Macrophage-mediated destruction of malignant tumor cells and new strategies for the therapy of metastatic disease

Inhibition of liver metastasis in mice by blocking hepatocyte lectins with arabinogalactan infusions and D-galactose

According to our hypothesis, organ-specific lectins (e.g., the D-galactose-specific hepatic binding protein) play an important role in the organ location of metastatic malignant cells. The rapid clearance and uptake by the liver of tritiated alpha 1-acid-(asialo)glycoprotein from the circulation of Balb/c mice was markedly delayed after preinjection of D-galactose or arabinogalactan. The preinjection (1 h) and regular application (for 3 days after tumor cell inoculation in Balb/c mice) of the receptor blocking agents D-galactose and arabinogalactan prevented the settling of sarcoma L-1 tumor in the liver completely, but did not influence the settling in the lung. Other galactans, dextrans, and phosphate-buffered saline showed no effect. Therefore, when lectins were blocked with competitive-specific glycoconjugates, colonization was prevented.

Characterization of liposomes containing the chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine (FMLP) and their interaction with mouse macrophages

The purpose of these studies was to characterize liposomes containing the chemotactic peptide N-formyl-methionyl-leucyl-phenylalanine (FMLP) and to examine their interaction with mouse peritoneal macrophages. Because of its hydrophobic properties, FMLP can be readily incorporated into both phosphatidylcholine (PC) and PC/phosphatidylserine (PS) (7:3, mol ratio) multilamellar vesicles (MLV). The association of FMLP with the MLVs is stable in medium with or without serum. The size distribution and aqueous trap volume of MLV consisting of PC or PC/PS that contain FMLP at a 10:1 molar ratio do not differ from that of MLV without FMLP. MLV containing equimolar phospholipid-to-FMLP ratios are smaller but have larger aqueous trap volumes. MLV consisting of PC and FMLP (10:1 molar ratio) are phagocytosed more efficiently than PC MLV, and their phagocytosis is mediated via the FMLP receptors on macrophages. In contrast, MLV consisting of PC/PS or PC/PS with FMLP (7:3:1 molar ratios) are phagocytosed to the same extent, and so the FMLP receptors on macrophage surface do not influence this process. The delivery of FMLP in liposomes to macrophages is very efficient: up to 25-fold more FMLP delivered by MLV is internalized into macrophages than when it is in free form. Once associated with the macrophages in either free form or via liposomes, some FMLP is degraded and released into the medium, but the majority of FMLP molecules remain with the macrophages to be degraded at a steady rate. The present findings may explain how liposome-FMLP produces tumoricidal properties in macrophages.

Recruitment of exogenous macrophages into metastases at different stages of tumor growth

The endogenous tumor-associated macrophage content and recruitment of labeled peritoneal exudate cells into experimental murine B16 melanoma metastases has been examined at different stages in the progressive growth of metastatic lesions. The recruitment of thioglycollate-elicited peritoneal exudate cells and peritoneal exudate cells activated in vitro with muramyl dipeptide was studied. Tumor-associated macrophages and labeled peritoneal exudate cells were identified in paraffin sections by specific histochemical staining and their density in individual metastases measured morphometrically. The density of tumor-associated macrophages and exogenously recruited peritoneal exudate cells was high in very small lesions but decreased rapidly as a function of enlargement of metastases, MD:An; where MD is macrophage density, A is the cross-sectional area of the lesion and n is a negative number. No significant difference was observed in the recruitment of activated and nonactivated peritoneal exudate cells. These results suggest that decreased recruitment of macrophages from the circulation may explain the decrease in the density of tumor-associated macrophages as metastases grow and indicate that macrophage activation is not accompanied by enhanced localization and/or uptake of macrophages into metastases.

Macrophage infiltration in non-Hodgkin's lymphomas: a quantitative in situ study

The frequency and distribution pattern of macrophages within 93 non-Hodgkin's lymphomas (NHL) were evaluated in situ by immunomorphometry using stereological methods. For the identification of macrophages (M phi), several antibodies (Mono 1, Mono 2, OKM 1) reactive with surface antigens on cells of the monocyte-macrophage series and cytochemical staining for acid phosphatase were applied. The average number of macrophages within lymph node tissue of NHL was 6,299 +/- 760 cells/microliter (similar to reactive lymphatic tissue: 6,559 +/- 1,027). The highest number of infiltrating macrophages was detected in immunoblastic NHL (17,306 +/- 2,773), differing significantly from other histological subtypes and reactive lymphatic tissue (p less than 0.005). The possible impact of tumor-infiltrating macrophages on lymphoma cell proliferation and differentiation is discussed.

Head circumference, brain weight, and tumor burden

The failure to diagnose an ependymoma at an appropriately early age led to an incomplete excision and a tumor burden too great for radiotherapy to control. The development of normative curves for brain weight correlated with head circumference allowed for the estimates of the extremes of possible growth rates of the tumor. The importance of early diagnosis and combined surgical and radiotherapies is emphasized since ependymomas are locally recurrent but curable by x-irradiation (and host macrophages?) if the tumor burden is small.

Antagonism of interferon induction in spleen and adherent peritoneal cells of mice by the lipophilic antiviral muramyl peptide MTP-PE

The immunomodulator muramyl tripeptide-phosphatidylethanolamine (MTP-PE) has been shown to enhance host resistance against a variety of experimental infections and to cure influenza virus infection in mice when given in a single dose, even at a late stage of the disease. Tests of its capacity to induce alpha/beta- and gamma-interferon (IFN-alpha/beta and -gamma) in vitro demonstrated that it is neither an inducer nor a primer of IFN synthesis. On the contrary, we found that it inhibits the induction of IFN-alpha/beta and -gamma by poly(rI:rC), Newcastle disease virus, lipopolysaccharide, or concanavalin A in adherent cells from the peritoneal cavity and spleen of mice. The antiviral activity of already induced or exogenously added murine IFN was, however, not impaired.

Optimization and limitations of systemic treatment of murine melanoma metastases with liposomes containing muramyl tripeptide phosphatidylethanolamine

The purpose of these studies was to determine the optimal conditions and limitations for the eradication of spontaneous melanoma metastases by the systemic administration of liposomes containing MTP-PE. Mice whose primary melanoma had been excised were given i.v. injections of liposomes at various schedules. Optimal treatment was achieved by twice weekly administration for 4 weeks (eight i.v. injections). Bioassays failed to reveal the presence of melanoma cells in lungs of mice surviving to day 250 of the experiment. The success of liposome treatment of metastases diminished when the first i.v. injection of liposomes-MTP-PE commenced on day 10 after surgical excision of the local melanoma, as compared with day 3 or day 7 after surgery. We conclude that the major limitation for macrophage-mediated destruction of metastases is the number of tumor cells in the lesions. Because of this limitation, it is unlikely that the systemic activation of macrophages could be used as a single modality for treatment of advanced metastases.

Augmentation of NK activity and/or macrophage-mediated cytotoxicity in the liver by biological response modifiers including human recombinant interleukin 2

Administration of several biological response modifiers (BRMs) to mice strongly augmented natural killer (NK) activity of leukocytes isolated from the liver. This augmentation of NK activity was induced by two synthetic molecules (MVE-2 and poly ICLC), by two BRMs of bacterial origin (formalin-fixed Propionibacterium acnes: P. acnes and a streptococcal cell wall preparation designated OK-432), as well as a single injection of human recombinant interleukin-2 (hrIL 2). All of these BRMs augmented NK activity in the liver to a greater degree than in the spleen. In addition, adherent leukocytes (greater than 90% macrophages) isolated from the liver following P. acnes administration also exhibited augmented macrophage-mediated cytotoxicity. This cytotoxicity was characterized as macrophage mediated and distinguished from NK activity, on the basis of adherence purification, kinetics of cytotoxicity, and target cell selectivity. The results demonstrate that a variety of BRMs induce augmented natural immunity in the liver and suggest that such organ-associated immune responses may play an important role in the antimetastatic effects of BRMs.

A microassay for the rapid and selective binding of cells from solid tumors to mouse macrophages

A microassay was developed to study the rapid binding characteristics of murine macrophages activated by gamma interferon and muramyl dipeptide to adherent neoplastic or nonneoplastic target cells. The binding of tumor cells to both activated and nonactivated macrophages was time- and temperature-dependent, and independent of tumor cell type. Activated macrophages bound more tumor cells than nonactivated macrophages. The initial binding of macrophages to target cells did not necessarily lead to lysis. First, primed macrophages bound tumor cells but did not lyse them, and second, nonactivated macrophages bound nontumorigenic cells without subsequent lysis. The rapid binding assay described here could prove useful in investigating the recognition mechanism(s) between macrophages and tumor cells derived from solid primary and metastatic cancers.

Influence of adoptively transferred thioglycollate-elicited peritoneal macrophages on metastasis formation in mice with depressed or stimulated NK activity

The effect of thioglycollate-elicited macrophages (TG-M phi) on natural killer (NK)-cell activity and metastases formation in mice was investigated. Intravenously (i.v.) inoculated TG-M phi inhibited spleen NK activity of normal mice and abrogated polyinosinic: polycytidylic (poly I:C) induced augmentation of NK cell function. TG-M phi also inhibited the clearance of i.v.-injected radiolabeled B16 melanoma cells from the lungs of normal or poly I:C stimulated mice. Formation of experimental B16 melanoma metastases was dramatically increased in mice pretreated with TG-M phi. Administration of TG-M phi increased metastasis formation to a greater extent than anti-asialo GM1 serum, while anti-asGM1 serum was more efficient than TG-M phi in depressing spleen NK cell activity. When mice with low NK reactivity (beige mice or mice treated with anti-asialo GM1 serum) were inoculated with TG-M phi, there was a substantial additive augmenting effect on metastasis formation in the lungs. Treatment with poly I:C elevated NK-cell activity and had profound antimetastatic effects in normal but not in TG-M phi pretreated mice. The metastasis augmenting effect of TG-M phi was fully expressed in poly I:C-treated mice as well as in athymic nude mice. Inoculation of proteose peptone-elicited macrophages (PM phi), unlike TG-M phi, did not depress NK activity or augment metastasis formation in normal or poly I:C-treated mice. However, since the inhibition of NK activity in TG-M phi-treated mice was relatively weak, and a substantial additional increase in metastases was observed in NK-depressed mice after transfusion of TG-M phi, it seems unlikely that the TG-M phi-induced inhibition of NK reactivity is entirely responsible for the augmented formation of metastases. Further studies revealed that i.v. inoculation of TG-M phi, but not PM phi, induced intravascular inflammatory reactions, and damage to endothelial cells and basement membrane of the lung vasculature. These reactions may contribute to increased tumor cell extravasation and metastasis formation in mice pretreated with TG-M phi.

Inhibition of growth and metastasis of syngeneic transplantable tumours by an aromatic retinoic acid analogue. 2. T cell dependence of retinoid effects in vivo

An aromatic retinoic acid analogue (Ro 10-9359) previously shown to be capable of inhibiting the growth and metastasis of immunogenic sarcomas and carcinomas (see accompanying paper) was tested for its anti-tumour effects in various categories of immune-deprived mice. 'Non-specific' immunosuppression evoked by sub-lethal whole-body X-irradiation abolished the inhibition of tumour growth induced by Ro 10-9359 in immunocompetent syngeneic hosts. Also, retinoid treatment of three categories of T-lymphocyte-deprived mice (nu/nu; thymectomized-irradiated; and cyclosporin A-treated) was ineffective in reducing the local growth rate or inhibiting spontaneous metastasis of their tumours; in fact, regardless of retinoid treatment the tumours grew faster and metastasized more widely in immunosuppressed animals than in controls. Silica and carrageenan (which are toxic to mononuclear phagocytes) did not interfere with the inhibitory effects of Ro 10-9359 on tumour growth, and did not themselves potentiate metastasis; however, both agents prevented the abolition of DM6 carcinoma metastasis by retinoids. APD, which inhibits the 'accessory cell' function of macrophages did not reduce the effectiveness of Ro 10-9359 against local tumours. However, in contrast to silica and carrageenan this agent did increase the incidence of metastasis of DM6 carcinoma from 40% to 60%, but in the presence of retinoids only 20% of mice succumbed to secondary disease. These results suggest an essential role for T lymphocytes in retinoid-induced local tumour growth inhibition, and a further contribution of mononuclear phagocytes to the prevention of metastatic disease.

Therapy of peritoneal murine cancer with biological response modifiers

We have used a murine renal adenocarcinoma of spontaneous origin (Renca) inplanted in the peritoneal cavity to study the therapeutic potential of biological response modifiers (BRMs) used alone or in conjunction with chemotherapy. This tumor model is therapeutically challenging since following intraperitoneal (i.p.) injection, the tumor grows progressively with hemorrhagic ascites, abdominal metastases to lymph nodes, liver, spleen, most serous membranes, and, in some animals, metastases to extra-abdominal sites (lungs). In the absence of therapy, death invariably occurs within 36 +/- 2 days. The tumor is efficiently lysed in 4 hours by peritoneal cells isolated from mice treated with BRMs. Both MVE-2 and rIL-2 significantly increased the survival time of tumor-bearing mice, but only treatment with MVE-2 led to definite cures of i.p. Renca. A single i.p. injection of MVE-2 cured 20% of the tumor-bearing mice, while repeated i.p. administration of this drug at 12 day intervals cured 70% of i.p. Renca-bearing mice. Combined therapy with doxorubicin hydrochloride and a single dose of MVE-2 cured 90% of tumor-bearing animals. The superior therapeutic efficiency of MVE-2 compared to that of the rIL-2 may be due to its ability, after i.p. inoculation, to generate and maintain high levels of cytotoxic effector cell activity for an elevated period of time within the peritoneal cell population. Additionally, MVE-2 augments effector cell activity in the liver, lungs, spleen, and blood and may therefore more efficiently interfere with metastasis formation in those compartments. The additive effects of MVE-2 and the chemotherapeutic agent suggest that more effective therapy may be achieved by the combination of immunotherapy with BRMs with chemotherapeutic drugs.

Induction of tumouricidal leucocytes by the intranasal application of MTP-PE, a lipophilic muramyl peptide

Single intranasal applications of MTP-PE, a lipophilic muramyl peptide, induce tumouricidal and tumouristatic leucocytes in the lungs of rats. In ex vivo assays the tumouristatic activity was detectable for 8 days after drug administration. By separation of the effector cells on Ficoll-Hypaque gradients, it was shown that both neutrophils and macrophages are responsible for this activity. Using the B16/BL6 melanoma system in mice, there was a high survival rate after repeated intranasal applications of MTP-PE.

The in situ activation of cytotoxic properties in murine Kupffer cells by the systemic administration of whole Mycobacterium bovis organisms or muramyl tripeptide

We determined whether the systemic administration of viable Mycobacterium bovis organisms (BCG) or a lipophilic derivative of muramyl tripeptide (MTP-PE) would lead to the activation of antitumor properties in murine Kupffer cells (KC). KC-mediated tumor cytolysis was determined by the release of radiolabeled nuclear breakdown products of target cells. KC harvested from either C57BL/6 or C3H/HEN mice treated with saline exhibited no cytotoxicity against syngeneic B16 melanoma or UV-2237 fibrosarcoma cells. In contrast, KC harvested from BCG or MTP-PE-injected mice were highly cytotoxic against the tumor targets, as measured by an in vitro radiorelease assay. The demonstration that the administration of macrophage activators can generate in situ tumoricidal activity in KC suggests that these cells can be important in the control of hepatic micrometastases.

Minactivin: a human monocyte product which specifically inactivates urokinase-type plasminogen activators

Culture supernatants from monolayers of human peripheral monocytes strongly inhibited colorimetric assays of urokinase in which plasmin was measured by esterolysis. This inhibitory activity of monocyte culture supernatant was enhanced after culture with muramyl dipeptide. Inhibition was specific for plasminogen activators of Mr 52 000 and 36 000, as shown by three methods: (1) inhibition of plasminogen-dependent fibrinolysis; (2) inhibition at the level of plasminogen activation in a colorimetric assay; (3) the irreversible loss of plasminogen-activating activity, as evidenced by electrophoresis, after preincubation with culture media. The factor responsible for this inactivation (which we propose to call minactivin) had an apparent Mr of 66 000 on Sephacryl S300 gel chromatography and interacted with enzyme in a biphasic manner: a rapid partial inhibition (reversible by sodium dodecyl sulphate) was followed by slow inactivation (irreversible by sodium dodecyl sulphate). It is proposed that secretion of minactivin by monocytes may contribute to regulation of extracellular proteolysis at sites of tissue injury.

Antitumor monoclonal antibodies for radioimmunodetection of tumors and drug targeting

Developments in hybridoma technology leading to the production of monoclonal antibodies recognizing tumor-associated antigens are providing new approaches for the radioimmunodetection of, and drug targeting to, metastases. These developments are illustrated in a series of studies on the in vivo localization of an antihuman-osteogenic sarcoma monoclonal antibody (791T/36) in human tumor xenografts maintained in immunodeprived mice. 131I-labelled 791T/36 antibody localized specifically in osteogenic sarcomas but not in xenografts of other tumors, such as bladder carcinoma T24, which do not express the antigen identified by this antibody. Developing from these studies, various parameters influencing antibody localization in tumors were examined including the kinetics of antibody uptake, the relationship between tumor size and antibody binding, and the site of antibody deposition. This provides a basis for considering the potential of antitumor monoclonal antibodies for targeting antitumor agents. Of particular importance here is the observation that antibody is principally located at the periphery of tumors since this will influence the population of cells within a tumor which can be attacked by antibody-drug or antibody-toxin conjugates. Experiments with human tumor xenografts demonstrate tumor localization of radioisotope-labelled 791T/36 monoclonal antibody. Tumor localization by external gamma camera imaging of osteogenic sarcoma xenograft-bearing mice was also demonstrated. These studies illustrate the potential of antitumor monoclonal antibodies for imaging primary and metastatic tumors. This approach is further emphasized by the radioimmunodetection of primary and metastatic colorectal carcinomas.