Heparan sulfate proteoglycans (HSPGs) abundantly exist in the ECM and on the cell surface of most cells [1, 2]. They are composed of a restricted set of core proteins to which one or more glycosaminoglycan chains (GAG) are covalently attached. It is established that HSPGs are involved in a wide range of biological functions through their ability to bind to and modify the activities of a diverse repertoire of ligands including growth factors, morphogens, cytokines, chemokines, proteases, lipases, matrix proteins, and cell adhesion molecules [3–6]. Heparan sulfate (HS) chains and heparins (chemical analogues of these chains) consist of repeating disaccharide units of glucuronic/iduronic acid and glucosamine that are subject to a complex set of modifications involving deacetylation, epimerization, and sulfation. Four different sites of sulfation are found in heparin/HS: the N-, 3-O, and 6-O positions of glucosamine and the 2-O position of the iduronic acid residue. Heparin is highly-sulfated throughout the polymer chain whereas in HS these modifications are concentrated mainly in the S-domains, which consist of contiguous clusters of N-sulfated disaccharide units variably sulfated at the other positions [7–9]. Interspersed with the S domains are regions with low (transition zones) or zero sulfation; however, the transition zones contain a considerable fraction of the 6-O-sulfates in HS . The ligand binding activities of HSPG/heparin depend on patterns of sulfation along the chains .
HS chains are dynamically regulated in development and during tumor progression [11, 12]. Since these changes are central to the ligand binding properties of HSPGs, there is considerable interest in mechanisms that generate diversity of the chains. One such mechanism is through regulated expression of enzymes involved in the biosynthesis of heparan sulfate, for example, the sulfotransferases that modify HS chains . Another potential mechanism is through the action of extracellular endosulfatases that remove specific sulfation modifications from intact GAG chains. The first enzyme identified in this category was QSulf-1, which was discovered in quail embryo . Stimulated by this work, we cloned cDNAs encoding Sulf-1 and a new member of the family, Sulf-2, in mouse and human . We showed that both Sulf-1 and Sulf-2 are secreted into conditioned medium when they are expressed in Chinese hamster ovary (CHO) cells. Both possess endoglucosamine-6-sulfatase activity against intact heparin with an optimum at neutral pH. The enzymes liberate sulfate groups from the C-6 position of glucosamine residues on trisulfated -IdoA(2-OSO3)-GlcNSO3(6-OSO3)- disaccharide units of heparin. A similar activity for QSulf-1 has now been confirmed on HS chains [16, 17]. This trisulfated disaccharide structure occurs within the S-domains and is known to be a key element in many of the protein ligand interactions of heparin and HS (see below).
The transcripts corresponding to QSulf-1 and its rat orthologue (RFP-Sulf-1) demonstrate complex spatiotemporal regulation during embryonic development [14, 18] and QSulf-1 plays an essential role in Wnt-dependent differentiation of somites into muscle in quail . In addition, in vitro assays have demonstrated that QSulf-1 promotes both Wnt  and bone morphogenetic protein  signaling via its sulfatase activity. Less is known about the distribution and function of the Sulfs in adult tissues. However, some interesting correlations have been revealed in tumors. Lai and colleagues have reported downregulation of Sulf-1 transcripts in human ovarian cancer and a subset of hepatocellular carcinomas [19–21]. These workers and others have stressed the potential role of the enzyme in down-modulating certain signaling pathways involved in cell proliferation, since over-expression of Sulf-1 reduces signaling by FGF-2, HB-EGF, or HGF [19–22]. In striking contrast to the results in ovarian cancer, increased levels of SULF1 or SULF2 transcripts are observed in other human cancers including breast and pancreatic carcinomas [23–25]. Upregulation of Sulf-2 at both the transcript and protein levels has been established in two mouse models of mammary carcinoma . Furthermore, cultured human breast carcinoma cells release enzymatically active Sulf-2 into conditioned medium. The upregulation of Sulf-2 raises the possibility that it may be involved in promoting tumor progression [24, 25]. Indeed, we have shown that Sulf-2 has pro-angiogenic activity, a relevant function for contributing to tumorigenesis .
Even before the discovery of the Sulfs, the glucosamine-6-sulfate modification of HS was the focus of considerable interest. Thus, it has been shown that suppressing the expression of heparan sulfate 6-O-sulfotransferases via gene silencing strategies has pronounced developmental consequences in both Drosophila and Zebrafish development [26, 27]. Secondly, binding studies with heparin/HS fragments and chemically modified heparins point to an essential contribution of the glucosamine-6-sulfate modification to their interaction with various protein ligands, including FGF-1 , FGF-10 , PDGF , VEFG , hepatocyte growth factor , lipoprotein lipase , herpes simplex glycoprotein C , noggin , and L- and P-selectin . Here, we have taken advantage of recombinant Sulf-2 generated in 293 cells and of native Sulf-2 obtained from conditioned medium of a breast carcinoma cell line to explore the range of activities for this enzyme. We have established an ELISA to examine the effects of Sulf-2 from these two sources on the interaction of several growth factors and chemokines with immobilized heparin/HS. We found that pre-treatment of heparin with Sulf-2 significantly reduced its reactivity with certain ligands but not others ("pre-binding effect") and the enzyme also dissociated heparin-ligand complexes ("post-binding effect") raising the possibility that it may be involved in mobilizing HS-bound ligands in the ECM and basement membranes.