Natural killer cells vs cytotoxic T cells in the peripheral blood of virus-infected mice

The cell-mediated immune response of mice against various enveloped RNA and DNA viruses expressed by immune lymphocytes from the spleen and the peripheral blood (PBL) were compared. PBL from mice of various strains infected with vaccinia virus, vesicular stomatitis virus (VSV), or lymphocytic choriomeningitis virus (LCMV) were tested on histocompatible or incompatible target cells infected with the homologous virus. PBL from immune mice showed clear H-2 restriction, but additionally, they expressed high natural killing (NK) activity on YAC-1 cells. The high NK-cytolytic activity of PBL on YAC-1 differed significantly from that expressed by splenic lymphocytes. In both lymphocyte populations lysis was detected as early as 1 day after infection; NK activity decreased in the spleen after day 4 post infection, whereas that of PBL persisted at high levels for up to 10 days after infection. Treatment of mice with anti-asialo GM1 in vivo abrogated NK activity in PBL effector cells tested in vitro. These results may explain some of the difficulties to observe MHC-restricted cytotoxic T cells in PBL from humans or primates during primary infections with virus.     

Antiviral T cell competence and restriction specificity of mixed allogeneic (P1 + P2----P1) irradiation chimeras

Mixed irradiation bone marrow chimeras were prepared by reconstituting lethally irradiated C57BL/10 (B10) or B10.D2 mice with T cell-depleted bone marrow cells of B10 plus B10.D2 origin. These chimeras were healthy and survived well under conventional housing conditions and after experimental laboratory infections. Of a total of 17 chimeras tested, 2 died spontaneously or from the injected virus. Twelve of fifteen chimeras mounted a measurable cytotoxic T cell response to virus. Despite approximately equal percentages of B10 and B10.D2 lymphocytes in chimeras, cytotoxic T cell responses to vaccinia virus and lymphocytic choriomeningitis virus were mediated variably by either syngeneic or allogeneic donor lymphocytes; thus the H-2 type of effector T cells frequently did not correspond to the 50:50 distribution of spleen or peripheral blood lymphocytes. Cytotoxic responses were restricted exclusively to recipient H-2 type. All mixed chimeras examined were able to mount a good IgG response to vesicular stomatitis virus. These results confirm previous data suggesting that such mixed chimeras are healthy and immunocompetent and demonstrate strict recipient-determined restriction specificity of effector T cells; they also suggest that if T help is necessary for induction of virus-specific cytotoxic T cells, it does not require host-restricted interactions between helper T cells and precursor cytotoxic T cells.     

Generation of cytotoxic lymphocytes in mixed lymphocyte reactions II. Importance of private and publicH-2 alloantigens on the expression of cytotoxicity

MLC were established to test for the generation of specific cytotoxic effector cells in CML. The target cell used to assay for CML in the five combinations tested was of a differentH-2 haplotype from the stimulating cell population. Cytotoxicity was observed against this target only when it shared private alloantigens (antigens that are specific for theH-2D andH-2K region of differentH-2 haplotypes) with the stimulating cell population. Very weak or no Cytotoxicity was found when such alloantigens were not shared, although cross-reactive publicH-2 specificities were. These findings indicate that T cells display a cytotoxic potential against privateH-2 antigens in a primary response in vitro and are not capable of responding to publicH-2 specificities to the same level.BSS balanced salt solution - CML cell-mediated lympholysis - GPC guinea pig complement - ¹²⁵IUdR ¹²⁵I-iodo-deoxyuridine - MLC mixed lymphocyte culture - SE standard error     

Cytotoxic T lymphocytes recognize determinants on the BALB/c-H-2Ld molecule controlled by α1 and α2 but not α3 external domains

We have shown that cytotoxic T lymphocytes (CTL) raised in H-2 ^dmice use H-2L^d but not H-2D^d or H-2K^d antigens as restricting elements in lymphocytic choriomeningitis virus (LCMV) and vesicular stomatis virus (VSV) infections. To localize the regions of H-2L^d protein recognized by CTL, we constructed a recombinant H-2L ^d/D ^dgene encoding a hybrid antigen with 1 and 2 external domains of H-2L^d and 3, transmembrane and cytoplasmic domains of H-2D^d. The recombinant gene was transfected into mouse cells and the hybrid molecules were characterized serologically, biochemically and functionally. In all assays, H-2L^d/D^d molecules were recognized by LCMV- and VSV-specific H-2L^d-restricted CTL in a manner similar to that of wild-type H-2L^d antigens. Analogous results were obtained with alloreactive CTL. Hybrid antigens containing the 3 domain of H-2L^d fused to 1 and 2 domains of a Qa-2,3 region-encoded antigen were not used as restricting elements by LCMV-specific CTL. These results suggest that H-2L^d-restricted CTL directed against LCMV and VSV recognize determinants controlled by the 1 and/or 2 domains of the H-2L^d molecule.Abbreviations used in this paper CTL cytotoxic T lymphocytes - VSV vesicular stomatitis virus - LCMV lymphocytic choriomeningitis virus - tk thymidine kinase - HAT hypoxanthine, aminopterine, thymidine - HSV herpes simplex virus - FCS fetal calf serum - SAC Staphylococcus aureus Cowan I strain - TM transmembrane - CYT cytoplasmic     

Antigen presentation by nonhemopoietic cells amplifies clonal expansion of effector CD8 T cells in a pathogen-specific manner

Professional APCs of hemopoietic-origin prime pathogen-specific naive CD8 T cells. The primed CD8 T cells can encounter Ag on infected nonhemopoietic cell types. Whether these nonhemopoietic interactions perpetuate effector T cell expansion remains unknown. We addressed this question in vivo, using four viral and bacterial pathogens, by comparing expansion of effector CD8 T cells in bone marrow chimeric mice expressing restricting MHC on all cell types vs mice that specifically lack restricting MHC on nonhemopoietic cell types or radiation-sensitive hemopoietic cell types. Absence of Ag presentation by nonhemopoietic cell types allowed priming of naive CD8 T cells in all four infection models tested, but diminished their sustained expansion by approximately 10-fold during lymphocytic choriomeningitis virus and by < or =2-fold during vaccinia virus, vesicular stomatitis virus, or Listeria monocytogenes infections. Absence of Ag presentation by a majority (>99%) of hemopoietic cells surprisingly also allowed initial priming of naive CD8 T cells in all the four infection models, albeit with delayed kinetics, but the sustained expansion of these primed CD8 T cells was markedly evident only during lymphocytic choriomeningitis virus, but not during vaccinia virus, vesicular stomatitis virus, or L. monocytogenes. Thus, infected nonhemopoietic cells can amplify effector CD8 T cell expansion during infection, but the extent to which they can amplify is determined by the pathogen. Further understanding of mechanisms by which pathogens differentially affect the ability of nonhemopoietic cell types to contribute to T cell expansion, how these processes alter during acute vs chronic phase of infections, and how these processes influence the quality and quantity of memory cells will have implications for rational vaccine design.     

Antigen-specific helper T cells recognize determinants encoded in the H-2D region of the H-2 gene complex

Observations have frequently been interpreted as showing that the helper T cells which collaborate with alloantigen-specific cytotoxic T-cell precursors can only recognize antigens encoded in the I region of the H-2 gene complex. An experimental system is described here that allows analysis of the recognition repertoire of these helper cells. CBA helper T-cell precursors can be primed in vitro to antigens encoded in the H-2 ^b gene complex. These helpers can then be tested for the existence of a subset of helper cells which recognize antigens encoded in the D region of H-2 ^b haplotype. CBA thymocytes were used as a source of cytotoxic T-cell precursors that respond poorly in the absence of exogeneous helper activity. The source of alloantigen was varied by using irradiated spleen cells from various (BALB/c × recombinant)F₁ hybrid mice as stimulator cells. When the stimulator cell bears BALB/c determinants recognized by the cytotoxic T-cell precursor and also bears only the D region antigens of the H-2 ^b haplotype, an anti-BALB/c cytotoxic response is generated only if the anti-H-2^b helper population contains cells able to recognize H-2D^b. A positive cytotoxic response was obtained, indicating that helper cells are not limited to recognition of I region antigens and can efficiently recognize antigens encoded in the D region of the H-2 gene complex. This was confirmed by the demonstration of helpers specific for H-2D^d. We were unable to detect any evidence for Ia-restricted recognition of the H-2D alloantigens, suggesting that, as for cytotoxic T lymphocytes (CTL), helper cell recognition of class I alloantigens is an unrestricted event.     

Bi-specific MHC Heterodimers for Characterization of Cross-reactive T Cells

T cell cross-reactivity describes the phenomenon whereby a single T cell can recognize two or more different peptide antigens presented in complex with MHC proteins. Cross-reactive T cells have previously been characterized at the population level by cytokine secretion and MHC tetramer staining assays, but single-cell analysis is difficult or impossible using these methods. In this study, we describe development of a novel peptide-MHC heterodimer specific for cross-reactive T cells. MHC-peptide monomers were independently conjugated to hydrazide or aldehyde-containing cross-linkers using thiol-maleimide coupling at cysteine residues introduced into recombinant MHC heavy chain proteins. Hydrazone formation provided bi-specific MHC heterodimers carrying two different peptides. Using this approach we prepared heterodimers of the murine class I MHC protein H-2K^b carrying peptides from lymphocytic choriomeningitis virus and vaccinia virus, and used these to identify cross-reactive CD8+ T cells recognizing both lymphocytic choriomeningitis virus and vaccinia virus antigens. A similar strategy could be used to develop reagents to analyze cross-reactive T cell responses in humans.     

Loss of reactivity of a myeloma tumor with H-2 compatible thymus-derived lymphocytes.

It was previously shown that MOPC-315-EL, a subline of the BALB/c myeloma tumor MOPC-315, reversibly alters its reactivity with T cells that recognize H-2^d, 2,4,6-trinitrophenyl, and minor histocompatibility antigens. This report demonstrates (a) that CTLs directed against vesicular stomatitis virus and Sendai virus are unable to recognize viral antigens associated with the unreactive tumor cell, and (b) that incorporation of Sendai virus antigens into the same membrane as the H-2 gene products is required for effective recognition by cytotoxic T lymphocytes.     

The spatial relationship of the viral and H-2 antigens recognized by anti-viral CTLs

With the use of monospecific rabbit anti-G protein and mouse monoclonal anti-H-2K^k, we have analyzed the spatial relationship of the serologically defined H-2K^k antigens and the major surface glycoprotein (G protein) of vesicular stomatitis virus (VSV) to those antigens recognized by B10.A (k, d) anti-VSV cytotoxic T lymphocytes (CTLs). The ability of monoclonal anti-H-2K^k or rabbit anti-G protein to inhibit specifically the cytolytic activity of B10.A anti-VSV CTLs indicates that the G protein and the H-2K^k molecules are in close proximity to the viral and H-2K^k antigens recognized by the anti-VSV (CTLs. By the method of sequential immunoprecipitation, we also demonstrated that only 10–30 percent of the serologically defined G and H-2K^k molecules are in theG-H-2K ^k complexes.Abbreviations used in this paper Con A Concanavalin A - cpm counts per minure - CTLs cytotoxic T lymphocytes - E: T ratio effector: target ratio - G major surface glycoprotein of VSV - MHC major histocompatibility complex - MOI multiplicity of infection - NP40 Nonidet-P40 - PAGE polyacrylamide gel electrophoresis - PBS phosphate-buffered saline - SaCI Staphylococcus aureus, Cowan I strain - SDS sodium dodecyl sulfate - UV ultraviolet light     

Effects of treatment with IL-2 receptor specific monoclonal antibody in mice. Inhibition of cytotoxic T cell responses but not of T help

Contribution of IL-2R-bearing activated lymphocytes to antiviral host defense was investigated in C57BL/6 mice by treatment in vivo with IL-2R-specific mAb PC61. When treated on days 0 and 1 with respect to infection with either vaccinia virus, lymphocytic choriomeningitis (LCM) virus (LCMV) or vesicular stomatitis virus, 6-day immune mice had low numbers of CD8+ T cells that were reduced to about 10% of the values found for infected but otherwise untreated controls. In contrast, the number of CD4+ T cells was within normal ranges. Correspondingly, induction of strictly T help-dependent antiviral neutralizing IgG antibody titers remained unaffected by the mAb treatment, whereas generation of antiviral cytotoxic T cell activity was abrogated. Anti-IL-2R treatment of thymectomized mice 14 and 15 days after infection prevented generation of secondary antiviral cytotoxic T cells in restimulation cultures in vitro initiated 24 days later. Treatment with IL-2R-specific mAb was comparable to treatment with CD8-specific mAb in preventing mice to eliminate virus. Because of the involvement of antiviral cytotoxic T cells in disease manifestations, treatment with IL-2R-specific mAb protected mice from lethal LCM after intracerebral infection with LCMV and inhibited the footpad swelling reaction caused by local infection with the same virus.     

Virus-specific T-cell sensitization

Syngeneic, semiallogeneic, or allogeneic spleen lymphocytes were transferred intonu/nu BALB/c mice, which were infected with vaccinia virus. Specific Sensitization of transferred thymus-derived cells was determined in vivo by mean survival time and virus titer in the spleen six days after infection, and in vitro by cell-mediated cytolysis of vaccinia virus-infected syngeneic target cells. Virus-specific Sensitization took place only after transfer of syngeneic or semiallogeneic spleen lymphocytes; allogeneic lymphocytes had no influence on mean survival time or virus titer and showed no virus-specific cytolytic activity in vitro. Infection of mice with vaccinia virus-strain WR, Elstree, DIs, or DIs-infected syngeneic fibroblasts resulted in the generation of virus-specific effector cells, while injection of a high amount of inactivated virus particles caused no Sensitization. These results suggest H-2 homology for production of virus-specific effector cells. Propagation of virus is not necessary, since early surface antigens, combined with syngeneic H-2 antigens, suffice for Sensitization of cytolytic T lymphocytes.Abbreviations used in this paper are as follows CMC cell-mediated cytolysis - CTL cytolytic T lymphocyte - LCM lymphocytic choriomeningitis - MHC major histocompatibility complex - MST mean survival time - T cell thymus-derived cell - TCID₅₀ 50 percent tissue culture infective dose     

H-2-associated specificity of virus-immune cytotoxic T cells fromH-2 mutant and wild-type mice: M523 (H-2K ka ) and M505(H-2K bd ) do,M504 (H-2D da ) and M506(H-2K fa ) do not crossreact with wild-typeH-2K orH-2D

B10.A(3R) (H-2K ^b ) mice infected with lymphocytic choriomeningitis virus (LCMV) or vaccinia virus generate cytotoxic T cells capable of specifically lysing virus-infected macrophage target cells fromH-2K ^b mutant mice M505 (H-2K ^bd ), and vice versa. Similarly, virus-immune B10.A(4R) (H-2K ^k ) T cells specifically lyse infected targets from M523 (H-2K ^ka ), and vice versa. In contrast, virus-specific cytotoxic T cells from neither M504 (H-2D ^da ) and B10.A(5R) (H-2D ^d ) nor M506 (H-2K ^fa ) and B10.M(11R) (H-2K ^f ) mutually crossreact at the cytotoxic effector-cell level. As far as tested, the crossreactivity patterns between wild-type and mutantK orD specificities are identical for LCMV- and vaccinia virus-immune spleen cells. Although this finding is no proof for either the altered self nor the dual recognition concept of T-cell recognition, it may be compatible with the latter model.     

Does the major histocompatibility complex serve as a specific receptor for Semliki Forest virus?

Murine F9 and PCC4 teratoma cells do not express H-2 major transplantation antigens according to virus-specific T-lymphocyte cytotoxic or serological assays. However, such cells can be infected with and readily replicate many types of viruses (coxsackie B 3, mouse hepatitis, Sindbis, Semliki Forest [SFV], lymphocytic choriomeningitis, Pichinde, vesicular stomatitis, herpes simplex type 1) to the same extent as do murine F12 teratoma cells and mouse embryo fibroblasts, all of which express the H-2 determinants. In contrast, F9 and PCC4 cells are not productively infected with murine cytomegalovirus, whereas F12 and mouse embryo fibroblast cells are. In addition to replicating in H-2-negative murine teratoma cells, SFV replicates in H-2-negative murine lymphoblastoid cells. The ability of SFV to infect cells without H-2 antigens and then to effect viral antigenic expression in the cells' cytoplasm and on their surface with similar kinetics and in equivalent amounts as cells with H-2 antigens indicates that the H-2 receptor is not needed for SFV infection. Daudi cells, which lack HLA antigens, block the replication of SFV. This occurs at some point after receptor binding, as demonstrated by diminished viral mRNA. In addition, a possible membrane defect precludes viral exit in Daudi cells transfected with SFV infectious RNA. These results indicate that a cell's possession of H-2 antigens is not a requirement for SFV infection and that major histocompatibility complex antigens are not specific receptors for this virus.     

Activation of natural killer (NK) cells in vivo with H-2 and non-H-2 alloantigens

Intraperitoneal inoculation of allogeneic lymphoid cells rapidly activates cytotoxic cells in the peritoneum which are nonadherent and express the NK-1, asialo-GM₁, and Thy-1 antigens. Allogeneic spleen cells were very efficient at activating these natural killer (NK) cells, while allogeneic thymocytes were much less effective. Heat-killed allogeneic cells or sonicates also could augment NK activity. — Incompatibility atH-2K, H-2I-A, orH- 2D readily evoked NK cell activity, whileH-2S- andH-2I-E/C-associated disparities did not. Non-H- 2 differences also stimulated NK activity and augmentation was particularly evident inMls-disparate combinations. Thus, the same alloantigens which efficiently activate T cells also activate NK cells.     

H-2 I alloantigens and recall of memory cytotoxic responses

AQR mice were immunized with H-2K and H-2 I encoded alloantigens presented by (Ax6R)F₁ splenocytes. Spleen cells from these alloimmune mice were subsequently restimulated in vitro with B10.A lymphocytes and/or B10.T(6R) lymphocytes, thus presenting them with the immunizing H-2K and H-2 I alloantigens independently. When stimulated with B10.A lymphocytes, alloimmune lymphocytes develop significant cytotoxicity against the immunizing H-2K target antigens. When stimulated with a similar number of B10.T(6R) spleen cells, alloimmune lymphocytes undergo a prominant proliferative response, but develop little, if any, cytotoxicity against the immunizing H-2 K target antigens. The most efficient restimulation of cytotoxicity occurs when the alloimmune spleen cells are simultaneously restimulated by B10.A and B10.T(6R) lymphocytes. Stimulation with the immunizing H-2 I alloantigens alone is not sufficient for regeneration of detectable cytotoxic responses from alloimmune spleen populations. Stimulation with the immunizing H-2K alloantigens alone appears to be both necessary and sufficient to stimulate alloimmune cytotoxic responses. Although the immunizing H-2 I alloantigens are apparently not required to generate alloimmune cytotoxic responses, they markedly potentiate the cytotoxic responses induced by the immunizing H-2K alloantigens.     

Induction of alloreactive cytotoxic T cells by acute virus infection of mice

Alloreactive cytotoxic T lymphocytes (CTL) distinct from virus-specific CTL and activated natural killer (NK) cells were generated during acute lymphocytic choriomeningitis virus (LCMV) infection of C57BL/6J mice. The alloreactive CTL shared similar antigenic markers (Thy-1.2+, Lyt-2.2+, and asialo GM1-) with the virus-specific CTL that appeared at the same time 7 days postinfection, but had different target specificities. These alloreactive CTL lysed allogeneic but not syngeneic or xenogeneic targets. These were distinct from activated NK cells, which lysed all target cell types, peaked 3 days postinfection, and had a phenotype of asialo GM1+, Thy-1 +/-, Lyt-2.2-. Cold target competition studies indicated that there were several subsets of alloreactive T cells with distinct specificities, and that these alloreactive T cells were not subsets of the virus-specific T cells. Similar types of alloreactive CTL were induced at much lower levels in C3H/St mice. This may indicate that the generation of this "aberrant" T cell activity is under genetic control. Hence, the LCMV infection of C57BL/6J mice induces several cytotoxic effector populations including alloreactive CTL, activated NK cells, and virus-specific CTL. Virus infections therefore have the ability not only to polyclonally stimulate B cells, as previously described, but also to stimulate CTL.     

Antiviral protection by CD8+ versus CD4+ T cells. CD8+ T cells correlating with cytotoxic activity in vitro are more efficient in antivaccinia virus protection than CD4-dependent IL.

T cell-mediated protection against a recombinant vaccinia virus was evaluated in mice with respect to the relative contributions of CTL vs that of T cell-dependent IL and of CD4+ cells. H-2b mice primed with the wildtype of vesicular stomatitis virus serotype Indiana (VSV-IND wt) mount an in vitro measurable cytotoxic response against the nucleoprotein (NP) of VSV-IND and are protected against a challenge infection with a vaccinia-VSV recombinant virus expressing the NP of VSV-IND (vacc-IND-NP). Their protective mechanism was highly susceptible to in vivo depletion of CD8+ T cells, but resistant to CD4+ depletion or treatment with anti-IFN-gamma and anti-TNF-alpha. Surprisingly, also VSV-CTL nonresponder H-2k mice were protected against a challenging infection with vacc-IND-NP when primed with VSV-IND wt. In contrast to the CTL responder H-2b mice, this protection was highly susceptible to CD4+ T cell depletion and to anti-IFN-gamma or anti-TNF-alpha treatment, but resistant to CD8+ T cell depletion. Antibodies were not responsible because they failed to transfer protection; in contrast CD4+ T cells conferred significant protection. VSV-CTL responder H-2b and nonresponder H-2k mice were protected almost equally well against a challenge dose of 10(3) pfu vacc-IND-NP inoculated intracerebrally. However, after intracerebral challenge with 5 x 10(6) pfu vacc-IND-NP, the CTL nonresponder mice died, whereas the CTL responder mice eliminated the virus by day 5. These results collectively show that CD4+ T cell-dependent IL may mediate antiviral protection, but their efficiency is relatively weak compared with CD8-mediated protection correlating with cytotoxic activity in vitro.     

Viral infections induce abundant numbers of senescent CD8 T cells.

Viral infections are often accompanied by extensive proliferation of reactive CD8 T cells. After a defined number of divisions, normal somatic cells enter a nonreplicative stage termed senescence. In the present study we have identified the inhibitory killer cell lectin-like receptor G1 (KLRG1) as a unique marker for replicative senescence of murine CD8 T cells. KLRG1 expression was induced in a substantial portion (30-60%) of CD8 T cells in C57BL/6 mice infected with lymphocytic choriomeningitis virus (LCMV), vesicular stomatitis virus, or vaccinia virus. Similarly, KLRG1 was found on a large fraction of LCMV gp33 peptide-specific TCR-transgenic (tg) effector and memory cells activated in vivo using an adoptive transfer model. Transfer experiments with CFSE-labeled TCR-tg cells into LCMV-infected hosts further indicated that induction of KLRG1 expression required an extensive number of cell divisions. Most importantly, KLRG1(+) TCR-tg effector/memory cells could efficiently lyse target cells and secrete cytokines, but were severely impaired in their ability to proliferate after Ag stimulation. Thus, this study demonstrates that senescent CD8 T cells are induced in abundant numbers during viral infections in vivo.     

A novel approach to visualize polyclonal virus-specific CD8 T cells in vivo

Recent technical breakthroughs in generating soluble MHC class I-peptide tetramers now allow the direct visualization of virus-specific CD8 T cells after infection in vivo. However, this technique requires the knowledge of the immunodominant viral epitopes recognized by T cells. Here, we describe an alternative approach to visualize polyclonal virus-specific CD8 T cells in vivo using a simple adoptive transfer system. In our approach, C57BL/6 (Thy1.2) mice were infected with lymphocytic choriomeningitis virus, vesicular stomatitis virus, or vaccinia virus to induce virus-specific memory T cells. Tracer T cells (2 x 106) from these virus-immune mice were adoptively transferred into nonirradiated (C57BL/6 x B6.PL-Thy-1a)F1 mice. After infection of the F1-recipient mice with the appropriate virus, the transferred cells expanded vigorously, and on day 8 postinfection 60-80% of total CD8 T cells were of donor T cell origin. Under the same conditions memory CD4 T cells gave rise to at least 10 times less cell numbers than memory CD8 T cells. The transfer system described here not only allows to visualize effector and memory CD8 T cells in vivo but also to isolate them for further in vitro characterization without knowing the epitopes recognized by these Ag-specific CD8 T cells.     

Antigen-presenting T cells. II. Clonal responses of alloreactive and virus-specific self-restricted human cytotoxic T cell responses stimulated by T lymphoblasts

The capacity of various stimulator cell types to present alloantigens or viral antigens to resting human CD8+ cytotoxic lymphocyte precursors (CLP) was analyzed in a limiting dilution culture system. Cell sorter-separated T lymphoblasts of both CD4+ and CD8+ phenotypes but not resting T cells were found to efficiently stimulate the clonal development of allogeneic CD8+ CLP. Thus, 5000 CD4+ T lymphoblasts activated as many (one out of 200 to one out of 300) allogeneic CLP as 50,000 peripheral blood mononuclear stimulator cells. This potent stimulator activity was found in CD4+ and CD8+ T lymphoblasts activated by mitogen, anti-T3 monoclonal antibody, or mixed leukocyte reactions. Cytotoxic T cells generated in this system were highly specific for HLA class I antigens. Furthermore, T lymphoblasts infected with mumps virus efficiently induced development of autologous CLP into CTL clones that were virus specific and self-HLA restricted, as shown by split-well analysis. The possible in vivo significance of antigen-presenting T lymphoblasts is discussed.