Effect of Toxoplasma infection on the sensitivity of mouse thymocytes to natural killer cells.

Thymocytes from mice 2 weeks after infection with Toxoplasma gondii resisted natural killer (NK) cell-mediated cytolysis in contrast to the high sensitivity of normal mouse thymocytes. The infected mouse thymocytes also failed to form conjugates with effector cells and to compete for cytolysis of NK sensitive targets. These effects were mediated, at least in part, by interferon-gamma because normal thymocytes became NK insensitive after incubation in the infected mouse serum which contained significant amount of interferon-gamma, and pH 2 treatment of the serum abolished the effect. An alternate possibility for the reduced NK sensitivity of the infected mouse thymocytes was the elimination of NK-sensitive cells from the thymus, since histopathological studies showed marked atrophy and clearance of NK-sensitive thymocytes in the cortex of thymuses of infected mice. Although T. gondii induced augmentation followed by suppression of the host splenic NK activity, it seems unlikely that this altered NK activity was responsible for the lowered NK sensitivity of the thymocytes.     

Surface markers on natural killer cells of the mouse.

Rabbit antiserum against mouse brain tissue (anti-brain-associated T cell antigen, anti-BAT) was capable of killing splenic natural killer (NK) cells of CBA/J, BALB/c, C 57 Bl/6J, C 3 H/He and nude mice, which were detected with Molony virus-induced lymphoma (YAC-1) and radiation-induced leukemia (RL male 1) cells as targets. The same antiserum abolished T cell functions, e.g. carrier-specific helper function and the responsiveness to concanavalin A, but not B cell functions, e.g. immunological memory for the secondary antibody response and the responsiveness to lipopolysaccharide. After absorption of the anti-BAT with thymocytes, the ability to kill T cells was completely abrogated, leaving the activity to kill NK cells intact. No other heterologous and isologous antisera, i.e. rabbit anti-mouse thymocyte antiserum, goat antiserum against antigens shared by thymus and B cells, anti-Thy-1.2 and anti-Ia antisera, could eliminate NK function regardless of their definite reactivity against T or B cells. The results indicate that the absorbed anti-BAT can distinguish NK cells from other known subsets of T and B cells.     

Human fetal thymus and bone marrow contain target cells for natural killer cells

Previous studies in the mouse natural killer (NK) system have indicated that NK cells may be involved in lysing normal, primary hematopoietic tissues. In the present report, this was analyzed in the human NK system using fetal bone marrow (BM) cells and thymocytes as well as adult BM cells from healthy donors as target cells in a conventional 6 to 12-h ⁵¹Cr-release assay. Adult BM cells showed low but significant levels of sensitivity, which could be increased by using interferon (IFN)-activated peripheral blood leukocytes (PBL) as effector cells. BM cells from 16 to 19-week-old fetuses consistently showed higher sensitivity for lysis than adult BM, and also fetal thymocytes proved very sensitive for lysis, in contrast to what has previously been reported for adult thymocytes. When used as effector cells against K562 targets, adult BM showed a clear lytic activity which could be further activated with IFN. In contrast, fetal BM was totally NK-inactive also after IFN activation. Among healthy adult donors, autologous BM was lysed as efficiently as allogeneic BM, and when different NK cell donors were used, the same classification of these as NK high or low reactive cells was seen regardless of the source of BM targets. IFN could augment lysis in autologous as well as in allogeneic effector BM target combinations. IFN could also protect adult BM cells from NK lysis, but no protection was seen with fetal BM cells. The highest NK activity against BM targets was seen among nylon-nonadherent, E rosette-negative PBL and, therefore, the effector cell seemed to be of the same nature as that active against continuous cell lines. These results show that in the human NK system, NK cells can lyse normal BM cells and thymocytes. The higher sensitivity expressed by fetal BM cells and thymocytes as compared to adult cells may suggest an increased frequency of a particular primitive NK-sensitive cell type in these tissues, a finding which is in line with results seen in the mouse NK system.     

Self-generating density gradients of Percoll provide a simple and rapid method that consistently enriches natural killer cells

The separation or enrichment of natural killer (NK) cells from the heterogeneous cell populations in murine spleen or bone marrow is a vital step for the study of NK cells. We report in this study a simple and rapid method for the enrichment of NK cells from B cell-depleted spleen cells, using a self-generating density gradient of polyvinyl pyrrolidone-coated silica (Percoll). Nylon wool-passed spleen cells are suspended in Percoll that is isotonic and isosmotic with mouse blood at a density of 1.087 g/ml and ultracentrifuged at 30 000 × g for 10 min. This method consistently enriches for NK cell cytotoxic activity, in spleen cells of both unstimulated and interferon-stimulated mice, as measured in the chromium release assay. There is a concomitant enrichment for cells bearing the NK marker asialo GM-1 and depletion of L3T4 or Lyt-2-bearing T cells. In contrast to discontinuous, step-wise gradients, the self-generating Percoll gradient, which relies on the intrinsic property of Percoll to form a continuous density gradient, appears to provide the cells with a physiological environment both before and during the centrifugation step.     

In vitro binding of natural killer cells to Cryptococcus neoformans targets.

Nylon wool-nonadherent splenic cells from 7- to 8-week-old CBA mice were further fractionated on discontinuous Percoll gradients. Enrichment of natural killer (NK) cells in Percoll fractions 1 and 2 was confirmed by morphological examination, by immunofluorescent staining, and by assessing the cytolytic activity of each Percoll cell fraction against YAC-1 targets in the 4-h51Cr release assay. Cells isolated from each Percoll fraction were tested for growth-inhibitory activity against Cryptococcus neoformans, a pathogenic yeastlike organism, by using an in vitro 18-h growth inhibition assay. The results showed that NK cell enrichment was concomitant with enrichment of anti-Cryptococcus activity in Percoll fractions 1 and 2. Cells from NK cell-rich fractions formed conjugates with the mycotic targets similar to the conjugates reported in NK cell-tumor systems. In addition, the percentage of effector cell-Cryptococcus conjugates was directly proportional to the level of the C. neoformans growth-inhibitory activity of the effector cells used. Scanning electron microscopy of the effector cell-Cryptococcus conjugates showed direct contact between the effector cells and the cryptococcal targets. An immunolabeling method combined with scanning electron microscopy was used to demonstrate that the effector cells attached to C. neoformans were asialo GM1 positive and, therefore, had NK cell characteristics.     

Enrichment of mouse splenic natural killer cells using discontinuous polyvinylpyrrolidone silica (Percoll) gradients.

A simple and rapid method is reported here for enriching murine spleen cells with natural killer (NK) function as assessed by short-term cytolysis assay of 51Cr-labelled YAC-1 lymphoma target cells. The established method used for the enrichment of NK reactive cells, including large granular lymphocytes (LGL) from human and rat peripheral blood lymphocytes, does not substantially enrich for mouse splenic NK cell activity. A reproducible procedure for enriching mouse splenic NK cells has been developed using a four- or five-step discontinuous Percoll gradient in the density range of 1.062 g/ml (top) to 1.092 g/ml (bottom) and osmolarity (310-340 mOsm/kg) nearer to that of mouse blood and tissue. A four- to 25-fold (usually about nine-fold) increase in NK cell activity, consisting of 50-100% of the recovered lytic unit activity, is found in which are the cells forming band 3, approximately 10% of the recovered cell number. This cell population with enriched NK cell activity has a characteristic density less than or equal to 1.077 g/ml, but more than 1.070 g/ml when centrifuged under appropriate conditions. Similar enrichment was obtained with a four-step gradient at an uniform osmolarity of 320 mOsm/kg throughout. Although the lymphocytes in band 3 show relatively little heterogeneity in appearance, only a minor population of the cells contain granules.     

Toxoplasma-induced activities of peritoneal and spleen natural killer cells from beige mice against thymocytes and YAC-1 lymphoma targets.

Homozygous (bg/bg) and heterozygous (bg/+) beige mice were infected with Toxoplasma gondii, and splenic and peritoneal natural killer (NK) cell activities were assayed against YAC-1 lymphoma (NK-YAC) and thymocyte (NK-THY) targets. Although uninfected bg/bg mice were devoid of NK-YAC activity when compared with bg/+ mice, NK-THY activity was at a completely normal level. Both effector cells showed NK-1.2+ Thy-1.2 +/- asialo GM1+ asialo GM2+ phenotype. T. gondii infection induced a marked augmentation in splenic NK-YAC activity of bg/+ mice, whereas a slight increase was shown in the bg/bg mouse spleen cells. On the other hand, the infection did not change the splenic NK-THY activity of either strain of mice. An increased expression of Thy-1.2 antigen was shown on both NK-THY and NK-YAC effector cells from the infected mouse spleen. The T. gondii-induced augmentation was dramatic in the peritoneal cavity of the both mice. The activated peritoneal NK cells were of the NK-1.2- Thy-1.2+ asialo GM1 +/- asialo GM2+ phenotype and were considered to be generated from functionally inactive peritoneal cells. Splenic effector cells obtained from the infected mice were selectively depleted with target cell monolayer, whereas peritoneal cells from the infected mice were strongly absorbed by the target monolayers without selectivity. A weak but significant interferon (IFN) titer was detected in the peritoneum, but not in the spleen, of the infected mice. Most of the IFN titer was acid labile. Treatment with anti-IFN-alpha/beta resulted in partial decline of both NK and IFN activities of bg/bg mice, but not bg/+ mice. Thus, involvement of both IFN-alpha/beta and IFN-gamma in the generation of peritoneal NK cells and IFN-independent augmentation of splenic NK cells in toxoplasmosis were suggested.     

Monoclonal antibodies (G1.4, D7.5) induce secretion of lytic factors from natural killer and T lymphocytes

Monoclonal antibodies were made by standard procedures using mononuclear cells isolated from hamster lung tissue to immunize the mice. The Moabs (G1.4, D7.5), secreted by the selected hybridomas, were isotyped as IgG2a, Kappa and were lytic in the presence of complement for 20% lymphocytes from lymph nodes or lung tissue; 40-50% of lymphocytes from spleen or bone marrow; and 70-80% of thymocytes or nylon wool nonadherent lung or lymph node lymphocytes. Functional studies showed that prior incubation of either Moab with enriched natural killer (NK) and T lymphocyte subpopulations did not interfere with effector:target conjugate formation but did block the ability of the effector cells to lyse NK-sensitive targets. We conclude that the Moabs bind to a molecule(s) on NK and T lymphocytes. Further, the preincubation medium acquired the capacity to lyse preferentially NK-sensitive targets but not NK-insensitive targets. Interestingly, the Moabs were able to induce secretion of lytic factors from thymocytes and to induce resting peripheral T lymphocytes to acquire NK activity and release lytic factors. The ability of the Moab to signal release of lytic factors was retained even when monovalent Fab fragments were used; therefore, the Moabs initiate secretion by mechanisms that do not apparently involve crosslinking or Fc receptor interaction. G1.4 and D7.5 Moabs were functionally and antigenically crossreactive with human, rat, and hamster lymphocytes but not with mouse lymphocytes. Biochemical and immunochemical analysis showed that the G1.4 and D7.5 antigen on thymus cells and lung mononuclear cells has an apparent molecular weight of 27 kD. Although it is unlikely that there is an universal mechanism for secretion used by cells in general, it is interesting that Moab treatment of three different subpopulation of lymphocytes induced secretion of lytic factors and suggests that under physiologic conditions a common molecule might participate in induction of the secretion pathway. These studies lead to the hypotheses that 1) NK and T lymphocytes share a cell surface molecule that participates in signals that initiate secretion and that 2) the expression of the G1.4, D7.5 antigens precedes thymic differentiation.     

Dissociation of natural killing and luminol-dependent chemiluminescence

A long lasting luminol-dependent chemiluminescence was seen when mouse NK cell preparations or human NK cell preparations from Percoll gradients were mixed with NK susceptible targets. This CL could be abolished by extensive removal of adherent cells though normal NK cell activity remained. In addition, the NK cell line HY 3-Ag3 did not show any CL; although it expressed a very strong cytolytic activity. Thus, we conclude that there are two cells reacting with NK susceptible targets. First, the NK cell whose cytolytic activity does not necessarily depend on the formation of oxygen metabolites detectable by CL, and second, a more adherent cell population, found in NK cell preparations obtained after Percoll gradients, that reacts with NK-sensitive targets and leads to luminol-dependent CL. The observed CL paralleled the cytotoxicity measured by 51Cr release. The time course of the CL signal was similar in human and mouse. The maximal CL-signal obtained was about 10(5) cpm at 37 degrees C when 10(6) human effector cells were used at a ratio of 5:1 effector to target cells. The CL was shown to be highly temperature dependent.     

Natural cytotoxic T cells responsible for anti-CD3-induced cytotoxicity in mice.

T cells obtained from normal mouse spleen cells showed significant cytotoxic activity against Fc receptor positive tumor cells in the presence of anti-CD3 monoclonal antibody (mAb). This activity was designated as natural cytotoxic T cell (NCT) activity and compared with natural killer (NK) activity. Considerable levels of NCT activity were detected in mouse strains with both high and low NK activity. NCT cells were distributed in both lower and higher density fractions of Percoll discontinuous density gradients, while NK cells were enriched in the lower density fraction of Percoll gradients. Moreover, NCT activity was resistant to in vivo anti-asialo GM1 treatment, in contrast to NK cells. These results indicate that NCT cells, which have different characteristics from NK cells, are present in normal, nonimmunized mouse spleen cells. Unexpectedly, CD4+ T cells sorted from normal mouse spleen T cells revealed significant NCT activity, as did CD8+ T cells. It was also demonstrated that NCT cells require the LFA-1 molecule to lyse tumor cells in the presence of anti-CD3 mAb.     

Higher level expression of lymphocyte function-associated antigen-1 (LFA-1) on in vivo natural killer cells

Normal mouse spleen cells express low levels of lymphocyte function-associated antigen-1 (LFA-1) as well as other lymphoid cells. However, fractionation of spleen cells with Percoll discontinuous gradients resulted in the appearance of lymphocytes expressing high levels of LFA-1 molecule (LFA-1 high lymphocytes) in parallel with the enrichment of natural killer (NK) activity. Lower density spleen cells (fractions 1 and 2) expressed higher level of LFA-1 antigen than unfractionated spleen cells and showed a higher NK activity. In contrast, higher density spleen cells (fractions 3 and 4) expressed lower levels of LFA-1 antigen and revealed lower NK activity. LFA-1 high lymphocytes possessed a high level of asialo GM1, which was the cell surface marker for NK cells. Moreover, sorting of LFA-1 high lymphocytes from spleen cells caused a great enrichment of NK cells. These results demonstrated that in vivo NK cells expressed higher levels of LFA-1 molecule, which was an important adhesion molecule for NK cell-mediated cytotoxicity.     

The chemiluminescence response of human natural killer cells. II. Association of a decreased response with low natural killer activity

Low natural killer (NK) responders selected from a panel of 600 normal, healthy volunteers exhibited 5- to 10-fold less cytotoxicity against the human erythroleukemic cell line K562, compared with high NK responders. Antibody-dependent cellular cytotoxicity against tumor cells, which is mediated by similar or identical effectors, was also depressed in low NK donors whereas lectin-dependent T cell killing and monocyte-mediated cytolysis of tumor cells was normal. Low NK donors exhibited normal frequencies of cells expressing the HNK-1 marker of human NK cells and highly enriched NK fractions were not impaired in their ability to recognize and bind to NK-sensitive target cells. Interferon partially activated low responder NK cells but did not restore the response to normal levels. The burst of chemiluminescence that is generated by NK cells within seconds of target cell contact was markedly impaired in low NK responder donors. We have previously shown that chemiluminescence detects reactive oxygen intermediates which are necessary but not sufficient for the activation of the NK cytolytic pathway.     

Human thymocyte development in mouse organ cultures

A novel system to study human thymocyte development is described in which embryonic mouse thymic rudiments are seeded with human precursor cells in vitro. In these cultures human thymocytes proliferate extensively (greater than 20-fold increase in cell number) and mature, as evidenced by the accumulation of double and single positive (CD4+ and/or CD8+) cells. Data presented here suggest that the survival and ordered development of the mature human thymocytes in chimeric thymuses is dependent on human stromal elements. Immature CD4-CD8- human thymocytes failed to colonize or minimally recolonized mouse thymic lobes unless provided with high density (greater than 1.077 g/ml) human thymic cell fractions. These fractions contain multicellular complexes of epithelial/nurse cells, thymocytes, and dendritic cells/macrophages which dramatically enhanced the recolonizing capacity of purified CD4-CD8- thymocytes. The chimeric organ culture system described here provides not only a new approach for studying human T cell ontogeny but also a direct means for the future dissection of stromal interactions necessary for successful transition of precursor cells (CD4-CD8-) to immature double positive (CD4+CD8+) and mature single positive cells (CD4+ or CD8+) in the thymus.     

Lectin-binding characteristics of human natural killer cells.

Human natural killer (NK) cells separated initially by density centrifugation of lymphocytes (E+) forming rosettes with sheep red blood cells (SRBC), were further fractionated on gradients of bovine serum albumin (BSA). Low density fractions contained effector cells which displayed high cytotoxicity against the NK-sensitive erythroleukaemic cell line, K562. These low density cells, which expressed receptors for Fc and the monoclonal antibody OKMI, showed enhanced cytotoxicity when treated with lymphoblastoid interferon (IFN-alpha). They also showed an increased response to phytomitogen in comparison with unseparated cells or those recovered from high density fractions. Two lymphocyte subsets one of high and one of low lectin binding capacity were identified in the E+ populations by their reactivity with Lens culinaris agglutinin (LCA). High LCA binding was observed only in low density fractions and was associated with a marked enrichment of NK activity. This property was used to separate the NK active population in E+ cells by fluorescence-activated cell sorting (FACS). These data add a new dimension to the cell surface properties of human NK cells and suggest the presence of LCA-reactive glycoproteins which are either enriched in, or uniquely associated with, cells of the NK subset. The experiments indicate that lectins can serve as useful probes of lymphocyte function and provide the basis for effective cell sorting.     

In vitro effects of natural killer cells against Paracoccidioides brasiliensis yeast phase.

Recently, data have been reported suggesting natural killer (NK) cells may function in natural resistance against a fungus, Cryptococcus neoformans. The primary objective of this study was to examine the reactivity of murine splenic cells against another fungus, Paracoccidioides brasiliensis. Levels of NK activity in effector cell pools were varied by: (i) removing nylon wool-adherent cells, (ii) fractionating splenic cells on Percoll discontinuous gradients, (iii) using old and young effector cell donor mice, (iv) using donors from different strains, and (v) pretreating donors with NK-augmenting and -depressing agents. The various effector cell pools were simultaneously used in the 4-h 51Cr release assay with YAC-1 targets to determine the NK reactivity and in the in vitro growth inhibition assay against P. brasiliensis yeast phase targets. In each case, the level of NK reactivity correlated with the ability of the effector cells to inhibit the in vitro growth of P. brasiliensis. NK activity and P. brasiliensis growth-inhibiting ability could be augmented by fractionation of splenic cells through nylon wool or Percoll gradients. The effector cells responsible for the NK activity and P. brasiliensis growth inhibition were characterized as being nylon wool nonadherent, being found in the low-density fractions from Percoll discontinuous gradients, and having no detectable Thy-1 antigen or immunoglobulin but having asialo GM1 on their surface. These data support the contention that NK or NK-like cells are responsible for limiting the in vitro growth of P. brasiliensis.     

Development of natural killer cells in human thymocyte culture: regulation by accessory cells

In vitro culture of human thymocytes resulted in the development of cells with natural killer (NK) activity and the acquisition of a pan-NK antigen (NKH1) by a large number of thymocytes. The ability to kill the NK-sensitive target, K562, was restricted to thymocytes expressing the NKH1 antigen. All NKH1+ thymocytes displayed a mature T cell phenotype, T3+T11+T8+T4-. Both the acquisition of NK activity and the development of cells with the NKH1+ phenotype could be suppressed by culturing thymocytes in the presence of adherent mononuclear cells. These results suggest that adherent accessory cells have the ability to regulate the development of T cell lineage NK cells.     

Relationship between large granular lymphocytes and NK-1.2+ cells from normal and poly(inosinic:cytidylic acid) (poly(I:C]-treated mice

The present work analyzes the relationship between large granular lymphocytes (LGL), NK-1.2+ cells, and natural killer (NK) activity of C3H/HeN mice. Different hematic cell fractions were obtained according to their nylon-wool adherence and density on Percoll gradients. NK-1.2+ cells (8% of nucleated cells) were more numerous than LGL (3% of nucleated cells) in the input blood population. Eighty-five percent of LGL were recovered from the sorted NK-1.2+ cell fraction. After incubation on nylon-wool column, 63% of LGL and 36% of NK-1.2+ were eluted in the nonadherent fraction. Eighteen percent of NK-1.2+ cells were recovered from the most adherent elutable cell fraction. After the discontinuous Percoll gradient most LGL were present in the low-density fractions while 20% of NK-1.2+ cells were recovered from the highest-density fraction. NK activity was significant both in the nylon-wool-nonadherent and -adherent fractions. After the Percoll gradient most NK activity was present in the low-density fractions. In the present experimental conditions treatment poly(inosinic:cytidylic acid) (poly(I:C] did not increase the numbers of LGL and NK-1.2+ cells either in the blood or in the spleen. However it increased significantly the NK activity of the input cell populations and of the nonadherent and low-density fractions. Similarly, exposure of specific pathogen-free (SPF) mice to non-SPF conditions stimulated NK cytotoxicity but did not alter the percentage of LGL in the blood or in the spleen. Poly(I:C) treatment induced a shift of LGL and NK-1.2+ cells toward the low-density fractions. In poly(I:C)-treated mice images of granule secretion from LGL were detected. Taken together, the present results indicate that LGL and NK-1.2+ cell populations do not totally overlap. Moreover subpopulations of LGL and NK-1.2+ cells can differ in NK activity, morphology, density, adherence to nylon wool, and response to poly(I:C).     

Interleukin-2-induced activation of natural killer activity in spleen cells from old and young mice.

Generation of natural killer (NK) activity in response to a partially purified preparation of rat interleukin-2 (IL-2) was compared in spleen cells derived from young (8-10 weeks old) and old (greater than 2 years old) female C57BL/6 mice. Significant NK activation was observed in both young and old mouse spleen cells incubated with 100 U IL-2/ml for 1-4 days, but the levels of cytotoxic activity generated in old mouse spleen cells was always lower than those of similarly treated young mouse spleen cells. Differences in IL-2-induced NK activation in old and young mouse spleen cells was obtained irrespective of the concentration of IL-2 used (25-400 U/ml). Quantitative comparisons indicated that old spleen cells activated by 3 day incubation with IL-2 acquired about two-fold higher NK activity than fresh young mouse spleen cells but still had only one-fourth of the levels of NK activity attained by IL-2-activated young mouse spleen cells. Cytotoxic activity of IL-2-activated young or old mouse spleen cells were totally abrogated by anti-asialo GM-1 antiserum + C but not by anti-Ly-2 + C treatment, indicating that the activated cytotoxic cells fell in the NK cell category. An analysis of NK precursor (NK-p) frequency by limiting dilution assay indicated that the NK-p frequency was about 4-fold higher in young as compared to old mouse spleen cells. The level of cytotoxic activity attained per NK-p cell was not significantly different for NK-p cells of old or young mice.     

Syngeneic pregnancy induces overexpression of natural killer T cells in maternal liver

Conditions such as stress, infection, autoimmune disease, etc. elevate the number and function of extrathymic T cells that are generated mainly in the liver. As primitive, self-reactive clones of T cells that coexpress receptors of the natural killer (NK) lineage, they mediate cytotoxicity against altered self, malignant and infected cells and have the unique potential to rapidly secrete large amount of T helper 1 (Th1) or Th2 cytokines. To elucidate whether some of these changes occur even during the syngeneic pregnancy, we made phenotypic and functional characterization of mononuclear lymphatic cells (MNLCs) isolated from the liver and spleen of pregnant C57BL/6 mice, testing their cytotoxicity against syngeneic thymocytes as well as against NK- and lymphokine-activated killer (LAK)-sensitive targets. The data have shown that on the sixteenth day of syngeneic pregnancy TCRint, NK1.1+ and IL-2Rbeta+ cells were accumulated in the liver, while the quantities of CD4+ and CD8+ T cells and total number classical NK (NK1.1+CD3- or IL-2Rbeta+CD3-) cells were increased in the spleen. Pregnancy-activated hepatic and splenic MNLCs were more cytotoxic against syngeneic thymocytes, YAC-1 and P815 targets, suggesting that the maternal liver is a main producer of autoreactive NKT clones, which subsequently augment NK- and LAK cell-mediated cytotoxicity in the liver and spleen.     

Rat natural killer cell and cytotoxic T cell lysis of H-2-negative murine embryonal carcinoma cells

H-2-lacking murine embryonal carcinoma (EC) cells have been proposed as universal targets for natural killer (NK) effectors from different species because their killing appeared to be uncomplicated by potential T cell effector mechanisms (Stern, P. L. et al., Int. J. Cancer 1981. 27:679). While some previous studies had shown that murine cytotoxic T cells were unable to lyse EC cells, rat T killers are shown here to be active against these targets and to be distinguishable from NK cells. Percoll density fractionation of rat peripheral blood lymphocytes enriches in parallel for NK-mediated lysis of both EC or YAC target cells. These NK cells unlike T cells, do not mediate lectin-dependent and cell-mediated cytotoxicity (LDCC) of NK-insensitive target cells. This procedure is thought to reveal the total cytolytic potential of stimulated T cell populations, regardless of specificity. In contrast to previous results with mice, we found that allogeneically primed rat cytotoxic T cells can kill murine EC cells in LDCC and, further, that rat cytotoxic T cells, generated by stimulation with mouse spleen cells in vitro, can lyse murine EC cells directly. This demonstration of T cell lysis of EC cells suggests that either there is a novel mechanism of lysis operating without requirement for major histocompatibility complex (MHC) structures, or EC cells express some hitherto unidentified MHC-like structures on their cell surface.