Clonal analysis of cytolytic T lymphocyte specificity. I. Phenotypically distinct sets of clones as the cellular basis of cross-reactivity to alloantigens.

The cellular basis of the cytolytic cross-reactivity observed in primary allogeneic (C56BL/6 anti-DBA/2 and C57BL/6 anti-C3H/He) mixed-leukocyte cultures (MLC) was investigated by analysis of the specificity of clonal progeny derived from individual cytolytic T lymphocyte (CTL) precursor cells (CTL-P) contained within these populations. A sensitive mixed-leukocyte microculture (micro-MLC) technique was used with limiting dilution analysis by Poisson statistics to determine the frequency of CTL-P reactive against both specific and third-party (P815 and AKRA) target cells, to calculate the probability that each micro-MLC was a clone derived from a single CTL-P, and to examine the specificity of each micro-MLC assayed separately against both target cells. A total of 287 phenotypically specific, heteroclitic, and cross-reactive micro-MLC from the 2 different strain combinations were observed with a relative frequency of 81, 11, and 8%, respectively, and were calculated to have mean clone probability of 90 and 99% when based, respectively, upon the frequencies of CTL-P reactive against the specific and third-party target cells. These clones were estimated to have an approximate size of 6 X 10(4) cells, which corresponded to roughly 16 cell doublings during the 7 d of culture. 22 clones were successfully subcloned and in virtually every case, the subclones retained the specificity phenotype of the original clone from which they were derived. These results provide direct evidence for three phenotypically distinct sets of CTL as the cellular basis of cross-reactivity in MLC populations assayed against two different target cells.

Separation of stages in the development of the "T" cells involved in cell-mediated immunity.

Density distribution analysis in continuous gradients of albumin has been used to study the development of cytotoxic lymphocytes (CL), to separate and characterize the progenitors of CL, and to determine their relationship to subpopulations of T cells. CL progenitors in the thymus were a homogeneous, medium-density population, distinct from the typical, dense, thymus small-lymphocyte. Activity seemed to be associated with one minor subpopulation of cells with surface antigenic properties characteristic of peripheral T cells (high levels of H-2 antigen, low levels of theta-antigen). CL progenitors in the spleen differed from those in the thymus and normally had the high buoyant density of typical small T lymphocytes. In states of antigenic stimulation, some lighter-density CL progenitors were found in the spleen. The buoyant density of the CL population developing in the spleens of immunized animals showed progressive changes with time. Early, "immature" CL had the light-density characteristics of large, activated lymphocytes. As the response developed, the density of the CL population increased, and finally approached that of CL progenitors and of typical small lymphocytes. The data suggest that density subpopulations of T cells represent stages in the development of immunocompetent cells. Possible differentiation pathways of T lymphocytes in the thymus and in the spleen are discussed.

The immunogenicity of antigen bound to the plasma membrane of macrophages.

Macrophages were cultured for several hours after a brief exposure to radio-iodinated keyhole limpet hemocyanin. Most of the hemocyanin taken up by the macrophages was rapidly catabolized and eliminated from the cell. A few molecules were retained on the plasma membrane of the cells for prolonged periods and were not subject to endocytosis and catabolism. These few molecules of hemocyanin bound to the plasma membrane were identified by observing the fixation of antibody fragments to macrophages at low temperature. The membrane-bound antigen, which could be removed by trypsin or EDTA, was of large molecular size, though heterogeneous. A great part of the immune responses of mice to hemocyanin bound to live macrophages could be abrogated by treatment of the macrophages in vitro with antibody or trypsin. Hence, most of the immunogenicity of hemocyanin bound to macrophages was attributed to the few molecules of antigen bound to the plasma membrane.

IL-2-mediated T cell proliferation in humans is blocked by a monoclonal antibody directed against monomorphic determinants of HLA-DR antigens.

We have studied the effect on the interleukin (IL-) 2-dependent human T cell growth of two distinct monoclonal antibodies (Mab), D1-12 and 4F2, with specificity for common determinant of human Ia antigens and for a differentiation antigen expressed on all activated T cells, respectively. Strong inhibition of cell growth was found in cultures supplemented with the anti-Ia D1-12 Mab but not in cultures supplemented with 4F2 Mab. These results were obtained when either total mixed leukocyte culture (MLC) T cells or an MLC-derived T cell clone were used as indicator cell systems for IL-2 activity. The inhibition of cell growth appears to be mediated by a direct interaction of D1-12 Mab with the cells and not by a direct inactivation of the growth factor, as addition of the antibody to murine MLC T cells, which do not express the determinant defined by D1-12 Mab, resulted in no inhibition of their proliferation induced by the same source of human IL-2.

Antigenic specificity of the cytolytic T lymphocyte response to murine sarcoma virus-induced tumors. III. Characterization of cytolytic T lymphocyte clones specific for Moloney leukemia virus-associated cell surface antigens.

Cytolytic T lymphocyte (CTL) clones specific for Moloney leukemia virus (MoLV)-derived tumor cells were generated by placing limiting numbers of C57BL/6 responder cells into mixed leukocyte-tumor cell microcultures. Under appropriate conditions (presence of stimulating tumor cells, accessory cells, and T cell growth factor), such cloned CTL could readily be expanded to provide large numbers of homogeneous, highly cytolytic CTL populations for further characterization. Using four target-cell types, three specificity patterns were observed: one reactive with the syngeneic MoLV-derived tumor only, one cross-reactive with an allogeneic MoLV-derived tumor, and one cross-reactive with normal allogeneic cells. Subclones derived from these three types of clones exhibited a high degree of stability in terms of lytic activity and specificity over a 4-mo period of observation. Three clones analyzed by flow cytofluorometry using monoclonal antibodies were all found to be of the Lyt-1+2+ phenotype. Furthermore, lysis of target cells by all of six clones tested was inhibited by anti-H-2Db (but not by anti-H-2Kb) monoclonal antibodies, demonstrating H-2Db-restriction at the clonal level.

Synthetic peptides as antigens and competitors in recognition by H-2-restricted cytolytic T cells specific for HLA.

The specificity of peptide recognition by a number of Kd-restricted CTL clones specific for HLA-CW3 or HLA-A24 was investigated. The CTL clones were derived from DBA/2 (H-2d) mice immunized with syngeneic P815 mouse cells transfected with genes encoding HLA-CW3 or HLA-A24 class I molecules. We had previously shown that CTL clones that lysed P815-CW3 transfectant target cells could lyse P815 (HLA-) target cells incubated with synthetic CW3 peptides corresponding to the COOH-terminal end of the alpha 2 domain. In the present study, we found that Kd-restricted CTL clones that lysed P815-A24 transfectant target cells recognized a synthetic peptide from the same region (residues 170-182) of the A24 molecule. CW3 and A24 differ by only one amino acid within this region. Recognition of CW3 or A24 peptides corresponded exactly with lysis of P815-HLA transfectants both for clones that mutually exclusively lysed CW3 or A24 transfectant target cells and for CW3/A24 crossreactive CTL clones. The latter CTL clones that lysed both CW3 and A24 transfectant target cells showed a clear preference for the peptide corresponding to the immunizing HLA allele. The homologous CW3 and A24 peptides could compete with each other for recognition, in contrast to a peptide from the same region of HLA-B7. Peptides from the corresponding region of the endogenous Kd and Dd/Ld molecules could also inhibit recognition of CW3 and A24 peptides. Competition with peptides apparently occurred at the level of the target cell. These results are consistent with a model whereby MHC class I molecules position protein fragments or peptides for specific recognition by T cells.

Generation of cytotoxic T lymphocytes in vitro. I. Response of normal and immune mouse spleen cells in mixed leukocyte cultures.

Mouse cytotoxic T lymphocytes (CTL) were generated in mixed leukocyte cultures (MLC) using spleen cells as responding cells and irradiated allogeneic spleen cells as stimulating cells. Cytotoxicity was assessed by a quantitative (51)Cr assay system and the relative frequency of CTL in individual cell populations was estimated from dose-response curves. Inclusion of 2-mercaptoethanol in the MLC medium resulted in a 20-40-fold increase in the relative number of CTL generated at the peak of the response. Under these culture conditions, cell-mediated cytotoxic activity was detectable in MLC populations as early as 48 h after the onset of the cultures. When spleen cells from mice immunized with allogeneic tumor cells 2-4 mo previously were cultured with irradiated spleen cells of the same alloantigenic specificity (MLC-Imm), it was found that the cell-mediated cytotoxic response was detectable earlier and reached higher levels than that observed in a primary MLC. At the peak of the response, MLC-Imm populations were observed to lyse up to 50% of the target cells within 3 h at a lymphocyte: target cell ratio of 0.3:1. Immunological and physical characterization of the effector cells generated in MLC-Imm indicated that they were medium to large-sized T lymphocytes. Altogether, these studies suggested the existence of an anamnestic cell-mediated cytotoxic response in MLC-Imm.

Generation of cytotoxic T lymphocytes in vitro. III. Velocity sedimentation studies of the differentiation and fate of effector cells in long-term mixed leukocyte cultures.

Separation of cells by velocity sedimentation at unit gravity was utilized to investigate the physical properties of cytotoxic thymus-derived lymphocytes (CTL) generated in long-term mixed leukocyte cultures (MLC). In kinetic studies, CTL were found almost exclusively in the large cell fractions at the peak of the response on day 4, whereas the majority of CTL in day 14 MLC had the sedimentation properties of small lymphocytes. Reculture until day 14 of cells fractionated on the basis of size on day 4 indicated that the small CTL were derived exclusively from cells which had been large on day 4. Re-exposure of day 14 MLC cells to the original stimulating alloantigens resulted in significant cell proliferation and rapid regeneration of CTL activity. Cell fractionation experiments demonstrated that the cells in the day 14 MLC population which responded to the secondary allogeneic stimulus were small T lymphocytes, and that these cells rapidly developed into large, highly cytotoxic CTL following stimulation. Moreover, by restimulating on day 14 fractions which were selected on the basis of size on day 4, it was found that the responding small lymphocytes were themselves the progeny of cells which were large at the peak of the response. Since CTL and CTL progenitors showed concomitant changes in physical properties with time, the possibility exists that they belong to the same cell lineage, and hence that CTL can differentiate into cells which are no longer cytotoxic, but capable of mounting an anamnestic response.

Effects of epitope modification on T cell receptor-ligand binding and antigen recognition by seven H-2Kd-restricted cytotoxic T lymphocyte clones specific for a photoreactive peptide derivative.

We tested for antigen recognition and T cell receptor (TCR)-ligand binding 12 peptide derivative variants on seven H-2Kd-restricted cytotoxic T lymphocytes (CTL) clones specific for a bifunctional photoreactive derivative of the Plasmodium berghei circumsporozoite peptide 252-260 (SYIPSAEKI). The derivative contained iodo-4-azidosalicylic acid in place of PbCS S-252 and 4-azidobenzoic acid on PbCS K-259. Selective photoactivation of the N-terminal photoreactive group allowed crosslinking to Kd molecules and photoactivation of the orthogonal group to TCR. TCR photoaffinity labeling with covalent Kd-peptide derivative complexes allowed direct assessment of TCR-ligand binding on living CTL. In most cases (over 80%) cytotoxicity (chromium release) and TCR-ligand binding differed by less than fivefold. The exceptions included (a) partial TCR agonists (8 cases), for which antigen recognition was five-tenfold less efficient than TCR-ligand binding, (b) TCR antagonists (2 cases), which were not recognized and capable of inhibiting recognition of the wild-type conjugate, (c) heteroclitic agonists (2 cases), for which antigen recognition was more efficient than TCR-ligand binding, and (d) one partial TCR agonist, which activated only Fas (C1)95), but not perforin/granzyme-mediated cytotoxicity. There was no correlation between these divergences and the avidity of TCR-ligand binding, indicating that other factors than binding avidity determine the nature of the CTL response. An unexpected and novel finding was that CD8-dependent clones clearly incline more to TCR antagonism than CD8-independent ones. As there was no correlation between CD8 dependence and the avidity of TCR-ligand binding, the possibility is suggested that CD8 plays a critical role in aberrant CTL function.

Role of CD8 in aberrant function of cytotoxic T lymphocytes.

Using H-2Kd-restricted photoprobe-specific cytotoxic T lymphocyte (CTL) clones, which permit assessment of T cell receptor (TCR)-ligand interactions by TCR photoaffinity labeling, we observed that the efficiency of antigen recognition by CTL was critically dependent on the half-life of TCR-ligand complexes. We show here that antigen recognition by CTL is essentially determined by the frequency of serial TCR engagement, except for very rapid dissociations, which resulted in aberrant TCR signaling and antagonism. Thus agonists that were efficiently recognized exhibited rapid TCR-ligand complex dissociation, and hence a high frequency of serial TCR engagement, whereas the opposite was true for weak agonists. Surprisingly, these differences were largely accounted for by the coreceptor CD8. While it was known that CD8 substantially decreases TCR-ligand complex dissociation, we observed in this study that this effect varied considerably among ligand variants, indicating that epitope modifications can alter the CD8 contribution to TCR-ligand binding, and hence the efficiency of antigen recognition by CTL.

An approach to the quantitation of immunogenic antigen.

Using passively administered isotope-labeled anti-KLH to suppress the antibody response of rabbits to KLH, we have attempted to estimate the amount of antigen actually involved in stimulating antibody formation. Single and paired label tracer studies of passively administered anti-KLH IgG indicated that from 0.7 to 2.9 microg were utilized or involved by the antigen in the course of a 90% suppression of the response to 2 mg KLH. Tracer studies of labeled anti-KLH F(ab')(2) fragments revealed the retention of from 2 to 3 microg of these fragments in the entire rabbit during a 60% suppression of the response to 1 mg KLH. Based on previously determined ratios of mixtures of KLH and suppressive amounts of anti-KLH in adjuvant, the antibody utilization data were converted to the probable amount of antigen or immunogen involved. It appears that after an injection of 2 mg of KLH approximately 0.2-0.5 microg of antigen persisted and reacted with antibody given 24 hr later. Since all of this persisting, reactive antigen may not be immunogenic, the above estimate of immunogen is probably high, but may serve to establish upper limits for the amounts of immunogen involved in stimulating antibody formation and provide a meaningful frame of reference for antigen tracer studies.

Thymus-derived (T) cell immunoglobulins. Presence of a receptor site for IgG and absence of large amounts of "buried" Ig determinants on T cells.

Quantitation of surface and total cell Ig obtained after lysis by detergent, urea-acid treatment, and freeze-thawing were determined on spleen cells, thymus cells, and spleen cells specifically depleted of B cells. A two- to four-fold increase in measurable Ig was found after cell lysis. All cell populations showed a similar increase in measurable Ig indicating that no discordantly large amounts of buried Ig determinants were associated with the surface of T cells. The lack of appreciable amounts of T cell Ig was confirmed by immunoprecipitation of radioiodinated cells. A theta-positive lymphoma was described which, when grown in culture, lacked detectable surface Ig but contained a receptor site for IgG. This resulted in appreciable amounts of surface IgG being associated with the tumor line when isolated from ascitic fluid of tumor-bearing mice or after preincubation of cultured cells with either heat-aggregated IgG or normal mouse serum.

Generation of cytotoxic T lymphocytes in vitro. IV. Functional activation of memory cells in the absence of DNA synthesis.

Re-exposure of day 14 mixed leukocyte culture (MLC) cells to the original stimulating alloantigens (secondary response) has previously been shown to result in significant proliferation and in rapid reappearance of high levels of cytolytic T-lymphocyte (CTL) activity within the next 4 days. Moreover, evidence has been presented that CTL precursor cells in day 14 MLC populations, while they derived from cells were large at peak of the primary response (day 4) were themselves small lymphocytes which developed into large CTL after restimulation. In this study, inhibition of DNA synthesis by cytosine arabinoside (ARA-C) was used to investigate whether CTL formation could be dissociated from proliferation during the secondary response. It was found that within the first 24 h after restimulation (a) CTL activity increased 6-to-20-fold, (b) 60-70% of the small T lymphocytes became medium- to large-sized cells, and (c) both events were independent of DNA synthesis. By using two successive cell separations by velocity sedimentation at unit gravity, before and after stimulation of day 14 MLC cells for 24 h in the presence or absence of ARA-C, direct evidence was obtained that small CTL precursor cells developed into large CTL, irrespective of DNA synthesis. The presence of ARA-C for periods longer than 24 h inhibited any further increase in CTL activity, in contrast to a parallel increase in lytic activity and cell number from day 1 to day 4 in control restimulated cultures. Taken together with the finding that 90% of the medium- and large-sized lymphoid cells in control restimulated cultures underwent DNA synthesis within 24 h, these results thus suggest that during a secondary MLC response there is initially a differentiation step leading to the formation of CTL which, although it can be clearly dissociated from DNA synthesis, is under normal conditions followed by proliferation of these effector cells.

Comparison of the immune responsiveness of NZB and NZB X NZW F1 hybrid mice with that of other strains of mice.

The immune responsiveness of (NZB x NZW) F(1) hybrid mice (NZB/W) has been compared with that of three other strains of mice, A/J, BALB/c, and CBA/J. The antigens used included sheep red blood cells (SRBC), keyhole limpet hemocyanin (KLH), bovine serum albumin (BSA), and human gamma-globulin (HGG). It was found that important strain differences existed in the amount of antibody produced, but the relative immune responsiveness depended very much upon the nature of antigen. By comparison with the other strains tested, NZB/W mice had a higher antibody production to some antigens (SRBC and BSA) but were low responders to others (KLH). Induction of unresponsiveness to HGG by treatment with ultracentrifuged HGG was studied in the strains cited above. NZB/W mice became tolerant after injection of HGG ultracentrifuged at 100,000 g for 2 hr. Similar experiments carried out with another preparation of HGG (centrifuged at 20,000 g for 30 min) failed to reveal any abnormal behavior of NZB/W mice as compared to BALB/c or A/J mice. These results do not support the concept that NZB/W mice possess a general immune hyperreactivity or a relative inability to be made tolerant to protein antigens. However, they do not rule out the possibility that these mice have a genetically determined hyperresponsiveness to some antigens, in particular to nuclear antigens.

Cellular and humoral response to transplantation antigens. I. Development of alloantibody-forming cells and cytotoxic lymphocytes in the graft-versus-host reaction.

After transfer into heavily-irradiated allogeneic mice, spleen cells were found to produce two types of effector cells directed against the recipient alloantigens, namely alloantibody plaque-forming cells (PFC) and cytotoxic lymphocytes (CL). Both types of effector cells were detectable in vitro by virtue of their lytic effect on target cells carrying the recipient alloantigens. Alloantibody PFC activity was dependent on the presence of an exogenous source of complement and could be inhibited by the addition of heterologous antisera to mouse micro-chain or Fab fragment in the assay system. CL activity was independent of added complement, was not affected by anti-immunoglobulin antisera, but was inhibited by the addition of antibody against target cell alloantigens. Treatment of the transferred spleen cells with anti-theta-serum and complement before in vitro assays for PFC and CL completely abolished the CL activity but had no effect on alloantibody-plaque formation. These results indicate that the two types of effector cells can be differentiated in vitro by virtue of their susceptibility to anti-theta-serum and the mechanisms by which they cause cell lysis.

Quantitation and clonal isolation of cytolytic T lymphocyte precursors selectively infiltrating murine sarcoma virus-induced tumors.

A limiting dilution mixed leukocyte-tumor cell microculture system was used to quantitate cytolytic T lymphocytes and their precursors (CTL-P), which infiltrate tumors induced by injection of Moloney sarcoma-leukemia virus (MSV-MoLV) complex into C57BL/6 mice. Leukocyte populations obtained from tumors collected on day 10 after virus injection were found to contain significantly higher frequencies of operationally defined (tumor-specific) CTL-P than either peripheral blood leukocytes (PBL) or spleen cells from the same animals. When these frequencies were normalized according to the content of Lyt-2+ T cells in each tissue, average CTL-P frequencies were found to be 1/9 in tumor-infiltrating cells vs. 1/41 in PBL. These results directly demonstrate selective accumulation of CTL-P in the tumor mass. A number of clonal isolates obtained from tumor-infiltrating leukocyte populations were expanded and studied for cytolytic activity and specificity. Of 11 isolates, 10 were found to have high cytolytic activity, leading to 50% lysis of the syngeneic MoLV-derived tumor target cells in 3.5 h at lymphocyte:target cell ratios ranging from 0.5:1 to 3.2:1. Furthermore, five randomly selected clones showed H-2 restriction by their selective lytic activity against MoLV-derived syngeneic MBL-2 target cells and their lack of activity against either MoLV-derived allogeneic (LSTRA) tumor cells or against syngeneic (EL4) or allogeneic (P815) target cells unrelated to MoLV.

Clonal heterogeneity in the functional requirement for Lyt-2/3 molecules on cytolytic T lymphocytes: analysis by antibody blocking and selective trypsinization.

While it is well established that murine cytolytic T lymphocytes (CTL) express the Lyt-2/3 molecular complex on their surface, conflicting results have been reported concerning the role of this complex in CTL activity. In the present study this question was reinvestigated at the clonal level. Although different (H-2b anti-H-2d) CTL clones expressed comparable amounts of Lyt-2/3 molecules, as assessed by quantitative flow microfluorometry, the activity of some clones was inhibited by low doses (10 ng) of monoclonal anti-Lyt-2 or anti-Lyt-3 antibodies (in the absence of complement), whereas other clones were not inhibited by either antibody at doses as high as 5 microgram. Treatment of these clones with doses of trypsin sufficient to cleave Lyt-2/3 antigenic determinants from the cell surface resulted in a similar dissociation: clones that were inhibited by antibodies lost cytolytic activity, whereas "uninhibited" clones were unaffected by trypsin treatment. Moreover, the dissociation observed among different alloreactive clones could be demonstrated with self-H-2-restricted (H-2b anti-MSV) clones exhibiting cross-reactivity with normal H-2d products. The lytic activity of these clones against the relevant syngeneic target cells was unaffected by anti-Lyt-2 antibodies or trypsin, whereas their cross-reactivity on H-2d target cells was abolished by either treatment. These results provide direct evidence for clonal heterogeneity in the requirement for Lyt-2/3 molecules in CTL-mediated lysis. It is proposed that the function of Lyt-2/3 molecules is to stabilize the interaction between CTL receptors and the corresponding antigens on the target cells and that the requirement for such a stabilization is correlated with low number and/or affinity of CTL receptors.

Direct transfer of antigen-specific cytolytic activity to noncytolytic cells upon fusion with liposomes derived from cytolytic T cell clones.

Murine cytolytic T lymphocytes (CTL) clones were solubilized in Nonidet P-40 detergent, and the matrix and membrane proteins separated from the nuclear constituents. These proteins, in combination with exogenous lipids and Sendai virus envelope proteins, were used to construct liposomes that were then fused with noncytolytic cloned T cell recipients. The resultant fusion products were found to be highly cytolytic and appeared to express the same specificity as the original donor clone. Further analysis showed that the liposomal transfer process was extremely efficient. Moreover, in addition to noncytolytic T cell clones, three transformed T cell lines and one B cell line were found to express specific cytolytic activity after fusion with appropriate liposomes. Inhibition experiments using monoclonal antibodies against target cell antigens, as well as analysis of the lytic reactivity pattern of the fusion products, confirmed the high degree of specificity conferred to the recipient cells. This study thus indicates that the two characteristics typically associated with CTL, namely antigen-specific recognition and cytolytic activity, can be solubilized from CTL and transferred to recipient cells that do not express these characteristics.

Antigen recognition by H-2-restricted cytolytic T lymphocytes is not mediated by two independent receptors.

Detergent-solubilized murine cytolytic T lymphocytes (CTL) clones were incorporated into Sendai virus-containing synthetic liposomes. When these liposomes were then fused with other CTL clones possessing a different non-cross-reacting specificity, the fusion products were observed to lyse target cells recognized by both parental CTL clones. This method was then used with two H-2-restricted CTL clones of different, non-cross-reacting specificities (anti-H-2b-H-Y or anti-H-2b Moloney leukemia virus). Once again, the fusion products were found to be lytic against both target cells recognized by the parental clones, but in no instance was there any observable lysis of target cells bearing the same nominal antigen in the context of different H-2 molecules. These results provide strong evidence that antigen recognition by H-2-restricted CTL is not mediated by two independent antigen receptors.

Generation of cytotoxic T lymphocytes in vitro. II. Effect of repeated exposure to alloantigens on the cytotoxic activity of long-term mixed leukocyte cultures.

Mouse cytotoxic T lymphocytes (CTL) were generated in unidirectional mixed leukocyte cultures (MLC) using normal C57BL/6 spleen cells as responding cells and irradiated DBA/2 spleen cells as stimulating cells. Cytotoxicity was assayed on (51)Cr-labeled P-815 (DBA/2) target cells, and the relative frequency of CTL in individual cell populations was estimated from dose-response curves. Upon inclusion of 2-mercaptoethanol in the culture medium, it was found that significant CTL activity could be detected for as long as 3 wk in primary MLC. Reexposure of MLC cells to the original stimulating alloantigens after 14-41 days in culture resulted in significant cell proliferation and rapid regeneration of high levels of immunologically specific cytotoxicity. CTL activity in these secondary cultures increased dramatically within the first 24 h and reached higher peak levels than those found at the peak of the primary response. Furthermore, proliferation and reappearance of CTL activity could be demonstrated following each of as many as four sequential alloantigenic stimulations of the same initial cell population at 20-day intervals. Interestingly, cells recovered from MLC at the peak of the primary response on day 4 were insensitive to further allogeneic stimulation. Taken together, these results are consistent with the hypothesis that CTL differentiate in MLC to become long-lived memory cells which gradually lose their cytotoxic activity. Upon reexposure to specific alloantigen, such memory CTL rapidly regain their functional activity and proliferate to generate an expanded CTL population.