Characterization of a heparan sulfate 3-O-sulfotransferase-5, an enzyme synthesizing a tetrasulfated disaccharide.

Heparan sulfate d-glucosaminyl 3-O-sulfotransferases (3-OSTs) catalyze the transfer of sulfate from 3'-phosphoadenosine 5'-phosphosulfate (PAPS) to position 3 of the glucosamine residue of heparan sulfate and heparin. A sixth member of the human 3-OST family, named 3-OST-5, was recently reported (Xia, G., Chen, J., Tiwari, V., Ju, W., Li, J.-P., Malmstrom, A., Shukla, D., and Liu, J. (2002) J. Biol. Chem. 277, 37912-37919). In the present study, we cloned putative catalytic domain of the human 3-OST-5 and expressed it in insect cells as a soluble enzyme. Recombinant 3-OST-5 only exhibited sulfotransferase activity toward heparan sulfate and heparin. When incubated heparan sulfate with [35S]PAPS, the highest incorporation of35S was observed, and digestion of the product with a mixture of heparin lyases yielded two major35S-labeled disaccharides, which were determined as DeltaHexA-GlcN(NS,3S,6S) and DeltaHexA(2S)-GlcN(NS,3S) by further digestion with 2-sulfatase and degradation with mercuric acetate. However, when used heparin as acceptor, we identified a highly sulfated disaccharide unit as a major product. This had a structure of DeltaHexA(2S)-GlcN(NS,3S,6S). Quantitative real-time PCR analysis revealed that 3-OST-5 was highly expressed in fetal brain, followed by adult brain and spinal cord, and at very low or undetectable levels in the other tissues. Finally, we detected a tetrasulfated disaccharide unit in bovine intestinal heparan sulfate. To our knowledge, this is the first report to describe not only the natural occurrence of tetrasulfated disaccharide unit but also the enzymatic formation of this novel structure.

S was observed, and digestion of the product with a mixture of heparin lyases yielded two major 35 S-labeled disaccharides, which were determined as ⌬HexA-GlcN(NS,3S,6S) and ⌬HexA(2S)-GlcN(NS,3S) by further digestion with 2-sulfatase and degradation with mercuric acetate. However, when used heparin as acceptor, we identified a highly sulfated disaccharide unit as a major product. This had a structure of ⌬HexA(2S)-GlcN(NS,3S,6S). Quantitative real-time PCR analysis revealed that 3-OST-5 was highly expressed in fetal brain, followed by adult brain and spinal cord, and at very low or undetectable levels in the other tissues. Finally, we detected a tetrasulfated disaccharide unit in bovine intestinal heparan sulfate. To our knowledge, this is the first report to describe not only the natural occurrence of tetrasulfated disaccharide unit but also the enzymatic formation of this novel structure.
Heparan sulfate proteoglycans (HSPGs) 1 are ubiquitously present on the cell surface and in the extracellular matrix and have divergent structures and functions (1)(2)(3). Many biological functions of HSPGs are mediated by interactions between the heparan sulfate chain and a variety of proteins, including protease inhibitors, heparin-binding growth factors, extracellular matrix components, protease, and lipoprotein lipase (4 -8). Moreover, some pathogens exploit heparan sulfate of the host cell surface, which binds to coat proteins or cell surface proteins of pathogens, at the invasion (9,10). Most of the interactions between heparan sulfate and various functional proteins occur in certain regions of the heparan sulfate chain with specific sulfated monosaccharide sequences such as binding sites for antithrombin III (AT III) (11), acidic fibroblast growth factor (12), basic fibroblast growth factor (13)(14)(15)(16), and hepatocyte growth factor (17,18). The ability of cells to produce heparan sulfate with such sequences depends on the specific mechanisms of heparan sulfate biosynthesis (19,20). Heparan sulfate is initially synthesized as a polymer of a disaccharide repeat sequence, -glucuronic acid-␤1,4-N-acetylglucosamine-␣1,4-. This polymer is then N-deacetylated/N-sulfated and subsequently undergoes epimerization of glucuronic acid (GlcA) to iduronic acid (IdoA), 2-O-sulfation of uronic acid, and 6-Osulfation of glucosamine residues. Additionally, a rare but functionally important modification, 3-O-sulfation of the glucosamine residue, also occurs (21). Most of the enzymes involved in the biosynthesis of heparan sulfate have been purified and cloned (22,23). Some of them have been shown to be present as isoforms. These isoforms play important roles in generating the specific and diverse structures of heparan sulfate. Four isoforms of N-deacetylase/N-sulfotransferase have been reported. * This work was supported by the R&D Project of the Industrial Science and Technology Frontier Program (R&D for Establishment and Utilization of a Technical Infrastructure for Japanese Industry) supported by the New Energy and Industrial Technology development Organization (NEDO). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
In this study, we cloned human 3-OST-5, and focused on the characterization of reaction products. Our results indicate novel substrate specificities and tissue distribution of 3-OST-5. Finally, we demonstrated the novel tetrasulfated disaccharide structure in natural heparan sulfate.
Construction and Purification of 3-OST-5 Protein Fused with FLAG Peptide-The putative catalytic domain of 3-OST-5 (amino acids 63-346) was cloned and expressed as a secreted protein fused with a FLAG peptide in insect cells according to the instruction manual of GATE-WAY TM Cloning Technology (Invitrogen, Groningen, Netherlands). Because the catalytic domain is encoded by a single exon (35), an ϳ1-kb DNA fragment containing the entire exon II was amplified by PCR using the human genomic DNA (Clontech), as a template, and two primers, 5Ј-ACTGGGGAACCAGAAAAATGAAAAG-3Ј and 5Ј-GTGT-CTCCAGGCACAACACATAGTG-3Ј. The PCR product was used as a template in a nested PCR to amplify the DNA fragment containing the putative catalytic domain. The forward primer was 5Ј-GGGGACAAG-TTTGTACAAAAAAGCAGGCTTCTTTAAGCGTGGCCTGCTGCACGA-G-3Ј and the reverse primer was 5Ј-GGGGACCACTTTGTACAAGAA-AGCTGGGTTTAGGGCCAGTTCAATGTCCTCCC-3Ј. The amplified fragment was cloned into a pDONR TM 201 vector (Invitrogen) and subsequently cloned into a vector pFBIF linearized by NcoI to yield pFBIF containing the catalytic domain of 3-OST-5 (pFBIF-3OST). pF-BIF is an expression vector derived from pFastBac1 (Invitrogen) and contains a fragment encoding a signal peptide of human immunoglobulin (MHFQVQIFSFLLISASVIMSRG), the FLAG peptide (DYKD-DDDK) and the conversion site for the GATEWAY TM system. DH10 BAC competent cells (Invitrogen) were transformed with the pFBIF-3OST and then bacmid DNA was prepared from the cells. Recombinant virus was prepared according to the instruction manual of BAC-TO-BAC Baculovirus Expression Systems (Invitrogen). Sf21 insect cells (BD Pharmingen, San Diego, CA) were infected with the recombinant virus and incubated at 27°C until the survival rate was less than 50% to yield secreted recombinant 3-OST-5 proteins fused with the FLAG peptide. The secreted enzyme was purified using anti-FLAG M1 monoclonal antibody agarose affinity gel (Sigma). The culture media and affinity gel mixed overnight at 4°C were centrifuged for 5 min, and the supernatant was aspirated. The affinity gel was washed twice with 50 mM TBS(50 mM Tris-HCl, pH 7.4, and 150 mM NaCl) containing 1 mM CaCl 2 , and resuspended in 50 mM TBS to obtain a 50% slurry. The immobilized enzyme was stable at Ϫ80°C for at least six months.
Assay for Sulfotransferase Activity-The standard reaction mixture contained 50 mM imidazole-HCl, pH 6.8, 75 g/ml of protamine chloride, 0.5 mM (as hexosamine) acceptor substrates, 1.5 M [ 35 S]PAPS (about 5 ϫ 10 5 dpm), and 1 l of immobilized enzyme in a final volume of 50 l. Acceptor substrates used to examine the substrate specificity were heparan sulfate, heparin, chondroitin sulfate A, chondroitin sulfate C, dermatan sulfate, keratan sulfate, and hyaluronic acid. After incubation at 37°C for 20 min, the reaction was stopped by heating at 100°C for 1 min. The reaction mixture was filtrated with an Ultrafree-MC (MILLIPORE, Bedford, MA) and 60 g of chondroitin sulfate C was added as a carrier. 35 S-Labeled substrates were precipitated with 3 volumes of ethanol containing 1.3% potassium acetate and separated completely from [ 35 S]PAPS and its degradation products by gel filtration HPLC on a TSKgel G2500PW column (0.75 ϫ 30 cm) equilibrated with 0.2 M NaCl at a flow rate of 0.6 ml/min and the column temperature of 35°C. Fractions of 0.3 ml were collected and the radioactivity was determined by liquid scintillation counting.
Preparation of Oligosaccharides from Heparin-It was reported that the heparinase digestion of heparin produces about 45% each disaccharide and tetrasaccharide, and a few percent hexasaccharide (36,37). Two milligrams of heparin was digested with 0.2 units of heparinase and applied to a BioGel P-4 column (1.5 ϫ 140 cm) equilibrated with 0.1 M ammonium bicarbonate. Elution positions of each oligosaccharide were monitored at 232 nm. Disaccharide, tetrasaccharide and hexasaccharide fractions were pooled and desalted by lyophilization.
HPLC Analysis on an Anion-exchange Column and a Gel Filtration Column-Anion-exchange (SAX) separation was performed on a Car-boPac TM PA1 column (4 ϫ 250 mm) as described (38). A combination of five linear LiCl gradients was used, from 30 to 180 mM (0 -5 min), from 180 to 570 mM (5-8 min), from 0.57 to 1.14 M (8 -15 min), from 1.14 to 2.1 M (15-20 min), and from 2.1 to 2.28 M (20 -28 min) followed by 2.28 M (from 28 min). The flow rate was 0.8 ml/min, and the column temperature was 40°C. Fractions were collected and the radioactivity was determined by liquid scintillation counting. To determine the molecular size of oligosaccharides, gel filtration HPLC was performed on a Superdex TM Peptide HR10/30 column (1 ϫ 30 cm, ϫ 2 columns in series) as described (39). The column was equilibrated with 0.2 M NaCl at a flow rate of 0.8 ml/min, at room temperature.
Preparation of 35 S-Labeled Heparan Sulfate and Digestion with Heparin Lyases-One milligram of heparan sulfate from bovine kidney and [ 35 S]PAPS (6.6 ϫ 10 7 dpm) was incubated with 0.5 ml of standard reaction mixture containing 67 l of immobilized enzyme at 37°C for 3 h. After the incubation, the reaction mixture was filtrated, and 35 Slabeled heparan sulfate was precipitated with ethanol as described above. The precipitate was dried at room temperature. The 35 S-labeled heparan sulfate was digested with a mixture of 0.5 units of heparinase, 0.3 units of heparitinase I, and 0.2 units of heparitinase II in 0.3 ml of 20 mM sodium acetate buffer (pH 7.0) containing 2 mM calcium acetate at 37°C for 2 h. The reaction was stopped by heating at 100°C for 1 min, and the mixture was filtrated. The digested products were subjected to chromatography on a BioGel P-4 column (1.5 ϫ 140 cm) equilibrated with 0.1 M ammonium bicarbonate at a flow rate of 5 ml/h. Fractions of 2 ml were collected. The fractions indicated by horizontal bars in Fig. 2, named P4-1 and P4-2, were pooled, lyophilized and completely desalted with a gel filtration column.
Digestion with 2-Sulfatase-The 35 S-labeled disaccharides were digested with 4 milliunits of 2-sulfatase in 20 mM sodium acetate buffer (pH 6.5) containing 1.5 mg/ml of bovine serum albumin. After incubation at 37°C for 2 h, the reaction was stopped by heating at 100°C for 1 min. The digested products were analyzed by HPLC on a SAX column as described above.
Determination of Glucosamine Residue-The 35 S-labeled disaccharides were treated with mercuric acetate to remove unsaturated uronic acid residues as described (40). The disaccharide fraction was mixed with an equal volume of 70 mM mercuric acetate (pH 5.0) and incubated at room temperature for 10 min. The reaction products, [ 35 S]sulfated glucosamines, were reduced with sodium borohydride as described (41) and analyzed by HPLC on a SAX column as described above.
Preparation of 35 S-Labeled Heparin and Digestion with Heparin Lyases-35 S-Labeled heparin was prepared by incubating 0.3 mg of heparin with [ 35 S]PAPS (2.3 ϫ 10 7 dpm) and 20 l of immobilized enzyme in a standard reaction mixture. The 35 S-labeled heparin was precipitated with ethanol and digested with a mixture of heparin lyases as described above. The digested products were subjected to HPLC on a SAX column as described above. Fractions of 0.24 ml were collected and the radioactivity was determined. A peak fraction at 30.1 min (indicated by a white arrow in Fig. 6) was applied to a Cellulofine G-25-sf column (1 ϫ 28 cm) equilibrated with distilled water to remove LiCl. Radioactive fractions were pooled and concentrated under vacuum.
Quantitative Analysis of the 3-OST-5 Transcripts in Human Tissues by Real-time PCR-For quantification of 3-OST-5 transcripts, we employed the real-time PCR method, as described in detail previously (42). Total RNAs derived from various human tissues were purchased from Clontech and cDNAs were synthesized with the SuperScript TM Firststrand Synthesis System for RT-PCR (Invitrogen). To obtain the control DNA of 3-OST-5, a DNA fragment containing exon I and the 5Ј-terminal region of exon II was amplified by PCR using the Marathon Ready TM cDNA, as a template, and two primers, 5Ј-TCTATCGCAGGCTGCAG-CAGTCCT-3Ј and 5Ј-GGAACTCGTGCAGCAGGCCACGC-3Ј. Standard curves for the 3-OST-5 and the endogenous control, glyceraldehyde-3phosphate dehydrogenase (GAPDH) cDNAs, were generated by serial dilution of the control DNA of 3-OST-5, and a pCR2.1 vector (Invitrogen) containing the GAPDH gene. The primer sets and the probes for 3-OST-5 were as follows: the forward primer was 5Ј-GCCAGAGTTGG-GAGCTTGG-3Ј, the reverse primer was 5Ј-ACCCAGTCGACCT-TCAATGG-3Ј, and the probe for 3-OST-5 was 5Ј-TAGGCTACAAC-CCATTTG-3Ј with a minor groove binder (43). PCR products were continuously measured with an ABI PRISM 7700 Sequence Detection System (Applied Biosystems, Foster City, CA). The relative amounts of 3-OST-5 transcripts were normalized to the amount of GAPDH transcript in the same cDNA.
Analysis of Bovine Intestinal Heparan Sulfate-Bovine intestinal heparan sulfate (0.2 mg) was digested with a mixture of heparin lyases as described above. The digested products were reduced with sodium [ 3 H]borohydride and subjected to HPLC on a SAX column as outlined above. 35 S-Labeled tetrasulfated disaccharide obtained from 3-OST-5modified heparin in this study was reduced with sodium borohydride and used as a standard.

RESULTS
Specificity for Acceptor Substrate-The putative catalytic domain of 3-OST-5 was cloned and expressed, and the sulfotransferase activity was examined as described in "Experimental Procedures." As expected, 3-OST-5 exhibited sulfotransferase activity toward heparan sulfate and heparin. On the other hand, it exhibited no activity toward chondroitin sulfate A, chondroitin sulfate C, dermatan sulfate, keratan sulfate, and hyaluronic acid. The activity of 3-OST-5 for heparan sulfate is presented as 100%, and the other activities are given as relative values in Table I.
Digestion of 35 (Fig. 1A). Several radioactive peaks were observed, but no peak was identified at the position of those six units (Fig. 1B). Therefore, the 35 S-labeled products were different from these six disaccharides. The digestion products were applied to a gel filtration column of BioGel P-4, and two major 35 S radioactive peaks were obtained (Fig. 2). The fractions indicated by a horizontal bar in Fig. 2, named P4-1 and P4-2, were pooled and desalted for further analysis. To confirm the molecular size, these two fractions were analyzed by HPLC on a gel filtration column. The elution profile of standard oligosaccharides prepared from heparin is shown in Fig. 3A. 35 S-Labeled substances, P4-1 and P4-2 in Fig. 2, were eluted at the position of the tetrasaccharide and disaccharide, respectively (Fig. 3, B and C). These two fractions were analyzed by HPLC on a SAX column, and each fraction gave two components. Judging from the elution position, P4-1 and P4-2 corresponded to peaks 1 and 3, and peaks 2 and 4, respectively, which were four major peaks observed in Fig. 1B. To confirm the resistance to enzyme digestion, P4-1 was treated with the heparin lyase mixture again. No change was observed in the appearance of peak 1 and peak 3 during HPLC analysis on a SAX column. Moreover, this fraction is also resistant to nitrous acid degradation. The P4-1 was not characterized further, since these tetrasaccharides  encountered some difficulty in further identification using enzymes (see "Discussion").
Structural Analysis of 35 S-Labeled Disaccharides from Heparan Sulfate-The 35 S-labeled disaccharide fraction (P4-2) was digested with 2-sufatase as described under "Experimental Procedures," and analyzed by HPLC on a SAX column. Fig. 4A shows the analysis of P4-2. With the enzyme digestion (Fig.  4B), only one component, peak 2 in panel A, shifted its retention time to an earlier position. The result indicates that peak 2 has a ⌬HexA(2S) residue, but not peak 1. To confirm the location of sulfate group of glucosamine residues, P4-2 was treated with mercuric acetate to remove unsaturated uronic acids. The reaction products were reduced with sodium borohydride and analyzed by HPLC on a SAX column (Fig. 5A). By comparison with 3 H-labeled standards (Fig. 5B), peak a and peak b were identified as reduced forms of GlcN(NS,3S) and GlcN-(NS,3S,6S). Judging from peak area, peak a and peak b originated from peak 2 and peak 1 in Fig. 4A, respectively. From these data, peak 1 and peak 2 in Fig. 4A were determined as ⌬HexA-GlcN(NS,3S,6S) and ⌬HexA(2S)-GlcN(NS,3S), respectively. Fig. 9, A and B show schematic drawings of the analysis described above.
Highly Sulfated Structure of 35 S-Labeled Heparin-35 S-Labeled heparin was prepared by incubating heparin with [ 35 S]PAPS and the recombinant 3-OST-5 and digested with a mixture of heparin lyases as described under "Experimental Procedures." When the digested products were separated by HPLC on a SAX column, major radioactivity was detected at the retention time of 30.1 min, later than the positions of trisulfated disaccharides (Fig. 6). The fraction indicated by a white arrow in Fig. 6 was pooled and desalted for further analysis. The 35 S-labeled product was confirmed to be a disaccharide by HPLC on a gel filtration column (data not shown). On 2-sulfatase digestion, the 35 S radioactive peak exhibited a shift in retention time to an earlier position (Fig. 7), indicative of a ⌬HexA(2S) residue. The retention time of the 2-sulfatasetreated product corresponded to that of a disaccharide isolated from 3-OST-5-modified heparan sulfate (peak 1 in Fig. 4A), its structure being determined as ⌬HexA-GlcN(NS,3S,6S) as described above. Therefore, ⌬HexA(2S)-GlcN(NS,3S,6S) was expected. To confirm the location of sulfate group of glucosamine residue, 35 S-labeled disaccharide was treated with mercuric acetate, reduced with sodium borohydride, and analyzed by HPLC on a SAX column (Fig. 8A). The position of the 35 S radioactive peak corresponded to that of 3 H-labeled GlcN (NS,3S,6S) standard (Fig. 8B). From these data, the 35 35 S-Labeled heparan sulfate digested with a mixture of heparin lyases was applied to a BioGel P-4 column as described under "Experimental Procedures." Fractions of 2 ml were collected, and radioactivity was determined. The fractions indicated by horizontal bars, named P4-1 and P4-2, were pooled.

FIG. 3. HPLC analysis on a gel filtration column of oligosaccharides derived from 35 S-labeled heparan sulfate.
Standard oligosaccharides were prepared from heparin as described in "Experimental Procedures." Then, 5 g of each oligosaccharide, hexasaccharide (Hexa), tetrasaccharide (Tetra), and disaccharide (Di), was applied to a Superdex TM Peptide HR10/30 column (panel A). The conditions for HPLC were as described under "Experimental Procedures." Panels B and C shows analysis of radioactive fractions obtained from the BioGel P-4 column, P4-1 and P4-2 in Fig. 2, respectively. as ⌬HexA(2S) -GlcN(NS,3S,6S). Fig. 9C shows a schematic drawing of the analysis described above.
Quantitative Analysis of the 3-OST-5 Transcripts in Human Tissues-We determined the tissue distribution and expression levels of the 3-OST-5 transcripts by a real-time PCR method. The expression levels of 3-OST-5 in various tissues were shown as a relative amount compared with the GAPDH transcripts (Fig. 10). The transcripts were highly expressed in fetal brain, followed by adult brain and spinal cord. Cerebellum, colon and skeletal muscle expressed the 3-OST-5 transcripts at a relatively low level. The expression levels in the remaining tissues were very low or undetectable.
Identification of Tetrasulfated Disaccharide in Bovine Intes-tinal Heparan Sulfate-To confirm the presence of the natural tetrasulfated disaccharide unit in heparan sulfate, we chose bovine intestinal heparan sulfate as a natural source because of the availability of highly purified material. The bovine intestinal heparan sulfate was digested with a mixture of heparin lyases as described under "Experimental Procedures." The digested products were reduced with sodium [ 3 H]borohydride and subjected to HPLC on a SAX column (Fig. 11). Six major 3 H radioactive peaks were identified as known disaccharides by comparing with disaccharide standards reduced with sodium borohydride. 35 S-Labeled tetrasulfated disaccharide obtained from 3-OST-5-modified heparin as described above was reduced with sodium borohydride and used as a standard. Its elution position is shown in Fig. 11 as ⌬Di-tetraS R . Weak but apparent radioactivity was detected at this position (inset of Fig. 11). To confirm the molecular size, the peak fraction (indicated by a white arrow) was analyzed by the HPLC on a gel filtration column. The 3 H-labeled substance was eluted at the position of the tetrasulfated disaccharide standard (data not shown). DISCUSSION The cloning and characterization of a sixth member of the human 3-OST family, named 3-OST-5, was recently reported. Xia et al. (35) employed low pH nitrous acid degradation to analyze the reaction products and identified three 3-O-sulfated disaccharides, GlcA-aMan(3S,6S), IdoA(2S)-aMan(3S), and IdoA(2S)-aMan(3S,6S). In the case of 3-O-sulfated glucosamine, both N-sulfated and N-unsubstituted residues are susceptible to low pH nitrous acid degradation, therefore the Nsubstituents of the glucosamine residue was unidentified (31). They also confirmed that the 3-OST-5-modified heparan sulfate binds to AT III and gD. In addition, transfection of the plasmid expressing 3-OST-5 makes CHO cells susceptible to HSV-1. From these results, they concluded that 3-OST-5 has the activities of both 3-OST-1 and 3-OST-3. The reaction specificity of 3-OST-1 and 3-OST-3 is the formation of GlcA-GlcN-(NS,3S,Ϯ6S) and Ido(2S)-GlcN(3S,Ϯ6S), respectively (31,44).
In the present study, we employed enzyme digestion to analyze the reaction products. 35 S-Labeled heparan sulfate, which was prepared by incubating heparan sulfate with [ 35 S]PAPS and recombinant 3-OST-5, was digested with a mixture of heparin lyases. Two major 35 S-labeled disaccharides obtained from the digests were determined as ⌬HexA-GlcN(NS,3S,6S) and ⌬HexA(2S)-GlcN(NS,3S). These two disaccharides appear to resemble the reaction products of 3-OST-1 and 3-OST-2, respectively, although the types of uronic acids are unidentified.
It has been believed that the glucosaminidic linkage adjacent to a disaccharide unit containing a 3-O-sulfated glucosamine residue is resistant to heparin lyases (45,46). To date, unsaturated disaccharides containing a 3-O-sulfated glucosamine were generated by serial dilution of control DNA or plasmid DNA as described under "Experimental Procedures." The expression level of 3-OST-5 was normalized to that of the GAPDH transcript, which was measured in the same cDNAs. Data were obtained from triplicate experiments and are given as the mean Ϯ S.D. PBMC, peripheral blood mononuclear cell. residue have not been isolated. On the other hand, Rhomberg et al. (47) reported that a decasaccharide containing a 3-O-sulfated glucosamine residue was depolymerized by heparinase II, and confirmed that the reaction product contained an unsaturated 3-O-sulfated disaccharide unit in the non-reducing terminal. They also examined a shorter oligosaccharide resistant to heparinase II and concluded that the length of the oligosaccharide determines the susceptibility to the enzyme. Accordingly, it is considered that the resistance of tetrasaccharides isolated in this study to the enzyme digestion may be due to the short length of the chain. Interestingly, the enzyme digestion of 3-OST-3-modified heparan sulfate produces no disaccharide containing a 3-O-sulfated glucosamine residue (31). At this time, the reaction specificity of heparin lyases especially for the 3-O-sulfated structure is still obscure. A more detailed analysis of the sulfated monosaccharide sequence around the modification site of 3-OST-5 will be helpful for solving this problem.
In addition to heparan sulfate, we analyzed 3-OST-5-modified heparin. HPLC analysis of 35 S-labeled heparin digested with a mixture of heparin lyases showed a very different elution profile compared with heparan sulfate. A prominent peak was observed at the later position of trisulfated disaccharides, and its structure was determined as ⌬HexA(2S)-GlcN (NS,3S,6S). As far as we know, this is the first report to show the enzyme catalyzes the reaction to form the tetrasulfated disaccharide unit. A small peak detected at the elution position of tetrasulfated disaccharide in Fig. 1B (peak 5) indicates enzymatic formation of this structure with heparan sulfate as an acceptor substrate.
3-OST-5 was highly expressed in fetal brain, followed by adult brain and spinal cord. Some other tissues, including skeletal muscle, also expressed 3-OST-5 at a relatively low level. Xia et al. (35) did not examine fetal brain in their Northern blot analysis. They also did not detect the expression of 3-OST-5 in the brain, and concluded that 3-OST-5 is predominantly expressed in skeletal muscle. This discrepancy is unexplained, although real time-PCR is a more quantitative method than Northern blot analysis. 3-OST-2 and 3-OST-4 were also predominantly expressed in human brain. 3-OST-1 and 3-OST-3 were widely expressed in human tissues (34). Heparan sulfate in the nervous system is predominantly expressed on proteoglycans of either the syndecan family (48) or the glypican family (49). A number of investigations have revealed that the development of the central nervous system is controlled by the spatiotemporal expression of these proteoglycans (50). Therefore, it can be expected that the 3-OSTs specifically expressed in the central nervous system play an important role in forming specific domain structures of heparan sulfate to bind some functional proteins necessary to control the central nervous system.
In the present study, we demonstrated the enzymatic formation of a tetrasulfated disaccharide unit in vitro. No evidence of this novel structure in natural heparan sulfate or heparin has been reported as far as we know. Therefore, we focused on the detection of the tetrasulfated disaccharide unit in natural heparan sulfate. We chose bovine intestinal heparan sulfate as a natural source because of the availability of highly purified material, even though the expression level of 3-OST-5 is relatively low in human intestine. As the result of analysis, bovine intestinal heparan sulfate was found to contain about 0.15% tetrasulfated disaccharide (mol/mol of total disaccharides). The retention time of tetrasulfated disaccharide on a SAX-column shifted markedly to an early position on reduction with sodium borohydride, although the retention times of six major disaccharides were not changed significantly by the reduction. Interestingly, the same phenomenon was observed with ⌬HexA-GlcN(NS,3S,6S) and ⌬HexA(2S)-GlcN(NS,3S). Therefore this is a characteristic of disaccharides containing a 3-O-sulfated glucosamine residue.
Although 3-OST-5 is the only enzyme known to catalyze the formation of the tetrasulfated disaccharide unit at this time, there is no evidence that the novel structure detected in this study was synthesized by the 3-OST-5 in bovine intestine. The generation of transgenic and gene knockout animal models could provide not only the answer to this question but also important information about the biological role of 3-OST-5 or its reaction products including tetrasulfated disaccharide. FIG. 11. HPLC analysis on a SAX column of 3 H-labeled disaccharides derived from bovine intestinal heparan sulfate. Heparan sulfate from bovine intestine was digested with a mixture of heparin lyases as described under "Experimental Procedures." The digested products were reduced with sodium [ 3 H]borohydride and separated by HPLC on a SAX column. The conditions for HPLC were as described under "Experimental Procedures." Fractions of 0.24 ml were collected, and radioactivity was determined. Six known disaccharides were identified in comparison with the disaccharide standards reduced with sodium borohydride. Abbreviations of disaccharides are shown in the legend of Fig. 1. A letter R attached to the name means reduced form. 35 S-Labeled ⌬HexA(2S)-GlcN(NS,3S,6S) obtained from 3-OST-5modified heparin as described in the present study was reduced with sodium borohydride and used as a standard. Its elution position is indicated by ⌬Di-tetraS R . sodium [ 3 H]borohydride was eluted at the retention time of 7 min. The inset represents an enlarged vertical axis with the same horizontal scale. The fraction indicated by a white arrow was used for further analysis.