Genetic control of immune responses in vitro. V. Stimulation of suppressor T cells in nonresponder mice by the terpolymer L-glutamic acid 60-L-alanine 30-L-tyrosine 10 (GAT)

CD8+ suppressor T cells resurrected

This review focuses on the role of antigen-specific T cells that mediate active inhibition of immune responses over the past 35 years since their initial description. The field has experienced several changes in the accepted paradigm of such suppressor/regulatory T cells, from initial indications that such cells were CD8(+), to the view that such cells did not exist, to the identification of the transcription factor Foxp3 as a key orchestrator of inhibitory function. Although most Foxp3(+) cells in a resting animal are CD4(+)CD25(+) cells, Foxp3 expression and inhibitory function can be induced by antigens in the periphery by selective cytokine conditions, particularly TGF-beta. Such induced T cells occur within both the CD4 and the CD8 T-cell lineages and appear to mediate suppression by inhibiting the costimulatory activity of antigen-presenting cells and the production of inhibitory cytokines. Recent data generated by analysis of TCR Tg T cells that do not select many Foxp3-positive cells during thymic development are reviewed, emphasizing the pattern of "linked suppression" and focus of the relative potency of different mechanisms of suppression.

Special regulatory T-cell review: A rose by any other name: from suppressor T cells to Tregs, approbation to unbridled enthusiasm

In the early 1970s a spate of papers by research groups around the world provided evidence for a negative regulatory role of thymus-derived lymphocytes (T cells). In 1971, Gershon and Kondo published a seminal paper in Immunology entitled 'Infectious Immunological Tolerance' indicating that such negative regulation could be a dominant effect that prevented otherwise 'helpful' T cells from mediating their function. Over the next decade, suppressor T cells, as these negative regulatory cells became known, were intensively investigated and a complex set of interacting cells and soluble factors were described as mediators in this process of immune regulation. In the early 1980s, however, biochemical and molecular experiments raised questions about the interpretation of the earlier studies, and within a few years, the term 'suppressor T cell' had all but disappeared from prominence and research on this phenomenon was held in poor esteem. While this was happening, new studies appeared suggesting that a subset of T cells played a critical role in preventing autoimmunity. These T cells, eventually dubbed 'regulatory T cells', have become a major focus of modern cellular immunological investigation, with a predominance that perhaps eclipses even that seen in the earlier period of suppressor T cell ascendancy. This brief review summarizes the rise and fall of 'suppressorology' and the possibility that Tregs are a modern rediscovery of suppressor T cells made convincing by more robust models for their study and better reagents for their identification and analysis.

Special regulatory T-cell review: Suppressors regulated but unsuppressed

The rise-and-fall and reincarnation of suppressor T cells is reviewed from the perspective of a participant in the field.

Special regulatory T-cell review: T-cell dependent suppression revisited

The concept of T-cell dependent regulation of immune responses has been a central tenet of immunological thinking since the delineation of the two cell system in the 1960s. Indeed T-cell dependent suppression was discovered before MHC restriction. When reviewing the data from the original wave of suppression, it is intriguing to reflect not just on the decline and fall of suppressor T cells in the 1980s, but on their equally dramatic return to respectability over the past decade. Hopefully their resurgence will be supported by solid mechanistic data that will underpin their central place in our current and future understanding of the immune system. Cannon to right of them, Cannon to left of them, Cannon in front of them Volley'd and thunder'd Storm'd at with shot and shell, Boldly they rode and well, Into the jaws of Death, Into the mouth of Hell, Rode the six hundred (suppressionists). (Adapted from The Charge of the Light Brigade, Alfred, Lord Tennyson)

A T cell receptor antagonist peptide induces T cells that mediate bystander suppression and prevent autoimmune encephalomyelitis induced with multiple myelin antigens

Experimental autoimmune encephalomyelitis (EAE) induced with myelin proteolipid protein (PLP) residues 139-151 (HSLGKWLGHPDKF) can be prevented by treatment with a T cell receptor (TCR) antagonist peptide (L144/R147) generated by substituting at the two principal TCR contact residues in the encephalitogenic peptide. The TCR antagonist peptide blocks activation of encephalitogenic Th1 helper cells in vitro, but the mechanisms by which the antagonist peptide blocks EAE in vivo are not clear. Immunization with L144/R147 did not inhibit generation of PLP-(139-151)-specific T cells in vivo. Furthermore, preimmunization with L144/R147 protected mice from EAE induced with the encephalitogenic peptides PLP-(178-191) and myelin oligodendrocyte protein (MOG) residues 92-106 and with mouse myelin basic protein (MBP). These data suggest that the L144/R147 peptide does not act as an antagonist in vivo but mediates bystander suppression, probably by the generation of regulatory T cells. To confirm this we generated T cell lines and clones from animals immunized with PLP-(139-151) plus L144/R147. T cells specific for L144/R147 peptide were crossreactive with the native PLP-(139-151) peptide, produced Th2/Th0 cytokines, and suppressed EAE upon adoptive transfer. These studies demonstrate that TCR antagonist peptides may have multiple biological effects in vivo. One of the principal mechanisms by which these peptides inhibit autoimmunity is by the induction of regulatory T cells, leading to bystander suppression of EAE. These results have important implications for the treatment of autoimmune diseases where there are autopathogenic responses to multiple antigens in the target organ.

Establishment of stable CD8+ suppressor T cell clones and the analysis of their suppressive function

Stable CD8+ suppressor T cell (Ts) clones were established by a relatively simple method. Keyhole limpet hemocyanin (KLH)-primed spleen cells from C3H mice were depleted of B cells and CD4+ T cells by panning and cytotoxic treatment, and the resulting CD8+ T cells were periodically stimulated with antigen and irradiated syngeneic spleen cells followed by manifestation in interleukin-2 (IL-2) containing medium. T cell clones with a definite suppressor function were established by limiting dilution. They were defined as classical effector type Ts of CD8+ phenotype as they had constant and definite suppressor functions in antigen-induced T cell proliferation and specific antibody response against T cell-dependent antigens without detectable cytotoxic activity against both antigen presenting cells (APC) and helper T cells (Th). They showed no helper activity for B cells and produced no detectable helper type lymphokines such as IL-2 and IL-4. CD8+ Ts clones were able to inhibit the antigen-induced IL-2 production of normal and cloned T cells. Their suppressive activity was antigen-nonspecific and major histocompatibility complex-unrestricted. CD8+ Ts clones were also able to suppress the proliferative response of Th clones induced by immobilized anti-T cell receptor (TcR) and anti-CD3 mAbs but not the response induced by concanavalin A (ConA) and IL-2. All the CD8+ T cell clones established independently utilized the TcR V beta 8 gene. Syngeneic antigen presenting cells could induce proliferation of these CD8+ clones, which was blocked by anti-CD8 and anti-I-Ak monoclonal antibody (mAb) but not by anti-class I mAbs. The stimulation of CD8+ Ts clones with immobilized anti-CD3 resulted in the release of a suppressor factor(s) that potently inhibited the antigen-induced proliferation of CD4+ Th clones and the in vitro secondary antibody formation.

In vivo-induced suppression of T cell proliferation: the relationship between the specificity of induction and control

We have previously shown that sc immunization of C57BL/10 (H-2b) mice with the tobacco mosaic virus protein (TMVP) or with its tryptic peptide number 8, representing residues 93-112 of TMVP, induces T cells which proliferate in vitro in response to TMVP and to peptide 8. In contrast, immunization of B10.BR (H-2k) mice either with TMVP or with peptide 8 induces T cells which respond in vitro to the homologous but not the heterologous Ag. In the present article , we report that in the B10.BR (H-2k) strain, ip prepriming with (TMVP) 7 days prior to sc immunization with peptide 8 causes a drastic reduction in the in vitro proliferative response of peptide 8-specific T cells while no such effect is seen in the congenic C57BL/10 (H-2b) strain. This suppression of T cell responsiveness can be transferred with TMVP-primed spleen cells. Moreover, deleting T cells from the transferred spleen cells abrogates the suppressive effect. In both H-2 haplotypes, ip prepriming with peptide 8 causes suppression of the proliferative T cell response induced by subsequent immunization with peptide 8. This prepriming has no effect on the response to TMVP immunization of B10.BR mice but does effect the response of C57BL/10 mice. Using various synthetic peptides to analyze the specificity of the suppression, we have determined that (1) T cells involved in the suppression of the proliferative T cell response to a single peptide determinant do not suppress the proliferative T cell response to other determinants on the protein antigen and (2) these T cells with suppressor function, and proliferating T cells which are ultimately regulated, can exhibit specificity for the same epitope. These studies suggest that there may exist fundamental differences as to how T cells which participate in suppression an proliferating T cells (which include mainly T helper cells) recognize protein antigens.

Solid-phase antigen binding by purified immunoproteins from antigen-specific monoclonal T cell hybridomas

We have developed two distinct solid-phase immunoassays for the detection of antigen binding activity by products of antigen binding T cell hybridomas in the absence of MHC. Two suppressor T cell hybridomas studied (34s-18 and 34s-704) are specific for keyhole limpet hemocyanin, a protein antigen, and the other suppressor T cell hybridoma (51H7D) binds specifically to the arsonate hapten. We have adapted these hybridomas to growth in serum-free medium and have isolated molecules with antigen binding activity both from the cell membranes and from the culture fluid in which the cells had been grown. The antigen binding molecules (ABM) produced by the KLH-specific hybridomas bound best to native hemocyanin; binding was decreased when KLH was denatured by reduction and alkylation and no binding was found to an arthropod (Limulus) hemocyanin. The arsonate binding hybridoma, on the other hand, produced molecules specific for this hapten; they showed no capacity to bind KLH. The antigen binding molecules affinity-purified from all three T hybridomas have intact masses of either 145,000, 67,000 or 48,000 when run in SDS-PAGE under non-reducing conditions. Following reduction, ABM resolve in SDS-PAGE into a complex of polypeptide chains having apparent masses of 65,000, 56,000 and 49,000, with either a pair of bands at 26,000 and 22,000, or with a single band at 32,000, which is consistent with the size of translation products of mRNA previously isolated from these hybridomas. Two of the hybridomas, 34s-18 and 34s-704, used for isolation of antigen binding products in this study, were previously reported to lack detectable rearranged gamma or beta genes and therefore to lack expression of the alpha/beta or gamma/delta heterodimers. The antigen binding molecules react in solid-phase immunoassay with some antibodies specific for variable (first framework) region and joining (J) region peptide sequences predicted from T cell receptor gene sequence. Furthermore, the affinity-purified antigen binding molecules from mouse T cell hybridomas cross-react in ELISA with goat anti-rabbit IgG and not with protein G, thus allowing the use of these commercially available reagents in standard laboratory assays. Interestingly, ABM anchored in intact cell membranes, which could be shown to specifically bind antigen, did not cross-react with goat anti-rabbit IgG, indicating that the cross-reactive moiety is not detectable when the ABM are in this situation.(ABSTRACT TRUNCATED AT 400 WORDS)

Regulation of IgG responses by helper and suppressor T cells activated by pneumococcal capsular polysaccharides

Type 2 antigens are usually unable to prime the helper T cells (TH) required for secondary IgG antibody responses. However, previous results from this laboratory indicated that low doses of the type 2 antigen polyvinylpyrrolidone (PVP) could activate T cells which provided help to PVP-primed B cells for the production of PVP-specific IgG antibody. Therefore, it was of interest to determine if other type 2 antigens may also be able to activate TH. Low doses of S19 or S3 (subimmunogenic for a primary IgM response) activated TH capable of providing help to S19- or S3-CRBC-primed B cells for a secondary IgG response. Higher doses of these antigens (optimally immunogenic for a primary IgM response) activated suppressor T cells (TS). Removal of these TS prior to transfer of T cells to recipient mice resulted in expression of TH function. Therefore, the preferential activation of TH versus TS was dependent on the dose of antigen used for priming. TH activated by low doses of S19 expressed Thy 1 and L3T4 and were antigen specific. In contrast to the ability of low doses of PVP to prime B cells for secondary IgG responses, low doses of S3 and S19 did not prime capsular polysaccharide-specific IgG memory B cells. High doses of S3 were able to prime B cells if TS precursors were first removed by treatment of mice with cyclophosphamide (Cy), whereas high doses of S19 did not prime B cells for secondary IgG responses in either Cy-treated or control mice. These results are discussed in relation to the general observations that type 2 antigens may not activate antigen-specific TH.

Antigen processing for presentation to T lymphocytes: function, mechanisms, and implications for the T-cell repertoire

Antigen processing encompasses the metabolic events that a protein antigen must undergo in or on the antigen-presenting cell before it can be recognized by the T lymphocyte. It appears that a primary goal of these events is to unfold the protein to expose residues that are buried in the native conformation, which is designed to be soluble in water. The APC usually accomplishes this task by proteolytic cleavage of the protein, but we have found that artificial unfolding without proteolysis is sufficient. The purpose of unfolding may be to allow different faces of the antigenic site to bind simultaneously to the T-cell receptor and the MHC molecule on the APC, or to interact with other structures on the membrane of the APC. This requirement for unfolding appears to apply to everything from small peptides to large multimeric proteins. We have found that the way the antigen is processed and the structure of the fragments produced can greatly affect the availability of antigenic sites. For instance, some antigenic sites are not recognized when the native protein is used as immunogen, despite the fact that immunization with a small peptide corresponding to that site reveals both the ability of the site to bind to MHC molecules of the animal in question and the presence of a T-cell repertoire specific for that site. The antigenic site is not destroyed by processing, since it can be presented by the same F1 APC to T cells of another MHC type. Similarly, cross-reactivity between homologous epitopes of related proteins may occur at the peptide level even though the native proteins do not crossreact for the same T-cell clone. Since these events occur with monoclonal T cells, they cannot be due to suppressor cells specific for other sites on the native molecule. The best explanation is that the products of natural processing of the protein are larger than the peptides corresponding to the minimal antigenic sites, and contain hindering structures that interfere with binding to some MHC molecules and not others, or to some T-cell receptors and not others. Thus, antigen processing is a third factor that can lead to apparent Ir gene defects - in addition to MHC specificity and holes in the T-cell repertoire - and can significantly influence which antigenic sites are immunodominant.(ABSTRACT TRUNCATED AT 400 WORDS)

Murine transfer factor. IV. Studies with genetically regulated immune responses

Transfer factor-containing dialysates from mice that were either high or low responders to GAT10, GLA5, or ovalbumin were assayed for their ability to transfer delayed hypersensitivity to murine recipients of either high or low responder phenotype. Dialysates from high responder strains contained transfer factor that would transfer delayed hypersensitivity to both high and low responder recipients. These transfers were not restricted by disparities at the MHC or Igh loci. Identically prepared materials from low responder donors contained little or no transfer factor activity and would not transfer delayed hypersensitivity to either high or low responder recipients. Thus, administration of transfer factor transfers the high responder phenotype to low responder recipients. The data also suggest that production of transfer factor is regulated by Ir genes but that the immunologic activities of transfer factor are not.

Influences of antigen processing on the expression of the T cell repertoire. Evidence for MHC-specific hindering structures on the products of processing

Two lines of evidence in the current study indicate that antigen processing is a major factor, in addition to MHC binding and T cell repertoire, that determines Ir gene responsiveness and epitope immunodominance. First, immunization with synthetic peptides of myoglobin sequences revealed new reactivities that had not appeared after priming with native myoglobin. For example, B10.S mice (H-2S) immune to equine myoglobin predominantly responded to peptide 102-118, whereas there was little, if any, response to this peptide in B10.BR (H-2k) mice immunized with native equine myoglobin. However, after immunization with the 102-118 peptide, both strains responded to the peptide. After in vitro restimulation, B10.BR T cells responded as well as B10.S T cells. Similarly, some individual 102-118-specific T cell clones from mice of both haplotypes showed similar dose responses and fine specificity patterns. Thus, low responsiveness to this site is due neither to a hole in the repertoire nor to a failure to bind to the appropriate MHC molecule. An alternative explanation was suggested by the observation that, whereas B10.S T cells from peptide 102-118-immune mice responded almost as well to whole myoglobin as to the peptide, the B10.BR T cells from peptide immune mice, while responding well to peptide, were poorly stimulated by whole myoglobin. Thus, the product of natural processing of equine myoglobin probably has hindering structures in the regions flanking the core epitope 102-118 that interfere with presentation by I-Ak but not I-AS. The second line of evidence that processing of native myoglobin may influence the apparent specificity of the T cell response was obtained using the I-Ad-restricted sperm whale myoglobin 102-118-specific clone 9.27. This clone discriminated readily between whole sperm whale myoglobin and equine myoglobin, but it did not distinguish between peptides corresponding to 102-118 of the sperm whale and equine sequences. This distinction between equine peptide and native equine myoglobin could be overcome by artificial "processing" of equine myoglobin with cyanogen bromide. In both sets of experiments, F1 APCs that present the same epitope well to T cells of another haplotype failed to overcome the defect, which was therefore not due to the availability of different processed cleavage fragments in APC of different haplotypes, as would be expected if there were MHC-linked processing. Thus, the differential responses to peptides versus native molecule for both I-Ad- and I-Ak-restricted clones appeared to depend on the restricting molecule used.(ABSTRACT TRUNCATED AT 400 WORDS)

The functional link between the immune suppression gene and Mhc class II molecules

The immune response to bovine insulin (BI) in the rat is controlled by the major histocompatibility complex (Mhc)-linked immune response gene (Ir-BI) and immune suppression gene (Is-BI). In the present study, we investigated the low responsiveness to BI in the WKAH rat (RT1k) and attempted to explore the functional link between Is-BI and Mhc class II molecules. Lymph node cells (LNC) from the low responder (WKAH) rats responded well to BI when a large amount of antigen was added to the culture in vitro or after OX8-bearing (OX8+) T cells were eliminated. These LNC, after the elimination of OX8+ cells, could show the RT1.Dk-restricted proliferative response upon in vitro challenge with BI, BI-B chain, or pork insulin. In addition, OX8+ T cells, which were activated with BI and antigen-presenting cells (APC) in vitro, suppressed the anti-BI response of W3/25-bearing proliferating T cells from BI-immunized rats. The results have demonstrated that proliferating T-cell repertoires do exist to BI, which recognize BI-B chain in the context of RT1.Dk molecules in the WKAH rat, and that the state of low responsiveness is mediated to a great extent by antigen-specific OX8+ suppressor T (Ts) cells. Furthermore, the elimination of APC or the addition to RT1.Bk-specific monoclonal antibody in the in vitro secondary activation culture of Ts cells diminished the suppressive activity of OX8+ Ts cells. In the induction phase of Ts cells it therefore seems to be necessary for these cells to recognize BI together with RT1.Bk molecules on APC.

Overlapping T cell antigenic sites on a synthetic peptide fragment from herpes simplex virus glycoprotein D, the degenerate MHC restriction elicited, and functional evidence for antigen-Ia interaction

Analysis of the B10.A T cell response to synthetic peptides representing the NH2-terminal 23 amino acids from the HSV glycoprotein D sequence revealed two antigenic determinants for T cells: one localized between residues 1-16 and the other between residues 8-23. The 1-16 site, which is helical, was recognized in the context of the Ia molecule, whereas the 8-23 site, which is nonhelical, was recognized in the context of the I-E molecule. The I-E-restricted response was found to be highly MHC degenerate in that T cell hybridomas specific for the 8-23 peptide responded to antigen on APCs derived from B10.A, B10.A(5R), and B10.A(9R) mice and showed differences in antigenic fine specificity with APCs of different haplotypes. These data support the idea of antigen-Ia interaction.

Molecular genetic characterization of the mRNA coding for an inducible suppressor factor specific for L-glutamic acid60-L-alanine30-L-tyrosine10

The suppressor T-cell hybridoma 1556A2.1 can be induced by the monoclonal L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT)-specific suppressor inducer 372B3.5 and soluble GAT to synthesize a disulfide-linked heterodimeric protein (GAT-TsF2), which directly suppresses a primary in vitro immune response to GAT. Induction and synthesis of the GAT-TsF2 protein is correlated with the appearance of specific mRNA, as detected by translation in vitro in a wheat germ cell-free extract of RNA isolated at various times after induction. The mRNA coding for the polypeptide chain that bears a serologically defined I-J determinant (I-J+ chain) appeared 8 hr after induction, whereas the mRNA coding for the antigen-binding chain (AB+ chain) was not detected until 16 hr after induction. The mRNAs coding for the individual chains sedimented as different species, suggesting that the two-chain factor is the product of two genes. The AB+ chain of the 1556A2.1 GAT-TsF2 was synthesized on membrane-bound polysomes, whereas the I-J+ chain was translated on free polysomes. The AB+ chain was synthesized from two independent mRNA species sedimenting at 10 S and 28 S, whereas a single 16S mRNA encoded the I-J+ chain. The in vitro translated I-J+ chain was bound by a monoclonal antibody against the I-J+ determinant of only the appropriate H-2 haplotype. These results suggest that posttranslational modification, including glycosylation, is not required for biological activity or for expression of the I-J epitope on the GAT-TsF2 molecule.

Strain-dependent protective effect of adult thymectomy on murine infection by Mycobacterium lepraemurium

C57BL/6, DBA/2, BALB/c and CBA mice were thymectomized as adults, or sham-thymectomized, and infected subcutaneously with 10(6) MLM. The number of MLM in the spleen and in the inoculated footpad was measured after 1 year of infection as well as the DTH reactions and the IgM and IgG antibody levels to MLM. Non-thymectomized mice exhibited a broad spectrum of resistance to MLM infection and of T cell mediated immunity grading from the highly resistant C57BL/6 strain to the highly susceptible CBA strain. In between, DBA/2 was found more resistant than BALB/c mice. Adult thymectomy reduced by 100 times the MLM number in the spleen of infected DBA/2 mice, without affecting that measured in the inoculated footpad, and significantly decreased DTH reaction in the same strain. No effect of adult thymectomy was observed in any other strain, except for an increase of anti MLM antibodies in BALB/c mice. These results may suggest that the medium-resistant DBA/2 strain develops after MLM infection suppressor T cells which favour MLM dissemination and are sensitive to adult.

Ir gene-controlled response to haptenated hen ovomucoid: isotypic specificity and dominant nonresponsiveness

We have examined the isotypic pattern of the response of mice to the Ir gene-controlled antigen, dinitrophenyl-ovomucoid (DNP-OM). H-2 kappa mice are high responders (HR); (H-2b,d mice are low responders (LR). The isotype patterns of HR and LR strains differ both quantitatively and qualitatively. In the primary response to doses of 20-100 micrograms DNP-OM, HR strains produce IgM and IgG antibodies, whereas LR strains produce only IgM. Background genes modify the kinetics of the IgGl primary response in HR strains, but no background was found which allowed an IgG response in a LR strain. In secondary responses, priming with 0.2 microgram DNP-OM increases secondary responses in HR strains, and decreases them in LR strains. Control of this response maps to I-A, and is not altered by the bm 12 I-Ab mutation. The LR phenotype is dominant in (HR X LR)F1 mice.

HLA-D restriction of "naturally occurring" MLR suppressor cells in acquired common variable hypogammaglobulinemia

Cells with capacity to suppress the mixed lymphocyte response (MLR) were detected in two patients with acquired common variable hypogammaglobulinemia (ACVH). No specificity with respect to the stimulating HLA type was observed. In one case of ACVH available for extensive study, there was evidence for HLA-D restriction of the suppressor cells. The patient's lymphocytes specifically suppressed the MLR of subjects who carried the same HLA-D type. Family studies confirmed that the suppressor activity was restricted to HLA-D, not DR, and segregated with the appropriate HLA haplotype. These observations suggest that an immune suppressor gene mapping in the HLA region may be involved in the pathogenesis of common variable hypogammaglobulinemia.

Differences in cellular immune response to L-glutamic acid60, L-alanine30, L-tyrosine10 (GAT) in full sibling embryo transfer cattle: examination of requirements for cell proliferation

Pairs of full sibling embryo transfer cattle that expressed identical MHC class I and II products were tested for their in vitro proliferative response to GAT. Peripheral blood mononuclear cells from these cattle were either high or low responders to GAT. Cells from certain pairs of MHC identical siblings gave opposite responses. Low responder animals were further tested to determine if they might respond to GAT with different kinetics, with secondary in vitro restimulation, or with exogenous help provided by interleukin 2. Also, the role of antigen presenting cells and suppressor T cells from low responder animals was investigated. Using appropriate in vitro conditions, cells from all animals tested could respond to GAT. However, MHC identical animals tested under similar conditions exhibited differences in their response to GAT which suggests the proliferative immune response was influenced by factors in addition to MHC coded products.

Mechanisms of genetic control of immune responses. II. Nonresponsiveness in BALB/c GT-specific cell-mediated immune responses does not correlate with the absence of functional T cells or the induction of suppressor T cells

The mechanisms underlying Ir gene control of CMI were addressed by examining the DTH and Tprlf responses specific for the synthetic polymers GT, GAT, and GA. We show that BALB/c mice (GAT/GA responders, GT nonresponders) primed with GT fail to develop DTH and Tprlf responses specific for GT, GAT, or GA. GAT immunization resulted in DTH responses that could be elicited not only with GAT and GA but also with GT, demonstrating that GT-specific TDH are present in nonresponder mice. GT-specific DTH was transferred with Thy-1+ Lyt-1+2-, H-2 I-restricted, nylon wool nonadherent cells. GA-primed BALB/c mice developed GAT- and GA-, but not GT-specific DTH responses, indicating that GA and GT do not cross-react at the T-cell level. The ability of GAT [but not a mixture of GA plus GT, or GT electrostatically complexed to the immunogenic carrier MBSA (GT-MBSA)] to induce GT-specific DTH suggested a requirement for covalent linkage of stimulatory 'GA' and nonstimulatory 'GT' determinants present on the GAT molecule. Similarly, GT-specific in vitro Tprlf responses could be demonstrated in GAT-primed mice exhibiting significant levels of GT-specific DTH but not in GT- or GT-MBSA-primed mice. Tolerization experiments also suggested that GT-specific Th were involved in the development of GT-specific DTH in GAT-primed mice. The GT nonresponsiveness of BALB/c mice for DTH and Tprlf responses could not be reversed by treatments designed to abrogate Ts activity (priming with GT-MBSA and CY injection), nor could GT-primed cells be shown to inhibit the development or elicitation of GT-specific CMI in GAT-primed mice during the afferent and/or efferent stages of DTH. Our results suggest that GT nonresponsiveness does not result from the absence of GT-specific T cells or preferential induction of Ts. The results are discussed in the context of hole-in-the-repertoire and antigen presentation (determinant selection) models of Ir gene control.

Epitope-specific regulation in Ir gene systems. I. Conditions for the induction of epitope-specific suppressors to poly(Glu60Ala30Tyr10)

Injection of responder mice with poly(Glu60Ala30Tyr10) (GAT) followed by immunization with GAT-methylated bovine serum albumin (GATMBSA) selectively suppresses anti-MBSA plaque-forming cell (PFC) and delayed hypersensitivity (DTH) reactions. Conversely, MBSA injection followed by GATMBSA immunization suppresses anti-GAT PFC and DTH, while anti-MBSA responses remain intact. Suppression occurs for doses of antigen which are optimally immunogenic. The suppression is specific and does not act in a bystander fashion. These results demonstrate that epitope-specific regulation is reciprocal, is not limited to humoral responses, and is not limited to molecules of low molecular weight.

Features of KLH-induced suppression in vivo: characterization of two pathways of suppression

Keyhole limpet hemocyanin (KLH) given at high dose (4 mg ip) in mice induced a state of unresponsiveness related to the activation of suppressor T cells. An early pathway of suppression is observed within the first 24 hr following KLH injection and is characterized by its cyclophosphamide (CPM) sensitivity and by the specificity of its effector phase, at the level of KLH helper T cells. A late pathway of suppression occurs at Day 3 following KLH injection and is characterized by its CPM resistance and the nonspecificity of its effector phase acting at the B-cell level. Indeed the anti-FLu antibody response to FLu Ovalbumin or thymus-independent antigen FLu LPS were found altered when these antigens were given with TNP KLH. These two pathways of suppression were found to last 8 months. These results suggest that KLH can trigger in an independent manner two pathways of suppression characterized by different CPM sensitivity and different target cells.

Intrinsic and extrinsic factors in protein antigenic structure

Recent advances in the preparation of synthetic peptide vaccines and the use of synthetic peptides as probes of antigenic structure and function have led to renewed interest in the prediction of antigenic sites recognized by antibodies and T cells. This review focuses on antibodies. Features intrinsic to the antigen, such as hydrophilicity and mobility, may be useful in the selection of amino acid sequences of the native protein that will elicit antibodies cross-reacting with peptides, or sequences which, as peptides, will be more likely to elicit antibodies cross-reactive with the native protein. Structural mobility may also contribute to protein-protein interactions in general. However, the entire accessible surface of a protein is likely to be detectable by a large enough panel of antibodies. Which of these antibodies are made in any individual depends on factors extrinsic to the antigen molecule, host factors such as self-tolerance, immune response genes, idiotype networks, and the immunoglobulin structural gene repertoire.

Repertoires of T cells directed against a large protein antigen, beta-galactosidase. II. Only certain T helper or T suppressor cells are relevant in particular regulatory interactions

11 cyanogen bromide (CB) peptides, comprising 70% of the large protein, Escherichia coli beta-galactosidase (GZ), were studied for their ability to induce T suppressor (Ts) cells capable of strongly suppressing the in vitro anti-fluorescein (FITC) response to GZ-FITC. Only CB-2 (amino acid residues 3-92) and CB-3 (residues 93-187) were found to bear such Ts-inducing epitopes. In examining the specificity of T helper cell (Th) targets susceptible to CB-2 and CB-3-specific Ts, it appeared that only nearly Th targets could be suppressed. Thus, CB-10-primed Th were not suppressed by either Ts; even CB-3-primed Ts did not suppress CB-2-specific Th, although CB-2-specific Ts were effective. Furthermore, analysis of the suppression pattern revealed a hierarchical use of potential epitopes on native GZ in triggering functional regulatory T cells. A dominant Th epitope near the amino terminus of GZ tops a hierarchy of potential Th, most of which are never engaged. The dominant determinant seems to exist on the peptide CB-2-3 (residues 3-187), and presumably is destroyed by its cleavage at Met 92; the Th cells that it induces are suppressible by each of the Ts-inducing peptides. In the GZ system, where the native antigen is quite large, the interactions between Th and Ts are highly circumscribed. This may be attributable to the topology of antigen fragments produced during processing; any relevant fragment must bear at least a Ts- and Th-reactive determinant to permit intercellular regulation. A final implication of these results is that, not only does the existence of a Th-inducing determinant depend on its being an appropriate distance from a B cell epitope, but the existence of Ts-inducing determinants likewise depends on the existence of a neighboring Th-B cell association.

Induction by monoclonal anti-idiotypic antibodies of an anti-poly(Glu60 Ala30 Tyr10) (GAT) immune response in GAT-responder and GAT-nonresponder mice

Two different monoclonal anti-idiotypic (Id) antibodies, HP-Id20 and HP-Id22, recognizing two discrete idiotopes characteristic of the anti-poly(Glu60 Ala30 Tyr10) (GAT) response were used to immunize BALB/c (GAT-responder) and DBA/1 (GAT-nonresponder) mice. The monoclonals were injected either copolymerized with keyhole limpet haemocyanin or polymerized with glutaraldehyde. The specific response was studied by two assays: (a) inhibition of binding of monoclonal anti-GAT antibody G5Bb2-2 to HP-Id20 and HP-Id22 and (b) GAT binding assays. In BALB/c GAT-responder mice, HP-Id20 and HP-Id22 immunization led to the preferential stimulation of immunoglobulin idiotypically related to anti-GAT antibodies (Ab1') and expressing anti-GAT activity. The results obtained with BALB/c nu/nu mice indicated that this response is T-cell-dependent. By means of the same experimental protocol GAT-nonresponder animals could be induced to produce anti-GAT antibodies after HP-Id immunization. This last result indicates that anti-Id immunization can bypass Ir gene control and does not preferentially stimulate the induction of GAT-specific T suppressor cells.

Regulatory mechanisms in immune responses to heterologous insulins. II. Suppressor T cell activation associated with nonresponsiveness in H-2b mice

Murine antibody responses to insulins are controlled by MHC-linked Ir genes. Although mice of the H-2b haplotype do not make antibody in response to pork insulin, we demonstrate in this communication that immunization with pork insulin stimulates radioresistant, Lyt-1+2- helper T cells that are capable of stimulating secondary antibody responses to pork insulin in vitro, but that this activity is masked by radiosensitive, Lyt-1-2+, I-J+ suppressor T cells. The suppressor T cells, present after immunization with pork insulin but not beef insulin, suppress the secondary response to pork but not beef insulin. The amino acid sequences of pork and beef insulins differ only at the A-chain loop; thus, pork insulin-specific suppressor T cells appear to recognize the A-chain loop determinant of pork insulin. The amino acid sequences of mouse and pork insulin are identical in the A-chain loop, which suggests that these suppressor T cells may be self-reactive. If this interpretation is correct, these suppressor T cells could be involved in the maintenance of self-tolerance to insulin. Nevertheless, these data clearly demonstrate that genetically determined nonresponsiveness in H-2b mice is conferred by activation of dominant, insulin-specific suppressor T cells (Ts), rather than by a defect in the stimulation of insulin-specific helper T cells (Th).

Immune response to the p-azobenzenearsonate-L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT) conjugate. III. Mechanisms of Ir gene-controlled phenotype conversion

Immunization of GT (random copolymer of L-glutamic acid51-L-tyrosine49) nonresponder animals with p-azobenzenearsonate (ABA) GT conjugates elicits an antibody response to both ABA and GT epitopes which is induced by ABA-specific T helper cells. Expression of these hapten-specific helpers is under the control of an I region gene which also regulates the proliferative T cell response to ABA. Conversion of the unresponsive phenotype to GT is, therefore, dependent on the ABA Ir gene and escapes the influence of the GT-specific I region-controlled suppressive pathway. Studies on the influence of ABA/polymer coupling ratio on T and B cell responses suggest that ABA-specific T cells, like conventional carrier-specific helpers, require linked interactions with B lymphocytes to provide helper signals. GAT (terpolymer of L-glutamic acid60-L-alanine30-L-tyrosine10) nonresponder animals immunized with ABA-GAT conjugates also develop an antibody response to ABA which is induced by ABA-specific T helper cells. Comparison of antibody affinity, specificity, isotypes and idiotypes in different mouse strains demonstrates that hapten-specific helper cells stimulate antibody responses to ABA which are qualitatively similar to those induced by GAT-specific helpers. However, ABA-specific helper cells do not permit the conversion of the I region gene-controlled nonresponder phenotype to GAT. The data suggests that high ABA density, which is required for optimal ABA help expression, extinguishes the immunogenicity of GAT determinants at both T and B cell levels.

Regulatory mechanisms of the immune response to heterologous insulins. I. Development and regulation of plaque-forming cell responses in vitro

We, as well as many other investigators, have been studying the regulation of immune responses to insulin as a model system of H-2 linked immune response (Ir) gene control. Although antibody responses by mice to heterologous insulins are qualitatively controlled, antibodies that are generated to one species of heterologous insulin cross react extensively with other species. The exquisite control of responsiveness is regulated by T cells that appear to recognize differences in the amino acid sequences of the A-chain loop of insulin. Our previous studies of the mechanism(s) by which Ir genes regulate T cell activity to insulin have been confined to an adoptive transfer model because traditional cell culture techniques using normal or insulin-primed spleen cells have failed to generate insulin-specific plaque-forming cell responses in vitro. In this communication we demonstrate that more vigorous immunization protocols and the use of lymph node T cells as a source of helper T cells can circumvent this problem. More importantly, all of the major features of the regulation of responses to insulin that have been observed in vivo are reflected in this in vitro system. Thus, these experiments provide the essential foundation for future dissection of the mechanism of Ir gene control of responses to insulin.

Immunoregulatory pathways in adult responder mice. I. Induction of GAT-specific tolerance and suppressor T cells for cellular and humoral responses

This report describes the alteration of helper-suppressor balances in an immune response (Ir) gene-controlled system by varying the route and form of antigen injection. Adult responder BALB/c mice develop Lyt 1+2-, T cells for delayed-type hypersensitivity (DTH), and T-cell proliferative (Tprlf) responses to subcutaneous injection of either poly(Glu60Ala30Tyr10) (GAT)-coupled syngeneic spleen cells (GAT-SP) or GAT emulsified in complete Freund's adjuvant. In contrast, intravenous injection of adult responders with GAT-SP results in specific unresponsiveness for DTH, Tprlf, interleukin-2, and plaque-forming cell (PFC) responses. This tolerance is mediated by both suppressor T cells (Ts) and a functional clonal inhibition. Lyt 1-2+ Ts suppress the induction (afferent limb) of GAT-specific DTH and PFC but not Tprlf responses. The reduced T-cell proliferation observed in GAT-tolerant mice is due to a non-transferable mechanism(s), possibly functional clonal inhibition. Our data are compatible with a multi-step pathway involving both proliferating and non-proliferating helper T (Th) cells. In addition, the fine specificity of tolerance induction for DTH and Tprlf responses was examined by using the related antigens poly(Glu60Ala40) (GA) and poly(Glu50Tyr50) (GT). Tolerance is exquisitely specific, as GA tolerizes responses to GA and GAT, whereas GT tolerizes GAT but not GA responses. Thus, both the route and form of antigen administration are important to the induction and regulation of immune response in Ir gene-controlled systems. Possible mechanisms governing the Th/Ts balance and the induction of GAT-specific tolerance and suppression for cellular and humoral responses in adult responders are discussed.

Mechanisms of genetic control of immune responses. I. Evidence for distinct multi-step helper T-cell pathways in cellular and humoral responses to GAT

We examined multiple genetically regulated humoral and cell-mediated immune (CMI) responses to poly( glu60ala30tyr10 ) (GAT) using a panel of mouse strains. We show that assignment of responder/nonresponder status depends upon the assay method. In addition, two distinct categories of nonresponder mice were found: (1) those which are unresponsive by all parameters tested (H-2q and H-2s haplotypes) and (2) those which are partially nonresponsive [H-2bm12 mutant strain--a low/nonresponder by splenic plaque-forming cell (PFC) and delayed-type hypersensitivity (DTH) responses, but exhibits B6 parental levels of high GAT-specific T-cell proliferation ( Tprlf ) and interleukin-2 production]. The distinction between these two nonresponder types was confirmed by complementation tests in which significant GAT-specific PFC and DTH responses were seen in (H-2q X H-2bm12)F1 hybrids, but not in (H-2q X H-2s)F1 hybrids. Suppressor T cells (Ts) also play a selective role in nonresponsiveness to GAT. Cyclophosphamide treatment of nonresponders (to eliminate Ts activity) as well as immunization with GAT coupled to the immunogenic carrier MBSA result in the development of GAT-specific humoral, but not CMI responses. Our results indicate that the T cell is the cellular site of Ir gene expression and that Tprlf responses do not correlate with functional helper T-cell activity and suggest distinct, multi-step Th/Ts regulatory pathways in the development of humoral and CMI effector functions.

Neonatal splenic suppressor cells in the chicken. I. In vivo suppression of the immune response to bovine serum albumin by normal and tolerized neonatal spleen cells

The presence of active splenic suppressor cells in neonatal chickens, either normal or tolerant to bovine serum albumin (BSA), was examined by assessment of their effect on both primary and adoptively transferred secondary responses to BSA or sheep red blood cells (SRC). Both normal and BSA tolerized spleen cells were shown to be highly suppressive of secondary anti-BSA responses generated by specifically primed adult spleen cells in inert recipients. Suppression of the secondary anti-BSA response by normal spleen cells was slightly less effective than that seen with BSA tolerant spleen cells. Transfer of BSA tolerant spleen cells into normal recipients, followed by BSA challenge, prevented any significant primary anti-BSA response. In contrast, transfer of normal spleen cells into normal recipients, followed by BSA challenge, failed to show any suppression of the resulting primary response. Neither normal nor BSA tolerant neonatal spleen cells were capable of suppressing either primary or secondary responses to SRC. Thus, chickens tolerized to BSA have suppressor cells specific for the tolerizing antigen. We present evidence that both the tolerance associated suppressors and the suppressors detected in normal neonatal chickens are T cells.

Antigen-specific suppressor T cell interactions. II. Characterization of two different types of suppressor T cell factors specific for L-glutamic acid50-L-tyrosine50 (GT) and L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT)

We have previously reported that two types of suppressor T cell factors (TsF) specific for L-glutamic acid50-L-tyrosine50 (GT) or L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT) can be distinguished based upon differences in their ability to suppress responses by allogeneic mice. Injection of GAT or GT induces a suppressor T cell subset that produces an antigen-binding, I-J+, genetically unrestricted, specific suppressor factor (TsF1). Injection of this factor plus small amounts of antigen induces a second-order suppressor T cell that produces an antigen-binding, I-J+, genetically restricted, specific suppressor factor (TsF2). In this report, we demonstrate that these two factors are also biochemically distinct. Monoclonal TsF1 molecules are composed of a single polypeptide chain that bears both the antigen-binding site and I-J determinant, whereas TsF2 molecules are composed of two disulfide-linked polypeptide chains, one of which is antigen-binding and I-J-, and the other, nonantigen-binding, I-J+. The antigen-binding chain must be added at culture initiation to achieve suppression, but the I-J+ chain can be added as late as day 3 with complete suppression observed. However, isolated chains from TsF2-producing hybridomas derived from three different haplotypes were unable to suppress immune responses when chains from heterologous TsF2 were mixed. Indirect evidence is presented that suggests that this restriction is because the chains fail to interact rather than the inability of the target cells to recognize both chains.

Role of contrasuppression in the adoptive transfer of immunity

The data presented in this paper show that the population of cells that adoptively transfer contact hypersensitivity are Lyt-1+ 2-, I-J- and nonadherent to V. villosa lectin. However, the adoptive transfer of immunity by this population of cells is successful only when the recipient has been treated in such a way as to impair the host immunosuppression mechanism. This population cannot, on its own, transfer immunity to adult, untreated naive recipients unless an additional population of immunoregulatory cells is present. This immunoregulatory population does not itself adoptively transfer immunity. This latter population is differentiated from the immune cells in that they are Lyt-1+ 2-, I-J+ and are adherent to V. villosa lectin. Both populations are required to adoptively transfer immunity to adult untreated naive recipients.

Hapten-reactive inducer T cells. I. Definition of two classes of hapten-specific inducer cells

Hapten-reactive inducer T cell clones can be divided into two groups based on their activation specificity. The first and largest group is conjugate specific. These clones are activated only by hapten coupled to the same carrier protein used for in vitro selection. The second group, which is quite rare, is hapten specific. Clones of this type are activated by hapten coupled to all foreign and autologus proteins tested. Both types of clones corecognize soluble antigen in association with products of the I-A locus. The hapten-specific cells were used to analyze the molecular basis of I-A vs. I-E gene control. The physiologic significance of hapten- and carrier-specific inducer T cells in the response to foreign antigens and autoantigens is discussed.

High frequency detection of different T-cell subsets in mice by a modified virus plaque assay

Different T-cell subsets participating in immune responses were detected at a high frequency by a modified virus plaque assay (VPA). By using the modified VPA, different activated T-cell subsets generated in primary immune responses, helper and suppressor T cells participating in antibody formation, and effector T cells involved in the delayed-type hypersensitivity (DTH) reaction were enumerated directly without in vitro antigen stimulation. The frequency of detection in immune systems used was 7.5-17.7 V-PFC/10(3) spleen cells. Although neither helper T cells for antibody formation nor effector T cells for DTH reaction were detected as V-PFC at a high frequency by the original VPA, it was also found in secondary immune response that Lyt 1 positive, antigen-specific helper and effector T-cell subsets, and cyclophosphamide (CY)-resistant precursors were enumerated at a high frequency by the modified VPA when the received in vitro antigen stimulation, and that the proliferative stage of these cells was critical for the development of V-PFC.

Strain differences in the immune response of mice to horseradish peroxidase

Hind footpads of mice of inbred strains were injected with horseradish peroxidase (PO) in Freund's complete adjuvant (FCA) and the response of the antibody-forming cells (AFC) in their popliteal lymph nodes was measured. Marked strain differences were found after the first injection and three types of responder strains were defined: high responders, with H-2^s, H-2^a and H-2^b haplotypes, low responders, with the H-2^d haplotype and an intermediate type of responder, observed in mice of the H-2^k and H-2^q haplotypes. After a second injection of PO in FCA, strain differences in the response disappeared and all strains responded equally well. F₁ hybrids from high and intermediate or low responders gave a mean AFC response between the two parental responses.

Immunoglobulin-forming cells (IFC) with no anti-PO function, appearing concomitantly with AFC in PO+FCA-immunized mice, were also counted. Compared with the IFC found in mice given FCA only, their number was always higher (two-five times more). The IFC response followed the same pattern as the AFC response and was high, intermediate and low in high, intermediate and low responders, respectively. The IFC:AFC ratio varied depending on how many days after the challenge it was measured, but was similar for all strains except the low responder strains, in which the ratio remained constant at approximately 1 throughout the immune response. Like the AFC response, the IFC counts in F₁ hybrids gave intermediate values, except in hybrids from the two low responders (BALB/c and DBA/2) in which there were five times more IFC than AFC.

We concluded that PO responsiveness in mice seems to be under genetic control governed by the H-2 locus and by non-H-2 genes and that the IFC which appear concomitantly with AFC are under the same genetic control.

H-2-determined kinetic differences for the induction of nucleoside-specific suppression

The kinetic and the H-2 requirements for the induction of nucleoside-specific suppression were examined in several strains of mice; specifically, whether adenosine (A)-coupled spleen cells given intravenously suppress the primary response to adenosine-KLH. The adenosine system was chosen because C57Bl/6 mice were originally found to be resistant to immune suppression when challenged 5 days after treatment with adenosine-coupled spleen cells. (Raps et al. J. Immunol. 126, 1542, 1981.) It was determined (i) whether A-specific nonresponsiveness is inducible in strains other than C57Bl/6; (ii) whether changes in hapten density on the A-conjugated spleen cells could alter C57Bl/6s ability to become nonresponsive, and (iii) whether there are interstrain differences in the time required to induce A-specific suppressor T cells (Ts). The results show that there are H-2-associated differences in the time required to induce A-specific immune suppression. While A-spleen cells failed to suppress the A-specific response in C67Bl/10 (H-2b), they did induce unresponsiveness in B10.D2 (H-2d on C57Bl/10 background). A 2.5-fold increase in epitope density of adenosine on cells did not influence the kinetics of suppression. C67Bl/6 were resistant to suppression on Day 5, but like the CB6F1, susceptible to unresponsiveness 10 days after treatment. Nonresponsiveness was T-cell-mediated and transferable across IgH-V barriers. Suppression induced by Balb/c donor mice is transferable to Igh-incompatible CAL-20 mice. These results are discussed in the context of genetic restrictions which regulate suppressor T-cell interactions.

Antigen-specific suppression in genetic responder mice to L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT). Characterization of conventional and hybridoma-derived factors produced by suppressor T cells from mice injected as neonates with syngeneic GAT macrophages

Spleen cells from C57BL/10 mice injected with syngeneic B10 L-glutamic acid60-L-alanine30-L-tyrosine10 (GAT)-pulsed macrophages (GAT-M phi) within 18 h of birth were unable to respond to soluble GAT, GAT-methylated bovine serum albumin, or B10 GAT-M phi as adults. Spleen cells from these neonatally treated mice responded at control levels to GAT presented in allogeneic M phi and to sheep erythrocytes. Partially purified T cells from these neonatally treated mice suppressed responses by syngeneic virgin, but not primed, spleen cells in an antigen-specific manner and acted during the early phases of the response. These responder GAT-specific suppressor T cells (GAT-TSR) were sensitive to anti-Thy-1 + C and 500-rad irradiation and have the phenotype Ly-1-2+, I-J+; GAT-TSR cells can only suppress responses by spleen cells syngeneic with the GAT-TSR cells at the I-J subregion of H-2. Restimulation of these Ts cells with syngeneic GAT-M phi induces an antigen-specific suppressor factor within the supernatant fluid. The factor, GAT-TsFR, is a glycoprotein with a molecular weight between 48,000 and 63,000, as determined by gel filtration chromatography using isotonic buffers; it bears serologically detectable determinants encoded by the I-J subregion of the H-2 complex, has an antigen-binding site for GAT and L-glutamic acid50-L-tyrosine50, and shares idiotypic determinants with anti-GAT antibodies. The presence of GAT-TsFR in the first 36 h of in vitro culture is required for significant suppression. Furthermore, only responses by spleen cell syngeneic with the cells producing GAT-TsFR at the I-J subregion are suppressed. The fusion of GAT-TsFR-producing cells with BW5147 resulted in generation of two hybridomas with properties and characteristics identical to those of the conventional GAT-TsFR with one exception: conventional and hybridoma 372.D6.5 GAT-TsFR only suppress responses by spleen cells of the I-Jb haplotype, whereas suppression mediated by the second hybridoma GAT-TsFR (372.B3.5) is genetically unrestricted. These hybridoma GAT-TsFR are compared with nonresponder GAT-Ts factor (GAT-TsF) and these responder and nonresponder GAT-TsF are considered in the context of suppressor pathways.