Production of multiple cytokines by clones of human large granular lymphocytes

Positive and negative regulation of Natural Killer cells: therapeutic implications

Natural Killer (NK) cells can mediate numerous anti-tumor and anti-viral effector functions as well as play important immunoregulatory roles in various disease states. Promoting the ability of NK cells to respond in an immunotherapeutic setting has often been sought by the addition of NK cell-stimulating factors. However, such therapies are often found to be insufficient, which may in part be due to the presence of inhibitory influences on the NK cell. NK cells can respond to a plethora of cytokines which are generated by numerous cell types and these interactions can markedly affect NK cell survival and activity. NK cells also possess multiple activating and inhibiting receptors which can alter their function. Whether the NK cell will become activated or not can depend on a complex balance of activating and inhibitory signals received by the cell and modulation of these signals may shift the balance on NK activation. This review discusses the various activating and inhibitory stimuli which can act on NK cells, and suggests that future NK cell-based therapies consider not only activating stimuli but also removal of possible inhibitory elements which could prevent optimal NK cell function and/or survival.

Effector cells derived from host CD8 memory T cells mediate rapid resistance against minor histocompatibility antigen-mismatched allogeneic marrow grafts without participation of perforin, Fas ligand, and the simultaneous inhibition of 3 tumor necrosis factor family effector pathways

Reduced-intensity conditioning regimens for transplant recipients have heightened awareness of immunologic resistance to allogeneic bone marrow transplants (BMT). Although T cell-mediated cytotoxicity has been assumed to play a role in the resistance against donor allogeneic hematopoietic stem and progenitor cell grafts, several studies have reported relatively unimpaired resistance by recipients who lack perforin, Fas ligand (FasL), and other cytotoxic mediators. This study compared the early kinetics of T cell-mediated resistance in B6 (H2b) cytotoxically normal versus deficient recipients after transplantation with major histocompatibility complex-matched, minor histocompatibility antigen (MiHA)-mismatched allogeneic marrow grafts. Wild-type B6 or cytotoxic double-deficient perforin-/-/gld+/+ (B6-cdd) mice were sensitized against major histocompatibility complex-matched BALB.B or C3H.SW (H2b) MiHA and transplanted with a high dose (1 x 10(7)) of T cell-depleted bone marrow. CD8 T memory cells were shown to be present in recipients before BMT, and anti-CD8 monoclonal antibody infusion abolished resistance, thus demonstrating that CD8 T cells are the host effector population. Donor-committed and high proliferative potential progenitor numbers were markedly diminished by 48 hours after transplantation in both wild-type B6 and B6-cdd anti-donor MiHA-sensitized recipients. These observations indicate that the resistance pathway used in the cytotoxic deficient mice was both potent and rapidly induced--consistent with a CD8 memory T-cell response. To examine the role of Tumor necrosis factor-like weak inducer of apoptosis (TWEAK)- and TL1A-mediated cytotoxicity in this strong resistance, newly generated monoclonal antibodies specific for these ligands were administered to B6-cdd recipients sensitized to donor antigens. Recipients of syngeneic B6-gfp bone marrow exhibited significant donor colony-forming unit numbers after BMT. In contrast, low or absent colony-forming unit levels were detected in allogeneic recipients, including those that lacked perforin and FasL and that received anti-TWEAK, anti-tumor necrosis factor-related apoptosis-inducing ligand, and anti-TL1A monoclonal antibodies. These findings extend previous observations by demonstrating the existence of a rapidly effected resistance pathway mediated by memory CD8 effector T cells independent of the 2 major pathways of cytotoxicity. Together with previous findings, these results support the notion that effector cells derived from memory CD8 T-cell populations can mediate strong resistance against donor allogeneic MiHA-disparate hematopoietic engraftment by using a mechanism that is independent of the contribution of perforin, FasL, and the known death ligand receptor pathways.

The immunobiology of natural killer cells and bone marrow allograft rejection

Natural killer (NK) cells mediate the acute rejection of bone marrow cell (BMC) allografts, but not solid tissue grafts, in lethally irradiated mice. However, the mechanisms underlying this capability for rejecting BMC remain unclear. NK cells express (1) inhibitory receptors specific for major histocompatibility complex (MHC) class I molecules and (2) activating receptors with diverse specificities. Inhibitory NK receptors confer to NK cells the ability to discriminate between MHC class I-positive and -negative target cells and are therefore involved in the control of NK cell tolerance to self, as well as in the elimination of cells that have downregulation of MHC class I molecules. Preclinical studies in mice have provided good evidence that subsets of NK cells that bear different combinations of both inhibitory and activating Ly49 receptors can interact with each other and target specific BMC rejection, as well as NK cell responses toward tumor cells. Recent clinical studies have also shown that the use of killer cell immunoglobulin-like receptor ligand incompatibility in patients with leukemia who received hematopoietic stem cell transplants correlated not only with the elimination of graft rejection, but also with eradication of tumor and prevention of graft-versus-host disease; this offers a significant advantage for survival. In this review, we attempt to bring together literature regarding the biology of NK cells and discuss the current issues in bone marrow transplantation and the potential clinical role of NK cell alloreactivity in the efficacy of this procedure for immunotherapy of cancer and infectious states.

Primary immune system effects of the orally administered cyclopentane neuraminidase inhibitor RWJ-270201 in influenza virus-infected mice

The cyclopentane derivative [1S,2S,3R,4R]-3-[(1S)-1-(acetylamino)-2- ethylbutyl]-4-[(aminoiminomethyl)amino]-2-hydroxy-cyclopentanecarboxylic acid (RWJ-270201) has been previously reported to be a potent and selective inhibitor of influenza virus neuraminidase, and to inhibit infections with this virus in vitro, in mice, and in clinical challenge studies. The effect of oral gavage therapy of 100 mg/kg/day of RWJ-270201 administered twice daily for 5 days beginning 16 h prior to virus exposure, on various immune factors of importance in response to primary influenza infection was determined in mice infected with influenza A/Shangdong/09/93 (H3N2) virus. Spleens taken from the mice 2 h after termination of treatment were processed for cytotoxic T lymphocytes (CTL) and natural killer (NK) cell activity and for enumeration of macrophages, T, T-helper, T-suppressor/cytotoxic, and B cells. Saline-treated mice and normal mice were run in parallel. Treatment had no significant effect on any immune parameter. In a second experiment, mice infected with influenza A/NWS/33 (H1N1) were treated similarly with RWJ-270201 beginning 4 h pre-virus exposure. Treatment prevented any deaths from occurring, and markedly lessened arterial oxygen decline, lung consolidation, and lung virus titers. The mice developed mean neutralizing antibody (NA) titers of 1:592, and six of seven rechallenged mice resisted rechallenge with the same virus, indicating the initial virus-inhibitory effect also did not prevent the animals from developing an adequate humoral immune response to the virus.

Lymphocytes Produce IL-1β in Response to Fcγ Receptor Cross-Linking: Effects on Parenchymal Cell IL-8 Release

Neutrophils mediate tissue injury in response to immune complexes, although the factors that induce their recruitment are incompletely understood. We have reported that lymphocytes may be important regulators of monocyte and macrophage IL-8 release in the presence of immobilized IgG. Since tissue parenchymal cells are important local producers of IL-8 but are not directly stimulated by FcγR cross-linking, we hypothesized that lymphocytes may also regulate parenchymal IL-8 release. Supernatants from lymphocytes incubated on immobilized IgG induced primary human fibroblasts and human mesangial cells to produce IL-8 (17 ± 3.5 and 44 ± 8 ng/ml, respectively). Fibroblast and mesangial cell IL-8 mRNA levels were similarly increased by the conditioned lymphocyte supernatant. Immobilized anti-human FcγRIII, but not FcγRI or FcγRII Abs, could stimulate this IL-8-inducing activity in lymphocytes, suggesting that FcγRIII-bearing lymphocytes were responsible. Supernatants from lymphocytes incubated on immobilized IgG contained 2.2 ± 0.8 ng/ml of IL-1β, while enriched monocyte preparations from the same donors incubated on immobilized IgG released only 0.1 ± 0.04 ng/ml of IL-1β (p = 0.05). Consistent with the identification of IL-1β as the lymphocyte factor, fibroblast or mesangial cell IL-8 release induced by the IgG-stimulated lymphocyte supernatants was inhibited by 1) the combination of IL-1R antagonist and soluble type II IL-1R, 2) an IL-1-converting enzyme inhibitor, or 3) anti-IL-1β but not preimmune Abs. These data suggest that targeted deposits of IgG can stimulate FcγRIII-bearing lymphocytes to produce IL-1β, which induces parenchymal cell IL-8 release.

Differential secretion of TNF-alpha and IFN-gamma by human peripheral blood-derived NK subsets and association with functional maturation

Natural killer cells can be separated into three major subsets (free, binder, and killer) based on their ability to bind and kill sensitive target cells. The nonbinder, nonkiller free cells are the most immature and can be activated to become binders and killers. Natural killer (NK) cells synthesize and secrete several cytokines that are intimately involved in NK activation. This study investigated the secretion of tumor necrosis factor-alpha (TNF-alpha) and interferon-gamma (IFN-gamma) by purified NK cells and NK subsets following activation by various stimuli. K562 target cells stimulated secretion of both TNF-alpha and IFN-gamma by both the binder and the killer subsets but not by the free subset. IFN-alpha activated the secretion of IFN-gamma only, whereas IL-2 activated the secretion of both TNF-alpha and IFN-gamma by the binder and killer subsets and secretion was augmented by the addition of K562 to the cultures. Phorbol myristate acetate (PMA) and ionophore stimulated TNF-alpha and IFN-gamma secretion in both the binder and the killer subsets, though IFN-gamma secretion was more pronounced in the binder subset. Activation of TNF-alpha and IFN-gamma secretion was dependent on de novo protein synthesis. Analysis at the single-cell level demonstrated that the binder subset had the highest frequency of cells secreting IFN-gamma. These results demonstrate that both the binder and the killer subsets can be activated to secrete TNF-alpha and IFN-gamma, whereas the free NK subset secretes little or no TNF-alpha and IFN-gamma following activation. These data suggest that the ability of NK cells to secrete TNF-alpha and IFN-gamma following activation correlates with the functional stage of maturation of NK cells.

Suppressive effects of volatile anesthetics on cytokine release in human peripheral blood mononuclear cells

This study investigated the effects of three volatile anesthetics (sevoflurane, isoflurane, and enflurane) on cytokine release by human peripheral mononuclear cells (PBMCs) stimulated by natural killer (NK)-sensitive tumor cells, K562, in vitro. PBMCs, as effector cells, obtained from 31 volunteers were randomly allocated to two groups in the first set of experiments. One group was incubated with K562 (n = 21) and the other with medium alone as a control (n = 10). In a second set of experiments, PBMCs from each volunteer (n = 21) were divided into three groups: nonanesthetic, 1.5-MAC, and 2.5-MAC groups (n = 7 for each anesthetic). After 2 h exposure to anesthetic gas or air, K562 cells were added to the effector cells. After 4 h incubation, interleukin-1 beta (IL-1 beta), interleukin-2 (IL-2), tumor necrosis factor-alpha (TNF-alpha), and interferon-alpha (INF-alpha) in the supernatant were assayed. IL-1 beta and TNF-alpha levels were significantly increased in comparison with those in the control group. IL-2 levels tended to be higher than those in the control group. No effect on IFN-alpha levels was found. After anesthetic exposure, the releases of IL-1 beta and the release of TNF-alpha were significantly inhibited compared with those after air exposure. None of the anesthetics inhibited IL-2 release. The anesthetics studied are capable of altering the release of cytokines by NK and NK-like cells in response to tumor cells.

Increase of plasma transforming growth factor beta (TGF beta) during immunotherapy with IL-2

Interleukin-2 (IL-2) is a lymphokine with pleiotropic activities on the immune system. When administered in vivo, besides inducing unrestricted tumor cytotoxicity, it is also responsible for the secondary release of other lymphokines, such as IL-1, TNF, and marrow growth factors, which may mediate some of the clinical toxicities (as well as therapeutic effects) seen during IL-2 immunotherapy. Among the clinical effects of IL-2, we previously reported thrombocytopenia and IL-2-induced in vitro inhibition of platelet aggregation accompanied by rapid secretion of alpha-granule components such as platelet factor 4 (PF4) and beta-thromboglobulin. Platelets constitute one of the largest storage forms of TGF beta. Preliminary evaluation of this factor in patients receiving IL-2 had indicated that plasma TGF beta activity increased in cancer patients following IL-2 therapy. We report a more detailed study of the quantitation of TGF beta activity in the plasma of 23 cancer patients treated with IL-2 immunotherapy. Of interest, we found that although elevation of the bioactive form of TGF beta occurred in most patients during IL-2 therapy, it was significantly higher in patients with clinical regression of tumor (p = .004). In the first 2 weeks of therapy increase of plasma TGF beta activity appeared to correlate with a decrease of platelet counts, suggesting that the factor may derive from the storage form of TGF beta contained therein.

Lymphokine-activated killer cell function of lymphocytes from regional lymph nodes in patients with gastric carcinoma

Lymphokine-activated killer (LAK) cells generated by culture of regional lymph node cells (LNC) with interleukin 2 (IL 2) for 4 and 11 days were examined for their functional capabilities in comparison with those of peripheral blood mononuclear cells (PBM) in 25 patients with gastric carcinoma. The cytotoxic activity of LAK cells induced from LNC for 4-day culture with IL 2 was significantly lower than that from PBM. However, the LNC-LAK cytotoxicity was markedly increased up to almost the same level as that of PBM after 11-day culture. The production of interferon-gamma (INF-gamma) and tumor necrosis factor-alpha (TNF-alpha) from nonadherent LAK cells in LNC was also significantly reduced as compared to that from PBM 4 days after culture, when stimulated with or without tumor target, Raji cells. After 11-day culture with IL 2, however, the levels of these cytokines produced by LNC-LAK cells either with or without stimulation by tumor target were comparable to those by PBM-LAK cells, although the release of these cytokines was markedly reduced when compared to that after 4-day culture. Phenotypic analysis revealed decreased proportion of cells mediating NK activity in LNC before and 4 days after culture. CD56+ and CD57+ cells in LNC were increased after 11-day culture, although the percentages of these cells were still low as compared to those in PBM. The proportions of OKIa1+ and CD25+ cells were uniformly increased after 4 and 11-day culture in both cell populations. Changes in subpopulations of CD4+ and CD8+ cells in LNC were not apparently different from PBM. These results indicated the differential LAK cell function of cells from regional lymph nodes from PBM in patients with gastric carcinoma.

Lymphokine-activated killer cell function of peripheral blood mononuclear cells, spleen cells and regional lymph node cells in gastric cancer patients

Lymphokine-activated killer (LAK) cells generated by culture of peripheral blood mononuclear cells (PBMC), spleen cells (SPC) and regional lymph node cells (LNC) with IL-2 for 4 days were examined for their functional capabilities in 29 patients with gastric carcinoma. The cytotoxic activity of LAK cells induced from LNC was significantly lower than that from either PBMC or SPC, although there was no difference between PBMC or SPC. The induction of mRNA of interferon-gamma (IFN-gamma) or tumour necrosis factor-alpha (TNF-alpha) and the production of these cytokines in the non-adherent LAK cells from LNC were also significantly reduced compared with those from PBMC or SPC. Further, the LAK cells from LNC secreted significantly lower levels of these cytokines when stimulated with tumour target, Raji cells, although the production of these cytokines was markedly increased by stimulation with the targets in all three cell populations. Phenotypic analysis of each cell population revealed a decreased proportion of the cells mediating natural killer (NK) activity, including CD16+, CD56+, and CD57+ cells in LNC either before or after culture, although OKIa1+ and CD25+ cells were uniformly increased in all cell populations after culture. Changes in subpopulations of CD4+ and CD8+ cells in LNC were not apparently different from PBMC or SPC. These results indicated the differential reactivity of each lymphocyte population to IL-2 and the reduced LAK cell function of LNC compared with PBMC or SPC in patients with gastric carcinoma.

CD8+ cells and natural cytotoxic activity among spleen, blood, and heart lymphocytes during the acute phase of Trypanosoma cruzi infection in rats

The infection developed by Wistar Furth rats inoculated with the Y strain of Trypanosoma cruzi was the experimental model used in our study. The results showed that this infection altered considerably the CD4/CD8 lymphocyte subset ratio and the natural cytotoxic activity of mononuclear cells in the spleen, blood, and myocardial tissue. Concomitantly, an expansion of the number of cells expressing major histocompatibility complex (MHC) class II antigens was observed, as well as spontaneous development of high levels of blast cells, mainly in the spleen. The inflammatory infiltration of the myocardium, made up essentially of CD8+ cells (cytotoxic/suppressor T cells, natural killer cells), was initially found at 9 days postinfection, spread continuously, and was observed until the death of the animals at about 18 days postinfection. T. cruzi infection also enhanced the natural killer activity of mononuclear cells in the blood, spleen, and myocardium. Sorting these cells by affinity columns showed that the natural killer function was performed exclusively by the CD8+ population, which did not express MHC class II antigens. It was shown that the polyclonal T-lymphocyte activation induced by T. cruzi infection results in a wide distribution of CD8+ cells with enhanced natural cytotoxic activity in the spleen, blood, and cardiac tissue.

Prolonged interleukin-2 (IL-2) treatment can augment immune activation without enhancing antitumor activity in renal cell carcinoma

Preliminary studies involving small numbers of patients have suggested that interleukin-2 (IL-2) administered by continuous infusion in repetitive weekly cycles using doses of 3 x 10(6) U/M2/day is immunologically active and can induce tumor responses in patients with renal cell carcinoma. This study was designed to examine both the immunological and clinical effects of prolonged infusion IL-2 given by repetitive weekly cycles; first at moderate doses for 4 weeks as an impatient followed by lower doses of IL-2 for up to 5 months. Prolonged IL-2 treatment was investigated because previous studies revealed that patients had a return to their baseline immune status within 4 weeks after completing IL-2 treatment. Twenty-five patients (including 18 with renal cell carcinoma) were treated with one of two regimens utilizing IL-2 as sole therapy. These regimens were designed to induce augmented and prolonged immune activation based upon in vitro and in vivo data. Though patients on both arms of the study demonstrated sustained lymphocytosis, increase in numbers of natural killer cells, and induction of lymphokine-activated killer activity with prolonged IL-2 administration, only 1 out of the 18 patients with renal cell carcinoma demonstrated a sustained partial antitumor response to therapy. Furthermore, several patients demonstrated profound immune activation, without any evidence of tumor regression. The lack of clinical responses in these patients showing marked activation of LAK cytotoxicity suggests that other variables must also influence the likelihood of antitumor effects for patients receiving IL-2 therapy.

In vivo migration and tissue localization of highly purified lymphokine-activated killer cells (A-LAK cells) in tumor-bearing rats

Our laboratory has previously reported that the adoptive transfer of highly purified lymphokine-activated killer cells (adherent-LAK, A-LAK) into Fischer 344 (F344) rats bearing established lung or liver micrometastases effectively reduced the resultant tumor growth more than 90%, leading to significant increases in animal survival (Cancer Res. 49, 1441, 1989). To begin to investigate the mechanism(s) by which A-LAK cells mediate this anti-tumor effect, we studied their migration patterns in F344 rats bearing experimentally induced lung and liver metastases as well as subcutaneous tumors. A-LAK cells which were phenotypically 95 to 100% natural killer cells/large granular lymphocytes were labeled with either 51Chromium or fluorescein diacetate (so as to be visualized microscopically). Intravenous injection of such labeled A-LAK cells did not show significant differences in their tissue distribution patterns in tumor-bearing versus normal rats, even when high levels of exogenous recombinant interleukin-2 (rIL-2) was administered. A-LAK cells first migrated to the lungs and then subsequently migrated to the liver and spleen as early as 2 to 6 hr following iv injection. The kinetics of exit of A-LAK cells from the pulmonary capillary beds was not significantly different in rats bearing 3-day micrometastases or 14-day macrometastases compared to normal rats. Moreover, the presence of metastases in the liver did not alter the extent or kinetics of entry of A-LAK cells into the liver even in the presence of exogenously administered rIL-2. Finally, in rats bearing subcutaneous tumors, no evidence could be obtained that A-LAK cells were selectively localized to the tumor site. Tissue sections of livers from metastases-bearing animals injected with fluorescein diacetate labeled A-LAK cells did not demonstrate significant numbers of A-LAK cells infiltrating tumor nests with or without the administration of exogenous IL-2. These data suggest that A-LAK cells may mediate tumor regression in vivo by direct and indirect mechanisms, possibly through the secretion of cytokines and/or the recruitment of secondary effector cells.

A feline large granular lymphoma and its derived cell line

A lymphoma cell line (MCC) was derived from an abdominal mass from a 13-yr-old castrated male cat. The cells resemble natural killer precursor cells, have membrane-bound granules, and are positive for chloroacetate esterase, alpha-naphthyl butyrate esterase, and tartrate-resistant acid phosphatase activities. The MCC cells are negative for rearranged feline T-cell receptor genes, negative for feline T-cytotoxic antigen, Ia, and surface mu, tau, and lambda chains and do not form E-rosettes. The MCC cell line is negative for the feline leukemia virus (FeLV); e.g., negative for exogenous FeLV (exU3) sequences, negative for cytoplasmic and surface FeLV major core protein of 27,000 daltons (p27) by indirect immunofluorescence assay, negative for helper FeLV by clone 81 assay, and negative for release of soluble FeLV p27 by enzyme-linked immunosorbent assay. Electron microscopy reveals budding type C retrovirus particles and MCC cells react with anti-RD-114 (anti-endogenous feline retrovirus) reference serum. After in vitro infection, MCC replicate FeLV readily, but replication is noncytopathic.

Cytokine activity after human bone marrow transplantation. III. Defect in IL2 production by peripheral blood mononuclear cells is not corrected by stimulation with Ca++ ionophore plus phorbol ester

Previous studies have shown that interleukin 2 (IL2) production by peripheral blood mononuclear cells (PBMC) is severely impaired post allogeneic bone marrow transplantation, whereas production of interferon-gamma (IFN-gamma) is at most marginally depressed. To investigate the mechanisms behind this apparently differential inhibition of lymphokine production, we stimulated PBMC from recipients of HLA-identical sibling bone marrow transplants with phytohaemagglutinin (PHA), PHA + phorbol ester (PMA) (to bypass accessory cell requirements) or Ca++ ionophore + PMA (to bypass both accessory cell and T cell surface receptor (CD2 and/or CD3/Ti interactions). Increasing the potency of the stimulus increased the amount of IL2 and IFN produced by PBMC from both normal volunteers and from marrow transplant recipients, but for each stimulus the amount of IL2 produced by marrow transplant recipient PBMC remained 10-100-fold lower than that produced by normal PBMC, suggesting an underlying defect in IL2 production by marrow transplant recipient T cells, not due to accessory cell or CD2 defects. Selection experiments showed that CD3+ cells were the primary IL2 producers, and we were unable to demonstrate presence of suppressor cells in marrow transplant PBMC. Statistical analysis of the clinical factors possibly affecting lymphokine synthesis showed that in vivo cyclosporin A did not affect the in vitro capacity of PBMC to produce cytokines, although steroid therapy had a negative effect on IL2 production. The only variable significantly affecting IL2 and IFN production in marrow transplant recipients was increasing time post transplant. It is suggested that the defect in IL2 but not IFN production could be due to either a selective reduction in the frequency of IL2 producing cells as opposed to IFN producing cells, or to a reduction in the amount of IL2 produced per cell in marrow transplant recipients.

Phenotypic and functional analysis of lymphokine-activated killer (LAK) cell clones. Ability of CD3+, LAK cell clones to produce interferon-gamma and tumor necrosis factor upon stimulation with tumor targets

Lymphokine-activated killer (LAK) cells are generated by the culture of peripheral blood lymphocytes with interleukin-2 (IL-2). A variety of cells, including T-lymphocytes and natural killer (NK) cells, can be activated by IL-2 to exhibit the ability to kill multiple tumor and "modified-self" targets. Recent reports indicate that culture conditions can determine the phenotype of cells expressing LAK activity. Using limiting dilution techniques, we first generated cloned LAK cells with three culture conditions: autologous human serum (AHS) + IL-2; AHS + IL-2 + 0.1 micrograms/ml phytohemagglutinin and fetal bovine serum and IL-2. We determined that all but one of the 47 LAK cell clones generated with the three culture conditions were CD3+ and T-cell like; one NK-like clone was observed. Clones that were cytotoxic for one target could generally kill multiple targets, and the absence of phytohemagglutinin did not significantly affect the ability of the LAK cell clones to kill multiple targets. The presence of phytohemagglutinin was, however, necessary for the long-term maintenance of proliferation and cytotoxic activity of the LAK cell clones. The mechanism by which LAK cells kill tumor targets is not known. We here demonstrate that LAK cells and LAK cell clones can produce interferon-gamma and tumor necrosis factor (TNF) when stimulated with an erythroleukemia cell, K562. Five of the six CD3+, LAK cell clones tested could be stimulated by K562 cells to produce both interferon-gamma and TNF. However, the ability of the cloned LAK cells to kill K562 cells, as measured in a 4-h 51Cr-release assay, did not correlate with their ability to produce these cytokines. Furthermore, specific antibodies that neutralize the cytotoxic activity of interferon-gamma and TNF did not inhibit killing of K562 cells by LAK cells as measured with a 4-h cytotoxic assay. The cytostatic and cytotoxic activities of interferon-gamma and TNF for tumor cells are well documented, but these cytolytic activities are slower acting and exhibit their maximum effect after 48-96 h. We here propose that LAK cells kill tumor targets by a combination of cell-to-cell-mediated killing and by the release of slower acting cytostatic/cytotoxic cytokines that can inhibit the growth of tumors some distance from the effector cells.

Stimulation of cytotoxic and non-cytotoxic functions of natural killer cells by bacterial membrane proteoglycans and ribosomes

The present study was designed to investigate the effect of membrane proteoglycans (MPG) from Klebsiella pneumoniae on the function of human natural killer (NK) cells. MPG combined with bacterial ribosomes from Klebsiella pneumoniae, Streptococcus pneumoniae, Streptococcus pyogenes and Haemophilus influenzae, constitute a bacterial immunomodulator (MS D 53), currently in clinical use. Human peripheral blood lymphocytes (PBL) exposed in vitro to MPG or MS D 53 for 20 h showed enhanced NK cytotoxicity. Augmentation of NK cytotoxicity depended upon a direct effect on NK cells, inasmuch as these compounds were also effective on highly purified large granular lymphocytes (LGL). We also studied the effects of MPG on non-cytotoxic functions of NK cells, namely in vitro locomotion and production of IL-1. MPG (and MS D 53) induced IL-1 release in LGL. Moreover, MPG-treated LGL showed enhanced locomotory activity, as assessed by measuring the penetration into nitrocellulose filters. The capacity of MPG (and MS D 53) to activate cytotoxic and noncytotoxic functions of NK cells may contribute to enhancement of nonspecific resistance in vivo after treatment with this agent.

Helper strategy in tumor immunology: expansion of helper lymphocytes and utilization of helper lymphokines for experimental and clinical immunotherapy

Two main kinds of immune strategy are possible against neoplasia. The first potentiates a selected effector arm. In vitro culture with exogenous interleukin-2 (IL-2) increases the activity of natural killer cells and leads to the expansion of T cytotoxic lymphocytes. Systemic reinfusion of both of these cells with high doses of IL-2 mediates the regression of a variety of murine and human tumors. In an alternative strategy, a few regulatory lymphocytes turn on immune reactivity by triggering a cascade of interconnected effector functions. The efficacy of this strategy rests on the repertoire of effector mechanisms moved to action. An effective immunoregulatory maneuver is the addition of helper determinants on the surface of tumor cells. Its power can be further increased by the pre-induction of helper T lymphocytes specific to the helper determinants. This approach can be achieved in mice by coupling muramyl dipeptides to tumor cells, along with eliciting T lymphocytes specifically reactive to Bacillus Calmette-Guerin. Noncytotoxic T helper lymphocytes produce factors which recruit nonspecific (macrophages) as well as specific (cytolytic T lymphocytes) anti-tumor attacking cells. In this way protection can be afforded against primary tumors and metastases, as well as leukemia cells. As the activity of helper lymphocytes rests mostly on lymphokine release, the use of molecularly defined lymphokines mimicking T-helper functions has also been attempted. In a few experimental models, the association of low doses of IL-2 with non-reactive lymphocytes from tumor-bearing mice promotes an effective anti-tumor reaction in the host. Moreover, the combination of distinct lymphokines can also build a molecularly defined helper system able to activate in sequence non-specific and specific anti-tumor reactions in vivo. Trials intended to evaluate the clinical impact of these helper approaches in the management of human tumors are being started or are already under way.

Purification and characterization of the human interferon-gamma receptor from placenta

Purification of the human interferon-gamma (IFN-gamma) receptor was facilitated by identification of human placenta as a large-scale receptor source. When analyzed in radioligand binding experiments, intact placental membranes and detergent-solubilized membrane proteins expressed 1.3 and 5.9 X 10(12) receptors per mg of protein, respectively, values that were 13-163 times greater than that observed for U937 membranes. Two protocols were followed to purify the IFN-gamma receptor from octyl glucoside-solubilized membranes: (i) sequential affinity chromatography over wheat germ agglutinin- and IFN-gamma-Sepharose and (ii) affinity chromatography over columns containing receptor-specific monoclonal antibody and wheat germ agglutinin. Both procedures resulted in fully active preparations that were 70-90% pure. Purified receptor migrated as a single molecular species of 90 kDa either when analyzed on silver-stained NaDodSO4/polyacrylamide gels or when subjected to electrophoretic transfer blot analysis using a labeled IFN-gamma receptor-specific monoclonal antibody. The identity of the 90-kDa component as the receptor was confirmed by demonstrating its ability to specifically bind 125I-labeled IFN-gamma following NaDodSO4/PAGE and transfer to nitrocellulose. Certain receptor preparations converted into a 55-kDa fragment either during purification or upon storage at 4 degrees C. On the basis of N-Glycanase digestion studies, the IFN-gamma receptor appeared to contain 17 kDa of N-linked carbohydrate. The ligand binding site, the epitope for the receptor-specific monoclonal antibody, and all of the N-linked carbohydrate could be localized to the 55-kDa domain of the molecule.

Phenotypic and functional analysis of human CD3+ and CD3- clones with "lymphokine-activated killer" (LAK) activity. Frequent occurrence of CD3+ LAK clones which produce interleukin-2

Clones capable of lysing fresh, uncultured tumor cells ("lymphokine-activated killer": "LAK" activity) were selected from microcultures derived from either E-rosette-positive or E-rosette-negative cell populations. All the selected clones displayed a strong cytolytic activity against the NK-sensitive K562 cell line. Two major phenotypic groups of clones could be identified: a first group expressed the CD3 differentiation antigen, present exclusively on mature T lymphocytes; however, in contrast to typical cytolytic T lymphocytes, the majority of these clones expressed the unusual CD4- CD8- phenotype, whereas the remainder were CD4- CD8+. A second group was represented by CD3- clones which, in most instances, expressed the T-cell-lineage-specific CD2 antigen. Following stimulation with phytohemagglutinin (PHA), most of the CD3+ LAK clones produced Interleukin-2 (IL-2) and interferon-gamma (IFN-gamma) whereas those expressing the CD3- phenotype did not. Since previous studies indicated that PHA may be inefficient in inducing lymphokine production by T-cell variants lacking the CD3/T cell receptor complex (TCR), CD3- clones were further stimulated with the calcium ionophore A23187 plus phorbol 12-myristate 13-acetate (PMA). Only 2/11 CD3- LAK clones produced small amounts of IL-2, whereas the majority released IFN-gamma. Given the peculiar phenotypic and functional properties of many CD3 + LAK clones, we suggest that they may belong to a T-cell subset distinct from typical CTLs.

Activation of lymphocyte anti-tumour responses in man: effector heterogeneity and the search for immunomodulators

Data continues to accumulate on the immunological reaction against solid human cancers. The evidence at the present time supports the view that rather than being immunologically invisible, tumour cell antigens are recognised by at least three lymphocyte subsets. Helper T cells can be induced to proliferate upon exposure to cells of the autologous tumour and to secrete detectable levels of interleukin 2 (IL-2). Cultured T cell lines and clones can be shown to respond in primed lymphocyte tests not only to autologous tumour cells but also to allogeneic tumour cells of the same histology and anatomic location. Cytotoxic T cells manifest specific reactivity against cells of the autologous tumour which is distinguishable from natural killing (NK) on the basis of specificity and organ distribution. Natural killer cells can lyse freshly isolated autologous tumour cells after purification on Percoll gradients or when activated by IL-2. There is thus a demonstrable heterogeneity of response to human cancer in unseparated lymphocyte populations and at the clonal level. In limiting dilution assays lymphocytes at the tumour site respond more frequently to autologous tumour relative to NK targets. For at least some tumours there is evidence that the expression of auto-tumour reactivity but not NK correlates with the clinical course of the disease and is a favourable prognostic indicator. The finding of these auto-tumour reactivities has important implications for the search for immunomodulating drugs for cancer treatment. However, it must be recognised that the response is heterogeneous and that the immune system comprises multiple interactive elements that exhibit both positive and negative control. Any treatment modality must take this into account and seek to focus on specific activation of the tumour lytic populations or the inhibition of negative regulatory elements as opposed to seeking a more general augmentation of immune reactivity which may, by stimulating suppressor cells, have a counterproductive effect.

Modulation of spontaneous immunoglobulin production by natural killer cells in rheumatoid arthritis

Synovial fluid (SF) mononuclear cells obtained from patients with rheumatoid arthritis (RA) spontaneously produce large amounts of immunoglobulin. In the rheumatoid joint, natural killer (NK) cell activity is reduced in comparison with that in the peripheral blood (PB). We examined the ability of SF NK cells to modulate the spontaneous production of Ig in RA SF, and we contrasted this with the activity in PB from RA patients and from normal subjects. We found that the spontaneous production of IgG was greater in RA SF than in RA or normal PB. The baseline NK activity was significantly lower in RA SF than in RA or normal PB (P less than 0.005). Incubation with anti-Leu-11b and complement reduced NK activity in PB, but not in SF, and it significantly (P less than or equal to 0.021) increased IgG production in both RA SF and RA PB. Lysis of NK cells in this manner also resulted in a significant increase (P less than 0.02) in IgM production in RA SF. These results suggest that NK cells with a Leu-11b phenotype down-regulate the ongoing synthesis of IgG and IgM in the rheumatoid joint.

Lymphokine production in T gamma lymphoproliferative disorders

We have studied five patients with chronic lymphocytosis consisting of large granular lymphocytes (LGL). The increased numbers of LGL in these patients had little or no natural killer activity, mediated antibody-dependent cellular cytotoxicity, and were induced to kill tumour lines after culture for 3 days with interleukin 2 (IL-2). Patients' LGL showed considerable reactivity with HNK-1 and AB8.28 monoclonal antibodies (MoAb), whereas positivity for OKM1 and N901 was found in only two subjects, and only one patient reacted with B73.1. No appreciable reactivity has been found with anti-Tac MoAb in the four patients tested. In the absence of stimulation, the patients' LGL produced no IL-2 and only minimal amounts of IL-1 and interferon (IFN). On stimulation with lipopolysaccharides (for IL-1) or phytohaemagglutinin A (PHA) (for IL-2 and IFN), they produced IL-1 and IFN in amounts similar to those produced by normal lymphocytes, but only modest levels of IL-2. These results indicated that proliferating LGL, like normal LGL, have a secretory capacity. The lack of constitutive lymphokine production, the lack of Tac receptor expression, and the defect in IL-2 production after PHA stimulation do not support the hypothesis of an autocrine proliferation sustained by a known growth factor.

'Supernatural' killer cells

The observation that lymphocytes from healthy individuals, without any known sensitization, could spontaneously lyse tumour target cells in vitro(1,2) was first thought to be an in-vitro artefact. Now the effector cells, known as natural killer (NK) cells, are accepted as a clearly distinct albeit heterogeneous subpopulation of lymphocytes. Yet their lineage remains controversial and despite much interest in the role of NK cells in protection against infection and cancer, their biological significance is far fiom clear(3-5). In this article Gordon Burns and his colleagues discuss recent research on NK cells which has illuminated the diverse effects of lymphokines, expanded knowledge on the mechanisms of cell recognition and killing by cytotoxic effector cells, and illustrated how a variety of leukocytes mediate more than one function - results of general interest to cell biologists.