Metastases Mouse Endostatin Inhibits the Formation ...

Mouse Endostatin Inhibits the Formation of Lung and Liver Metastases Sam S. Yoon, Hiroshi Eto, Cheng-mao Lin, et al. Cancer Res 1999;59:6251-6256.

Updated version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/59/24/6251

Cited Articles

This article cites by 17 articles, 4 of which you can access for free at: http://cancerres.aacrjournals.org/content/59/24/6251.full.html#ref-list-1

Citing articles

This article has been cited by 21 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/59/24/6251.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on June 1, 2013. © 1999 American Association for Cancer Research.

[CANCER RESEARCH 59, 6251– 6256, December 15, 1999]

Mouse Endostatin Inhibits the Formation of Lung and Liver Metastases1 Sam S. Yoon, Hiroshi Eto2, Cheng-mao Lin, Hideo Nakamura, Timothy M. Pawlik, Sun U. Song, Kenneth K. Tanabe3 Division of Surgical Oncology, Department of Surgery, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts 02114

ABSTRACT Angiogenesis is required for tumor formation. Several studies have demonstrated that tumor angiogenesis is regulated by a balance between proangiogenesis and antiangiogenesis factors and that this balance varies in different organ environments. To investigate whether expression of an angiogenesis inhibitor by cancer cells could alter this balance and prevent tumor formation in different organ environments, we engineered stable transfectants from RenCa mouse renal carcinoma cells and SW620 human colon carcinoma cells to constitutively secrete a mouse endostatin protein with c-myc and polyhistidine (His) tags. Production and secretion of the endostatin-c-myc-His fusion protein by endostatin-transfected cells were confirmed by immunofluorescence staining and Western blot analysis. The endostatin transfectants and control transfectants, stably transfected with a control plasmid, had similar in vitro growth rates compared with their parental cell lines. Conditioned medium from endostatin-transfected cells inhibited human umbilical vein endothelial cell proliferation by 36 –51% compared with conditioned medium from control cells. After inoculation into mice, flank tumors from endostatin-transfected cells were 73–91% smaller than flank tumors from control cells after 3 weeks. Inoculation of a cell mixture containing 25% endostatin-transfected cells and 75% control cells resulted in inhibition of flank tumor formation as effective as after inoculation of 100% endostatin-transfected cells. Formation of lung metastases by RenCa endostatin-transfected cells and formation of liver metastases by SW620 endostatin-transfected cells were dramatically inhibited compared with formation of metastases by control cells. These findings demonstrate that endostatin can inhibit tumor formation in different organ environments and that gene delivery of endostatin into even a minority of tumor cells may be an effective strategy to prevent progression of micrometastases to macroscopic disease.

capsule of nude mice produce 10 –20 times more bFGF4 mRNA than when implanted into s.c. tissue (4). Similarly, LS174T human colon carcinoma cells produce less VEGF mRNA when experimentally grown in the liver than when grown in s.c. tissue (5). In addition, the endothelial cells that form new tumor blood vessels differ according to the host organ. New tumor blood vessels in the lung originate from capillary endothelial cells (2), whereas new tumor blood vessels in the liver originate from sinusoidal endothelial cells (6) that, unlike lung capillary endothelial cells, are fenestrated and lack a basement membrane (7). Because of these differences, the influence of antiangiogenesis factors on tumor growth may vary in different host organ environments. However, few studies have examined this aspect of angiogenesis. One of the more potent antiangiogenesis agents, endostatin, is a cleavage product consisting of the COOH-terminal 184 amino acids of collagen XVIII (8). This Mr 20,000 protein inhibits endothelial cell migration and proliferation (8 –11), and induces the regression of a wide variety of tumors grown s.c. in mice (8, 10, 11). The efficacy of endostatin in organ environments other than s.c. tissue is not well established. Endostatin has been shown in two studies to inhibit the formation or growth of lung metastases (11, 12). To our knowledge, there are no published reports on the efficacy of endostatin against liver metastases. The aim of the present study was to determine the efficacy of endostatin in preventing tumor formation in different organ environments. We engineered stable transfectants from RenCa mouse renal carcinoma cells and SW620 human colon carcinoma cells that constitutively secrete mouse endostatin and determined the effect of endostatin on tumor formation in the flank, lung, and liver.

INTRODUCTION MATERIALS AND METHODS Paget noted in 1889 that certain tumors metastasize preferentially to specific organs and emphasized that growth of metastases is dependent upon not only the properties of the tumor cells but also upon the environment in which the tumors cells grow (1). Similarly, tumor angiogenesis, or the development of new tumor blood vessels, is critically dependent on the interaction between tumor cells and the host organ environment (2). This process is regulated by a balance between proangiogenesis and antiangiogenesis factors (3). Metastatic cells from a specific cancer may be able to establish a balance favorable for tumor angiogenesis in some organs but not in others. Tumor angiogenesis varies in different organ environments. For example, human renal carcinoma cells implanted under the renal Received 6/15/99; accepted 10/19/99. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work is supported by NIH Grants CA64454 (to K. K. T.), GM21700 (to C-M. L.), DK43352 (core facilities), CA76656 (to S. U. S.), and CA71345 (to S. S. Y.) and the Claude E. Welch Research Fellowship (to S. S. Y.). 2 Present address: Department of Urology, Kobe University School of Medicine, 7-5-1, Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan. 3 To whom requests for reprints should be addressed, at Division of Surgical Oncology, Department of Surgery, Massachusetts General Hospital, Cox 626, 100 Blossom Street, Boston, MA 02114. Phone: (617) 724-3868; Fax: (617) 724-3895; E-mail: ktanabe@ partners.org.

Plasmids. The plasmid pTBO1#8 (8) containing the cDNA for mouse endostatin was generously provided by Dr. J. Folkman (Harvard Medical School, Boston, MA). Mouse endostatin cDNA was PCR amplified using a standard protocol with pTBO1#8 as template and two oligonucleotide primers, 59-CACGGTACCCATACTCATCAGGACTTTCAG-39 and 59-AGGGGATCCTTGGAGAAAGAGGTCATGAAG-39. These primers introduced KpnI and BamHI restriction sites at the 59 and 39 ends, respectively. The resulting PCR product was digested with KpnI and BamHI and ligated into pSecTag2/HygroB (Invitrogen, Carlsbad, CA). This placed the mouse endostatin cDNA downstream of the CMV promoter and murine immunoglobulin k chain signal peptide and upstream of a c-myc epitope and polyhistidine (His) tag. This plasmid was designated pEndoSTHB. DNA sequence analysis confirmed that the mouse endostatin cDNA sequence was inserted in the proper reading frame and without mutations. Cell Lines. The mouse renal carcinoma cell line RenCa (13) was provided by Dr. K. Okumura (Jutendo University, Tokyo, Japan). The human colon carcinoma cell line SW620 was obtained from the American Type Culture Collection (Manassas, VA). RenCa and SW620 cells were maintained in DMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 mg/ml streptomycin. HUVECs were obtained from Clonetics (San Diego, CA) and maintained in EGM-2-MV medium (Clonetics). 4 The abbreviations used are: bFGF, basic fibroblast growth factor; VEGF, vascular endothelial growth factor; CMV, cytomegalovirus: mAb, monoclonal antibody; MTT, thiazolyl blue; HUVEC, human umbilical vein endothelial cell.

6251


ENDOSTATIN INHIBITS LUNG AND LIVER METASTASES

Fig. 1. A, construction of plasmid pEndoSTHB encoding the mouse endostatin. cDNA for mouse endostatin (Endo) was inserted into the mammalian expression plasmid pSecTag2/HygroB, which encodes the CMV promoter (CMV), murine immunoglobulin k chain signal peptide (Igk), c-myc epitope (myc), and polyhistidine tag (His). B, Western blot analysis of secreted mouse endostatin-c-myc-His fusion protein. Conditioned medium from RenCa control cells (RD0), RenCa endostatin-transfected cells (RDE15 and RDE17), SW620 control cells (SD0C1), and SW620 endostatin-transfected cells (SDED2 and SDED9) was concentrated, followed by detection with an anti-c-myc mAb. The size of the mouse endostatin-c-myc-His fusion protein is estimated based on peptide sequence analysis to be Mr 29,000.

Generation of Stable Transfectants That Secrete Mouse Endostatin. Endostatin transfectants were generated by transfection of pEndoSTHB into RenCa and SW620 cells using LipofectAMINE and LipofectAMINE PLUS Reagent (Life Technologies, Inc., Gaithersburg, MD) following the manufacturer’s instructions. Two days later, cells were placed under hygromycin selection at 800 mg/ml. Four weeks after transfection, hygromycin-resistant colonies were expanded and tested for endostatin-c-myc-His fusion protein production and secretion by immunofluorescence and Western blot analysis as described below. Control transfectants were generated in a similar manner,

except the parent plasmid pSecTag2/HygroB (without endostatin cDNA) was substituted for pEndoSTHB. Western Blot Analysis. One million cells were plated onto 60-mm plates and incubated for 24 h. The medium was replaced with 1 ml DMEM, and cells were incubated for 24 h. One ml of conditioned medium was concentrated in a Microcon 10 microconcentrator (Amicon, Beverly, MA) to 20 ml and subjected to electrophoresis under reducing conditions on an 18% polyacrylamide gel. Proteins were transferred onto a nitrocellulose membrane (Schleicher & Schuell, Keene, NH) and incubated overnight in 5% nonfat milk in PBS at 4°C. After briefly washing in 1% nonfat milk and 0.1% Tween 20 in PBS, the membrane was incubated with anti-c-myc mouse mAb (Sigma Chemical Co., St. Louis, MO) diluted 1:100. After three 10-min washes in 1% nonfat milk and 0.1% Tween 20 in PBS, membranes were incubated in horseradish peroxidase-conjugated antimouse immunoglobulin (Amersham, Arlington Heights, IL) diluted 1:4000. After three 10-min washes in TBS containing 0.05% Tween 20, proteins were detected using the enhanced ECL kit (Amersham). Immunofluoresence. Cells were grown on eight-well glass slides (LabTek, Naperville, IL), fixed in 4% paraformaldehyde in PBS, and permeabilized with 0.2% Triton X-100 in PBS. After washing in PBS, cells were incubated with either 1.25 mg/ml anti-His mouse mAb (Invitrogen, Carlsbad, CA) for RenCa cells or anti-c-myc mouse mAb (Sigma) diluted 1:200 for SW620 cells for 1 h. Cells were washed in PBS and incubated in 9.3 mg/ml FITCconjugated antimouse immunoglobulin (Biosource, Camarillo, CA) in 10% nonfat milk in PBS for 1 h. After washing in PBS, cells were mounted in Vectashield (Vector Laboratories, Burlingame, CA) and photographed with an Axiophot immunofluorescence microscope (Carl Zeiss, Thornwood, NY) at 360. Background staining using the anti-c-myc mAb on RenCa cells was too high for accurate discrimination. In Vitro Growth Rates. Cells were plated onto 96-well plates at a density of 2000 cells/well in quadruplicate. At designated time points, the numbers of cells were quantified using a colorimetric MTT assay as described previously (14). Endothelial Cell Proliferation Assay. Stable transfectants were plated onto 150-mm plates at a density of 5–10 million cells/plate and incubated for 48 –72 h. The cells were washed with PBS, and 5 ml of DMEM were added to each plate. Cells were incubated for an additional 24 h, and 5 ml of conditioned serum-free medium were collected and concentrated to 1 ml using a Centriplus

Fig. 2. Immunofluorescence of endostatin-c-myc-His fusion protein in endostatin-transfected cells. Anti-His mAb was applied to RD0 control transfectants (a) and RDE15 endostatin transfectants (b), followed by a FITC-conjugated secondary antibody. Anti-c-myc mAb was applied to SD0C1 control transfectants (c) and SDED9 endostatin transfectants (d), followed by a FITC-conjugated secondary antibody.

6252



Fig. 3. In vitro growth rates of stable transfectants. A, RenCa parental cells (RenCa), control cells (RD0), and endostatin-transfected cells (RDE15 and RDE17) were plated onto 96-well plates, and at specified time points, a colorimetric MTT assay was used to determine cell number, as measured by absorbance (OD). B, in vitro growth rates for the SW620 parental cells (SW620), control cells (SD0C1), and endostatin-transfected cells (SDED2 and SDED9) were determined in a similar fashion.

10 concentrator (Amicon). HUVEC cells were plated at a density of 12,500 cells/well in 24-well collagen-coated plates for 24 h, after which 250 ml of concentrated, conditioned medium were added to each well for 30 min. An equal volume of Medium 199 supplemented with 10% fetal bovine serum and 5 ng/ml bFGF (Sigma) was then added for 72 h. The numbers of cells were then quantified using a colorimetric MTT assay. Tests were performed in quadruplicate. Animal Studies. Animal studies were performed in accordance with the policies of the Massachusetts General Hospital Subcommittee on Research Animal Care. Flank tumors were generated by s.c. injection of 5 3 106 cells in 100 ml of HBSS into the right flank of immune-competent BALB/c mice for RenCa cell lines or athymic BALB/c (nu/nu) mice (Charles River Laboratories, Wilmington, MA) for SW620 cell lines (n 5 5/group). Tumor width (W) and length (L) were measured every 3– 4 days with calipers. The tumor volume (TV) was determined by the following formula: TV 5 (L 3 W2)/2. Experimental lung metastases were generated by injection of 5 3 105 cells in a single-cell suspension in 100 ml of HBSS without calcium or magnesium into the tail veins of immune-competent BALB/c mice (n 5 5/group). Mice were sacrificed after 4 weeks and examined for lung metastases. Experimental liver metastases were generated by injection of 5 3 106 cells in a single-cell suspension in 100 ml of HBSS without calcium or magnesium into the spleens of nude mice (n 5 5/group). Mice were sacrificed after 12 weeks to examine for liver metastases. Organs were fixed in 10% buffered formalin prior to being weighed and photographed. To generate flank tumors from lung metastases, lung metastases from recently sacrificed mice were resected from the lung and soaked in PBS. The metastases were implanted into the s.c. flanks of BALB/c mice using a trocar. A single lung metastasis of ;2–3 mm in diameter was implanted into each mouse (n 5 6/group). Tumor volume was determined every 4 days.

RESULTS Generation of Stable Transfectants That Secrete Mouse Endostatin. Mouse endostatin cDNA was inserted into a mammalian expression plasmid (pSecTag2/HygroB), and the resulting plasmid

was designated pEndoSTHB (Fig. 1A). pEndoSTHB or pSecTag2/ HygroB was transfected into RenCa and SW620 cells. Two endostatin-transfected RenCa clones (RDE15 and RDE17) that secreted the expected Mr 29,000 protein and one control clone (RD0) were selected (Fig. 1B). Similarly for the SW620 cells, two endostatin-transfected clones (SDED2 and SDED9) and one control clone (SD0C1) were selected. Stable transfectants were examined for endostatin expression by immunofluorescence. Endostatin-secreting RenCa clones (Fig. 2b) and SW620 clones (Fig. 2d) demonstrated cytoplasmic staining consistent with the presence of the endostatin-c-myc-His fusion protein prior to secretion. The control RenCa clone (Fig. 2a) and control SW620 clone (Fig. 2c) showed only minimal background staining. In Vitro Growth Rates of Stable Transfectants. Cells were seeded onto 96-well plates, and the number of cells was quantified at specific time points by MTT assay. RenCa stable transfectants and parental RenCa cells grew at similar rates in vitro (Fig. 3A). In addition, SW620 stable transfectants and parental SW620 cells displayed similar in vitro growth rates (Fig. 3B). Thus, endostatin secreted into the culture medium by endostatin-expressing clones had no inhibitory effect on the proliferation of these nonendothelial cells. Mouse Endostatin Inhibits Endothelial Cell Proliferation. The biological activity of secreted mouse endostatin-c-myc-His fusion protein was tested against HUVEC cells. Conditioned medium from RD0 control cells did not inhibit HUVEC cell growth as compared with DMEM alone (Fig. 4). Conditioned medium from RDE15 and RDE17 endostatin-transfected cells inhibited HUVEC proliferation by 51 and 40%, respectively, compared with conditioned medium from RD0 control cells. Similar results were obtained for conditioned medium from SW620 stable transfectants, with conditioned medium from SDED2 and SDED9 endostatintransfected cells inhibiting HUVEC proliferation by 39 and 36%, respectively, compared with conditioned medium from SD0C1 control cells. Mouse Endostatin Inhibits Formation of s.c. Tumors. To determine whether endostatin secreted from RenCa cells inhibited s.c. tumor formation, 5 3 106 cells from each RenCa clone were injected s.c. into the flanks of immune-competent BALB/c mice. Similarly, 5 3 106 cells from each SW620 clone were injected s.c. into the flanks of nude BALB/c mice. For RenCa clones, RD0 control cells formed tumors rapidly, but RDE15 and RDE17 endostatin-transfected

Fig. 4. Inhibition of endothelial cell proliferation by conditioned-medium from endostatin-transfected cells. Conditioned medium from RenCa control cells (RD0), RenCa endostatin-transfected cells (RDE15 and RDE17), SW620 control cells (SD0C1), and SW620 endostatin-transfected cells (SDED2 and SDED9) was concentrated and applied to HUVEC cells grown in 96-well plates. Three days later, cell number, as measured by absorbance (OD), was then quantified using a colorimetric MTT assay. HUVEC cells grown with DMEM added instead of conditioned medium were similarly quantified. Bars, SD. p, P , 0.05, compared with RD0; pp, P , 0.05 compared with SD0C1.

6253



Fig. 5. Inhibition of flank tumor formation by endostatin. A, RenCa control cells (RD0) or endostatin-transfected cells (RDE15 and RDE17) were inoculated s.c. into the flanks of BALB/c mice, and tumor volume was determined every 3– 4 days. B, similarly, SW620 control cells (SD0C1) or endostatin-transfected cells (SDED2 and SDED9) were inoculated into the flanks of nude mice. C, SD0C1 and SDED9 cells were mixed in the indicated ratios and inoculated into the flanks of nude mice. Bars, SD. p, P , 0.05, RD0 compared with RDE15 or RDE17; pp, P , 0.05, SD0C1 compared with SDED2 or SDED9.

cells formed s.c. tumors that were 73 and 84% smaller, respectively, compared with RD0 tumors after 21 days (Fig. 5A). SD0C1 control cells also formed tumors rapidly, whereas SDED2 and SDED9 endostatin-transfected cells formed tumors that were 91 and 82% smaller, respectively, after 21 days (Fig. 5B). Mice bearing RD0 and SD0C1 tumors were euthanized on day 21 because of tumor burden. Mice bearing RDE15, RDE17, SDED2, and SDED9 tumors were followed after 21 days, and their tumors eventually grew at an appreciable rate (data not shown). Seven to 8 weeks after initial tumor inoculation, mice with tumors from endostatin-transfected cells were euthanized because of tumor burden. Cell mixing experiments were performed to determine whether endostatin secretion by a subpopulation of cells could inhibit the growth of the remaining population of cells. SD0C1 control cells and SDED9 endostatin-transfected cells were mixed in various ratios before s.c. implantation into nude mice. Tumor growth was equally inhibited, independent of whether 100, 75, or 25% of the cells expressed endostatin (Fig. 5C). In a separate experiment, each clone was injected s.c. into mice to generate tumors for immunohistochemical analysis. During the first 3 weeks after inoculation, flank tumors from endostatin-transfected cells were too small to harvest and section; therefore, tumors from endostatin-

transfected cells were harvested between 3 and 5 weeks after inoculation, when they had grown beyond 2–3 mm in diameter. No immunohistochemical staining for endostatin-c-myc-His fusion protein was observed in any of the sections, suggesting down-regulation or loss of the endostatin transgene by the time these tumors had grown several mm in diameter (data not shown). Sections were also stained for tumor blood vessels using an anti-CD31 mAb. Blood vessel densities were determined and were not significantly different between tumors from endostatin-transfected clones and control clones (data not shown). Mouse Endostatin Inhibits Formation of Lung Metastases. To investigate the effect of endostatin on tumor formation in the lung, we injected 5 3 105 cells from each RenCa clone into the tail veins of BALB/c mice. The parental RenCa cells in this well-established model produce abundant lung metastases after 4 weeks (15). After sacrificing mice at 4 weeks, lungs from mice injected with RDE15 and RDE17 endostatin-transfected cells had few or no metastases, whereas lungs in mice injected with control cells had metastases that were too numerous to count (Fig. 6A; Table 1). Lungs from mice injected with RD0 cells also weighed significantly more than lungs from mice injected with RDE15 and RDE17 cells. This experiment was repeated in 15 more mice with similar results (data not shown). We found previously that delayed flank tumor growth after inocu-

Fig. 6. Inhibition of lung metastases formation by endostatin. A, RD0 control cells (top row), RDE15 endostatin-transfected cells (middle row), or RDE17 endostatin-transfected cells (bottom row) were injected into the tail veins of BALB/c mice. Mice were sacrificed 1 month later, and their lungs were fixed in 10% formalin. B, pulmonary metastases from mice previously injected into the tail vein with RenCa stable transfectants were resected, washed in PBS, and implanted into the flanks of BALB/c mice. Tumor volume was determined every 4 days. Bars, SD. p, P . 0.05, SD0C1 compared with SDED2 or SDED9.

6254



lation of endostatin-transfected cells was associated with loss of endostatin expression as assessed by immunohistochemistry. To test whether pulmonary metastasis formation by endostatin-transfected cells was associated with loss of endostatin biological activity, lung metastases generated from endostatin-transfected cells (RDE15 and RDE17) and control cells (RD0) were harvested and implanted s.c. into BALB/c mice. Although we had demonstrated previously that RDE15 and RDE17 endostatin-transfected cells grew poorly in the flank compared with RD0 control cells after s.c. inoculation of a cell suspension (Fig. 5A), we observed different results after s.c. implantation of harvested pulmonary metastases. There was no significant difference in the size of flank tumors grown from RD0 pulmonary metastases compared with those grown from either RDE15 or RDE17 pulmonary metastases (Fig. 6B). These data are consistent with downregulation or loss of expression as an explanation for eventual tumor formation by endostatin-transfected cells. Mouse Endostatin Prevents Formation of Liver Metastases. We next examined the effect of endostatin expression on formation of experimental liver metastases. Others have demonstrated previously that SW620 cells injected into the spleens of nude mice generate experimental liver metastases in 60% of mice after 3 months (16). We injected 5 3 106 cells from each SW620 stable transfectant into the spleens of nude mice. After 12 weeks, the mice were sacrificed, and their livers were harvested. Three of the five livers from mice injected with SD0C1 control cells developed liver metastases, whereas none of the 10 livers from mice injected with SDED2 or SDED9 endostatin-transfected cells developed liver metastases (Fig. 7A; Table 2). This experiment was repeated in 15 more mice with similar results (data not shown). DISCUSSION We have demonstrated in two different cancers, mouse renal carcinoma and human colon carcinoma, that secretion of mouse endostaTable 1 Formation of lung metastases from RenCa stable clonesa RenCa clone

Mean number of metastases (range)

Mean weight (g) 6 SD

RD0 RDE15 RDE17

.100 11–33 (18.4) 0–15 (3.0)

1.38 6 0.36 0.67 6 0.06b 0.73 6 0.13b

a A single-cell suspension (5 3 105 cells in 100 ml of HBSS) was injected into the tail veins of BALB/c mice (n 5 5/group). Mice were sacrificed after 1 month. This is one representative experiment of two. b P , 0.01, in comparison with RD0.

Table 2 Formation of liver metastases from SW620 stable clonesa SW620 clone

No. of livers with metastases/ Total number of livers

Mean weight (g) 6 SD

SD0C1 SDED2 SDED9

3/5 0/5 0/5

3.07 6 1.15 2.14 6 0.29b 2.36 6 0.26c

a A single-cell suspension (5 3 106 cells in 100 ml of HBSS) was injected into the spleens of athymic BALB/c mice (n 5 5/group). Mice were sacrificed after 3 months. b P 5 0.12, in comparison with SD0C1. c P 5 0.22, in comparison with SD0C1.

tin from these cancer cells inhibits tumor formation in three different organ environments, s.c. tissue, lung, and liver. Other studies have shown that endostatin can inhibit lung metastases (11, 12), but to our knowledge, this is the first study to demonstrate that endostatin is effective in the liver environment. These results demonstrate the ability of endostatin to inhibit tumor or metastasis formation in different organ environments and do not examine the efficacy of endostatin against established tumors. A recent study by Bergers et al. (17) found that in a transgenic mouse model of spontaneous b-islet cell tumors, endostatin was highly effective in preventing progression of hyperplastic nodules to small tumors but was ineffective in inhibiting the growth of established invasive carcinomas. Achieving regression of established tumors is a more difficult undertaking than inhibiting tumor formation. Endostatin has been demonstrated to induce regression of tumors grown s.c. on the flanks of mice (8, 10, 11), but actual regression, as opposed to growth inhibition, has not been demonstrated in the lung or liver environments. Gene transfer of endostatin and other angiogenesis inhibitors has been performed by other groups. Nguyen et al. (9) reported the use of an adeno-associated virus to transduce tumor cells with the endostatin gene in vitro and demonstrated that the secreted endostatin inhibits endothelial cell proliferation. In another study, plasmid DNA containing the endostatin cDNA was injected i.m. into mice and resulted in the inhibition of s.c. tumors and spontaneous pulmonary metastases (11). In addition, s.c. inoculation of fibrosarcoma cells stably transfected with the gene for another angiogenesis inhibitor, angiostatin, resulted in inhibition of flank tumor growth and spontaneous pulmonary metastases (18). For the purposes of this study, stable transfection of the endostatin gene into cancer cells, followed by inoculation of these cancer cells into mice, avoided some potential problems associated with systemic administration of endostatin. Production of large amounts of biologically active

Fig. 7. Inhibition of liver metastases formation by endostatin. SD0C1 control cells (top row), SDED2 endostatin-transfected cells (middle row), or SDED9 endostatin-transfected cells (bottom row) were injected into the portal system of nude mice. Mice were sacrificed after 3 months, and their livers were fixed in 10% formalin.

6255



recombinant endostatin to treat cohorts of mice can be difficult. In addition, the pharmacokinetics of endostatin absorption, distribution, and elimination after systemic administration is unknown and may result in different tissue levels in various organs, particularly in the liver, which is the primary site of metabolism of many drugs (19). Although host organ environment may have influenced the expression of the endostatin gene from our endostatin-transfected clones, expression was enough at all sites tested to inhibit tumor formation. Results of immunohistochemical staining studies of flank tumors indicated that endostatin expression was no longer detectable once tumors had grown beyond a small size. Endostatin expression may have been down-regulated or lost after tumor cell implantation into mice in the absence of hygromycin selection. Alternatively, proangiogenesis factors such as bFGF and VEGF may have eventually been up-regulated in these tumors, overcoming the effects of endostatin. After flank inoculation of a cell mixture consisting of only 25% of endostatin-transfected cells, flank tumor growth was inhibited as much as when the inoculum consisted of 100% endostatin-transfected cells. However, we also observed that lung metastases from endostatin-transfected cells, after implantation into the flank, formed flank tumors as rapidly as lung metastases from control cells. These results may be consistent because an endostatin-expressing cell line likely represents a heterogeneous population in which some cells express little or no endostatin. When this heterogeneous population is implanted s.c., endostatin secreted by the majority of cells appears to prevent tumor formation by any immediately adjacent cells that do not express endostatin. In contrast, when inoculated i.v. as a single-cell suspension, individual cells that express little or no endostatin may successfully establish pulmonary metastases. This scenario presumes that the endostatin-producing cells that lay dormant in the lung do not produce levels of endostatin high enough to prevent metastasis formation elsewhere in the lung. The observed reduction in tumor growth associated with endostatin secretion was most likely attributable to the biological activity of the fusion protein and not secondary to other causes. Clonal selection for a less tumorigenic phenotype is hypothetically possible but improbable for several reasons: (a) we used two different carcinoma cell lines and selected two endostatin-expressing clones from each cell line. We also selected hygromycin-resistant control clones that had been transfected with the same expression vector without the endostatin cDNA insert. The statistical likelihood that only the four clones of the six that we selected are incapable of in vivo growth for reasons unrelated to endostatin expression is exceedingly low; (b) we demonstrated that the expressed endostatin was biologically active against endothelial cells in vitro; (c) the control clones were clearly tumorigenic, yet implantation of a mixture of cells in which only 25% of cells expressed endostatin inhibited growth of the control cells. Host-immune response to the endostatin-c-myc-His fusion protein is also an unlikely cause of the observed results because significant antitumor effects were observed in both immune-competent hosts and congenitally athymic hosts. Ideally, we would have liked to perform experiments using endostatin-neutralizing antibodies. However, the absence of an endostatin antibody capable of neutralizing all of its biological activity precluded such an experiment. The anti-c-myc and anti-His mAbs reacted with the recombinant fusion protein but did not attenuate the ability of endostatin to inhibit HUVEC proliferation in vitro (data not shown). Accordingly, these antibodies could not be used for neutralization experiments. Many issues regarding endostatin and other angiogenesis inhibitors remain unanswered. The molecular mechanisms by which endostatin exerts its effects are not well understood. Recently, Knebelmann et al. (20) reported that endostatin inhibits activation of mitogen-activated protein kinase, which is a downstream target in the bFGF and VEGF

signaling pathways. Examination of the effect of endostatin on these pathways in different environments may shed more light on organspecific differences in endostatin function. One drawback to gene transfer of endostatin into cancer cells prior to inoculation into mice is that this strategy did not allow us to study the efficacy of endostatin against established tumors in different organ environments. We are currently examining the efficacy of endostatin against established tumors in various organ environments using baculoviral and herpes simplex viral vectors to deliver the endostatin gene. ACKNOWLEDGMENTS We are grateful to Dr. Judah Folkman for supplying the mouse endostatin expression plasmid. We thank Dr. Sandra Ryeom for technical advice and Drs. Rakesh Jain and Dai Fukumura for helpful discussions.

REFERENCES 1. Paget, S. The distribution of secondary growths in cancer in the breast. Lancet, 1: 571–573, 1889. 2. Fidler, I. J. Modulation of the organ microenvironment for treatment of cancer metastasis. J. Natl. Cancer Inst., 87: 1588 –1592, 1995. 3. Hanahan, D., and Folkman, J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell, 86: 353–364, 1996. 4. Singh, R. K., Bucana, C. D., Gutman, M. R., Fan, D., Wilson, M. R., and Fidler, I. J. Organ site-dependent expression of basic fibroblast growth factor in human renal cell carcinoma cells. Am. J. Pathol., 145: 365–376, 1994. 5. Fukumura, D., Yuan, F., Monsky, W. L., Chen, Y., and Jain, R. K. Effect of host microenvironment on the microcirculation of human colon adenocarcinoma. Am. J. Pathol., 151: 679 – 688, 1997. 6. Terayama, N., Terada, T., and Nakanuma, Y. A. morphometric and immunohistochemical study on angiogenesis of human metastatic carcinomas of the liver. Hepatology, 24: 816 – 819, 1996. 7. Junqueira, L. C., Carneiro, J., and Kelley, R. O. Basic Histology, Ed. 6, p. 514. Norwalk: Appleton & Lange, 1989. 8. O’Reilly, M. S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W. S., Flynn, E., Birkhead, J. R., Olsen, B. R., and Folkman, J. Endostatin: an endogenous inhibitor of angiogenesis and tumor growth. Cell, 88: 277–285, 1997. 9. Nguyen, J. T., Wu, P., Clouse, M. E., Hlatky, L., and Terwilliger, E. F. Adenoassociated virus-mediated delivery of antiangiogenic factors as an antitumor strategy. Cancer Res., 58: 5673–5677, 1998. 10. Dhanabal, M., Ramchandran, R., Volk, R., Stillman, I. E., Lombardo, M., IruelaArispe, M. L., Simons, M., and Sukhatme, V. P. Endostatin: yeast production, mutants, and antitumor effect in renal cell carcinoma. Cancer Res., 59: 189 –197, 1999. 11. Blezinger, P., Wang, J., Gondo, M., Quezada, A., Mehrens, D., French, M., Singhal, A., Sullivan, S., Rolland, A., Ralston, R., and Min, W. Systemic inhibition of tumor growth and tumor metastases by intramuscular administration of the endostatin gene. Nat. Biotech., 17: 343–348, 1999. 12. Sim, B. K. L., Fogler, W. E., Zhou, X. H., Liang, H., Madsen, J. W., O’Reilly, M. S., Panigrahy, D., and Fortier, A. H. Potent inhibition of experimental metastases and primary tumors by recombinant human endostatin that is suitable for human use. Proc. Am. Assoc. Cancer Res., 40: 620 – 621, 1999. 13. Murphy, G. P., and Hrushesky, W. J. A murine renal cell carcinoma. J. Natl. Cancer Inst., 50: 1013–1025, 1973. 14. Yoon, S. S., Carroll, N. M., Chiocca, E. A., and Tanabe, K. K. Cancer gene therapy using a replication-competent herpes simplex virus type 1 vector. Ann. Surg., 228: 366 –374, 1998. 15. Chakrabarty, A., Hillman, G. G., Maughan, R. L., Visscher, D. W., Ali, E., Pontes, J. E., and Haas, G. P. Influence of tumor site on the therapy of murine kidney cancer. Anticancer Res., 14: 373–378, 1994. 16. Zirvi, K. A., Keogh, J. P., Slomiany, A., and Slomiany, B. L. Transglutaminase activity in human colorectal carcinomas of differing metastatic potential. Cancer Lett., 60: 85–92, 1991. 17. Bergers, G., Javaherian, K., Lo, K. M., Folkman, J., and Hanahan, D. Effects of angiogenesis inhibitors on multistage carcinogenesis in mice. Science (Washington DC), 284: 808 – 812, 1999. 18. Cao, Y., O’Reilly, M. S., Marshall, B., Flynn, E., Ji, R. W., and Folkman, J. Expression of angiostatin cDNA in a murine fibrosarcoma suppresses primary tumor growth and produces long-term dormancy of metastases. J. Clin. Investig., 101: 1055–1063, 1998. 19. Benet, L. Z., Mitchell, J. R., and Sheiner, L. B. Pharmacokinetics: the dynamics of drug absorption, distribution, and elimination. In: A. G. Gilman, T. W. Rall, A. S. Nies, and P. Taylor (eds.), Goodman & Gilman’s The Pharmacological Basis of Therapeutics, Ed. 8, pp. 3–32. New York: Pergamon Press, 1990. 20. Knebelmann, B., Dhanabal, M., Ramchandran, R., Waterman, M., Lu, H., and Sukhatme, V. P. Endostatin inhibits VEGF and bFGF induced MAPK activation in endothelial cells. Proc. Am. Assoc. Cancer Res., 40: 414, 1999.

6256
