The presence of HS (heparan sulphate) proteoglycans on the cell surface and in the extracellular environment is critical to many physiological processes including the growth of new blood vessels from pre-existing vasculature (angiogenesis). A plethora of growth factors and their receptors, extracellular matrix molecules and enzymes bind to specific sites on the HS sugar chain. For example, HS proteoglycans have profound effects on the bioactivity of the key angiogenic factor VEGF (vascular endothelial growth factor) (VEGF165), affecting its diffusion, half-life and interaction with its tyrosine kinase receptors. A number of HS structural features that mediate the specific binding of VEGF165, including sulphation requirements, have been determined. In parallel, zebrafish embryos were used as a vertebrate model system to study the role in vascular development of the biosynthetic enzymes that create these specific binding sites on HS. It was discovered that knockdown of one of the HS 6-O-sulphotransferases in zebrafish with morpholino antisense oligonucleotides reduced vascular branching and corresponded to changes in the HS structure. The roles of the extracellular 6-O-sulphatase enzymes, the sulfs, in vascular development are now being investigated. Both oligosaccharides and small molecule biosynthetic enzyme inhibitors could be valuable HS-based strategies for controlling aberrant angiogenesis in diseases as diverse as cancer and heart disease.
Angiogenesis is the tightly controlled fundamental process by which new blood vessels arise from pre-existing vasculature [1,2]. It is of immense clinical interest as promoting angiogenesis could bypass blocked coronary arteries or aid the complications of diabetes, whereas inhibition of angiogenesis starves tumours of their essential blood supply.
Angiogenesis is a multi-step process. Localized breakdown of the basement membrane and extracellular matrix precedes the proliferation and migration of capillary endothelial cells into the surrounding tissue, and formation of new vessels. Angiogenesis contributes to development of the cardiovascular system in vertebrate embryos and is essential in adults for wound healing and the female reproductive cycle. The switch from normal quiescent vasculature to angiogenesis is induced by a change in the balance of many pro- and antiangiogenic factors that are released predominantly by surrounding pericytes and lymphocytes. A large number of these factors have been found to bind to HS (heparan sulphate) proteoglycans or the structurally related heparin, including key angiogenic growth factors such as the FGFs (fibroblast growth factors) and VEGFs (vascular endothelial growth factors) . However, the only growth factor observed almost ubiquitously at sites of angiogenesis and whose levels correlate most closely with the spatial and temporal events of blood vessel growth is VEGF A .
Roles of VEGF in angiogenesis
VEGF is a potent stimulator of angiogenesis both in vitro and in vivo (for a review, see [4,5]). Inactivation of even a single VEGF allele results in embryonic lethality due to abnormal blood vessel development. VEGF is an almost ubiquitous initiator of tumour angiogenesis. Its expression is potentiated by hypoxia, by activated oncogenes, and by a variety of cytokines. VEGF is able to regulate most, if not all, steps in the angiogenic cascade; it regulates protein expression, migration, division and apoptosis of endothelial cells. VEGF interacts with two type III receptor tyrosine kinases, VEGFR-1 (VEGF receptor-1; flt-1) and VEGFR-2 [flk-1/KDR (kinase insert domain-containing receptor)], and with the isoform-specific receptors, the neuropilins. In contrast with all other known angiogenic factors, the proliferative action of VEGF seems to be restricted almost entirely to endothelial cells, making it an excellent agent or target for angiogenesis-based therapies.
Interaction of VEGF with HS proteoglycans
HS proteoglycans are ubiquitous components of the cell surface (syndecans and glypicans) and extracellular matrix (e.g. Perlecan) . They consist of a protein core with between one and four polysaccharide chains, which can be transmembrane, GPI (glycosylphosphatidylinositol)-anchored or embedded in the matrix. It is mainly through the HS polysaccharide chains that they can bind to a variety of protein ligands, including growth factors, growth factor receptors and serine proteases. They have the ability to mediate angiogenesis driven by a number of growth factors, including all but one splice variant of VEGF [7,8]. A number of in vivo and in vitro studies have examined the roles of HS proteoglycans in VEGF activity: most importantly, HS seems to be essential to spatially restrict VEGF to create the appropriate gradient to allow blood vessel branching to occur . HS proteoglycans can also re-activate oxidation-damaged forms of VEGF, a function that may be essential in hypoxic sites where new vessels are required . They potentiate VEGF binding to its two specific signal-transducing receptors  and may interact directly with the receptors [11–13] analogous to the widely studied HS–FGF-2–FGF receptor signalling complex [14,15].
Analysis of the binding site on HS for VEGF
The HS polysaccharide chain is initially synthesized as an alternating polymer of GlcA (D-glucuronate) and GlcNAc (N-acetyl-D-glucosamine). Partial N-sulphation creates three types of domains within the chain: blocks of N-sulphated disaccharides [S-domains (N-sulphated domains)] are flanked by alternating regions of N-acetylated and N-sulphated disaccharides and these sulphated areas are separated by N-acetylated stretches. Further structural variability is generated by O-sulphation of some of the disaccharides within the S-domains and alternating regions. The fine structure and spacing of these S-domains appears to be the major determinant of specific binding to protein ligands (for reviews of HS, see [6,16]).
The binding site on HS for the common, most potent VEGF splice variant, VEGF165, has recently been analysed . The types of domains required for VEGF165 binding were assessed using structurally distinct fragments created by the enzymes heparanase I, heparinase III, platelet heparanase and the phage-derived enzyme K5 lyase. Unlike members of the FGF family , VEGF165 did not bind strongly to single S-domains isolated using heparinase III. In fact the only fragments with strong binding were those created by K5 lyase. This enzyme cleaves within the N-acetylated stretches of HS, leaving the S-domains and alternating regions intact. The minimum size of K5 lyase fragment with good VEGF binding was 22 saccharides. Modelling studies indicated that a single heparin-binding site on VEGF165 could be filled with a heptasaccharide. Therefore the 22+ oligosaccharides could probably stretch across both of the heparin-binding domains within the VEGF dimer and may consist of two S-domains separated by an alternating region (Figure 1).
Hypothetical model of a K5 lyase excised oligosaccharide binding to the VEGF165 dimer
The sulphate dependency of HS binding to VEGF165 was also investigated . Binding of specifically desulphated HS to VEGF165 suggested that N-, 2-O- and 6-O-sulphated disaccharides all contributed to the interaction. However, experiments where octasaccharide S-domains were bound to a single VEGF site at sub-physiological salt concentrations demonstrated that 6-O-sulphation was essential. This raised the question as to whether one or more of the enzymes that 6-O-sulphate HS may be critical for VEGF-mediated angiogenesis.
Role of HS biosynthetic enzymes in angiogenesis
Eleven families of enzymes are required to create the complex structure of HS (for a review, see ). Subsequent to chain initiation on serine residues in the proteoglycan protein core, the GlcNAc-GlcA bipolymer is synthesized. Partial N-sulphation occurs by the N-deacetylase/N-sulphotransferase enzyme family followed by conversion of some of the GlcA into IdoA (L-iduronate) within the S-domains. A single enzyme adds 2-O-sulphate groups to some of the IdoAs and GlcAs, whereas multigene families are responsible for 6-O-sulphation and 3-O-sulphation of the glucosamines. The incomplete action of these enzymes on the HS chain results in its structural diversity.
In recent years, reports of mutations in the HS biosynthetic enzymes in animal model systems have given important insights into the roles of HS in vivo. Mutations that completely prevented formation of HS affected many developmental processes. For example, mutation in the gene required for production of the UDP-glucuronic acid precursor to HS was found to prevent initiation of heart valve formation in zebrafish , and was critical for patterning mediated by Wingless (Wnt family member) and Decapentaplegic (related to the bone morphogenetic proteins) in Drosophila [20–22]. Strikingly, it was found that fine modification of HS-controlled specific physiological processes, for example 2-O-sulphation, is important for kidney development , and 6-O-sulphation of HS is essential for formation of the tracheal system in Drosophila , which in many ways is analogous to the vasculature in vertebrates. These findings suggested the exciting possibility that fine-tuning of HS structure, by influencing the activity of specific biosynthetic enzymes, could be used to control individual physiological processes.
The role of HS6STs (HS 6-O-sulphotransferases) in angiogenesis was investigated in zebrafish embryos . Zebrafish are an excellent model system in which to study angiogenesis due to their rapid development, ease of genetic manipulation and comparability of vascular development to mammals [26,27]. Morpholino antisense oligonucleotides were used to individually block the translation of two of the four HS6STs found in zebrafish and the decreased level of HS 6-O-sulphation was confirmed by a sensitive HPLC assay. Reduction in expression of HS6ST-2 but not HS6ST-1 decreased vascular branching in the caudal vein plexus of the zebrafish embryos, a paradigm for remodelling stages of angiogenesis. The activity of the HS6ST-2 was thus shown to be important for VEGF-mediated angiogenesis.
Role of sulfs in angiogenesis
In the last couple of years, a novel family of enzymes called sulfs, which remove 6-O-sulphation from HS, were discovered by Charles Emerson's group . These are thought to fine-tune the structure of HS at the cell surface and were initially found to promote Wnt activity in quail . Subsequent work using the chick chorioallantoic membrane assay demonstrated that sulf1 activity could block FGF2-mediated angiogenesis . There are three sulf enzymes in zebrafish. Their expression patterns and activities, especially with respect to angiogenesis, are being investigated (S.E. Stringer, unpublished work).
The studies discussed here have demonstrated that HS 6-O-sulphation is important both for VEGF binding and VEGF-mediated angiogenesis in zebrafish. Inhibition of HS6STs or the sulf enzymes could be a valuable way of preventing or promoting angiogenesis respectively. Promising results are already emerging from groups screening for small molecules that affect HS enzyme activity, while others are investigating the use of specific oligosaccharides to control HS activity. Such compounds may have clinical potential in diseases as diverse as heart disease and cancer.
Cytokine–Proteoglycan Interactions: Biology and Structure: Biochemical Society Focused Meeting held at Royal Holloway University of London, Egham Hill, U.K., 9–10 January 2006. Organized and edited by B. Mulloy (NIBSC, U.K.) and C. Rider (Royal Holloway University of London, U.K.).