Cell types required for H-2-restricted cytotoxic responses generated by trinitrobenzene sulfonate-modified syngeneic cells or trinitrophenyl-conjugated proteins.

Murine spleen cells were fractionated over nylon wool or Sephadex G-10 columns, and the cell types involved in the generation of trinitrophenyl (TNP)-specific, H-2 restricted (TNP-self) cytotoxic effector cells were studied from cultures stimulated with trinitrobenzene sulfonate (TNBS)-modified syngeneic cells, TNP-conjugated soluble proteins such as bovine gamma-globulin (TNP-BGG), or bovine serum albumin (TNP-BSA). Unfractionated or nylon nonadherent responding cells generated such effectors, irrespective of whether the cultures were stimulated with TNBS-modified cells or TNP-conjugated proteins. TNP-modified T lymphocytes, B lymphocytes, and phagocyte-enriched spleen cells were all capable of stimulating TNP-self effectors. TNP-self effectors. TNP-self as well as allogeneic cytotoxic responses were dependent on the presence of a radioresistant non-T cell that was removed by Sephadex G-10 fractionation and was replaced by irradiated, Thy 1.2-negative, glass adherent spleen cells, enriched in phagocytic cells. Results obtained by using glass adherent cells that were allogeneic or semi-syngeneic to the responding cells indicated that H-2 homology was not required for efficient glass adherent cell function, and that the H-2 restriction of TNP-self effectors is not determined by these glass adherent cells.     

Fractionation of lymphocyte subpopulations which regulate mixed lymphocyte reactions.

Four days after injection of allogeneic lymphocytes BALB/c splenic T cells suppress proliferation of syngeneic cells in mixed lymphocyte reactions (MLR). Conversely, lymph node cells from the same mice amplify MLR responses. To further characterize these functional subpopulations, alloantigen-primed lymphocyte suspensions from both organs were fractionated by velocity sedimentation at unit-gravity. After fractionation MLR suppressor cells from spleens localized exclusively in rapidlly sedimenting fractions of large cells. MLR suppressor activity of cells from these fractions, as well as that of unfractionated spleen cell suspensions, was abolished by treatment with anti-Thy-1.2 serum and complement. Spleen cell fractions of similar sedimentation velocity also secreted a soluble MLR suppressor into culture supernatants. Although inhibitory of MLR, spleen cells of rapid sedimentation velocity did not suppress responses to T cell mitogens. In marked contrast with the effects of spleen cells, large 4-day-alloantigen-primed lymph node cells had no suppressive activity in MLR. MLR amplifier cells of uncertain derivation were found in fractions of medium sedimentation velocity from both spleens and lymph nodes. Fractionation of alloantigen-primed lymph node cell suspensions did reveal, however, a subpopulation of small cells with MLR suppressor acitivty which was unaffected by treatment with anti-Thy-1 serum and complement. The data thus indicate that large alloantigen-activated lymphocytes are not intrinsically suppressive nor are cells which suppress MLR necessarily large. We consequently conclude that regulation of MLR responses by alloantigen-primed lymphocytes involves a complex interaction between distinct functional subpopulations of cells which are separable both by physical and biologic properties.     

In vitro generation of suppressor cell activity: suppression of in vitro induction if cell-mediated cytotoxicity.

It was observed that when normal mouse spleen cells were cultured alone in vitro (precultured) for 3 to 7 days, these cells lost the ability to generate cell-mediated cytotoxicity (CML) during subsequent in vitro sensitization with allogeneic spleen cells, trinitrophenyl (TNP)-modified syngeneic spleen cells, or syngeneic tumor cells. These precultured cells, which were themselves unable to generate CML, were also shown in mixing experiments to suppress, actively, the generation of CML by freshly explanted spleen cells. Suppression occurred at the sensitization phase of CML, and not at the effector level; supernatants from suppressive precultured cells were not suppressive. Suppression was totally abrogated by the treatment of spleen cells with a T cell-specific rabbit anti-mouse brain serum and complement (RalphaMB+C) either before or after preculturing, suggesting that a T cell eas essential both to the generation of suppressor activity and to its expression. Suppressor activity was entirely absent in precultured nylon wool column-nonadherent spleen cells, a T cell-enriched population containing most of the RalphaMB+C-sensitive cells in the spleen. Precultured nylon column-adherent cells (T cell-depleted) did have suppressive activity, and a mixture of nylon-adherent and nylon-non-adherent cells was a suppressive after preculture as the precultured unseparated spleen. Moreover, the ability of nylon-adherent spleen cells to generate suppressive activity during preculturing was abrogated by treatment with RalphaMB+C. Thus, the "spontaneous" generation of CML-suppressive activity was dependent upon a limited subpopulation of splenic T cells isolated in the nylon column-adherent fraction. The relationship of these data to a previously described synergy between subpopulations of normal spleen in the generation of CML is discussed, and the findings related to other suppressor systems described in the literature.     

Suppressor cells for natural killer activity in carcinomatous pleural effusions of cancer patients

Summary Carcinomatous pleural effusions of 25 of 32 patients with lung cancer, which had markedly low or no natural killer (NK) activity against K562 cells in a 4 h chromium release assay, contained cells capable of suppressing the lytic function of blood NK cells from normal donors and cancer patients. Suppressor cells were found to be Sephadex G-10- and serum coated plastic dish-adherent monocyte/macrophages in 21 of 25 patients and nylon wool-nonadherent lymphocytes in the other four cases. Nonmalignant pleural effusions did not contain any type of suppressor cells. Twenty-four-hour preincubation of suppressor cells with effector cells was required for mediation of the suppressor function. Neither culture supernatants of effusion cells and NK cells nor effusion supernatants suppressed NK activity. The presence of indomethacin during the preincubation and cytotoxicity assay did not abrogate suppressor function. Suppressor cells did not reduce the number of lymphocyte/K562 conjugates. Contaminating tumor cells were not responsible for the suppression of cytotoxic activity. NK cells precultured with suppressor cells were not able to show cytotoxic function even after removal of the suppressor cells. When effusion mononuclear cells were passed through a Sephadex G-10 column and then preincubated for 24 h, these cells showed a significant increase in NK activity. The results suggest that carcinomatous pleural effusions contain at least two classes of suppressor cells for NK activity, monocyte/macrophages, and nylon wool-nonadherent lymphocytes, which could be one of the causes of impaired NK activity in carcinomatous pleural effusions.     

Alloantigen-specific suppressor T cells are not inhibited by cyclosporin A, but do require IL 2 for activation

Alloantigen-specific suppressor T cells are activated from normal murine spleen cells in mixed lymphocyte reactions (MLR). These T cells are radioresistant and suppress the activation of cytotoxic T lymphocytes (CTL) in second primary MLR cultures. This report demonstrates that cyclosporin A (CsA) blocks the activation of these suppressor cells at a dose of 1 microgram/ml. However, reconstitution of CsA blocked cultures with IL 2 restores the activation of the suppressor T cells, but fails to significantly restore the activation of CTL in these same cultures. This differential activation requirement was used to establish T cell lines that demonstrate enriched suppressor cell activity but depletion of CTL activity. These findings are discussed in terms of the mechanism of action of CsA in these distinct T cell subsets and the relevance to models of allograft unresponsiveness.     

Macrophages suppress CTL generation in rat mixed leukocyte cultures.

The generation of CTL in rat MLC was actively suppressed by a cell population present in spleen cell preparations from normal rats. These suppressor cells were characterized by using a variety of cell fractionation techniques. Suppressor cells were removed by passage of spleen cells through nylon wool columns or by treatment with carbonyl iron. Suppressive activity was present in the mononuclear cell fraction of spleen cells obtained by Ficoll-Hypaque density gradient centrifugation. After velocity sedimentation at unit gravity, enrichment of suppressive activity was demonstrated in the fractions containing large cells as compared to the fractions containing small cells. Populations rich in macrophages were shown to have similar suppressive activity upon CTL induction in MLC. These studies suggest that macrophages present in normal rat spleen cell preparations account for the difficulty in generating CTL in MLC prepared with rat cells.     

Regulation of experimental allergic encephalomyelitis. II. Appearance of suppressor cells during the remission phase of the disease

Lewis rats are susceptible to experimental autoimmune encephalomyelitis (EAE). Most rats recover from paralysis and are subsequently resistant to the disease. In an adoptive transfer system, we found that lymph node cells (LNC) from rats that had recovered from EAE protect syngeneic recipients from the disease when the latter are challenged with encephalitogenic myelin basic protein and adjuvant after receiving donor cells. Suppression is antigen-specific and requires viable LNC. In contrast to the suppressor cells we previously studied in tolerized rats, which were nonadherent T lymphocytes, the suppressor cells found in rats that have recovered from EAE adhere to glass wool. However, they are not retained on Sephadex G-10 columns to which macrophages adhere. Suppressor activity is enriched in the nylon wool-adherent LNC population (which consists of approximately 80% Ig+ cells). Our findings suggest that activation of adherent suppressor cells may be implicated in recovery from EAE. These may be adherent T cells, or B cells that produce anti-BP blocking antibodies.     

Further characterization of suppressor lymphocytes induced by fetal calf serum in murine lymphoid cell cultures: comparison with in vitro-generated cytotoxic lymphocytes.

Nonspecific suppressor cell (SPC) activity has been induced in vitro by preculturing splenocytes from normal mice in the presence of fetal calf serum (FCS) for 3 days or more. In adoptive transfer experiments in vivo, these precultured SPC were shown to reduce the humoral response of mice to SRBC and the cell-mediated cytotoxic (CMC) response to allogeneic tumor cells. In mixing experiments in vitro, using freshly explanted splenocytes, the precultured splenocytes abrogated the generation of specific cytotoxic lymphocytes (CL) in primary mixed lymphocyte-tumor cell cultures (MLTC). By contrast, secondary cytotoxic response was only marginally affected. Supernatants of precultured cells were also inhibitory, although to a lesser degree than whole cells. The induction of suppressive activity was abolished by addition of mitogenic amounts of concanavalin A to the preculturing medium.By the use of cell fractionation techniques it was found that both specific CL and nonspecific SPC lack an Fc receptor, do not adhere to nylon wool, and cannot be separated from each other by density sedimentation on a discontinuous BSA gradient. However, precursors of SPC and CL differed in their susceptibility to cyclophosphamide, hydrocortisone, and irradiation. The data presented does not exclude the possibility that suppressive activity exerted by FCS-induced SPC is mediated through a cytotoxic effect.     

Inhibition of in vitro antibody synthesis by cyclophosphamide-induced suppressor cells

A population of suppressor lymphocytes appears in the spleens of mice 5 to 14 days after treatment with a high dose of cyclophosphamide (100–200 mg/kg body wt). Removal of carbonyl iron adherent cells or Ig^? cells from cyclophosphamide (CP)-treated spleen cells does not abolish suppressive activity. These suppressors are, however, sensitive to removal by treatment with anti-Thy-1.2 and rabbit complement. CP-treated spleen cells can suppress the in vitro primary response of normal spleen cells to the soluble hapten-protein conjugate DNP-MON or the particulate antigen HRBC when added at time of culture initiation or up to the second day of culture. CP-treated spleen cells can themselves respond in vitro to DNP-MON, as well as to HRBC, but with altered kinetics from that of normal spleen cells. Collectively, the data suggest that the CP-induced suppressors act late in the in vitro antibody response, possibly by prematurely shutting off antibody synthesis by B cells.     

Human C3a-mediated suppression of the immune response. I. Suppression of murine in vitro antibody responses occurs through the generation of nonspecific Lyt-2+ suppressor T cell

C3a derived from the third component of human complement was found to suppress in vitro murine anti-SRBC responses. C3a-mediated suppression occurs through the generation of nonspecific Lyt-2+ suppressor T cells. The generation of suppressor cells occurs at an early phase in the response because incubation of naive T cells with C3a for as little as 30 min results in suppression of the anti-SRBC response. The generation of suppressor T cells requires the interaction of T cells, C3a, and a Sephadex G-10-adherent cell, presumably a macrophage. Although the mechanism of action of these suppressor cells has not been elucidated, several possibilities have been eliminated. C3a-suppressor T cells do not apparently release inhibitory lymphokines, nor is helper cell activity inhibited by a 2-day co-culture with these suppressor cells. The observation that interleukin 2 (IL 2)-containing lymphokine preparations could overcome C3a-induced suppression led us to investigate the interaction of the suppressors with IL 2 producer cells. However, neither C3a nor C3a-generated suppressor T cells can block the synthesis of IL 2.     

Clinical studies on cell-mediated immunity in patients with malignant disease

Summary In vitro mitogenic responses of lymphocytes of patients with advanced cancer of the stomach or lung were determined and the cells involved in the depressed responses of patients characterized. Proliferative responses to phytohemagglutinin and concanavalin A of lymphocyte rich mononuclear cells of the patients were impaired, but increased after in vitro unstimulated culture for seven days. Mitogenic responses were enhanced by depletion of monocytes using a Sephadex G-10 column, and further enhanced by removal of nylon wool nonadherent cells. Nylon wool nonadherent cells of the patients suppressed the mitogenic responses of autologous and allogeneic lymphocytes, but lost the suppressive activity during seven days' in vitro culture. Nonadherent cells of normal donors did not inhibit mitogenic responses. The results suggest that peripheral blood mononuclear cells of cancer patients may contain at least two populations of suppressor cells for mitogenic responses, monocytes and nylon wool nonadherent cells, which could be one of the causes of impaired mitogenic responses in these cancer patients.     

Suppression of antitumor immunity by macrophages in spleens of mice bearing a large MOPC-315 tumor

We had shown previously that progression of MOPC-315 plasmacytoma growth is associated with an increase in the percentage of macrophages in the spleen as well as a decrease in the ability of tumor-bearer spleen cells to mount an antitumor cytotoxic response upon in vitro immunization. Here we provide evidence that macrophages in the MOPC-315 tumor-bearer spleen are responsible at least in part for the suppression of the generation of antitumor cytotoxicity. Accordingly, removal of most macrophages by depletion of phagocytic cells or Sephadex G-10-adherent cells from spleens of mice bearing a large tumor resulted in augmented antitumor immune potential. Also, Sephadex G-10-adherent spleen cells from tumor-bearing (but not normal) mice drastically suppressed the in vitro generation of antitumor cytotoxicity by normal spleen cells. The suppressive activity of these adherent cells did not reside in contaminating suppressor T cells, since it was not reduced by treatment with monoclonal anti-Thy 1.2 antibody plus complement. The Sephadex G-10-adherent cell population from the tumor-bearer spleen suppressed the in vitro generation of antitumor cytotoxicity against autochthonous tumor cells but not against allogeneic EL4 tumor cells, and hence the suppression was apparently specific. The suppressive activity of the Sephadex G-10-adherent cell population from tumor-bearer spleens was overcome by treatment of the tumor-bearing mice with a low curative dose of cyclophosphamide. This immunomodulatory effect of a low dose of the drug in overcoming the suppression mediated by the Sephadex G-10-adherent cell population enables the effector arm of the immune system of tumor-bearing mice to cooperate effectively with the drug's tumoricidal activity in tumor eradication.     

Con A-inducible suppression of MLC: evidence for mediation by the TH2 + T cell subset in man.

Human peripheral lymphoid cells pretreated with Concanavalin A for 48 hr can markedly suppress the proliferative response of untreated autologous lymphoid cells in MLC. Isolation studies with Sephadex G-200 anti-F(ab')2 affinity chromatography, nylon adherence, and E rosetting indicate that the Con A-induced suppressor cell is a T cell. Further fractionation into TH2+ and TH2- cell subsets with an equine-anti TH2 serum show that both subsets can be activated by Con A to an equivalent degree. After activation only the TH2+ subset can suppress autologous responder cells in MLC. The TH2- subset, which comprises 80% of peripheral human T cells, although induced by Con A to proliferate, cannot itself suppress the MLC response. Nevertheless, the TH2- subset can be shown to modulate the generation of suppressor TH2+ cells at 24 hr but not at 48 hr. These studies support the notion that the Con A-induced suppressor cell is confined to a distinct T cell subset in man and that T-T interactions are important in the overall expression of the immune response.     

Interaction between T cells and non-T cells in suppression of cytotoxic lymphocyte responses.

Generation of cytotoxic T lymphocytes (CTL) in mixed leukocyte cultures was suppressed by a factor elaborated by alloantigen-activated T cells. This suppressor factor, CTL-TsF, in contrast to a factor that suppresses proliferative responses in mixed leukocyte reactions (MLR-TsF), was effective only when added during the first 24 hr of a 6-day-culture period. Moreover, removal of CTL-TsF 24 hr after culture initiation failed to restore CTL responses. CTL activity could be rescued from suppressed cultures, however, by addition of 2-mercaptoethanol on days 3 or 4. Similarly, transfer of nonadherent cells at 3 or 4 days from cultures treated with CTL-TsF to cultures of adherent cells initiated in control factor restored CTL responses. Mixing experiments with cells pulsed with CTL-TsF for 4 hr at culture initiation identified a target of CTL-TsF as a Thy-1 negative cell that was adherent to plastic and to Sephadex G-10. Suppression was not due to interference with physiologic accessory cell function, but more likely was accomplished via a negative signal from CTL-TsF-pulsed cells. The results thus suggest that CTL-TsF acts early, but reversibly, in the CTL differentiative process via a second suppressor effector cell, possibly a macrophage.     

Suppressor T lymphocyte induction by a factor released from cultured blastocysts

For the analysis of immunologic escape mechanisms of embryos during the implantation period in mice, the effects of culture supernatant of blastocysts on in vitro responsiveness to alloantigen of mice was investigated. Blastocyst-cultured conditioned medium was prepared by culturing late blastocysts of outbred ICR mice for 5 days. The addition of culture supernatant containing four or eight blastocysts to allogeneic mixed lymphocyte culture inhibited both the MLR responses and the generation of cytotoxic T lymphocytes (CTL). Preincubation of the culture supernatant with lymphocytes syngeneic to the responder cells of MLR induced potent suppressor cell activity in the MLR. The supernatant did not inhibit the activity of CTL at the effector phase, but preinduced suppressor cells obtained by incubation of splenocytes with the supernatant showed almost complete suppression of CTL activity at the effector phase. Both of the suppressor cells, active on MLR and at the generation phase of CTL as well as active at the effector phase, had a surface phenotype of Thy-1+ and Ig-. The suppressive material could be extracted from the eight-cell stage of fertilized ova or blastocysts but not from unfertilized ova, indicating that the production of the factor(s) is dependent on the stages of early embryogenesis. These results suggest that the active induction of suppressor T lymphocytes by the factor(s) released from implanted embryos is one of the protective mechanisms from maternal immunologic attack.     

Suppression of antitumor immunity by macrophages in spleens of mice bearing a large MOPC-315 tumor

Summary We had shown previously that progression of MOPC-315 plasmacytoma growth is associated with an increase in the percentage of macrophages in the spleen as well as a decrease in the ability of tumor-bearer spleen cells to mount an antitumor cytotoxic response upon in vitro immunization. Here we provide evidence that macrophages in the MOPC-315 tumor-bearer spleen are responsible at least in part for the suppression of the generation of antitumor cytotoxicity. Accordingly, removal of most macrophages by depletion of phagocytic cells or Sephadex G-10-adherent cells from spleens of mice bearing a large tumor resulted in augmented antitumor immune potential. Also, Sephadex G-10-adherent spleen cells from tumor-bearing (but not normal) mice drastically suppressed the in vitro generation of antitumor cytotoxicity by normal spleen cells. The suppressive activity of these adherent cells did not reside in contaminating suppressor T cells, since it was not reduced by treatment with monoclonal anti-Thy 1.2 antibody plus complement. The Sephadex G-10-adherent cell population from the tumor-bearer spleen suppressed the in vitro generation of antitumor cytotoxicity against autochthonous tumor cells but not against allogeneic EL4 tumor cells, and hence the suppression was apparently specific. The suppressive activity of the Sephadex G-10-adherent cell population from tumor-bearer spleens was overcome by treatment of the tumor-bearing mice with a low curative dose of cyclophosphamide. This immunomodulatory effect of a low dose of the drug in overcoming the suppression mediated by the Sephadex G-10-adherent cell population enables the effector arm of the immune system of tumor-bearing mice to cooperate effectively with the drug's tumoricidal activity in tumor eradication.This paper was presented in part at the annual meeting of the American Association of Immunologists, Chicago, Illinois, 10–15 April 1983     

The differentiation of suppressor cell populations as revealed by studies of the effects of mitogens on the mixed lymphocyte reaction and on the generation of cytotoxic lymphocytes.

The regulation by concanavalin A (Con A) and bacterial lipoloysaccharide (LPS) of the mixed lymphocyte reaction (MLR) and of the generation of cytotoxic lymphocytes (CL) was studied in congenic resistant mice using cortisone resistant thymocytes as the responding cells. LPS enhances the generation of CL selectively when suboptimal numbers of allogeneic cells are present in mixed lymphocyte cultures and also results in the augmentation of the MLR. Mitogenic concentrations of Con A on the other hand suppress the generation of CL regardless of alloantigen dose. The mechanism of suppression cannot be ascribed to the presence of suppressor T cells, since the addition to the cultures of syngeneic cortisone resistant thymocytes activated by Con A does not change the immune response. However, prospective suppressor cells that can be activated by Con A are located in secondary lymphoid organs such as spleen and lymph node. Suppressor activity by those cells is abolished by anti θ plus complement. Con A activated spleen cells suppress the MLR, whereas Con A activated thymocytes amplify the proliferation of responding cells.     

Subpopulation of T Cells Sensitive to Natural Thymocytotoxic Autoantibody (NTA) of New Zealand Mice: I. Distinct Cytotoxic Sensitivity of Functional T Cell Subsets to NTA and Anti-Thy-1 Antibodies

NZB mice produce a natural thymocytotoxic autoantibody (NTA) capable of specifically injuring thymocytes and T cells. NTA-reactive antigen (NTA-A) shows a different density distribution among T cells, and partial killing with NTA and complement can eliminate T cells bearing NTA-A in high density. Thy-1 antigen is similar to NTA-A in this respect. To determine the effects of NTA and anti-Thy-1 on distinct functional subsets of T cells, Con A-induced suppressor T cell (Con A-Ts) activity against the allogeneic mixed lymphocyte reaction (MLR), responding T cell (T_MLR) activity in the allogeneic MLR, and Con A-induced cytotoxic T cell (Con A-Tc) activity were examined simultaneously in BALB/c spleen cells before and after partial elimination of NTA- and anti-Thy-1-sensitive T cells. Treatment with NTA and complement resulted in a marked reduction in Con A-Ts activity, a significant increase in T_MLR activity and a slight and inconstant decrease in Con A-Tc activity. Since Con A-generated Ts were much less sensitive to NTA, the NTA-sensitive T cells involved in Con A-Ts activity appear to be precursors or promoters of the Con A-Ts. In contrast, the precursors of Con A-Tc seem to be relatively resistant to NTA. The increase in T_MLR activity caused by NTA suggests the possibility that NTA is less cytotoxic for T_MLR and cytotoxic for some suppressor T cells in allogeneic MLR. The monoclonal anti-Thy-1 antibody showed no such preferential cytotoxic effects on the three T cell functions. The NTA-sensitive T cells, in contrast to anti-Thy-1-sensitive T cells, were reduced gradually during Con A stimulation. All these findings indicate that NTA-A not only differs from Thy-1 antigen but that it appears to be a unique T cell antigen.     

Suppressor factor produced by neonatal mouse spleen cells

Spleen cells from 5- to 6-day- old BDF₁ mice produce a soluble suppressor factor (SF) when cultured for 2–4 days in the presence of 10% fetal calf serum. This suppressor factor inhibits the mixed lymphocyte reactivity of adult mouse spleen cells as well as the in vitro generation of cytotoxic cells. The SF which is not H₂-restricted or antigen specific is most effective when added in the early phase of the culture period. The SF is resistant to heat and uv treatment and appears to consist of a large and small component. It is resistant to treatment with pronase or trypsin. The SF appears to be produced by neonatal spleen cells which are not adherent to plastic or Sephadex G-10 and are insensitive to treatment with anti-Thy 1.2 and complement. Incubation of SF with peritoneal exudate cells reduces suppressor activity.     

Alloantigen-specific T suppressor-inducer and T suppressor-effector cells can be activated despite blocking the IL-2 receptor

To determine IL-2 requirement for activation of suppressor cells, PBMC were primed in one-way MLR in the presence of 10 micrograms/ml anti-IL-2R beta-chain antibody 2A3 (CD25) or control antibody, then irradiated and added as regulators in a fresh MLR. Cells primed in the presence of antibody 2A3 suppressed the proliferative response to fresh autologous lymphocytes to specific alloantigen but had no effect on the response to cells from third party donors. Priming in the presence of an antibody of irrelevant specificity induced only limited suppressor activity. Activated suppressor cells did not show cytolytic activity specific for the stimulators when tested at the time of the suppressor cell assay. To identify the subset(s) responsible for suppression, cells primed in the presence of antibody 2A3 were separated into CD4+/CD45RA+, CD4+/CD45RA-, and CD8+ subsets, which were irradiated and then tested. The suppressive activity was found predominantly in the CD4+/CD45RA+ subset, whereas CD8+ cells had some activity and CD4+/CD45RA- cells had none. No subset suppressed the response of autologous cells to third-party cells. When primed CD4+/CD45RA+ cells were cocultured with fresh autologous lymphocytes depleted of CD8+ cells, no suppression was observed, indicating that, although the CD4+/CD45RA+ cells can function as inducers of suppressors, they cannot function as suppressor-effectors. Conversely, CD8+ cells activated in MLR in the presence of 2A3 caused suppression, regardless of whether the fresh autologous responder population contained CD8+ cells. CD4+/CD45RA+ and CD8+ subsets isolated after priming in the presence of 2A3 also demonstrated Ag-specific suppression in the generation of cytotoxic T lymphocytes whereas CD4+/CD45RA- cells had no activity. Our data are consistent with the model that suppression of alloreactivity requires the cooperation of two types of cells, a CD4+/CD45RA+ suppressor-inducer and a CD8+ suppressor-effector population. Activated Tsi and fresh Tse or activated Tse alone can suppress lymphocyte proliferation and generation of CTL in response to specific Ag. Activation of Ag-specific T suppressor-inducer and T suppressor-effector cells appears to be relatively IL-2 independent and presumably require one or more other growth factors.