DNP conjugated to polymerized flagellin (DNP-POL) (or the structurally related bacterial flagella) elicited a primary anti-DNP response in vitro. Other DNP proteins, such as DNP-monomeric flagellin (DNP-MON), were capable of inducing secondary responses in vitro. The capacity of DNP-POL to immunize spleen cell suspensions devoid of thymus-derived cells was the reason for the greater immunogenicity of DNP-POL, since even large numbers of flagellin-reactive activated thymus cells did not increase the anti-DNP response of normal spleen cells immunized with DNP-POL, whereas the thymus-dependent response to DNP-MON was markedly increased.
The capacity of various batches of DNP-POL to immunize normal spleen cells in vitro varied markedly, depending on the number of DNP groups conjugated. DNP-POL with few DNP groups conjugated was immunogenic, but even at very high concentrations did not induce tolerance. In contrast, highly conjugated DNP-POL did not immunize, but readily induced tolerance. DNP-POL with intermediate degrees of conjugation were, like unconjugated polymeric flagellin, capable of inducing both immunity and tolerance. Since DNP-POL immunizes bone marrow-derived lymphocytes (B cells) directly the reduced response was not due to a masking of carrier determinants, necessary for cell collaboration. By using mixed DNP-5-(dimethylamino)-1-naphthalyl (dansyl)-POL conjugates, it was found that the inhibitory effect of a high degree of hapten conjugated was hapten specific. Depolymerization of DNP-POL to DNP-MON, which does not induce primary anti-DNP responses, was excluded by centrifugation analysis and electron microscopy.
The relationship of the degree of hapten conjugation on DNP-POL to the capacity to induce tolerance and immunity in B cells has clarified the mechanism of immunological triggering of these cells. A model of the mechanism of "signal" discrimination between immunity and tolerance in B cells, based on these findings, is proposed.