Rapid killing of actinomycin D-treated tumor cells by mononuclear phagocytes: characterization of effector cells in mice

Pretreatment with Actinomycin D (Act D, 1 microgram/ml for 3 hr) rendered WEHI 164 tumor cells susceptible to killing by mouse resident or peptone-induced peritoneal exudate cells (PEC) in a 6-hr 51Cr release assay. Cytotoxicity was attributed to cells of the monocyte macrophage lineage on the basis of tissue distribution, separation by adherence on plastic and carbonyl iron, membrane antigens, and expression in mice with defective T cell- or NK cell-mediated immunity. Macrophages from four strains of mice (C3H/HeJ, A/J, P/J, C57B1/10 ScCR) previously shown to have defective "classical" nonspecific tumoricidal activity were examined for killing of Act-D-treated WEHI 164 cells. C3H/HeJ peritoneal macrophages had little or no DDCC, whereas cells from A/J, P/J, and C57B1/10 ScCR mice had normal levels of this reactivity. Tumor cells exposed to ActD were heterogenous in their susceptibility to killing by PEC, with five lines showing significant, though variable, lysis, whereas 12 tumor lines, normal fibroblasts, and lymphoblasts were not appreciably killed under the same conditions. Macrophage-mediated DDCC was also detectable in a colony assay. DDCC could explain how macrophages contribute to the antitumor activity of selected chemotherapeutic agents in murine tumor models.     

Rapid killing of actinomycin D-treated tumor cells--cytotoxicity of cell-free monocyte supernatants

Human monocytes kill Actinomycin D-treated WEHI 164 sarcoma cells in a 6 h 51Cr release assay (drug dependent cellular cytotoxicity, DDCC). In the present study we have investigated and characterized the human monocyte production of a cytotoxic factor which mediates DDCC. Cell-free supernatants obtained culturing monocytes for 4-5 h kill Actinomycin D-treated WEHI 164 cells but not untreated tumor cells. A series of antiproteases inhibits the cytotoxic activity of cell-free monocyte supernatants, whereas scavengers of reactive oxygen intermediates were ineffective. The lytic activity was destroyed treating supernatants at 100 degrees C for 5 min or by exposure to acid pH or to proteinase K, whereas it was unaffected by heating at 56 degrees C for 30 min. Upon gel filtration on Sephacryl S200, cytolytic activity eluted in the 33,000 molecular weight range.     

Cytotoxicity on tumour cells of human mononuclear phagocytes: defective tumoricidal capacity of alveolar macrophages.

Human cells of the monocyte-macrophage lineage were isolated by adherence from peripheral blood, peritoneal exudate, non-neoplastic ascites, benign ovarian cystic fluid and bronchoalveolar lavages. Cytolytic activity was measured as 3H-thymidine release from prelabelled mKSA TU5 tumour cells over 48-72 hr and cytostasis was evaluated in a 72-hr spectrophotometric assay. Mononuclear phagocytes from the various anatomical sites examined, except lung alveolar spaces, were significantly cytolytic and cytostatic on target cells. Unlike other cells of the monocyte-macrophage lineage, alveolar macrophages were not cytocidal, but significantly inhibited tumour cell proliferative capacity. Peripheral blood monocytes and peritoneal macrophages showed enhanced cytotoxicity in the presence of partially purified human fibroblast interferon or of lymphokine supernatants from mitogen-stimulated lymphocytes. In contrast, interferon did not affect the cytotoxic potential of alveolar macrophages, whereas lymphokines augmented their cytostatic activity and rendered them weakly cytolytic.     

Antibody-dependent and -independent cytotoxicity of human mononuclear phagocytes: defective stimulation of tumoricidal activity in milk macrophages.

Human peripheral blood monocytes, milk macrophages and peritoneal exudate macrophages were purified by adherence. antibody-dependent cellular cytotoxicity (ADCC) was measured using the murine TLX9 lymphoma pre-labelled with 3H-thymidine. Direct, antibody-independent tumoricidal activity was measured against the murine TU5 line pre-labelled with 3H-thymidine. All mononuclear phagocyte populations tested were similarly effective in mediating ADCC against TLX9 cells. In the absence of deliberate stimulation blood monocytes and peritoneal macrophages had appreciable spontaneous cytotoxicity against the susceptible TU5 line. In contrast, four out of 10 milk macrophage preparations lacked detectable spontaneous killing activity on this target. In vitro exposure to partially purified fibroblast interferon (IFN) or to lymphokine supernatants from PHA stimulated lymphocytes augmented the direct tumoricidal activity of blood monocytes and peritoneal exudate macrophages. Milk macrophages were completely unresponsive to IFN and lymphokines. therefore the capacity to mediate antibody-dependent and -independent cytotoxicity against tumour cells can be dissociated to some extent in human mononuclear phagocyte populations from diverse anatomical sites.     

Cytolysis of actinomycin D-treated target cells by cell-free supernatants from human monocytes

The lytic mechanism of human peripheral blood monocytes was studied by using as targets actinomycin D-treated WEHI-164, an NK-insensitive murine fibrosarcoma cell line. Monocytes, but not lymphocytes, lysed WEHI-164 target cells pre-treated with actinomycin D within 6 h in 51Cr-release assays. Because cytolysis could not be inhibited competitively by unlabeled WEHI/D target cells, contact-independent mechanisms of cytolysis were investigated. Cell-free supernatants collected from monocytes cultured for 4-6 h at 37 degrees C lysed target cells as effectively as effector cell preparations of monocytes. Supernatants from lymphocytes cultured in parallel were not cytolytic. Cytolytic activity was not detected in supernatants from preparations of monocytes that were held on ice. However, monocytes produced cytolytic activity whether they were isolated by adherence or remained unseparated in suspensions of mononuclear cells. The cytolytic activity in cell-free supernatants (CFS) from monocytes was unaffected by incubation with protease inhibitors. CFS activity was destroyed by heat. Storage of CFS at 37 degrees, 22 degrees, 4 degrees, or -20 degrees C for 24 h decreased cytolytic activity; however, loss of cytotoxicity was minimized by storage at 4 degrees C. The cytolytic substance detected in 4-h CFS from monocytes appeared to be a protein(s) based on the sensitivity of the cytolytic activity to proteases. Cytolytic activity of CFS eluted from Sephacryl 200 in a single peak with an apparent molecular weight between 25,000 and 45,000 daltons.     

Interactions of Escherichia coli and Proteus mirabilis with mouse mononuclear phagocytes

Five strains of enterobacteria (three of Escherichia coli and two of Proteus mirabilis) were studied to assess and compare their phagocytic uptake and intracellular killing by mouse macrophages. Each strain was injected intraperitoneally into separate groups of mice and peritoneal exudate cells were harvested after 3 min for phagocytosis to occur in vivo. Acridine orange staining showed that there were approximately 10-fold fewer intracellular P. mirabilis than E. coli cells. The average numbers of viable intracellular bacteria per leucocyte were 0.03 and 0.02 for P. mirabilis strains M13 and H1, respectively, and 0.48, 0.45, and 0.28 for E. coli strains M14, A-D M5 and H40. Thus, both P. mirabilis strains were ingested less readily than any of the three E. coli strains (p less than 0.01). The rates of in-vitro intracellular killing were similar for all five strains of bacteria. The intracellular killing constants (Kk) for the three mouse isolates were 0.017, 0.016 and 0.020 min for E. coli M14 and A-D M5, and P. mirabilis M13, respectively; the Kks for the two human isolates were 0.026 and 0.029/min for E. coli H40 and P. mirabilis H1, respectively. The Kks for all five strains were not significantly different. Assuming that the numbers of viable intracellular bacteria at the beginning of the assay represented 100% viability, 6-17% of the intracellular bacteria remained viable after 2 h, reflecting log10 3.9-5.6 bacteria (6-8) x 10(6) peritoneal exudate cells. Intravenous injection of these five strains into separate groups of mice demonstrated that the P. mirabilis strains were more virulent than the E. coli strains. Injection of each P. mirabilis strain was associated with ruffled fur and death, whereas mice given any of the three E. coli strains remained visibly healthy and none died. Consistent with these observations, quantitation of viable bacteria in the liver and spleen showed that greater numbers of P. mirabilis M13 than of E. coli M14 or A-D M5 persisted in these organs; similarly greater numbers of P. mirabilis H1 than of E. coli H40 persisted in the liver and spleen. Because the rates of intracellular killing of these five strains were similar, the relative virulence of both strains of P. mirabilis appeared to be associated with decreased phagocytic uptake rather than differences in intracellular survival.     

Involvement of tumour necrosis factor in monocyte-mediated rapid killing of actinomycin D-pretreated WEHI 164 sarcoma cells.

Human and murine monocyte-macrophages kill actinomycin D (ActD)-treated WEHI 164 sarcoma cells in a 6-hr 51Cr-release assay (drug-dependent cellular cytotoxicity, DDCC). In this study, we have investigated the cytotoxic activity of human recombinant tumour necrosis factor (hrTNF) against untreated and ActD-treated WEHI 164 sarcoma cells. Human recombinant TNF when added to the 6-hr 51Cr-release assay killed ActD-treated targets at doses ranging from 33 to 0.33 ng/ml, whereas untreated targets were resistant to lysis. The kinetics of lysis of ActD-treated targets was similar for hrTNF and blood monocytes. The protease inhibitors phenyl-methyl-sulphonyl-fluoride (PMSF) and N-alpha-p-tosyl-L-lysine chloromethyl ketone (TLCK) reduced the DDCC activity of monocytes, monocyte supernatants and hrTNF. Killing of drug-sensitized target cells by monocyte supernatants was totally inhibited by a rabbit anti-TNF serum. These, as well as previous data on the physicochemical properties of the soluble cytotoxic factor released by monocytes, suggest that rapid monocyte-mediated killing of ActD-pretreated WEHI 164 sarcoma cells involves TNF or TNF-like molecules.     

Phagocytosis and killing of bacteria by professional phagocytes and dendritic cells

Dendritic cells (DC) represent a class of professional antigen-presenting cells whose primary function is to alert the immune system, not to clear invading microorganisms. The objective of our study was to compare the abilities of polymorphonuclear neutrophilic leukocytes (PMN), monocytes, monocyte-derived macrophages (MDM), monocyte-derived immature DC (imDC), and mature DC (maDC) to ingest and destroy Staphylococcus aureus and Escherichia coli. Acridine orange staining and fluorescence microscopy demonstrated that MDM, followed by monocytes, imDC, and PMN, internalized bacteria well but that maDC exhibited less pronounced phagocytic activity. PMN, monocytes, and MDM exhibited a much higher capacity to kill ingested bacteria than both imDC and maDC. In summary, these data are in agreement with the generally accepted idea that different types of leukocytes fulfill specialized tasks in antigen presentation and killing of pathogens.     

Ultrastructure of mouse mononuclear phagocytes in bone marrow colonies grown in vitro.

Recently a new method was developed to culture bone marrow cells in a liquid medium on a glass surface. Two kinds of colonies develop in these cultures, namely mononuclear phagocyte and granulocyte colonies. The study of the ultrastructure of the cells in the mononuclear phagocyte colonies was the primary aim of the present study. The architecture of the mononuclear phagocyte colonies appeared to be quite different from that of the granulocyte colonies, since in the latter, the cells lie close together in dense clusters, whereas in mononuclear phagocyte colonies the cells are more loosely dispersed with the highest cell density at the center and stellate orientation of the cells at the periphery. However, both kind of colonies grow entirely separate from each other and mixed colonies are not observed. Electron microscopy showed that there are three types of cell in the mononuclear phagocyte colonies, i.e., monoblasts, promonocytes, and macrophages. The ultrastructure of the promonocytes and macrophages of the colonies is similar to that of the same types of cell isolated directly from the mouse. The monoblast, the most immature cell of mononuclear phagocyte colonies has not been characterized before. The ultrastructure of this cell is clearly distinct from that of the promonocyte in having a round contour without pseudopods, a nuclear to cytoplasmic ratio greater than one, a cytoplasm that contains many polyribosomes, a few small granules, and a small Golgi complex surrounded by a few short strips of rough endoplasmic reticulum.     

Immunohistological analysis of human mononuclear phagocytes and dendritic cells by using monoclonal antibodies

Mononuclear phagocytes and dendritic cells were analyzed in detail with a panel of monoclonal antibodies by immunoalkaline phosphatase and immunofluorescence techniques. The panel included two antibodies (KB90 and SHCL3) against the two chain membrane constituent, P150,95, an antibody (Mo1) against C3bi receptor, an antibody (To5) against C3b receptor, and two antibodies (Mo2 and UCHM1) against a single chain glycoprotein with a molecular weight of 55 to 60 previously found on monocyte/macrophages. In addition, the panel contained several antibodies including EBM11, LeuM3 and FMC32 whose molecular specificities are not known at present. Frozen sections of spleen, bone marrow, thymus, lymph node, liver, brain, lung, kidney, colon, and skin were examined. Several populations of mononuclear phagocytes and dendritic cells could be distinguished on the basis of their reactivity with the antibody panel. Populations of macrophages resident in parenchymal organs had reaction patterns distinct from those of circulating macrophages. Many of the antibodies in the panel reacted with both macrophages and dendritic cells suggesting a close cytogenic, if not functional, relationship between these cells. There was wide variability in the reactions of macrophages and dendritic cells in different tissue and less variability among cells within individual anatomic locations.     

Separation of natural and antibody-dependent cytotoxicity by differential reactivity of human mononuclear cell Fc receptors with IgG.

Nylon wood non-adherent, human peripheral blood mononuclear cells which are reactive with the OKM1 monoclonal antibody could be separated into two subpopulations based on their Fc receptor reactivity with human monomeric IgG (FcR-IgG) using flow cytometry. The majority of natural killer cells was found primarily in the OKM1+ subset with low FcR reactivity. In contrast, effector cells for antibody-dependent cytotoxicity (K cells) were found in subsets with both high and low FcR-IgG reactivity.     

Divergent effects of interleukin-10 on cytokine production by mononuclear phagocytes and endothelial cells

Interleukin-10 (IL-10), a product of T helper type 2 (TH2) cells and monocytes, inhibits cytokine production in mononuclear phagocytes. Given the similarities and interrelationship between cells of the monocyte-macrophage lineage and endothelial cells, we examined the effects of IL-10 on vascular endothelium. Murine IL-10 induced low levels of IL-6 production and amplified induction of IL-6 by lipopolysaccharide (LPS) or IL-1 in the murine tEND.1 endothelioma line, used for these studies because it retains properties of normal endothelium. The effect was more evident after prolonged (48–72 h) exposure to IL-10. IL-10 had similar activity on other endothelioma lines, whereas it inhibited IL-6 production by peritoneal macrophages. Induction and amplification of cytokine production by IL-10 was associated with higher levels of mRNA, which were maintained longer (up to 48 h) than in controls. In addition to IL-6, murine IL-10 induced or amplified expression of the chemoattractant cytokines monocyte chemotactic protein-1 (MCP-1) and KG Human IL-10 inhibited IL-6 release by LPS-stimulated human peripheral blood mononuclear cells, whereas it did not interfere with cytokine production by LPS- or IL-1-stimulated human umbilical vein endothelial cells. The selective inhibitory action of IL-10 on mononuclear phagocytes versus endothelial cells may play a role in the pathophysiology of TH2-directed responses.     

Increased IgE-dependent cytotoxicity by blood mononuclear cells of allergic patients.

Peripheral blood mononuclear cells from 14 healthy donors and 22 allergic patients were incubated with 51Cr-labelled chicken erythrocytes coated with an IgE myeloma protein or rabbit IgG antibodies. Mononuclear cells from patients with severe atopic disorders released a significantly greater percentage of 51Cr (P less than 0.001) from IgE-coated target cells than mononuclear cells from healthy controls, patients with mild atopic disease, or patients with severe atopic disease taking oral prednisone. Specific 51Cr-release from IgE-coated target cells was directly correlated to the percentage of monocytes (latex-ingesting cells) with Fc receptors for IgE (r = 0.87, P less than 0.01) as detected by a rosette assay employing ox erythrocytes coated with IgE. Mononuclear cells from patients and normals released similar amounts of 51Cr from IgG-sensitized target cells. Depletion of monocytes from mononuclear cell preparations from two severe atopic patients decreased 51Cr-release from IgE-coated target cells to levels seen in healthy donors or patients with mild allergic disease. These results demonstrate that mononuclear cells from severely allergic patients have a significantly increased cytotoxicity toward IgE-coated targets coated target cells and that this cytotoxicity correlates highly with the percentage of monocytes with Fc receptors for IgE in these mononuclear preparations.     

Reduced LAK cytotoxicity of peripheral blood mononuclear cells in patients with bladder cancer: Decreased LAK cytotoxicity caused by a low incidence of CD56+ and CD57+ mononuclear blood cells

The cytotoxicity of unstimulated peripheral blood mononuclear cells (US-PBMC), phytohemagglutinin (PHA)-stimulated PBMC (PS-PBMC) and interleukin-2 (IL-2)-activated PBMC (LAK cells) was assessed in patients with noninvasive and invasive transitional-cell bladder cancer and compared with those determined in healthy controls. The differences in the cytotoxicities were correlated with specific changes in the subsets of peripheral blood mononuclear cells (PBMC). PBMC from 37 patients and 13 healthy controls were tested against the bladder cancer cell line T24 in⁵¹Cr-release assays. The PBMC subsets were analyzed using monoclonal antibodies against T cells, natural killer (NK) -cells, monocytes, and activation markers. The cytotoxicities of US-PBMC, PS-PBMC, and LAK cells were all significantly lower in the cancer patients than in the controls (P<0.05). The percentages of PBMC positive for the NK-cell markers CD56 and CD57 were lowest in the patients and were correlated to the decrease in cytotoxicity. Depletion of CD56⁺ or CD57⁺ cells from PBMC prior to or after 2 days stimulation with IL-2 demonstrated that these cells are the major source of LAK-cell cytotoxicity and showed that the reduced ability of bladder cancer patient PBMC to develop LAK-cell cytotoxicity is a result of a low incidence of CD56⁺ and CD57⁺ cells in the blood. These findings indicate that IL-2 therapy alone might not be a sufficient therapy of bladder cancer patients.     

Neutrophil--mediated cytotoxicity against tumour cells: state of art.

When activated by a variety of stimuli, human neutrophils become capable of lysing tumour cells. The main physiologic cytotoxic trigger in neutrophils is represented by the surface receptor for the Fc domain of IgG (FcR). This receptor confers specificity to the target recognition by neutrophils and cooperates with adhesion glycoproteins (CD11-CD18) in the neutrophil-target conjugate formation. Although the FcR-dependent pathways of signal transduction remain to be clarified, the neutrophil responses involve both the production of oxidants and the release of granule constituents. These molecules are responsible for the target lysis, whose extent also depends on the target structural and metabolic characteristics. The use of monoclonal antibodies specific for tumour cell antigens, coupled with appropriate cytokines, may provide rational basis for designing trials to employ the neutrophil cytotoxic potential as adjuvant therapy in cancer patients.     

Rapid killing of monocytes in vitro by Candida albicans yeast cells.

To study the interaction between Candida albicans blastoconidia and human phagocytes, we incubated peripheral leukocytes with fungi for 1 h at 37 degrees C and stained the cells with fluorescent vital stains ethidium bromide (EB) and fluorescein diacetate. Fungi that had been phagocytosed showed little staining; however, some leukocytes containing blastoconidia exhibited nuclear staining with EB, even though their cell membranes showed no signs of penetration by fungi. The number of EB-positive leukocytes was related to viability of the yeast cells and the temperature at which they were maintained before use. Because efforts to quantitate EB-positive leukocytes microscopically were frustrated by cell aggregation, we labeled the leukocytes with 51Cr and measured isotope release. We determined that leukocytes incubated with viable fungi released significantly more isotope than cells incubated alone or with killed blastoconidia. Furthermore, 51Cr release correlated directly with concentration of fungi in the assay, time of incubation, and temperature at which fungi were maintained before use. Using a number of isolates of C. albicans and several other species of Candida, we found that all exhibited cytotoxic activity against leukocytes, but the level of activity varied among organisms. Finally, we depleted or enriched peripheral leukocytes for specific cell populations and determined that only monocytes released more 51Cr after incubation with viable blastoconidia. Blastoconidia can lyse phagocytic cells through germination and penetration of cell membranes within 1 to 2 h, but the cytotoxic phenomenon we describe occurs within 15 to 30 min after yeast cells have been phagocytosed. Therefore, this capacity may represent a more immediate response by blastoconidia against phagocytosis and killing by monocytes.     

Generation of procoagulant activity by mononuclear phagocytes: optimisation of an in vitro assay

Leucocyte procoagulant activity (PCA) has previously been advocated as a useful measurement of cell-mediated immune reactivity. The assay is, however, susceptible to inter- and intra-experimental variation. This investigation identifies several factors which influence the stability and reproducibility of the test. Major factors which affect the recalcification time of plasma, include plasma lability, medium/buffer pH and both the nature and concentration of the indicator cells used in the assay. C. parvum-induced mouse peritoneal exudate cells have been used as a novel source of mononuclear phagocytes in the generation of PCA. Their sensitivity as indicator cells has been demonstrated by their responsiveness to stimulation by phytomitogen, endotoxin, and lymphokine (macrophage procoagulant-inducing factor). A simple test based on antigen-induced PCA of these cells, has provided an in vitro index of in vivo sensitisation to sheep red blood cells. Optimisation and standardisation of conditions detailed in this report for estimating PCA, renders the assay of value in monitoring lymphocyte and macrophage activation.     

Specific killing of mouse leukemic cells with ricin A-chain immunotoxin

Monoclonal IgM antibody against L1210V leukemia was coupled with ricin A-chain using N-succinimidyl 3-(2-pyridyldithio) propionate (SPDP) as a cross--linking agent. The coniugate had potent concentration--dependent cytotoxicity against L1210V, L1210 and RL male 1 cells being completely non toxic to EL-4, P388, RPC-5 and mouse bone marrow cells. The minimum time required for killing L120V leukemia cells was 30h of in vitro exposure, at a concentration 10(-6) M (as assessed by trypan blue test). However, 1h contact of L1210V cells with immunotoxin was sufficient to completely inhibit proliferation of leukemic cells subsequently inoculated into compatible mice. The toxicity could be potentiated by addition of NH4Cl, that shortened minimum exposure time to 18h and 45 min respectively.     

Activation of cytotoxic lymphocytes by interferon-alpha: role of oxygen radical-producing mononuclear phagocytes

A significant part of the therapeutic benefit of interferon-alpha (IFN-alpha) therapy in malignant diseases and in chronic viral infections is assumed to result from activation of lymphocytes with natural killer (NK) and T cell phenotype. In tumor tissue and in chronically infected tissue, the function and viability of these lymphocytes are frequently impaired. Mononuclear phagocyte (MP)-derived reactive oxygen species (ROS) have been proposed to contribute to the lymphocyte suppression in these tissues. Here, we report that three types of human cytotoxic lymphocytes of relevance to immunoactivation by IFN-alpha, CD3epsilon+/8+/56- T cells, CD3epsilon-/56+ NK cells, and CD3epsilon+/56+ NK/T cells became anergic to IFN-alpha induction of the cell-surface activation marker CD69 after exposure to autologous MPs in vitro. In addition to their incapacity to express CD69, cytotoxic lymphocytes acquired features characteristic of apoptosis after incubation with MPs. The lymphocyte apoptosis and nonresponsiveness to IFN-alpha were prevented by two inhibitors of reduced nicotinamide adenine dinucleotide phosphate oxidase-dependent formation of ROS in MPs, histamine dihydrochloride and diphenylene ionodonium, as well as by catalase, a scavenger of ROS. We conclude that MP-derived ROS may negatively affect IFN-alpha-induced immunostimulation and propose that ROS inhibitors or scavengers may be useful to improve lymphocyte activation during treatment with IFN-alpha.