Hybrid resistance to EL-4 lymphoma cells. I. Characterization of natural killer cells that lyse EL-4 cells and their distinction from marrow-dependent natural killer cells

Correlation of growth of tumours in NC-cell-depleted mice with NC- and NK-cell-mediated lysis in vitro

The anti-tumour surveillance activity of natural cytotoxic (NC) cells was studied in vivo using the transplantable tumours WEHI-164 fibrosarcoma, MPC-II plasmacytoma, WEHI-7 T-lymphoma, B16 melanoma and EL-4 thymoma in syngeneic and semi-allogeneic mice. Experimentally, mice were treated with the anti-NC-I.I monoclonal antibody (MAb) IC4 to abrogate splenic NC activity. This was followed by s.c. inoculation of MTD100 doses of the tumours. Comparison of the diameters of the tumours in the anti-NC-I.I-treated mice with control mice using non-parametric statistics showed significantly faster growth of WEHI-164 (p less than 0.01), MPC-II (p less than 0.05) and WEHI-7 (p less than 0.05) when the mean tumour diameters were less than 15 mm in the anti-NC-I.I-treated mice. Significantly faster growth was also observed in anti-NC-I.I-treated mice with the B16 tumour (p less than 0.05), but at a later stage of growth, when the tumour diameter was greater than 15 mm. In vitro, WEHI-164, MPC-II and WEHI-7 were shown to be predominantly sensitive to lysis by mouse splenic NC cells, while B16 was predominantly lysed by splenic natural-killer (NK) cells. Anti-NC-I.I treatment of mice did not affect the growth of EL-4 in vivo and in vitro experiments with anti-NK-I.I and anti-NC-I.I MAb indicated that this tumour was lysed by sub-sets of NK and NC cells distinct from those which lysed the other tumours. We conclude that, in mice at least, NC cells have an in vivo role in controlling the growth of some transplantable tumours, and this correlates with the in vitro NC cell lysis of these same tumours.

Murine trophoblast resists cell-mediated lysis. II. Resistance to natural cell-mediated cytotoxicity

The susceptibility of murine trophoblast cells to natural cell-mediated cytotoxicity has been assessed. Primary short-term cultures of murine trophoblast cells isolated from 14-day placentas were found to be resistant to endogenous and interferon-activated natural killer (NK) cells and natural cytotoxic cells. That the relevant target structures are expressed on the surface of trophoblast cells and accessible to the effectors was demonstrated by their ability to inhibit the lysis of NK-sensitive target cells (YAC-1) in a dose-dependent manner. The lytic resistance of trophoblast cells was unaffected by neuraminidase treatment, inhibition of protein synthesis, or extending the assay time to 12 hr. Moreover, trophoblast cells were resistant to antibody-dependent cell-mediated cytotoxicity when coated with an alloantibody capable of mediating their lysis in the presence of heterologous complement. Neither the preincubation of effector cells in concentrated trophoblast culture supernatants nor the direct exposure of effectors to monolayers of trophoblast cells inhibited their NK lytic activity, indicating that the secretion of a suppressive factor or the direct inactivation of the NK cells was not responsible for the observed resistance to lysis. These observations, together with previous results showing the resistance of trophoblast to cytotoxic T cell-mediated lysis, reveal that murine trophoblast cells possess a resistance mechanism against several forms of cell-mediated lysis. This feature of trophoblast cells at the maternal-fetal interface is likely to play an important role in protecting the fetoplacental allograft from immune rejection.

The effect of bearing tumours on the ability of mice to reject bone marrow transplants

To examine the possible relationship between the cells mediating resistance to tumour cells and those mediating rejection of foreign bone marrow transplants (BMT), the effect of tumour-bearing on BMT rejection has been measured by means of the spleen colony assay. Moderate doses (less than 5.0 X 10(6] of bone marrow cells from DBA/2 strain mice, which normally produce few haemopoietic spleen colonies in gamma-irradiated (950R) CBA/J strain mice, gave numerous (confluent) colonies when given soon after injection of Ehrlich ascites tumour (EAT) cells. Transfusion of [3H]UdR-labelled DBA/2 bone marrow cells demonstrated that the increased spleen colony formation in tumour-bearing mice was not due simply to changes in the total number of injected cells homing to the spleen. Injection of EAT ascites fluid alone, given to CBA/J mice before 950R + BMT, transiently reduced spleen colony development, the effect being more marked when fluid from older tumors was used. Supernatant fluid from EAT cells grown in vitro also depressed growth of BMT in vivo. The results reveal two processes in progress in mice bearing an ascites tumour: (1) an early reduction in the natural resistance to BMT allowing successful grafting and spleen colony formation, and (2) a progressive production by the tumour cells of short-acting soluble factors tending to suppress the proliferation of colony forming bone marrow cells in the transplant. The effect of the tumour-bearing state in weakening the natural resistance to foreign BMT strongly suggests that both tumour and foreign marrow graft resistance are mediated by the same or closely related effector cells.

In vivo natural antitumor resistance against murine EL-4 lymphoma cells in lethally irradiated syngeneic C57Bl/6 mice

Natural resistance has been detected in lethally irradiated C57Bl/6 (B6) mice inoculated intravenously with the ascites form of a syngeneic B6 leukemia. EL-4 cells were injected into lethally irradiated (800 R) B6 mice and tumor cell proliferation was evaluated by 125IUdR uptake in different organs 4 days after the challenge. Differential growth of lymphoma cells was observed when young mice were injected as compared with older mice and when mice were treated with agents known to interfere with natural resistance (e.g., poly(I:C), FLV-P, carrageenan, cyclophosphamide, high doses of irradiated cells). Similar results were obtained by measuring rapid clearance of 125IUdR-labeled EL-4 cells from lungs of intact B6 mice. In vivo cold competition studies, employing EL-4 and several other tumor lines of the same or different haplotype, showed that only EL-4 and RBL-5 cells were capable of inhibiting syngeneic resistance against EL-4 tumor. On the contrary, YAC-1 lymphoma cells, the most susceptible target to natural killer-mediated cytotoxicity in vitro, did not compete. These results suggest that EL-4 cells express membrane determinants not detectable on normal H-2b parental bone marrow cells and are susceptible to natural resistance against hemopoietic tumor cells in lethally irradiated syngeneic B6 mice.

Influence of interferon on the functional expression of natural killer target structures of murine lymphoma cells

Murine lymphoma cells (YAC-1), induced by Moloney leukemia virus, nontreated (YAC) or pretreated in vitro with interferon (YAC-IF), were tested for their susceptibility to natural killer (NK)-mediated cytolysis. In line with previous reports YAC-IF were less susceptible to NK lysis than YAC cells. In cold competition assay, YAC-IF inhibited cytotoxicity to a lesser extent than YAC lymphoma when labeled target YAC cells were used. However, when radioactive YAC-IF cells were used as targets, cold competition attained with both YAC and YAC-IF was essentially the same. Furthermore, effector splenocytes, depleted of NK effector cells through immunoabsorption on YAC monolayer, were inactive against both YAC and YAC-IF targets. On the other hand, effector lymphocytes, absorbed on YAC-IF monolayer, retained NK activity against YAC cells but not against YAC-IF targets. These results are compatible with the hypothesis that interferon (IF) modulates negatively a subset of "interferon-susceptible" (IFS) NK target structure(s) (TS) of YAC cells, which would then express membrane determinants not functionally present on YAC-IF cells. On the other hand YAC and YAC-IF cells share "interferon-resistant" (IFR) TS not affected by pretreatment with IF. In order to test whether IFS X TS and IFR X TS are present on the same cell or clonally distributed, YAC cells were cloned and tested for NK susceptibility following IF pretreatment. The results did not support the hypothesis of a clonal distribution of both IFS X TS and IFR X TS since IF pretreatment of all clones, obtained by limiting dilution, resulted in a net impairment of target susceptibility to NK effector cells.

Separation of human natural killer cell subpopulations differentially responsive to interferon potentiation

Subsets of human natural killer (NK) cells were identified that differed in the capacity to be activated by interferon (IFN) or the IFN-inducer, polyriboinosinic:polyribocytidylic acid [poly(I):poly(C)]. These subsets, which represented effectors of both spontaneous and antibody-dependent cellular cytotoxicity, were physically separable on the basis of cell buoyant density changes induced by exposure of lymphocytes to hyperosmolar Ficoll-Hypaque solutions or by centrifugation of lymphocytes through hyperosmolar (350 mOs/kg) Percoll gradients. Hyperosmolar conditions per se altered neither cell viability, NK cell cytolytic activity, nor the capacity of NK cells in unseparated lymphocyte preparations to be activated by IFN. IFN-unresponsive NK cells, separated by centrifugation through a 350 mOs/kg Percoll layer of 1.069-1.070 g/cm3 specific density, constituted 20 +/- 4% of all active NK cells identified at the single cell level and, per active NK cell, killed comparably to unstimulated IFN-responsive NK cells in 51Cr release assays. Thus, the IFN-unresponsive phenotype was probably not attributable to NK cells that were in an activated state prior to IFN treatment. Surface marker analysis of active NK cells at the single cell level identified comparable proportions in each subfraction to be of the OKM1+, OKT8+, or OKT11+ phenotypes and few, if any, in either subfraction to be of the OKT3+ phenotype. The human IFN-unresponsive NK cell phenotype, in contrast to the corresponding phenotype in the mouse, was therefore not linked to expression of T-cell-associated membrane differentiation antigens.

Mediation of mouse natural cytotoxic activity by tumour necrosis factor

Natural cell-mediated cytotoxic activity in the mouse has been associated with two types of effector cells, the natural killer (NK) cell and the natural cytotoxic (NC) cell, which seem to differ with regard to their patterns of target selectivity, cell surface characteristics and susceptibility to regulatory factors. During studies on the mechanism of action of cytotoxic molecules, it became evident that WEHI-164, the prototype NC target cell, was highly susceptible to direct lysis by both human and mouse recombinant tumour necrosis factor (TNF). Here we show that NC, but not NK activity mediated by normal splenocytes, is abrogated by rabbit antibodies to recombinant and natural TNF, respectively. Thus, the cell-mediated activity defined as NC is due to release of TNF by normal spleen cells and does not represent a unique natural effector mechanism.

Antimicrobial activity of various immunomodulators: independence from normal levels of circulating monocytes and natural killer cells

The effects of 89Sr treatment on the natural host resistance of CD-1 mice and the enhancement of resistance by immunomodulators to infection with Listeria monocytogenes or herpes simplex virus type 2 (HSV-2) were determined. In the CD-1 mouse, single-dose treatment with 89Sr caused a profound decrease in the number of circulating monocytes (Mo), lymphocytes, and polymorphonuclear leukocytes (PMN) within 1 week. There was also marked functional impairment of the Mo inflammatory response, as well as markedly decreased spontaneous and activatable cytotoxicity by splenic natural killer (NK) cells. Despite this profound cellular suppression, there was no significant change in natural resistance of CD-1 mice to L. monocytogenes or HSV-2 infection. Furthermore, prophylactic treatment of mice with the biologic immunomodulator Corynebacterium parvum or the synthetic immunomodulators maleic anhydride-divinyl ether or avridine in liposomes resulted in comparable enhancement of resistance in 89Sr-treated and normal mice. These data indicate that natural and immunomodulator-enhanced resistance of CD-1 mice to microbial infections do not depend on normal levels of Mo, PMN, or NK cells. The resistance enhancement may rely on activated tissue macrophages (M phi). In contrast to the early changes in circulating leukocytes, the resident peritoneal cell populations were not markedly altered until after day 30. There then was a distinct decline in lymphocytes and a gradual decline in M phi; the change in M phi was apparently due to the lack of an age-related increase in the peritoneal M phi population in 89Sr-treated mice in comparison with a slight increase in resident M phi in normal mice. After CD-1 mice were treated with 89Sr, the number of PMN and the function of NK cells generally recovered by about day 50 and was followed by partial recovery of circulating Mo, unless a second dose of 89Sr was administered.

Influences of sodium diethyldithiocarbamate, DTC (imuthiolR) on T cell defective responses of aged BALB/c mice

The effect of chronic imuthiol treatment, 25 mg/kg weekly for 4 months initiated at the age of 12 months, on T-cell functions of aged female BALB/c mice was investigated. Imuthiol restored to normal value the impaired response to Concanavalin A (Con A) and enhanced the proliferative response to phytohemagglutinin (PHA). Impaired cytotoxic T-cell activity (CTL), was restored near to the value of young controls by imuthiol. Serum thymic factor (FTS) levels in serum of treated aged animals outpassed those of untreated young mice. Delayed-type hypersensitivity (DTH) reaction to oxazolone was increased. In contrast, the graft-vs-host (GVH) mortality induced by injecting H-2 histoincompatible cells to irradiated recipients, which GVH was impaired by aging, was not significantly modified by imuthiol. The excessive cytotoxicity for chicken red cells of macrophages (ADCC) from aged mice was reduced, as well as macrophage cytotoxicity for tumor cells. Natural killer cell activity remained unchanged. This finding confirms that imuthiol enhanced effectively T cell-dependent responses but the data on GVH reaction suggest that its effects are under a complex mode of action. Restoration of a normal production of FTS may be one mechanism by which imuthiol acts on the reinduction of the T-cell differentiating pathway in aged mice.

The role of NK cell activity in age-dependent resistance of mice to murine cytomegalovirus infection

The resistance of mice to cell culture passaged murine cytomegalovirus (CC-MCMV) infection developed with age. In parallel with this finding, augmentation of the splenic NK cell activity in older mice was always higher than that of younger mice. The splenic NK cell activity reached the maximum level at 6 day post infection (PI) in 2-4-week-old mice while in 6-8-week-old mice it peaked at 4 days PI. When the dose of CC-MCMV was increased, the NK cell activity was potentiated accordingly. However, it was decreased on the infection with increased doses of the salivary gland passaged MCMV (SG-MCMV). NK cells augmented by MCMV infection actually inhibited in vitro replication of MCMV when they were added to mouse embryonic fibroblast (MEF) monolayers infected with CC-MCMV. Splenic and peritoneal macrophages inhibited in vitro replication of MCMV, but their activities were less potent than those of NK cells.

Inhibition of bone marrow colony formation by human natural killer cells and by natural killer cell-derived colony-inhibiting activity

Incubation of human peripheral blood lymphocytes with bone marrow cells resulted in significant inhibition of colony formation by committed myeloid and erythroid cells. Using positively selected homogeneous natural killer (NK) cell preparations and lymphocyte subpopulations depleted of or enriched for NK cells, we definitively characterize as NK cells the cells in normal peripheral blood that are responsible for inhibition of bone marrow colony growth. The inhibitory effect of NK cells on hematopoiesis can be mediated by a soluble factor that is produced only by NK cells upon culture with HLA-DR+ hematopoietic cells and with NK-sensitive cell lines. Both NK cells and the NK-produced, colony-inhibiting activity (NK-CIA) are suppressive for allogeneic and autologous bone marrow CFU-GEMM (colony-forming units, granulocyte, erythroid, monocyte, megakaryocyte), CFU-E (CFU, erythroid), and early CFU-GM (CFU, granulocyte, monocyte), but not for either BFU-E (burst-forming units, erythroid) or late CFU-GM. [3H]Thymidine incorporation was inhibited by NK-CIA-containing supernatants in HLA-DR+ but not HLA-DR- bone marrow cell populations stimulated to proliferative by colony-stimulating factor (CSF). These data suggest that the NK cell-mediated inhibitory effect on proliferation and differentiation of hematopoietic precursor cells is mediated in part or completely by the secreted NK-CIA. The concentration of NK-CIA reached in the supernatant of the mixture of NK cell-containing lymphocyte populations with bone marrow cells is sufficient to account for the inhibitory effect mediated by NK cells. Our data support the hypothesis that human NK cells play a major role in the control of hematopoiesis, down-regulating it under conditions in which the NK cells are functionally activated.

Deficiency of interferon-alpha generating capacity is associated with susceptibility to opportunistic infections in patients with AIDS

Acquired immune deficiency syndrome (AIDS) is a newly described syndrome in which patients are susceptible to certain malignancies and opportunistic infections (OI) usually found only in immunosuppressed individuals. Patients with AIDS have been found to have deficiencies of virtually all of their host defense systems. In this report, the natural resistance systems have been discussed. Although a deficiency of NK-cell function has been found in many patients with AIDS, this deficiency failed to distinguish patients susceptible to OI or malignancy from male homosexual controls. A deficiency of interferon-alpha generation by mononuclear cells upon exposure to HSV-1 infected fibroblasts was the best correlate with susceptibility to OI in AIDS patients. This deficiency failed to correlate with serum levels of acid-labile interferon-alpha in these patients. Although the interferon generating deficiency may be caused by the infections in these patients, it is more likely that the deficiency lays the groundwork for the establishment of the opportunistic infections.

Paradoxical presence of T cell anergy during successful T cell-dependent tumour immunotherapy: characterization of a state of T cell 'amnaesia' following systemic administration of C. parvum

Systemic administration of Corynebacterium parvum causes T cell-dependent regression of an established methylcholanthrene-induced murine fibrosarcoma beginning 10 days after Cp injection. At this time, tumour specific effector T cell responses measured by reactivity in a T helper cell assay or in a Winn assay disappear only to return later. We refer to this temporary lapse in T cell reactivity as immunological 'amnaesia'. Antigen specific T cell responses within all lymphoid organs appear to be affected. The 'amnaesic' state is characterised by the presence of primed T cells but the absence of T effector cells and suppressor cells. The differentiation of the primed T cells is blocked probably as a result on the non-delivery of a differentiation signal. There are several possible mechanisms which could account for this; the one we prefer is that cells are prevented from entering T cell-dependent cell interaction areas within lymphoid organs. This state of T cell 'amnaesia' may underlie anergy in some inflammatory, infectious and neoplastic diseases. The apparent paradox of T cell-dependent tumour regression occurring in mice with depressed T cell responses is discussed.

Natural resistance of lethally irradiated F1 hybrid mice to parental marrow grafts is a function of H-2/Hh-restricted effectors

The natural resistance of F1 hybrid mice against parental bone marrow grafts is thought to be mediated by natural killer (NK)-like effector cells. However, unlike the NK cell activity against a wide range of tumors and normal cells, hybrid resistance is characterized by the immunogenetic specificity controlled by a set of unique noncodominant genes denoted as Hh. Two alternative hypotheses can account for the specificity. Thus, the specificity may reflect either the Hh restriction of effectors or the Hh gene control of mechanisms regulating non-Hh-restricted effector activity. In this study, therefore, we tested the recognition specificity of putative effectors mediating hybrid resistance in lethally irradiated H-2b/d and H-2b/k F1 hybrid mice to the engraftment of parental H-2b bone marrow. As a direct means of defining the effector specificity, rejection of parental bone marrow grafts was subjected to competitive inhibition in situ by irradiated tumor cells. Of the 16 independent lines of lymphoma and other hemopoietic tumor cells tested, the ability to inhibit hybrid resistance was the exclusive property of all tumors derived from mice homozygous for the H-2Db region, regardless of whether the tumor cells were susceptible or resistant to NK cell-mediated cytotoxicity in vitro. Four cell lines heterozygous for the H-2Db were noninhibitory, including one that is susceptible to natural killing. Pretreatment of the F1 hosts with an interferon inducer augmented the resistance with no alteration in the recognition specificity of effector cells. Therefore, natural resistance to parental H-2b bone marrow grafts was mediated by effectors restricted by the H-2Db/Hh-1b gene(s), and not by the nonrestricted NK cells detectable in conventional in vitro assays.

Lysis of FLD-3 Friend erythroleukemia cells in vitro and in vivo: effect of 89Sr treatment and Friend virus infection

The effector cells from non-immunized mice capable of lysing 51Cr-labelled FLD-3 BALB/c Friend virus-induced erythroleukemia cells in vitro and cells capable of clearing FLD-3 cells labelled with 5-iodo-2'-deoxyuridine-125I (125IdUrd) from the lungs in vivo were characterized and compared with natural killer (NK) cells reactive against YAC-I lymphoma cells. Unlike NK cells, the cells capable of lysing FLD-3 cells in vitro were insensitive to antibodies directed against NK-2.1 or Thy-1.2 antigens (plus complement) and to pretreatment of mice in vivo with silica particles, 89Sr or estradiol. Heat-killed C. parvum organism stimulated anti-FLD-3 effector cells without changing the slow rate (24 h) of lysis in vitro. The ability to clear FLD-3 and YAC-1 cells from the lung was normal and defective, respectively, in C57BL/6-bg/bg(beige) mice and in mice pretreated with 89Sr or estradiol. We conclude that natural cytotoxic (NC) cells lyse FLD-3 cells, Fv-2, which regulates resistance to leukemia induction by Friend virus, does not regulate NC(FLD-3) activity, and the virus does not affect NC(FLD-3) activity during the first several days of infection of normal genetically susceptible mice. However, infection of 89Sr-treated mice inhibits NC(FLD-3) function owing to the activation of suppressor cells. These data suggest (but do not prove) that effector cells similar or identical to NC(FLD-3) cells may function in vivo to resist the proliferation/survival of certain leukemia cells.

A genetic analysis of natural resistance to nonsyngeneic cells: the role of H-2

The genetic control of natural resistance in vivo to four natural killer (NK) cell-resistant H-2 homozygous lymphoid tumor cell lines was investigated by following the survival and organ distribution of cells prelabeled with radioactive iododeoxyuridine. Backcross mice derived from DBA/2J and CBA/J parents were injected with H-2d tumor cells and tumor cell elimination was lowest in H-2d homozygotes. Natural killer cell activity was also reduced in mice with the H-2d haplotype, but no direct correlation between NK cell levels against YAC-1 or SL2-5 lymphoma cells and natural resistance in vivo was demonstrable. Analysis of 23 BXD recombinant inbred strains indicated that natural resistance to H-2d tumors was restricted to H-2b strains. There was no direct association of NK cell activity with H-2 type in the BXD strains and NK cell levels did not correlate with tumor survival in vivo. By comparing natural resistance to H-2d and H-2b tumors in DBA/2, C57BL/6, B6D2F1, and B10.D2 mice we found that H-2 nonidentity between the tumor and the host, rather than the host H-2 haplotype, determined whether natural resistance occurred. Again, NK cell activity against YAC-1 cells was not predictive of tumor survival in these strains. These results provide genetic evidence that NK cells alone cannot account for natural resistance to H-2 nonidentical cells of hemopoietic origin.

Natural resistance against hematopoietic cells in lethally-irradiated mice infected with Friend leukemia virus

The influence of in vivo infection with the polycythemic substrain of Friend leukemia virus on noninducible ('natural') resistance against allogeneic normal or malignant grafts was studied in lethally irradiated mice. Parallel studies were performed on the NK system in the same experimental conditions. The results indicate that FLV-P infection of mice with full (DBA/2) vs partial (BALB/c and CD2F1) susceptibility did not suppress their in vivo natural resistance against bone marrow or El-4 leukemia cells. On the other hand, a decline in NK activity paralleled the progression of leukemic disease in the more susceptible DBA/2 hosts.

Murine malaria: dissociation of natural killer (NK) cell activity and resistance to Plasmodium chabaudi

Striking differences in the resistance to P. chabaudi infection among different inbred mouse strains have previously been correlated with the level of both the spontaneous and the infection-induced enhanced level of NK cell activity. We have examined this putative correlation in individual animals of backcross progeny derived from A/J (malaria-susceptible, low NK cell activity) and B10.A (malaria-resistant, high NK cell activity) progenitors. We have found that NK cell activity and resistance to malaria segregated independently. Furthermore, C57BL/6-bg/bg mice which are deficient in NK cell activity were found to be as resistant to malaria as their heterozygous C57BL/6-bg/+siblings. We conclude that low NK cell activity, characteristic of A/J strain mice, is not a sufficient determinant of the exquisite susceptibility of these animals to infection with Plasmodium chabaudi.

Murine non-lymphoid tumors are lysed by a combination of NK and NC cells

Natural cell-mediated cytotoxicity (NCMC) against a variety of tumor targets is mediated by a heterogeneous group of effector cells with the natural killer (NK) and natural cytotoxic (NC) cells being the predominant prototypes in mice. This report shows that non-lymphoid tumor targets, mostly derived from chemically induced fibrosarcomas, are susceptible to either (1) NK-mediated lysis with all the activity being the function of a poly-IC augmentable Qa-5+ effector cell; (2) NC-mediated lysis with all activity being the function of a Qa-5- cell not augmented by poly-IC; and (3) a combination of NK-and NC-mediated lysis with activity being the function of both Qa-5+ and Qa-5- cells, the NK (Qa-5+) augmented by poly-IC. These studies further support the view that murine NC and NK cells are distinct and collectively make up the NCMC system, and also that the previous association of NK cells with lymphoid tumor lysis and NC cells with non-lymphoid tumor lysis is not a valid one.

Rat natural killer (NK) cells are not bone marrow dependent

Rats were treated with various doses of 89Strontium. At a dose effecting depletion of the bone marrow (200 muCi), NK activity of spleen, peritoneal and peripheral blood lymphocytes was unaltered or only slightly decreased. At higher doses (400 muCi), a general impairment of the immune system, including T-cell functions, was observed. After application of 700 muCi 89Strontium, rats died within three weeks.

Natural killer (NK) cells in the rat: "heterogeneity" as reflection of the activation status

Rat natural killer (NK)/natural cytotoxic (NC) cell activity was tested in a 4-h and a 20-h 51-Cr-release assay. In the 4-h assay, NK activity was high with peritoneal cells (PC), medium with peripheral blood lymphocytes (PBL) and spleen cells and low with bone marrow (BM), thymus and lymph node (LN) cells. The differences in lytic capacity were less pronounced in a 20-h assay. This led to the suggestion that NK/NC cells may get activated upon coculture with tumor cells during the long-term assay and that the short-term assay actually reflects the activation status of NK/NC cells in different organs rather than non-identity of cells with natural cytotoxic potential. This was supported by the following observations: 1. After in-vivo or in-vitro activation of NK cells, increased reactivity was observed in a 4-h assay especially with lymphoid cells of organs with low or medium NK activity and the differences to a priori highly active organs were significantly diminished. 2. Neither by biophysical (density) nor by biochemical (surface markers) methods differences between NK versus NC cells could be substantiated. 3. Closely related patterns of cytotoxicity were observed using adherently growing or lymphoid tumor target cells.

Modulation of natural killer (nk) cell activity during FLV-P virus infection of mice

We analyzed the effects of a polycythemic substrain of Friend leukemia virus, i.e. the FLV-P virus, on splenic NK activity of DBA/2 susceptible mice. One day after virus injection a significant increase of NK activity was found, which persisted until day 10. On the other hand, 14-21 days after virus injection a marked and significant depression of activity was measured. This depression was associated with the appearance of suppressor cells able to inhibit the lytic activity of untreated splenocytes when mixed in vitro in the 4 h 51Cr-release assay. The suppressor cell population was insensitive to treatment with anti-Thy 1.2 plus complement, was adherent to Sephadex G-10 and nylon, but did not adhere to plastic, suggesting it is neither a T-cell nor a typical macrophage. The possible relevance of NK activity modulation in relation to the induction of leukemia is discussed.

Aging of natural and acquired immunity of mice. II. Decreased T cell responses to syngeneic tumor cells and parental-strain spleen cells

Cytolytic T lymphocyte (CTL) functions were compared in mice between the ages of 2 and 30 months. The stimulator cells were H-2 allogeneic spleen or tumor cells, parental-strain spleen cells, or syngeneic tumor cells. Effector cells capable of lysing syngeneic tumor cells were shown to be T cells and not NK cells. The cell-mediated lympholysis (CML) responses by spleen cells of aged mice were near normal against H-2 allogeneic spleen or tumor cells but were defective against syngeneic tumor cells or parental-strain spleen cells. The defective syngeneic tumor CML response was observed at various responder:stimulator ratios and at various days of incubation. The defect was in the nonadherent, and not in the adherent, fraction of spleen cells. Suppressor cells were detected in spleens of 30 month, but not of 18 month old mice. Aged mice were more susceptible than young mice to small inocula of syngeneic C57BL EL-4 lymphoma cells. The immunogenicity of irradiated spleen cells of old mice had not changed for the F1 antiparent CML response. Splenic CML responses of young mice treated with 89Sr demonstrated a similar pattern, i.e., good responses to H-2 allogeneic stimulator cells but poor responses to syngeneic tumor cells or to parental-strain spleen cells. This loss of certain CTL functions influenced by marrow dependent cells can partially explain the increased susceptibility of old animals to tumors.

Natural cytotoxic cells against solid tumors in mice. IV. Natural cytotoxic (NC) cells are not activated natural killer (NK) cells

Natural cell-mediated cytotoxicity (NCMC), measured against a variety of tumors, is mediated by at least two sub-populations of effector cells: natural cytotoxic (NC) and natural killer (NK). The studies described in this report show that target lysis by NC cells requires a prolonged (18- to 24 h) assay period, whereas NC-susceptible targets can be lysed in short-term (4h) [3H]-proline assays using allo-sensitized CTL or mitogen-activated cytotoxic populations. NC cells are not "activated" during the long-term assay as indicated by: (1) demonstrating that lysis of NC-susceptible targets in long-term (20 h) 51Cr assays is still the function of a Qa-5- effector cell (NC) and (2) the fact that preincubation of the NC effector cell with susceptible targets for 18 h did not result in the activation of an NK-like population (kinetics of target lysis were comparable to those noted with fresh NC cell preparations). We show that NC cell activity is preserved in both the beige mutant and the PL/J mouse strains, both of which exhibit low NK cell activity, even in long-term assays. These combined studies support the view that NC and NK cell activities are the function of distinct cell types and not the property of a single cell class under different states of activation.

Apparent irrelevance of NK cells to resolution of infections with Babesia microti and Plasmodium vinckei petteri in mice

Increased natural killer (NK) cell activity was found in the spleen and peritoneal cavity of mice infected with Babesia microti or Plasmodium vinckei petteri. This increased activity appeared not to be associated with the effectiveness of the host response against these parasites, since it reached its maximum when the parasitaemia was still low, and had decreased by the time the parasites reached peak densities. In addition mice pretreated with 89Strontium of 17 beta-oestradiol experienced the same pattern of infection as did control mice, yet the infections induced much lower levels of NK activity in the pretreated mice. The course of infection with B. microti was also unaltered in beige mice, which are genetically deficient in NK cells. Thus we consider it unlikely that NK cells are of primary importance in non-specific immunity to these haemoprotozoan infections.

Analysis of low natural killer cell activity in 89Sr-treated mice

Treatment of mice with the long-lived bone-seeking radioisotope 89Sr results in the selective irradiation and destruction of the bone marrow. This is accompanied by a marked reduction in natural killer cell activity against YAC-1 lymphoma [NK(YAC-1)]. To test for the presence of cellular suppressors of NK(YAC-1) in 89Sr-treated mice, in vitro and in vivo cell mixture protocols were used. In vitro, we did not observe any specific inhibitory effect of spleen cells from 89Sr-treated mice on NK(YAC-1) activity of normal spleen cells. The NK(YAC-1) activity of 89Sr-treated mice, measured in vivo by their ability to clear radiolabeled YAC-1 cells from the lungs, was impaired. However, spleen cells from 89Sr-treated mice, when adoptively transferred with normal spleen cells, failed to inhibit the NK(YAC-1) activity of the latter in the lung clearance assay. Further, when normal spleen cells were injected into 89Sr-treated mice, the ability of the transferred cells to mediate in vivo activity was not suppressed in the 89Sr-treated host. These experiments support the suggestion that the low NK(YAC-1) activity in 89Sr-treated mice is not mediated by suppressor cells, but may be due to the destruction of the marrow microenvironment which is essential for the generation of functional NK(YAC-1) cells.

Murine cytomegalovirus stimulates natural killer cell function but kills genetically resistant mice treated with radioactive strontium

Treatment of C3H/St mice with 100 microCi of 89Sr weakened their genetic resistance to murine cytomegalovirus (MCMV) infection. The criteria utilized to detect increased susceptibility were: (i) survival of mice; (ii) numbers of MCMV-infected cells in the spleens and liver; and (iii) serum glutamic pyruvic transaminase levels. The natural killer (NK) cell activity of spleen cells from mice treated with 89Sr is very low. However, the NK activities of spleen cells of both normal and 89Sr-treated mice were greatly augmented 3 days after infection with MCMV. These NK cells lysed a variety of tumor cells and shared several features with conventional NK cells, but were not lysed by anti-Nk-1.2 serum (specific for NK cells) plus complement. Splenic adherent cells did not lyse tumor cells themselves but were necessary for the stimulation of NK cells by MCMV. The paradox of high NK cell function and poor survival in 89Sr-treated mice infected with MCMV was a surprise. We conclude that these augmented NK cells, of themselves, cannot account for the genetic resistance of C3H/St mice to infection with MCMV.

On the heterogeneity of murine natural killer cells

The heterogeneity of cells capable exerting spontaneous cytotoxicity in vitro was explored using antisera to several genetically determined surface markers on mouse lymphocytes. Four phenotypes of cells derived either from fresh or cultured murine lymphoid tissue were found to exert natural killer (NK) activity in vitro. One affector cell subset, termed NKI cells, had the serological phenotype of Thy-1-, Lyt-2-, Qa5+, and lysed measles virus persistently infected target cells (HeLa-Ms) but not P815 mastocytoma cells. It corresponds with the NK cells described in most systems in which lymphoma targets are commonly used. A second subset, with the same target cell specificity, termed NKT is a thymus-independent cell with the phenotype Thy-1+, Lyt-2-, Qa-5+, Ly-5+. A third subset of NK cells, termed T killer (TK) cells deriving from cultures of conventional but not nude mouse spleens, mediated spontaneous cytotoxicity of P815 mastocytoma cells, but not of virus-infected targets. It has a phenotype of Thy-1+, Lyt-2+, Qa-5-, Ly-5+, apparently identical with that of conventional, antigen-specific cytotoxic T lymphocytes. The fourth phenotype of NK cells, termed NKM, derived primarily from cultures of bone marrow, is cytotoxic for HeLa-measles but not P815, and expresses only Ly-5+ among the various markers tested. Beige mice possess normal TK and NKM activities, but had normal NKI, NKT as well as NKM activity. All NK cell subsets express the Ly-5 surface marker. The existence of four phenotypically distinct NK effector cells was strengthened by studies on selective regulation of their activity by two different biological factors. Interferon (IFN) augmented NK activity of primarily one of the subsets examined, the NKI cell; the activity of IFN on NKT cells could not be directly tested, but IFN was without positive effect on TK or NKM cells. In contrast, partially purified IFN-free interleuken 2 (IL-2) augmented the activities of both the TK and NKT subsets, but not of NKI or NKM cell. IL-2 was active in augmenting NK activity in spleen cells obtained from both conventional and nu/nu mice, but was without effect on spleens of nu/nu mice depleted of Thy-1+ cells. These and other data suggest that IL-2 acts primarily, if not exclusively, on THy-1+ cells. These results strengthen the view that natural cytotoxicity in vitro can be mediated by several distinct cell populations under different genetic and regulatory control and indicate the importance of defining and delineating the cell lineages of each and the role of the independent subsets in resistance to virus infections and tumors in vivo.

Heterogeneity of natural killer cells in the mouse

Mice were treated with the bone-seeking isotope, 89Sr, cyclophosphamide, and short-term lethal irradiation in vivo, and murine spleen cells are treated with anti-Nk-1.2 plus complement (C) in vitro. Fresh spleen cell suspensions from the above groups and from beige and neonatal mice were subsequently tested for natural killer (NK) cell activity against a panel of lymphoid and nonlymphoid tumor cell target. NK cell reactivities against YAC-1, MPC-11, and Cl.18 tumors were markedly and consistently reduced in (a) mice treated with 89Sr, (b) spleen cells treated with anti-Nk-1.2 plus C, and (c) C57BL/6 bg/bg mice. In contrast, NK activities against FLD-3 and WEHI-164.1 tumors were usually normal in mice treated with 89Sr, in beige mutant mice, and in spleen cells after treatment with anti-Nk-1.2 antibody and C. It appears, therefore, that two major groups of NK cells exist in fresh mouse spleen cells suspensions. NK-A cells are marrow dependent, Nk antigen positive, and deficient in beige mice; these lyse YAC-1, MPC-11, and Cl.18 tumors. NK-B cells, which are responsible for the lysis of WEHI-164.1 and FLD-3, are Nk antigen negative, marrow independent, and unaffected by the bg/bg mutation. Other features of NK-B cells, suggest that these NK cells, although they share the characteristics mentioned above, differ among themselves especially with respect to age of maturation and susceptibility to cyclophosphamide and total body irradiation. The NK-B group may therefore induce subsets that remain to be defined.

Multiple activities of a cloned cell line mediating natural killer cell function

A special class of immunologic cells can lyse or damage a variety of target cells, notably malignant cells in vitro. These cells have been called natural killer (NK) cells because lysis does not require deliberate immunization by tumor cells. Although these cells can be distinguished from conventional T cells, B cells, and phagocytic cells, they have been difficult to define. We describe a representative cloned cell line that was obtained by cloning Ig -Ly-5+ cells from spleen. This clone, Cl.Ly-1-2-NK-1+/11, displays Thy-1, Ly-5, Qat-4, Qat-5 and NK-1 cell surface antigens and lyses the NK-sensitive YAC-1 lymphoma cells, but does not lyse RL-12 cells, an NK-resistant lymphoma. In addition, this clone lysed the P815 mastocytoma, EL4 lymphoma, and lipopolysaccharide-activated B lymphocyte targets. This cloned population therefore combined information for a unique display of cell surface antigens and specialized function similar to "activated" NK cells. Because this cloned population forms conjugates with susceptible but not resistant target cells, it may prove useful to identify the structure of cell surface molecules that recognize foreign cells. Finally, cells of this clone also specificity lysed target cells coated by antibodies to determinants on the target cell surface, demonstrating that a single cloned cell population can mediate two specialized immunologic functions: antibody-dependent cellular cytotoxicity and NK cell lysis.

Hybrid resistance to EL-4 lymphoma cells. II. Association between loss of hybrid resistance and detection of suppressor cells after treatment of mice with 89Sr

(C57BL/6 X DBA/2)F1 hybrid (B6D2F1) mice resist the growth of parental-strain (B6) EL-4 lymphoma cells inoculated intraperitoneally; that is, B6D2F1 mice survive longer than B6 mice and do not develop ascites. As compared with B6 mice, B6D2F1 mice have higher levels of natural killer (NK) activity against 51Cr-labelled EL-4 cells in their lymphoid organs. B6D2F1 mice treated with 89Sr lose NK activity for certain lymphoma cell targets, e.g. YAC-1, but NK(EL-4) function is usually intact. However, 89Sr-treated mice had lost hybrid resistance to EL-4 cells in vivo, as determined by survival by irradiated or unirradiated EL-4 cells, Corynebacterium parvum, or polyinosinic:polycytidylic acid (pI:pC) in spleens of normal B6D2F1 mice, but NK(EL-4) activity was depressed within 3 days by such treatment in B6D2F1 mice previously injected with 89Sr. Suppressor cells for NK(EL-4) but not for NK(YAC-1) effectors were easily detected in spleens of 89Sr-treated mice "challenged' with C. parvum. Thus, agents capable of stimulating NK cell function in normal mice may lead to suppression of that activity in mice depleted of marrow-dependent cell function by 89Sr. Spleen cells of 89Sr-treated B6D2F1 mice were also unable to generate anti-EL-4 cytotoxic T lymphocytes in a cell-mediated lympholysis system; this defect appeared also to be mediated by suppressor cells. Lymphoid cells depleted by 89Sr-induced marrow aplasia may have two functions in host defences against tumours (especially lymphomas): they may lyse tumour cells directly and they may "down-regulate' suppressor cells capable of inhibiting other "natural' or "induced' immune functions.

Homozygosity at the major histocompatibility complex is required for optimal immunogenicity of bone marrow cell allografts in irradiated rats

Hemopoietic histocompatibility (Hh) Genes associated with the H-2 region control the antigenicity of hemopoietic cell grafts in the mouse. We have tested for similar genes in rats. Wistar Furth (WF, RTlu) or Lewis (LEW RTl1) bone marrow cell grafts did not proliferate in spleens of lethally irradiated (WFxLEW) F1 hybrid rats as assessed by measuring the incorporation of 5-iodo-2' deoxyuridine - 125I (IUdR) into recipient spleens 5 days after transplantation. In contrast, (WFxLEW) F1 hybrid marrow cells grew well in both WF and LEW parental strain hosts. (WFxLEW) F1 or (WFxLEW) F1 hybrid rats were backcrossed to WF parental strain rats to produce progeny, either homozygous, or heterozygous for the MHC. The RTl type of 46 individual backcross progeny was determined using a 5 day mixed-lymphocyte reaction (MLR). Correlation between RTl type and growth of marrow grafts of individual backcross rats were determined by using each rat as a bone marrow donor for irradiated LEW hosts. Marrow grafts from rats heterozygous for RTl were accepted in all 25 cases, whereas, grafts from 19 to 21 homozygous donors were rejected by the LEW hosts. Thus, homozygosity, for Hh determinants in or near the RTl region appears to be necessary for optimal immunogenicity of bone marrow allografts.

Protective effect of low-dose interferon against neonatal murine cytomegalovirus infection

Mice were injected with 10 to 5,000 reference units of interferon intraperitoneally or subcutaneously within 24 h of birth and reinoculated intraperitoneally 24 h later with 200 plaque-forming units of murine cytomegalovirus. Mock interferon and virus diluent were the control inocula. Infection of mock interferon-treated mice resulted in significant retardation of growth, accompanied by tissue injury and a depressed blastogenic response of splenic lymphocytes. Prophylactic administration of interferon prevented growth retardation and resulted in lower tissue viral titers and diminished injurious effects of the virus. Intraperitoneal inoculation of interferon was more protective than was subcutaneous, and 10 U of interferon was often as effective as 5,000 U. Accelerated maturation and enhanced activity of lymphoid elements were observed histologically in spleens and lymph nodes of interferon-treated mice; supportive of these findings was the greater incorporation of [3H]thymidine of splenocytes from interferon-treated mice. The protective effect of interferon may, therefore, be due to stimulation or accelerated maturation of cellular immune functions.

Normal levels of natural cytotoxic cells against solid tumours in NK-deficient beige mice

Natural cell-mediated cytotoxicity (NCMC) capable of in vitro lysis of various lymphoid and non-lymphoid tumours has been described in mice and other species, including man. NCMC has been proposed as a first level of defence against tumour growth in vivo, one which does not need the priming of the conventional immunological response. The effector cells of NCMC seem to belong to a special category of lymphoid cells, being neither classical T or B cells nor macrophages; natural killer (NK) cells have been proposed as the prototype effector cell, although some heterogeneity among effector cells seems to exist, depending on the target cells used for testing. Two main subgroups of NCMC effector cells have been defined: NK cells directed against lymphoma targets and natural cytotoxic (NC) cells directed against solid non-lymphoid tumours. We describe here another distinction between the two systems: while NK activity is low in mice homozygous for the beige (bg) gene NC activity in spleen cell preparations from these animals is comparable with that observed in the appropriate controls (bg/+ and +/+ littermates). The bg syndrome of mice affects lysosome, melanosome and enzymatic functions and is a homologue of the Chediak--Higashi syndrome of man. Defective NK activity in blood lymphocytes has also been reported in patients with Chediak--Higashi syndrome. We also show that several mouse strains which have low NK activity, have normal or high levels of NC functions, expanding our previous observation that NC and NK cells are under distinct genetic control.

Immunosuppression by human plasma alpha 1-acid glycoprotein: importance of the carbohydrate moiety

Human plasma alpha 1-acid glycoprotein (orosomucoid) and derivatives of this protein produced by sequential enzymatic cleavage of the glycosidic residue (sialic acid, galactose, N-acetylglucosamine, mannose) were tested for the ability suppress a number of immune functions of mouse spleen cells in vitro at physiological concentrations. alpha 1-Acid glycoprotein and especially the agalacto/asialo derivative suppressed (i) the mitogenic responses to concanavalin A, lipopolysaccharide, and alloantigens, (ii) antibody responses to sheep erythrocytes, and (iii) induction of cell-mediated lympholysis against allogeneic target cells. The native protein and its derivative did not inhibit the proliferation of EL-4 lymphoma cells and did not suppress the lysis of YAC-1 lymphoma cells by natural killer cells. The enhanced potency of the agalacto/asialo derivative indicates that the nature of the carbohydrate moiety exposed on the protein determines the effectiveness of the glycoprotein and its deglycosylated derivatives may function to regulate immune responses in various physiological and pathological conditions.