Purification and Characterization ofN-Acetylgalactosamine 4-Sulfate 6-O-Sulfotransferase from the Squid Cartilage*

N-Acetylgalactosamine 4-sulfate 6-O-sulfotransferase (GalNAc4S-6ST), which transfers sulfate from 3′-phosphoadenosine 5′-phosphosulfate (PAPS) to position 6 of N-acetylgalactosamine 4-sulfate in chondroitin sulfate and dermatan sulfate, was purified 19,600-fold to apparent homogeneity from the squid cartilage. SDS-polyacrylamide gel electrophoresis of the purified enzyme showed a broad protein band with a molecular mass of 63 kDa. The protein band coeluted with GalNAc4S-6ST activity from Toyopearl HW-55 around the position of 66 kDa, indicating that the active form of GalNAc4S-6ST may be a monomer. The purified enzyme transferred sulfate from PAPS to chondroitin sulfate A, chondroitin sulfate C, and dermatan sulfate. The transfer of sulfate to chondroitin sulfate A and dermatan sulfate occurred mainly at position 6 of the internal N-acetylgalactosamine 4-sulfate residues. Chondroitin sulfate E, keratan sulfate, heparan sulfate, and completely desulfated N-resulfated heparin were not efficient acceptors of the sulfotransferase. When a trisaccharide or a pentasaccharide having sulfate groups at position 4 ofN-acetylgalactosamine was used as acceptor, efficient sulfation of position 6 at the nonreducing terminalN-acetylgalactosamine 4-sulfate residue was observed.

Sulfotransferases capable of producing GalNAc(4,6-bis-SO 4 ) residues have been examined in squid cartilage (24), quail oviduct (25), and human serum (26,27). The sulfotransferase from the quail oviduct and human serum mainly catalyzed sulfation of position 6 of nonreducing terminal GalNAc(4-SO 4 ) residues (25)(26)(27), whereas sulfotransferase partially purified from the squid cartilage mainly catalyzed transfer of sulfate to position 6 of internal GalNAc(4-SO 4 ) residues (24,27). However, strict substrate specificities of these sulfotransferases are still obscure because no homogeneous preparations of these sulfotransferases have been obtained. In the present study, we purified to apparent homogeneity a sulfotransferase from squid cartilage that transfers sulfate to position 6 of the GalNAc(4-SO 4 ) residue of chondroitin sulfate. This sulfotransferase was found to transfer sulfate to position 6 of both internal and nonreducing terminal GalNAc(4-SO 4 ) residue. Thus, this enzyme may be described as an N-acetylgalactosamine 4-sulfate 6-O-sulfotransferase (GalNAc4S-6ST).
‡ To whom correspondence should be addressed.
linked 6% agarose with covalently coupled dye) was from Millipore. Fresh squids, Ommastrephes sloani pacificus, were obtained locally. Human meniscus CS-C was a gift from Dr. H. Habuchi, Institute for Molecular Science of Medicine, Aichi Medical University.
Step 1. Preparation of Crude Extracts-Squid cranial cartilage was dissected, freed of soft tissues by wiping with cotton cloth and cut into slices with a razor. The slices were placed in 3 volumes of ice-cold 0.5 M NaCl in buffer A (10 mM Tris-HCl, pH 7.2, 20 mM MgCl 2 , 2 mM CaCl 2 , 10 mM 2-mercaptoethanol, 0.5% Triton X-100, 20% glycerol) containing protease inhibitors (5 M N ␣ -p-tosyl-L-lysine chloromethyl ketone, 3 M N-tosyl-L-phenylalanine chloromethyl ketone, 30 M phenylmethyl sulfonyl fluoride, and 3 M pepstatin A), and homogenized with a Polytron homogenizer 5 times at speed 7 for 30 s. After gentle stirring for 24 h, the homogenate was centrifuged at 100,000 ϫ g for 60 min. The clear supernatant fraction was used as the crude extract.
Step 2. Protamine Precipitation-Proteoglycans contained in the crude extract could be removed as insoluble complexes with protamine sulfate. Protamine sulfate was added to the crude extract with continuous stirring to a final concentration of 0.80 mg/mol uronic acid. The precipitates formed were removed by centrifugation at 100,000 ϫ g for 30 min. The apparent enzyme activity was markedly increased after protamine precipitation. The degree of the purification in each step was thus expressed as the relative value of the specific activity of each fraction compared with the value of the protamine fraction (Table I).
Step 3. Chromatography on Matrex Gel Red A-One-fifth volume of the clear solution from step 2 was applied to a Matrex gel red A column (2.7 ϫ 32.5 cm) equilibrated with 0.5 M NaCl in buffer A. The column was washed with 2000 ml of buffer A containing 0.5 M NaCl and was eluted with 3000 ml of buffer A containing 1 M NaCl. Most of GalNAc4S-6ST activity was eluted in the 1 M NaCl fractions. The fractions of the eluate containing GalNAc4S-6ST activity were pooled and dialyzed against 0.05 M NaCl in buffer A. This chromatography was repeated 5 times, and the dialyzed fractions were pooled.
Step 4. First Chromatography on Heparin-Sepharose CL-6B-Half of the dialyzed solution from step 3 was applied to a column of heparin-Sepharose CL-6B (2.2 ϫ 26.5 cm) equilibrated with 0.05 M NaCl in buffer A. The column was washed with 1000 ml of 0.05 M NaCl in buffer A. The adsorbed materials were eluted with a linear gradient formed from 500 ml each of 0.05 M and 1 M NaCl in buffer A. The fractions containing the sulfotransferase activity were pooled, dialyzed against 0.05 M NaCl in buffer B (10 mM Tris-HCl, pH 7.2, 10 mM 2-mercaptoethanol, 0.5% Triton X-100, 5% glycerol). This chromatography was repeated twice, and the dialyzed fractions were pooled.
Step 5. First Chromatography on 3Ј,5Ј-ADP-Agarose-The dialyzed fraction from Step 4 was applied to a 3Ј,5Ј-ADP-agarose column (1.2 ϫ 8.5 cm) equilibrated with buffer B containing 0.05 M NaCl. The column was washed with 150 ml of buffer B containing 0.05 M NaCl. The sulfotransferase activity was eluted with a linear gradient formed from 150 ml each of 0.05 M and 5 M NaCl in buffer B. After the gradient elution, the column was eluted with 150 ml of 5 M NaCl in buffer B. The fractions containing sulfotransferase activity (indicated by a horizontal bar in Fig. 2) were pooled and dialyzed against 0.05 M NaCl in buffer B.
Step 6. Second Chromatography on 3Ј,5Ј-ADP-Agarose-The dialyzed solution from step 5 was applied to a 3Ј,5Ј-ADP-agarose column. The column size and the conditions for the elution were the same as those of the first 3Ј,5Ј-ADP agarose chromatography. The second 3Ј,5Ј-ADP agarose fraction was dialyzed against 0.05 M NaCl in buffer A.
Step 7. Second Chromatography on Heparin-Sepharose CL-6B-The dialyzed solution from Step 6 was applied to a column of heparin-Sepharose CL-6B (0.9 ϫ 1.4 cm) equilibrated with 0.05 M NaCl in buffer A. The adsorbed sulfotransferase was eluted with 15 ml of buffer A containing 1.0 M NaCl. The collected fraction was dialyzed against buffer A containing 0.05 M NaCl. This step was adopted for the concentration of the enzyme solution. The purified enzyme was stored at Ϫ20°C.
Assay of Sulfotransferase Activity-GalNAc4S-6ST activity was assayed by the method described previously (24,32). The standard reaction mixture contained, in a final volume of 50 l, 2.5 mol of imidazole-HCl, pH 6.8, 1 mol of CaCl 2 , 1 mol of reduced glutathione, 25 nmol (as galactosamine) of CS-A, 50 pmol of [ 35 S]PAPS (about 5.0 ϫ 10 5 cpm), and enzyme. The reaction mixtures were incubated at 25°C for 20 min, and the reaction was stopped by immersing the reaction tubes in a boiling water bath for 1 min. After the reaction was stopped, 35 S-labeled glycosaminoglycans were isolated by precipitation with ethanol followed by gel chromatography with a Fast Desalting Column as described previously (32), and radioactivity was determined. One unit of enzyme activity is defined as the amount required to catalyze the transfer of 1 pmol of sulfate per min. For determining the activity for various glycosaminoglycans, CS-A was replaced with 25 nmol (as galactosamine for chondroitin sulfate and DS or glucosamine for heparan sulfate, CDSNS-heparin, and keratan sulfate) of glycosaminoglycans. For determining the position of sulfate transferred to CS-A and DS, 35 S-labeled glycosaminoglycans were digested with chondroitinase ACII, chondroitinase ABC, chondroitinase ACII plus chondro-6-sulfatase, or chondroitinase ABC plus chondro-6-sulfatase. The radioactive products formed after the enzymatic digestion were separated with HPLC using a Whatman Partisil-10 SAX column as described below, and 35 S radioactivity was determined. When oligosaccharides were used as acceptors, the reaction mixtures were applied directly to the Superdex 30 column as described below, and the 35 S-labeled oligosaccharides were separated from [ 35

S]sulfate and [ 35 S]PAPS.
Digestion with ␤-Glucuronidase, Chondroitinase ACII, Chondroitinase ABC, and Chondro-6-sulfatase-Digestion with ␤-glucuronidase was carried out for 3.5 h at 37°C in a reaction mixture containing, in a final volume of 40 l, 35 S-labeled or -unlabeled tetrasaccharide (40 nmol as galactosamine), 2 mol of sodium acetate buffer, pH 4.5, 20 nmol of 2-acetamido-2-deoxy-D-galactonic acid-1,4-lactone, 0.8 mol of sodium fluoride, and 18 units of ␤-glucuronidase. Under these conditions, removal of the nonreducing terminal GlcA was complete, and no release of inorganic sulfate was observed. Digestion with chondroitinase ACII or chondroitinase ABC was carried out for 4 h at 37°C in the reaction mixture containing, in a final volume of 25 l, 35 S-labeled glycosaminoglycans, 1.25 mol of Tris acetate buffer, pH 7.5, 2.5 g of bovine serum albumin, and 50 milliunits of chondroitinase ACII or chondroitinase ABC. After the reaction with chondroitinase ACII or chondroitinase ABC was over, the reaction mixtures were immersed in a boiling water bath for 1 min. Chondro-6-sulfatase (75 milliunits) was added to the reaction mixtures, and the incubation at 37°C was continued for a further 30 min for the complete degradation of ⌬Di-diS E, or for 5 h for the complete degradation of GalNAc (4,6-bis-SO 4 ).
SDS-Polyacrylamide Gel Electrophoresis-Polyacrylamide gel electrophoresis of proteins in SDS was carried out on 10% polyacrylamide gels under reducing or nonreducing conditions as described (33). Protein bands were detected by silver stain or Coomassie Brilliant Blue.
Gel Chromatography of the Sulfotransferase on Toyopearl HW-55-A Toyopearl HW-55 column (1.4 ϫ 96 cm) was equilibrated with a buffer containing 2 M NaCl, 10 mM Tris-HCl, pH 7.2, 20 mM MgCl 2 , 2 mM CaCl 2 , 0.1% Triton X-100, and 20% glycerol. Fractions (1.2 ml) were collected at a flow rate of 2.0 ml/h. Superdex 30 Chromatography and HPLC-A Superdex 30 16/60 column was equilibrated with 0.2 M NH 4 HCO 3 and run at a flow rate of 1 ml/min. 1-ml fractions were collected. Separation of the degradation products formed from 35 S-labeled glycosaminoglycans was carried out by HPLC using a Whatman Partisil-10 SAX column (4.6 mm x 25 cm) equilibrated with 10 mM KH 2 PO 4 . The column was developed with 10 mM KH 2 PO 4 for 10 min followed by a linear gradient from 10 to 450 mM KH 2 PO 4 as indicted in Fig. 8. Fractions (0.5 ml) were collected at a flow rate of 1 ml/min and a column temperature of 40°C.
Protein Assay-Protein was determined by the method of Bradford using bovine serum albumin as a standard (34). Protein assay reagent was obtained from Bio-Rad. Protein concentration of the second heparin-Sepharose CL-6B fraction was too low to be determined directly; therefore, samples were concentrated as described previously (32).
Other Methods-The galactosamine and glucosamine contents of glycosaminoglycans and oligosaccharides were determined by the Morgan-Elson method as modified by Strominger et al. (35) after hydrolysis with 6 M HCl at 100°C for 4 h. Uronic acid was determined by the method of Bitter and Muir (36).

Preparation of Crude Extract and Removal of Proteoglycans
from the Crude Extract-When sliced squid cartilage was homogenized with a Polytron homogenizer, addition of sodium chloride to buffer A at a final concentration of 0.5 M augmented the yield of the activity by 1.35-fold. A clear supernatant solution (crude extract) obtained after centrifugation of the homogenate contained a significant amount of proteoglycan as judged from the content of uronic acid (about 7 mol/ml as glucuronic acid). Most of the proteoglycans contained in the crude extract were removed as insoluble protamine-proteoglycan complexes by centrifugation after addition of protamine sulfate. In contrast, the sulfotransferase activity was markedly increased after the removal of proteoglycans (Fig. 1, Table I).
Purification of GalNAc4S-6ST by Affinity Chromatography-GalNAc4S-6ST was purified to apparent homogeneity about 20,000-fold over the specific activity of the supernatant fraction after protamine precipitation. Table I summarizes the purification of the sulfotransferase from 1750 g of squid cartilage.
As observed in most of the glycosaminoglycan sulfotransferases, GalNAc4S-6ST was adsorbed to heparin-Sepharose CL-6B. GalNAc4S-6ST was eluted at about 0.7 M NaCl from the column. The NaCl concentration was much higher than the concentration at which C6ST and C4ST were eluted from the same column. On 3Ј,5Ј-ADP-agarose chromatography, GalNAc4S-6ST was eluted at about 5 M NaCl (Fig. 2). The concentration of NaCl required for the elution of GalNAc4S-6ST from 3Ј,5Ј-ADP-agarose was much higher than that required for the elution of glucosaminyl 3-O-sulfotransferase (37). Purification of GalNAc4S-6ST was achieved after rechromatography on 3Ј,5Ј-ADP-agarose. The second heparin-Sepharose CL-6B fraction was used for the later experiments on the purified GalNAc4S-6ST.
Purity of GalNAc4S-6ST-The different fractions in the purification of GalNAc4S-6ST were separated by SDS-PAGE under nonreducing conditions and stained with silver nitrate (Fig.  3, lanes 1-6). A broad protein band of 63 kDa was the predominant band in the second heparin-Sepharose CL-6B fraction (lane 6). When the second heparin-Sepharose CL-6B fraction was treated with sample buffer containing 2-mercaptoethanol and was stained with Coomassie Brilliant Blue, the mobility of the protein was not altered (lane 9). The second heparin-Sepharose CL-6B fraction also showed a broad band around the origin of the gel (lane 6). This band appeared to be protein aggregates formed under the nonreducing conditions because this band became weaker under the reducing conditions when stained with Coomassie Brilliant Blue (lane 9) or silver nitrate (data not shown). The protein bands of 63 kDa disappeared after N-glycosidase F digestion, and new protein bands of 38 and 40 kDa appeared (lane 7), indicating that the purified protein contained N-linked oligosaccharides. To confirm that the protein bands observed in SDS-PAGE corresponded to GalNAc4S-6ST, the purified GalNAc4S-6ST was applied to a Toyopearl HW-55 column, and elution profiles of the GalNAc4S-6ST activity and protein were determined (Fig. 4). The protein band of 63 kDa appeared almost exclusively when the peak fraction (tube number 40 in Fig. 4A) was subjected to SDS-PAGE before N-glycosidase F digestion (Fig. 4B). When the peak fraction was subjected to SDS-PAGE after N-glycosidase F digestion, protein bands of 38 and 40 kDa appeared almost exclusively (Fig. 4C). A protein band with slightly higher molecular weight was observed when the fraction 39 was subjected to SDS-PAGE before N-glycosidase F digestion (Fig. 4B); however, the heterogeneity in the molecular weight observed before N-glycosidase F digestion seemed to be caused by microheterogeneity of N-linked oligosaccharides attached to the enzyme protein because only protein bands of 38 and 40 kDa were also detected in fraction 39 after N-glycosidase F digestion (Fig. 4C). A weak band of 54 kDa was observed in Fig.  4, B and C, but this band was thought to be an impurity that was artificially brought into the sample during the Toyopearl HW-55 chromatography, because this band was not observed in the purified fraction (Fig. 3, lanes 6, 7, 9). Molecular mass of the peak fraction determined from the elution position in the Toyopearl HW-55 chromatography (Fig. 4A) was 66 kDa, which agreed well with molecular mass determined by SDS-PAGE. These results suggest that GalNAc4S-6ST may behave as a monomer. The sulfotransferase activity toward DS and the sulfotransferase activity toward CS-A were eluted from this column at exactly the same position, suggesting that sulfation of CS-A and DS may be catalyzed by the same protein.
Properties of the Purified GalNAc4S-6ST-The pH optimum for the GalNAc4S-6ST was around 6.2. This value is similar to the pH optimum of C6ST (38) but slightly lower than that of C4ST (39). The apparent K m for PAPS was 5.0 ϫ 10 Ϫ7 M; this value is the same order of magnitude as the K m values for C6ST (38) and C4ST (39). The K m values for CS-A and DS (expressed as the concentration of galactosamine) were 1.1 ϫ 10 Ϫ6 M and 1.3 ϫ 10 Ϫ7 M, respectively. GalNAc4S-6ST was activated 2.5 to 3.5-fold with various divalent ions such as Mn 2ϩ , Mg 2ϩ , Ca 2ϩ , Sr 2ϩ , Ba 2ϩ , and Co 2ϩ at 20 mM. About 2-fold activation was observed in the presence of 0.1 M NaCl, 0.1 M KCl, or 0.15 mg/ml protamine. Reduced glutathione caused 33% stimulation at 20 mM. Other sulfhydryl compounds such as dithiothreitol and 2-mercaptoethanol showed only minimal effects on GalNAc4S-6ST activity.
Sulfation of Glycosaminoglycans with GalNAc4S-6ST-For determining acceptor specificity, the purified GalNAc4S-6ST was incubated with different glycosaminoglycans in the presence of varying amount of CaCl 2 . The disaccharide composition of these glycosaminoglycans are shown in Table II. Fig. 5 shows that the purified GalNAc4S-6ST was able to transfer sulfate to CS-A, CS-C, and DS. A low activity was observed toward squid skin chondroitin. CS-E, keratan sulfate, and heparan sulfate, and CDSNS-heparin did not serve as acceptor. The optimum concentration of CaCl 2 was 20 mM for CS-A and CS-C, but 100 FIG. 1. Removal of proteoglycans from the crude extract with protamine sulfate. To the crude extract from the squid cartilage, different amounts of protamine sulfate were added. The precipitates formed were removed by centrifugation at 10,000 ϫ g for 10 min, and GalNAc4S-6ST activity (q) and uronic acid contents (E) of the supernatant fractions were determined as described under "Experimental Procedures." mM for DS. When two CS-C preparations were compared, CS-C from shark cartilage showed higher activity than CS-C from human meniscus. The ratios of GlcA␤1-3GalNAc(4-SO 4 )/ GlcA␤1-3GalNAc(6-SO 4 ) of CS-C from shark cartilage and CS-C from human meniscus were 0.49 and 0.17, respectively (Table II). The lower ratio may result in the lower acceptor activity. The incorporation into CS-A was markedly inhibited by CS-E. In the presence of an equal amount of CS-E, incorporation into CS-A was decreased to 35% of control (indicated by an open rhombus in Fig. 5). Sulfation of CS-A with the purified GalNAc4S-6ST was inhibited by dermatan sulfate in a dosedependent manner (Fig. 6A) and sulfation of DS was also inhibited by CS-A (Fig. 6B) suggesting that the sulfation of both CS-A and DS are catalyzed by the same enzyme or at least the same catalytic site.
To determine the position of the sulfate groups transferred to CS-A and DS, we digested 35 S-labeled glycosaminoglycans with chondroitinase ACII or chondroitinase ABC and analyzed the digestion products by Superdex 30 gel chromatography (Fig. 7) and Partisil-10 SAX HPLC (Fig. 8). When the 35 S-labeled CS-A was digested with chondroitinase ACII, the major radioactive peak was detected at the position of ⌬Di-diS E (Figs. 7B and 8A). On the other hand, the 35 S-labeled DS was hardly depolymer-

FIG. 2. Affinity chromatography on 3,5-ADP-agarose.
The fractions eluted from the first heparin-Sepharose CL-6B were pooled, dialyzed, and applied to a 3Ј,5Ј-ADP-agarose column as described under "Experimental Procedures". After the column was washed with buffer B containing 0.05 M NaCl, GalNAc4S-6ST was eluted with a linear gradient of 0.05-5 M NaCl in buffer B. After gradient elution, the column was eluted with 5 M NaCl in buffer B. GalNAc4S-6ST activities (q) and the protein concentrations (E) of each fraction were assayed. The fractions indicated by a horizontal bar were pooled and used for further purification. ized by chondroitinase ACII digestion (Fig. 7D). When the 35 S-labeled DS was digested with chondroitinase ABC, major radioactivity was detected at the position of ⌬Di-diS E in Superdex 30 chromatography (Fig. 7E) and Partisil-10 SAX HPLC (data not shown). To determine which sulfate group of ⌬Di-diS E had 35 S radioactivity, the degradation products obtained after chondroitinase ACII were further digested with chondro-6-sulfatase and subjected to SAX HPLC (Fig. 8B). After digestion with chondro-6-sulfatase, 35 S radioactivity was shifted to the position of inorganic sulfate. Essentially the same results were obtained from 35 S-labeled DS after chondroitinase ABC and chondro-6-sulfatase digestion (data not shown). These results clearly indicate that 35 S-labeled SO 4 was transferred to position 6 of GalNAc(4-SO 4 ) residues in GlcA␤1-3GalNAc(4-SO 4 ) units of CS-A or IdoA␣1-3GalNAc(4-SO 4 ) units of DS. When the 35 S-labeled CS-A was digested with chondroitinase ACII, small amounts of radioactivity were detected at the position of GalNAc (4,6-bis-SO 4 ) (Fig. 8A). The radioactivity observed at the position of GalNAc (4,6-bis-SO 4 ) did not disappear after digestion with chondro-6-sulfatase for 30 min (Fig. 8B) but disappeared after digestion for 5 h (data not shown).
Squid skin chondroitin served as acceptor for the purified GalNAc4S-6ST. Because squid skin chondroitin was mainly composed of GlcA␤1-3GalNAc units and the contents of 4-sul-fated disaccharide units were about 0.2% of total repeating disaccharide units (Table II), the sulfate group transferred to squid chondroitin was expected to be located at position 6 of GalNAc residues. But, when the sulfated products were digested with chondroitinase ACII and subjected to SAX-HPLC, 35 S radioactivity was detected in GalNAc(4,6-bis-SO 4 ) and ⌬Di-diS E , and no radioactivity was observed in ⌬Di-6S or GalNAc(6-SO 4 ) (Fig. 8C).
Sulfation of 4-Sulfated Oligosaccharides with GalNAc4S-6ST-We determined whether 4-sulfated trisaccharide, tetrasaccharide, or pentasaccharide could serve as acceptor. To investigate the purity of the 4-sulfated oligosaccharides used as acceptor, we digested the oligosaccharides with chondroitinase ACII and subjected the digest to SAX-HPLC. For analysis of the 4-sulfated tetrasaccharide, the tetrasaccharide was digested with ␤-glucuronidase before digestion with chondroiti-

FIG. 5. Incorporation of [ 35 S]sulfate from [ 35 S]PAPS into glycosaminoglycan acceptors by the purified GalNAc4S-6ST.
Incorporation into the polysaccharide fraction was determined as described under "Experimental Procedures" except that CS-A was replaced with various glycosaminoglycans (25 nmol as galactosamine or 25 nmol as glucosamine) and the concentration of CaCl 2 was varied. CS-A (q), CS-C from shark cartilage (E), CS-C from human meniscus (OE), DS (‚), squid skin chondroitin (Ⅺ), and CS-E (f) were used as acceptors. Incorporation into keratan sulfate, heparan sulfate, and CDSNS-heparin was below the incorporation into CS-E. Incorporation into CS-A in the presence of 25 nmol of CS-E (as galactosamine) was determined at 20 mM CaCl 2 (᭛).

FIG. 6. Inhibition of sulfation of CS-A with DS and inhibition of sulfation of DS with CS-A.
The sulfotransferase reaction was carried out as described under "Experimental Procedures" using various concentrations of CS-A and DS, which are indicated under the graph. The concentrations of these glycosaminoglycans were expressed as the concentration of galactosamine. After the sulfotransferase reaction, 35 S-labeled glycosaminoglycans were isolated and digested with chondroitinase ACII or chondroitinase ABC. To the digests, 3 volumes of ethanol containing 1.3% potassium acetate was added, and the mixtures were centrifuged. 35 S radioactivity, which became soluble in ethanol after chondroitinase ACII or chondroitinase ABC digestion was determined. A, the amounts of sulfate transferred to chondroitin sulfate A were calculated from the radioactivity that became soluble in ethanol after chondroitinase ACII. B, the amounts of sulfate transferred to dermatan sulfate were calculated from the difference between the radioactivity that became soluble in ethanol after chondroitinase ABC and the radioactivity that became soluble in ethanol after chondroitinase ACII. Bars indicate mean Ϯ S.D. of three determinations. nase ACII (Fig. 9). The elution profiles detected by the absorption at 210 nm showed that after chondroitinase ACII digestion the trisaccharide (Fig. 9B), the tetrasaccharide treated with ␤-glucuronidase (Fig. 9C) and the pentasaccharide (Fig. 9D) gave GalNAc(4-SO 4 ) and ⌬Di-4S as main products, although a small peak of ⌬Di-6S (Fig. 9, B and D) and GalNAc(6-SO 4 ) (Fig.  9D) were detected. The molar ratio of monosaccharides/disaccharides was calculated from the elution profiles (Table III). These data appear to confirm that the length of these oligosaccharides. GalNAc4S-6ST was found to catalyze the efficient sulfation of 4-sulfated oligosaccharides. The rates of sulfation of trisaccharide, tetrasaccharide, and pentasaccharide were 90, 94, and 87%, respectively, of the rate of sulfation of CS-A under the standard assay conditions. 35 S-Labeled trisaccharide, tetrasaccharide, and pentasaccharide were eluted from the Superdex 30 column at 80, 77, and 72 min, respectively, in the symmetrical peaks. These elution times were 2-3 min earlier than the elution time of the respective acceptor oligosaccharides (data not shown). We showed previously that monosulfated sialyl N-acetyllactosamine trisaccharide was eluted from this column 3 min earlier than sialyl N-acetyllactosamine trisaccharide (40). From the comparison of the elution time between the 35 S-labeled products and the respective acceptor oligosaccharides, only one sulfate group appeared to be transferred to each oligosaccharide acceptor. When the 35 S-labeled trisaccharide and 35 S-labeled pentasaccharide were digested with chondroitinase ACII and applied to SAX-HPLC, the radioactivity appeared at the position of GalNAc(4,6-bis-SO 4 ) and ⌬Di-diS E . The proportion of GalNAc(4,6-bis-SO 4 ) was 27 and 40% of the total radioactivity for 35 S-labeled trisaccharide and 35 S-labeled pentasaccharide, respectively (Fig. 10, A and C). When 35 S-labeled tetrasaccharide was digested with chondroitinase ACII after ␤-glucuronidase digestion, the radioactivity appeared mainly at the position of ⌬Di-diS E with a small peak at the position of GalNAc(4,6-bis-SO 4 ); the proportion of GalNAc(4,6-bis-SO 4 ) was 11% of the total radioactivity. These observations suggest that the rate of sulfation of the penultimate GalNAc(4-SO 4 ) residue of the tetrasaccharide is lower than the rate of sulfation of nonreducing terminal GalNAc(4-SO 4 ) residue of the trisaccharide. To establish the position to which 35 S-labeled SO 4 was transferred, we digested 35 S-labeled trisaccharide with chondro-6-sulfatase after digestion with chondroitinase ACII. Although GalNAc(4,6-bis-SO 4 ) was refractory to chondro-6-sulfatase digestion when ⌬Di-diS E was completely converted to ⌬Di-4S (Fig. 11B), prolonged digestion with chondro-6-sulfatase allowed nearly quantitative conversion of GalNAc(4,6-bis-SO4) to GalNAc(4-SO 4 ) (Fig. 11C). When 35 S-labeled trisaccharide was subjected to the prolonged digestion with chondro-6-sulfatase after chondroitinase ACII digestion, the radioactivity of not only ⌬Di-diS E but also Gal-NAc(4,6-bis-SO4) disappeared and was shifted to the position of inorganic sulfate (Fig. 10B). These results clearly indicate that GalNAc4S-6ST is able to transfer sulfate to position 6 of both reducing and nonreducing terminal GalNAc(4-SO 4 ) residues of the 4-sulfated trisaccharide. The K m value expressed as the concentration of galactosamine for the trisaccharide was 1.6 ϫ 10 Ϫ5 M. When affinities for reducing-end GalNAc(4-SO 4 ) and nonreducing-end GalNAc(4-SO 4 ) were determined separately from the radioactivity of GalNAc(4,6-bis-SO 4 ) and ⌬Di-diS E , respectively, formed after chondroitinase ACII digestion, there was no significant difference in the affinity between the reducing and nonreducing terminal GalNAc(4-SO 4 ) residue of the trisaccharide (data not shown). The K m value for the tetrasaccharide could not be determined because the Lineweaver-Burk plot did not show a straight line. DISCUSSION In the present study, we have purified GalNAc4S-6ST to apparent homogeneity. The purified sulfotransferase showed two protein bands on SDS-PAGE after N-glycosidase F digestion as observed in C6ST (32), C4ST (39), heparan sulfate 6-sulfotransferase (41), and heparan sulfate 2-sulfotransferase (42). Although the identity of the two protein bands is not clear at present, it is possible that both protein bands correspond to GalNAc4S-6ST because the protein bands were both coeluted with GalNAc4S-6ST activity from Toyopearl HW-55. GalNAc4S-6ST activity was increased after removal of endogenous proteoglycans. The observed increase in the sulfotransferase activity after the protamine precipitation may be due mainly to the removal of proteoglycans with inhibitory activity to the sulfotransferase because the purified GalNAc4S-6ST was strongly inhibited with CS-E from squid cartilage (Fig. 5). Affinity chromatography on 3Ј,5Ј-ADP-agarose was critical for the purification. The concentration of NaCl required for the elution of the enzyme from 3Ј,5Ј-ADP-agarose was very high, suggesting that the affinity of GalNAc4S-6ST for 3Ј,5Ј-ADPagarose is higher than for any other glycosaminoglycan sulfotransferases so far purified.
GalNAc4S-6ST activity was activated by Ca 2ϩ ; the optimum concentration of Ca 2ϩ was 20 mM for CS-A and 100 mM for DS. , ⌬Di-4S (4), GalNAc(4,6-bis-SO 4 ) (5). A small peak observed at 17 min seems to be derived from the elution buffer because this peak was also observed in the elution profile of the standard materials. Another small peak observed in the digests of trisaccharide (25.5 min in B and C) was not identified.

TABLE III
Analysis of 4-sulfated oligosaccharides used for acceptors The purified trisaccharide and pentasaccharide (40 nmol as galactosamine) were digested with chondroitinase ACII. The purified tetrasaccharide (60 nmol as galactosamine) was digested with ␤-glucuronidase, purified with Superdex 30 chromatography, and then digested with chondroitinase ACII. The digested materials were subjected to SAX-HPLC. Monosaccharides and unsaturated disaccharides were monitored by absorption at 210 nm (Fig. 9). From the elution profiles shown in Fig. 9, composition of monosaccharides and unsaturated disaccharides were determined. At 210 nm, the observed ratio (molecular absorption of monosaccharides)/(molecular absorption of unsaturated disaccharides) was 0.32. When digested materials were monitored at 232 nm, unsaturated disaccharides other than ⌬Di-4S and ⌬Di-6S were not detected. a The possible reason why analytical data of these oligosaccharides were not identical is that tetrasaccharide used for acceptor and tetrasaccharide used for the preparation of trisaccharide acceptor were purified under different conditions. b Not detected.

Oligosaccharides
FIG. 10. HPLC separation of the degradation products obtained from 35 S-labeled oligosaccharides after digestion with chondroitinase ACII or chondroitinase ACII plus chondro-6sulfatase. 35 S-Labeled trisaccharide (A), 35 S-labeled pentasaccharide (C), and 35 S-labeled tetrasaccharide treated with ␤-glucuronidase (D) were digested by chondroitinase ACII and separated by SAX-HPLC. Another part of the 35 S-labeled trisaccharide was separated after digestion with chondroitinase ACII plus chondro-6-sulfatase (B). Chondro-6sulfatase digestion was carried out as described under "Experimental Procedures" for 5 h. Elution positions of the standard materials indicated by arrows are the same as those described in the legend to Fig. 8.
Although the reason that DS requires a higher concentration of Ca 2ϩ than CS-A is not clear, it is possible that the difference in the optimal concentration of Ca 2ϩ between DS and CS-A may reflect the different conformation of these glycosaminoglycans induced by the addition of Ca 2ϩ . Reduced glutathione was found to activate GalNAc4S-6ST activity as observed in C4ST (39) but the degree of the activation was much lower than that observed for C4ST.
Both CS-A and CS-C served as good acceptors. The incorporation of 35 S-labeled SO 4 into these glycosaminoglycans was closely related to the ratio of GlcA␤1-3GalNAc(4-SO 4 )/GlcA␤1-3GalNAc(6-SO 4 ); the lower ratio resulted in the lower activity, suggesting that GlcA␤1-3GalNAc(6-SO 4 ) unit may have an inhibitory effect on the activity of GalNAc4S-6ST. CS-E was not an acceptor but was an inhibitor for GalNAc4S-6ST, suggesting that GalNAc4S-6ST may be inhibited by the reaction products as suggested previously (24). DS was sulfated at IdoA␣1-3GalNAc(4-SO 4 ) units because 35 S-labeled DS was not degraded by chondroitinase ACII but was degraded by chondroitinase ABC. Such specificity of GalNAc4S-6ST contrasts clearly with the specificity of C4ST; C4ST preferentially sulfated position 4 of GalNAc residue contained in glucuronic acid-rich regions of desulfated dermatan sulfate (39,43). Sulfation of CS-A and DS seems to be catalyzed by the same enzyme because two activities were coeluted with Toyopearl HW-55, and the sulfation of CS-A and the sulfation of DS were inhibited by DS and CS-A, respectively. However, at present, the possibility remains that each of the two protein bands observed after N-glycosidase F digestion may have a preference for either of these glycosaminoglycans. The K m for DS was much smaller than the K m for CS-A. Glycosaminoglycans containing IdoA␣1-3GalNAc(4,6-bis-SO 4 ) units were reported to be present in hag fish notochord (18), rat glomeruli (16), and rat mesangial cells (17). Sulfotransferases involved in the biosynthesis of these glycosaminoglycans may have substrate specificity similar to that of GalNAc4S-6ST.
The purified GalNAc4S-6ST was shown to transfer sulfate efficiently to position 6 of nonreducing terminal GalNAc(4-SO 4 ) residues when a 4-sulfated trisaccharide and a 4-sulfated pentasaccharide were used as acceptors. However, sulfation of nonreducing terminal GalNAc(4-SO 4 ) residue appears to be suppressed by the substitution of a GlcA residue because the proportion of 35 S radioactivity of GalNAc(4,6-bis-SO 4 ) released from the 35 S-labeled tetrasaccharide after digestion with ␤-glucuronidase and chondroitinase ACII was much smaller than the proportion of the radioactivity of GalNAc(4,6-bis-SO 4 ) released from the 35 S-labeled trisaccharide or pentasaccharide. The K m for the trisaccharide was higher than the K m for CS-A by about ten-fold, indicating that molecular size of the acceptor may contribute to the affinity between the acceptors and the enzyme.
Squid skin chondroitin was sulfated, but sulfate was transferred to position 6 of internal or nonreducing terminal Gal-NAc(4-SO 4 ) residues, which were minor components of chondroitin. These results clearly indicate that GalNAc4S-6ST absolutely requires the presence of GalNAc(4-SO 4 ) residues for the activity. Proportion of ⌬Di-4S released after chondroitinase ACII digestion of squid skin chondroitin to the total unsaturated disaccharides was about 0.2%. The concentration of internal GalNAc(4-SO 4 ) residue in the reaction mixtures was thus calculated as 1 ϫ 10 Ϫ6 M. Because the K m for CS-A was 1.1 ϫ 10 Ϫ6 M, sulfation of GalNAc(4-SO 4 ) residues contained in chondroitin may also occur.
We have previously purified and cloned C6ST (44). The purified C6ST transferred sulfate to position 6 of GalNAc residues of chondroitin and Gal residues of keratan sulfate (38) or sialyl N-acetyllactosamine oligosaccharides (40). However, C6ST could not catalyze the sulfation of GalNAc(4-SO 4 ) residues even when CS-A was used as acceptor. In contrast, the purified GalNAc4S-6ST from squid cartilage could not transfer sulfate to the GalNAc residues of chondroitin.
At present, it is not clear whether sulfotransferases with the substrate specificity similar to the specificity of squid GalNAc4S-6ST are present in the vertebral tissues. GalNAc4S-6ST activity was reported in the human serum (26,27) and quail oviduct (25), but these sulfotransferases may be distinguished from the squid GalNAc4S-6ST in the substrate specificity because when chondroitin sulfate was used as acceptor, both human and quail enzymes transferred sulfate preferentially to position 6 of nonreducing terminal GalNAc(4-SO 4 ) residues, whereas the purified squid GalNAc4S-6ST transferred sulfate mainly to the internal GalNAc(4-SO 4 ) residues. Molecular cloning of the squid GalNAc4S-6ST would offer important clue about the molecular nature of the sulfotransferase involved in the biosynthesis of CS-E in the vertebral tissues.