Suramin A POTENT INHIBITOR OF MELANOMA HEPARANASE AND INVASION*

Suramin, a polysulfonated naphthylurea, has anti- reverse transcriptase and anti-proliferative activities and inhibits the binding of various growth factors to their cell surface receptors. This drug is used in the treatment of acquired immunodeficiency syndrome and several types of cancers. Increased levels of circulating glycosaminoglycans have been observed in suramin-treated cancer patients, suggesting that it may inhibit glycosaminoglycan catabolism. Melanoma-derived heparanase, a heparan sulfate-specific endo- 0-D-glucuronidase that plays an important role in metastatic melanoma cell invasion through basement mem- branes, is inhibited by suramin in a dose-dependent manner: 100% inhibition was observed at a concentration of -100 PM. Structurally related polysulfonated compounds, such as trypan blue and Evans blue, had lower heparanase inhibitory activities: the concentra- tions required for 50% heparanase inhibition (IDso) were 310-320 PM and six times higher than for suramin (IDso = 46 PM). Oversulfated heparin tetrasaccha- ride, whose average molecular size is similar to suramin, had also much lower heparanase inhibitory activ- ity than suramin. The inhibition constants (Ki) for suramin and oversulfated

Suramin, a polysulfonated naphthylurea, has antireverse transcriptase and anti-proliferative activities and inhibits the binding of various growth factors to their cell surface receptors. This drug is used in the treatment of acquired immunodeficiency syndrome and several types of cancers. Increased levels of circulating glycosaminoglycans have been observed in suramin-treated cancer patients, suggesting that it may inhibit glycosaminoglycan catabolism. Melanomaderived heparanase, a heparan sulfate-specific endo-0-D-glucuronidase that plays an important role in metastatic melanoma cell invasion through basement membranes, is inhibited by suramin in a dose-dependent manner: 100% inhibition was observed at a concentration of -100 PM. Structurally related polysulfonated compounds, such as trypan blue and Evans blue, had lower heparanase inhibitory activities: the concentrations required for 50% heparanase inhibition (IDso) were 310-320 PM and six times higher than for suramin (IDso = 46 PM). Oversulfated heparin tetrasaccharide, whose average molecular size is similar to suramin, had also much lower heparanase inhibitory activity than suramin. The inhibition constants (Ki) for suramin and oversulfated heparin tetrasaccharide were 48 and 290 PM, respectively. Suramin had a remarkable inhibitory activity against B16 melanoma cell invasion through reconstituted basement membranes (IDs0 e10 WM). The inhibitory effects of suramin on melanoma heparanase and cell invasion appeared to be completely independent of its antiproliferative activity, because significant effects on melanoma cell growth were not observed at the concentrations of suramin used in this study. The results suggest that the antimetastatic effects of suramin may be due to its antiinvasive rather than antiproliferative activities.
Suramin ( M , 1429), a polysulfonated naphthylurea, has been widely used for the treatment of onchocerciasis and trypanosomiasis. It is a potent competitive inhibitor of reverse transcriptase (1) and blocks in vitro the infectivity and cyto-R01-CA41524 (to M. N.) and R35-CA44352 to (G. L. N.). The costs * This work was supported by National Institutes of Health Grants of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 (2). Following these observations, clinical trials were initiated to treat patients with acquired immunodeficiency syndrome (3,4). Suramin has also been tested for anticancer activity and found to be an active agent in the treatment of metastatic cancers (5).
Suramin blocks the cell surface binding of various growth factors, such as platelet-derived growth factor, epidermal growth factor, and transforming growth factor$, and prevents cell proliferation (6)(7)(8)(9). It has also been found to inhibit many enzymes including phosphoinositol and diacylglycerol kinases (9) and DNA polymerases (lo), suggesting that its anticancer effects may be mediated through its antiproliferative activities.
In animals and patients high doses of suramin dramatically increase tissue glycosaminoglycans and, through elevations in the concentration of circulating heparan sulfate (HS)' and dermatan sulfate (DS), diminishes blood coagulation (11,12). The HS and DS increases appear to be due, in part, to inhibition of iduronate sulfatase, one of the lysosomal enzymes responsible for HS and DS intracellular degradation (11).
We investigated the effect of suramin on glycosaminoglycan degrading endoglycosidases produced by malignant tumor cells and the ability of suramin to inhibit heparanase, an endo-P-D-glucuronidase that specifically degrades HS (13) and participates in the degradation of basement membranes by invasive tumor and normal cells (14). Indeed, heparanase activities in malignant cells, including melanoma (15), Tlymphoma (16), fibrosarcoma (17), rhabdomyosarcoma (18), and colon carcinoma (19), correlates with the metastatic potentials of these tumors. Furthermore, natural and synthetic inhibitors of heparanase are inhibitors of lung colonization of B16 melanoma cells in their syngenic host (20-23).
Here we demonstrate that suramin is one of the most potent inhibitors of melanoma heparanase and that suramin can inhibit melanoma cell-mediated degradation of subendothelial extracellular matrix (ECM). We also report that suramin greatly inhibits melanoma cell invasion through reconstituted basement membranes and that this effect is independent of suramin's antiproliferative activity. In some experiments B16-BL6 cells were grown in DMEM/F-12 supplemented with 5 pg/ml insulin, 5 pg/ml transferrin, and 25 nM sodium selenite (Sigma) or only with 0.1% bovine serum albumin (RI grade, proteinase-free, Sigma). Rat lung endothelial clone 8 (RLE c1.8) cells were isolated as described previously (24) and grown in DMEM/F-12 supplemented with 10% plasma-derived horse serum and 100 pg/ml endothelial mitogen (Biomedical Technologies, Inc., Cambridge, MA) and 5 pg/ml porcine intestinal heparin (Sigma). Heparanase Assay-Heparanase was assayed as described previously (19,22,25). Briefly, HS purified from bovine lung (10 mg) was labeled with 5 mCi of [3H]acetic anhydride (100 mCi/mmol, ICN Radiochemicals) in 0.4 M sodium acetate, pH 6.5, and [3H]HS with a M , -34,000 was purified using a Sephacryl S-200 column (Pharmacia LKB Biotechnology Inc.) equilibrated with 0.2 M pyridine acetate, pH 5.0. B16 melanoma heparanase was prepared as described peviously (25). Briefly, B16-BL6 cells were extracted at 4 "C in 50 mM Tris-HCI, pH 7.5, containing 1 mM phenylmethylsulfonyl fluoride, 5 mM N-ethylmaleimide, 0.5% Triton X-100, and 0.05% sodium azide. The cell extracts were subjected to heparanase purification by sequential chromatography using columns of heparin-Sepharose, concanavalin A-Sepharose (Pharmacia LKB Biotechnology Inc.), and Bio-Si1 SEC-400 (Bio-Rad). In the assay ['HIHS was mixed with the heparanase in 100 pl of 0.2 M sodium acetate, pH 5.0, and incubated a t 37 "C for 6 h. Incubation was terminated by heating at 100 "C for 5 min followed by a 5-min centrifugation at 18,000 X g. A 50-pl aliquot of the supernatant was analyzed by size exclusion chromatography using a Waters 600E system (Waters, Milford, MA) equipped with a Bio-Gel TSK-30 XL column (Bio-Rad). Elution was performed at 23 "C with 12.5 mM Tris-HC1, 150 mM sodium chloride, pH 7.5, at a flow rate of 0.5 ml/min, and the eluents were collected every 30 s and counted in a liquid scintillation counter. Heparanase activity was determined by measuring the decrease in area of the first one-half of Subendothelial Matrix Degradation-RLE c1.8 cells were grown in 1.0% gelatin-coated plastic culture plates (16-mm diameter; Costar, Cambridge, MA), and confluent RLE c1.8 cells were incubated for 48 h with 50 pCi/ml [36S]sulfate (ICN Radiochemicals, Irvine, CA). ECM was isolated by sequential treatment with 0.2% Triton X-100 and 10 mM NH,OH as described previously (22,24). The RLE-ECM was incubated in medium containing 10% fetal bovine serum for 2 h.
B16-BL6 melanoma cells were harvested from subconfluent cultures and suspended in 5% fetal bovine serum in DMEM/F-12. Cell suspensions (5 X lo5 cells/900 pl) and 100-pl aliquots of DMEM/Fisolated RLE-ECM and incubated over 90 h at 37 "C in a CO, 12 containing suramin at various concentrations were placed on incubator. Culture supernatants were withdrawn and centrifuged at 18,000 X g for 10 min, and the [35S]sulfate radioactivity released into 400 p1 supernatant was determined.
Degradation products of the RLE-ECM were analyzed by size exclusion chromatography using a Bio-Gel TSK-40 XL column. Elution was performed at 23 "C with 12.5 mM Tris-HC1, 150 mM NaC1, pH 7.5, at a flow rate of 1 ml/min, the eluents collected every 30 s, and radioactivity determined on a liquid scintillation counter. Various degradation products were tested for their susceptibility to Flavobacterium heparitinase (Seikagaku Amerika, St. Petersburg, FL) digestion.
Invasion Assay-Invasion assay was performed as described previously (23, 26), using Costar 6.5-mm Transwell'" chambers equipped with 8.0-pm pore size polycarbonate membranes. The upper surface of the membrane was coated with Matrigel'* (200 pg of protein, Collaborative Research, Bedford, MA). The bottom chamber was filled with 600 p1 of a solution containing laminin (100 pg/ml, Collaborative Research) and fibronectin (50 pg/ml, Sigma) in DMEM/F-12 supplemented with 5 pg/ml insulin, 5 pg/ml transferrin, and 25 nM sodium selenite. B16-BL6 melanoma cells grown as subconfluent cultures were harvested by a brief treatment with 2.5 mM EDTA and suspended in DMEM/F-12 plus 5 pg/ml insulin, 5 pg/ml transferrin, and 25 nM sodium selenite. Melanoma cells (5 X lo3 to 2 X lo4) were seeded on the reconstituted Matrigel'" in the upper Transwell'" chamber. After a 60-h incubation, cells that penetrated through the polycarbonate membrane were harvested from the bottom chamber by trypsin-EDTA treatment and counted.

Suramin Inhibition of Melanoma Heparame-When [3H]
HS ( 5 mg) was incubated at 37 "C with 25 mg of the partially purified melanoma heparanase (specific activity, 60 mg HS/ mg protein/h), the intact [3H]HS peak decreased with time, and degradation fragments of characteristic molecular sizes (Mr -8000 and -5400) appeared on the chromatogram (13). This degradation was almost completely inhibited by 100 p~ suramin (Fig. 2), but it was not affected by 100 pM concentration of related polysulfonated compounds, such as trypan blue and Evans blue. The concentration of suramin required for 50% inhibition (ID50) of the melanoma heparanase was 46 pM under the conditions used, and the ID50 of trypan blue and Evans blue for heparanase activity were 310-320 pM, or six times higher than that of suramin.
We have reported that heparin and its nonanticoagulant derivatives are potent competitive inhibitors of melanoma heparanase (20,21). Here we employed a heparin tetrasaccharide and its oversulfated derivative with molecular sizes similar to suramin and compared their effects on heparanase. Heparin tetrasaccharide had no significant effect on heparanase activity even at the high concentration of 500 pM, whereas the oversulfated tetrasaccharide inhibited heparanase in a dose-dependent manner (Fig. 3). The ID50 of oversulfated heparin tetrasaccharide for heparanase activity was 435 p~, almost 10 times higher than that of suramin.
The Dixon plots (27) of the data from inhibition experiments using different concentrations of the substrate demonstrated that the heparanase inhibition by suramin is primarily noncompetitive, whereas the heparanase inhibition by oversulfated heparin tetrasaccharide is a mixed type of com-
Suramin Inhibition of Endothelial Cell ECM Degradation by B16 Melanoma Cells-Highly metastatic B16 melanoma cells are capable of degrading basement membrane-like endothelial ECM in vitro and producing characteristic size fragments of HS proteoglycans (15,29). When B16-BL6 cells were seeded on [35S]sulfate-labeled RLE-ECM, [35S] radioactivity in the culture medium increased with time (Fig. 5). During the first 24 h of incubation, concentrations of suramin between 1 and 100 PM slightly increased the spontaneous release of [35S]sulfate-labeled materials from RLE-ECM. However, over a 90-h incubation period suramin effectively inhibited B16-BL6 cell-mediated degradation of [35S]sulfatelabeled ECM components in a dose-dependent manner. The ID50 of suramin for melanoma cell-mediated ECM degradation was approximately 80 pM. The released [35S]sulfatelabeled materials were analyzed by size exclusion chromatography using a Bio-Gel TSK-40 XL column. Large molecular size HS proteoglycans (Mr >200,000) were found in the first peak, and HS degradation fragments of M, -10,000 and -5,000 were identified in the second and third peaks (Fig. 6) by their susceptibility to Flavobacterium heparitinase digestion. Suramin at a concentration of 100 PM markedly reduced the production of small HS degradation fragments (Fig. 6). Antiproliferative Activity of Suramin on Melanoma Cells-In serum-free culture medium supplemented with insulin and transferrin, the growth of B16-BL6 melanoma cells was not significantly affected by suramin of up to 100 p~ during 72 h of culture. Only a slight reduction (13%) in cell growth was observed in the presence of 100 p~ suramin (Fig. 7). Other compounds used in this study such as trypan blue, Evans blue, heparin tetrasaccharide, and oversulfated heparin tetrasaccharide a t concentrations as high as 100 p~ had no significant effect on B16-BL6 cell growth over a 72-h incubation period.
Suramin Inhibition of Melanoma Cell Invasion-We showed previously that some of the natural and chemically modified heparanase inhibitors inhibit B16 melanoma cell experimental pulmonary metastasis and invasion through reconstituted basement membranes (20-23). When B16-BL6 cells were

Inhibition of B16 melanoma cell invasion by suramin
The invasion assay was performed as described under "Experimental Procedures" using Costar 6.5-mm Transwell'" chambers equipped with 8.0-pm pore size polycarbonate membranes. Cells (1 X 10' or 5 X lo3) were seeded on the reconstituted Matrigel'" in the upper Transwell'" chamber. After a 60-h incubation, the B16 cells that penetrated through the Matrigel'" and polycarbonate membrane was harvested from the bottom chamber by trypsin-EDTA treatment and counted. seeded onto Matrigel'", the cells migrated into this matrix but did not penetrate through the matrix layer. Fibronectin and laminin (50 and 100 pg/ml, respectively) were added to the bottom chamber to enhance melanoma cell invasion. The number of cells penetrating through the Matrigel'" and underlying polycarbonate membrane was dependent on the total number of cells seeded as well as the quality of the Matrigel'" and conditions used. In general, approximately 0.5% of the total cells penetrated through both the Matrigel'" and underlying polycarbonate membrane during a 60-h incubation ( Table I). This invasion was effectively inhibited by suramin: the IDso of suramin for melanoma cell invasion in four independent experiments varied from 0.8 to 9.9 pM. The results from two such experiments are shown in Table I. DISCUSSION Suramin has been employed in the treatment of malignant diseases, primarily because it inhibits tumor cell growth (3)(4)(5)(6)(7)(8)(9). It has been reported that suramin is useful for the treat-

Melanoma Heparanase
Inhibition by Suramin 9665 ment of some metastatic cancers, such as adrenal ( 5 ) and prostate carcinomas (30) and lymphomas (31), that are unresponsive to conventional cytotoxic chemotherapy. Here we examined the effect of suramin on metastatic murine melanoma cell heparanase, an HS-specific endo-fi-D-glucuronidase, whose activity correlates with metastatic potential of melanoma cells and is involved in their invasive degradation of subendothelial ECM. The effect of suramin was compared with those of related polysulfonated compounds as well as oversulfated and unmodified heparin tetrasaccharides of similar size. Suramin inhibited heparanase activity in a dosedependent manner, and its IDs0 was significantly lower than that of the other compounds tested.
Jentsch et al. (32) reported that molecular size and steric hindrance seemed to be more important for the reverse transcriptase inhibitory activity of the suramin-related compounds and that the inhibitory activity of these compounds on reverse transcriptase did not correlate with its effect against filariae or trypanosomes. In our study the effect of suramin on melanoma heparanase appeared to be specific, and both the molecular size and the number of sulfonic acid groups seemed to be important for the heparanase inhibitory activity. However, these factors do not account for the difference in heparanase inhibitory activity between suramin (Mr 1429) and oversulfated heparin tetrasaccharide (Mr 1530).
Analysis of the kinetic data using the Dixon plot (27) and the Cornish-Bowden plot (28) indicated that heparanase inhibition by suramin is primarily noncompetitive. This interpretation is supported by the report (11) that both lysosomal iduronate sulfatase and /3-glucuronidase are noncompetitively inhibited by suramin. In contrast, heparanase inhibition by heparin is a mixed type of competitive and uncompetitive inhibition. Nonetheless, we conclude that suramin is one of the most potent inhibitors of heparanase.
Suramin inhibits a large number of enzymes, including urease, hexokinase, acid phosphatase, serine proteases in the complement system, DNA polymerases, and kinases involved in phosphoinositide metabolism (9,10,11,(33)(34)(35). The enzymes involved in the catabolism of glycans, such as ganglioside neuraminidase, P-glucuronidase, hyaluronidase, and iduronate sulfatase, have also been found to be inhibited by suramin, perhaps accounting for the aberrant accumulation of sialogangliosides and glycosaminoglycans, including HS, DS, and hyaluronic acid, in various organs of suramin-treated patients and animals (11,12,36). Since normal and tumor cell heparanases are the major endoglycosidases that initiate the sequential cleavage of H S linked to HS-proteoglycan core peptides (13,37,38), heparanase inhibition is much more critical to the inhibition of total HS degradation than the inhibition of exoglycosidases and sulfatases. Thus, the inhibition of heparanase in the liver and blood vasculature is thought to be a major cause of increased levels of plasma H S in suramin-treated patients. In fact, the suramin levels in treated patients' plasma are generally two to three times higher than the concentration required for 100% inhibition of heparanase activity ( 5 ) .
The antiinvasive effects of suramin appeared to be completely independent of its antiproliferative activity, because no significant effect on cell growth was observed at the concentrations used in this study. Zabrenetzky et al. (39) have recently reported that suramin inhibits human melanoma cell adhesion and chemotaxis to laminin and thrombospondin without affecting cell growth. Inhibition of B16 melanoma cell invasion through reconstituted basement membranes by suramin is probably due to inhibition of both ECM degradation and cell migration. Suramin did not alter the production of heparanase in B16-BL6 cells (data not shown). It is therefore plausible that the direct inhibition of heparanase activity could be a major cause of ECM degradation inhibition. Suramin may also modify other tumor-secreted enzyme activities. It is noteworthy that one of the heparanase inhibitors, 6-0sulfated carboxymethyl chitin, effectively reduced type IV collagenolysis by B16-BL6 cells (23). We are currently studying the effect of suramin on other enzymes and endogenous inhibitors secreted by melanoma cells.
In conclusion, this study has documented that suramin is a potent inhibitor of melanoma cell heparanase and invasion. The IDs0 for these activities are very low compared to the suramin levels detected in plasma of patients during the treatment of trypanosomiasis, onchocerciasis, acquired immunodeficiency syndrome, or malignant diseases ( 5 ) . These results support further investigation of the possible usefulness of suramin as a therapeutic agent for metastatic disease.