Molecular Cloning and Characterization of a Human β-Gal-3′-sulfotransferase That Acts on Both Type 1 and Type 2 (Galβ1–3/1–4GlcNAc-R) Oligosaccharides*

A novel sulfotransferase gene (designated GP3ST) was identified on human chromosome 2q37.3 based on its similarity to the cerebroside 3′-sulfotransferase (CST) cDNA (Honke, K., Tsuda, M., Hirahara, Y., Ishii, A., Makita, A., and Wada, Y. (1997) J. Biol. Chem. 272, 4864–4868). A full-length cDNA was obtained by reverse transcription-polymerase chain reaction and 5′- and 3′-rapid amplification of cDNA ends analyses of human colon mRNA. The isolated cDNA clone predicts that the protein is a type II transmembrane protein composed of 398 amino acid residues. The amino acid sequence indicates 33% identity to the human CST sequence. A recombinant protein that is expressed in COS-1 cells showed no CST activity, but did show sulfotransferase activities toward oligosaccharides containing nonreducing β-galactosides such asN-acetyllactosamine, lactose, lacto-N-tetraose (Lc4), lacto-N-neotetraose (nLc4), and Galβ1–3GalNAcα-benzyl (O-glycan core 1 oligosaccharide). To characterize the cloned sulfotransferase, a sulfotransferase assay method was developed that uses pyridylaminated (PA) Lc4 and nLc4 as enzyme substrates. The enzyme product using PA-Lc4 as an acceptor was identified as HSO3-3Galβ1–3GlcNAcβ1–3Galβ1–4Glc-PA by two-dimensional 1H NMR. Kinetics studies suggested that GP3ST is able to act on both type 1 (Galβ1–3GlcNAc-R) and type 2 (Galβ1–4GlcNAc-R) chains with a similar efficiency. In situ hybridization demonstrated that the GP3ST gene is expressed in epithelial cells lining the lower to middle layer of the crypts in colonic mucosa, hepatocytes surrounding the central vein of the liver, extravillous cytotrophoblasts in the basal plate and septum of the placenta, renal tubules of the kidney, and neuronal cells of the cerebral cortex. The results of this study indicate the existence of a novel β-Gal-3′-sulfotransferase gene family.

Sulfated glycoconjugates, whose sulfate groups are biologically relevant, occur in a wide range of biological compounds (reviewed in Ref. 1), including glycoproteins, proteoglycans, glycolipids, and polysaccharides. The enzymes responsible for the sulfation of these compounds, sulfotransferases, utilize in common the sulfate donor 5Ј-phosphoadenosine 3Ј-phosphosulfate (PAPS). 1 We recently reported on the purification of the glycolipid 3-sulfotransferase (cerebroside sulfotransferase (CST); galactosylceramide sulfotransferase, EC 2.8.2.11) to homogeneity from human renal cancer cells (2) and the cloning of the human CST cDNA on the basis of the amino acid sequence of the purified enzyme (3). The deduced amino acid sequence shows no overall homology to other sulfotransferases, except for the PAPS-binding motifs, suggesting that CST has a different evolutionary origin (3).

Materials-[
The nucleotide sequence reported in this paper has been submitted to the DDBJ/GenBank TM /EBI Data Bank with accession number AB040610.
Cloning of Human GP3ST cDNA-One microgram of total RNA isolated from various human tissues (OriGene Technologies, Rockville, MD) was reverse-transcribed with random hexamers and then subjected to PCR. PCR was carried out using Taq DNA polymerase and two pairs of primers: CF2 (5Ј-TCACCAACATCATGTTC-3Ј, corresponding to human GP3ST cDNA nucleotides 284 -300 in Fig. 1A) and CR2 (5Ј-TGAGGCAGGTTGAACCT-3Ј, corresponding to nucleotides 490 -506); and CF1 (5Ј-CCAACGACACCTTCTAC-3Ј, corresponding to nucleotides 524 -540) and CR1 (5Ј-GGAACGGGATGTTCTTG-3Ј, corresponding to nucleotides 1302-1318), which are located in the putative exons of the CST-like candidate gene in the CEB1 cosmid (GenBank TM / EBI accession number AF048727). The PCR products were electrophoresed on 1.5% agarose gel. DNA fragments with the expected sizes of 223 and 795 base pairs, respectively, were excised and subcloned into pT7 Blue(R) (Novagen, Madison, WI) and sequenced as described below. The determined sequences were identical to those in the CEB1 cosmid, and the cloned DNA fragments were designated LIKE-C1 and LIKE-C2.
The 5Ј-end of human GP3ST cDNA was cloned by the rapid amplification of cDNA ends (RACE) (21) using a 5Ј-RACE system kit (Life Technologies, Inc.). According to the manufacturer's recommended protocol, 1 g of total RNA from human colon (OriGene Technologies) was reverse-transcribed with the GP3ST gene-specific antisense primer CR2. The first strand cDNA was tailed at the 3Ј-end by terminal deoxynucleotidyltransferase with dATP and then subjected to PCR. PCR was accomplished using Taq DNA polymerase, a 3Ј-RACE adapter primer (5Ј-GGCCACGCGTCGACTAGTACTTTTTTTTTTTTTTTTT-3Ј), and the GP3ST gene-specific antisense primer CR3 (5Ј-GACCCCACGC-CTTCCACGTA-3Ј, corresponding to nucleotides 436 -455). The PCR product was electrophoresed on 1.5% agarose gel, and the DNA was transferred to a nylon membrane (Roche Molecular Biochemicals, Tokyo) and hybridized with a digoxigenin-labeled DNA probe synthesized using LIKE-C1 as a template. Southern hybridization was carried out at 42°C overnight in hybridization buffer containing 50% formamide, 5ϫ SSC, 2% blocking reagent (Roche Molecular Biochemicals), 0.1% N-lauroylsarcosine, and 0.02% SDS. Positive bands were cut out, subcloned into pT7 Blue(R), and sequenced.
The 3Ј-end of human GP3ST cDNA was cloned using a 3Ј-RACE system kit (Life Technologies, Inc.) in a similar manner. According to the manufacturer's protocol, 1 g of human colon total RNA was reverse-transcribed with the 3Ј-RACE adapter primer and subsequently subjected to PCR. PCR was performed using Taq DNA polymerase, the abridged universal amplification primer (5Ј-GGCCACGCGTCGAC-TAGTAC-3Ј), and the GP3ST gene-specific sense primer CF5 (5Ј-TCAA-GAACCACACGCAGATC-3Ј, corresponding to nucleotides 1115-1134). The PCR product was electrophoresed on 1.5% agarose gel and subjected to Southern blotting as described above, except that the hybridization probe was synthesized using LIKE-C2 as a template.
DNA Sequencing-The subcloned DNAs were sequenced by the dideoxy chain termination method using a Dye Terminator cycle sequencing kit (PerkinElmer Life Sciences) with a DNA sequencer (Applied Biosystems Model 377).
Overexpression of Human GP3ST in COS-1 Cells and Sulfotransferase Assay-To obtain a GP3ST cDNA clone that contained the entire open reading frame, RT-PCR was accomplished using total RNA from human colon as a template. PCR was performed using Taq DNA polymerase and a set of primers (CF3, 5Ј-CCAGAGGCCAAGATGATG-3Ј, corresponding to nucleotides 121-138 in Fig. 1A); and CR6, 5Ј-GGAGAGAGGAGCTGGTGT-3Ј, corresponding to nucleotides 1362-1379). The PCR product with the expected size of 1259 base pairs was subcloned into pT7 Blue(R) and sequenced as described above. The obtained clone was digested with EcoRI and SalI, and the GP3ST open reading frame fragment was inserted into the EcoRI and SalI sites of pSVK3 (Amersham Pharmacia Biotech), yielding pSV-GP3ST. COS-1 cells (1 ϫ 10 6 ) that had been pre-cultured for 1 day in a 10-cm diameter dish were transfected with 10 g of pSV-GP3ST and 30 l of Lipo-fectAMINE (Life Technologies, Inc.). After 72 h, the cells were washed twice with 10 ml of cold phosphate-buffered saline, harvested with 1 ml of phosphate-buffered saline using a silicon scraper, centrifuged at 1500 rpm for 2 min, sonicated in 0.3 ml of ice-cold Tris-buffered saline containing 0.1% Triton X-100, and assayed for sulfotransferase activity according to the method for CST activity (2). When oligosaccharides were used as acceptor substrates, water was employed as the eluant in DEAE A25 column chromatography as a substitute for the organic solvent.
GP3ST Activity Assay Using PA-Lc4 and PA-nLc4 -The standard incubation mixture contained the following components in a total volume of 25 l: 100 mM MES (pH 6.2), 10 mM MnCl 2 , 1% Lubrol PX, 1 mM PAPS, 0.2 mM PA-Lc4 or PA-nLc4, and 7.5 l of enzyme source. After incubation at 37°C for 2 h, the reaction was terminated by boiling for 2 min. The sample was centrifuged at 15,000 rpm for 5 min, and 20 l of the supernatant was transferred to a new tube, followed by dilution with 180 l of water; and 50 l of the diluted solution was injected onto a TSKgel SuperQ-5PW column (7.5 ϫ 75 mm; Tosoh) equipped with a Shimadzu LC-VP HPLC system. Elution was performed with a linear gradient of 0 -0.2 M ammonium acetate (pH 9.0) and monitored with a fluorescence spectrophotometer (excitation, 320 nm; and emission, 400 nm). 1 H NMR Analysis of the Reaction Product-For preparation of the NMR sample, a large-scale GP3ST reaction was carried out using PA-Lc4 as a substrate, and the reaction product was isolated by anionexchange chromatography as described above. The eluate of the anionexchange chromatography was then directly injected onto a TSKgel ODS-80TM column (4.6 ϫ 75 mm; Tosoh). Elution was carried out at 55°C with 20 mM ammonium acetate and 0.01% 1-butanol (pH 4.0) at a flow rate of 1.0 ml/min and monitored as described above. A single peak was detected. The effluent corresponding to the peak was collected and dried. Approximately 17 nmol of the reaction product, as evaluated by fluorescence intensity, was deuterium-exchanged; dissolved in 0.1 ml of D 2 O; and subjected to 1 H NMR analysis as described previously (22,23), except for the use of microtubes (Shigemi, Tokyo). The 1 H signals of the pyridylaminated oligosaccharides were assigned based on the 1 H-1 H DQF-COSY, two-dimensional HOHAHA, and two-dimensional ROESY experiments in a Joel GX-400 spectrometer at 30°C. In the case of the DQF-COSY and two-dimensional HOHAHA experiments, the data matrices of 256 (t1) ϫ 4096 (t2) points, acquired with a spectral width of 2000 Hz, were zero-filled to 1048 ϫ 4096 data points. The mixing time of the two-dimensional HOHAHA was set at 100 ms, and the twodimensional ROESY spectrum was acquired with a 200-ms spin-lock period with a spectral width of 5 kHz.
In Situ Hybridization of GP3ST Transcripts-To detect GP3ST transcripts in normal human tissues, in situ hybridization was carried out using an RNA probes with nonradioactive labeling as described previously (24). GP3ST gene-specific nucleotide sequence (nucleotides 720 -861 in Fig. 1A) was amplified by PCR using a 5Ј-primer (5Ј-CCCAAG-CTTCGGCTTCGACCCCAA-3Ј) and a 3Ј-primer (5Ј-CGGAATTCCAGC-GCAGCCGGCGC-3Ј) (HindIII and EcoRI sites are underlined). This amplified DNA fragment was cloned into the HindIII and EcoRI sites of pGEM-3Zf (ϩ) (Promega), and the resultant vector was used for construction of the RNA probe. A digoxigenin-labeled antisense RNA probe was obtained using a HindIII-cut template and T7 RNA polymerase with a DIG RNA labeling mixture (Roche Molecular Biochemicals). Similarly, a sense probe was prepared for negative control experiments using an EcoRI-cut template and SP6 RNA polymerase with the DIG RNA labeling mixture.
Paraffin blocks of normal human tissues, including colon, liver, placenta, kidney, and brain, fixed for 24 h in 20% buffered Formalin (pH 7.4) were selected from the pathology file of the Central Clinical Laboratories of the Shinshu University Hospital (Matsumoto, Japan). The deparaffinized tissue sections were immersed in 0.2 M HCl for 20 min and then digested with 100 g/ml proteinase K at 37°C for 20 min, followed by post-fixation with 4% paraformaldehyde. These slides were rinsed with 2 mg/ml glycine, acetylated for 10 min in 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8.0), and then defatted with chloroform and air-dried. After prehybridization with 50% deionized formamide and 2ϫ SSC for 1 h at 45°C, these slides were hybridized with 0.5 mg/ml antisense or sense probe in 50% deionized formamide, 2.5 mM EDTA (pH 8.0), 300 mM NaCl, 1ϫ Denhardt's solution, 10% dextran sulfate, and 1 mg/ml brewers' yeast tRNA at 45°C for 48 h.
After hybridization, the slides were washed with 50% formamide and 2ϫ SSC for 1 h at 45°C and digested with 10 mg/ml RNase A at 37°C for 30 min. After sequential washing with 2ϫ SSC and 50% formamide at 45°C for 1 h and 1ϫ SSC and 50% formamide at 45°C for 1 h, the sections were subjected to immunohistochemistry for detection of the hybridized probes using an alkaline phosphatase-conjugated antidigoxigenin antibody (Roche Molecular Biochemicals). The alkaline phosphatase reaction was visualized with 5-bromo-4-chloro-3-indolyl phosphate and nitro blue tetrazolium in the presence of levamisole and 10% polyvinyl alcohol. A control study using the sense probe showed no specific reactivity.  1. cDNA and deduced amino acid sequences of human GP3ST. A, the predicted amino acid sequence is indicated by the single-letter amino acid code below the nucleotide sequence. The putative transmembrane portion is double-underlined. The potential N-glycosylation site is indicated by an asterisk. The presumed PAPS-binding sites are boxed. B, shown is a comparison of the amino acid sequences of human GP3ST and human CST (hCST) (3). Asterisks below the sequences show identical residues.

RESULTS
cDNA Cloning and the Predicted Protein Sequences of a Novel Sulfotransferase-When the GenBank TM /EMBL/DDBJ DNA Data Bank was searched for a sequence homologous to the coding region of the human CST gene (3), a cosmid clone containing the minisatellite CEB1 (GenBank TM /EBI accession number AF048727), located on human chromosome 2q37.3, included a significantly similar sequence (25). The CEB1 cosmid includes two exons with open reading frames that together cover 1076 base pairs of the human CST mRNA (25). To investigate the issue of whether the CEB1 candidate gene actually encodes a functional sulfotransferase, we elected to clone its cDNA from human tissues. First, the cDNA fragments LIKE-C1 (nucleotides 284 -506 in Fig. 1A) and LIKE-C2 (nucleotides 524 -540), which were found in the putative exons of the CEB1 candidate gene, were obtained by RT-PCR using total RNAs from various human tissues. Since the RT-PCR product was strongly detected in the colon (Fig. 5), we employed human colon as a source of mRNA for subsequent cDNA cloning. The 5Ј-and 3Ј-ends of the cDNA of the sulfotransferase candidate were obtained by RACE analyses as described under "Experimental Procedures." Fig. 1A shows the DNA and deduced amino acid sequences of the cloned cDNA (termed GP3ST). The deduced protein is a type II membrane protein composed of 398 amino acids with a molecular mass of 46,092 Da and has six potential N-glycosylation sites. Its sequence exhibits 33% identity to the human CST sequence (Fig. 1B) and contains PAPS-binding motifs (26), suggesting that the cDNA encodes a sulfotransferase.
Substrate Specificity of GP3ST-To investigate the issue of whether the isolated cDNA encodes a functional sulfotransferase and, if so, on what substrates the sulfotransferase acts, the cloned cDNA was inserted into a mammalian expression vector, pSVK3 (pSV-GP3ST) and overexpressed in COS-1 cells. We performed a substrate specificity experiment using [ 35 S]PAPS as the sulfate donor, various carbohydrate compounds as acceptors, and cellular lysates of pSV-GP3ST-introduced COS-1 cells as an enzyme source (Table I). A recombinant protein expressed in COS-1 cells showed no CST activity, but did show significant sulfotransferase activities toward oli-gosaccharides containing nonreducing ␤-galactosides such as N-acetyllactosamine, lactose, Lc4, nLc4, and Gal␤1-3Gal-NAc␣-benzyl (O-glycoside core 1 oligosaccharide). On the other hand, the COS-1 cells transfected with pSV-CST (3) or with pSVK3 alone did not show sulfotransferase activity toward any of the oligosaccharides examined (data not shown).
Sulfotransferase Assay Using PA-Lc4 or PA-nLc4 as an Acceptor-To further characterize the cloned sulfotransferase, we developed a novel sulfotransferase assay system using pyridylaminated oligosaccharides as acceptors. This system was originally developed for use in the structural analysis of oligosaccharides (27) and was later applied to glycosyltransferase assays (28). To separate the acidic products from the neutral substrates, anion-exchange chromatography was employed. A typical elution pattern of the reaction products when PA-Lc4 or PA-nLc4 was used as an acceptor is shown in Fig. 2. The arrows a The concentration of acceptor substrates was 1 mM, except for galactosylceramide (GalCer) and lactosylceramide (LacCer), which were used at 0.2 mM. b Ten micrograms of COS-1 cell lysates that had been transfected with pSV-GP3ST was used for each assay. When N-acetyllactosamine (LacNAc) was used as a substrate, 0.75 nmol of [ 35 S]sulfate was incorporated for 1 h. The values represent the percentage of the activity compared with that of N-acetyllactosamine.
indicate the elution positions of the products. When COS-1 cell lysates transfected with pSV-GP3ST were used as an enzyme source, a robust peak was detected at the positions indicated by the arrows (Fig. 2, b and d), whereas COS-1 cells transfected with the pSVK3 vector alone showed no significant peak at the indicated positions (Fig. 2, a and c). The yield of GP3ST products was increased relative to incubation time (over a period of 2 h) in a linear fashion, as well as at enzyme concentrations up to 1 mg/ml (data not shown).
Identification of the Reaction Product-To investigate the position where the sulfate group was transferred, a 1 H NMR analysis of the enzyme product was performed. For this, a large-scale reaction was performed using PA-Lc4 as a donor, and the product was isolated by sequential HPLC with anionexchange chromatography and C 18 reverse-phase chromatography as described under "Experimental Procedures." H1-H4 of the Gal residues in PA-Lc4 and the sulfation product (PA-Lc4-S) were assigned by DQF-COSY and twodimensional HOHAHA. As shown in Fig. 3, the H2-H4 signals of the terminal Gal (IV) residue in PA-Lc4-S were shifted to the lower field by 0.14, 0.67, and 0.38 ppm, respectively, indicating unequivocally that the hydroxyl group at C-3 was esterified by a sulfate. On the other hand, the signals of the internal Gal (II) residue in PA-Lc4-S remained unchanged relative to those in PA-Lc4. To investigate whether or not C-6 was sulfated, the H5 protons were assigned from the two-dimensional ROESY spectra in which the cross-peak between H1 and H5 was observed. The difference in chemical shift of the H5 proton in Gal (IV) was 0.06 ppm (Table II), which is typical for the sulfation at C-3 of the terminal residue (22,29). The additional downfield shift by ϳ0.2 ppm, which should result from the sulfation at C-6 (22, 30), was not observed. These results clearly show that GP3ST is a ␤-Gal-3Ј-sulfotransferase that transfers a sulfate group to C-3 of the nonreducing ␤-galactosyl residue.
Some Properties of GP3ST-The optimal pH was found to be between 6.0 and 6.5 (Fig. 4) when a MES buffer was used. The effect of divalent cations on GP3ST activity is summarized in Table III. GP3ST essentially has no requirement for divalent cations; and in fact, Cu 2ϩ and Zn 2ϩ strongly inhibited the sulfotransferase activity. The K m values for PA-Lc4, PA-nLc4, and PAPS were found to be 220, 480, and 240 M, respectively (Table IV). Although the K m value for PA-nLc4 was higher than that for PA-Lc4, the V max for PA-nLc4 was higher than that for PA-Lc4, and V max /K m values were nearly equal for PA-Lc4 and PA-nLc4, suggesting that GP3ST is able to act on both type 1 and type 2 chains with a similar efficiency.
Tissue-specific Expression of the GP3ST Gene-Northern blot and RT-PCR analyses using total RNAs from various human tissues consistently showed that GP3ST is ubiquitously transcribed (Fig. 5). Among the examined organs, GP3ST mRNA was found to be highly expressed in heart, stomach,   colon, liver, and spleen.
Expression of GP3ST Transcripts in Normal Human Tissues-Since the Northern blot analysis revealed that GP3ST is ubiquitously transcribed, we performed in situ hybridization using an RNA probe specific for GP3ST to determine the cell types that actually express GP3ST transcripts (Fig. 6). Among the various human tissues examined, a specific signal for GP3ST was detected in epithelial cells lining the lower to middle layer of the crypts in colonic mucosa, hepatocytes surrounding the central vein of the liver, extravillous cytotrophoblasts in the basal plate and septum of the placenta, renal tubules of the kidney, and neuronal cells of the cerebral cortex. DISCUSSION We report herein the identification of a novel functional sulfotransferase gene (named GP3ST) based on the similarity of its DNA sequence to that of the CST gene (3). The primary structure of GP3ST showed an overall similarity to that of CST with 33% identity. GP3ST catalyzes the transfer of a sulfate group to C-3 of the nonreducing ␤-galactoside in oligosaccharides such as CST. The similarities of structure and function between CST and GP3ST argue that they are members of the ␤-Gal-3Ј-sulfotransferase family. Similarly, several isoforms of this carbohydrate sulfotransferase have been found, e.g. heparan-sulfate N-deacetylase/N-sulfotransferase (31)(32)(33), heparan-sulfate glucosaminyl-3-O-sulfotransferase (34,35), heparan-sulfate 6-O-sulfotransferase (36), chondroitin-sulfate/ keratan-sulfate 6-O-sulfotransferase (17,37), and GlcNAc-6sulfotransferase (18 -20). The distinct regulation of gene expression and the different substrate specificities of these isoforms may give rise to the specificity and diversity of sulfated glycans that are often found in a particular cell type.
CST acts on various nonreducing terminal ␤-galactosides in glycolipids such as galactosylceramide, lactosylceramide, galactosyl-1-alkyl-2-acyl-sn-glycerol, and galactosyldiacylglycerol, but it does not act on oligosaccharides that are not associated with lipid (Ref. 2 and this study), suggesting that CST recognizes the nonreducing ␤-galactoside of an oligosaccharide chain attached to a lipid moiety. On the other hand, GP3ST acts only on the nonreducing terminal ␤-galactoside of an oligosaccharide without lipid, suggesting that this enzyme is involved in the biosynthesis of sulfated glycoproteins. The substrate specificity of GP3ST is similar to that of the mucin 3Ј-sulfotransferase activity in rat colonic mucosa (15), although GP3ST prefers N-acetyllactosamine to the O-glycan core 1 structure. Since the sulfation of core 1 prevents the branching reaction to form a core 2 structure (15), GP3ST may regulate the biosynthesis of core 2. GP3ST also acts on methyl-␤-galactoside to some extent, like the mucin 3Ј-sulfotransferase in human respiratory mucosa (13). Since N-acetyllactosamine is a good substrate for GP3ST, it may catalyze the sulfation of the N-acetyllactosamine structures of N-glycans like the calf thyroid 3Ј-sulfotransferase (4). The issue of whether GP3ST actually acts on N-glycans, O-glycans, or both remains to be solved.
The expression of the GP3ST gene was found to be ubiqui-  The sizes of the molecular markers are indicated on the right. b, RT-PCR analysis. Next, 1 g of total RNA from the indicated organs was reverse-transcribed with a random primer. The resulting cDNA was amplified by PCR using a set of GP3ST primers: CF2 (nucleotides 284 -300 in Fig. 1A) and CR2 (nucleotides 490 -506). The PCR product was electrophoresed, blotted, and hybridized with a digoxigenin-labeled DNA probe synthesized using the GP3ST cDNA as a template. The observed PCR products showed the predicted size of 223 bases. tous, with a relatively high level in the colon, where a considerable amount of mucin is generated and secreted. GP3ST may be involved in the biosynthesis of sulfated mucin in the human colon. A decreased level of mucin sulfation has been observed in colon carcinomas and ulcerative colitis (43)(44)(45)(46). This decrease in mucin sulfation is associated with reduced Gal-3Ј-sulfotransferase activity (14,47). In some human breast cancer cell lines, the similar sulfotransferase activity is much lower than in normal mammary cells (16). It would be interesting to study the expression of the GP3ST gene in tumor and normal tissues of these secreting organs. The high level of GP3ST expression in heart might be involved in the production of the tumor necrosis factor-␣-inducible sulfated ligand for L-selectin in human cardiac microvascular endothelial cells (48).
To measure the enzyme activity of a large number of samples, it is essential to establish a simple and sensitive assay method. The method developed in this study does not require radioisotopes and is a convenient and very sensitive assay method for GP3ST using a unique fluorescent-labeled pyridylaminated oligosaccharide as an acceptor substrate. The method is sufficiently sensitive for the detection of 0.5 pmol of reaction product and can be applied to other sulfotransferase assays. In conclusion, we report the cloning of the cDNA of ␤-Gal-3Ј-sulfotransferase that is involved in the biosynthesis of glycoprotein and propose a novel ␤-Gal-3Ј-sulfotransferase gene family.