Search

To examine the influence of Ag presentation by B cells on immune responses, we have used mice transgenic for an Ig heavy chain from a monoclonal anti-azobenzenearsonate (Ars) Ab to deliver Ag to B cells during immunization. A large proportion of transgene-expressing B cells in these mice binds Ars, while transgenic serum Ig shows poor Ars binding. Transgenic B cells present Ars proteins better than their nonhaptenated counterparts. This is associated with an increase in the proliferative responses of transgenic T cells to Ars protein immunization. Although B cell numbers in the transgenic mice are lower, many B cells in them show an activated phenotype, as identified by altered surface levels of peanut agglutinin reactivity, CD23, CD24, CD44, CD62L, and CD86. Even against nonhaptenated immunogens, transgenic responses show significant enhancement in the relative proportions of the Th1 cytokine IFN-gamma over the Th2 cytokines IL-4 and IL-10. Haptenated immunogens further enhance the predilection of transgenic mice to produce relatively more IFN-gamma. Consistent with this, there is an increase in IgG2a/IgG1 ratios in serum Abs in response to haptenated immunogens in transgenic mice. Adoptive transfer of primed hapten-specific secondary B cells into nontransgenic mice also induces an increase in relative levels of IFN-gamma in response to haptenated immunogens. Thus, presentation of immunogen in vivo by activated Ag-binding B cells contributes to enhanced immunogenicity and a Th1 cytokine bias.

Antigen-specific molecules secreted by murine T cell hybrids (TABM) specific for 4-hydroxy 3-nitrophenyl (NP) or azobenzenearsonate (ABA) were purified from ascitic fluid by ion exchange chromatography and/or affinity for antigen. Partial amino acid sequence of reduced Mr 72,000 NP-specific polypeptides and Mr 20,000 peptides prepared by treatment of the ABA-specific immunoprotein with cyanogen bromide was obtained and a septapeptide of the NP-specific TABM shared 3/7 residues with the ABA-specific TABM. Both TABM shared residues present in 95% T cell receptor for antigen (TCR) V alpha subgroup I and 83-96% murine immunoglobulin V kappa Fr3. These results provide evidence that extracellular antigen-specific T cell proteins are soluble analogues of TCR alpha chains and belong to the immunoglobulin supergene family.

Antigen-specific molecules secreted by a murine T cell hybrid specific for azobenzene arsonate (ABA) were purified from ascites fluid by ion exchange chromatography and affinity for antigen. The antigen-specific proteins were purified 250 fold and were resolved predominantly as Mr 110,000 polypeptides by reduction and SDS-polyacrylamide gel electrophoresis. The ability of these molecules to bind antigen was analyzed by an ELISA using antigen-coated microtiter trays. Binding of the T cell proteins to antigen was detected with antisera specific for the proteins. Antigen binding to ABA-ovalbumin but not ovalbumin was optimal at 37 degrees C and protein derived from another T cell hybrid did not bind ABA-ovalbumin. Solid phase antigen binding was inhibited specifically by soluble ABA-ovalbumin, indicating that these T cell-derived proteins bind nominal antigen in the solid or liquid phase. It is suggested that these proteins represent a soluble, antigen specific manifestation of some T cell function.

Some T cells produce membrane-associated or soluble molecules which bind nominal antigen specifically (TABM) and effect immunoregulation or events similar to cell-mediated hypersensitivity. We have used polyclonal antisera raised against an azobenzene arsonate (ABA)-specific TABM secreted by an ABA-specific T cell hybrid or against TNP-specific polypeptides produced by immunoregulatory T cells to identify the expression of soluble (secreted) or membrane-associated TABM. Ascites fluid or culture medium containing a T cell hybrid or T cell lines, respectively, contain TABM recognized only by an antiserum specific for the secreted T cell hybrid (ABA-specific) derived TABM. Conversely, an antiserum that recognized the TNP-specific polypeptides detected cell-membrane associated TABM but did not bind TABM secreted by the T cell hybrid or cell lines.

We have compared by functional assays the relative preference among Ia molecules for their ability to present tyrosine-azobenzenearsonate (ABA-Tyr) to T cells. Immunization of B10.BR mice (IAk, IEk) with ABA-Tyr resulted in the induction of IAk-restricted T cells only. Immunization of B10.A(5R) mice (IAb, IEb/k) gave only IEb/k-restricted T cell clones even though IAb-restricted responses could be induced in C57BL/6 mice (IAb). These results indicated that IAk was preferred over IEk and IEb/k was preferred over IAb for presentation of ABA-Tyr. A comparison between IAk and IEb/k made by immunizing [B10.BR x B10.A(5R)]F1 mice (IAk, IEk, IAb, IEb/k), showed that IEb/k was favored over IAk. No IAb- or IEk-restricted response was seen. Further attempts were made to compare Ia preference for ABA-Tyr presentation by competitive inhibition assays. It could be shown that the presence of IEb/k molecules on an accessory cell interfered with the ability of IAb molecules on the same cell to present ABA-Tyr to an IAb-restricted T cell clone by direct competition. Such a competition was not observed between IAk and IEk. Finally, it could be shown that addition of ABA-Tyr inhibited the presentation of moth cytochrome-c peptide (81-103) by IEb/k but did not influence its presentation by IEk. From these functional studies we suggest that the binding affinity of ABA-Tyr with the Ia molecules will fall in the order: IEb/k greater than IAk greater than IAb greater than IEk.

The T cell response to L-tyrosine-azobenzenearsonate (ABA-tyr) has been studied using T cell lines and clones derived from three different mouse strains, B10.BR, B10.A (5R) and C57B1/6. In all cases, the arsonate group in conjunction with the amino group of tyrosine formed the functional T cell epitope. Molecules without any one or both of these groups are non-stimulatory. The hydrophobic moiety consisting of the azo-linked benzene rings forms the agretope of the molecule, as is evident from competitive inhibition of T cell stimulation by non-stimulatory analogues lacking the epitope. Substitutions on the benzene ring at ortho or meta positions resulted in decreases in ability to compete, indicating the likelihood of steric inhibition of binding of the agretope with the Ia molecule. This pattern was observed for clones and lines restricted by IAk, IAb and IEb/k MHC class II molecules. Peptides from lambda repressor protein, P84-98 and P73-88, showed haplotype specificity in their ability to inhibit ABA-tyr-induced proliferation of T cell clones, BRTC-4 and B6TC, respectively. The binding constants of ABA-tyr analogues were considered to be comparable to those of lambda repressor peptides because equimolar concentrations resulted in similar levels of competition. A cluster of aromatic amino acids on the floor of most MHC class II molecule binding sites might provide strong hydrophobic interaction with azo-linked benzene rings of ABA-tyr, thus accounting for its immunogenicity in all strains of mice studied.

The suppressor pathway that regulates the T cell response to p-azobenzenearsonate (ABA)-coupled cells has been studied. It has been found that the ability of anti-idiotypic second-order T suppressor cells (Ts2) to inhibit T cell-dependent delayed-type hypersensitivity (DTH) responses depended upon the presence of cross-reactive-idiotype (CRI)-bearing T cells present in ABA-primed mice. This suppressor T cell subset, termed Ts2, so exists with CRI-negative T cells that mediate DTH in vivo. It appears that antigen-activated CRI+ Ts3 require signals from the anti-CRI Ts2 subset to suppress DTH reactions in an idiotype-nonspecific manner. The relevance of these observations to a comprehensive scheme of T and B cell regulation is discussed.

Biologic activities and immunochemical characteristics of an azobenzenearsonate- (ABA)specific suppressor T cell factor produced by a longterm T cell hybridoma, F12, were studied. In vivo administration of F12 culture supernatant resulted in the suppression of ABA-specific delayed-type hypersensitivity (DTH) responses and the inhibition of priming for ABA-specific cytotoxic T lymphocyte responses. Moreover, F12 induced a second set of suppressor cells that act in the efferent phase of DTH. The active material in the F12 culture supernatant expressed major cross-reactive idiotypic (CRI) determinants of anti-ABA antibodies of A/J mice and I-J subregion-coded specificities but not express determinants of immunoglobulin constant regions. These results demonstrated that F12 is a functioning hybrid cell line of the first-order suppressor T cell subset.

We have used a fractionation procedure involving bovine serum albumin gradient floatation, adherence to glass, and rosetting with antibody-coated sheep erythrocytes to purify an accessory cell fraction from Lewis rat spleens. The fraction which is of light buoyant density, nonadherent, and FcR- is markedly Ia+, lacks phagocytic ability, is nonspecific esterase negative and under scanning electron microscopy has a heterogeneous morphology with a variety of protuberences described for rat dendritic cells. This putative dendritic cell preparation is very effective at stimulating proliferative responses with concanavalin A and allogeneic cells. When used to reconstitute the reactivity of peritoneal exudate cells of rats immunized with azobenzenearsonate-N-acetyl-L-tyrosine (ABA-Tyr) and subsequently depleted of Ia+ cells, it was shown to be highly effective with ABA conjugates of tyrosine, Ficoll, and polystyrene beads. Thus, despite the apparent absence of phagocytic or degradative abilities, this cell was very efficient at presenting large soluble and insoluble antigens. It is suggested that processing may occur at the cell surface without requiring internalization.

The antigen-induced proliferative response of lymph node cells (LNC) from mice sensitized to the monofunctional antigen L-tyrosine-p-azobenzenearsonate (ABA-Tyr) was used to monitor genetic control. All strains tested mounted significant responses, but those that were H-2(b) at both the I-A and I-E loci [B10., B6., B10.A(18R), A.BY, and C3H.SW] gave consistently weaker responses than other haplotypes. The F(1) progeny of matings between high and low responder phenotype parents (DBA/2 and B6, respectively) were high responders, establishing the dominance of the responder trait. Proliferative responses of LNC to ABA-Tyr were blocked by the appropriate anti-Ia monoclonal reagents. For example, B10.A(4R) LNCI (I-A(k), I-E(b)) were blocked by anti-I-A(k), whereas B10.A(3R) LNC (I-A(b), I-E(k)) were blocked by anti-I-E(k). Long-term cultures of T cell lines specifically reactive to ABA-Tyr were established from LNC of A/J mice immunized with ABA-Tyr and were cloned by limiting dilution. The proliferative responses to ABA-Tyr of 14 out of 15 clones tested were I-A restricted on the basis of activation by antigen-presenting cells from appropriate recombinant strains and the blocking activity of the monoclonal anti-Ia antibodies. The response of the other clone was I-E restricted. The fine antigen specificity of the clones was studied using structural analogs of the homologous antigen to induce proliferation. The clones could be divided into three types with respect to responsiveness to ABA-histidine (ABA-His). One group responded about equally well to ABA-His and ABA-Tyr. A second set responded less strongly to ABA-His than to ABA-Tyr, while the third showed no response above background to ABA- His. In all instances, the ABA-His-responding clones discriminated exquisitely between the 2-azo and 4-azo histidine isomers, responding only to the 4-azo compound. These T cell clones provide extremely useful tools for studies of T cell specificity, antigen recognition and lymphoid cell interaction systems.

Azobenzenearsonate (ABA)-specific T cell-mediated suppression has been studied in aging mice. ABA-specific suppressor T cells were induced in young and old mice by injection of ABA conjugated to syngeneic spleen cells (ABA-SC). These suppressor cells were tested for their ability to suppress the in vitro anti-trinitrophenyl (TNP) antibody response of lymph node cells obtained from ABA-keyhole limpet hemocyanin (KLH)-primed young or old mice and cultured with TNP-ABA-KLH. Suppressor T cells were found to be more easily induced in old than in young mice but to suppress less efficiently the antibody response of cells from old than from young mice. The increased inducibility of antigen-specific suppressor T cells in old mice is compatible with the age-dependent decline of immune responsiveness to exogenous antigens. The loss of cell sensitivity to antigen-specific immunosuppression as well as the lack of evidence for increased nonspecific suppression in old mice is consistent with the age-related increase in autoimmune disorders. These findings provide a unifying explanation for the most relevant immunological phenomena of senescence.

Anti-p-azobenzenearsonate (ABA) antibodies, coupled covalently to normal syngeneic spleen cells and then given intravenously to normal animals, were found to be potent tolerogens for delayed-type hypersensitivity (DTH) to ABA. The ability of the antibody-coupled cells to induce tolerance was determined to be a result of the cross-reactive idiotype (CRI+) fraction of the antibodies, because anti-ABA antibodies lacking the CRI+ components when coupled to spleen cells were unable to cause any significant inhibition. Furthermore, genetic analysis revealed that the ability of CRI-coupled cells to inhibit ABA-specific DTH is linked to Igh-1 heavy chain allotype, in as much animals which possess heavy chain allotypes similar to that of A/J were sensitive to this inhibition. Adoptive transfer experiments provided evidence that CRI-coupled cells induce suppressor cells, and spleen cells or thymocytes from animals received CRI-coupled cells were able to transfer suppression to naive recipients. In addition, treatment with anti-Thy1.2 serum plus complement completely abrogated their ability to transfer suppression. Thus, this active suppression is a T-cell-dependent phenomenon. In investigating the specificity of these suppressor T cells, it was found that they functioned in an antigen-specific manner and were unable to suppress the development of DTH to an unrelated hapten 2,4-dinitro-1-fluorobenzene.

The homologous, monoclonal antiidiotope, MB, induced idiotope suppression that was remarkably stable and could be transferred by B lymphocytes. Marked depletion of T cell function, confirmed by limiting diluting analysis, did not affect the ability of MB to suppress the corresponding idiotope. Suppression induced by MB appears to result from direct interaction with idiotope-positive B cells, without the intervention of idiotope-specific T suppressor cells.

Strain A/J mice immunized with azobenzenearsonate (ABA)-mouse IgG conjugates develop suppression for anti-trinitrophenyl(TNP) responses to doubly conjugated (ABA,TNP) proteins. This suppression is specific for the ABA epitope and is mediated by T cells in cell transfer experiments. ABA-binding T cells from suppressed animals were purified by a two-stage procedure in which B cells were removed from spleen cell populations by adherence to plastic surfaces coated with anti-mouse Ig antibody, followed by binding the nonadherent population (more 95 percent Thy-1-positive) to surfaces coated with ABA-protein conjugates. Approximately 90 percent of the cells recovered by temperature-dependent elution from the ABA plates (similar to 2 percent of the spleen cells) bound antigen immediately afterward, and up to 50 percent of the cells bound anti-cross-reactive idiotype antibody. On the other hand, the nonadherent T-cell population was completely negative in the antigen- binding and idiotype assays. Another distinguishing feature of the two T-cell populations was that 78 percent of the adherent cells, but only 2 percent of the nonadherent cells, were Ia positive, although the specific I-region marker(s) expressed on the cells was not identified. The biological function of the antigen-binding T cells was investigated using a standard cell transfer protocol. Suppressor cells were enriched in the adherent population by a factor of at least 25, establishing that functional, epitope-specific, idiotype-bearing T cells can be significantly purified by this procedure. Note Added in Proof. We have recently isolated two types of ABA-binding molecules biosynthetically labeled with (35)S-methionine from NP-40 lysates of purified antigen-specific T cells. The molecules were purified by adsorption onto an ABA-Sepharose immunoadsorbent followed by elution with 9 M urea. Autoradiograms of SDS-PAGE of the eluates revealed components with tool wt of approximately 60,000 and 33,000 dahons. These molecules were not present in eluates from a bovine IgG-Sepharose control immunoadsorbent and thus represent specific ABA-binding products synthesized by T cells.

It has been firmly established that helper T lymphocytes respond to antigen only when it is presented in the context of class II major histocompatibility complex molecules. This has fostered the concept that the nominal antigen becomes physically associated with class II molecules on antigen-presenting cells (APCs). However, the structural components of antigen molecules that actually interact with class II molecules on APCs (agretopes) and with antigen receptors of T lymphocytes (epitopes) have not been precisely defined. This question was addressed using murine T cell clones specific for the model synthetic antigen L-tyrosine-p-azobenzenearsonate (ABA-tyr), which induces T cell responses despite its small size and simple structure, and a series of analogues of the homologous immunogen. Two experimental approaches were taken. First, APCs were pulsed with analogues and used to stimulate T cell proliferation. The patterns of stimulation segregated the clones into two specificity groups and indicated that the epitope recognized by the T cell included the arsonate group and elements of the side chain of tyrosine. Second, analogues that failed to stimulate were used to block the presentation of ABA-tyr in an effort to define the agretope. The blocking protocol involved serially pulsing APCs with analogue and with ABA-tyr prior to culturing with T cells. Compounds containing the core azo-linked ring structure blocked presentation of ABA-tyr in a dose-dependent fashion, whereas p-arsanilic acid and L-tyrosine were ineffective. The blocking was specific inasmuch as the compounds had no effect on the antigen-induced proliferative responses of unrelated T cell clones.(ABSTRACT TRUNCATED AT 250 WORDS)

Infection of mice with vesicular stomatitis virus (VSV) at the time of immunization failed to enhance T cell-mediated immune response to azobenzenearsonate-(ABA) conjugated spleen cells as measured by delayed-type hypersensitivity and by in vitro proliferation and in vitro generation of ABA-specific cytotoxic T cells. However, mice infected with VSV are incapable of responding to signals from suppressor T cells or their soluble factors. Further analysis revealed that VSV infection does not interfere with the induction of Ts-1 or Ts-2 cells. Because infection of Ts-1 or Ts-2 donors had no effect on the subsequent response seen in the recipients of antigen and suppressor T cells, the most likely candidate for the target of VSV infection is therefore the Ts-3 cell or another T cell interacting with Ts-3. This is supported by our observation that it is possible to bypass the VSV effect by providing the recipients of VSV with normal Lyt-2+-bearing T cells.

Strain A/J mice made secondary indirect plaque-forming cell (PFC) responses to azobenzenearsonate (ABA) conjugates of giant keyhole limpet hemocyanin (KLH), a thymic-dependent antigen, but not to conjugates of Ficoll, a T-independent antigen. ABA-Ficoll was also unable to elicit a response in animals primed with ABA-KLH, which have an expanded anti-ABA memory cell pool. On the other hand, ABA-Ficoll rendered mice unresponsive to ABA-KLH when administered before priming or boosting with the T-dependent immunogen. Hence, the T-independent antigen was able to tolerize but unable to trigger B-memory cells responsive to the T-dependent antigen. A/J mice immunized with dinitrophenyl conjugates of Ficoll or bovine IgG (BGG) made vigorous IgM and IgG PFC responses. PFC responses to ABA-KLH and 2,4-dinitrophenyl (DNP)-BGG were abrogated by depleting mice of C3 with cobra venom factor, whereas the IgM and IgG PFC responses to DNP-Ficoll were unaffected. B lymphocytes were fractionated on the basis of receptors for C3 and the subpopulations were assayed for in vitro PFC responses to DNP-Ficoll. Very little response was obtained from complement receptor lymphocyte [CRL(+)] B cells, whereas CRL(-) cells were more responsive than unfractionated B cells. Both populations responded to a polyclonal B-cell mitogen (lipopolysaccharide). On the other hand, the in vitro PFC response to a T-dependent antigen (sheep erythrocytes) correlated with the presence of CRL(+) B cells in the cultures. However, a minor component of this response, sensitive to anti-Thy-1 serum, was made by CRL(-) B cells, indicating the existence of subpopulations of T-dependent B cells with different signalling requirements. The results suggest that most B cells responsive to T-dependent antigens possess receptors for C3 and that C3 plays an obligatory role in the response of these cells. A distinct subpopulation of B cells which lack C3 receptors respond to T-independent antigens. The precursors of PFC for the ABA epitope reside largely or exclusively in the CRL(+) compartment in A/J mice, whereas precursors for the DNP determinant are found in both compartments.

An I-Ab-restricted, L3T4+ Ly2- T-cell clone, 5R-4F3, specific for ABAtyr was established in culture from a B10.A(5R) mouse. Since b haplotype mice respond weakly to ABAtyr compared to other haplotypes, this is a candidate clone of low responder phenotype. In support of this contention, 5R-4F3 grew very poorly under conditions that supported the vigorous growth of E beta bE alpha k-restricted T-cell clones from the same mouse. The I-A (low responder) and I-E (high responder) restricted T-cell clones also differed in their responses to apc pre-pulsed with antigen, compared to apc with antigen present continuously during culture. The low and high responder clones responded comparably to IL-2. Attempts to elevate the response of C57BL/6 mice to ABAtyr in vivo by injecting them with human recombinant IL-2 and antigen together were only partially successful: C57BL/6 mice treated in this way showed a 3-5-fold increase in their proliferative responses to ABAtyr, which was at best only one quarter of the level of response shown by high responder A/J mice to the same antigen dose.

B cell-deficient, rabbit anti-mouse IgM-treated mice were compared with normal or normal rabbit immunoglobulin-treated controls in their ability to develop proliferative T cell responses, delayed hypersensitivity, and primary or secondary cytotoxic T cell responses. Immunization with hapten-coupled autologous spleen cells resulted in anti-mu-treated mice generating only marginal T cell responses. This decreased responsiveness was shown to be attributable not to an intrinsic T cell defect or to changes in the ability of macrophages from anti-mu-treated mice to present soluble antigen, but rather to the greatly diminished capacity of B cell-deficient spleen cells to present antigen. The results support the concept that B cells play a significant role in antigen presentation required for T cell activation.

The possibility that the failure of anti-mouse immunoglobulin (Ig) antibody to induce antibody synthesis by B cells might be due to reversible receptor blockade was investigated. Murine spleen cells were cultured for 3 days in the presence of minute quantities of intact of (Fab') fragments of rabbit anti-mouse Ig antibody. Thereafter, the cells were washed and either trypsin treated or not before reculturing for 18 hr. Only cells that had been trypsinized after culturing with either intact or fragments of anti-Ig gave a vigorous polyclonal antibody response. This response was extremely T dependent, since T cells or culture supernatants from Con A-activated T cells were required for the B cell response. Moreover, anti-delta was much more effective than anti-mu in inducing antibody synthesis. Finally, the use of three different anti-idiotypic antisera rather than anti-Ig reagents selectively activated the specific idiotype in each instance. The findings demonstrate that anti-Ig reagents can potentiate the response of B cells to signals delivered by T cells.

In an effort to study T cell functions in Lewis rats immunized with ABA-N-acetyl-L-tyrosine (ABA-tyr), we developed an antigen that provides a sensitive assay of ABA-specific helper function that is read as an increase in TNP-specific plaque-forming cells (PFC). This antigen has ABA coupled to AECM-Ficoll by virtue of a tripeptide (tyr-ala-ala) spacer and TNP coupled to the AECM side chains. At subimmunogenic doses, this antigen induced 400 anti-TNP PFC/10(6) spleen cells in ABA-tyr-immunized rats. As many as 8000 PFC/10(6) spleen cells were induced with larger doses of antigen (200 micrograms). By contrast, only 490 PFC/10(6) spleen cells could be induced with 1 mg of the conventional doubly haptenated protein carriers such as ABA-BSA-TNP. Both direct and indirect PFC were induced by this antigen in primed rats. The use of this antigen and passive transfer techniques to study ABA-specific helper activity revealed some differences from ABA-specific delayed-type hypersensitivity (DTH) and in vitro proliferation, which were studied previously. Cells responsible for helper activity appeared sooner after immunization and were found most prominently in peritoneal exudates but also significantly in spleen where the cells responsible for DTH or in vitro proliferative responses were never found. By contrast, helper cells were not seen in lymph nodes, where some proliferative activity could be found. Of these three ABA-specific T cell functions, helper activity was least easily suppressed by the previously used regimens of ABA-tyr in incomplete freunds adjuvant (IFA). Moreover, helper activity appears after injection of ABA-tyr in IFA, a method that has never in our hands yielded detectable DTH or in vitro proliferative responses. Despite these differences, phenotyping with monoclonal antibodies indicated that cells responsible for helper and proliferative activities were both W3/25+ and OX8-.

All A/J mice immunized with a conjugate of p-azophenylarsonate groups to keyhole limpet hemocyanin produce antibodies against azophenylarsonate, some of which share a crossreactive idiotype. The appearance of the idiotype can be suppressed, without reducing the response against azophenylarsonate, by injecting rabbit anti-idiotypic antibodies prior to immunization. We have now observed that mice suppressed in this way, or by adoptive transfer of leukocytes from other suppressed mice, and then immunized with the hemocyanin-azophenylarsonate conjugate, possess high proportions (up to 14%) of lymphocytes that form rosettes with A/J erythrocytes coated with Fab fragments possessing the idiotype. Idiotypic specificity was demonstrated by various experiments. Most of all of the rosette-forming lymphocytes appear to be thymus-derived lymphocytes (T cells). Treatment of T cells with trypsin eliminated the capacity to form rosettes, which was restored on standing overnight in medium. Thus, the receptors are synthesized by the cells and are not passively adsorbed. Treatment of mice with anti-idiotypic antiserum without antigenic stimulation did not elicit substantial numbers of rosette-forming cells. The requirement for antigen suggests that antigen-idiotype complexes may be a stimulatory agent. A prolonged rest period after immunization of suppressed mice was required for the induction of high percentages of rosette-forming cells. Rosette formation provides a convenient method for studying factors that induce the formation of idiotype-specific T cells.

Delayed type hypersensitivity to the hapten azobenzenearsonate (ABA) can be induced and suppressed by the administration of hapten-coupled syngeneic spleen cells by the appropriate route. Suppressor T cells stimulated by the intravenous administration of ABA-coupled spleen cells have been shown to produce a discrete subcellular factor(s) which is capable of suppressing delayed type hypersensitivity to azobenzenearsonate in the mouse. Such suppressor factors may be produced by the mechanical disruption of suppressor cells or by placing such suppressor cells in culture for 24 h. The suppressor factor(s) (SF) derived from ABA-specific suppressor cells exhibit biological specificity for the suppression of ABA delayed type hypersensitivity (DTH), but not trinitro-phenyl DTH, as well as the capacity to bind to ABA immunoadsorbents. Passage of suppressor factor(s) over reverse immunoadsorbents utilizing a rabbit anti-mouse F(ab')2 antiserum demonstrated that the antigen-specific T-cell derived SF does not bear conventional immunoglobulin markers. The suppressor factor(s) are not immunoglobulin molecules was further demonstrated by the inability of anti-ABA antibodies to suppress ABA DTH. Gel filtration of ABA suppressor factor(s) showed that the majority of the suppressive activity was present in a fraction with molecular weight ranging between 6.8 x 10(4) and 3.3 x 10(4) daltons. We also analyzed for the presence of determinants encoded by the H-2 major histocompatibility complex (MHC) and found that immunoadsorbents prepared utilizing antisera capable of interacting with gene products of the whole or selected gene regions of H-2 MHC, i.e., B10.D2 anti-B10.A and B10 anti-B10.A immunoadsorbents, retained the suppressive activity of ABA-SF. Elution of such columns with glycine HCl buffers (pH 2.8) permitted recovery of specific suppressive activity. Taken collectively such data supports the notion that suppressor T-cell-derived ABA suppressor factors have antigen-binding specificity as well as determinants controlled by the K end of the H-2 MHC. The distribution of strains capable of making SF has also been analyzed. The relationship of the antigen-binding specificity to VH gene products is discussed in this and the companion paper.

The T cell antigen L-tyrosine-p-azobenzenearsonate is unique in being a simple determinant that can be presented in the context of both I-A and I-E. I-E-restricted T cell clones derived from B10.A(5R) mice were found to fall into three groups: Type I clones recognized antigen only in the context of syngeneic apcs, Type II clones recognized antigen with the same highly specific major histocompatibility complex restriction but in addition proliferated in response to allogeneic stimuli; Type III clones were "degenerate" in their major histocompatibility complex-restricted recognition of antigen and proliferated when antigen-presenting cells bearing Eb beta Ek alpha (syngeneic), Ek beta Ek alpha, or Ed beta Ed alpha were used. These observations allow some conclusions to be drawn about sites on the I-E molecule that may be functionally significant in the presentation of this antigen. By using the B cell hybridoma LK35.2 as target cells, some of these T cell clones act as cytotoxic cells in the Class II-restricted manner predicted from the results of proliferative assays. Class II-restricted cytotoxicity can therefore be controlled by both I-A and I-E mouse Ir gene loci.

The T cell repertoire of BALB/c mice contains clones capable of recognizing p-azobenzenearsonate (ABA)-tyrosine (Tyr) in association with both I-A and I-E-encoded class II molecules. Immunization of BALB/c animals with ABA-GAT (terpolymer of L-Glu60-L-Ala30-L-Tyr10) or ABA-GLT (terpolymer of L-Glu51-L-Lys34-L-Tyr15) instead of ABA-Tyr reduces the secondary proliferative response to ABA-Tyr in vitro. Limiting dilution experiments indicate that this situation corresponds to the recruitment of fewer ABA-specific T cells in vivo. The same experiments, performed in A.TH mice, which are nonresponder to both GAT and GLT, demonstrate that the number of ABA-specific T cells stimulated in vivo with ABA conjugates depends on the Ir gene-controlled immunogenicity of the carrier rather than on the ABA epitope density on the immunogen. Although GAT is preferentially recognized in association with A, and GLT with E, ABA-GAT and ABA-GLT stimulate both A and E-restricted ABA T cells in vivo and in vitro. The ABA-Tyr-specific T cell repertoire is not qualitatively affected by the carrier. This demonstrates that the inhibition of hapten-specific T cell expression upon immunization with ABA conjugates does not result from a competition between hapten and carrier-specific T cells for epitope recognition in association with the same Ia molecule on antigen-presenting cells.

Pretreatment of Lewis rats with a single i.p. injection of ABA-N-acetyl-tyrosine in incomplete Freund's adjuvant induced an unresponsiveness for delayed-type hypersensitivity to subsequent immunization with the same antigen in complete Freund's adjuvant. Complete suppression of in vitro antigen-induced proliferative responses required repeated pretreatment. Passive transfer of lymphoid cells from spleen and lymph nodes but not sera from suppressed rats induced unresponsiveness of hapten-specific T cell functions. Nylon wool-nonadherent cells and cells panned on F(ab')2 of rabbit anti-Lewis rat Ig plates suppressed the induction of DTH and in vitro antigen-stimulated proliferation. Adult thymectomy increased DTH and failed to abolish the induction of suppression.

The T lymphocyte repertoire consists of clones recognizing foreign antigens together with self histocompatibility molecules. Diversification of the receptor is believed to arise by somatic mechanisms during ontogeny. MHC gene products are essential for this process as well as for antigen recognition and expression of T cell functions. Yet, the antigen-specific T cell receptor is not encoded by MHC genes. Little is still known concerning the nature and the genetic origin of this receptor despite numerous experimental approaches. Although the T cell repertoire is mainly determined, in a single individual, by the alleles expressed at the MHC locus, one can postulate that it could also be influenced by the existence of alleles of the germ line gene(s) encoding the T cell receptor. If so, an analysis of the T cell fine specificity in mice of the same H-2 haplotype with different background genes might permit the mapping of the genes coding for this receptor. Such an experimental approach requires the use of an antigen consisting of only one major determinant. Several recent observations suggested to us that the hapten p-azobenzenearsonate (ABA) was a suitable model for such investigations. Thus, we decided to compare the specific pattern of responses to ABA-tyrosine, ABA-histidine and to free ABA in different inbred mouse strains. We report here that the lymph node T cell proliferative response to these molecules is under the control of an ABA-specific Ir gene. The ABA-Tyr conjugate is the most potent immunogen of the three in vivo as well as in vitro. High responder strains to ABA-His or ABA are included in the group of high responders to ABA-Tyr suggesting that the response to the three molecules is under the control of the same Ir-gene. The pattern of the response is also influenced by background gene(s). One of these can be localized on chromosome 12 using congenic mice. No close linkage to IgCH markers or VH idiotypes can be demonstrated but a linkage of this gene(s) to the Pre-1 locus seems possible. B lymphocytes do not seem to account for the involvement of Chr.12-genes in the response since; in our experimental system, they do not present ABA to T cells nor do they proliferate in the assays. Similarly, ABA-Tyr-antibody complexes do not enhance macrophages presentation of ABA to T cells, which supports the conclusion that IgCH or VH gene products are not involved in the control of the response.(ABSTRACT TRUNCATED AT 400 WORDS)

All A/J mice produce anti-p-azophenylarsonate (anti-Ar) antibodies, some of which share a cross-reactive idiotype. The idiotype can be suppressed by treatment with anti-idiotypic antiserum before immunization, although normal concentrations of anti-Ar antibodies are synthesized. We have previously reported that such suppressed mice, if hyperimmunized and then allowed to rest, contain up to 10% of splenic T cells which form rosettes with autologous RBC coated with Fab fragments of anti-Ar antibodies bearing the idiotype. Our present results indicate that the rosette-forming T cells include the idiotype-specific suppressor T-cell population. The suppressive activity is largely depleted by removal of the rosette-forming lymphocytes, and the rosettes themselves are highly suppressive. The data do not establish whether all of the idiotype-specific rosette-forming cells are suppressor cells. The system may provide a source of large numbers of suppressor cells for further study, and facilitate investigation of the mechanism of generation of idiotype-specific suppressor cells.

Serological analysis of idiotypic specificities present in azobenzenearsonate (ABA)-specific first-order suppressor T cell factors (TsF1) from C.AL-20 and BALB/c mice revealed a significant difference between TsF from these two strains of mice. The idiotypic composition of TsF1 from BALB/c mice appears to be more heterogeneous, and at least two different fractions can be readily identified. One bears the characteristic BALB/c-associated CRI(C) (crossreactive idiotype) determinants, and the other is non-CRI(C)-bearing. Analysis of ABA-specific TsF1 from animals lacking B cells uncovered a fundamental change in the expression of their idiotypic specificities. TsF from rabbit anti-mouse IgM (anti-mu)-treated C.AL-20 mice failed to express the characteristic CRI(A) determinants. Instead, they express CRI(C) specificities. Similarly, TsF1 from anti-mu-treated BALB/c mice did not express their characteristic CRI(C) specificities, but rather express CRI(A) determinants. These experiments provide strong evidence that the Igh restriction specificity of TsF is dictated by the particular idiotypic specificities expressed. They also clearly demonstrate that B cells and their products play an important role in establishing the idiotypic composition and repertoire of suppressor T cells.

The administration of azobenzenearsonate-modified syngeneic spleen cells (ABA-SC) intravenously induces a population of first order hapten-specific inducer suppressor T cells (Ts1), which downregulate various aspects of T-cell-mediated immune responses via a well defined suppressor-T-cell pathway. In this study, we investigated the effects of these suppressor cells on the generation of ABA-specific cytolytic T lymphocytes (CTL) and helper T cells (Th) in vivo. We found evidence for functional impairment of ABA-activated Th and ABA-specific CTL precursors (CTLp) in the suppressed animals by a number of different in vitro criteria. Functional analysis of ABA-specific CTLp and ABA-activated Th in suppressed animals revealed that ABA-specific Ts inhibit the generation of CTL by impairing the antigen-specific activation of Th, which may in turn, prevent the clonal expansion of antigen-specific CTLp. The significance of these findings in relationship to our understanding of the cellular interactions necessary for the generation of CTL and the mode of action and mechanisms of suppressor T cells is discussed.

To study T cell idiotype expression at the functional level, we developed a hapten-specific delayed-type hypersensitivity (DTH) system by which we avoid the complication of anti-hapten antibody and which is specific only for the immunizing hapten, and not for conjugate specific determinants. Immunization with ABA-Tyr and challenge with ABA diazonium induced footpad swelling with the characteristics of DTH. Anti-ABA antibodies did not contribute to this reaction, as they were undetectable in mice immunized with ABA-Tyr. Furthermore, this ABA-Tyr-specific DTH was under Ir gene control identical to that reported for ABA-Tyr-specific lymphocyte proliferation. All mouse strains tested responded to ABA-Tyr except those of the b haplotype across the entire Ia region. In contrast, contact sensitivity induced by ABA diazonium was not under apparent Ir gene control, probably reflecting 1) different specificities of the induced T cells and 2) the production of anti-ABA antibodies that contribute to the footpad swelling via an Arthus reaction. Having shown that ABA-Tyr can induce T cells mediating DTH, we then examined ABA-Tyr-reactive T cell clones, propagated in vitro, for their ability to mediate DTH. Such clones elicited a response identical to that seen with in vivo immunization with respect to dose dependency, I-Ak restriction, and antigen specificity.

Delayed-type hypersensitivity (DTH) to the azobenzenearsonate (ABA) hapten can be readily induced in A/J mice injecting ABA-coupled syngeneic spleen cells subcutaneously. To further characterize this T-cell-dependent immunological phenomenon, the effect of passively administered anti-cross-reactive idiotype common to anti-ABA antibodies of A/J mice (CRI) antibodies on the development of ABA-specific DTH was investigated. Animals given daily injections (of minute amounts) of anti-CRI antibodies subsequent to immunization with ABA-coupled cells show significant reduction of ABA specific responses. This inhibition is antigen specific and requires the intact immunoglobulin molecule, as F(ab')2 treatments were ineffective in suppressing the reaction. Investigations of the mechanism of the anti-CRI-induced suppression of ABA DTH revealed that the observed suppression is a result of the activation of suppressor cells. Spleen cells taken from animals which received anti-CRI antibodies were able to adoptively transfer suppression to naive recipients. This suppression was shown to be mediated by T cells, as anti-Thy1.2 plus complement completely abrogated the transfer of suppression. In addition, animals pretreated with low doses of cyclophosphamide were not suppressed by the administration of anti-CRI antibodies. The genetic restriction of anti-CRI-induced suppression was demonstrated. Antibodies to the major cross-reactive idiotype, (CRI) associated with anti-ABA antibodies in A/J mice were unable to suppress the development of DTH to ABA in BALB/c mice (H-2d, Igh-1a). Such antibodies were, however, fully active in suppressing ABA DTH in the allotype-congenic C.AL-20 strain which has an allotype (Igh-1d) similar to that of A/J (Igh-1e) on a BALB/c background, and which produces humoral antibodies with the CRI.

Our previous study demonstrated (1) that the presence of charged groups (amino and carboxyl groups) at the alpha-carbon of tyrosine is essential for activation of azobenzenearsonate-L-tryosine (ABA-L-Tyr specific T cells, and (2) that T cells recognizes ABA-L-Tyr in association with macromolecules on syngeneic spleen cells used as antigen presenting cells (APC). The present study was undertaken to confirm that the macromolecules on APC are Ia molecules, by using L cells transfected with A beta k and A alpha k genes as APC. I-Ak restricted ABA-L-Tyr specific cloned T cells, and T hybridoma cells were activated by ABA-L-Tyr in the presence of the L cell transfectants, of which expression of I-Ak molecules had been proven by specific binding of anti-I-Ak monoclonal antibody (MAb) 10.2.16 on the cell surface. The pattern of responses of I-Ak restricted ABA-L-Tyr specific T cells to various ABA-Tyr derivatives presented by the I-Ak expressing L cell transfectants was similar to the pattern obtained by using H-2k spleen cells as APC. Thus, ABA-L-Tyr and ABA-Tyr derivatives, which have both amino and carboxyl groups at the alpha-carbon of Tyr, presented by the L cell transfectants triggered good response of I-Ak restricted ABA-L-Tyr specific T cells. By contrast, ABA-Tyr derivatives, which lack the amino or carboxyl group, or both groups, at the alpha-carbon of Tyr, presented by the L cell transfectants could not activate the ABA-L-Tyr specific T cells at all. Furthermore, anti-I-Ak MAb, but not anti-I-Ek MAb, inhibited completely the response of I-Ak restricted ABA-L-Tyr specific T cells to ABA-L-Tyr presented by the L cell transfectants. These results indicate strongly that the macromolecules on APC which associate with ABA-L-Tyr are A beta k A alpha k gene products, i.e., I-Ak molecules.

Despite recent advances in our understanding of T cell antigen receptor structure, relatively little is known about the role of this receptor in MHC-restricted antigen recognition. To study this problem, we have developed a panel of ABA-Tyr-reactive, I-Ak-restricted T cell clones that differ in their ability to recognize structural analogs of ABA-Tyr. Three fine specificity groups have been defined. In each group, ABA-Tyr elicited the strongest response of any of the antigens tested. Group I clones responded to ABA-conjugated hydroxyphenyl-ethanol (ABA-HPE). Group II clones responded to ABA-conjugated hydroxyphenyl-methanol (ABA-HPM) but not to ABA-HPE, and group III clones responded only to ABA-Tyr. These studies show that differences as small as a single methylene group can dramatically affect fine specificity. Because these clones are all I-Ak-restricted, it was possible to correlate receptor serology with fine specificity. To this end, monoclonal anti-clonotypes were made against clone 16-F2 from group I and used to study the relationship between fine specificity and clonotype expression. A panel of 15 T cell clones studied with four anti-clonotype antibodies showed a strict correlation between clonotype expression and fine specificity. Taken together, these data suggest that the structure recognized by the anti-clonotype antibodies is a determinant of receptor fine specificity.

A series of antigens was synthesized in which peptide spacers were inserted between the ABA group and TNP-Ficoll. When these antigens were used to assess helper activity in Lewis rats immunized with ABA-tyr, it was found that an increase in the ABA-epitope density and an increase in the peptide spacer size both increased the efficacy of the antigen in eliciting ABA-specific help, manifest by an enhancement of anti-TNP PFC. Substitution of D- for L-amino acids progressively decreased the ability of these conjugates to elicit help until D-tyr-D-ala-D-ala, which when used for coupling ABA to the TNP-Ficoll resulted in a nonimmunogenic molecule. When these same Ficoll conjugates were used to study ABA-specific in vitro proliferative responses, it was found that introduction of even a single D-amino acid into the spacer greatly reduced reactivity. By contrast the ABA-peptides, free of the Ficoll backbone, were all equivalent on a molar basis in their ability to elicit ABA-specific in vitro proliferation, regardless of their content of D-amino acids. These results suggest some form of digestion is required in the processing of these antigens before they can elicit helper activity. This processing can occur only if one or more L-amino acid residues are present. If the ABA-peptides are free of the Ficoll backbone, they are all capable of stimulating T cell proliferation without apparent further processing.

The genetic restrictions of the activation of third-order suppressor cells (Ts3) were studied in mice, using two different types of anti-azobenzenearsonate (ABA)-immune responses, namely delayed-type hypersensitivity (DTH) and cytotoxic T lymphocyte (CTL) generation. Ts2 cells were induced in several different strains of mice by injecting monoclonal T hybridoma molecules or first-order suppressor factors (TsF1) originating in A/J (H-2a, Igh-1e) mice and then testing the TsF2 molecules derived from these Ts2 in A/J and A.By (H-2b, Igh-1e) or (A/J X A.By)F1 (H-2a/b, Igh-1e) and (C57Bl/6 X A/J)F1 (H-2b/a, Igh-1e) mice. It was shown that the activity of TsF2 was restricted to the I-J of the strain in which Ts2 was induced. By genetic analysis, restriction was shown to be due to the requirement of H-2 identity between ABA-coupled cells used for Ts3 activation and the strain of the TsF2 origin. Moreover, by using H-2-congenic ABA-coupled cells, we were also able to precisely map and demonstrate that ABA-coupled cells I-J identical to TsF2 induced in various strains were necessary for effective suppression to occur. This selective activation of Ts3 suggested the existence of I-J-related antigen presentation for suppression as the counterpart of I-A or I-A-I-E-restricted antigen presentation for positive immune responses.

Treatment of responder cells with monoclonal anti-Ly-1,2 antibodies plus complement in vitro completely eliminated their ability to generate azobenzenearsonate (ABA)-specific cytolytic T lymphocytes (CTL). However, addition of the concanavalin A-stimulated supernatants of rat spleen cells (Con A-Sup) can fully reconstitute the response. Therefore, Lyt-1,2-bearing T cells are required for the generation of ABA-specific CTL, and such requirement can be replaced by factors present in the Con A- sup. Suppressor T cells (Ts), when adoptively transferred into naive recipients, will inhibit the in vivo priming of CTL. This inhibition can also be reversed by in vitro addition of Con A-Sup. furthermore, mice serving as donors of Ts also show profound unresponsiveness when primed and restimulated in vitro. In contrast to the Ts-mediated inhibition, in vitro addition of Con A-Sup was unable to abolish the unresponsiveness observed in these cultures. Thus, we identified two unresponsive states in a hapten-specific killing system that differ in their ability to be reconstituted by Con A-Sup.

T-cell derived suppressor factor(s) (SF) specific for azobenzenearsonate (ABA) were prepared by the mechanical disruption of suppressor cells. Such suppressor factors were adsorbed to and recovered from immunoadsorbents prepared from the F(ab')2 fragments of rabbit immunoglobulin directed against the cross-reactive idiotype of A/J anti-ABA antibodies. These ABA-suppressor factors were not retained on Sepharose 4B immunoadsorbent columns which had been coupled with F(ab')2 fragments or normal rabbit immunoglobulins prepared from prebleeds of rabbits used to make anti-idiotypic antiserum. The specificity of the F(ab')2 rabbit anti-idiotypic serum was established by direct idiotypic-binding assays and by affinity purification over an immunoadsorbent consisting of CRI+ anti-ABA immunoglobulin from A/J mice. ABA-suppressor factors were shown to be specifically absorbed and eluted from F(ab')2 anti-idiotypic columns. Futhermore, the eluted suppressor factor can be specifically reabsorbed and recovered from a second anti-idiotypic immunoadsorbent. The concordance between antigen-binding specificity and the presence of idiotypic determinants was demonstrated by adsorbing ABA SF to antigen columns and then fractionating the ABA-specific factor on anti-idiotypic immunoadsorbents. ABA-suppressor factors were shown to be specifically retained on immunoadsorbents directed against major histocompatibility complex (MHC) determinants. Factor eluted from anti-MHC columns could then be specifically adsorbed to anti-idiotypic immunoadsorbents. This suggests that the same molecular complex that is recognized by the H-2 alloantiserum is specifically adsorbed to an anti-idiotypic immunoadsorbent. Genetic analysis of the expression of CRI+ suppressor factor was performed using the C.AL-20 mouse strain which has the AL/N allotype and produces CRI+ anti-ABA immunoglobulins. The implication of these findings to the nature of T-cell-derived regulatory molecules is discussed.

In a previous report characterizing the arsonate (ABA)-specific plaque-forming cell (PFC) responses of A/J mice induced by ABA-KLH, two interesting characteristics of the idiotypic (Id) profile were noted: (1) an apparent Id selectivity in the isotype switch since the earliest appearing IgG PFC in the primary response were significantly more "cross-reactive Id" (CRI)-dominant than the IgM PFC population, and, (2) a temporal waning of CRI dominance with time among IgG PFC, from 75-100% CRI+ PFC to about 25-45% CRI+ PFC in secondary responses. Experiments were performed to determine whether these effects are largely attributable to T or to B cells. Mice were immunized with a T-independent (TI) form of ABA (ABA-Brucella abortus) and apparent Id selectivity was observed; the earliest IgG PFC averaged 75% CRI+ while IgM PFC were only 39% CRI+. Due to the TI nature of the Ag, this provides suggestive, but not conclusive, evidence that the Id asymmetry in the isotype switch may be attributable to the direct interaction of Ag with B cells. Other studies addressed the temporal shift in CRI dominance. First, it was found that preexposure of mice to either KLH or to ABA (on an irrelevant carrier) resulted in diminished CRI dominance in subsequent "primary" responses to ABA-KLH. Secondly, adoptive transfer experiments with B and T cells from virgin mice (Bv, Tv) or ABA-KLH-primed mice (Bp, Tp) showed that recipients of Bv + Tp or Bp + Tv generated anti-ABA PFC responses with intermediate CRI levels. The Tv cells had some preferential tendency to activate CRI+ clones in the Bp population. The results demonstrate that CRI levels are jointly determined by the immune status of both B and T cells. A simple model is offered which accounts for early Id dominance and its gradual decline and has as its central postulate the assumption that CRI+ B cells in the virgin ABA-specific repertoire have an affinity advantage over CRI- clones.

A ubiquitous nonimmunoglobulin molecule that binds p-azobenzenearsonate (ABA) has been detected in the cytoplasm of several murine cell lines, including T cell hybridomas as well as in normal liver and spleen. Similar to many recently described antigen-specific T cell factors, this ABA-binding protein has a 62,000 mol wt, and, when analyzed by direct binding, the molecule reacts with several different rabbit anti-idiotypic antisera specific to the ABA system. The presence of this antigen-specific, "idiotype positive" molecule in many different cells indicates that it is not an important immunoregulatory molecule.

Immunization of mice with the ABA-GAT conjugate stimulates GAT-specific T helper cells in GAT-responder animals and ABA-specific helpers in nonresponders. Unexpectedly, immunization of (responder X nonresponder) F1 mice, which have the GAT-responder phenotype, leads to the recruitment of both ABA- and GAT-specific clones of T helper lymphocytes. The GAT-reactive population is restricted to the haplotype of the responder parent (Iak), whereas ABA-specific T cells are mostly restricted to the nonresponder one (Ias). This is demonstrated by the ability of monoclonal antibodies to parental la antigens to inhibit T cell proliferation to GAT or ABA-Tyr in vitro. Consistently, ABA-GAT-primed F1 T cells can only activate nonresponder B cells to proliferate in the presence of ABA-Tyr and responder B lymphocytes in the presence of GAT. Furthermore, F1 T cells seem to recognize both ABA and GAT epitopes only in association with molecules encoded by the I-A subregion. Analysis of ABA-specific F1 T cell lines generated by in vitro stimulation with ABA-Tyr or ABA-GAT demonstrates a competition between GAT- and ABA-specific T cells present in the hybrid T cell repertoire and restricted to the same parental I-Ak molecule. The results indicate that F1 macrophages can present both ABA and GAT epitopes to T cells in association with the two parental and hybrid Ia determinants. It seems unlikely that the absence of GAT-specific T cells restricted to the nonresponder I-A in the F1 is due to suppressor T cells. Thus, the competition model that we propose, to explain the selective F1 T cell response to ABA-GAT, leads us to believe that GAT nonresponder animals may lack clones capable of recognizing, with a high affinity, I-As + GAT.