The capacity of various angiogenic factors to interact with heparin/HS raise the possibility that molecules able to interfere with this interaction may act as angiogenesis inhibitors. Two classes of compounds can be envisaged to have the capacity to exert angiostatic activity via this mechanism of action (Fig. 6):

1) heparin-binding, polycationic compounds able to compete with heparin-binding growth factors for HSPG interaction;

2) heparin-like, polyanionic compounds able to compete with HSPGs for heparin-binding growth factor interaction. In both cases, the binding of angiogenic factors to endothelial cell surface would be hampered with consequent inhibition of their angiogenic capacity.
 
 

Fig. 6: Mechanisms of inhibition of the HSPG-dependent interaction between growth factors and TK receptors. HSPGs facilitate GF/TK receptor interaction (1). This can be prevented by polycationic compounds that bind to and mask HSPGs (2) or by polyanionic compounds that bind the GF hampering its interaction with HSPGs (3).

Here it follows a brief description of natural and synthetic angiostatic compounds whose putative mechanism of action is based on their capacity to affect growth factor-heparin/HS interaction.

4.1.1. Protamine.

Protamine is a small sperm-derived, DNA-binding cationic protein (Mr = 4,300). It inhibits neovascularization in embryogenesis, inflammation, immune reaction, and tumor growth, even though its efficacy is limited by toxicity at high doses (99). In vitro, protamine inhibits the mitogenic activity of FGF-2 (100) and hampers the capacity of heparin to prevent trypsin digestion of the growth factor (95). This suggests that protamine may compete with heparin-binding growth factors for interaction with sulfated GAGs.

4.1.2 Chemokines.

Chemokines belong to a superfamily of low molecular weight chemotactic proteins that are active on different leukocyte populations. They are divided into three structural subclasses depending on whether the first two of the four invariant cysteine residues are adjacent (C-C chemokines), separated by an intervening residue (C-X-C chemokines), or whether the first cysteine is missing (C chemokines) (101-103). Moreover, C-X-C chemokines are subdivided further depending on the presence or the absence of the sequence Glu-Leu-Arg (the ELR motif) (101).

Chemokine receptors are members of the G protein-coupled 7 transmembrane family and they bind either C-C or C-X-C chemokines (104). An exception to this ligand specificity is the promiscuous Duffy antigen/erythrocyte chemokine receptor (DARC) which can bind members of both the C-X-C and C-C class (105).

Besides a high affinity receptor interaction, chemokines bind sulfated glycosaminoglycans including heparin and HS (106) and thus can interact with HSPGs of the cell membrane and extracellular matrix. HSPGs on endothelial cells have been shown to present some chemokines to leukocytes in the multistep process of recruitment (107).

Recent work indicates that various chemokines may affect endothelial cell function. In particular, platelet factor-4 (PF-4) and other C-X-C chemokines lacking the ELR motif (ELR- C-X-C chemokines) exert an angiostatic activity in vivo and inhibit FGF-2-induced proliferation and migration in cultured endothelial cells (108, 109). Recombinant human PF-4 has been tested in clinical trials that include Kaposi's sarcoma, colon and kidney carcinomas, and melanoma (110). In particular, intralesional injection of PF-4 is effective in the treatment of Kaposi's sarcoma (111).

The role of HSPG interaction in mediating the anti-angiogenic activity of chemokines has not been clearly established and some contradictory findings have been reported. For instance, IP-10 and PF-4, that share a specific HSPG binding site in SV-40-transformed murine endothelial cells, cause inhibition of FGF-2-induced proliferation in human umbilical vein endothelial cells which is abrogated by soluble heparin (112). Similarly, inhibition of FGF-2 activity by PF-4 is overcome by exogenous heparin in 3T3 fibroblasts (113).

Nevertheless, a PF-4 mutant called rPF-4-241, in which the heparin-binding site has been mutated such that the molecule no longer binds to heparin, retains a full anti-angiogenic activity (114).

Also, the ELR+ C-X-C chemokines GROa/MGSA and GROb, but not GROg, have been shown to inhibit the mitogenic activity of FGF-2 in bovine capillary endothelial cells but their antagonist action is not reversed by heparin (115). In the same study GROb reduced FGF-2-stimulated mouse corneal neovascularization and tumor growth. In contrast, a second study (108) has shown that GROa/MGSA, GROb, and GROg can stimulate chemotaxis in bovine capillary endothelial cell and neovascularization of the rat cornea (limited to GROa/MGSA). Others have failed to observe any activity of GROa/MGSA on endothelial cells (107).

4.1.3. Tecogalan sodium.

D-gluco-D-galactan sulfate (DS-4152 or tecogalan) is a sulfated polysaccharide isolated from Arthrobacter species. It exerts a potent inhibitory activity against tumor-induced angiogenesis and the growth of solid tumors (116). It also inhibits FGF-2-induced endothelial cell growth, chemotaxis, and morphogenesis by preventing the binding of the growth factor to HSPGs and high affinity TK receptors (117, 118).

In vivo, tecogalan inhibits neovascularization induced by FGF-2 in the chick embryo chorioallantoic membrane and in the rabbit cornea (119, 120) and the development of Kaposi's sarcoma-like lesions in nude mice (121). Eckhardt et al. (122) have found that the plasma levels of tecogalan sodium achieved in patients with advanced malignancies or HIV-related Kaposi's sarcoma are in the range of concentrations biologically active in experimental models and stabilization of the disease has been reported.

Alpha2-Macroglobulin.

a2-Macroglobulin is a major serum protein originally characterized as a protease inhibitor. It binds several heparin-binding proteins, including FGF-2 and VEGF, thus neutralizing their receptor binding capacity and biological activity in cultured endothelial cells (123, 124). Heparin completely prevents the binding of VEGF and TGF-b to a2-macroglobulin, but not that of FGF-2. Since a2-macroglobulin does not interact with heparin, these results suggest that both molecules compete for the binding to the heparin-binding domain of the growth factor.

4.2.1. Suramin derivatives.

Suramin is a polysulfonated naphthylurea originally developed for the treatment of trypanosomiasis and onchocerciasis. Recent studies have shown that suramin possess a variety of biological effects. The compound has been evaluated for antiviral therapy in the acquired immunodeficency syndrome (AIDS) because of its capacity to inhibit reverse transcriptase and to prevent HIV entry into the cell (125).

More recently, suramin has been used experimentally in the treatment of cancer (126). In vitro, suramin blocks the activity of several growth factors by inhibiting their binding to cognate receptors (127-129). Also, suramin is internalized into the cell (130) where it may affect the activity of various key enzymes involved in the intracellular transduction of mitogenic signals including protein kinase C, DNA and RNA polymerases, topoisomerase II, phosphoinositol and diacylglycerol kinases (130-133). As observed for different heparin derivatives, suramin inhibits the activity of heparanase (134) and of urokinase-type plasminogen activator (135), thus exerting an anti-invasive effect.

The anti-proliferative and anti-invasive action of suramin, together with its inhibitory activity on cell adhesion and migration (136), may explain, at least in part, the capacity of this molecule to inhibit tumor growth and metastasis in different experimental models (137-138). On this basis, suramin has been employed in patients unresponsive to conventional chemotherapy and anti-tumor activity has been reported in the treatment of adrenocortical carcinoma and prostate carcinoma (126).

However, a limitation on the clinical use of suramin is represented by the serious toxic side-effects consequent to the administration of the high doses of the molecule required to achieve anti-tumor activity. For instance, in vivo administration of suramin, that shows a plasma half-life equal to 30-50 days (139), dramatically increases tissue GAGs, leading to mucopolysaccharidosis-like pathologic conditions (140), and elevates the concentration of circulating HS and dermatan sulfate, thus inducing coagulopathy (126).

As an angiogenesis inhibitor, suramin has been demonstrated to inhibit the activity exerted by FGFs and VEGF on cultured endothelial cells by preventing their interaction with cell-surface HSPGs and TK receptors and to block their angiogenic activity in different animal models (see ref. 141 and references therein).

This is due, at least in part, to the capacity of suramin to bind to the heparin-binding region of the growth factor via one or more of its sulfate groups. Indeed, suramin is ineffective against angiogenesis elicited by non-heparin binding growth factors (141). Accordingly, suramin is able to mimic heparin/HS for the capacity to protect FGF-2 from trypsin digestion. Interestingly, the same capacity is observed for the related polysulfonated compound trypan blue (95).

In order to improve the therapeutic ratio of this class of compounds, various polysulfonated naphthylureas structurally related to suramin have been investigated for the capacity to inhibit the activity exerted by FGF-2 on cultured endothelial cells and in a rat sponge angiogenesis assay (128, 142). Also, a series of sulfonated distamicyn A derivatives structurally related to suramin have been developed (143, 144).

The results demonstrate that the number of sulfate groups and modifications of the backbone of the molecule significantly affect the activity of these compounds. In particular, an extended multiple ring structure with at least two aromatic groups intervening between the two terminal naphthyl rings confers to suramin derivatives a reduced toxicity without affecting, or even improving, their FGF-2 antagonist capacity.

4.2.2. Pentosan polysulfate.

Pentosan polysulfate is a polymer of xylose hydrogen sulfate (Mr = 3100) and contains two sulfate groups per carbohydrate monomer.

It binds FGFs as well as other heparin-binding growth factors (145, 146) and it has been shown to interact also with the heparin-binding site of FGFR-1 (69). It inhibits the growth of SW13 adrenocortical cells transfected with FGF-4 (147) and tumorigenicity of MCF-7 breast carcinoma cells transfected with FGF-1 or FGF-4 (148).

Even though the contribution of a possible angiostatic effect of pentosan polysulfate was not investigated in these studies, independent observations have shown that cultured endothelium is sensitive to inhibition by pentosan polysulfate. Interestingly, microvascular endothelium appears to be more sensitive to inhibition by pentosan polysulfate and suramin derivatives than large-vessel endothelium (128).

Also, some observations have suggested that pentosan polysulfate may be more effective in inhibiting FGF-4-dependent than FGF-2-dependent cell proliferation. If confirmed, these findings support previous observations about different structural requirements in FGF-4-heparin interaction when compared to FGF-2-heparin interaction (58).

Pentosan polysulfate has been found to inhibit the growth of Kaposi's sarcoma-derived spindle cells in vitro (121). As observed for suramin, pentosan polysulfate is also a potent anti-HIV agent in vitro (149). These observations suggested that pentosan polysulfate might be worth exploring as a potential agent for the treatment of Kaposi's sarcoma. A trial in patients with HIV-related Kaposi's sarcoma has shown that the maximally tolerated dose of pentosan polysulfate given by continuous venous infusion is 3 mg/kg per day. No patient had an objective clinical anti-tumor response to either systemic or intralesional pentosan polysulfate administration; however, three patients had stable Kaposi's sarcoma for 3-27 weeks. Dose-limiting toxic effects were characterized by anticoagulation and thrombocytopenia and were reversible (126, 150, 151).

4.2.3. Miscellanea.

Besides the molecules described above, various polysulfonated compounds have been described as potential angiostatic drugs. A non-comprehensive list includes: chemically sulfated malto-oligosaccharides (152); sulfated chitin derivatives (153, 154); beta-cyclodextrin tetradecasulfate in combination with angiostatic steroids (155); cortisol-linked heparin adipic hydrazide (156).

Malto-oligosaccharides and sulfated chitin derivatives exert they inhibitory activity by acting as heparin-like, polyanionic compounds. As observed for suramin derivatives and pentosan polysulfate, they inhibit cultured endothelial cell proliferation, chemotaxis, and/or morphogenesis. In vivo, they inhibit neovascularization in different assays and suppress tumor angiogenesis and growth.

In contrast, the rationale for the use of beta-cyclodextrin tetradecasulfate in combination with angiostatic steroids and of cortisol-linked heparin adipic hydrazide was based on pioneer observations on the capacity of defined batches of commercial heparin preparations to magnify the angiostatic capacity of corticosteroids devoid of glucocorticoid and mineralocorticoid activity (157, 158). The mechanism of action of these compounds is unclear, even though their primary action appears to be exerted on basement membrane integrity and synthesis (159).

Recently, the capacity of phosphorothioate oligodeoxynucleotides to interact with heparin-binding growth factors, including FGF-2, FGF-1, FGF-4, and VEGF, has been demonstrated. This interaction is reversed by heparin or suramin (160, 161). Phosphorothioate oligodeoxynucleotides are isoelectronic congeners of phosphodiester oligodeoxynucleotides that contain internucleoside linkages in which one of the nonbridging oxygen atoms has been replaced by a sulfur atom. In analogy with other polyanions, they inhibit the binding of FGF-2 to HSPGs and TK receptors, thus preventing its mitogenic activity. Interestingly, the antagonist activity of phosphorothioate oligodeoxynucleotides depends only in part on their size and sequence and is independent of P-chirality (162).

In contrast, RNA ligands with defined consensus sequences and secondary structures have been described to interact specifically with FGF-2 with nanomolar affinity (163). Also in this case, interaction with the growth factor was prevented by heparin.

At present, no data are available on the effect of phosphorothioate oligodeoxynucleotides and RNA ligands in angiogenesis-related assays in vitro and in vivo.