Comp. Biochem. Physiol. Vol. 88B, No. I, pp. 313-316, 1987 0305-0491/87 $3.00+0.00 Printed in Great Britain © 1987 Pergamon Journals Ltd

COMPARATIVE STUDY OF SPECIFIC ~x-I,2-GLUCOSIDASE ACTIVITY TOWARD GLUCOSYL GALACTOSYL

HYDROXYLYSINE IN VARIOUS ANIMAL SPECIES

IKUKO ISHII, HITOO IWASE, HIDEAKI HAMAZAKI and KYOKO HOTTA Department of Biochemistry, School of Medicine, Kitasato University, Sagamihara, Kanagawa, 228,

Japan

(Received 10 November 1986)

Abstract--1. Assay of the activity of ~-1,2-glucosidase was completed within 10 min using reversed-phase high performance liquid chromatography and purified dansylated glucosyl galactosyl hydroxylysine as the substrate.

2. A comparative study was made on the enzyme activity of liver homogenate from eight animal species, mouse, frog, chicken, rabbit, pig, rat, human, bovine and that of a spinach leaf homogenate. ct-l,2-glucosidase activity in human and bovine liver was very low, and that of a-l,2-glucosidase could not be detected in the spinach homogenate as expected.

3. 1-deoxynojirimycin, a well known potent inhibitor of *t-l,2-glucosidases which act on the N- glycosidic type carbohydrate chain, also inhibited ~,-l,2-glucosidase acting specificity on glucosyl galactosyl hydroxylysine derived from the collagen molecule.

INTRODUCTION

Glucosyl galactosyl hydroxylysine (GGH) is a well known giycopeptide derived from collagen by alka- line treatment (Spiro, 1967). This type of carbo- hydrate chain is distributed widely in vertebrates and invertebrates (Spiro, 1969; Spiro and Fukushi, 1969; Katzman et aL, 1972) and found also in a com- plement, Clq (Shinkai and Yonemasu, 1979). In spite of its wide distribution, the biological significance of the disaccharide unit is still not known.

Recently Hamazaki and Hotta (1979) reported on the degradation enzyme, r,-l,2-glucosidase, which acts specifically on Glc-Gal-Hyl. This enzyme is of interest in the catabolic process of collagen since its action on collagen molecules possibly causes ex- posure of the galactose residue which functions as a signal for the degradation of serum glycoprotein (Kawasaki and Ashwell, 1976).

lwase et aL (1987) have devised a means for fractionating the dansyl derivative of the gly- copeptide, Glc-Gal-Hyl based on the method for fractionating the N-glycosidic type glycopeptide (Iwase et aL, 1983, 1984, 1985; Mutsaers et al., 1985). Mono-dansyl GGH (C~H-DNS) was fractionated into four peaks, designated as a, b, e and d, and di-dansyl GGH (GGH-DNS2) into two peaks, e and f by reversed-phase high performance liquid chro- matography (HPLC). The GGH-DNS fractions cor- responding to a and b were found to serve well as substrates for detecting specific ~-glucosidase activity in crude enzyme preparations.

In the present research, the method for assaying ~-glucosidase activity was further developed and a study was made of differences in enzyme activity according to species and age.

MATERIALS AND METHODS

Materials

The following compounds and materials were obtained commercially: dansyl chloride (5-dimethyl-amino-naphtha- lene-l-sulfonyl chloride), Pierce Chemical Company, USA; acetonitrile, l-butanol and acetic acid, Wako Pure Chemical Industries, Ltd., Japan; Whatmann 3MM paper, W and R Balston Ltd., UK. All solvents for HPLC were filtered through Toyo Roshi filter (pore size 0.45/zm). All other reagents were of the highest grade available and used without further purification.

High-performance liquid chromatography

The HPLC apparatus used consisted of a Hitachi solvent delivery system equipped with a Hitachi 655-71 Processor B, proportioning valve and Rheodyne sample injector 7125, an Inertsil ODS (5#m; 25cm × 4.6mm i.d.) protected by a Brownlee Lab MPLC guard column (30 × 4.6 mm i.d.) and a FP-210 (Japan Spectroscopic Co., Ltd.) fluorescence monitor. For the HPLC analysis, about 1 nmol of dansyl glycopeptide was injected into the column. Elution was then carried out at a flow rate of 1.0 ml/min using a mobile phase containing of 20% acetonitrile in 25 mM sodium borate buffer, pH 7.0. The effluent was monitored by fluorescence (excitation at 325 nm, emission at 530 nm).

Preparation of dansylated glucosyl galactosyl hydroxylysine as the enzyme substrate

Purified glucosyl galactosyl hydroxylysine (GIc-GaI-Hyl or GGH) from the alkaline degradation of edible medusae was dansylated. Briefly, the glycopeptide was dissolved in 0.2 M sodium bicarbonate to a concentration of 2.5 mg/ml. To this solution was added an equal volume of dansyl chloride solution (10 mg DNS-C1/mi acetone) and the reac- tion mixture was allowed to stand for 12hr at room temperature. The mixture was spotted on Whatmann 3MM paper and developed with the solvent, 1-butanol:acetic add:water = 3:1:1 (v/v/v). A broad yellowish fluorescence

313

314 IKUKO IsmI et al.

band near the origin was cut out from the paper and the fluorescent material was eluted from the paper strip with water and designated as GP-I. In a previous paper, the mono-dansyl derivative (GP-I) was fractionated into four components designated as a, b, c and d by HPLC, and component a was found the best substrate for the ~-l,2-glucosidase examined here. Consequently, this com- ponent was purified from the mono-dansyl derivative by HPLC to serve as the substrate using 16% aeetonitrile-water as the mobile phase according to the method we developed.

Preparation of crude enzyme Crude liver enzymes from frog (Bufo bufo), chicken

(Gallus domesticus), mouse (Mus musculus), rat (Rattus noruegicur), rabbit (Oryctolagus cuniculus), bovine (Bos taurus), pig (Sus scrota) and human were prepared by homogenizing the liver samples (20 g) in 50 mM phosphate buffer, pH 7.0 (80 ml) with a Polytoron homogenizer. A spinach leaf (40 g) was homogenized in 40 ml of the same buffer using a Warring blender. Each homogenate was centrifuged at 11,000 g for 20 min and the supernatant thus obtained was stored as an enzyme preparation at -20°C until use.

Assay of ~x-glucosidase activity A standard reaction mixture containing 20/zl of GGH-

DNS (a) (0.5/amol/ml), 20 #1 of 0.2 M acetate buffer, pH 5.0 and 20 #1 of the crude enzyme was incubated at 37°C for a specified period of time. Following incubation, the reaction was terminated by heating the mixture at 100°C for 5 rain. The insoluble material thus produced was removed by centrifugation at 8500g for 10min. An aliquot of the supernatant was then analyzed by HPLC.

A time course experiment was conducted as follows: 900/~1 of above standard mixture were prepared. During incubation of the mixture, 50/al aliquots were removed at 0, 15, 30, 45, 60, 90, 120, 180 and 240min. The human, bovine and spinach enzymes, were incubated for a pro- longed period of 14 hr. An assay of the enzyme activity in each aliquot was performed according to the procedure stated above.

RESULTS

Assay of ~-1,2-glucosidase activity by reversed-phase high performance liquid chromatography

A typical elution profile of GGH-DNS (substrate) and GH-DNS (product) obtained by the a-l,2-glucosidase assay is shown in F ig . 1. Crude enzyme from frog liver was used as a source of a-glucosidase since its activity was highest among all the animal enzymes examined. As shown in Fig. 1A, purified GGH-DNS was eluted with a single peak by HPLC. As a control experiment, a reaction mixture without a substrate was analyzed in the same manner (Fig. 1B). Fluorescence peaks from the enzyme itself appeared at 3.1 and 5.9 min and had no effect on the results of the enzyme assay since they both eluted faster from the column than those of the substrate (GGH-DNS) and product (GH-DNS) with retention times 7.4 and 9.4m in, respectively (Fig. 1C). En- zymes from other sources showed essentially the same patterns as those in Fig. 1B and had no effect on the results of the assay of activity. Enzyme activity was calculated following the assay method reported for endo-/~-N-acetylglucosaminidase (Iwase et al., 1981). On the basis of peak height, Hs and Hp of the substrate and product on the chromatogram,

hi U Z LIJ 0

h i n- O

. - I LL

h i

I-- < J h i E

A

Cl

LU

z

m

i

0 10

C

I

0 10

Ci

i , i

20 20 0 10 20

T I M E ( M I N )

Fig. l. Assay of *(-giucosidase activity by reversed-phase high performance liquid chromatography. Each sample was analyzed by HPLC using a mobile phase containing 20% acetonitrile in 25 mM borate buffer, pH 7.0. A: HPLC profile of substrate (GGH-DNS). Peak a in the figure indicated the substrate. B: HPLC profile of enzyme pre- pared from frog liver. C: HPLC profile of incubated stan- dard reaction mixture. Peak a and Pa in the figure indicated substrate (GGH-DNS) and product (GH-DNS), re-

spectively.

100 x Hp/ (Hs+ Hp) was calculated as percent of hydrolysis.

o~-glucosidase activity differences according to species and age and individual differences in the same species

Curde enzymes from five 7-week-old rats were prepared as described in Materials and Methods. ~-l,2-glucosidase activity determined for these ani- mals following incubation at 37°C for 1 hr was as per cent of hydrolysis 12.2, 12,0, 11.7, 11.9 and 11.5. At the same time, the effects of age on enzyme activity was also examined. The enzyme preparation from retired rats (18-month-old) used as a model of aging, showed slightly low (about 80% the control activity) • -glucosidase activity compared with that of 7-week-old rats (Fig. 2).

To examine variation in activity according to spe- cies, assays were conducted on frog, chicken, mouse, rat, rabbit, bovine, pig, human and spinach leaf. A time course experiment was conducted for each en- zyme by the method described in Materials and Methods. The enzymes appeared fairly stable during incubation since hydrolysis rate of the substrate remained constant until the hydrolysis exceeded 30% (Fig. 2). ~-l,2-Glucosidase activity from bovine and human liver were much lower than that in samples from the other animals. As expected, the crude enzyme from the spinach leaf failed to show any ~-l,2-glucosidase activity even with a prolonged in- cubation period of 14 hr, the same as the incubation time for the bovine and human liver enzymes. The percentage hydrolysis for human and bovine enzymes was 21.3 and 9.1%, respectively, following this period of incubation. Liver enzyme activity of the other

r,- 1,2-glucosidase activity 315

100

>,

-1-

_ _ 1 I ~S 1

0 60 120 )80 240 840

TIME (MIN)

Fig. 2. Time course of a-l,2-gtucosidase activity toward GGH-DNS in various animal species. Each sample was analyzed by HPLC under the same conditions as in Fig. 1. ~-l,2-Glucosidase activity was expressed as percent of hy- drolysis calculated as in the text. Symbols indicate mouse, &--&; frog, O- -O; chicken, V- -V; rabbit, A - - A ; pig, II--I1; rat, ~7--~; retired rat, ~7--~7; human, O- -Q;

bovine, I"l--I"q.

animals decreased in the order of mouse, frog, chicken, rabbit, pig and rat. The enzyme activity of the frog and mouse was particularly high, consti- tuting over 90% that of GH-DNS after 4hr of incubation.

DISCUSSION

The method for asssaying a-l,2-glucosidase was further developed for greater clarification of collagen catabolic processes. The complete separation of the product, GH-DNS from the substrate, GGH-DNS was acheived within 10 min by HPLC using a mobile phase that was 20% acetonitrile. The time required for analysis could be shortened by increasing the acetonitrile concentration. However, a concentration exceeding 25%, resulted in poorer separation as well as shortened column life.

Rats of the same age were used to detect individual differences in enzyme activity, following the same procedure that included homogenization and centrif- ugation for preparation of the crude enzyme. Re- producible results were always obtained. This is indication that the enzyme is possibly recovered in soluble form and not as a membrane-bound enzyme (Hamazaki and Hotta, 1979). The fairly low activity noted in the retired rats is possibly an indication of the relation of age to decrease in ~-glucosidase activity. Such a relationship is in good agreement with the report that collagen content in various tissues increases with age (Takeda et al., 1975). Differences in activity according to species may be summarized as follows: the smaller the animal, the higher is the ~-glucosidase activity, except in the case of the rat.

There is another ~-l,2-glucosidase which acts on the common glucosyl-l,2-glucose linkage of an inter- mediate produced in the biosynthesis of the N- glycosidic type carbohydrate chain. A spinach leaf certainly appears to contain this type of carbohydrate chain (Lehle, 1981), but not collagen molecules. As expected, no :~-l,2-glucosidase activity could be de-

tooted in the spinach leaf homogenate. The present results indicate the possible presence of two different kinds of ~-1,2-glucosidase---a soluble (Hamazaki and Hotta, 1979) and a membrane bound enzyme (Kilker et al., 1981) already reported, and different from each other. 1-Deoxynojirimycin, a potent inhibitor of the processing enzyme of the N-glycosidic type carbo- hydrate chain (Fuhrmann et al., 1985), also inhibited the activity of ~-glucosidase toward Glc-Gal-Hyl (Ki = 2.3 × 10-6).

Kawasaki and Ashwell (1976) reported that the non-reducing terminal galactose residue which func- tions as the signal for the degradation of serum glycoprotein, determines the life span of gly- coprotein. The sialic acid generally present at the non-reducing end of oligosaccharide was removed by the action of neuraminidase, thus exposing the galac- tose residue. The glycoprotein thus produced was immediately taken by the receptor on the surface of liver cell and degraded. Similarly, exposure of the galactose residue produced by the action of ~-glucosidase on Glc-Gal-Hyl may be an important step in the catabolic process of collagen.

It was previously reported that the quantitative measurement of the dansyl derivatives of glycosylated hydroxylysine in urine makes possible estimation of the metabolic turnover rate of collagen molecules (Moro et al., 1984), and that, on the basis of the measurement of Gal-Hyl/Glc-Gal-Hyl ratio in urine, the types of collagen molecules catabolized can be determined (Segrest et al., 1970). Differences in glu- cosidase activity according to species and age indicate that special care should be taken when investigating catabolic processes by analysis of glycosylated hy- droxylysine.

In the present research, it was clarified that ~-glucosidase activity exists in various animals but not in plants. This enzyme has also been reported in several organs of the rat (Hamazaki and Hotta, 1980). It is likely that this enzyme and the collagen molecule are present together. The optimum pH of 5.0 of this enzyme and its 90% recovery in soluble form as reported by other investigators are actually features of the lysosomal and cytoplasmic enzymes, respectively (Hamazaki and Hotta, 1979, 1980). Elu- cidation of the biological role of ~-l,2-glucosidase may possibly provide an explanation for this discrep- ancy.

Acknowledgements--We thank Professor Okudaira for kindly providing human liver. The authors express their sincere appreciation to Miss Urata for the technical assis- tance she rendered. This work was supported in part by a grant for special research from Kitasato University, School of Medicine, No. 83502 and by a Grant-in-Aid from the Japanese Ministry of Education, Science and Culture.

REFERENCES

Fuhrmann U., Bause E. and Ploegh H. (1985) Inhibitor of oligosaceharide processing. Biochim. biophys. Aeta 825, 95-110.

Hamazaki H. and Hotta K. (1979) Purification and charac- terization of an ~-glucosidase specific for hydroxylysine- linked disaccharide of collagen. J. biol. Chem. 254, 9682-9687.

316 IKUKO IsmI et al.

Hamazaki H. and Hotta K. (1980) Enzymatic hydrolysis of disaccharide unit of collagen-isolation of 2-O-g-D- glucopyranosyl- O-~8-D-galactopyranosyl- hydroxylysine glucohydrolase from rat spleens. Eur. J. Biochem. 111, 587-591.

Iwase H., Morinaga T., Li Y.-T. and Li S.-C. (1981) Analysis of endo-~-N-acetylglucosaminidase activity of high pressure liquid chromatography on a silica-based chemically bonded octadecyl column. Analyt. Biochem. 113, 93-95.

Iwase H., Li S.-C. and Li Y.-T. (1983) Fractionation of DNS-glycopeptides by reversed-phase high-performance liquid chromatography. J. Chromatogr. 267, 238-241.

Iwase H., Kato Y. and Hotta K. (1984) Comparative study of the carbohydrate chain of ovalbumin from various avian species by high pressure liquid chromatography. Comp. Biochem. Physiol. 77B, 743-747.

Iwase H., Kato Y. and Hotta K. (1985) Fractionation of peptic glycopeptides from chicken ovalbumin by reversed-phase high-performance liquid chromatography. J. Chromatogr. 320, 426-429.

Iwase H., Ishii I., Kato Y., Hamazaki H., Hotta K., Umezawa A. and Kanzaki T. (1987) Subfractionation of the dansylated derivatives of glucosyl galactosyl hydroxy- lysine by high performance liquid chromatography and its application to a specific ~-1,2-glucosidase assay. J. Chro- matogr. 416, 37-44.

Katzman R. L., Halford M. H., Reinhold V. N. and Jeanloz R. W. (1972) Isolation and structure determination of glucosyl galactosyl hydroxylysine from sponge and sea anemone collagen. Biochemistry 11, 1161-1167.

Kawasaki T. and AshweU G. (1976) Chemical and physical properties of an hepatic membrane protein that specificially binds asialoglycoproteins. J. biol. Chem. 251, 1296-1302.

KJlker R. D., Jr., Saunier B., Tkacz J. S. and Herscovics A. (1981) Partial purification from Saccharomyces cerevisiae of a soluble glucosidase which removes the terminal glucose from the oligosaccharide Glc3Man9GlcNAc 2. J. biol. Chem. 256, 5299-5303.

Lehle L. (1981) Plant cells synthesize glucose-containing lipid-linked oligosaccharides similar to those found in animals and yeast. FEBS Lett. 123, 63-66.

Moro L., Modricky C., Stagni N., Vittur F. and Bernard B. D. (1984) High-performance liquid chromatographic analysis of urinary hydroxylysyl glycosidases as indicators of collagen turnover. Analyts 109, 1621-1622.

Mutsaers J. H. G. M. Halbeek H. V., Vliegenthart J. F. G., Iwase H., Kato Y. and Hotta K. (1985) Structural analysis of dansyl glyco-asparaglnes from quail oval- bumin. Biochim. biophys. Acta 843, 123-127.

Segrest J. P. and Cunningham L, W. (1970) Variations in human urinary O-hydroxylysyl glycoside levels and their relationship to collagen metabolism. J. clin. Invest. 49, 1497-1509.

Shinkai H. and Yonemasu K. (1979) Hydroxylysine-linked glycosides of human complement subcomponent Clq and of various collagens. Biochem. J. 177, 847-852.

Spiro R. G. (1967) Studies on the renal glomerular basement membrane. J. biol. Chem. 242, 1923-1932.

Spiro R. G. (1969) Characterization and quantitative deter- mination of the hydroxylysine-linked carbohydrate units of several collagens. J. biol. Chem. 244, 602-612.

Spiro R. G. and Fukushi S. (1969) The lens capsule-studies on the carbohydrate units. J. biol. Chem. 244, 2049-2058.

Takeda T., Suzuki Y. and Yao C.-S. (1975) Experimental studies on the effect of aging and endocrine control on collagen formation in various organs. Acta Path. Jap. 25, 135-151.