6483.full-text.pdf - The Journal of Immunology

of January 10, 2022.This information is current as

Melanoma CellsTumor Formation in NontumorigenicProtein-1 Stimulation of Monocytes Leads to Low-Level Monocyte Chemoattractant

Meenhard HerlynMark Nesbit, Helmut Schaider, Thomas H. Miller and

http://www.jimmunol.org/content/166/11/6483doi: 10.4049/jimmunol.166.11.6483

2001; 166:6483-6490; ;J Immunol 

Referenceshttp://www.jimmunol.org/content/166/11/6483.full#ref-list-1

, 25 of which you can access for free at: cites 66 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists All rights reserved.Copyright © 2001 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on January 10, 2022

http://ww

w.jim

munol.org/

Dow

nloaded from

Low-Level Monocyte Chemoattractant Protein-1 Stimulation ofMonocytes Leads to Tumor Formation in NontumorigenicMelanoma Cells1

Mark Nesbit,2 Helmut Schaider,2 Thomas H. Miller, and Meenhard Herlyn3

Tumors commonly produce chemokines for recruitment of host cells, but the biological significance of tumor-infiltrating inflam-matory cells, such as monocytes/macrophages, for disease outcome is not clear. Here, we show that all of 30 melanoma cell linessecreted monocyte chemoattractant protein-1 (MCP-1), whereas normal melanocytes did not. When low MCP-1-producing mel-anoma cells from a biologically early, nontumorigenic stage were transduced to overexpress the MCP-1 gene, tumor formationdepended on the level of chemokine secretion and monocyte infiltration; low-level MCP-1 secretion with modest monocyte infil-tration resulted in tumor formation, whereas high secretion was associated with massive monocyte/macrophage infiltration intothe tumor mass, leading to its destruction within a few days after injection into mice. Tumor growth stimulated by monocytes/macrophages was due to increased angiogenesis. Vessel formation in vitro was inhibited with mAbs against TNF-a, which, whensecreted by cocultures of melanoma cells with human monocytes, induced endothelial cells under collagen gels to form branching,tubular structures. These studies demonstrate that the biological effects of tumor-derived MCP-1 are biphasic, depending on thelevel of secretion. This correlates with the degree of monocytic cell infiltration, which results in increased tumor vascularizationand TNF-a production. The Journal of Immunology,2001, 166: 6483–6490.

Solid human tumors are often infiltrated by host immuneand inflammatory cells comprised mainly of lymphocytesand cells of the mononuclear lineage (1). Whereas in-

creased levels of lymphocyte infiltration into primary tumors de-crease tumor recurrence and death rates (2), the presence of in-flammatory cell infiltrates has not been clearly correlated withdisease outcome. Infiltration of tumors with host cells is regulatedby tumor-derived chemokines, a superfamily of proinflammatorycytokines that is responsible for the selective recruitment and ac-tivation of mononuclear cells (2). Chemokines induce directed mi-gration of leukocytes and stimulate their adhesion and transendo-thelial migration (3, 4). Due to the large number of chemokinesproduced by human tumors and the broad spectrum of their bio-logical functions, their precise roles in tumor development andprogression remain undefined.

Monocyte-chemoattractant protein-1 (MCP-1)4 is the prototype ofthe CC family of chemokines (5). It can recruit monocytes (6), NKcells (7), and subpopulations of T lymphocytes (8), which all expresshigh-affinity receptors (9, 10), predominantly CCR2 (11, 12). BecauseMCP-1 secretion results in tissue infiltration of monocytes and T lym-phocytes, the cytokine plays a major role in autoimmune disease

pathogenesis. The role of MCP-1 in tumor development and progres-sion is less clear. Expression has been reported for melanoma (13),glioma (14, 15), sarcoma (16, 17), leukemia (18), hemangioma (19),and carcinomas of breast (20), cervix (21, 22), and ovary (23). Themalignant cells express MCP-1, apparently due to the constitutiveproduction of activating growth factors and cytokines such as IL-1(24), TGF-b (25), and platelet-derived growth factor (26, 27). MCP-1can be protective in some tumor models but destructive in others;murine colon carcinoma cells expressing MCP-1 fail to metastasizewhen injected into mice (28), whereas other carcinoma cells showenhanced metastasis (29). Overexpression of MCP-1 by tumor cellscan lead to their destruction by an infiltrate of activated mononuclearcells (30–33). The potential tumoricidal activity of monocytes/mac-rophages has been used previously as a therapeutic strategy by en-hancing their activity with muramyl dipeptides (34, 35). However,despite promising results in experimental animals, clinical studieshave been disappointing. The lack of clinical success is apparentlydue to the potential positive effect of MCP-1 on tumor growth. MCP-1expression results in the infiltration of macrophages that secrete stim-ulatory factors either for the tumor cells or the vasculature (1, 36).

Infiltration of macrophages/monocytes into cutaneous malignantmelanomas may be critical for progression of melanomas toward anaggressive phenotype (37). Most melanomas from primary and met-astatic lesions produce MCP-1 (38), and macrophage infiltration ap-pears to correlate with tumor stage and angiogenesis (39). We hy-pothesized that monocyte recruitment depends on the level of MCP-1secretion by melanoma cells and that the effect of monocytes on tumorgrowth depends on their level of infiltration. We constructed a repli-cation-defective adenoviral vector for MCP-1 overexpression and es-tablished a MCP-1 gradient before injection into SCID mice. Wedemonstrate that intermediate levels of MCP-1 elicit an angiogeniceffect mediated through monocyte activation that results in tumorgrowth, whereas high levels of 3MCP-1 lead to massive monocyte/macrophage accumulation and tumor destruction. Monocytes/macrophages activated by tumor cells that secrete MCP-1 releaseTNF-a, which may induce angiogenesis. Thus, there is a delicate,

The Wistar Institute, 3601 Spruce Street, Philadelphia, PA 19104

Received for publication August 10, 2000. Accepted for publication March 16, 2001.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby markedadvertisementin accordance with18 U.S.C. Section 1734 solely to indicate this fact.1 This work was supported by National Cancer Institute Grants CA-25874,CA-80999, and CA-10815. H.S. was supported by the Austrian Science Foundation(FWF-1652-Med) and by the Max Kade Foundation of New York.2 M.N. and H.S. contributed equally to this study.3 Address correspondence and reprint requests to Dr. Meenhard Herlyn, The WistarInstitute, 3601 Spruce Street, Philadelphia, PA 19104. E-mail address: [email protected] Abbreviations used in this paper: MCP-1, monocyte chemoattractant protein-1; pAb,polyclonal Ab; PECAM-1, platelet endothelial cell adhesion molecule-1; VEGF, vas-cular endothelial growth factor; bFGF, basic fibroblast growth factor; Ad, adenovirus.

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

concentration-dependent balance for the biological function ofMCP-1, which may result in either tumor enhancement or destructionby infiltrating monocytes/macrophages.

Materials and MethodsCells

SBcl2 cells (obtained from Dr. B. Giovanella, St. Joseph’s Hospital CancerCenter, Houston, TX) were isolated from a primary cutaneous melanoma.These cells are nontumorigenic in immunodeficient mice, grow poorly insoft agar, and require exogenous growth factors for proliferation (40). Allother melanoma cell lines were isolated and are maintained at the WistarInstitute (Philadelphia, PA) (41). They were grown in melanoma growthmedium W489, consisting of MCDB 153 medium (Sigma, St. Louis, MO)and LeibovitzL-15 medium (Sigma) at a 4:1 (v/v) ratio (40) and supple-mented with insulin at 5mg/ml (Sigma) and 2% heat-inactivated FCS (Ir-vine Scientific, Irvine, CA) unless otherwise stated. Normal human mela-nocytes were obtained from newborn foreskin as described (42). They werecultured in medium W489, supplemented with 2 mM CaCl2, 2% FCS and5 mg/ml insulin (Sigma), 10 ng/ml epidermal growth factor, 140mg/mlbovine pituitary extract, and 10 ng/ml 12-O-tetradecanoylphorbol-13-ace-tate. HUVECs were grown on gelatin-coated plastic dishes in M199 me-dium (Life Technologies, Carlsbad, CA) supplemented with 10% FCS,endothelial cell growth factor (150mg/ml), and heparin (5 U/ml) as pre-viously described (43). Cells were used between the second and eighthpassages. The 293 E1A-transformed human embryonic kidney cells(American Type Culture Collection, Manassas, VA) were grown inDMEM supplemented with 10% FCS.

Human peripheral blood monocytes were isolated essentially as described(44) using only endotoxin-free reagents. Briefly, human peripheral bloodmonocytes from the blood of healthy volunteers were separated on a Ficoll-Paque (Pharmacia Biotech, Uppsala, Sweden) gradient and resuspended inRPMI 1640 medium (Sigma) supplemented with 10% human AB serum andpolymyxin-B (10mg/ml) at 43 106 cells/ml. Tissue culture dishes (150 mm;Corning Glass, Corning, NY) were coated with 5 ml of 2% gelatin in physi-ological saline and incubated for 2 h at 37°C, after which the gelatin wasaspirated and the dishes were left to dry. Autologous serum (10 ml) was added,and the dishes were incubated for 60 min at 37°C. After removal of the serum,dishes were rinsed with Mg21- and Ca21-free PBS, 30 ml of mononuclear cellsuspension was added per dish, and they were incubated for 45 min at 37°C.Nonadherent cells were aspirated, and adherent cells were rinsed with pre-warmed (37°C) RPMI 1640 medium. A 10-ml mixture (1:1) of 10 mM EDTAand PBS (Mg21- and Ca21-free) was added for 15 min to remove adherentcells. Cells were centrifuged and resuspended in RPMI 1640 with 10% FCSand analyzed by flow cytometry. Monocyte yields were calculated to be.70%with .90% cell viability.

Adenovirus (Ad) vector

A plasmid containing the 741-bp human MCP-1 cDNA was used to con-struct the adenoviral vector MCP-1-Ad5 using previously described tech-niques (45). Briefly, the open reading frame of MCP-1 cDNA (;400 bp)was subcloned into a modified pSL301 vector (Vector Core, Institute forHuman Gene Therapy, University of Pennsylvania, Philadelphia, PA) us-ing EcoRI andPstI digestion of both MCP-1 cDNA and pSL301. ThepSL301 containing the MCP-1 cDNA was excised withNotI, pAdCMV(Vector Core), linearized withNotI at the unique restriction site, and li-gated. Sense orientation of the insert was determined by restriction analysisusingEcoRI and sequencing. MCP-1 cDNA was under the control of theCMV immediate/early enhancer-promoter element and the SV40 polyad-enylation signal. Recombination was done in 293 cells, and the rAd wasplaque-purified, expanded in 293 cells, and purified by cesium chloridegradient centrifugation. The adenoviral control vectorLacZ-Ad5 express-ing b-galactosidase (45) was produced using the same techniques.

Production of rMCP-1, mAb, and polyclonal Ab (pAb)

The rMCP-1 was produced fromEscherichia colias a GST fusion proteinand affinity-purified on glutathione-Sepharose beads (Pharmacia, Piscat-away, NJ). The cDNA was cloned into the PGEX-2T vector, and recom-binant protein was induced, purified, and cleaved with thrombin accordingto the manufacturer’s instructions (Pharmacia). Western analysis indicatedreactivity of a 13-kDa protein with mouse pAb specific for MCP-1 (AB-479-NA; R&D Systems, Minneapolis, MN). BALB/c mice were immu-nized s.c. with 50mg of recombinant protein in CFA followed by threeinjections of MCP-1 in IFA at biweekly intervals. Three days before fu-sion, 50mg of rMCP-1 was injected i.v. without adjuvant. Murine my-eloma cells SP/20 were used for fusion, and three hybridomas, MCP-1-

D10-1, MCP-1-D10-3, and MCP-31-B, were selected for specific bindingto rMCP-1 in enzyme-linked immunoadsorbent assays. All three mAbswere of IgG1,k isotype. Western blotting confirmed MCP-1 binding of themouse mAb. For production of pAbs, rabbits were immunized using thesame protocol as for mice, except rabbits received a total of six injectionsand were bled 10 days after the final injection. Abs were purified withSepharose B-bound protein A using standard protocols.

Immunoblotting

To demonstrate MCP-1 production from rAd, 53 107 infected 293 cells in20 ml of DMEM with 10% FCS were left to develop 50% cytopathiceffects. Aliquots (50ml) of the supernatants were heated to 100°C for 10min in the same amount of SDS sample buffer (10% SDS, 100 mM Tris(pH 6.8), 1% glycerol, 125 mg/ml bromophenol blue, with or without 5%(v/v) 2-ME (12.5 M)), separated on 12% SDS-polyacrylamide gels, andtransferred onto polyvinylidene difluoride membranes (Millipore, Bedford,MA) overnight at 4°C with 20 V constant voltage. Nonspecific binding tothe polyvinylidene difluoride membranes was blocked with PBS containing3% BSA for 60 min at room temperature. Between incubations, mem-branes were washed three times for 5 min each with PBS containing 0.05%Tween 20. Membranes were probed with mouse mAb MCP-1-D10-1 (60.8mg/ml; 1:20) or rabbit pAb against rMCP-1 (1:200) for 1 h. Membraneswere then incubated with IgG goat anti-mouse phosphatase (Jackson Im-munoResearch Laboratories, West Grove, PA) or donkey IgG anti-rabbitphosphatase (Jackson ImmunoResearch Laboratories) as secondary Absfor 2 h at room temperature. Immunoreactive bands were visualized using5-bromo-4-chloro-3-indolyl-phosphate/nitro blue tetrazolium (Promega,Madison, WI) in alkaline phosphatase buffer.

Tumor formation in SCID mice and immunohistochemistry

SBcl2 melanoma cells were infected with MCP-1-Ad5 orLacZ-Ad5 atdefined PFU per cell. At 48 h after transduction, 23 106 SBcl2 cells in 100ml of PBS were injected s.c. into five SCID mice per group. To inhibittumor growth after injection of SBcl2 transduced cells at 0.5 PFU/cell,SCID mice were treated daily i.p. with 150mg of a neutralizing rabbit pAbagainst MCP-1 starting 1 day before s.c. injection of transduced SBcl2 cellsuntil day 4. Tumor growth was evaluated 4, 8, and 14 days later. Forhistological examination, tumor lesions were fixed in formalin, dehydratedthrough graded alcohol and xylene, and embedded in paraffin. Fresh frozensamples were embedded in OCT embedding medium (Sakura Finetek, Tor-rance, CA). Serial 5-mm sections were cut and stained with hematoxylinand eosin. Immunohistochemistry was performed on serial cryosections byan immunoperoxidase technique using an avidin-biotin-peroxidase com-plex system (Vector Labarotories, Burlingame, CA) and 3,39-diaminoben-zidine as chromagen. Tissue sections were acetone-fixed for 10 min at 4°C,incubated with primary Ab overnight at 4°C, thoroughly rinsed with PBS,and overlaid with biotinylated anti-mouse or anti-rabbit IgG for 30 min atroom temperature. After three washings, avidin-biotin-peroxidase complexwas added for 45 min. Slides were rinsed well with PBS, developed with3,39-diaminobenzidine, and counterstained lightly with hematoxylin. Thefollowing mAbs were used: anti-human Ki67 proliferation marker (Immu-notech, Westbrook, ME) at 10mg/ml, anti-mouse CD11b (Mac-1a-chain)(BD PharMingen, San Diego, CA) at 25mg/ml to detect mouse macro-phages, anti-mouse TNF-a (BD PharMingen) at 10mg/ml, and anti-mouseCD31 (platelet endothelial cell adhesion molecule-1 (PECAM-1); BDPharMingen) at 25mg/ml to show vessel formation. For immunofluores-cence, an FITC-conjugated goat anti-rat IgG Ab (Jackson ImmunoRe-search) was used.

Chemotaxis assay

Chemotaxis assays were conducted using filter inserts with 3-mm pores(Millipore) in triplicate 24-well plates. SBcl2 melanoma cells were trans-duced with MCP-1 orLacZ, and, 24 h later, the growth medium waschanged to serum- and growth factor-free medium for 72 h, after whichsupernatants were collected. Freshly isolated monocytes (33 106 cells/insert) preincubated with human IgG (1mg/ml per 106 monocytes) for 15min were placed in the upper chambers of inserts, and supernatants ofcocultures or recombinant human MCP-1 (50 ng/ml) were added to thelower chamber. After 3 h, monocytes, which had migrated through the filterto the lower chamber, were collected and viable cells confirmed by trypanblue exclusion were counted. Chemotaxis was inhibited by adding mAbMCP-1-D10-1 (5mg/ml) or a rabbit pAb against MCP-1 (50mg/ml) to thelower chamber.

6484 MCP-1 EITHER STIMULATES OR INHIBITS MELANOMAS

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

Coculture assays, ELISA, and radioimmunoassay

SBcl2 melanoma cells were infected with the adenoviral vectors forMCP-1 orLacZ 36 h before coculture with freshly isolated human mono-cytes in DMEM with 5% FCS for 18 h. Supernatants were tested for IL-4,IL-8, IL-10, basic fibroblast growth factor (bFGF), vascular endothelialgrowth factor (VEGF), GM-CSF, and MCP-1 using ELISA kits (Quan-tikine, obtained from R&D Systems; and Endogen, Woburn, MA). TNF-awas quantitated by radioimmunoassay as previously described (46).Briefly, mAb-coated plates were washed four times with PBS-Tween 20,and 50ml of sample in replicates or standard was added to each plate(detection limit $ 1 pg/ml). The assay was repeated three times. Afterincubation at 4°C overnight, plates were washed four times with PBS-Tween 20, and 1mg of 125I-labeled mAb was added to 10 ml of PBS-5%milk with 100 ml being placed in each well. Plates were incubated over-night at 4°C, washed, and radioactivity was measured.

Modulation of HUVEC phenotype

Changes in endothelial morphology were assessed as previously described(43) with modifications. HUVECs were removed from plastic flasks withtrypsin and EDTA, which were neutralized with FCS and M199 medium.

Cells were washed and seeded in quadruplicate in 96-well plates at 23 104

cells/well. When confluent, cells were washed with HBSS, incubated for4 h with conditioned or control medium, and overlayered with an acellularcollagen mix of M199 medium supplemented with heparin, glutamine, en-dothelial cell growth factor, sodium bicarbonate, and bovine collagen (col-lagen type I; Organogenesis, Canton, MA) (final concentration of 1 mg/ml). The mixture was left to gel, 200ml of conditioned or control mediumwas added, and cultures were incubated at 37°C for 18 h. Morphologicalchanges in cells were evaluated microscopically.

Statistics

Comparisons between groups were made by the Studentt test. A differencebetween groups ofp , 0.05 was considered significant.

ResultsMCP-1 production by melanoma cells

ELISA screening of normal melanocytes and melanoma cells re-vealed constitutive production of MCP-1 in all 30 melanoma celllines but in none of five melanocyte cultures (Fig. 1A). About half

FIGURE 1. Secretion of MCP-1 by melanocytes, nevi, and primary and metastatic melanoma.A,MCP-1 levels in supernatants of cultures after 72 h. Secretionof biologically active MCP-1 by melanocytic cells, as detected by ELISA. Amount of MCP-1 produced by melanocytic cells in milliliters per 106 cells per 72 h.B, MCP-1 levels in supernatants after infection with MCP-1 orLacZ adenoviral vectors at different PFU per cell. Control, nontransduced SBcl2 cells. Levels weredetermined by ELISA. The assay was repeated three times.C, Detection of MCP-1 by Western blot analysis using mAb MCP-1-D10-1. Equal amounts of lysates(50 mg) were loaded.Lane 1, MCP-1-Ad5-transduced SBcl2 cells (20 PFU/cell);lane 2,LacZ-transduced SBcl2 cells; lanes3 and4, 5 mg, respectively, ofMCP-1-GST fusion protein; andlane 5, 5mg of GST protein.D, Chemotaxis was measured as migration of human monocytes in the upper chamber through filterinserts into the lower chamber containing 72-h supernatants of MCP-1-Ad5-transduced SBcl2 cells. R, Chemotaxis in response to recombinant human MCP-1 at50 ng/ml. Supernatants of SBcl2 cells, which were not transduced, served as control. Chemotaxis is given as percentage of control (mean6 SD) of triplicate assaysperformed twice.E, Same asD, except that culture supernatants contained mAb MCP-1-D10-1 (5mg/ml). p, p , 0.05, significant inhibition of chemotaxis ascompared with supernatant without Ab.

6485The Journal of Immunology

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

of the melanoma cell lines produced 5–100 ng MCP-1/ml per 106

cells in 72 h, with three of these cell lines producing between 200and 400 ng/ml. A low-producer primary melanoma cell line,SBcl2, representing a biologically early radial growth-phase pri-mary melanoma, was selected for MCP-1 transduction with anadenoviral vector. As shown in Fig. 1B, nontransduced SBcl2 cellssecreted;4–9 ng/ml over a 72-h period, andLacZ-transducedcells secreted only marginally more compared with MCP-1-trans-duced cells. In cells transduced with the viral vector MCP-1-Ad5,MCP-1 production increased with increasing PFU, reaching amaximum of 6000 ng/ml at a dose of 50 PFU/cell.

Western blotting using mAb MCP-1-D10-1 (Fig. 1C) confirmedthe production of MCP-1 protein by transduced SBcl2 cells (lane 1),whereas MCP-1 was not detected inLacZ-transduced cells (lane 2).The mature protein migrated at 14.5 kDa, with a smaller band appar-ently the result of incomplete glycosylation. The MCP-1-GST fusionprotein was detected at 34 kDa (lanes 3and4). Chemotaxis assaysconfirmed the biological activity of adenoviral vector-induced MCP-1(Fig. 1D). The migration of human monocytes in response to theculture supernatants of transduced SBcl2 cells depended on the doseof MCP-1-Ad5 used for transduction. Supernatants of melanoma cellstransduced with MCP-1 at 50 PFU/cell increased chemotaxis.4-foldcompared with control, whereas rMCP-1 was 2-fold more chemotac-tic, similar to the response to supernatants ofLacZ-transduced cells at50 PFU/cell, which was due to the increase in MCP-1 production afterLacZ transduction (see Fig. 1B). There was a significant difference inchemoattraction between 50 PFU/cell and 0.5 or 0.05 PFU/cell (p ,0.05) as well asLacZ-transduced cells and 0.05 PFU/cell (p , 0.05),but there was not a significant difference with rMCP-1 and 0.05 PFU/cell. Chemotaxis in response to supernatants of MCP-1-Ad5-trans-duced SBcl2 cells was significantly inhibited with mAb againstMCP-1, except when 50 PFU/cell of MCP-1-Ad5 was used for trans-duction (Fig. 1E). No inhibition was observed with a nonspecific mu-rine IgG (data not shown).

In vivo survival and growth of SBcl2 melanoma cells isdependent on low-level MCP-1 production

SBcl2 cells injected into SCID mice (23 106 cells/mouse) did notsurvive and grow; after 4 days, the tumor nodule was no longervisible at the injection site, nor were viable tumor cells seen inhistological sections of tumor cell debris at the injection site. Thesame results were obtained with SBcl2 cells transduced with MCP-1-Ad5 at 0.005 PFU/cell. Following transduction of SBcl2 cellswith LacZ-Ad5 at 50 PFU/cell, the tumor remained palpable, andhistochemical analysis indicated a moderate inflammatory reac-tion, necrotic cells, and only a few surviving melanoma cells at day4 (Fig. 2C). At day 14, the tumor had disappeared. SBcl2 cellstransduced with MCP-1-Ad5 at 50 PFU/cell underwent rapid ne-crosis, with large infiltrates of inflammatory cells, presumablymononuclear cells, in tumor sections (Fig. 2D). Lesions had dis-appeared by day 14. In contrast, tumor growth and survival wasobtained using SBcl2 cells transduced with MCP-1-Ad5 at 0.5PFU/cell. After 4 days, the lesions were very well circumscribedwith a mild inflammatory reaction and a small necrotic area in themiddle of the tumor (Fig. 2A). Viable cells and mitotic figureswere abundant. Tumor growth continued, and, by day 14, the le-sion was highly vascularized (Fig. 2B). Transduction of SBcl2cells at 0.05 and 5 PFU/cell resulted in a somewhat intermediatetumor phenotype. Growth of tumors formed by SBcl2 cells trans-duced with MCP-1-Ad5 at 0.5 PFU/cell was almost completelyinhibited by daily i.p. injection of a rabbit pAb against MCP-1,with tumor sections revealing some inflammatory reaction but lit-tle growth by day 10 (Fig. 2E). Fig. 2F shows a lesion from a

mouse treated without Ab. Injection of mice with a nonspecificrabbit IgG showed the same results.

MCP-1-induced macrophage migration leads to increased vesselformation and production of TNF-a

To further characterize the inflammatory infiltrate observed at sitessurrounding the tumor and to show an increase in vessel formation,melanoma tissue sections were analyzed immunohistochemically(Fig. 3). After injection of SBcl2 cells transduced with MCP-1-Ad5 at 50 PFU/cell, a strong infiltration of macrophages, as de-tected with mAb Mac-1, was observed within and around the le-sions on day 4 (Fig. 3A). With MCP-1-Ad5 at decreasing PFU/cell,fewer macrophages were seen. Tissue sections from SBcl2 cellstransduced withLacZ-Ad5 at 50 PFU/cell revealed only a fewmacrophages (Fig. 3B). Tumor growth 14 days after injection ofSBcl2 cells transduced with MCP-1-Ad5 at 0.5 PFU/cell was con-firmed by staining with the proliferation marker Ki67 (Fig. 3C),and tumor vasculature within the tumor area had increased, asindicated by immunofluorescence analysis for mouse PECAM-1(CD31) (Fig. 3D). No such increase in vessel formation was ob-served after injection of SBcl2 cells transduced with MCP-1-Ad5at higher PFU per cell or withLacZ-Ad5-transduced cells. Non-transduced cells could not be evaluated because they did not sur-vive. A significant increase in tumor vessels in lesions of SBcl2cells transduced with MCP-1 at 0.5 PFU/cell on day 14 (5.561.2/mm2) was observed if compared withLacZ at 50 PFU/cell(1.8 6 0.53/mm2) or MCP-1 at 50 PFU/cell (2.56 0.87/mm2) onday 4. Murine TNF-a was produced on day 4 at sites of infiltrationwith macrophages in sections of MCP-1-Ad5-transduced SBcl2cells, as shown for 50 PFU/cell (Fig. 3E), whereas, at sites ofLacZ-Ad5-transduced cells, hardly any positive cells were found(Fig. 3F).

Production of TNF-a in cocultures of SBcl2 melanoma cells andhuman monocytes

Conditioned medium of SBcl2 cells transduced at different PFUper cell and cocultured overnight with freshly isolated humanmonocytes was analyzed for production of cytokines and growthfactors. Levels of IL-4, IL-8, IL-10, GM-CSF, bFGF, and VEGFremained unchanged in supernatants of transduced SBcl2 cellsalone or under coculture conditions. However, TNF-a levels wereincreased 6-fold in MCP-1-transduced SBcl2 cells cocultured withhuman monocytes (Fig. 4A). No TNF-a production was found insupernatants of transduced SBcl2 cells alone or in supernatants ofmonocytes activated overnight with recombinant human MCP-1(100 ng/ml). Screening of 20 melanoma cell lines revealed no con-stitutive TNF-a production.

To test for the biological significance of TNF-a produced duringcocultures of human monocytes and melanoma cells, HUVECs werecultured under collagen type I in the presence of conditioned mediumfrom cocultures (Fig. 4,B–J). Only culture supernatants containingTNF-a induced a branching-like network resembling tubule forma-tion in endothelial cells (236 4.3 vessel-like structures per 27 mm2)(Fig. 4G). This network was inhibited with a neutralizing Ab againstTNF-a (7 6 2.1 vessel-like structures per 27 mm2) (Fig. 4H). Noinhibition was observed with an unspecific human IgG (data notshown). The rTNF-a also induced the same morphologic changes(12 6 2.8 vessel-like structures per 27 mm2) (Fig. 4I), which wereinhibited with the same neutralizing Ab (56 3.2 vessel-like structuresper 27 mm2) (Fig. 4J) but not with a control Ab (data not shown). Noother culture conditions tested (Fig. 4,B–F) induced circular struc-tures, suggesting that TNF-a secreted under coculture conditions ismost likely responsible for the angiogenic activity of MCP-1-inducedtumors.

6486 MCP-1 EITHER STIMULATES OR INHIBITS MELANOMAS

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

DiscussionThe contribution of tumor-derived chemokines in either sup-porting tumor growth or suppressing it is still controversial.Transfection of tumor cells with MCP-1 can prevent tumor for-mation (30) and decrease metastasis (28) but can also increasetumorigenicity and lung metastasis if fewer cells are injectedinto animals (47). The adenoviral vector-mediated transfer ofhuman MCP-1 into a melanoma cell line that is immortalizedbut nontumorigenic (40) is ideally suited to establish a gradient-dependent expression system. SBcl2 cells are derived from anearly, primay cutaneous melanoma and do not grow upon in-jection into SCID mice at 23 106 cells. When infected at 0.5PFU/cell with the rAd, melanoma cells showed a 10-fold in-crease in MCP-1 production as compared with noninfected con-trols. The higher production levels after 0.5 PFU/cell transduc-tion corresponded to the constitutive levels in about half of the30 melanoma cell lines tested. These data suggest that MCP-1-mediated recruitment of monocytes into melanoma lesions maybe important in many cases for the critical progression step,

when the melanoma cells begin to proliferate to form an ex-panding vertical growth-phase tumor (48). However, experi-ments in transgenic mice have shown that MCP-1 expressionalone does not cause inflammatory activation of cells (49).Thus, the stimulation of monocytes is crucial for the observedbiological effects. Lower MCP-1 concentrations have little ef-fect, whereas higher production levels appear to lead to massiveinfiltration of monocytes/macrophages capable of tumor de-struction. To prove the concept of tumor destruction by mac-rophages at high levels of MCP-1, we transduced an aggressivehuman melanoma cell line (WM9), which is metastatic in SCIDmice, at 50 PFU/cell. As expected, these cells did not grow invivo (data not shown). Because human MCP-1 binds to themurine receptor with;50% affinity (50), the production levelsby patients’ tumors to attract the critical number of monocytesfor tumor growth stimulation may be lower.

In mice injected with SBcl2 cells that were transduced with MCP-1-Ad-5 at 0.5 PFU/cell, tumor-associated murine macrophages werefound mainly as a peritumoral infiltration, whereas the 50 PFU/cell

FIGURE 2. MCP-1-dependent growth of SBcl2 cell tumors in vivo. SBcl2 cells were infected with MCP-1-Ad5 orLacZ-Ad5 and 2 days later were injecteds.c. into SCID mice (23 106 cells/mouse). Sections were stained with hematoxylin and eosin.A, SBcl2 cells 4 days after injection of MCP-1-transduced cells at0.5 PFU/cell (magnification,340). At higher magnification (A1; magnification,3600) mitosis is visible (indicated by arrowhead). The bar represents 0.5 mm.B,Same asA, except 14 days after injection (magnification,340). Higher magnification (B1; magnification,3400) reveals abundant blood vessels (indicated byarrowhead), with a single vessel shown inB2 (magnification,3600).C, SBcl2 cells transduced withLacZ at 50 PFU/cell. No tumor growth 4 days after injection(magnification,340). Higher magnifications show only few surviving tumor cells (C1; magnification,3400) and apoptosis (�; C2; magnification,3600).D,SBcl2 cells transduced with MCP-1-Ad5 at 50 PFU/cell 4 days after injection. A strong cellular infiltrate and tumor necrosis are visible (magnification,340).Higher magnifications show tumor necrosis (D1; magnification,3600) and a mononuclear cell infiltrate (D2; magnification,3600).E, Inhibition of tumor growthof SBcl2-transduced cells at 0.5 PFU/cell at day 10 after daily i.p. injection with 150mg of rabbit pAb against MCP-1 starting 1 day before s.c injection oftransduced SBcl2 cells until day 4 (magnification,380), with some surviving tumor cells (E1; magnification,3160).F, Same asE but without Ab against MCP-1(magnification,380), showing tumor growth (F1; magnification,3160). The same results were obtained with nonspecific rabbit IgG.

6487The Journal of Immunology

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

infection rate resulted in intratumoral as well as peritumoral infiltra-tion patterns. Recruitment of peritumoral macrophages was likelybeneficial, whereas intratumoral infiltration led to macrophage-medi-ated cytotoxicity. Due to the transient nature of adenoviral-mediatedgene transfer and the decline in MCP-1 levels when tumor cells di-vide, the observation period in our experiments was 2 wk, whichrestricts overall conclusions for a longer period of time, especiallywith respect to sustained tumor growth and progression. After 3 wk,when Ad-driven MCP-1 production had ceased, tumors disappeared,suggesting that continuous stimulation by mouse macrophages is nec-essary to maintain tumor survival and growth.

MCP-1-mediated macrophage attraction appears to be essen-tial for tumor growth at 0.5 PFU/cell because i.p. injection oftumor-bearing mice with a neutralizing pAb against MCP-1 in-hibited the recruitment of macrophages and abrogated tumorgrowth. Similar results have been obtained in a tumorigenicmelanoma cell line, which recruited large numbers of macro-phages within the tumor mass. Treatment with a neutralizing Abagainst MCP-1 resulted in reduced numbers of intratumoralmacrophages and, subsequently, in significantly higher tumorgrowth (33).

Tumor-associated macrophages play a pivotal role in tumor angio-genesis (51), thus enabling tumor cells to survive and proliferate. Ac-tivated macrophages can release growth factors (VEGF, platelet-de-rived growth factor, insulin-like growth factor-1, bFGF, GM-CSF)and cytokines (IL-1, IL-6, IL-8, and TNF-a), some of which are can-didates for melanoma growth stimulation in the 0.5 PFU/cell trans-duction group. However, MCP-1 overexpression in melanoma cellsdid not increase production of VEGF and bFGF, the two most likelycandidates inducing tumorigenicity in biologically early melanomacells (45 and our unpublished observations), nor was their productionincreased in cocultures of human monocytes and melanoma cells.MCP-1 can also trigger adhesion of monocytes to vascular endothe-lium (3) and can regulate expression of adhesion molecules and cy-tokines, in particular, thea-chains of two members of theb2 integrinfamily (52, 53). MCP-1 stimulates monocytes to produce IL-1 andIL-6 but not TNF-a (52). In our investigations, TNF-a appears to be

the pivotal cytokine in inducing tumor growth; production of TNF-awas increased up to 6-fold in cocultures of monocytes and melanomacells, depending on the batch of isolated monocytes, whereas produc-tion in either melanoma cells or monocytes alone was unchanged.

TNF-a is expressed at low levels in nevi and primary andmetastatic melanoma in situ throughout the progression ofmelanocytic lesions (54). In vitro, only a few melanoma celllines express TNF-a RNA transcripts (55). Therefore, thesource of the TNF-a increase observed in this study may be dueto the activation of monocytes through contact with SBcl2 cellsor indirectly through soluble factors. Staining for murineTNF-a in sections revealed that TNF-a is indeed produced bymacrophages infiltrating (Fig. 3E) or surrounding tumor lesions.This is in accordance with recent findings that describe the pro-duction of murine TNF-a after stimulation with melanoma-con-ditioned medium (56) or IL-18 (57). TNF-a production bymonocytes/macrophages after activation is well known (51).The reasons for the dramatic increase in TNF-a production incocultures of melanoma cells and monocytes are unclear. It ispossible that MCP-1-stimulated monocytes produce IL-6, andmelanoma cells in vitro express functional IL-6 receptors (54,58). However, biologically early melanoma cells are inhibitedby IL-6, and only metastatic melanoma cells are stimulated(58).

The reasons for the increase in the vascularization of the tumorsobserved at intermediate concentrations of MCP-1 remain specula-tive. In the rabbit corneal angiogenesis assay, MCP-1 was capable ofinducing neovascularization, although the angiogenic process waslinked to the recruitment of macrophages (59), suggesting that mac-rophages were a prerequisite for this process and that the effect ofMCP-1 was only indirect. However, a direct involvement of MCP-1in angiogenesis has been reported recently (60). In our study, nochanges in HUVEC morphology were observed when rMCP-1 of 100ng/ml was added to the medium, whereas conditioned medium fromcocultures or rTNF-a induced a branching-like network. The pheno-typic alterations were partially inhibited by a TNF-a neutralizing Ab.Thus, it seems likely that the angiogenic phenotype is initiated

FIGURE 3. Immunohistochemical staining of mela-noma tissue sections. Human SBcl2 melanoma cells weretransduced with MCP-1 orLacZ using adenoviral vectorsand injected s.c. into SCID mice (magnification,3160).A, Staining for mouse macrophages with mAb Mac-1 4days after s.c. injection of SBcl2 cells transduced withMCP-1 at 50 PFU/cell.B, Melanoma cells transducedwith LacZ at 50 PFU/cell; sections stained with mAbMac-1.C,Tumor growth 14 days after injection of SBcl2cells transduced with MCP-1 at 0.5 PFU/cell; stainingwith mouse mAb against the human proliferation markerKi67. D, SBcl2 cells transduced with MCP-1-Ad5 at 0.5PFU/cell; staining of a tumor section 14 days after injec-tion of the transduced SBcl2 cells using rat mAb againstmouse endothelial cell marker PECAM-1 (CD31). Assecondary Ab, FITC-conjugated goat anti-rat IgG wasused.E, Staining for murine TNF-a with mAb MP6-XT22 4 days after s.c. injection of SBcl2 cells transducedwith MCP-1 at 50 PFU/cell.F, SBcl2 cells transducedwith LacZ at 50 PFU/cell; sections stained for murineTNF-a.

6488 MCP-1 EITHER STIMULATES OR INHIBITS MELANOMAS

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

through activated macrophages producing TNF-a. In turn, activatedmacrophages may produce proangiogenic factors (61–63). AlthoughTNF-a was shown to inhibit the growth of endothelial cells (64), lowdoses of TNF-a stimulated migration of endothelial cells and inducedtubule-like structures (65). The dual role of TNF-a in angiogenesishas been demonstrated by Fajardo et al. (66) through direct stimula-tion and modulation of angiogenic factors such as IL-8, VEGF, andbFGF (63). Together, our results point to the pivotal role of tumor-infiltrating inflammatory cells for melanoma progression at a stagewhen the tumor cells are still susceptible to cytotoxicity by host cells.

Acknowledgments

We thank A. Garfall and F. Reisman for excellent technical assistance andDrs. A. Mackiewicz, M.Wysocka, and C. Brando for helpful comments.

References

1. Mantovani, A., B. Bottazzi, F. Colotta, S. Sozzani, and L. Ruco. 1992. The originand function of tumor-associated macrophages.Immunol. Today 13:265.

2. Rollins, B. 1997. Chemokines.Blood 90:909.3. Gerszten, R. E., E. A. Garcia-Zepeda, Y. C. Lim, M. Yoshida, H. A. Ding,

M. A. Gimbrona, A. D. Luster, F. W. Luscinskas, and A. Rosenzweig. 1999. MCP-1

and IL-8 trigger firm adhesion of monocytes to vascular endothelium under flowconditions.Nature 398:718.

4. Strieter, R. M., S. L. Kunkel, V. M. Elner, C. L. Martonyi, A. E. Koch, P. J. Polverini,and S. G. Elner. 1992. Interleukin-8, a corneal factor that induces neovascularization.Am. J. Pathol. 141:1279.

5. Rollins, B. J., A. Walz, and M. Baggiolini. 1991. Recombinant human MCP-1/JEinduces chemotaxis, calcium flux, and the respiratory burst in human monocytes.Blood 78:1112.

6. Valente, A. J., D. T. Graves, C. E. Vialle-Valentin, R. Delgado, and C. J. Schwartz.1988. Purification of a monocyte chemotactic factor secreted by nonhuman primatevascular cells in culture.Biochemistry 27:4162.

7. Allavena, P., G. Bianchi, D. Zhou, J. van Damme, P. Jilek, S. Sozzani, andA. Mantovani. 1994. Induction of natural killer cell migration by monocyte chemo-tactic protein-1, -2 and -3.Eur. J. Immunol. 24:3233.

8. Carr, M. W., S. J. Roth, E. Luther, S. S. Rose, and T. A. Springer. 1994. Monocytechemoattractant protein 1 acts as a T-lymphocyte chemoattractant.Proc. Natl. Acad.Sci. USA 91:3652.

9. Yoshimura, T., and E. J. Leonard. 1990. Identification of high affinity receptors forhuman monocyte chemoattractant protein-1 on human monocytes.J. Immunol. 145:292.

10. Charo, I. F., S. J. Myers, A. Herman, C. Franci, A. J. Connolly, and S. R. Coughlin.1994. Molecular cloning and functional expression of two monocyte chemoattractantprotein 1 receptors reveals alternative splicing of the carboxyl-terminal tails.Proc.Natl. Acad. Sci. USA 91:2752.

11. Kurihara, T., G. Warr, J. Loy, and R. Bravo. 1997. Defects in macrophage recruit-ment and host defense in mice lacking the CCR2 chemokine receptor.J. Exp. Med.186:1757.

FIGURE 4. Production of TNF-a in cocultures ofhuman monocytes with SBcl2 melanoma cells and in-duction of morphologic changes in HUVECs.A, Pro-duction of TNF-a by human monocytes coculturedovernight with SBcl2 melanoma cells infected withMCP-1-Ad5 at different PFU per cell,LacZ-Ad5 at 50PFU/cell, or not transduced (control). Values are pre-sented as means of duplicate assays for TNF-a pro-duction.B–J,In vitro angiogenesis using HUVECs un-der collagen gel and the following conditions:B,HUVECs cultured in M199 with 10% FCS, EGF, andheparin (optimal growth medium);C, HUVECs cul-tured in DMEM with 5% FCS (less vigorous growth inmedium for monocyte maintenance but no overallchanges in morphology);D, DMEM with 5% FCS andrMCP-1 (100 ng/ml) (growth similar to that observedin C); E, conditioned medium from an overnight cul-ture of human monocytes in DMEM with 5% FCS(growth similar to that observed inC); F, HUVECscultured in M199 medium without supplements (de-creased growth);G, conditioned medium from cocul-tures of human monocytes and SBcl2 cells transducedwith MCP-1-Ad5 at 0.5 PFU/cell in DMEM with 5%FCS (arrowheads indicate network-like structures ofendothelial cells);H, same as that observed inG but inthe presence of neutralizing Ab against TNF-a (100mg/ml); I, M199 medium and rTNF-a (50 ng/ml) (ar-rowheads indicate tubule-like structures); andJ, sameas that observed inI but in the presence of neutralizingAb against TNF-a (100mg/ml).

6489The Journal of Immunology

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from

12. Han, K. H., R. K. Tangirala, S. R. Green, and O. Quehenberger. 1998. Chemokinereceptor CCR2 expression and monocyte chemoattractant protein-1-mediated che-motaxis in human monocytes: a regulatory role for plasma LDL.Arterioscler.Thromb. Vasc. Biol. 18:1983.

13. Graves, D. T., and A. J. Valente. 1991. Monocyte chemotactic proteins from humantumor cells.Biochem. Pharmacol. 41:333.

14. Desbaillets, I., M. Tada, N. de Tribolet, A. C. Diserens, M. F. Hamou, andE. G. Van Meir. 1994. Human astrocytomas and glioblastomas express mono-cyte chemoattractant protein-1 (MCP-1) in vivo and in vitro.Int. J. Cancer58:240.

15. Leung, S. Y., M. P. Wong, L. P. Chung, A. S. Chan, and S. T. Yuen. 1997. Monocytechemoattractant protein-1 expression and macrophage infiltration in gliomas.ActaNeuropathol. 93:518.

16. Bottazzi, B., F. Colotta, A. Sica, N. Nobili, and A. A. Mantovani. 1990. Chemoat-tractant expressed in human sarcoma cells (tumor-derived chemotactic factor, TDCF)is identical to monocyte chemoattractant protein-1/monocyte chemotactic and acti-vating factor (MCP-1/MCAF).Int. J. Cancer 45:795.

17. Jiang, Y., A. J. Valente, M. J. Williamson, L. Zhang, and D. T. Graves. 1990. Post-translational modification of a monocyte-specific chemoattractant synthesized by gli-oma, osteosarcoma, and vascular smooth muscle cells.J. Biol. Chem. 265:18318.

18. Selvan, R. S., J. H. Butterfield, and M. S. Krangel. 1994. Expression of multiplechemokine genes by a human mast cell leukemia.J. Biol. Chem. 269:13893.

19. Isik, F. F., R. P. Rand, J. S. Gruss, D. Benjamin, and C. E. Alpers. 1996. Monocytechemoattractant protein-1 mRNA expression in hemangiomas and vascular malfor-mations.J. Surg. Res. 61:71.

20. Valkovic, T., K. Lucin, M. Krstulja, R. Dobi-Babic, and N. Jonjic. 1998. Expressionof monocyte chemotactic protein-1 in human invasive ductal breast cancer.Pathol.Res. Pract. 194:335.

21. Davidson, B., I. Goldberg, and J. Kopolovic. 1997. Inflammatory response in cervicalintraepithelial neoplasia and squamous cell carcinoma of the uterine cervix.Pathol.Res. Pract. 193:491.

22. Conti, P., C. Feliciani, R. C. Barbacane, S. Frydas, F. C. Placido, I. Cataldo, andM. Reale. 1999. Monocyte chemotactic protein-1 gene expression and translation informed granulomatous calcified tissue in vivo.Calcif. Tissue Int. 64:57.

23. Negus, R. P., G. W. Stamp, M. G. Relf, F. Burke, S. T. Malik, S. Bernasconi,P. Allavena, S. Sozzani, A. Mantovani, and F. R. Balkwill. 1995. The detection andlocalization of monocyte chemoattractant protein-1 (MCP-1) in human ovarian can-cer.J. Clin. Invest. 95:2391.

24. Hall, D. J., C. Brownlee, and C. D. Stiles. 1989. Interleukin-1 is a potent regulator ofJE and KC gene expression in quiescent BALB/c fibroblasts.J. Cell. Physiol. 141:154.

25. Takeshita, A., Y. Chen, A. Watanabe, S. Kitano, and S. Hanazawa. 1995. TGF-binduces expression of monocyte chemoattractant JE/monocyte chemoattractant pro-tein 1 via transcriptional factor AP-1 induced by protein kinase in osteoblastic cells.J. Immunol. 155:419.

26. Rollins, B. J., E. D. Morrison, and C. D. Stiles. 1988. Cloning and expression of JE,a gene inducible by platelet-derived growth factor and whose product has cytokine-like properties.Proc. Natl. Acad. Sci. USA 85:373.

27. Freter, R. R., J. C. Irminger, J. A. Porter, S. D. Jones, and C. D. Stiles. 1992. A novel7-nucleotide motif located in 39 untranslated sequences of the immediate-early geneset mediates platelet-derived growth factor induction of the JE gene.Mol. Cell. Biol.12:5288.

28. Huang, S., R. K. Singh, K. Xie, M. Gutman, K. K. Berry, C. D. Bucana, I. J. Fidler,and M. Bar-Eli. 1994. Expression of the JE/MCP-1 gene suppresses metastatic po-tential in murine colon carcinoma cells.Cancer Immunol. Immunother. 39:231.

29. Nakashima, E., N. Mukaida, Y. Kubota, K. Kuno, K. Yasumoto, F. Ichimura,I. Nakanishi, M. Miyasaka, and K. Matsushima. 1995. Human MCAF gene transferenhances the metastatic capacity of a mouse cachectic adenocarcinoma cell line invivo. Pharm. Res. 12:1598.

30. Rollins, B. J., and M. E. Sunday. 1991. Suppression of tumor formation in vivo byexpression of the JE gene in malignant cells.Mol. Cell. Biol. 11:3125.

31. Hirose, K., M. Hakozaki, Y. Nyunoya, Y. Kobayashi, K. Matsushita, T. Takenouchi,A. Mikata, N. Mukaida, and K. Matsushima. 1995. Chemokine gene transfection intotumor cells reduced tumorigenicity in nude mice in association with neutrophilicinfiltration. Br. J. Cancer 72:708.

32. Manome, Y., P. Y. Wen, A. Hershowitz, T. Tanaka, B. J. Rollins, D. W. Kufe, andH. A. Fine. 1995. Monocyte chemoattractant protein-1 (MCP-1) gene transduction:an effective tumor vaccine strategy for non-intracranial tumors.Cancer Immunol.Immunother. 41:227.

33. Zhang, L., T. Yoshimura, and D. T. Graves. 1997. Antibody to Mac-1 or monocytechemoattractant protein-1 inhibits monocyte recruitment and promotes tumor growth.J. Immunol. 158:4855.

34. Zunic, M., G. M. Bahr, G. C. Mudde, J. G. Meingassner, and C. Lam. 1998. MD-P(lysyl)GDP, a nontoxic muramyl dipeptide derivative, inhibits cytokine productionby activated macrophages and protects mice from phorbol ester- and oxazolone-induced inflammation.J. Invest. Dermatol. 111:77.

35. Kricek, F., M. Zunic, C. Ruf, G. DeJong, P. Dukor, and G. M. Bahr. 1997. Suppres-sion of in vivo IgE and tissue IL-4 mRNA induction by SDZ 280.636, a syntheticmuramyl dipeptide derivative.Immunopharmacology 36:27.

36. Leek, R. D., C. E. Lewis, R. Whitehouse, M. Greenall, J. Clarke, and A. L. Harris.1996. Association of macrophage infiltration with angiogenesis and prognosis in in-vasive breast carcinoma.Cancer Res. 56:4625.

37. Brocker, E. B., G. Zwadlo, L. Suter, M. Brune, and C. Sorg. 1987. Infiltration ofprimary and metastatic melanomas with macrophages of the 25F9-positive pheno-type.Cancer Immunol. Immunother. 25:81.

38. Graves, D. T., R. Barnhill, T. Galanopoulos, and H. N. Antoniades. 1992. Expressionof monocyte chemotactic protein-1 in human melanoma in vivo.Am. J. Pathol.140:9.

39. Torisu, H., M. Ono, H. Kiryu, M. Furue, Y. Ohmoto, J. Nakayama, Y. Nishioka,S. Sone, and M. Kuwano. 2000. Macrophage infiltration correlates with tumor stageand angiogensis in human malignant melanoma: possible involvement of TNF-alphaand IL-7-alpha.Int. J. Cancer 85:182.

40. Satyamoorthy, K., E. DeJesus, A. J. Linnenbach, B. Kraj, D. L. Kornreich, S. Rendle,D. E. Elder, and M. Herlyn. 1997. Melanoma cell lines from different stages ofprogression and their biological and molecular analyses.Melanoma Res. 7:35.

41. Herlyn, D., D. Iliopoulos, P. J. Jensen, A. Parmiter, J. Baird, H. Hotta, K. Adachi,A. H. Ross, J. Jambrosic, H. Koprowski, and M. Herlyn. 1990. In vitro properties ofhuman melanoma cells metastatic in nude mice.Cancer Res. 50:2296.

42. Horikawa, T., D. A. Norris, T. Zekman, and J. G. Morelli. 1996. Effective eliminationof fibroblasts in cultures of melanocytes by lowering calcium concentration in TPAdepleted medium following geneticin treatment.Pigm. Cell. Res. 9:58.

43. Zimrin, A. B., B. Villeponteau, and T. Maciag. 1995. Models of in vitro angiogenesis:endothelial cell differentiation on fibrin but not matrigel is transcriptionally depen-dent.Biochem. Biophys. Res. Commun. 213:630.

44. Freundlich, B., and N. Avdalovic. 1983. Use of gelatin/plasma coated flasks forisolating human peripheral blood monocytes.J. Immunol. Methods 62:31.

45. Nesbit, M., H. K. Nesbit, J. Bennett, T. Andl, M. Y. Hsu, E. Dejesus, M. McBrian,A. R. Gupta, S. L. Eck, and M. Herlyn. 1999. Basic fibroblast growth factor inducesa transformed phenotype in normal human melanocytes.Oncogene 18:6469.

46. D’Andrea, A., M. Aste-Amezaga, N. M. Valiante, X. Ma, M. Kubin, andG. Trinchieri. 1993. Interleukin 10 (IL-10) inhibits human lymphocyte interferong-production by suppressing natural killer cell stimulatory factor/IL-12 synthesis inaccessory cells.J. Exp. Med. 178:1041.

47. Bottazzi, B., S. Walter, D. Govoni, F. Colotta, and A. Mantovani. 1992. Monocytechemotactic cytokine gene transfer modulates macrophage infiltration, growth, andsusceptibility to IL-2 therapy of a murine melanoma.J. Immunol. 148:1280.

48. Clark, W. H., D. E. Elder, D. Guerry, M. N. Epstein, M. H. Greene, andM. A. Van Horn. 1984. Study of tumor progression: the precursor lesions of super-ficial spreading and nodular melanoma.Hum. Pathol. 15:1147.

49. Gunn, M. D., N. A. Nelken, X. Liao, and L. Williams. 1997. Monocyte chemoat-tractant protein-1 is sufficient for the chemotaxis of monocytes and lymphocytes intransgenic mice but requires an additional stimulus for inflammatory activation.J. Im-munol. 158:376.

50. Luini, W., S. Sozzani, J. Van Damme, and A. Mantovani. 1994. Species-specificityof monocyte chemotactic protein-1 and -3.Cytokine 6:28.

51. Sunderkoetter, C., K. Steinbrink, M. Goebeler, R. Bhardwaj, and C. Sorg. 1994.Macrophages and angiogenesis.J. Leukocyte Biol. 55:410.

52. Jiang, Y., D. I. Beller, G. Frendl, and D. T. Graves. 1992. Monocyte chemoattractantprotein-1 regulates adhesion molecule expression and cytokine production in humanmonocytes.J. Immunol. 148:2423.

53. Vaddi, K., and R. C. Newton. 1994. Regulation of monocyte integrin expression byb-family chemokines.J. Immunol. 153:4721.

54. Moretti, S., C. Pinzi, A. Spallanzani, E. Berti, A. Chiarugi, S. Mazzoli, M. Fabiani,C. Vallecchi, and M. Herlyn. 1999. Immunohistochemical evidence of cytokine net-works during progression of human melanocytic lesions.Int. J. Cancer 84:160.

55. Mattei, S., M. P. Colombo, C. Melani, A. Silvani, G. Parmiani, and M. Herlyn. 1994.Expression of cytokine/growth factors and their receptors in human melanoma andmelanocytes.Int. J. Cancer 56:853.

56. Eisengart, C. A., J. R. Mestre, H. A. Naama, P. J. Mackrell, D. E. Rivadeneira,E. M. Murphy, P. P. Stapleton, and J. M. Daly. 2000. Prostaglandins regulate mel-anoma-induced cytokine production in macrophages.Cell. Immunol. 204:143.

57. Netea, M. G., B. J. Kullberg, I. Verschueren, and J. W. Van Der Meer. 2000. Inter-leukin-18 induces production of proinflammatory cytokines in mice: no intermediaterole for the cytokines of the tumor necrosis factor family and the interleukin-1b. Eur.J. Immunol. 30:3057.

58. Lu, C., M. F. Vickers, and R. S. Kerbel. 1992. Interleukin 6: a fibroblast-derivedgrowth inhibitor of human melanoma cells from early but not advanced stages oftumor progression.Proc. Natl. Acad. Sci. USA 89:9215.

59. Goede, V., L. Brogelli, M. Ziche, and H. G. Augustin. 1999. Induction of inflam-matory angiogenesis by monocyte chemoattractant protein-1.Int. J. Cancer 82:765.

60. Salcedo, R., M. L. Ponce, H. A. Young, K. Wasserman, J. M. Ward, H. K. Kleinman,J. J. Oppenheim, and W. J. Murphy. 2000. Human endothelial cells express CCR2and respond to MCP-1: direct role of MCP-1 in angiogenesis and tumor progression.Blood 96:34.

61. Koch, A. E., P. J. Polverini, S. L. Kunkel, L. A. Harlow, L. A. Di Pietro, V. M. Elner,S. G. Elner, and R. M. Strieter. 1992. Interleukin-8 as a macrophage-derived mediatorof angiogenesis.Science 258:1798.

62. Ryuto, M., M. Ono, H. Izumi, S. Yoshida, H. A. Welch, K. Kohno, and M. Kuwano.1996. Induction of vascular endothelial growth factor by tumor necrosis factor-a inhuman glioma cells: possible role of Sp-1.J. Biol. Chem. 271:28220.

63. Yoshida, S., M. Ono, T. Shono, H. Izumi, T. Ishibashi, H. Suzuki, and M. Kuwano.1997. Involvement of interleukin-8, vascular endothelial growth factor, and basicfibroblast growth factor in tumor necrosis factora-dependent angiogenesis.Mol. Cell.Biol. 17:4015.

64. Sato, N., T. Goto, K. Haranaka, N. Satomi, H. Nariuchi, Y. Mano-Hirano, andY. Sawasaki. 1986. Actions of tumor necrosis factor on vascular endothelial cells:morphologic modulation, growth inhibition, and cytotoxicity.J. Natl. Cancer Inst.76:1113.

65. Leibovich, S. J., P. J. Polverini, H. M. Shepard, D. M. Wiseman, V. Shively, andN. Nuseir. 1987. Macrophage-induced angiogenesis is mediated by tumour necrosisfactor-a. Nature 329:630.

66. Fajardo, L. F., H. H. Kwan, J. Kowalski, S. D. Prionas, and A. C. Allison. 1992. Dualrole of tumor necrosis factor-a in angiogenesis.Am. J. Pathol. 140:539.

6490 MCP-1 EITHER STIMULATES OR INHIBITS MELANOMAS

by guest on January 10, 2022http://w

ww

.jimm

unol.org/D

ownloaded from