J. Biochem., 79, 27-34 (1976)

Two Glycosulfatases from the Liver of a Marine Gastropod , Charonia lampas

Partial Purification and Properties1

Hiroshi HATANAKA, Yoko OGAWA , and Fujio EGAMI Mitsubishi-Kasei Institute of Life Sciences , Minamiooya, Machida-shi, Tokyo 194

Received for publication , July 9, 1975

Two glycosulfatases [EC 3.1 .6.3], I and ‡U, were purified 31 .3- and 33.9-fold respectively, from a crude extract of the liver of Charonia lampas . The purification was carried out by the following chromatographic procedures ; phosphocellulose , S ephadex G-150, Concanavalin A-Sepharose and isoelectric focussing . The preparations obtained were practically free from [EC 3.1.6.1] contami nation. Both glycosulfatases are probably glycoproteins differing in their carbohydrate moieties. The molecular weights of glycosulfatase I and H were estimated to be

about 112,000 and 79 ,000, respectively. They had the same optimum pH of 5.5, and the same Km value of 25.0mM for 6-sulfate .

As a result of recent investigations on sul yet been fully investigated. Glycosulfatase was fatase reactions, the physiological significance found first in molluscs by Soda and co-workers of arylsulfatase [EC 3. 1. 6. 1] is gradually being (10) and later in microorganisms, but its oc- elucidated. The hydrolysis of naturally oc currence in higher animals is doubtful (11). curring sugar-sulfate derivatives by arylsul In the present investigation, we studied fatase has been reported in the cases of sulfo the partial purification and some properties of galactolipids (1-5), UDP - N- acetylgalacto two glycosulfatases from the liver of a marine samine 4-sulfate (6, 7), and ascorbate 2-sulfate gastropod, Charonia lampas. (8, 9). Each derivative is apparently one of the physiological substrates of arylsulfatase. EXPERIMENTAL In spite of its name, glycosulfatase does not seem to participate in the hydrolysis of these Materials-D-Glucose 6-sulfate (potassium sugar-sulfates, as found in the case of Charonia salt) was purchased from Seikagaku Kogyo. lampas glycosulfatases [EC 3.1.6.3] (5). In p-Nitrophenyl sulfate (potassium salt) obtained any event, the glycosulfatase reaction, includ from Boehringer was recrystallized from 60% ing its enzymological characterization, has not aqueous ethanol. Glucose oxidase [EC 1.1.3.4] and standard proteins for molecular weight

1 This paper is No . ‡Y of the series " Ascorbate-2- estimation were purchased from Boehringer.

sulfate Sulfohydrolase." The preceding paper in Concanavalin A was obtained from Sigma. this series is Ref. 5. Eight other lectins were gifts from Prof. T.

Vol. 79, No. 1, 1976 27 28 H. HATANAKA, Y. OGAWA, and F. EGAMI

Osawa (University of Tokyo). Methyl ƒ¿-D- Enzyme Assays- Glycosulfatase activity

glucoside and methyl ƒ¿-D-mannoside were pur was usually assayed by measuring the liberated chased from Merck. Ascorbate 2-sulfate sulfate according to the method previously de (potassium salt) was synthesized by the method scribed (8). For particular purposes, the described previously (8, 12). amount of liberated glucose was measured by Other chemicals were of reagent grade. the use of glucose oxidase. Instead of termi Preparation of Glycosulfatases I and ‡U nation by the addition of 0.4ml of 0.15N HCl For the preparation of crude extract, the liver and 0.4ml of barium chloride-gelatin solution (hepatopancreas) (100g) of C. lam pas was (14), the reaction was terminated by putting homogenized as described previously (8). The the container in ice-water and adding 0.1ml dialyzed 105,000•~g supernatant was applied of 0.2M sodium-potassium phosphate buffer, to a phosphocellulose (Brown) column (2.65•~43 pH 5.8. An aliquot of the mixture was added cm, 237ml) equilibrated with 0.01M sodium to the preconditioned glucose oxidase reaction acetate-acetic acid buffer, pH 5.0. The frac mixture, in a final volume of 3ml, containing tions of the main peak containing both glyco 2ml of 0.2M sodium-potassium phosphate and arylsulfatase activities which buffer, pH 5.8, and 0.1mg of glucose oxidase eluted at 0.2M sodium chloride were pooled. protein (of fungal origin ; Boehringer grade I, These fractions were concentrated using an approx. 210U/mg). The rate of oxygen con Amicon model 402 pressure-filtration cell with sumption during the oxidation of glucose was a PM-10 membrane. The sample was applied measured with an oxygen probe (Yellow Spring to a Sephadex G-150 (Pharmacia) column (2.1•~ Inst., type YSI 5331). The reaction was car 132cm, 454ml). The fractions which eluted ried out at 37? with constant stirring.

just behind the void peak contained most of Arylsulfatase activity was measured by the the glycosulfatase and arylsulfatase activities, method described in previous papers (8, 15). forming partially overlapping peaks. They Molecular Weight Estimation-The method were collected together. of Andrews (16) was followed for the estima Concanavalin A-Sepharose (Pharmacia) col tion of the molecular weight of glycosulfatases umn chromatography was used for the separa I and ‡U by gel filtration. A column of Sepha tion of the two glycosulfatases. The gel was dex G-200 (1.25•~48cm, 58.9ml) eluted with

equilibrated with 0.02 M sodium acetate-acetic 0.05M Tris-HC1 buffer, pH 7 .5, containing 0.1 acid buffer, pH 5.5, containing 0.5M sodium M potassium chloride was used. The follow chloride, 0.5mM MnC12, and 0.5 mM CaC12, ing standards were utilized : horse heart cyto and packed in a small column (1.25•~8.5cm, chrome c, bovine serum albumin , yeast glucose- 10.4ml). The sample was adjusted to 0.5M 6 - phosphate dehydrogenase [EC 1. 1. 1. 49] , sodium chloride, 0.5mM MnC12, and 0.5mM rabbit muscle aldolase [EC 4 .1.2.13], beef liver CaC12, and applied at a flow rate of 3.0ml/hr. catalase [EC 1.11.1.6], and beef liver glutamate Fractions of 2ml were collected. Elution was dehydrogenase [EC 1.4 .1.3]. done with the same buffer containing 0.5M Precipitation of Glycosulfatases I and II methyl ƒ¿-D-glucoside and 0.1M methyl ƒ¿-D with Various Lectins - The precipitation of

mannoside. glycosulfatases with various lectins was carried For the separation of glycosulfatase ‡U and out at 4? in a screw-capped polycarbonate tube

arylsulfatase retained together on the Con (1.6•~7.6cm) using the following mixture ; 0.14 canavalin A-Sepharose gel, isoelectric focussing M Tris-acetic acid buffer , pH 7.5, 11mM CaC12, was done with an LKB model 8100 Ampholine 11mM MnC12, 0.11M sodium chloride , about unit. The ampholyte, isoelectric between pH 0.5mg/ml each of nine species of lectins , and 3.0 to 7.0, was used in a total volume of 110 16.1ƒÊg of protein/ml of glycosulfatase I or 10 .9ƒÊ Ml. g of protein/ml of glycosulfatase ‡U. The final Determination of Protein-The method of volume was 0.5ml in each case . After 1 day, Lowry et al. (13) was used routinely with bo the mixture was centrifuged at 105 ,000•~g for vine serum albumin as a standard. 60min using a Spinco type 50 Ti rotor . An

J. Biochem. C. lampas GLYCOSULFATASES 29 aliquot of the resulting supernatant was used graphic profile. The passed-through fractions for the determination of glycosulfatase activ contained about half of the glycosulfatase ac ity. The remainder of the supernatant and tivity, but no arylsulfatase activity. Arylsul precipitate were mixed again. These suspen fatase activity together with the rest of the sions were kept standing at 4? for 2 days more, glycosulfatase activity was retained on the -and were then centrifuged in the same man Concanavalin A-Sepharose gel. This result ner. The enzyme activity in the supernatant suggests that the two retained , i.e., was measured. arylsulfatase and half of the glycosulfatase, namely glycosulfatase ‡U, may be glycoproteins.

RESULTS The passed-through glycosulfatase, name ly glycosulfatase I, was subjected to a second Partial Purification of Glycosulfatases I Concanavalin A-Sepharose column chromatog and ‡U-After the phosphocellulose and Sepha raphy, in order to minimize arylsulfatase con dex G-150 column chromatographies, the sam tamination. The purification procedures for

ple was applied to a Concanavalin A-Sepharose glycosulfatase I are summerized in Table I. column. Figure 1 represents the chromato-

Fig. 1. Concanavalin A-Sepharase column chroma

tography of the eluate from the Sephadex G-150 column. The broken line with filled circles repre sents the absorbance at 280nm. Enzyme activities

of glycosulfatase and arylsulfatase, represented as solid lines with open triangles and open circles,

respectively, were determined as follows: The assay mixture for glycosulfatase, in a total volume

of 200 pl, contained 5ƒÊmoles of D-glucose 6-sulfate, 20ƒÊmoles of sodium acetate-acetic acid buffer, pH 5.7, and the enzyme. After reaction for 60 min at

37?, the resulting turbidity produced by adding barium chloride-gelatin solution was measured at

360nm. The assay mixture for arylsulfatase, in a total volume of 300ƒÊl, contained 3ƒÊmoles of p- nitrophenyl sulfate, 30ƒÊmoles of Tris-acetic acid

buffer, pH 7.5, and the enzyme. The reaction was carried out at 37? in a Gilford model 2400-2 Tecord

ing spectrophotometer. One unit of each enzyme represents 1ƒÊmole/min.

TABLE I. Purification of glycosulfatase I from the liver of C. lampas.

Vol. 79, No. 1, 1976 30 H. HATANAKA, Y. OGAWA, and F. EGAMI

Glycosulfatase I was purified 31.3 fold, and glycosulfatase I seemed to be attributable to was practically free from arylsulfatase. As p - nitrophenyl sulfate- hydrolyzing activity, mentioned below (shown in Fig. 3), the very which had an optimum pH of 5.0. small activity of arylsulfatase, measured at pH The glycosulfatase ‡U retained on the Con 7.5, contained in the purified preparation of canavalin A-Sepharose gel was subjected to isoelectric focussing to separate it from aryl sulfatase, as shown in Fig. 2. Glycosulfatase

(pI 6.3) was detected from fractions 85 to 95 and was separated from arylsulfatase activity with pI 5.0. For further minimization of the

glycosulfatase I and arylsulfatase contamina tion, the fractions with a large amount of

glycosulfatase ‡U were applied to a second Con canavalin A-Sepharose gel and a further iso electric focussing column. The purification

procedures for glycosulfatase ‡U are summarized in Table ‡U. Glycosulfatase ‡U was purified 33.9-fold, and was also practically free from arylsulfatase. As a byproduct of the above purification

procedure for glycosulfatases I and ‡U, we ob tained an arylsulfatase preparation practically free from glycosulfatases. The purified aryl sulfatase had 5.93 units total activity and 7.38 Fig. 2. Isoelectric focussing column chromatography units/mg protein specific activity. The extent of the fractions retained on concanavalin A-Sepharose of purification was 77.7-fold. However, this gel. Electrophoresis was carried out in 1.1% preparation had a slight hydrolytic activity for ampholite (pH 3.0-7.0) for 48 hr at 300volts in an glucose 6-sulfate at pH 5.5; specific activity, LKB 110ml column. Anode ; 1.4% H2SO4 at the 0.032 unit/mg protein. We could not decide column bottom, cathode ; 2.0% ethylenediamine. whether this activity might be attributable to Fractions of 1 ml were collected after electrophoresis and the pH was measured (represented as dots). glycosulfatase contamination or to very weak Enzyme activities were determined by the method activity of arylsulfatase against glucose 6-sul described in the legend to Fig. 1. fate.

TABLE II. Purification of glycosulfatase ‡U from the liver of C. lampas.

J. Biochem. C. lampas GLYCOSULFATASES 31

Effect of pH on the Enzyme Activities- were estimated by the gel filtration The pH dependence of glycosulfatase activity method. A plot of elution volume/void volume in the preparations of glycosulfatases I and ‡U against the logarithm of molecular weight is is presented in Fig. 3. Both showed an opti presented in Fig. 4. From this plot, values mum at pH 5.5. In these preparations, a weak of 112,000 and 79,000 for glycosulfatases I and hydrolyzing activity toward p-nitrophenyl sul II, respectively, were obtained. fate with an optimum pH of 5.0 was observed Effect of Concentration on the

(Fig. 3). The activity was not attributable to Activities of Glycosulfatases I and ‡U-Figure contamination by arylsulfatase, which had a 5 shows the effect of increasing concentration

pH optimum of 7.5. of glucose 6-sulfate on the two enzyme ac Molecular Weight Determination-The ap tivities. From Lineweaver-Burk (16) plots, proximate molecular weights of the two glyco- the same Km value, 25.0mM, was obtained for both glycosulfatases I and ‡U. Slightly dif ferent Vmax values, 5.88 and 6.54 units/mg of

protein were obtained for glycosulfatase I and ‡U , respectively. Comparison of Precipitation Effects due to the Addition of Various Lectins to Glycosul

fatases I and ‡U Glycosulfatase ‡U and aryl sulfatase had an affinity for concanavalin A, a jack bean agglutinin, as shown in Fig. 1. However, glycosulfatase I passed through con canavalin A-Sepharose gel. We examined the effects of various lectins on glycosulfatases I and ‡U. Table ‡V shows the results. Among nine species of lectins examined, concanavalin A, specific to mannopyranoside (17 ), and

Fig. 3. pH-Activity curves of glycosulfatase I and ‡U with glucose 6-sulfate and p-nitrophenyl sulfate

as substrates. The enzyme preparations used were

glycosulfatase I (above) and ‡U (below). The assay mixture with glucose 6-sulfate (solid line) as a sub strate was the same, except for the buffer, as that

described in the legend to Fig. 1. The assay Fig. 4. Molecular weight estimation of glycosul mixture with p-nitrophenyl sulfate (broken line) as fatases I and ‡U by gel filtration on a calibrated a substrate was also the same except for the buffer, Sephadex G-200 column. Glycosulfatase I and ‡U, but after reaction at 37? for 60min, 0.5ml of 1 N represented as horizontal arrows I and ‡U, respec NaOH was added and the resulting optical density tively, and the other standard proteins (1, cyto at 400nm was measured. These enzyme activities chrome c; 2, serum albumin; 3, glucose-6-phosphate were measured in sodium acetate-acetic acid buffer dehydrogenase ; 4, aldolase ; 5, catalase ; 6, glutamate (filled circles, pH 4.0-5.5), Tris-acetic acid buffer dehydrogenase) were applied to a Sephadex G-200 (half-filled circled, pH 6.0-7.5) and Tris-HCl buffer column. (open circles, pH 7.5-9.0).

Vol. 79, No. 1, 1976 32 H. HATANAKA , Y. OGAWA, and F. EGAMI

Ricinus communis agglutinin, specific to ga lactopyranoside (17, 18), had precipitating ac tivity toward glycosulfatases ‡U and I, respec

tively. This result suggests that glycosul fatases I and ‡U are glycoproteins which have different carbohydrate moieties. Effects of Various Compounds on the Ac tivities of Glycosulfatases I and ‡U-Phosphate, known to be an inhibitor of sulfatase reactions in general, was a potent inhibitor of glyco sulfatases I and ‡U, too. The concentrations of phosphate ions necessary for 50% inhibition were 1.2 and 0.75 M for glycosulfatase I and

‡U, respectively. Sulfate also inhibited these

Fig. 5. Effect of glucose 6-sulfate on glycosulfatase I and ‡U. Enzyme activities were determined by

the method described in the legend to Fig. 1. Glycosulfatase I, open circles ; glycosulfatase ‡U, closed circles.

TABLE ‡V. Precipitation effects of various lectins on the activities of glycosulfatases I and ‡U. The fol

lowing phytohemagglutinins, except for concanavalin A, were gifts of Prof. T. Osawa ; jack bean concana valin A, Glycine max (soybean) hemagglutinin, Bauhinia purpurea hemagglutinin, wheat-germ agglutinin,

Sophora japonica agglutinin, Arachis hypogoea (peanut) agglutinin, Ricinus communis (castor bean) agglutinin, Pisum sativum (garden pea) agglutinin, and Lens culinaris (common lentil) agglutinin. The precipitation

reaction was carried out at 4? in a solution, in a final volume of 0.5ml, containing 0.14M Tris-acetic buffer, pH 7.5, 11mM CaC12, 11mM MnC12, 0.11M sodium chloride, the indicated amount of each lectin,

and 8.04ƒÊg protein of glycosulfatase I or 5.46ƒÊg protein of glycosulfatase‡U.

J. Biochem. C. lampas GLYCOSULFATASES 33

activities. In this case, the assay was carried the preparation procedure. Most of the glyco out by determination of the liberated glucose sulfatase activity was retained on the phos using glucose oxidase. Five mM sulfate in phocellulose column in this paper, differing hibited 33.5% and 25.1% of the activity of from the previous result (8). glycosulfatase I and II, respectively. The activities of glycosulfatase I and ‡U Ascorbate 2-sulfate showed no inhibition were rather unstable during the purification of these glycosulfatases at a concentration of steps using concanavalin A-Sepharose and iso 6mM. electric focussing columns (shown in Tables I and ‡U). As can be seen in Table ‡V, both

DISCUSSION activities were also unstable at pH 7.5 in the presence of Ca2+ and Mn2+ cations. However, During the course of biodegradation studies of the purified and concentrated enzyme prepa ascorbate 2-sulfate, we first found that the pro rations were reasonably stable. No loss of these tein molecule expressing ascorbate-2-sulfate activities was observed during storage at 4? at sulfohydrolase activity might be identical with 0.02M sodium acetate-acetic acid buffer, pH that expressing arylsulfatase activity, but not 5.5, for at least two months. The substrate with that of glycosulfatase (8). Meanwhile, specificity of these glycosulfatases remains to the hydrolysis by arylsulfatase of other sugar be studied. sulfate derivatives, i.e., sulfated glycolipids With the purified preparations of glyco and UDP-N-acetylgalactosamine 4-sulfate, has sulfatases I and ‡U, a small hydrolytic activity been reported (1-7). On the other hand, the (about one order of magnitude lower than glyco physiological significance of the glycosulfatase sulfatase activity) was observed with p-nitr- is still unknown. As a first step to elucidat phenyl sulfate. It remains to be elucidated ing this, we tried to obtain glycosulfatase free whether this was due to contamination by an from arylsulfatase. arylsulfatase or due to the wobbly specificity Two species of glycosulfatase activities, of glycosulfatases. Taking into consideration namely I and ‡U, were purified 31.3- and 33.9- the changes of glycosulfatase activity and p- fold, respectively, from a crude extract of nitrophenyl sulfate-hydrolyzing activity (opt. Charonia lampas liver. These enzyme prepa pH 5.0) in the course of purification, we are rations were practically free from arylsulfatase. rather inclined to the latter view. Glycosulfatase I and ‡U are probably glycopro Sulfatases are classified by the nature of teins with a different carbohydrate moiety and the residue R in the substrate, R-OS03- : aryl different moleclar weights. However, the ca sulfatase, glycosulfatase, etc. However, the talytic activities of these two enzymes were substrate specificity of sulfohydrolases seems almost the same ; the same Km 25.0mM ; the not to be absolute, but relative. Thus aryl same optimum pH, pH 5.5. It remains to be sulfates such as p-nitrophenyl sulfate might be elucidated whether the difference between the slightly hydrolyzed by glycosulfatase. This two glycosulfatases is due to the different carbo cannot be attributed to " high sulfate transfer hydrate moiety only, or to the protein moiety potential" (19), because ascorbate 2-sulfate as well. with a similar " high sulfate transfer potential " The crude extract of C. lampas liver used was not hydrolyzed by the glycosulfatase prep- in this investigation had a larger amount of arations (5). glycosulfatase by a factor of about 30 than The authors wish to express their gratitude to Prof. that used in the previous paper (8). This T. Osawa (University of Tokyo) for a generous difference, observed in spite of the use of the supply of lectin samples. same batch of the liver, could not be explained. The present value of glycosulfatase content in REFERENCES the liver was reproducible. Perhaps the very small content of the crude extract might pre 1. Mehl, E. & Jatzkewitz, H. (1968) Biochim. Biophys. Acta 151, 619-627 viously have led to enzyme denaturation during

Vol. 79, No. 1, 1976 34 H. HATANAKA, Y. OGAWA, and F. EGAMI

2. Jerfy, A. & Roy, A.B. (1973) Biochim. Biophys. 10. Soda, T. (1936) J. For. Sc., Imp. Univ. Tokyo Acta 293, 178-190 Sec. 1, 3, 149-222 3. Yamato, K., Handa, S., & Yamakawa, T. (1974) 11. Roy, A.B. (1971) in The Enzymes (Boyer, P.D., J. Biochem. 75, 1241-1247 ed.) Vol. 5, pp. 1-19, Academic Press, New York 4. Fluharty, A.L., Stevens, R.L., Miller, R.T., & 12. Hatanaka, H., Ogawa, Y., & Egami, F. (1974) Kihara, H. (1974) Biochem. Biophys. Res. J. Biochem. 75, 851-866 Commun. 61, 348-354 13. Lowry, O.H., Rosebrough, N.J., Farr, A.L., & 5. Hatanaka, H., Ogawa, Y., Egami, F., Ishizuka, Randall, R.L. (1951) J. Biol. Chem. 193, 265-275 I., & Nagai, Y. (1975) J. Biochem. 78, 427-429 14. Dodgson, K.S. (1961) Biochem. J. 78, 312-319 6. Tsuji, M., Hamano, M., Nakanishi, Y., Ishihara, 15. Hatanaka, H., Ogawa, Y., & Egami, F. (1975) K., & Suzuki, S. (1974) J. Biol. Chem. 249, 879- J. Biochem. 77, 801-806 886 16. Andrews, P. (1965) Biochem. J. 96, 595-606 7. Fluharty, A.L., Stevens, R.L., Fung, D., Peak, 17. Sharon, N. & Halina, L. (1972) Science 177,. S., & Kihara, H. (1975) Biochem. Biophys. Res. 949-959 Commun. 64, 955-962 18. Irimura, T., Kawaguchi, T., Terao, T., & 8. Hatanaka, H., Ogawa, Y., & Egami, F. (1975) Osawa, T. (1975) Carbohyd. Res. 39, 317-327 J. Biochem. 77, 353-359 19. Bond, A.D., McClelland, B.W., Einstein, J.R., 9. Roy, A.B. (1975) Biochim. Biophys. Acta 377, & Finamore, F.J. (1972) Arch. Biochem. Biophys. 356-363 153, 207-214

J. Biochem.