Lymphotoxin- Receptor Immune Interaction Promotes Tumor Growth

Home , Lymphotoxin

[CANCER RESEARCH 62, 4034–4040, July 15, 2002] Lymphotoxin-␤ Receptor Immune Interaction Promotes Tumor Growth by Inducing Angiogenesis1

Thomas Hehlgans, Benjamin Stoelcker, Peter Stopfer, Peter Mu¨ller, Grigore Cernaianu, Markus Guba, Markus Steinbauer, Sergei A. Nedospasov, Klaus Pfeffer, and Daniela N. Ma¨nnel2 Departments of Pathology/Tumor Immunology [T. H., B. S., P. S., P. M., D. N. M.] and Surgery [G. C., M. G., M. S.], University of Regensburg, D-93042 Regensburg, Germany; Institute of Medical Microbiology, Immunology, and Hygiene, Technical University of Munich, D-81675 Munich, Germany [K. P.]; and Laboratory of Molecular Immunology, Engelhardt Institute of Molecular Biology, Russian Academy of Sciences, and Belozersky Institute of Physico-Chemical Biology, Moscow State University, 119991 Moscow, Russia [S. A. N.]

ABSTRACT activated when the balance of angiogenic inhibitors to stimulators is shifted toward a proangiogenic milieu (9). Potent inducers of angio- Growth of solid fibrosarcoma tumors in mice was inhibited by the genesis have been identified, e.g., bFGF, VEGF, and angiopoietins release of a soluble lymphotoxin-␤ receptor inhibitor (LT␤R-immuno- (reviewed in Ref. 10). There is great interest in identifying and globulin fusion protein) from the tumor cells. Tumor growth arrest in ␣ ␤ modulating antiangiogenic pathways and in antiangiogenic drug de- mice deficient in the ligand LT 1 2 demonstrated the requirement for ␤ ␣ ␤ velopment for therapeutic purposes. activation of the LT R on the tumor cells by host cell-derived LT 1 2. Activation of the LT␤R resulted in enhanced release of macrophage Here we provide evidence that activation of LT␤R on fibrosarcoma inflammatory protein-2. Blocked angiogenesis was revealed in LT␤R tumor cells is necessary for angiogenesis and solid tumor growth. ␣ ␤ ␤ inhibitor-producing tumor nodules by immunohistochemistry and in vivo Prevention of LT 1 2-LT R signaling inhibited tumor angiogenesis microscopy. The growth arrest of LT␤R inhibitor-producing fibrosarco- and neovascularization, and resulted in tumor growth arrest. In addi- mas was overcome by forced MIP-2 expression in the tumor cells. Thus, tion, we show that LT␤R activation on the tumor cells induced LT␤R activation on tumor cells by activated host lymphocytes can initiate enhanced release of MIP-2, an angiogenic CXC chemokine (11). a novel proangiogenic pathway leading to organized tumor tissue devel- Thus, our studies identify the interaction of activated LT␣ ␤ -carying opment. 1 2 lymphocytes with LT␤R-expressing tumor cells as initiating event for a previously unknown proangiogenic pathway. INTRODUCTION ␣ ␤ 3 The membrane-bound heterotrimer consisting of LT 1 2 binds as MATERIALS AND METHODS a functional ligand specifically to the LT␤R, a member of the TNF receptor family. Interestingly, LT␤R is expressed on many cell types Animals. Male 20–25 g C57BL/6 mice were purchased from Charles River with the exception of lymphocytes, whereas expression of the ligand (Sulzfeld, Germany). Mice deficient for the LT␤R (6) and LT␣/LT␤-deficient is apparently restricted to activated lymphocytes (1). Currently the mice4 had been backcrossed more than six times to the C57BL/6 background. biological functions of this ligand-receptor system are not completely Experiments were performed with age- and sex-matched mice. understood. It has been shown that signaling through the LT␤R Tumor Cells. The methylcholanthrene-induced fibrosarcoma (BFS-1) was generated in a female C57BL/6 mice as described (12). The tumor cells were induced death in some human adenocarcinoma tumor lines (HT-29 ␥ maintained in vitro in DMEM high glucose medium (Life Technologies, Inc., and WiDr) in the presence of IFN- (2). Combined in vivo treatment Karlsruhe, Germany) supplemented with 5% heat inactivated FCS (Life Tech- of human adenocarcinoma cells (WiDr), which form solid tumors in nologies, Inc.) and 0.05 ng/ml gentamicin (PAA Laboratories, Linz, Austria). immunocompromised mice, with an agonistic anti-LT␤R antibody Tumor cells (1.5 ϫ 106 in 50 ␮l) were inoculated i.d. on the back of mice, and and human IFN-␥ resulted in tumor growth arrest. In addition, sig- tumor growth was measured as described recently (13). naling through LT␤R has been reported to induce NF␬B activation DNA Constructs and Transfectants. The mp55TNFR-Fc and mLT␤R-Fc and chemokine production in a cell type restricted manner (3, 4). expression constructs were generated by insertion of the extracellular domains Additional results have established a constitutive requirement for of the mp55TNFR or the mLT␤R into the Signal pIG plus vector (R&D LT␣ ␤ signaling in maintaining normal levels of secondary lymph- Systems, Wiesbaden, Germany). Stable transfectants expressing the 1 2 ␤ oid tissue chemokine and B lymphocyte chemoattractant in lymphoid p55TNFR-Fc or LT R-Fc fusion protein or MIP-2 were prepared by transfection of BFS-1 cells with the corresponding expression construct using N-[1- tissue (5). Recent studies using genetically modified mice indicated ␤ (2,3-dioleoyloxyl)propyl]-N,N,N-trimethylammoniummethyl sulfate (Roche that LT R is critically involved in lymphoid organogenesis and in the Diagnostics, Mannheim, Germany), following the manufacturer’s instructions. generation of adaptive humoral immune responses (6–8). Transfected cells were selected and maintained in G418 (0.8 mg/ml) or It is generally accepted today that tumor growth is angiogenesis- hygromycin (0.02 mg/ml; PAA Laboratories). dependent and that every increment of tumor growth requires an Flow Cytometry. Expression of mLT␤R on BFS-1 cells was detected by increment of vascular growth. Tumors lacking angiogenesis remain flow cytometry on a FACStar Plus (Becton Dickinson, San Jose, CA) using a dormant indefinitely and rapid logarithmic growth follows the acqui- specific rat anti-mLT␤R monoclonal antibody (mLTR1C5, IgG2a) followed sition of blood supply. The tumor angiogenic switch seems to be by a FITC-conjugated mouse antirat IgG at concentrations of 10 ␮g/ml or an irrelevant isotype-matched rat IgG as control. The mLTR1C5 had been generated by immunizing a rat with mLT␤R-Fc fusion protein, fusing the spleen Received 12/17/01; accepted 5/7/02. The costs of publication of this article were defrayed in part by the payment of page cells with SP2 myeloma cells (14), and screening the resulting hybridoma charges. This article must therefore be hereby marked advertisement in accordance with supernatants for positive staining of LT␤R-transfected Chinese hamster ovary 18 U.S.C. Section 1734 solely to indicate this fact. cells. 1 S. A. N. is International Research Scholar of Howard Hughes Medical Institute. 2 Western Blot Analysis. Supernatants of BFS-1 cells stably transfected To whom requests for reprints should be addressed, at Department of Pathology/Tumor ␤ Immunology, University of Regensburg, F.-J.-Strauss-Allee 11, D-93042 Regensburg, Ger- with the p55TNFR-Fc or the mLT R-Fc expression construct were incubated many. Phone: 49-941-944-6626; Fax: 49-941-944-6602; E-mail: daniela.maennel@klinik. with 10 mg protein G-Sepharose and incubated at 4°C for 1 h. Precipitated uni-regensburg.de. proteins were resolved on 12% SDS-PAGE and blotted on polyvinylidene 3 ␣ ␤ ␣ The abbreviations used are: LT 1 2, one molecule lymphotoxin- and two molecules lymphotoxin-␤;LT␤R, lymphotoxin-␤ receptor; TNF, tumor necrosis factor; NF␬B, nuclear factor ␬B; bFGF, basic fibroblast growth factor; VEGF, vascular endothelial 4 T. Plo¨tzel, M. B. Alimzhanov, D. Kuprash, S. A. Nedospasov, and K. Pfeffer. growth factor; MIP, macrophage inflammatory protein; MVD, microvascular density; Generation and characterization of mice with simultaneous inactivation of LT␣ and LT␤ ROI, region of interest; IL, interleukin; TNFR, tumor necrosis factor receptor. genes, manuscript in preparation. 4034

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2002 American Association for Cancer Research. TUMOR ANGIOGENESIS INDUCED BY LT␤ RECEPTOR ACTIVATION difluoride membrane (Immobilon-P, 0.45 ␮m; Millipore, Eschborn, Germany). shown by flow cytometric analysis using a monoclonal anti-mLT␤R The membrane was blocked with 5% dry milk in Tris-buffered saline for 1 h antibody (mLTR1C5; Fig. 1a). The expression of LT␤R was also at ambient temperature followed by incubation with antihuman IgG1 (1 mg/ confirmed at mRNA level by reverse transcription-PCR using specific ml) in incubation buffer (1% dry milk in Tris-buffered saline) for 1 h. After primers for mLT␤R (data not shown). washing, bound antihuman IgG1 was reacted with goat antimouse-conjugated BFS-1 cells stably transfected with cDNA of LT␤R-Fc or horseradish peroxidase (Sigma Biochemicals, Deisenhofen, Germany) in in- p55TNFR-Fc expressed comparable amounts of the corresponding cubation buffer for1hatambient temperature. The membrane was washed again and developed with the enhanced chemiluminescence kit (Energene, fusion proteins, whereas in the supernatants of nontransfected tumor Regensburg, Germany) following the manufacturer’s instructions. cells no fusion protein was detectable by Western blot analysis (Fig. Determination of MIP-2 and VEGF. MIP-2 and VEGF production was 1b). Binding of both secreted Fc fusion proteins to their corresponding determined using the corresponding ELISA kits (R&D Systems) according to ligands was verified by flow cytometric analysis in staining activated ␣ ␤ the manufacturer’s instructions. PMN1 cells, which express LT 1 2 on their surface (1) with super- Immunohistochemical Staining. Polyclonal antibodies to collagen type natants of BFS-1 cells secreting the LT␤R inhibitor and by immuno- IV (Novotec, Lyon, France) were used for immunohistochemical staining of precipitation of biotinylated rmTNF with supernatants of BFS-1 cells blood vessels as described recently (15). secreting the p55TNFR-Fc fusion protein, respectively (data not ␤ Determination of Cell Viability. Cell viability of wild-type or LT R-Fc shown). The in vitro growth characteristics of nontransfected BFS-1 expressing BFS-1 cells was tested by seeding the cells in 96-well plates at the cells and LT␤R inhibitor-expressing BFS-1 cells were not different indicated numbers overnight and subsequent culture in medium either with or ␤ without the agonistic anti-mLT␤R monoclonal antibody (10 ␮g/ml) for 24 h. even in the presence of an agonistic anti-LT R monoclonal antibody Cell viability was determined by adding 10 ␮l 3-(4,5-dimethyl-thiazol-2-yl)- 2,5-diphenyltetrazolium bromide dye (5 mg/ml; Sigma) at 37°C for 4 h and lysing the cells by adding 70 ␮l of 10% SDS per well. The plates were kept at 37°C overnight, and absorbance was measured at 540 nm. Each experiment was repeated at least three times. Intravital Microscopy. Tumor angiogenesis was quantified in the trans- parent dorsal skin fold as described in detail (16, 17). Mice were shaved on the back, anesthetized with ketamine hydrochloride (7.5 mg/100 mg body weight; Parke, Davis & Company, Mu¨nchen, Germany) and xylazine (2.5 mg/100 mg body weight; Bayer AG, Leverkusen, Germany), and placed on a heating pad. One pair of titanium frames was implanted in a dorsal skin fold parallel to the dorsum so as to sandwich the stretched double layer of skin. One layer of the dorsal skin was removed in a circular area of 10-mm diameter using an operation microscope to facilitate implantation. The underlying thin layer of striated skin muscle, s.c. tissue, and epidermis was sealed with a coverslip enclosed in one of the frames. Two days later the coverslip was removed and BFS-1 cells transfected with either the LT␤R-Fc or the p55TNFR-Fc construct (3 ϫ 105 cells/animal; n ϭ 6 animals/group) were carefully placed on the upper tissue layer as a pellet and the chamber closed again. Tumor-bearing animals were monitored in regular time intervals for the development of tumor vessels (MVD ϭ total vascular length/observation area). For intravital microscopy the mice were placed in a Plexiglas tube with the chamber extending from a longitudinal slit and immobilized on a platform. At days 2 and 6 after tumor implantation, 10 ROIs were defined and examined by intravital microscopy (8 ROIs distributed clockwise at the tumor margin and 2 ROIs in the tumor center) after relocation by a computer-controlled stepping motor system. In vivo microscopy was performed using a modified Axiotech Vario microscope (Zeiss, Oberkochen, Germany) equipped with a filter set 09 (BP 450–490, FT 510, and LP 520; Zeiss). Observations were made using ϫ2.5 long distance, and ϫ10 and ϫ20 water immersion working objectives (Zeiss) resulting in a magnification of ϫ53, ϫ213, and ϫ425, respectively. Observations of the window chambers were carried out in white light transillumination technique and in epi-illumination technique after injection of 0.05 ml 5% FITC dextran 500 (mW ϭ 500,000) for contrast enhancement. Images were recorded through a CCD video camera (PCO, Kehlheim, Germany) on S-VHS tapes for later offline analysis. MVD was determined as described (18) using a modified custom-made image software (IDL). Statistical Analysis. Data represent one of at least three independent experiments and are shown as the mean ϩ SD. Statistical significance of data were determined using Student’s t test.

RESULTS Fig. 1. Expression of mLT␤R and fusion proteins by BFS-1 tumor cells. a, flow To dissect the biological consequences of LT␤R signaling in tumor cytometric analysis of fibrosarcoma cells (BFS-1) using a monoclonal anti-mLT␤R development the growth of syngeneic LT␤R-positive fibrosarcoma antibody (white area) or an isotype-matched control antibody (shaded area). b, immu- noprecipitation of LT␤R-Fc fusion proteins from supernatants of untransfected BFS-1 cells expressing a LT␤R-Fc fusion protein acting as an inhibitor of cells (Lane 1), transfected with LT␤R-Fc (Lane 2), or p55TNFR-Fc (Lane 3) expression LT␤R activation was compared with growth of tumor cells expressing constructs and subsequent detection using antihuman IgG1 antibody. c, cell viability of BFS-1 wild-type (circles) and LT␤R inhibitor-expressing cells (squares) in the presence a p55TNFR-Fc fusion protein as control. Fibrosarcoma cells (BFS-1) (closed symbols) or absence (open symbols) of an agonistic monoclonal anti-LT␤R syngeneic to C57BL/6 mice expressed LT␤R on the cell surface as antibody. 4035

indicating that the growth inhibitory effect of LT␤R inhibitor was dose dependent. To: (a) support the observation that activation of the LT␤Ris necessary for BFS-1 tumor growth; (b) clearly identify ligand speci- ␣ ␤ ficity; and (c) additionally determine the source of LT 1 2, wild-type BFS-1 tumor cells were inoculated into normal syngeneic C57BL/6 and into LT␣/LT␤-deficient mice. Comparable retardation of tumor growth was observed in LT␣/LT␤-deficient mice as with LT␤R inhibitor-expressing BFS-1 in normal mice (Fig. 3a). This result ␣ ␤ ␤ clearly identified LT 1 2 as a functional ligand for the LT R activation and also determined the host as the source for the ligand. Additionally, by using reverse transcription-PCR we were able to detect mRNA for both ligand molecules LT␣ and LT␤ in explanted fibrosarcoma tumor tissue (data not shown). To identify the responsible receptor-carrying cells, LT␤R-deficient mice were inoculated with wild-type BFS-1 cells. Tumor growth in such LT␤R-deficient mice was indistinguishable from tumor growth in normal wild-type mice (Fig. 3b) demonstrating that LT␤Ronthe tumor cells themselves is required for unimpaired growth. Histological examination of tissue sections from tumors on day 9 revealed no obvious difference between wild-type and LT␤R inhibitor-expressing BFS-1 tumors. No significant difference was observed in numbers of infiltrating lymphocytes, cells in the mitotic phase, Tec-3 positive cells, and cells with pyknotic nuclei as a sign for dying cells (Fig. 4a; data not shown). However, staining for collagen type IV as a marker for vessel formation (15) indicated reduced vascular- ization in those tumors consisting of LT␤R inhibitor-expressing BFS-1 cells (Fig. 4b). Numbers as well as size of newly formed vessels were reduced in LT␤R inhibitor-producing BFS-1 tumor nodules.

Fig. 2. Tumor growth of BFS-1 transfectants. a, solid tumor growth of syngeneic fibrosarcoma cells BFS-1 (E), BFS-1 cells expressing the LT␤R inhibitor (F), or the p55TNFR-Fc fusion protein (f) in C57BL/6 mice (n ϭ 5). b, solid tumor growth of BFS-1 cells consisting of 100% (F), 75% (ࡗ), 50% (Œ), 25% (f), or 0% (E)ofLT␤R-Fc expressing cells in C57BL/6 mice (n ϭ 5). indicating that neither overexpression of LT␤R inhibitor nor stimulation of the LT␤R interfered with BFS-1 cell proliferation and survival in vitro (Fig. 1c). Solid tumor growth in vivo of BFS-1 cells expressing LT␤R inhibitor was inhibited compared with growth of wild-type or control p55TNFR-Fc fusion protein-expressing tumor cells (Fig. 2a). Com- parable results were obtained from independent transfection experiments and also by using pool transfectants or four individual stably transfected BFS-1 clones (data not shown). To investigate the possi- bility of whether rejection of the LT␤R-Fc transfectants may account for this inhibition of tumor growth the growth pattern of wild-type and transfected tumor cells in allogeneic mice was compared. Wild-type tumors, LT␤R inhibitor-, and p55TNFR-Fc-expressing tumors were completely eradicated by days 9–12 after inoculation (data not shown). In contrast, LT␤R inhibitor-expressing tumor cells on syngeneic mice grew to a palpable size of ϳ2 mm in diameter and later on showed a highly suppressed growth characteristic for Ͼ40 days (Fig. 2a; data not shown). To test whether the amount of secreted LT␤R inhibitor influences the growth of BFS-1 tumor cells, mixtures of tumor cells containing increasing percentages of LT␤R inhibitor- expressing BFS-1 were inoculated. Tumor growth was inhibited when 50% or more of the tumor inoculate consisted of LT␤R inhibitor- Fig. 3. Tumor growth in wild-type, LT␣/␤-deficient, or LT␤R-deficient mice. a, solid tumor growth of BFS-1 cells (E), BFS-1 cells expressing the LT␤R inhibitor (F)in expressing BFS-1 cells (Fig. 2b). The inhibitor produced from 25% of C57BL/6 mice, and BFS-1 cells (Ⅺ)inLT␣/␤-deficient mice (n ϭ 4). b, solid tumor the inoculated tumor cells was only sufficient to delay tumor growth growth of BFS-1 cells in C57Bl/6 mice (E) and in LT␤R-deficient mice (F; n ϭ 5). 4036

Fig. 4. Histology of BFS-1 tumors and LT␤R inhibitor-transfected BFS-1 tumors. a, H&E staining of tumor tissue derived from BFS-1 cells (A) and from BFS-1 cells expressing the LT␤R inhibitor (B). b, immunohistochemical staining of blood vessels in tumor tissue derived from BFS-1 cells (A) and from BFS-1 cells expressing the LT␤R inhibitor (B) using collagen type IV polyclonal antibodies.

To follow angiogenesis in growing tumors, intravital microscopy the LT␤R inhibitor-expressing tumors was impaired, MIP-2 cotrans- was performed with the two tumor cell transfectants. Hereby it be- fected tumors grew much better (Fig. 6c). Regaining of growth by came clear that implantation of LT␤R inhibitor-expressing BFS-1 MIP-2 expression clearly indicated that activation of the LT␤R and cells led to inhibited sprouting of new vessels from the surrounding ensuing up-regulation of MIP-2 expression is important for growth of skin area into the direction of the growing tumors from days 3 to 4 on BFS-1 fibrosarcoma tumors. (Fig. 5a). This observation was substantiated by quantitative evaluation of the angiogenesis on day 6 (Fig. 5b). DISCUSSION The supernatant of BFS-1 cells stimulated with the agonistic monoclonal anti-LT␤R antibody mLTR1C5 was tested for the presence of Tumor cell cytotoxicity induced by LT␤R activation has been the angiogenic mediators VEGF and MIP-2. Whereas the amount of described for a very limited range of tumor cells (2). However, VEGF generated into the supernatant by the tumor cells was not activation of the LT␤R on BFS-1 tumor cells by an agonistic mono- affected by the addition of the antibody, increased amounts of MIP-2 clonal antibody mLTR1C5 did not affect tumor cell proliferation in were released from BFS-1 cells after LT␤R stimulation (Fig. 6a). our experiments. The tumor cells did not show any signs of cell death Comparison of MIP-2 production ex vivo from explanted wild-type when stimulated in vitro with the antibody neither in the presence nor and LT␤R inhibitor-expressing BFS-1 tumor tissue, respectively, on in the absence of IFN-␥ (data not shown). Apart from the direct day 8 revealed a significantly higher MIP-2 expression in wild-type antitumoral effect other biological functions of LT␤R activation have tumors (Fig. 6b). This finding correlated with the impaired MIP-2 been reported as well. NF␬B activation and the release of IL-8 and production in vitro and the reduced tumor growth in vivo. RANTES from A375 human melanoma cells have been described as To additionally investigate whether MIP-2 production as a result of results of LT␤R stimulation (3, 4). Similar functional activity was LT␤R stimulation is important for growth of BFS-1 tumors and ascertained in the mouse system in this report by documenting MIP-2 because preliminary experiments with systemic neutralization of production from BFS-1 cells after stimulation with the agonistic MIP-2 failed, LT␤R inhibitor-expressing BFS-1 cells were cotrans- antibody mLTR1C5. Also, NF␬B activation was induced by fected with a MIP-2 expression construct. Whereas, again, growth of mLTR1C5 in LT␤R-positive BFS-1 and L929 cells (data not shown). 4037

Fig. 5. Blocked angiogenesis in tumors of LT␤R inhibitor-transfected BFS-1 cells. a, intravital microscopy of tumor angiogenesis on day 9 after implantation of BFS-1 cells expressing the LT␤R inhibitor (A)orthe p55TNFR-Fc fusion protein (B). b, quantitative evaluation of angiogenesis in tumors on day 6 after implantation of BFS-1 cells expressing ;(P Ͻ 0.013 ء) the LT␤R inhibitor or the p55TNFR-Fc fusion protein bars, ϮSD.

Our experiments with LT␣/b gene-deficient mice clearly dem- lular domains and share a TRAF-2 binding domain, which has been onstrate that host cells are the source for the stimulating ligand. found to be responsible for NF␬B activation (20). The increase of Because LT␤, the membrane anchoring part of the heterotrimeric MIP-2 production after stimulation of the LT␤R might be explained ␣ ␤ ligand LT 1 2, is exclusively produced by activated lymphocytes by the fact that the promoters of both IL-8 and MIP-2, which is the ␣ ␤ and natural killer cells, the cellular source of LT 1 2 needs to be murine functional homologue of IL-8 (21), contain binding sites for determined in future experiments with tissue-restricted LT␣/b regulatory elements, e.g., NF␬B, which are activated after LT␤R gene-deficient mice. Obviously, LT␤R activation is not only re- stimulation (22, 23). quired for lymphoid organogenesis but also for the development of The observation that angiogenesis correlates with tumor malig- solid tumor tissue. This implies new significance for the function nancy is well accepted (10). Our finding of a dose-dependent inhib- of this interesting member of the TNF receptor family, which itory effect of the LT␤R inhibitor produced by the tumor cells fits the stands at the interface of the immune system interacting with idea of angioneogenesis as a limiting factor for solid tumor growth. cellular systems indispensable for organogenesis. Implantation of LT␤R inhibitor-expressing tumor cells on one side of The fact that activation of the LT␤R can initiate cytotoxic as well the back of the mouse failed to affect wild-type tumor growth on the as cell protective signaling pathways, depending on the cell type or the other side of the back demonstrating the local restriction of the action experimental conditions, is reminiscent of the signaling pathways (data not shown). Enhanced MIP-2 production from the explanted following activation of the TNFR type 2. These two receptors are growing tumor compared with LT␤R inhibitor-producing tumors was more closely related than the two TNFR with each other (19). In demonstrated. Therefore, in our model the CXC chemokine MIP-2 addition, both receptors lack classical death domains in their intracel- could potentially attract and activate neutrophils and lymphocytes, 4038

might be envisaged as supported by data obtained in the human system. The human IL-8 is known as a multifunctional cytokine that has both angiogenic effects and potent leukocyte chemotactic and activation properties (24–27). IL-8 has been shown to stimulate chemotaxis and proliferation of human umbilical vein endothelial cells similar to bFGF, and was found to be proangiogenic in vivo (28). In addition, IL-8 has been found in tissues of many neoplastic diseases (29–31), and to be functional for neovascularization, both IL-8 and IL-8 receptors must be present within the tumor microenvironment (32). The presence of IL-8 receptors on endothelial cells lining the vessels in human tumor tissue, which is critical for the recognition of IL-8 as an angiogenic factor, has been described recently (32). Pres- ence of IL-8 receptors on the tumor cells themselves could result in enhanced tumor cell proliferation. MIP-2, which is structurally ho- mologous to the human CXC chemokines GRO-␤/␥, represents the murine homologue of IL-8 in terms of function (21, 33) and binds to the mouse homologue of the human IL-8 receptor (34, 35). It has been reported recently that like IL-8 MIP-2 is also chemotactic for endothelial cells and induces neovascularization (11). Whether the mouse IL-8 receptor homologue is expressed on endothelial cells or on BFS-1 cells is presently not known. The reconstitution of tumor growth in the presence of the LT␤R inhibitor by MIP-2 supplementation firmly supports the idea that activation of the LT␤R on the BFS-1 tumor cells by interaction with activated host lymphocytes is a prerequisite for enhanced MIP-2 production. This angiogenic chemokine might, in a para- crine manner, diffuse to the nearest blood vessel signaling, either directly or indirectly, the endothelial cells to start the angiogenic process. Very clearly, different tumor types have developed different mechanisms to induce angiogenesis. In regard to potential therapeutic intervention strategies the finding that in some, tumors angiogenesis is initiated by activated host lymphocytes activating the LT␤R on tumor cells for subsequent MIP-2 production cer- tainly deserves additional investigations.

ACKNOWLEDGMENTS

We thank M. Alimzhanov for outstanding contribution to the development of LT␣/LT␤ double-deficient mice, B. Ruhland for excellent technical assist- ance, Drs. R. C. Krieg and O. Gleich for photographic documentation, and Dr. D. Breitkreutz for helpful discussions.

REFERENCES 1. Force, W. R., Walter, B. N., Hession, C., Tizard, R., Kozak, C. A., Browning, J. L., and Ware, C. F. Mouse lymphotoxin-␤ receptor. Molecular genetics, ligand binding, and expression. J. Immunol., 155: 5280–5288, 1995. 2. Browning, J. L., Miatkowski, K., Sizing, I., Griffiths, D., Zafari, W., Benjamin, C. D., Meier, W., and Mackay, F. Signaling through the lymphotoxin ␤ receptor induces the death of some adenocarcinoma tumor lines. J. Exp. Med., 183: 1756–1762, 1997. 3. Degli-Esposti, M. A., Davis-Smith, T., Din, W. S., Smolak, P. J., Goodwin, R. G., and Smith, C. A. Activation of the lymphotoxin ␤ receptor by cross-linking induces chemokine production and growth arrest in A375 melanoma cells. J. Immunol., 158: 1756–1762, 1997. 4. Mackay, F., Majeau, G. R., Hochman, P. S., and Browning, J. L. Lymphotoxin ␤ Fig. 6. MIP-2 production correlates with tumor growth. a, enhanced MIP-2 release receptor triggering induces activation of the nuclear factor ␬B transcription factor in after LT␤R activation. VEGF and MIP-2 levels in supernatants of BFS-1 cells 24 h after some cell types. J. Biol. Chem., 271: 24934–24938, 1996. addition of a monoclonal anti-mLT␤R antibody. b, MIP-2 release of explanted BFS-1 5. Ngo, V. N., Korner, H., Gunn, M. D., Schmidt, K. N., Riminton, D. S., Cooper, M. D., wild-type (F)orLT␤R-expressing (Œ) tumor tissue at day 8 (P Ͻ 0.038). c, solid tumor Browning, J. L., Sedgwick, J. D., and Cyster, J. D. Lymphotoxin ␣/␤ and tumor growth of syngeneic fibrosarcoma cells expressing the LT␤R inhibitor (E) or cotrans- necrosis factor are required for stromal cell expression of homing chemokines in B fected with a MIP-2 expression construct (F) in C57BL/6 mice (n ϭ 5). and T cell areas of the spleen. J. Exp. Med., 189: 403–412, 1999. 6. Futterer, A., Mink, K., Luz, A., Kosco-Vilbois, M. H., and Pfeffer, K. The lymphotoxin ␤ receptor controls organogenesis and affinity maturation in peripheral lymphoid tissues. Immunity, 9: 59–70, 1998. which then might release angiogenic factors like VEGF and bFGF to 7. Rennert, P. D., James, D., Mackay, F., Browning, J. L., and Hochman, P. S. Lymph node genesis is induced by signaling through the lymphotoxin ␤ receptor. Immunity, induce vessel growth in the tumor microenvironment. Such support 9: 71–79, 1998. for tumor angiogenesis can become critical when tumor cells only 8. Mackay, F., and Browning J. L. Turning off follicular dendritic cells. Nature (Lond.), 395: 26–27, 1998. insufficiently prepare their microenvironment directly. 9. Hanahan, D., and Folkman, J. Patterns and emerging mechanisms of the angiogenic Alternatively, a direct local effect of MIP-2 on endothelial cells switch during tumorigenesis. Cell, 86: 353–364, 1996. 4039

10. Kerbel, R. S. Tumor angiogenesis: past, present and the near future. Carcinogenesis ␤, members of the chemokine superfamily of proinflammatory cytokines. J. Immu- (Lond.), 21: 505–515, 2000. nol., 150: 4996–5012, 1993. 11. Keane, M. P., Belperio, J. A., Moore, T. A., Moore, B. B., Arenberg, D. A., Smith, 23. Roebuck, K. A. Regulation of interleukin-8 gene expression. J. Interferon Cytokine R. E., Burdick, M. D., Kunkel, S. M., and Strieter, R. M. Neutralization of the CXC Res., 19: 429–438, 1999. chemokine, macrophage inflammatory protein-2, attenuates bleomycin-induced pul- 24. Matsushima, K., and Oppenheim, J. J. Interleukin 8 and MCAF: novel inflammatory monary fibrosis. J. Immunol., 162: 5511–5518, 1999. cytokines inducible by IL 1 and TNF. Cytokine, 1: 2–13, 1989. 12. Stoelcker, B., Ruhland, B., Hehlgans, T., Bluethmann, H., Luther, T., and Ma¨nnel, 25. Thelen, M., Peveri, P., Kernen, P., Walz, A., and Baggiolini, M. Mechanism of D. N. Tumor necrosis factor induces tumor necrosis via tumor necrosis factor receptor neutrophil activation by NAF, a novel monocyte-derived peptide agonist. FASEB J., type 1-expressing endothelial cells of the tumor vasculature. Am. J. Pathol., 156: 2: 2702–2706, 1988. 1171–1176, 2000. 26. Colditz, I., Zwahlen, R., Dewald, B., and Baggiolini, M. In vivo inflammatory activity 13. O’Reilly, M. S., Holmgren, L., Chen, C., and Folkman, J. Angiostatin induces and of neutrophil-activating factor, a novel chemotactic peptide derived from human sustains dormancy of human primary tumors in mice. Nat. Med., 2: 689–692, 1996. monocytes. Am. J. Pathol., 134: 755–760, 1989. 14. Shulman, M., Wilde, C. D., and Kohler, G. A better cell line for making hybridomas 27. Strieter, R. M., Kunkel, S. L., Elner, V. M., Martonyi, C. L., Koch, A. E., Polverini, secreting specific antibodies. Nature (Lond.), 276: 269–270, 1978. P. J., and Elner, S. G. Interleukin-8. A corneal factor that induces neovascularization. 15. Breitkreutz, D., Stark, H. J., Mirancea, N., Tomakidi, P., Steinbauer, H., and Fusenig, Am. J. Pathol., 141: 1279–1284, 1992. N. E. Integrin and basement membrane normalization in mouse grafts of human 28. Koch, A. E., Polverini, P. J., Kunkel, S. L., Harlow, L. A., DiPietro, L. A., Elner, keratinocytes–implications for epidermal homeostasis. Differentiation, 61: 195–209, V. M., Elner, S. G., and Strieter, R. M. Interleukin-8 as a macrophage-derived 1997. mediator of angiogenesis. Science (Wash. DC), 258: 1798–1801, 1992. 16. Endrich, B., Asaishi, K., Gotz, A., and Messmer, K. Technical report–a new chamber 29. Singh, R. K., Gutman, M., Radinsky, R., Bucana, C. D., and Fidler, I. J. Expression technique for microvascular studies in unanesthetized hamsters. Res. Exp. Med., 177: of interleukin 8 correlates with the metastatic potential of human melanoma cells in 125–134, 1980. nude mice. Cancer Res., 54: 3242–3247, 1994. 17. Leunig, M., Yuan, F., Menger, M. D., Boucher, Y., Goetz, A. E., Messmer, K., and 30. Weidner, N., Carroll, P. R., Flax, J., Blumenfeld, W., and Folkman, J. Tumor Jain, R. K. Angiogenesis, microvascular architecture, microhemodynamics, and in- angiogenesis correlates with metastasis in invasive prostate carcinoma. Am. J. terstitial fluid pressure during early growth of human adenocarcinoma LS174T in Pathol., 143: 401–409, 1993. SCID mice. Cancer Res., 52: 6553–6560, 1992. 31. Cohen, R. F., Contrino, J., Spiro, J. D., Mann, E. A., Chen, L. L., and Kreutzer, D. L. 18. Nolte, D., Zeintl, H., Steinbauer, M., Pickelmann, S., and Messmer, K. Functional Interleukin-8 expression by head and neck squamous cell carcinoma. Arch. Otolar- capillary density: an indicator of tissue perfusion? Int. J. Microcirc. Clin. Exp., 15: yngol. Head Neck Surg., 121: 202–209, 1995. 244–249, 1995. 32. Miller, L. J., Kurtzman, S. H., Wang, Y., Anderson, K. H., Lindquist, R. R., and 19. Crowe, P. D., VanArsdale, T. L., Walter, B. N., Ware, C. F., Hession, C., Ehrenfels, Kreutzer, D. L. Expression of interleukin-8 receptors on tumor cells and vascular B., Browning, J. L., Din, W. S., Goodwin, R. G., and Smith, C. A. A lymphotoxin- endothelial cells in human breast cancer tissue. Anticancer Res., 18: 77–81, 1988. ␤-specific receptor. Science (Wash. DC), 264: 707–710, 1994. 33. Wolpe, S. D., Sherry, B., Juers, D., Davatelis, G., Yurt, R. W., and Cerami, A. 20. Rothe, M., Wong, S. C., Henzel, W. J., and Goeddel, D. V. A novel family of putative Identification and characterization of macrophage inflammatory protein 2. Proc. Natl. signal transducers associated with the cytoplasmic domain of the 75 kDa tumor Acad. Sci. USA, 86: 612–616, 1989. necrosis factor receptor. Cell, 78: 681–692, 1994. 34. Bozic, C. R., Gerard, N. P., Uexkull-Guldenband, C., Kolakowski, L. R., Conklyn, 21. Driscoll, K. E., Hassenbein, D. G., Howard, B. W., Isfort, R. J., Cody, D., Tindal, M. J., Breslow, R., Showell, H. J., and Gerard, C. The murine interleukin 8 type B M. H., Suchanek, M., and Carter, J. M. Cloning, expression, and functional charac- receptor homologue and its ligands. Expression and biological characterization. terization of rat MIP-2: a neutrophil chemoattractant and epithelial cell mitogen. J. Biol. Chem., 269: 29355–29358, 1994. J. Leukoc. Biol., 58: 359–364, 1995. 35. Lee, J., Cacalano, G., Camerato, T., Toy, K., Moore, M. W., and Wood, W. I. 22. Widmer, U., Manogue, K. R., Cerami, A., and Sherry, B. Genomic cloning and Chemokine binding and activities mediated by the mouse IL-8 receptor. J. Immunol., promoter analysis of macrophage inflammatory protein (MIP)-2. MIP-1 ␣, and MIP-1 155: 2158–2164, 1995.

4040

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2002 American Association for Cancer Research. Lymphotoxin-β Receptor Immune Interaction Promotes Tumor Growth by Inducing Angiogenesis

Thomas Hehlgans, Benjamin Stoelcker, Peter Stopfer, et al.

Cancer Res 2002;62:4034-4040.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/62/14/4034

Cited articles This article cites 32 articles, 12 of which you can access for free at: http://cancerres.aacrjournals.org/content/62/14/4034.full#ref-list-1

Citing articles This article has been cited by 11 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/62/14/4034.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/62/14/4034. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.