Receptor-like role of HLA-class I antigens: regulation of T cell activation

Class I major histocompatibility antigens are known to restrict the cytotoxic activity of T lymphocytes. However, experiments using monoclonal antibodies against class I antigens showed that these antigens also play some role in the regulation of T cell activation. Three monoclonal antibodies, namely W6/32 (anti-class I HLA-A, B, C, antigens), 4E (anti-class I HLA-B antigens), and BBM.1 (anti-beta 2-microglobulin) significantly suppressed the phytohemagglutinin-induced T cell proliferation. The inhibitory effect of anti-class I antibody was found to depend on the presence of monocyte/macrophage-type adherent cells. In the presence of antibody, adherent cells released a factor that suppressed T cell proliferation. These results suggest that HLA class I antigens on Mo1+ monocyte/macrophage cells function like ligand-receptor molecules, and regulate the secretion of suppressor factor(s).     

Cross-reactivity between human and canine helper and suppressor T cell antigens using monoclonal antibodies RPA-T4 and HuLy-m8

The murine monoclonal antibodies RPA-T4 and HuLy-m8, specific for a framework determinant of human helper/inducer and suppressor/cytotoxic T cell antigens, cross-reacted with canine cell membrane molecules recognizing a biomolecular complex (50,000 to 55,000 daltons) similar to that described in humans. We investigated the distribution of these helper and suppressor T cell-like antigens on canine peripheral blood lymphocytes. With complement-mediated lymphocytotoxicity, 34% and 35% of the canine lymphocytes expressed the helper T cell-like antigens and the suppressor-like T cell antigens, respectively. When the lymphocytes were treated with RPA-T4 and HuLy-m8, the respective helper and suppressor function was significantly inhibited.     

Distinctive functional properties of human blood L lymphocytes: a comparison with T lymphocytes, B lymphocytes, and monocytes.

Human blood lymphocytes with high affinity Fc receptors have been operationally named L lymphocytes because of membrane-labile IgG markers. L lymphocytes lack membrane-incorporated immunoglobulin and do not form rosettes with sheep red blood cells coated with IgM antibody and mouse complement. These lymphocytes are capable of binding IgG in normal human serum at 4 degrees C and will form rosettes with human lymphocytes coated with Ripley IgG. In this study, functional in vitro properties of isolated L lymphocytes were compared with T lymphocytes, B lymphocytes, and monocytes. To obtain these mononuclear populations, first, plastic adherent monocytes were harvested. T lymphocytes were then isolated by centrifugation of E rosette-forming cells, and other rosetting techniques were employed to isolate L and B lymphocytes by negative selection. The functional properties of L lumphocytes were completely unlike those of T cells, B cells, or monocytes. L lymphocytes did not proliferate in response to mitogens, soluble antigens, or cell surface antigens. Moreover, this population could not replace monocytes in helping T lymphocytes respond to concanavalin A and pokeweed mitogen. Once T cells were supplemented with monocytes, however, the addition of L lymphocytes to the culture greatly enhanced the T lymphocytes proliferative response to phytohemagglutinin, concanavalinA, purified protein derivative (PPD), and streptokinase/streptodornase. L lymphocytes were not a subset of B cells. They did not spontaneously develop surface Ig in culture, and pokeweek mitogen could not induce them to transform and generate cytoplasmic Ig detectable by immunofluorescence. Mixtures of B cells and T cells responded to pokeweed mitogen better than do T cells alone. In contrast, enhanced reactivity with L and T cell combinations was not observed. Another sharp difference between these two populations was the stimulator capacity of each in mixed lymphocyte culture. When B and L lymphocytes were carefully monocyte-depleted, only B cells were effective stimulators of autologous and allogeneic lymphocytes. In comparison with T cells, B cells, and monocytes, L lymphocytes were the only effective killers of human blood lymphocytes sensitized with IgG. L lymphocytes, then, have cytotoxic potential, but cannot proliferate in response to various stimulants or become antibody-producing cells. These findings suggest that L lymphocytes comprise a third lymphocyte population.     

Evidence for specific association between class I major histocompatibility antigens and the CD8 molecules of human suppressor/cytotoxic cells

Human T lymphocytes, metabolically labeled with 35S-cysteine and 35S-methionine, were reacted with the homobifunctional cross-linking reagent, dithiobis (succinimidyl propionate) (DSP). When detergent lysates from these cells were immunoprecipitated with a monoclonal antibody reactive with the CD8 antigen, a radiolabeled protein of approximately 44 kd was coprecipitated with the CD8 molecule. Immunoprecipitates from detergent lysates prepared without prior chemical cross-linking contained only the 33 kd CD8 molecule. Similar results were obtained when T lymphocytes or a cytotoxic T cell clone (T4T8Cl) were radiolabeled with 32P-orthophosphoric acid. The 44 kd CD8-associated protein was identified as the heavy chain of the class I major histocompatibility antigen by depletion in preclearing experiments with anti-class I MHC antibody and by peptide mapping. Further analyses indicated that the CD8-class I MHC association is due, in part at least, to disulfide bonding, which may be susceptible to cleavage during processing of cell lysates.     

Phosphorylation and down-regulation of CD4 and CD8 in human CTLs and mouse L cells

The CD4 and CD8 molecules are rapidly phosphorylated following exposure of CD4⁺ or CD8⁺ human cytotoxic T lymphocytes (CTL) clones to B-lymphoblastoid cell lines bearing the relevant target alloantigens. Treatment of CD4⁺ or CD8⁺ CTL clones with phorbol myristate acetate (PMA), phytohemagglutinin, or mitogenic combinations of CD2-specific antibodies also resulted in CD4 or CD8 phosphorylation. Down-regulation of the surface expression of these molecules could be demonstrated in both CD4⁺ and CD8⁺ clones following exposure to the relevant alloantigen or PMA. Parallel experiments were conducted using mouse L cells in which the human CD4 or CD8 antigens were stably expressed. Exposure of these transfectants to PMA induced rapid phosphorylation of the CD4 and CD8 molecules. As in CD4⁺ CTL clones, rapid modulation of the CD4 antigen could be demonstrated in L cells following PMA treatment. In contrast, there was no demonstrable down-regulation of the CD8 antigen in PMA-treated CD8⁺ L cell transfectants. These studies demonstrate a significant differential property of the CD4 and CD8 antigens and suggest that down-regulation of the CD8 antigen may require its expression in a T-cell environment and/or the association of CD8 with the T-cell receptor or other T cell-specific molecules.     

Regeneration of T cell subpopulations after bone marrow transplantation: cytomegalovirus infection and lymphoid subset imbalance

Immune reconstitution was studied serially by using T lymphocyte cell surface differentiation antigens in 37 individuals receiving bone marrow transplants. Antigen expression was assessed by immunofluorescence analysis using monoclonal antibodies to T lymphocytes including Leu-3, which defines a T lymphocyte subpopulation in healthy individuals with T helper or inducer activity (L3+), and Leu-2, which defines a T lymphocyte subpopulation in healthy individuals with T helper or inducer activity (L3+), and Leu-2, which defines a T lymphocyte subpopulation with suppressor or cytotoxic activity (L2+). These studies demonstrated that the L2+ subpopulation regenerated more rapidly after transplant than did the L3+ subpopulation. Imbalances between these two T lymphocyte subpopulations, indicated by a decreased L3/L2 ratio, persisted for periods up to 12 mo post-transplant. Expression of a cell surface antigen associated with immature lymphocytes (OKT-10), and of HLA-DR (Ia-like) antigens was markedly increased during the post-transplant period. HLA-DR antigen expression did not appear related to immune activation in that increased reactivity was not detected with a monoclonal antibody (anti-TAC) specific for activated T cells. These observations in bone marrow transplant recipients and other disorders characterized by lymphoid restoration or immaturity indicate that inversion of the normal L3/L2 ratio and increased expression of OKT-10 and HLA-DR antigens may be features of a regenerating immune system. Furthermore, serial observation of individual patients indicated that infection with cytomegalovirus was associated with a progressive decrease in the L3/L2 ratio.     

Murine lymphocyte and embryonal carcinoma cell surface antigens recognised by rabbit anti-murine embryonal carcinoma serum

Some recent studies have suggested that murine embryonal carcinoma (EC) cells which lack classical H-2 molecules must express some novel class-I-MHC (major histocompatibility complex) type antigens. We have investigated whether rabbit anti-EC sera may recognize such components. Molecules of 40 and 48 kd were immunoprecipitated from EC cells and lymphocytes. The possibility that these molecules may be coded for by genes in the Qa-Tla locus which contains many MHC class-I genes or pseudogenes was investigated. The 40 and 48 kd molecules were not associated with beta 2-microglobulin on EC cells, nor with each other or other molecules through S-S bonds, although they appeared to have intradisulphide structures. Peptide map analysis suggested that the lymphocyte 40 and 48 kd molecules were related to the EC cell 40 and 48 kd molecules. However, these molecules were different from H-2, Ia or Qa-2 molecules from the same mouse strain.     

Qualitative and quantitative studies of antigen-presenting cell function by using I-A-expressing L cells

I-A-expressing transfected murine L cells were analyzed as model antigen-presenting cells. Four features of accessory cell function were explored: antigen processing, interaction with accessory molecules (LFA-1, L3T4), influence of Ia density, and ability to stimulate resting, unprimed T lymphocytes. I-A+ L cells could present complex protein antigens to a variety of T cell hybridomas and clones. Paraformaldehyde fixation before but not subsequent to antigen exposure rendered I-A+ L cells unable to present intact antigen. These results are consistent with earlier studies that made use of these methods to inhibit "processing" by conventional antigen-presenting cells. The ability of anti-L3T4 antibody to inhibit T cell activation was the same for either B lymphoma or L cell antigen-presenting cells. In striking contrast, anti-LFA-1 antibody, which totally blocked B lymphoma-induced responses, had no effect on L cell antigen presentation, measured as interleukin 2 (IL 2) release by T hybridomas, proliferation, IL 2 release, or IL 2 receptor upregulation by a T cell clone. I-A+ L cell transfectants were found to have a stable level of membrane I-A and I-A mRNA, even after exposure to interferon-gamma-containing T cell supernatants. In agreement with earlier reports, a proportional relationship between the (Ia) X (Ag) product and T cell response was found for medium or bright I-A+ cells. However, dull I-A+ cells had a disproportionately low stimulatory capacity, suggesting that there may be a threshold density of Ia per antigen-presenting cell necessary for effective T cell stimulation. Finally, I-A-bearing L cells were shown to trigger low, but reproducible primary allogeneic mixed lymphocyte responses with the use of purified responder T cells, indicating that they are capable of triggering even resting T cells. These studies confirm the importance of antigen processing and I-A density in antigen-presenting cell function, but raise questions about the postulated role of the LFA-1 accessory molecule in T cell-antigen-presenting cell interaction. They also illustrate the utility of the L cell transfection model for analysis and dissection of antigen-presenting cell function.     

Coexpression of the human HLA-A2 or HLA-B7 heavy chain gene and human beta 2-microglobulin gene in L cells

L cells expressing human HLA-A2 or HLA-B7 class I antigen heavy chains are not recognized by human cytotoxic T lymphocytes directed at HLA-A2 or HLA-B7 antigens. To test whether the absence of human beta 2-m was the cause of the lack of recognition by the human cytotoxic T lymphocytes, coexpression of the human beta 2-m gene and the HLA-A2 or HLA-B7 heavy chain in L cells ("double transfectants") was obtained. In addition, L cells expressing HLA-A2 or HLA-B7 antigens in association with human beta 2-m were obtained by an exchange reaction, in which human beta 2-m from serum replaced the endogenous murine beta 2-m. Both types of transfectant cells were used in 51Cr-release assays and cold target inhibition assays for human cytotoxic T cell clones which were directed at HLA-A2 or HLA-B7. Neither human CTL clones nor a mixture of CTL specific for HLA-A2 and HLA-B7 were able to recognize these cells. Several alternative explanations for these observations are discussed.     

L3T4+ cytotoxic T lymphocytes specific for class I H-2 antigens are activated in primary mixed lymphocyte reactions

Thy-1+, L3T4+, Ly-2- cytotoxic lymphocytes (CTL) are generated in a primary anti-H-2d mixed lymphocyte reaction, by using responders depleted of Ly-2+ cells. In addition to expressing the L3T4 marker, as detected by anti-L3T4 antibody and complement-mediated elimination, the L3T4+ CTL are inhibited by L3T4 antibody. The observation of these L3T4+ CTL in cells recovered from primary mixed lymphocyte reactions confirms the previous reports. However it is demonstrated for the first time that a subpopulation of these are class I-specific by their specific inhibition with an antiserum to class I antigens. The class I specificity of the CTL was further shown by their ability to kill class II antigen negative P815 tumor cells. The lysis of this target cell by L3T4+ CTL was also specifically blocked by the class I antiserum. The data is consistent with the presence also of a class II-specific population of L3T4+ cytotoxic cells. The fact that a level of L3T4+ cell-mediated cytotoxic activity comparable to Ly-2+ cytolytic activity is generated in a primary mixed lymphocyte response, even though the precursor frequency of L3T4+ killer cells is 10 times lower than for Ly-2+ killers, is suggestive of their physiologic significance. It was also shown that the activation of these cells is not dependent on the presence of xenogeneic serum components or exogenous helper or mitogenic factors in the culture medium. The findings provide further evidence against both the phenotype-function and phenotype-major histocompatibility complex antigen specificity models of T cell diversity.     

Purification, chain separation and sequence of the MRC OX-8 antigen, a marker of rat cytotoxic T lymphocytes.

The MRC OX-8 antigen is a marker of the rat cytotoxic T lymphocytes that consists of disulphide-linked chains of mol. wts. 37 and 32 kd. It is thought to be equivalent to the human T8 and mouse Lyt2,3 antigens (all now called CD8 antigens). MRC OX-8 antigen was purified from thymocytes using a monoclonal antibody column and because antigenicity was retained after reduction and alkylation the two polypeptide chains could be separated by a subsequent affinity chromatography step. Peptides were isolated from each chain and their sequences determined. A cDNA probe coding for the mouse CD8 antigen (pLY2C-1 provided by Dr L. A. Herzenberg) was used to obtain rat cDNA clones from which the sequence of the equivalent rat molecule was determined. Peptides from the 32-kd chain were identified in this translated sequence whereas peptides from the 37-kd chain were not. The 32-kd polypeptide sequence consisted of 210 amino acids and had one possible N-linked glycosylation site. The N-terminal part of the sequence was surprisingly different from both its mouse and human counterparts but, as in the other two species, it showed a clear relationship to Ig V domains.     

Participation of L3T4 in T cell activation in the absence of class II major histocompatibility complex antigens. Inhibition by anti-L3T4 antibodies is a function both of epitope density and mode of presentation of anti-receptor antibody

The recognition of many class II major histocompatibility complex (MHC)-associated antigens by T cells requires the participation of the L3T4 molecule. It has been proposed that this molecule acts to stabilize low affinity binding to antigen in association with MHC and thereby increases the avidity of T cell/antigen interactions. By using antibodies against the T cell antigen receptor (TCR) to activate T cells, thereby circumventing the requirement for antigen presenting cells and MHC-associated antigen, we have been able to study the function of L3T4 in the absence of class II MHC. We have used two monoclonal antibodies, KJ16-133.18 and F23.1, that recognize a determinant encoded by the T cell receptor V beta 8 gene family. These antibodies were used to select two clones of T cells with surface phenotype Thy-1.2+, L3T4+, Lyt-2-, KJ16-133.18+, F23.1+, IA-, IE-. One of these clones (E9.D4) was hapten-specific (anti-ABA + Iak), the other (4.35F2) was alloreactive (anti-Iak). Activation of these clones by antigen, concanavalin A (Con A) or by the F23.1 antibody was studied by assaying the production of interleukin 3 (IL 3). Both soluble and solid phase-coupled F23.1 induced T cell activation in the complete absence of class II MHC, immobilized antibody (either Sepharose-coupled or plastic-adsorbed) being more effective. The induction of IL 3 production by suboptimal doses of either Con A or plastic-adsorbed F23.1 was inhibited by the anti-L3T4 antibody GK1.5, as was the response to F23.1 coupled to Sepharose-4B beads. However, the responses to optimal or superoptimal doses of these stimuli were not inhibited. In contrast, weak responses to non-TCR cross-linking stimuli such as phorbol myristate acetate (PMA) or low concentrations of soluble F23.1 were not inhibited by GK1.5 (the latter response was usually slightly enhanced). These results show that anti-L3T4 antibodies are not inherently inhibitory, but require both ligation and cross-linking of the TCR for their effect. We propose a model whereby L3T4 interacts with the TCR during T cell activation. Anti-L3T4 antibodies sterically hinder the formation of TCR complexes and so prevent activation. However, by increasing the epitope density of the activating ligand, the avidity of the T cell/ligand interaction can be increased sufficiently to prevent this disruption.(ABSTRACT TRUNCATED AT 400 WORDS)     

In vitro induction of tumor-specific immunity. III. Lack of requirement for H-2 compatibility in lysis of tumor targets by T cells activated in vitro to oncofetal and plasmacytoma antigens.

Many recent studies have demonstrated that cytotoxic T lymphocytes (CL) activated to various antigens other than those of the H-2 complex, will lyse target cells only when H-2 compatibility exists between the CL and target cell. From these observations, it has been inferred that T lymphocytes might only be capable of responding to H-2 antigen or antigens that become associated with H-2 region gene products. Our results suggest that this is not the case, and that in some situations, cytotoxic T lymphocytes can specifically lyse target cells of different H-2 types. Two in vitro systems are described where primary induction of cytotoxic T lymphocytes to oncofetal and plasmacytoma antigens results in CL capable of lysing suitable targets bearing these antigens, of either syngeneic or allogeneic derivation. Thus it is proposed that although interaction antigens involving H-2 components may preferentially activate T lymphocytes, this does not imply a restriction on the recognition potential of T lymphocytes.     

Idiotype and antigen-specific T cell responses in mice on immunization with antigen, antibody, and anti-idiotypic antibody.

Idiotypic determinants of immunoglobulin molecules can evoke both CD4(+) and CD8(+) T responses and exist not only as the integral components of a bona fide antigen binding receptor but also as distinct molecular entities in the processed forms on the cell surface of B lymphocytes. The present work provides experimental evidence for the concept that regulation of memory B cell populations can be achieved through the presentation of idiotypic and anti-idiotypic determinants to helper and cytotoxic cell. The potential of B cells to present antigens to helper and cytotoxic T cells through class II and class I MHC suggests a mechanism by which both B and T cell homeostasis can be maintained. We provide evidence for the generation of idiotype- and antigen-specific Th and Tc cells upon immunization of syngenic mice with antigen or idiotypic antibody (Ab1) or anti-idiotypic antibody (Ab2). The selective activation and proliferation of the antigen-specific Th and Tc cells mediated by idiotypic stimulation observed in these experiments suggests a B-cell-driven mechanism for the maintenance of antigen-specific T cell memory in the absence of antigenic stimulation, under certain conditions.     

Function of the CD4 and CD8 molecules on human cytotoxic T lymphocytes: regulation of T cell triggering

The function of the T cell differentiation antigens CD4 (Leu-3/T4) and CD8 (Leu-2/T8) on human cytotoxic T lymphocytes (CTL) is presently seen only in conjugate formation between CTL and target cell via class II or class I MHC antigens rather than in the later killing steps. In this study, human CD4+ and CD8+ CTL clones were used to investigate the effects of monoclonal antibodies against these differentiation antigens on nonspecific triggering of cytotoxicity. Cytotoxicity was induced either by antibodies against the CD3 (T3) antigen or by the lectins Con A and PHA. Anti-CD4 or anti-CD8 antibodies specifically inhibited all types of cytotoxicity of CD4+ or CD8+ CTL, respectively, regardless of the specificity of the CTL for class I or class II HLA antigens and regardless of whether target cells expressed class I or class II antigens. These results are incompatible with an exclusive role of the CD4 and CD8 molecules in MHC class recognition and are discussed with respect to a function as negative signal receptors for these molecules on CTL.     

A labile antigen complex at the lymphocyte surface: the effect of the substratum on its expression

Human T lymphocytes carry a surface antigen, detectable by a monoclonal antifibronectin antibody, which appears to consist of 150 and 55 kd components as revealed by SDS-PAGE. After in vitro culture of the lymphocytes on a plastic substratum for 48 hr comparatively few cells (40 +/- 18% in separate individuals) express the antigen. In contrast, the vast majority of lymphocytes cultured on a collagen matrix for the same time period maintains surface expression of the antigen (76 +/- 14% in separate individuals). Conditioned media from lymphocytes on plastic contain substantial amounts of antigenicity detectable by the same antibody, whereas conditioned media from lymphocytes on collagen are devoid of such antigenicity. The expression at the cell surface of other T lymphocyte antigens (Leu-4, Leu-3, and OKT8) is identical during culture on plastic and collagen for 48 hr. Collagen does not activate the cells to DNA synthesis or expression of IL 2 receptors, and consequently the potentiation of antigen expression by this substratum cannot be attributed to a mitogenic effect. The composition of subsets of T lymphocytes and the viability of the cells are the same on plastic and collagen, which excludes that the substratum-dependent variations in antigenicity reflect selection or loss of antigen-bearing cells. Thus, substratum-dependent regulation of the expression at the cell surface appears to be a unique property of the 150/55 kd T cell surface antigen. Culture on collagen substrata augments the number of lymphocytes showing motile behavior two to four times compared with culture on plastic.     

Distinct roles for L3T4 in T cell activation

The mutually exclusive expression of L3T4 and Lyt-2 on murine T cells and the correlation of their expression to the major histocompatibility complex (MHC) restriction of the T cell antigen receptor (Ti) have led to the hypothesis that these surface molecules are related to recognition of class II and class I MHC antigens, respectively. It has been suggested that these T cell surface molecules interact with nonpolymorphic determinants on MHC antigens. We have studied the role of L3T4 in activation of an H-2Dd-specific T cell hybridoma. This novel hybridoma allowed the separate evaluation of the specificities of Ti and L3T4 and the examination of their roles in T cell activation. Antibody-blocking experiments have demonstrated that L3T4 was involved in triggering this T cell hybridoma only if the antigen-bearing cell expressed Ia. The apparent requirement for an L3T4-Ia interaction reflected the amount of available H-2Dd antigen. It appears that the L3T4-Ia interaction influences T cell activation during suboptimal antigenic stimulation. We have begun to examine the role of L3T4 in lectin and anti-Ti monoclonal antibody stimulation of the same T cell hybridoma. These experiments have suggested a distinct role for L3T4 in the absence of Ia, as a mediator of a negative signal for activation.     

Induction of antigen-specific regulatory T cells following overexpression of a Notch ligand by human B lymphocytes

In mice, activation of the Notch pathway in T cells by antigen-presenting cells overexpressing Notch ligands favors differentiation of regulatory T lymphocytes responsible for antigen-specific tolerance. To determine whether this mechanism operates in human T cells, we used Epstein-Barr virus-positive lymphoblastoid cell lines (EBV-LCL) as our (viral) antigen-presenting cells and overexpressed the Notch ligand Jagged-1 (EBV-LCL J1) by adenoviral transduction. The EBV-LCL J1s were cocultured with autologous T cells, and the proliferative and cytotoxic responses to EBV antigens were measured. Transduction had no effect on EBV-LCL expression of major histocompatibility complex (MHC) antigens or of costimulatory molecules CD80, CD86, and CD40. However, we observed a 35% inhibition of proliferation and a >65% reduction in cytotoxic-T-cell activity, and interleukin 10 production was increased ninefold. These EBV-LCL J1-stimulated T lymphocytes act as antigen-specific regulatory cells, since their addition to fresh autologous T cells cultured with autologous nontransduced EBV-LCL cells significantly inhibited both proliferation and cytotoxic effector function. Within the inhibitory population, CD4(+)CD25(+) and CD8(+)CD25(-) T cells had the greatest activity. This inhibition appears to be antigen-specific, since responses to Candida and cytomegalovirus antigens were unaffected. Hence, transgenic expression of Jagged-1 by antigen-presenting cells can induce antigen-specific regulatory T cells in humans and modify immune responses to viral antigens.     

Lectin-dependent and anti-CD3 induced cytotoxicity are preferentially mediated by peripheral blood cytotoxic T lymphocytes expressing Leu-7 antigen

Monoclonal antibodies against the CD3 antigen and certain lectins can induce interleukin 2 dependent antigen-specific T cell clones to mediate non-antigen specific cytotoxicity. On the basis of this observation, we predicted that it may be possible to identify cytotoxic T lymphocytes (CTL) in peripheral blood without knowing the antigen specificity of these in vivo primed CTL. By using this strategy, peripheral blood lymphocytes were separated into low and high-density fractions on Percoll gradients and were tested for cytotoxic activity in the presence or absence of concanavalin A (Con A) or anti-Leu-4 antibody. Lectin-dependent cellular cytotoxicity (LDCC) and anti-CD3 induced cytotoxicity against both natural killer (NK)-insensitive and NK-sensitive targets were exclusively mediated by low-density CD3+ T lymphocytes. Additional studies indicated that low-density CD3+ T lymphocytes co-expressing Leu-7 antigen preferentially mediated this activity, although in some individuals, significant activity was also observed in the low-density T cells lacking Leu-7. In contrast, high-density CD3+ T lymphocytes and CD16+ (Leu-11+) NK cells (both Leu-7 and Leu-7+) did not mediate nonantigen-specific cytotoxicity under these conditions. The finding that NK cell-mediated cytotoxicity was unaffected by these lectins refutes the hypothesis that lectin-dependent cellular cytotoxicity is simply a result of effector and target agglutination. T cell-mediated cytotoxicity was both lectin and antibody specific. Phytohemagglutinin, Con A, and pokeweed mitogen induced cytolytic activity in the Leu-7+ T cells, whereas wheat germ agglutinin did not. Of the antibodies against T cell-associated differentiation antigens (anti-Leu-2,3,4, and 5), only anti-Leu-4 induced cytotoxicity. This anti-CD3-induced cytotoxicity was essentially completely inhibited by the presence of anti-LFA-1 or anti-CD2 monoclonal antibodies, implicating these molecules in the triggering process. A proportion of the CD3+, Leu-7+ CTL expressed HLA-DR antigens, indicating possible in vivo activation. Because previous clinical studies have indicated that lymphocytes with this phenotype may be elevated in clinical situations associated with immunosuppression and chronic viral infection, this unique subset of CD3+ T lymphocytes may represent a population of in vivo primed CTL possibly against viral antigens.