Introduction

Heparin and its structural analogue, heparan sulphate (HS), are member of a family of polyanionic, polydisperse, linear polysaccharides called glycosaminoglycans (GAGs), which perform a variety of crucial biological functions in a number of physiological and pathological processes and have been extensively employed as therapeutic agents (Castelli et al. 2004; Casu 2005). Despite the widespread use of heparin as an anticoagulant, several aspects of the structure and physiological function of HSGAGs remained obscure for years. The molecular level understanding of glycosaminoglycans fine structure, for characterization of functionally active domains of heparin, controlling the diverse biological processes was possible by depolymerization of HSGAGs (Loganathan et al. 1990). The development of better defined heparin in terms of more or less uniform mass has been achieved in part by the generation of low molecular weight heparin (LMWH; M W 3,000–6,000 Da) by either gel filtration (Pangrazzi et al. 1985) or by partial enzymatic or chemical depolymerization of heparin (Guo and Conrad 1988; Linhardt 1992; Linhardt et al. 1986).

A variety of enzymes, both mammalian and bacterial in origin, are known to degrade HSGAGs (Ernst et al. 1995; Hopwood 1989; Linhardt et al. 1987). These mammalian and bacterial endolytic enzymes (endoglycosidases) cleave the glycosidic linkage between heparin and/or heparan sulphate residues at susceptible points within the polysaccharide chain. Mammalian endoglycosidases (or heparanases) cleave at the reducing end of glucuronic acid by a hydrolytic mechanism (Hopwood 1989; Pikas et al. 1998) whereas heparinases, from microbial sources, distinctively differ from heparanases as they depolymerise heparin and HS by β-eliminative cleavage. Heparanase cleaves the glucuronidic linkage between a non-sulphated glucuronic acid (GlcA) and N-sulpho-glucosamine (GlcN(S)) of heparin/HS. A highly sulphated glucosamine residue either 3-O-sulphated or 6-O-sulphated is critical for the enzyme action (Pikas et al. 1998). The GlcN(2-N-sulphate) structure on the reducing side and GlcN(6-O-sulphate) structure on the nonreducing side of the cleavage site are considerably important for the substrate recognition by the enzyme (Okada et al. 2002). The additional 2-N-sulphate group on the nonreducing GlcN or 6-O-sulphate group on the reducing GlcN appears to have a promoting effect on the heparanase action (Fig. 1).

Fig. 1
figure 1

Structures of the cleavage site heparinases I, II, III and human heparanase in heparin or heparan sulphate. GlcN glucosamine, GlcA glucuronic acid, S SO3, Ac COCH3. Dashed line indicates presence of either of the two groups at the site

Full size image

Heparanases are found in a variety of normal and malignant cells and tissues, among which are cytotrophoblasts, endothelial cells (ECs), platelets, mast cells, neutrophils, macrophages, T and B lymphocytes and lymphoma, melanoma and carcinoma (Dempsey et al. 2000a, b, c; Gonzalez-Stawinski et al. 1999; Nakajima et al. 1986; Rosenberg 1989). The recent studies of cloning of a single gene by several groups (Bernard et al. 2001; Fairbanks et al. 1999; Hulett et al. 1999; Kussie et al. 1999; Toyoshima and Nakajima 1999; Vlodavsky et al. 1999), together with biochemical studies (Baker et al. 1999; Hulett et al. 2000; Levy-Adam et al. 2005), suggests that various normal and malignant mammalian cells express primarily identical or highly homologous sequences of a single 65-kDa latent heparanase enzyme (Parish et al. 2001). The human heparanase gene maps to chromosome 4 at band 4q21.3, contains 14 exons and encodes a 65-kDa polypeptide. This proenzyme undergoes proteolytic cleavage to yield active heparanase, a heterodimer of 50- and 8-kDa polypeptides (Vlodavsky et al. 1999).

Heparanase is an important modulator of the extracellular matrix and associated factors, specifically by releasing angiogenic factors and accessory fragments of HS from the tumour microenvironment to induce an angiogenic response (Hulett et al. 2000). Indeed, the heparanase enzyme is preferentially expressed in human tumours, and its overexpression in low-metastatic tumour cells is reported to facilitate tumour cell invasion and vascularization, events leading to cancer progression. However, heparanase expression in non-invasive and non-immune tissue suggests a role for heparanase in tissue morphogenesis, regeneration and repair during embryonic development and in the adult human phase. Both zymogen and active forms of heparanase have been shown to play biological functions which include osteoblastogenesis (Smith et al. 2010), nervous system development and neural cellular differentiation (Navarro et al. 2008; Takahashi et al. 2007).

Distinction of heparanases from heparinases is multifaceted which includes, but is not limited to, the working mechanism, substrate specificity, molecular properties, expression pattern, cellular activation and localization of these enzymes. Heparinases from various microbial sources and their commercial, pharmaceutical and clinical applications are the main focus of this review.

Heparin/heparan sulphate lyases: the HSGAG degrading enzymes

Heparin lyases—sources and properties

Heparinases or heparin lyases are a class of enzymes that are capable of specifically cleaving the major glycosidic linkages in heparin and heparan sulphate. Three heparin lyases have been identified in Pedobacter heparinus (formerly known as Flavobacterium heparinum), a heparin-utilising organism that also produces exoglycuronidases, sulphoesterases and sulphamidases that further act on the lyase-generated oligosaccharide products (Galliher et al. 1981, 1982; Sasisekharan et al. 1995; Yang et al. 1985). These lyases are designated as heparin lyase I (heparinase, EC 4.2.2.7), heparin lyase II (heparinase II, no EC number) and heparin lyase III (heparitinase, EC 4.2.2.8). The three purified heparin lyases differ in their capacity to cleave heparin and heparan sulphate: Heparin lyase I primarily cleaves heparin, heparin lyase III specifically cleaves heparan sulphate and heparin lyase II acts equally on both heparin and heparan sulphate (Linhardt et al. 1986, 1990).

The heparin lyases of P. heparinus are the most widely used and the best studied heparin degrading enzymes. Linker and Hovingh (1965) first reported the production of a crude lyase enzyme from P. heparinus, which had shown characteristic cleavage of heparin and heparitin sulphate to produce unsaturated oligosaccharide. Later, the same group separated this crude lyase fraction into a heparinase (heparin lyase I) and a heparitinase (heparin lyase III). Both activities were purified by 50–100-fold, but no physical characterization of these enzymes was performed. Dietrich et al. (1973), Silva and Dietrich (1974), Silva et al. (1976), Ototani and Yosizawa (1978) and Ototani et al. (1981) isolated three lyases, a heparinase (heparin lyase I) and two heparitinases, from the same microorganism (P. heparinus). The heparinase acted on heparin to produce mainly tri-sulphated disaccharides (Dietrich and Nader 1974; Dietrich et al. 1971).

Linhardt et al. (1984) reported the purification of heparinase (heparin lyase I) to a single band on sodium dodecyl sulphate–polyacrylamide gel electrophoresis (PAGE). Affinity purification of heparin lyase I on heparin-Sepharose failed, apparently due to degradation of the column matrix. Sufficient quantities of pure heparin lyase I for detailed characterization studies and amino acid analysis were first prepared by Yang et al. (1985). Heparin lyase I was used to prepare polyclonal antibodies in rabbits for affinity purification of heparin lyase I, but excessively harsh conditions required to elute the enzyme resulted in substantial loss of activity (Linhardt et al. 1985). By a combination of hydroxylapatite chromatography and negative adsorption on QAE-Sephadex at pH 8.3, Yang et al. (1987) also described a large-scale method to prepare heparin lyase I.

McLean and Williamson (1985) described the specificity of a partially purified heparinase II. Although no evidence of homogeneity or any physical properties for heparinase II were presented, the broad specificity on various polymeric substrates identified the enzyme as heparin lyase II (Linhardt et al. 1990; Moffat et al. 1991). Nader et al. (1990) purified two heparitinases, called heparitinase I and II, possibly corresponding to heparin lyases II and III and characterized their substrate specificity towards heparin and heparan sulphate, although no physical properties of these enzymes were presented. Heparitinase I degraded both N-acetylated and N-sulphated heparan sulphate. A single purification method to purify all the three heparinases was developed and used to purify these enzymes to homogeneity by Lohse and Linhardt (1992). All the three heparinases from P. heparinus are periplasmic proteins and have been isolated either by osmotic shock or sonication (Lohse and Linhardt 1992; Zimmermann et al. 1991). Their physical and kinetic characteristics were studied thoroughly, and optimal reaction conditions were established and these are summarized in Table 1.

Table 1 Comparative chart of biochemical properties of heparinases produced and purified from various microbial isolates
Full size table

Some Bacteroides spp., Bacillus sp. and Sphingobacterium sp. (Nakamura et al. 1976, 1988; Salyers et al. 1977) are also reported to produce heparinases (Table 1). Most of the heparinase producing bacteria were isolated from soil, although it is not known why a soil bacterium would produce an acidic polysaccharide degrading lyase. It is possible that these bacteria could use these enzymes to degrade the GAGs from carcasses. Heparinases were purified to apparent homogeneity from Bacteroides stercoris HJ-15, isolated from human intestine (Kim et al. 2000, 2004). A heparinase has also been purified to apparent homogeneity from an unidentified soil bacterium (Bohmer et al. 1990a, b).

All the heparinases are positively charged at neutral pH (pI, 8.5–10), which may be attributed to the basic nature of heparinases required to degrade highly polyanionic substrates. These enzymes differ from those isolated from P. heparinus in its molecular weight, pI, amino acid composition and kinetic properties (Table 1). Furthermore, antibody assays, amino acid analysis/sequencing, peptide mass fingerprinting and Southern blotting between heparinases from different bacteria have shown no cross-reactivity (Bellamy and Horikoshi 1992; Sasisekharan et al. 1993), suggesting a low degree of sequence and structural homology among heparinases from different species. However, Yoshida et al. (2002a) have reported that Bacillus circulans heparinase gene and deduced amino acid sequence have partial similarity with enzymes belonging to the family of acidic polysaccharide lyases that degrade chondroitin sulphate and hyaluronic acid. Recently, Hyun et al. (2010) have reported that recombinant heparinase III from B. stercoris HJ-15 has 70 % homology to heparinase II from P. heparinus.

All the three heparinases from P. heparinus have been sequenced, cloned and expressed in Escherichia coli (Godavarti et al. 1996; Sasisekharan et al. 1993; Shaya et al. 2004). The heparinase genes were expressed in E. coli with intact biological function (Sasisekharan et al. 1993; Su et al. 1996). Molecular analysis of the three heparinases revealed no significant homology either at the DNA or protein levels, nor were they closely linked on the P. heparinus chromosome (Su et al. 1996).

Elucidation of the catalytically critical amino acids in the heparinases active site and substrate binding domains was useful for understanding their mechanism of action and for their development as molecular tools for heparin or heparan sulphate-like glycosaminoglycan (HLGAG) analysis. The heparin-binding site in heparinase I contains two Cardin–Weintraub heparin-binding consensus sequences, a calcium coordinating motif, a cysteine and a histidine residue (Godavarti and Sasisekharan 1996). The primary sequence of heparinase II reveals three cysteine residues (cysteine-164, cysteine-189 and cysteine-348) in contrast to heparinase I, which has two cysteine residues and heparinase III which does not have cysteines. Similarly, calcium is required for the enzymatic activity of heparinases I and III, but heparinase II is not only active in the absence of calcium but the presence of calcium inhibits enzyme activity (Lohse and Linhardt 1992; Sasisekharan et al. 1996b).

Heparinases—clinical and diagnostic applications

The HLGAGs have chemical heterogeneity and structural complexity, which has limited the development of effective tools and methods for a rapid sequencing. Sequencing methodology of HLGAGs is mainly dependent on the chemical or enzymatic degradation of the polysaccharide in a sequence-specific manner (Venkataraman et al. 1999). An important enzymatic tool in this sequencing process is the heparinases, including heparinases I, II and III, produced by P. heparinus. Each of the heparinases has its own unique HLGAG sequence at which it cleaves, making these enzymes valuable tools in obtaining sequence-specific information (Dongfang et al. 2002). Heparinases I, II and III from P. heparinus cleave heparin/HS with a high degree of substrate specificity, at the α-1–4-glucosaminidic linkages within HS/heparin and produce disaccharides with a nonreducing 4,5-unsaturated uronic acid residue (Godavarti and Sasisekharan 1996). Heparin lyase I cleaves the glucosaminidic linkage in GlcN(N-sulphate) α1-4 IdceA(2-sulphate) and tolerates C-6 sulphation of the hexosamine unit (Fig. 1). Conversely, heparinase III requires primarily an unsulphated uronic acid moiety and cleaves the glucosaminidic linkage in GlcN(N-sulphate or N-acetylate) α1-4 GlcA and tolerates C-6 sulphation of the hexosamine unit (Desai et al. 1993; Sugahara et al. 1995; Yamada and Sugahara 1998). Heparinase II displays a broader substrate specificity, possessing the ability to cleave all glycosaminidic linkages independent of O- and/or N-sulphation as well as the type of the uronic acid residue of heparin/HS (Shriver et al. 1998b) (Fig. 1). Structure determination of heparin or heparin sulphate involves depolymerisation of their chains into constituent disaccharide components by a detailed degradation with a single or a cocktail of heparinases. The structural information with regard to the hexuronic acid epimer present at the nonreducing end is lost by the creation of a C=C bond between C4 and C5 (Stringer et al. 2003). Accurate sequence assignment relies on the highly pure activity of GAG degrading enzymes as well as a detailed understanding of their action towards specific linkages. The substrate specificity and action pattern of the enzymes provide important constraints to reconstruct the GAG sequence.

The preparation of structurally defined pure heparin-derived oligosaccharides is crucial in understanding heparin’s interaction with proteins and in determining the precise structural requirements essential for the binding of heparin with the proteins. Structurally defined heparin-derived oligosaccharides can be obtained from the mixtures of heparin-derived oligosaccharides produced by depolymerization of heparin. The oligosaccharides prepared from heparin are purified and analysed by fractionation on the basis of molecular weight, charge density, solubility characteristics and affinity (usually to antithrombin) (Hook et al. 1976). The oligosaccharide mixtures are first fractionated by low-pressure gel permeation chromatography into size-uniform mixtures of disaccharides, tetrasaccharides, hexasaccharides and higher oligosaccharides. Each size-fractionated mixture is then separated further by gradient PAGE (Edens et al. 1992; Rice et al. 1987), capillary electrophoresis (CE) (Linhardt et al. 1993), strong anion exchange (Hileman et al. 1997) and reversed-phase ion-pairing high-performance liquid chromatography (Guo and Conrad 1988; Linhardt et al. 1989; Thanawiroon and Linhardt 2003). These provide important data on composition and domain structure but generally yield indirect and incomplete sequence information. Chromatographic and electrophoretic approaches share several limitations. The enzymatic cleavage is characterized based purely on gel mobility or HPLC elution, which are not completely predictable. Elution of GAGs from SAX or amine columns using salt could potentially bias the analysis towards species that have relatively higher sulphation and may not be sensitive to the separation of species with low sulphation. Both methodologies rely on end-labelling of saccharides requiring the introduction of excess labelling reagent. The ladder of peaks or gel shifts caused by treatment with exolytic enzymes is not fully comprehensible, as these substrate specificities have not been thoroughly characterized.

Sensitive methodologies based on CE and mass spectrometry (MS) have been developed to detect the femto-/picomolar amounts of GAG. MS has also been applied to the analysis of oligosaccharides. Fast-atom bombardment MS, electrospray ionization MS and matrix-assisted laser desorption/ionization MS are capable of determining the molecular weight of oligosaccharides (Mallis et al. 1989). However, one- and two-dimensional nuclear magnetic resonance spectroscopy provides for the accurate determination of the structure of small oligosaccharides (Desai and Linhardt 1994; Linhardt et al. 1986; Mikhailov et al. 1996).

The heparinases have also been used to investigate in great detail the role of HSGAGs in aspects of heparin-binding proteins naming a few interaction of growth factors, including the fibroblast growth factors (FGFs) involved in cell proliferation, differentiation and angiogenesis (Bernfield et al. 1999; Tumova et al. 2000); interaction of HSGAGs as coreceptor for intracellular pathogens entry into host cells (Ascencio et al. 1993; Bugatti et al. 2007; Chen et al. 1997; Crublet et al. 2008; Urbinati et al. 2009); chemokines, a family of over 40 structurally related glycoproteins that facilitate leukocyte migration, angiogenesis and breast cancer metastasis (Lantz et al. 1991; Nelson et al. 1993); inhibition of smooth muscle cell growth for antiproliferative properties (Benitz et al. 1986; Castellot et al. 1982, 1985a, b; Jackson et al. 1991; Reilly et al. 1989); anticomplementary activity through interaction of heparin with complement proteins (Edens et al. 1993; Yu et al. 2005); cell adhesion (Edens et al. 2001) and lipid metabolism (Engelberg 1996).

The most thoroughly studied heparin-binding protein is the serine protease inhibitor antithrombin III (AT III) that interacts with thrombin and factor Xa in the blood coagulation cascade. The molecular basis for the anticoagulant function of heparin was elucidated in the early 1980s when a distinct pentasaccharide sequence within the heparin chain was identified as being crucial for binding and activating antithrombin, leading to accelerated inhibition of the coagulation cascade (Lindahl et al. 1980; Petitou et al. 2003). These studies were triggered by the finding that only a fraction of heparin molecules were capable of binding with high affinity to antithrombin and further that this fraction essentially accounted for the anticoagulant activity of the unfractionated material. Oligosaccharides obtained by selective, partial depolymerization of heparin were fractionated on immobilized antithrombin, and the smallest high-affinity molecules recovered were subjected to structural analysis, in conjunction with various modification steps (Petitou et al. 2003). This region is composed of one glucuronic acid unit, one iduronic acid unit and three glucosamine units, two of which are invariably N-sulphated, whereas the remaining one may be either N-acetylated or N-sulphated (Bourin and Lindahl 1993).

Heparinases are certainly essential agents in studying structural, biochemical, physiological and pathological roles of HLGAGs. Further, heparinases have shown potential for several diagnostic as well as therapeutic applications.

Low molecular weight heparin production

Although heparin is highly efficacious in a variety of clinical situations and has the potential to be used in many others, the side effects associated with heparin therapy are many and varied. Side effects such as heparin-induced thrombocytopenia are primarily associated with the long chain of unfractionated heparin (UFH), which provides binding domains for various proteins. This has led to the explosion in the generation and utilisation of low molecular weight heparin as an efficacious alternative to UFH. Although attention has been focussed on LMWH as heparin substitutes due to their more predictable pharmacological action, reduced side effects, sustained antithrombotic activity and better bioavailability, there is at present limited ability to standardize the LMWH manufacturing process. Because the LMWH are derived from heparins and hence are polydisperse and microheterogenous, with undefined structure, they possess inherent variability, which currently prevents an efficient process for their manufacture. It would be of value both medically and scientifically to have a consistent, quality controlled, time efficient, concentration independent and highly reproducible method for producing heparin and other glycosaminoglycans (Sundaram et al. 2003).

The development of better defined heparin in terms of more or less uniform mass has been achieved in part by the generation of LMWH (M W 3,000–6,000 Da) by either gel filtration (Pangrazzi et al. 1985) or by partial enzymatic or chemical depolymerization of heparin (Guo and Conrad 1988; Linhardt 1992; Linhardt et al. 1986). LMWHs possess distinct pharmacologic profile that is largely determined by their composition and thus on their preparative method (Fareed et al. 2000). The oxidative instability of heparin, as observed during pharmaceutical grade heparin preparation, suggested the possibility of preparation of LMWHs by oxidative methods, oxygen radical processes and oxidative depolymerization through nitrous acid deamination. Besides oxidative breakdown, two β-eliminative methods, one enzymatic and the other chemical, are used to commercially prepare LMWHs (Table 2) (Linhardt 1994; Linhardt and Toida 1997). Enzymatic breakdown of UFH for generation of LMWHs is reported by heparin lyases (or heparinases I, II, III) of P. heparinus, specifically heparinases I and II (Langer et al. 1983; Linhardt et al. 1992; Shriver et al. 2000; Viskov and Mourier 2007). The extent of heparinase reaction is monitored by measuring the change in absorbance associated with the unsaturated uronic acid residue formed in each product molecule (Linhardt 1994; Lohse and Linhardt 1992). The depolymerization is stopped by removing or inactivating the enzyme. After recovery of the GAG from the enzyme and removal of very low molecular weight by-products (i.e. disaccharides and tetrasaccharides), a LMWH is obtained that has the desired molecular weight and activity properties (Linhardt and Gunay 1999). This method is used to prepare the clinically used LMWH product, tinzaparin sodium (Table 2).

Table 2 Commercially available low molecular weight heparins; their method of preparation and commercial manufacturer
Full size table

Heparin antagonist

Patients undergoing surgery with cardiopulmonary bypass (CPB) must receive systemic anticoagulation with intense antithrombin activity to prevent activation of the coagulation system by the artificial surfaces of the CPB apparatus. Heparin has long been used to temporarily render the blood incoagulable during extracorporeal circulation, cardiovascular surgery and other arterial interventions. But bleeding complications are especially common when the arterial tree is violated, occurring in as many as 10–15 % of cases. After the patient is separated from CPB, heparin’s anticoagulant effect must be neutralized to halt substantial bleeding. For cardiovascular surgery and many related interventions, protamine has long been the standard antagonist when acute and complete neutralization of heparin anticoagulant effect is necessary. Protamine is the only currently approved drug in the USA with antiheparin activity. Protamine’s efficacy is related in part to its total net cationic charge, but unfortunately so is its toxicity, including systemic hypotension, pulmonary vasoconstriction and anaphylactic reactions. For these reasons, there is renewed interest in developing heparin antagonists which will replace the use of protamine (Shenoy et al. 1999).

Heparinase I (Neutralase™; IBEX Technologies, Montreal, QC, Canada) that specifically inactivates heparin is a possible alternative to protamine. Heparinase neutralizes heparin by enzymatic cleavage of alpha–glycosidic linkages at the AT III binding site (Ammar and Fisher 1997). Several studies have examined the efficacy of heparinase I to reverse heparin-induced anticoagulation in vitro and compared heparinase I to protamine as an antagonist of heparin-induced anticoagulation in animal models like dogs and rabbits (Michelsen et al. 1996; Silver et al. 1998). Animal investigations demonstrated that heparinase I reverses heparin-prolonged activated clotting time (ACT) without significant hemodynamic changes. When given in doses up to 30 μg/kg, heparinase I successfully neutralized heparin’s anticoagulant effect in a dose-dependent fashion without significant adverse sequelae. Heres et al. (2001) assessed the heparin-neutralizing activity and safety profile of different doses of heparinase I in 49 patients undergoing coronary artery surgery. They found that heparinase I, 7 or 10 μg/kg, effectively restored the ACT after unfractionated heparin was given to patients undergoing CPB for coronary artery surgery. Heparinase I caused no clinically significant hemodynamic or other adverse responses. In addition, because heparinase I totally eliminated the antithrombin activity of heparin but only partially eliminated the anti-Xa activity, anti-Xa activity increased as additional heparin was metabolized by heparinase I. Heparinase I displayed a half-life of 12 min in patients with coronary artery disease. Its activity decayed accordingly, so that after approximately 36 min, little activity remained. However, another study examining the efficacy and safety of heparinase I as an alternative to protamine in patients undergoing aortocoronary bypass graft surgery found that though heparinase I reversed heparin anticoagulation but is not equivalent to protamine because of its inferior safety profile (Stafford-Smith et al. 2005). Further studies are required to compare heparinase I and protamine for clinically relevant outcome variables.

Anti-angiogenic activity of heparinase

Cancer development is characterized by uncontrolled tumour proliferation and secondary metastasis. The tumour tissue can be divided into three compartments, namely the tumour cell compartment, the endothelial cell compartment and the extracellular matrix compartment. ECM compartment interfaces with both tumour and endothelial cell compartments and regulates the overall development of cellular compartments. Heparan sulphate proteoglycans (HSPGs) along with structural proteins are key components of the cell surface–ECM interface. The strategic location of HSPGs facilitates regulation of cell proliferation and migration, the key components for the tumour growth and angiogenesis (Sanderson 2001). HSPGs on endothelial cell surfaces act as co-receptors for a variety of pro-angiogenic growth factors such as FGF and vascular endothelial growth factor (VEGF) and anti-angiogenic factors such as endostatin (Bikfalvi et al. 1997). There is sufficient experimental data to suggest that heparins may interfere with various aspects of cancer proliferation, angiogenesis and metastasis formation (Collen et al. 2000; Hasan et al. 2005; Weidner et al. 1991).

Vascularisation is the hallmark of malignant tumours without which solid tissue cannot grow beyond 1–2 mm (Pluda 1997). Tumour-associated neovascularization is a central pathogenic step in the process of tumour growth, invasion and metastasis. These complex processes involve multiple steps and pathways dependent on the local balance between positive and negative regulatory factors, as well as interactions among the tumour, its vasculature and the surrounding extracellular tissue matrix. Neovascularization is the process of generating new blood vessels mediated primarily by progenitor and/or endothelial cells leading to tube formation, resulting in a stabilized neovascular channel. Angiogenesis, the predominant form of neovascularization in carcinogenesis, is mediated by endothelial cells sprouting from postcapillary venules, leading primarily to new capillaries (Risau 1997).

The genetic background of the angiogenic switch during tumour progression is not fully understood, but discoveries of the endothelial mitogenic growth factors (VEGF, FGF, platelet-derived growth factor, hepatocyte growth factor) as the main angiogenic factors suggested that the switch is able to turn on the expression of the genes of these factors in tumours. The molecular interactions between endothelial mitogenic growth factors with their signalling or accessory receptors have become a major pharmacological target for development of anti-angiogenic drugs. The recognition of the common chemical nature of these endothelial growth factors, i.e. their heparin-binding potential, provides an easy though non-specific anti-angiogenic target. Besides other anti-angiogenic factors (Timar et al. 2001), heparinases I–III have been reported to reduce neovascularization by altering the action of FGF at the level of receptor binding (Dongfang et al. 2002; Raman and Kuberan 2010; Sasisekharan et al. 1994). It was found that heparinase treatment of capillary EC caused a significant loss in FGF binding capacity in these cells. Treatment with 125 nM heparinase I resulted in greater that 95 % loss in HSPG binding sites on ECs accompanied by an over 80 % decrease in receptor binding. The half maximal concentrations required to remove HSPG sites and reduce receptor binding were 0.7 and 1.5 nM, respectively. The concentrations of heparinase I required to reduce FGF binding to heparan sulphate and receptor sites by 50 % were 0.5 and 1.5 nM, respectively. For heparinase II, IC50 concentrations were 2 and 8 nM, respectively, for the heparan sulphate and the receptor binding sites. On the other hand, heparinase III was most potent in inhibiting the FGF binding, with an IC50 of 0.15 and 0.2 nM for the heparan sulphate and the receptor binding, respectively (Sasisekharan et al. 1996a).

Concluding remarks

The role of heparin as an anticoagulant was established much earlier to the discovery of heparinases. Although it is still in its infancy, clinical and diagnostic application of heparinases holds promise for unravelling the HLGAGs microstructure and biological activity along with a range of diverse applications. Deciphering their physiologic and clinical role will establish the evolutionary relationship of these proteins from various microbial sources but also having potential to improve and develop the role of heparinases in various clinical manifestations.