Tumoricidal activity and cytokine secretion by tumor-infiltrating macrophages

Murine macrophages from different anatomical sites were compared for their ability to become tumoricidal and to secrete interleukin-1 (IL-1) and tumor necrosis factor (TNF) following stimulation in vitro by several biological response modifiers (BRM). Peritoneal macrophages (PM), alveolar macrophages (AM), and tumor-infiltrating-macrophages (TIM), isolated from B16F10 melanoma colonies in the lung, were incubated overnight with BRM [recombinant murine interferon gamma (rMulFN-gamma), lipopolysaccharide (LPS), muramyl dipeptide (MDP)], either alone or in combination. PM exhibited an increased cytotoxic response following incubation with LPS or rMuIFN-gamma but not with MDP. Both AM and TIM were induced to become tumoricidal following incubation with rMuIFN-gamma plus LPS or rMuIFN-gamma plus MDP but not after stimulation with any BRM alone; the level of cytotoxicity obtained with TIM incubated with rMuIFN-gamma plus LPS was slightly lower than that observed with PM or AM, while with rMuIFN-gamma plus MDP both AM and TIM had lower cytotoxicity than PM. Secretion of IL-I and TNF was observed in PM stimulated with LPS or MDP but not with rMuIFN-gamma. Likewise, secretion of IL-I by AM or TIM was also induced with LPS, although less than that obtained with PM. AM stimulated with LPS secreted larger amounts of TNF than PM while TIM secreted very low amounts of TNF. However, this result may be a consequence of the enzymatic isolation procedure used to obtain TIM since TNF secretion was also impaired in LPS-stimulated normal lung macrophages isolated by a similar enzymatic procedure, or enzyme-treated PM. Our results suggest that TIM obtained from lung metastases share certain functional characteristics with normal AM and respond to BRM in like manner with respect to induction of tumoricidal activity and cytokine secretion.     

Nude mouse model to study passive humoral immunotherapy directed against B16 F10 murine melanoma

A model to study passive humoral immunotherapy of experimental melanoma was generated by subcutaneous injection of B16 F10 murine melanoma cells in the midtail of BALB/C nude (nu/nu) mice. Mice were challenged with melanoma cells pretreated: (1) with complete culture medium, (2) with 10% adjuvant control serum, (3) with 10% anti-fECA (formalinized extracellular antigens) immune serum, or (4) with a monoclonal antibody (mAb H2-3-3) specific for the B700 melanoma-associated antigen. All control mice challenged with melanoma cells pretreated either with culture medium or with medium containing adjuvant control serum (Groups I and II) died during the observation period of 84 days. At day 84, 60% of the mice challenged with melanoma cells pretreated with anti-fECA immune serum (Group III) survived, as did 100% of the mice challenged with cells pretreated with mAb H2-3-3 (Group IV). Injection of melanoma cells pretreated with mAb H2-3-3 was associated with the greatest reduction of subsequent local tumor growth and the lowest number of metastatic lung tumors. The inhibitory effects of immune sera in vivo also correlated with in vitro effects of anti-fECA immune serum and mAb H2-3-3, determined on B16 F10 melanoma target cells using assays for DNA synthesis and antibody dependant cellular cytotoxicity (ADCC). In sum, this nude mouse model for the study of passive humoral immunotherapy of experimental melanoma was utilized to demonstrate significant protective effects against B16 F10 melanoma cell challenge by treatment with anti-fECA immune sera or a melanoma-specific monoclonal antibody. © 1994 Wiley-Liss, Inc.     

Interactions with host macrophages and ability of human melanoma cell lines to grow in nude mice

The interactions of nude mouse macrophages with five human melanoma cell lines, characterized by their resistance to mouse NK activity and varying in their ability to grow s.c. in nude mice, were investigated. These lines were equally susceptible in vitro to both cytostatic and tumoricidal activities of activated peritoneal macrophages collected from nude mice inoculated 3 days previously with Brucella abortus B19R strains. I.p. injection of a poorly tumorigenic melanoma cell line (PTCL) in nude mice was followed by the local appearance of macrophages able to kill these cells in a 48-hr 3H-thymidine cytotoxicity assay. The level of tumoricidal macrophages was maximum for the first week and then slowly declined to disappear by the 4th week following PTCL inoculation. The use of an HTCL instead of a PTCL also induced macrophages able to kill HTCL cells, but the cytotoxicity level was lower and the activity disappeared more rapidly. In cross-experiments using PTCL-activated macrophages as effectors on HTCL targets, these cells were found to be less sensitive than PTCL cells when macrophages were taken at weeks 2 and 3 following PTCL inoculation. To investigate whether tumoricidal macrophages activated in vivo with human melanoma cells could also act in vivo, we inoculated these s.c. into nude mice, simultaneously with live HTCL cells. Peritoneal cells rich in melanoma-activated macrophages prevented HTCL growth in most recipients, whereas spleen cells from the same donor mice did not modify the tumor take. These data indicate that xenogeneic tumors could activate nude mouse macrophages in vivo and suggest that the ability of human tumors to grow in nude mice could be related to their capacity to activate host macrophages locally and to the susceptibility of human tumor cells to the tumoricidal activity of activated macrophages.     

Expression of a Mr 41,000 glycoprotein associated with thrombin-independent platelet aggregation in high metastatic variants of murine B16 melanoma

In the previous study, we generated a monoclonal antibody, 8F11, against NL-17, a high metastatic clone derived from a metastatic variant of murine colon adenocarcinoma 26. 8F11 inhibited platelet aggregation induced by NL-17 and recognized a Mr 44,000 membrane protein as antigen. In the present study, the reactivity of 8F11 to murine B16 melanoma and its metastatic variants was examined, and the antigen recognized by 8F11 on the cell surface was characterized. 8F11 was found to strongly react with 3 metastatic variants of B16 melanoma. In contrast, only slight reactivity was observed with parent B16 melanoma. The reactivity of the antibody to these cells was in the order B16F10 greater than B16BL-6 greater than B16F1 much greater than B16. Western blot analysis showed a Mr 41,000 protein as the antigen recognized by 8F11 on the cell surface of B16F10 cells. The Mr 41,000 antigen appeared to be a glycoprotein that bound to wheat germ agglutinin as has been observed for the Mr 44,000 antigen of NL-17. To elucidate the functional role of the Mr 41,000 antigen in B16 melanoma, platelet aggregation induced by B16 and B16F10 was compared. B16 was reported to stimulate platelet aggregation by the generation of thrombin, whereas B16F10 was found to activate platelet by at least 2 mechanisms: one dependent on thrombin and the other independent on thrombin. The activity of B16 and its metastatic variants to induce platelet aggregation in the presence of MD805, a synthetic antagonist of thrombin, well correlated with the reactivity of 8F11 to these cells. 8F11 blocked platelet activation by B16F10 under conditions preventing thrombin activity such as enzymatic formation of lysolecithin through the treatment of the cell surface with phospholipase A2 or in the presence of MD805. These data indicate that Mr 41,000 glycoprotein recognized by 8F11 on metastatic variants of B16 melanoma is involved in the thrombin-independent platelet aggregation. A positive correlation was observed between the levels of Mr 41,000 glycoprotein expression of B16 and its metastatic variants and their pulmonary metastasis after i.v. injection, suggesting Mr 41,000 glycoprotein, as well as other factors reported previously, may play an important role in the hematogenous spread of B16 melanoma.     

Direct and indirect effects of human recombinant gamma-interferon on tumor cells in a clonogenic assay

Purified, recombinant human gamma-interferon (rIFN-gamma) was tested at clinically achievable doses for direct and indirect antiproliferative activity against human tumor cell lines using a clonogenic assay. One-h treatment with rIFN-gamma showed direct dose dependent inhibition of tumor colony growth in cell lines established from human melanoma, myeloma, renal cell, and cervical cancers. Longer treatments resulted in suppression of ovarian and breast carcinoma clonogenicity. In order to test for indirect antiproliferative effects of rIFN-gamma, feeder cells were included in a separate agarose underlayer in the cloning assay. These feeder cells included mouse peritoneal macrophages, U-937 (human histiocytic lymphoma cell line), and adherent cells from human malignant ascites specimens. Colony growth of ovarian carcinoma and melanoma cell lines was stimulated by each of these feeder cell types. Cultures containing mouse peritoneal macrophages or U-937 cells showed the same antiproliferative responses to rIFN-gamma as did control cultures without feeder cells. In contrast, human adherent ascites cells (greater than 80% macrophages) became strongly inhibitory to tumor colony growth when treated with rIFN-gamma. These results suggest that human tumor associated macrophages may become tumoricidal under the influence of rIFN-gamma, producing a diffusable substance in agarose culture which causes the observed antiproliferative effects on tumor cells.     

Differential antiproliferative effects of combinations of recombinant interferons alpha and gamma on two murine tumor cell lines

The in vitro antiproliferative effects of recombinant interferons (IFNs) alpha and gamma against the murine reticulum-cell sarcoma M5076 and the malignant melanoma B16 F10 were evaluated using the human hybrid IFN-alpha A/D (rHuIFN-alpha A/D), which is active on murine cells, and recombinant murine IFN gamma (rMuIFN-gamma). An isobologram analysis was used to evaluate the interactive antiproliferative effects of the recombinant IFNs on these two tumor cell lines. The data, in contrast to prior reports, indicate that rHuIFN-alpha A/D and rMuIFN-gamma interact in an additive rather than a synergistic manner against M5076 cells. When a similar analysis was performed on B16 F10 cells, synergy was obtained. Thus, either a synergistic or an additive antiproliferative effect can be obtained by combining IFN-alpha and IFN-gamma, depending upon the cell line used in the assay.     

Overexpression of Merlin in B16F10 mouse melanoma cells reduces their metastatic activity: role of the cell surface heparan sulfate glycosaminoglycans

Merlin, the protein product of the neurofibromatosis type 2 gene (NF2) acts as a tumor suppressor in mice and humans. In this study, melanoma B16F10 cells were engineered to overexpress the NF2 gene by establishing stable transductants. A cell line overexpressing Merlin (B16F10-M) was generated. When compared to the parental cells, the B16F10-M cells demonstrated differences in their cell surface organization. The overexpressing strain changed its ability to grow in soft agar as well as its cell motility properties. B16F10-M cells were then examined in the in vivo mouse melanoma tumor growth and tumor metastasis models. While tumor growth was marginally affected, the presence of increased Merlin severely reduced the metastastatic ability of the cells. When isolated using specific enzymes with distinct substrate specificity, the cell surface heparan sulfate glycosaminoglycans (HSGAGs) from the overexpressing B16F10-M cells, inhibited the metastatic properties of the parental B16F10 cells. The results obtained provide a causal link between the reorganization/changes to the cell surface HSGAGs by the overexpression of Merlin and the inhibition of the metastatic activity of the mouse melanoma B16F10 cells in vivo.     

Differential effects of transforming growth factor beta 1 on the growth of poorly and highly metastatic murine melanoma cells

We have examined the effects of transforming growth factor beta 1 (TGF-beta 1) on the growth of paired murine melanoma cell clones that differ with respect to their experimental metastatic potential. Neither poorly (clone 16) nor highly (clone M2) metastatic cells were capable of anchorage-independent growth in 0.3% agar/Dulbecco's modified Eagle's medium in the absence of serum. However, both clones were capable of anchorage-independent growth in 0.3% agar/Dulbecco's modified Eagle's medium containing 10% calf serum. Colony formation in the presence of 10% calf serum was enhanced in a dose-dependent manner by TGF-beta 1 (half-maximal dose, 0.1 ng/ml) and was 5- to 10-fold greater than colony formation in the presence of 10% calf serum alone. Under anchorage-dependent (monolayer) conditions, neither clone grew in the absence of serum or in medium containing less than 1% calf serum. The monolayer growth of poorly metastatic cells (clone 16) was enhanced in a dose-dependent manner by TGF-beta 1 in medium supplemented with calf serum. Growth was 3.5-fold and 2.3-fold greater than untreated controls after 5 days in submitogenic (0.5%) and mitogenic (10%) concentrations of calf serum, respectively. In contrast, TGF-beta 1 had no effect on the monolayer growth of highly metastatic cells (clone M2) either in submitogenic (0.5%) or mitogenic (10%) concentrations of serum. TGF-beta 1 did not directly stimulate DNA synthesis by either poorly or highly metastatic cells when measured 24 h after TGF-beta 1 treatment. The ability of TGF-beta 1 to stimulate the anchorage-independent growth of metastatic melanoma cells suggests that this potent growth factor may play a role in the growth of these cells in vivo. In addition, the differential sensitivity of poorly and highly metastatic cells to TGF-beta 1 may be relevant to their metastatic potential in vivo. While the mechanism(s) by which TGF-beta 1 stimulates the growth of these cells remains unknown, these differentially metastatic clones of the K-1735 murine melanoma should provide a useful model in which to study the effects of transforming growth factor beta on the metastatic phenotype.     

Differentiation and the tumorigenic and metastatic phenotype of murine melanoma cells

Using B16 F10 murine melanoma cells and sublines generated from the JB/MS melanoma which exhibit various degrees of melanogenesis, the relationships among differentiation, tumorigenicity, and metastatic potential were examined. The effect of melanocyte-stimulating hormone (MSH), which specifically stimulates differentiation of melanocytes, was also studied. All melanoma lines tested were capable of growing as experimental pulmonary metastases but, surprisingly, the undifferentiated and amelanotic JB/MS-w cells failed to grow as primary subcutaneous tumors. JB/MS-w cells, which had few surface MSH receptors, did not respond to MSH with an increase in melanin production, unlike the other cell lines. Although in vitro treatment with MSH did not change the rates of growth of primary tumors by these cell lines, such treatment decreased the number of pulmonary metastases from B16 F10, JB/MS cells, JB/MS-b1 cells and JB/MS-w cells. Conversely, MSH treatment significantly increased the rates of pulmonary metastases from JB/MS-p cells. The expression of surface melanoma antigens, urokinase-type plasminogen activity and susceptibility to natural killer cells were examined. MSH did not significantly alter surface melanoma antigen expression, but increased the natural killer cell susceptibility of B16 F10, JB/MS and JB/MS-b1 cells, cells which possess abundant surface MSH receptors. There was an inverse correlation between differentiation (pigmentation) and proliferation in vitro, and the more pigmented melanoma cells (B16 F10, JB/MS and JB/MS-b1) expressed relatively lower levels of class-I MHC, relatively higher levels of class-II MHC and the highest metastatic capacity. These results demonstrate that MSH possesses the capacity to regulate not only melanogenesis, but also other factors critical to the metastatic growth of the cells.     

In vitro assay demonstrates similar invasion profiles for B16F1 and B16F10 murine melanoma cells

The variants of the B16 murine melanoma cell line were assayed for their invasive characteristics in the membrane invasion culture system (MICS) and concomitantly tested for their ability to form lung metastases in vivo. Specifically, the B16F1 (low metastatic variant) and the B16F10 (high metastatic variant) murine melanoma cell lines were examined for their ability to invade human amniotic basement membranes (BMs) in vitro and simultaneously examined for lung colony formation in vivo. The B16F1 and B16F10 cell lines both demonstrated similar invasion profiles over 72 h with the total percent invasion through the BMs for both cell lines not exceeding 5.0%. In vivo observations reconfirmed the significant difference in the metastatic capabilities of the 2 variants. These data suggest that tumor cells with differing metastatic propensities can invade an amniotic BM at similar rates, but their survival and metastatic lesion forming capabilities in vivo may vary considerably.     

Critical factors regulating site-specific brain metastasis of murine melanomas

The intracarotid injection of B16 melanoma cells syngeneic to C57BL/6 mice and K-1735 melanoma cells syngeneic to C3H/HeN mice results in site-specific brain metastasis in C57BL/6 x C3H/HeN F1 mice. The K-1735 cells produce lesions only in the brain parenchyma, whereas the B16 cells produce lesions only in the meninges and ventricles. To determine the mechanisms that regulate this site-specific brain metastasis, we transfected the melanoma cells with DNA from plasmids pSV2neo or pSV2hvgro, which confer resistance to the drugs neomycin and hygromycin, respectively. Hybrids between the B16 and K-1735 cells were obtained by fusion. Cells of the K-1735 x K-1735 hybrid produced lesions only in the brain parenchyma of C57BL/6 x C3H/HeN F1 mice, whereas all B16 x K-1735 hybrids produced lesions only in the meninges and the ventricles. Initial cell arrest in the meninges or the brain parenchyma, production of collagenolytic activity, motility, and expression of CD44 did not predict or correlate with site-specific brain metastasis. The response of the different melanomas and hybrid cells to transforming growth factor-beta (TGF-beta) correlated with growth in the brain parenchyma. B16 cells and B16 x K-1735 hybrids bound more TGF-beta than K-1735 cells. The in vitro growth of B16 cells and all B16 x K-1735 hybrid cells was significantly inhibited by TGF-beta1 and TGF-beta2, whereas the growth of K-1735 cells and K-1735 x K-1735 hybrids was enhanced. Since TGF-beta is abundant in brain tissue, the results suggest that the ability of melanoma cells to proliferate in the brain parenchyma determines the production of site-specific brain metastasis.     

Functional roles of tumor necrosis factor‐related apoptosis‐inducing ligand–DR5 interaction in B16F10 cells by activating the nuclear factor‐κB pathway to induce metastatic potential

Tumor necrosis factor‐related apoptosis‐inducing ligand (TRAIL) has been recognized as a promising target for cancer therapy because it can induce apoptotic cell death in tumor cells but not normal cells. Although TRAIL shows specific tumoricidal activity, resistance to TRAIL‐induced apoptosis in some tumor cells has been considered a clinical obstacle of its application. It has been shown that TRAIL provides inflammatory signals that may contribute to the TRAIL‐resistance of cancer cells; however, it is not known whether TRAIL itself is involved in malignant cancer cell behavior. In the present study, we examined the functional role of TRAIL in B16F10 mouse melanoma cells, which are totally insensitive to TRAIL‐induced apoptosis. By establishing B16F10 cells stably expressing the nuclear factor‐κB (NFκB)‐luciferase reporter gene, we found that TRAIL can activate NFκB through its death receptor DR5 in B16F10 cells. Furthermore, TRAIL–DR5 interaction not only promoted malignant behaviors of B16F10 cells, such as cell proliferation and MMP‐9 production, but also induced lung metastasis of B16F10 cells in vivo. These findings may imply a contrary role for the TRAIL–DR5 pathway in the inflammatory tumor microenvironment, in its ability to induce the metastatic potential of B16F10 melanoma cells instead of inducing apoptosis.     

Taurolidine induces apoptosis of murine melanoma cells in vitro and in vivo by modulation of the Bcl-2 family proteins

BACKGROUND: This study evaluates whether taurolidine, a novel antibiotic agent, induces murine melanoma cell apoptosis in vitro and in vivo. METHODS: Murine melanoma cells (B16 4A5 and B16 F10) were treated with taurolidine (0-100 microM) for 12 and 24 hr. Cell viability and apoptosis were assessed by MTT assay and FACScan analysis. Expression of the Bcl-2 family proteins was detected by Western blot analysis. In vivo, taurolidine-induced anti-tumor cytotoxicity was assessed in C57BL/6 mice. Therapeutic effectiveness, by intraperitoneal injection of taurolidine (15 mg/mouse) on alternate days for 2 weeks, was evaluated in mice bearing B16 4A5 tumor xenografts. Primary and metastatic tumor growth and intra-tumor apoptotic index were measured. RESULTS: Taurolidine induced cell apoptosis and reduced cell viability in murine melanoma cells. The pro-apoptotic protein Bax was enhanced, whereas the anti-apoptotic protein Bcl-2 was inhibited by taurolidine treatment. In vivo, systemic injection of 15-mg taurolidine was identified as the maximally tolerated dose. Administration of taurolidine at 15 mg/mouse significantly inhibited primary and metastatic tumor growth, which was mirrored by a significantly increased intra-tumor apoptotic index. CONCLUSIONS: These results demonstrate that taurolidine significantly attenuated melanoma tumor growth, which may result from taurolidine-induced apoptosis by modulation of the Bcl-2 family proteins.     

Herpes simplex virus-induced suppression of macrophage-mediated tumoricidal activity in murine macrophages

Herpes simplex virus (HSV) infections can enhance the progression of neoplastic diseases. Since macrophages can be activated to become tumorilytic and may figure prominently in host defense against cancer, the ability of HSV to modify macrophage-mediated tumoricidal functions was evaluated. Murine peritoneal macrophages treated with HSV could not be activated to a tumoricidal state by mouse recombinant gamma-interferon (IFN-gamma). Addition of HSV 4 h after treatment with IFN-gamma, at a time when the macrophages are fully committed to developing the cytotoxic phenotype, blocked macrophage-mediated lysis of syngeneic melanoma target cells. This inhibition of activation and cytotoxicity was not due simply to uptake of virus particles, because treatment with heat-inactivated HSV at 4-h posttreatment with IFN-gamma had no effect. In addition, HSV did not undergo a productive infection within macrophages, suggesting that the observed inhibitory activity might be due to a virus-induced product. In this regard, treatment of macrophages with recombinant alpha-interferon suppressed the activation of these cells by IFN-gamma, suggesting that virus-induced alpha-interferon may be mediating all or part of the suppressive activity. These studies suggest that enhancement of tumor progression following HSV infection may be related to the virus-induced suppression of macrophage-mediated tumoricidal activity.     

Rac-WAVE2 signaling is involved in the invasive and metastatic phenotypes of murine melanoma cells

WAVEs (WASP-family verprolin-homologous proteins) regulate the actin cytoskeleton through activation of Arp2/3 complex. As cell motility is regulated by actin cytoskeleton rearrangement and is required for tumor invasion and metastasis, blocking actin polymerization may be an effective strategy to prevent tumor dissemination. We show that WAVEs, especially WAVE2, are essential for invasion and metastasis of melanoma cells. Malignant B16F10 mouse melanoma cells expressed more WAVE1 and WAVE2 proteins and showed higher Rac activity than B16 parental cells, which are neither invasive nor metastatic. The effect of WAVE2 silencing by RNA interference (RNAi) on the highly invasive nature of B16F10 cells was more dramatic than that of WAVE1 RNAi. Membrane ruffling, cell motility, invasion into the extracellular matrix, and pulmonary metastasis of B16F10 cells were suppressed by WAVE2 RNAi. WAVE2 RNAi also had a profound effect on invasion induced by a constitutively active form of Rac (RacCA). In addition, ectopic expression of both RacCA and WAVE2 in B16 cells resulted in further increase in the invasiveness than that observed in B16 cells expressing only RacCA. Thus, WAVE2 acts as the primary effector downstream of Rac to achieve invasion and metastasis, suggesting that suppression of WAVE2 activity holds a promise for preventing cancer invasion and metastasis.     

In vivo activation of nude mouse macrophages by human melanoma cells

Human melanoma cell lines inoculated ip in outbred nude mice were found to activate locally macrophages, which became tumoricidal for the EL 4 target cells in a 48-hour [3H]thymidine cytotoxicity assay. However, the kinetics of this activation largely depended on the tumorigenicity of the cell line used. One week after inoculation with a poorly tumorigenic cell line (PTCL), peritoneal macrophages showed a maximal tumoricidal activity, which then slowly declined to disappear on the 4th week. Macrophages obtained after inoculation of a highly tumorigenic cell line (HTCL) were also activated, but the level of their tumoricidal activity was somewhat lower and decreased more rapidly. Irradiated melanoma cells were also able to activate peritoneal macrophages. The inoculation of a higher number of melanoma cells (less than or equal to 8 X 10(7) cells) resulted in a parallel increase in the cytotoxicity of peritoneal macrophages when activated by PTCL and in a parallel decrease when activated by HTCL. Activated macrophages taken 1 week after tumor cell inoculation and further kept in vitro without additional stimulation progressively lost their tumoricidal activity, within 48 hours after being harvested from PTCL-inoculated mice and within 24 hours after being collected from HTCL-inoculated animals. These data allied to the in vivo capacity of peritoneal cells rich in activated macrophages to prevent the growth of HTCL in nude mice strongly leaned toward the idea that macrophages are involved in the tumor growth control in the absence of a specific immune response. In addition, tumor-macrophage interactions are likely to vary from tumor to tumor and may contribute to the expression of the xenografting capacity of human tumor cells.     

Effects of three irreversible inhibitors of ornithine decarboxylase on macrophage-mediated tumoricidal activity and antitumor activity in B16F1 tumor-bearing mice

The objective of the present investigation was to compare the effects of three ornithine decarboxylase inhibitors on tumoricidal macrophage and antitumor activities in vivo. alpha-Difluoromethylornithine (DFMO), (2R,5R)-6-heptyne-2,5-diamine, and alpha-(fluoromethyl)dehydroornithine methyl ester (delta MFMOme) were administered continuously in drinking water starting on Day 1 to B16F1 tumor-bearing mice. DFMO, (2R,5R)-6-heptyne-2,5-diamine, and delta MFMOme reduced B16F1 tumor growth, measured on Day 18, up to 87, 79, and 95%, respectively. Similarly, all three ornithine decarboxylase inhibitors reduced B16F1 putrescine and spermidine levels. delta MFMOme was substantially more effective both as an antitumor agent and in reducing polyamines. Both DFMO and delta MFMOme augmented macrophage tumoricidal activity directed against B16F1 target cells. MAP had no effect on macrophage tumoricidal activity. Lipopolysaccharide-stimulated macrophages from delta MFMOme-treated mice also exhibited an increase in interleukin and tumor necrosis factor levels. Furthermore, treatment with a known macrophage activator, gamma-interferon, enhanced the antitumor activity of delta MFMOme. delta MFMOme did not alter natural killer cell activity; however, cytolytic T-lymphocyte induction was reduced by 40 to 50%. These results demonstrate that, in addition to their established antitumor activity, ornithine decarboxylase inhibitors may also potentiate specific tumoricidal effector cell generation in vivo.     

Increased experimental metastatic capacity of a murine melanoma following induction of differentiation

Because of the interest in possible links between defective differentiation and cellular malignancy, the effects were examined of induced cell differentiation upon the experimental metastatic potential of the sublines F1 and F10 of the B16 mouse melanoma. These cell lines normally have low and high rates, respectively, of colonization of the lungs of mice after i.v. injection. Cellular differentiation was assessed by pigmentation and tyrosinase activity. In both cell lines, low and high levels of differentiation could reproducibly be generated by culture, respectively, at a low extracellular pH and at a higher pH in the presence of a melanocyte-stimulating hormone. Surprisingly, in both lines the cells grown under conditions promoting differentiation showed a markedly higher rate of experimental metastasis, despite their slower proliferation in culture and in subcutaneous tumor implants, than the poorly differentiated cells. Radiolabeled well- and poorly pigmented cells were not initially deposited at significantly different rates in the lungs of mice after i.v. injection. However, subsequent retention in the lungs fell more quickly for the poorly differentiated cells. As indicated by tests in vitro, this difference appears not to be due to differential cytotoxicity by either host macrophages or natural killer cells, and it is under further study.     

Metastatic potential of murine B16 melanoma correlates with reduced surface heparan sulfate glycosaminoglycan

We studied the relationship between the metastatic potential of murine B16 melanoma cells and their surface expression of heparan sulfate glycosaminoglycan (HS-GAG) by using HepSS-1, a monoclonal antibody specific to HS-GAG. Firstly, among five B16 sublines, those capable of developing lung colonies with high efficiencies, such as B16-F10, had relatively low levels of surface HS-GAG. Secondly, a subline freshly prepared from metastatic lung colonies (F1) displayed a level of surface HS-GAG lower than that of injected B16 cells. Thirdly, in vitro selection of B16 cells with low surface HS-GAG by repeated HepSS-1 staining and cell-sorting resulted in cells with a higher metastatic efficiency than that of the original B16 cells. Collectively, B16 melanoma cells with high metastatic activities seem to have low surface HS-GAG. We also found that there was a positive correlation between the surface level of HS-GAG and the susceptibility to natural killer cells in eight B16 sublines.     

Generation of drug-resistant variants in metastatic B16 mouse melanoma cell lines

Genetic instability is recognized as an important aspect of the development of tumor heterogeneity and malignancy. In a previous study [Hill et al. Science (Wash. DC), 244:998-1001, 1984], we demonstrated that metastatic variants are generated at a more rapid rate in the highly metastatic B16F10 mouse melanoma cell line than in the less metastatic B16F1 cell line. The metastatic variants were phenotypically unstable, being generated and lost at high rates; consequently, we proposed a dynamic heterogeneity model of tumor metastasis which describes these properties quantitatively. As an extension of this work, we have examined the ability of these two melanoma cell lines to generate variants resistant to the drugs methotrexate and N-(phosphonacetyl)-L-aspartate. We observed that the highly metastatic B16F10 cell line generated variants resistant to a given concentration of methotrexate or N-(phosphonacetyl)-L-aspartate at higher rates than the B16F1 cell line. We conclude that B16F10 cells are genetically less stable than B16F1 cells and since resistance to methotrexate and N-(phosphonacetyl)-L-asparate usually results from gene amplification that B16F10 cells possess increased ability to amplify DNA. This higher rate of generation of drug-resistant variants corresponds to the higher rate of generation of metastatic variants we observed previously and suggests that a gene amplification mechanism may be involved in the generation of a metastic phenotype in B16 melanoma cells.