Heparan sulfate biosynthesis: 6-O-Sulfate groups surprisingly do not affect the action of glucosaminyl N-deacetylase-N-sulfotransferase

Abstract

Several biologically important growth factor-heparan sulfate (HS) interactions are regulated by HS sulfation patterns. However, the biogenesis of these combinatorial sulfation patterns is largely unknown. N-deacetylase/N-sulfotrasferase (NDST) converts N-acetyl-D-glucosamine residues to N-sulfo-D-glucosamine residues. This enzyme is suggested to be a gateway enzyme because N-sulfation dictates the final HS sulfation pattern. It is known that O-sulfation blocks C5-epimerase, which acts immediately after NDST action. However, it is still unknown whether O-sulfation inhibits NDST action in a similar manner. In this article we radically change conventional assumptions regarding HS biosynthesis by providing in vitro evidence that N-sulfation is not necessarily just a gateway modification during HS biosynthesis.

Introduction

Heparan sulfate proteoglycans (HSPGs) are composed of core proteins attached to sulfated glycosaminoglycan (GAG) side chains. They play important roles as co-receptors for several molecules including fibroblast growth factors (FGF), hepatocyte growth factor (HGF), glial cell derived neutrotrophic factor (GDNF), and vascular endothelial growth factors (VEGF).[1] By acting as co-receptors, they are involved in a variety of pathological and physiological processes such as tumor progression and embryonic development.[2-4] The role of HS as a co-receptor depends greatly on its sulfation pattern. For example, HS-FGF2 interactions require the presence of N-sulfated glucosamine units, 2-O sulfated iduronic acid units, and 6-O sulfated glucosamine units.[5,6]

To generate the diverse sulfation patterns present in HS chains, the biosynthesis of HS involves a complex interplay between several enzymes that are predicted to be co-localized within GAGOSOMEs.[7-12] HS biosynthesis begins with the formation of a linkage region composed of one xylose residue, two galactose residues and one glucuronic acid residue. Next, the HS chain is extended by Ext-1 and Ext-2. As the polymerization takes place, N-deacetylase/N-sulfotrasferase enzyme isoforms (NDST) convert N-acetyl-D-glucosamine residues in the nascent chain into N-sulfo-D-glucosamine residues. There are four known NDST isoforms that have different deacetylation/sulfation abilities and are expressed variably in different tissues.[13-17] The N-sulfate groups added by these enzymes determine the domain structure of HS because subsequent modifications by C5-epimerase, 2-O sulfotransferase, 3-O sulfotransferase, and 6-O sulfotransferase are dictated by the location of N-sulfate groups.[18] The order of modifications will have enormous implications in generating diverse structures of biological significance.[19] It is well known that C5-epimerase, which converts glucuronic acid residues to iduronic acid residues after NDST modification, is inhibited by O-sulfation.[20] However, it is unknown whether NDST, which confers the first modification of HS chains, is influenced similarly by the presence of nearby O-sulfate groups. Earlier studies have shown that NDST can act on a variety of substrates.[21-28] However, these studies have not elucidated the site of action of NDST within the larger heparanome (which contains non-sulfated (NA), heavily sulfated (NS) and mixed (NA/NS) domains). Earlier studies estimated the levels of radioactive [³⁵S]-sulfate incorporation into a variety of substrates in the presence of NDST without analyzing the corresponding disaccharides. Therefore, it is unknown whether O-sulfation inhibits NDST activity, i.e. NDST could just be modifying N-acetyl glucosamine (GlcNAc) residues in the non-sulfated NA domains within the heparanome.

This study utilizes NDST-2, a robust enzyme isoform which shows equal N-deacetylase and N-sulfotransferase activity, to determine the site of action of NDST on HS for the first time and to probe whether O-sulfation affects N-deactylation/N-sulfation. These in vitro results are obtained from studying the action of NDST-2 on a mature HS chain and on a synthetic enzymatically-generated substrate.

Materials and Methods

Materials

Heparan sulfate from bovine kidney was procured from Seikagaku. DEAE-Sepharose gel was purchased from Amersham Biosciences. [³⁵S]Na₂SO₄ and Ultima-FloAP scintillation fluid were purchased from Perkin Elmer Life and Analytical Sciences. [³²S]PAPS and HS disaccharide standards were purchased from Iduron and Sigma-Aldrich. SF-900 II SFM media was purchased from Invitrogen. [³⁵S]PAPS was prepared as reported earlier.[19] Heparitinase I, II, and Heparitinase III from flavobacterium heparinum were cloned and expressed as previously described. All other reagents and solvents were from Sigma-Aldrich.

Purification of NDST-2

15 ml of NDST-2 viral stock was added to 2 × 10⁹ Sf9 cells in 1 L of SF900 II SFM media. Infected cells were shaken at a rate of 150 rpm in a humidified incubator maintained at 28 °C for 4 days. After 4 days, the cell suspension centrifuged at 1000 × g for 30 min to pellet cells. To the supernatant, PIPES solution was added to a final concentration of 10 mM and the pH of the solution was adjusted to 7.0. Next, a solution containing phenylmethylsulfonyl fluoride in 10 ml isopropanol was added to a final concentration of 1 mM. The supernatant was then chilled on ice for 1 hr prior to centrifugation at 4000 × g for 1 hr. The supernatant was then diluted 1:2 with ddH₂O and filtered through a 0.45 μm pore-size PES membrane under vacuum. This solution was then loaded onto a 100-ml column of ToyoPearl AF heparin 650 M. The column was washed with 600 ml of PCG-50 (10 mM PIPES, pH 7.0, 2% glycerol, 0.6% CHAPS, 50 mM NaCl) and eluted with a 450-ml linear gradient of 50 – 1000 mM NaCl in PCG. Aliquots of selected fractions were then analyzed to check NDST-activity as described later in the methods section utilizing a heparan sulfate susbtrate instead of K5NS. Positive fractions were then pooled and concentrated using an Amicon YM-10 membrane.

Preparation of partially N-sulfated 6-O-sulfated K5 heparosan

All reactions were performed in a buffer consisting of 25 mM MES (pH 7.0), 0.02 % Triton X-100, 2.5 mM MgCl₂, 2.5 mM MnCl₂, 1.25 mM CaCl₂ and 0.75 mg/ml BSA. 480 μg of K5 heparosan was combined with 200 μg of [³²S]PAPS, 40 μl of NDST-2 in a 600 μl reaction volume. The product was analyzed after digestion with Heparitinases I, II, and III, using strong anion exchange HPLC (SAX-HPLC) with an inline UV detector measuring absorbance at 232 nm. 10 μg of partially N-sulfated K5 heparosan was incubated with 10 μl 6-OST-3 and 3 μl of [³⁵S]PAPS (0.5 × 10⁷ CPM) in a 50 μl reaction. An aliquot of the radioactive sample was then desalted through a 3000 MWCO filter, digested with Heparitinases I, II, and III, and analyzed using SAX-HPLC with an inline radiometry detector. Radioactive disaccharides were eluted with a linear gradient of 0 to 1 M NaCl in 70 mM phosphate buffer (pH 3.0) for 60 min and co-injected HS disaccharide standards were used to confirm their identity.

Enzymatic reactions of NDST-2

Heparan sulfate or K5NAc/NS6S were treated with NDST-2 by incubating 10 μg of substrate in 10 μl of NDST-2 with 3 μl of [³⁵S]PAPS (0.5 × 10⁷ CPM) in a 50 μl reaction overnight. The next day, an additional 10 μg of [³²S]PAPS was added along with 10 μl of NDST-2 to saturate all possible modification sites. The modified products were then desalted, digested, and analyzed.

Results and Discussion

To demonstrate that N-sulfation is not just a gateway modification and to elucidate the site of action of NDST enzymes on a mature HS chain, bovine kidney HS was incubated with NDST-2 in the presence of [³⁵S]PAPS. The resulting polymer chain was treated with heparitinase I, II, and III, and the resulting radio-labeled disaccharides were co-injected with disaccharide standards and analyzed by strong anion exchange chromatography (Figure 1a). A chromatogram of the HS from bovine kidney is included in the supplementary materials (Figure S1).

Figure 1.

Radiometric traces of disaccharides formed by digesting polysaccharides with Heparitinase I, II, and III. A) Disaccharides formed when HS is treated with NDST-2 and digested with Heparitinase I, II, and III. B) Disaccharides of partially N-sulfated, 6-O sulfated Heparosan, a synthetic substrate designed to test the activity of NDST-2 without neighboring group effects from other O-sulfates. The ratio of GlcNS6S/GlcNAc6S peaks is ~0.6. C) Disaccharides of NDST-2 treated partially N-sulfated, 6-O sulfated Heparosan. After NDST-2 addition, the GlcNS6S/GlcNAc6S ratio increases. Thus 6-O sulfation does not inhibit NDST-2 action. The peaks in the chromatograms were determined by comparison to co-injected standards.

After NDST-2 treatment of HS, the disaccharides generated by heparitinase treatment are: ΔUA-GlcN[³⁵S] and ΔUA-GlcN[³⁵S]6S Thus, only ΔUA-GlcNAc and ΔUA-GlcNAc6S residues on mature HS chains can be N-sulfated. Surprisingly, the HS chains did not have any additional modification sites. One would expect that small amounts of (GlcA/IdoA)2S-GlcNAc sequences would also be sulfated by NDST, however the expected disaccharide product, ΔUA2S-GlcN[³⁵S], was not found after digestion of the polysaccharides with Heparitinase I, II, and III. This observation points to the possibility that other O-sulfates may inhibit NDST action. However, such a claim is difficult to test because of the rare nature of N-acetylated 2-O sulfated disaccharide units and the difficulties involved in synthesizing such substrates using chemoenzymatic methods.

While NDST was found to act on 6-O sulfated residues in HS, it is difficult to analyze the true sites of modification because nearby sulfate groups can influence the action of the enzyme on the polymer chain. To further confirm the hypothesis that NDST modifies GlcNAc6S sites on HS, and to negate any such neighboring group effects, a synthetic substrate consisting of this residue was synthesized by 6-O-sulfation of 50% N-sulfated heparosan with 6-OST-3 in the presence of [³⁵S]PAPS (Figure 2, Polysaccharide 1). Treating this polysaccharide with NDST-2 in the presence of [³²S]PAPS then allowed us to observe the conversion of GlcNAc6[³⁵S] residues to GlcNS6[³⁵S] residues (Figure 1b and 1c). It is noteworthy that all GlcNAc6[³⁵S] residues were converted to GlcNS6[³⁵S] residues – confirming that 6-O-sulfation does not inhibit NDST enzymatic action.

Figure 2.

Modification of polysaccharide 1 by NDST. (*) Indicates that this sulfate group is [³⁵S] and the radiolabeled product, after digestion with Heparitinase I, II, and III, are resolved and visualized on the radiochromatogram shown in Figure 1.

Based on these results, it is evident that the conventional view of NDST, as a gateway enzyme, provides incomplete information. This study provides in vitro evidence showing that NDST acts on the mature HS chain and modifies it after it has been 6-O-sulfated – thus NDST-mediated modification is not necessarily just a gateway modification during HS biosynthesis. Extending the findings reported here to probe the action of NDST in vivo is beyond the scope of the current work. Nevertheless, one can conclusively state that current and previously reported in vitro experiments have undoubtedly helped demonstrate that O-sulfation differentially affects the action of two early biosynthetic enzymes, NDST and C5-Epimerase.[9,20] One can envision that in vivo evidence delineating the exact sequence of HS biosynthetic events and the exact action of biosynthetic enzymes may require complex cellular experiments involving isotope enriched molecules and sophisticated instruments.[29] In conclusion, the results of this study redefine the role of NDST during HS biosynthesis and provide new insights into the actions of GAGOSOMES.

Highlights.

Heparan Sulfate (HS) biosynthesis is a complex multi-step process

N-deacetylase-N-sulfotransferase (NDST) converts GlcNHAc to GlcNHSO₃⁻ residues

NDST is considered as a gateway enzyme because it carries out the first modification

We show that O-sulfation does not inhibit NDST and NDST can also catalyze maturation

We redefine the dynamic role of NDST in the GAGOSOME

Figure 3.

The comparative effect of O-Sulfation on the action of NDST and C5-Epimerase. After the formation of the linkage region and subsequent polymerization, NDST enzymes add on the first sulfate residues onto a nascent chain. C5-epimerase and several O-sulfotransferases then further modify the HS chain. C5-epimerase cannot act on this mature chain due to 2-O and 6-O sulfates which block its activity. However, NDST can still act on this substrate and thus N-sulfation is much more than just a gateway modification of heparan sulfate.