The ability of antigen-bearing syngeneic and allogeneic peptone-induced peritoneal exudate macrophages to support development of primary and secondary antibody responses by murine lymphoid or spleen cells in vitro has been investigated. The antigen used was the terpolymer of L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT). Syngeneic and allogeneic macrophages supported development of comparable primary antibody responses to GAT, indicating that genetic restrictions do not limit efficient macrophage-lymphocyte interactions in primary responses. By contrast, immunized spleen or lymphoid cells developed secondary antibody responses preferentially when stimulated in vitro with GAT on macrophages syngeneic to the macrophages used to present GAT during in vivo immunization. Thus, genetic restrictions regulate efficient macrophage-lymphocyte interactions in secondary antibody responses. These restrictions have been demonstrated from 2 to 8 wk after a single immunization with limiting quantities of GAT and are controlled by the H-2 gene complex. The implications that immune lymphocytes selectively recognize and respond to antigen presented in the context of the macrophage membrane-antigen complex which sensitized the lymphocytes initially are considered.

Concanavalin A, a nonspecific polyclonal activator of T lymphocytes, activates Lyl and Ly23 subclasses to the same degree. After activation, the Ly23 subclass, but not the Lyl subclass, has the following properties: (a) Suppression of the antibody response to sheep erythrocytes (SRBC) in vitro. (b) Production of a soluble factor that suppresses the anti-SRBC response in vitro. (c) Suppression of the generation of cell-mediated cytotoxicity to H-2 target cells in vitro. Con A-activated cells of the Lyl subclass, but not the Ly23 subclass, express helper function in the anti-SRBC response in vitro. Because the intact Con A-stimulated T-cell population contains both cell types, these cells do not exert detectable helper effects in an anti-SRBC system in vitro, because the helper effect of Lyl cells is masked by the suppressor effect of the Ly23 cells. Each function is revealed by eliminating one or the other population with the relevant Ly antiserum. The resting T-cell population, before activation by Con A, also contains already programmed Lyl and Ly23 cells with similar helper and suppressor potentials, respectively. This is revealed by experiments with Ly subclasses which have been separated from the resting T-cell population and then stimulated by Con A. Thus helper and suppressor functions, as expressed in these systems, are manifestations of separate T-cell-differentiative pathways and do not depend upon stimulation of the cells by antigen.

The I-J subregion of the mouse major histocompatibility complex has been reported to encode antigenic determinants expressed by suppressor T cells. Previously, cosmid clones were obtained from mouse sperm DNA that contain all of the sequences between the I-A and I-E subregions, where I-J has been mapped genetically. However, hybridization of these sequences to RNA prepared from several I-J-positive suppressor T-cell hybridomas did not reveal the presence of a transcript. In addition, no rearrangements in this DNA were detected in the suppressor T cells that we have analyzed. Our results indicate that the I-J polypeptides are not encoded between the I-A and I-E subregions of the major histocompatibility complex. We discuss several hypotheses concerning the possible location and expression of I-J genes.

In vivo, after antigenic stimulation, mice develop antibody responses representing all five major immunoglobulin classes (1-5). Evaluation of serum antibody responses to heterologous erythrocytes at the cellular level is possible with the hemolytic plaque technique which facilitates the screening of large numbers of lymphoid cells for those cells releasing antibodies against erythrocyte antigens. The technique as described initially by Jerne et al. (6, 7) enumerated only cells releasing high hemolytic efficiency, %,M antibodies (6-9). Modifications have permitted enumeration of those cells releasing antibodies of other immunoglobulin classes (10-12). By using heterologous anti-mouse globulins monospecific for each major immunoglobulin class, antibody-releasing cells from each class have been detected and the kinetics of their appearance after immunization have been described (1, 2, 5). 7G I plaque-forming cells (PFC) usually are detected later and reach maximum numbers later than do 7M PFC (1, 2, 5). "),M PFC are thought to develop by rapid division of precursor cells restricted to production of antibody of the 7M class (6,(13)(14)(15)(16)(17).

In vivo, the antibody response in mice to the random terpolymer L-glutamic acid50-L-alanine30-L-tyrosine10 (GAT) is controlled by a histocompatibility-linked immune response gene(s). We have studied antibody responses by spleen cells from responder and nonresponder mice to GAT and GAT complexed to methylated bovine serum albumin (GAT-MBSA) in vitro. Cells producing antibodies specific for GAT were enumerated in a modified Jerne plaque assay using GAT coupled to sheep erythrocytes as indicator cells. Soluble GAT stimulated development of IgG GAT-specific plaque-forming cell (PFC) responses in cultures of spleen cells from responder mice, C57Bl/6 (H-2b), F1 (C57 x SJL) (H-2b/s), and A/J (H-2a). Soluble GAT did not stimulate development of GAT-specific PFC responses in cultures of spleen cells from nonresponder mice, SJL (H-2s), B10.S (H-2s), and A.SW (H-2s). GAT-MBSA stimulated development of IgG GAT-specific PFC responses in cultures of spleen cells from both responder and nonresponder strains of mice. These data correlate precisely with data obtained by measuring the in vivo responses of responder and nonresponder strains of mice to GAT and GAT-MBSA by serological techniques. Therefore, this in vitro system can effectively be used as a model to study the cellular events regulated by histocompatibility-linked immune response genes.

A population of thymus-derived lymphocytes has been identified that, upon activation by the nonspecific plant mitogen concanavalin A, suppresses the development of plaque-forming cell responses in fresh or 48-h antigen-stimulated cultures of mouse spleen cells. Suppressor cells can inhibit both primary and secondary IgM and IgG responses in vitro. X-irradiation before activation of peripheral thymus-derived cells by concanavalin A abrogates generation of suppressor cells. After a 48 h activation period, however, the function of concanavalin A-activated suppressor cells is radioresistant. As yet uncertain is whether these suppressor cells are a population of cells distinct from thymus-derived "helper" cells. In certain important regards, the cells mediating these two opposing functions share similar characteristics; the effect observed may be determined by the circumstances of activation or the numbers of activated cells, and may consequently represent different functions of a single thymus-derived regulator cell population.

The synthetic terpolymer of L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT) fails to stimulate development of GAT-specific antibody responses in nonresponder strains of mice, but does stimulate the development of GAT-specific suppressor T cells that inhibit the development of normal anti-GAT antibody responses to GAT complexed to methylated bovine serum albumin (GAT-MBSA). Furthermore, extracts prepared from lymphoid cells of GAT-primed, but not control, nonresponder mice inhibit the development of antibody responses to GAT-MBSA by normal nonresponder mice. This suppression is specific, dose-dependent, and can be readily analyzed in vitro. The suppressive factor is a T-cell product. An extract from GAT-primed DBA/1 mice inhibits the response to GAT-MBSA by spleen cells from histoincompatible strains of mice that are nonresponders to GAT, but not strains that are responders to GAT.

In recent studies we have found that GAT not only fails to elicit a GAT-specific response in nonresponder mice but also specifically decreases the ability of nonresponder mice to develop a GAT-specific PFC response to a subsequent challenge with GAT bound to the immunogenic carrier, MBSA. Studies presented in this paper demonstrate that B cells from nonresponder, DBA/1 mice rendered unresponsive by GAT in vivo can respond in vitro to GAT-MBSA if exogenous, carrier-primed T cells are added to the cultures. The unresponsiveness was shown to be the result of impaired carrier-specific helper T-cell function in the spleen cells of GAT-primed mice. Spleen cells from GAT-primed mice specifically suppressed the GAT-specific PFC response of spleen cells from normal DBA/1 mice incubated with GAT-MBSA. This suppression was prevented by pretreatment of GAT-primed spleen cells with anti-θ serum plus C or X irradiation. Identification of the suppressor cells as T cells was confirmed by the demonstration that suppressor cells were confined to the fraction of the column-purified lymphocytes which contained θ-positive cells and a few non-Ig-bearing cells. The significance of these data to our understanding of Ir-gene regulation of the immune response is discussed.