THE JOURNAL OF BIOLO~IWL CHEMISTRY Vol. 250, No. 18, Issue of September 25, pp. 7106-7113., 1975

Printed in U.S.A.

Human Liver a-L-Fucosidase

PURIFICATION, CHARACTERIZATION, AND IMMUNOCHEMICAL STUDIES*

(Received for publication, November 14, 1974)

JACK A. ALHADEFF, ARNOLD L. MILLER,$ HEATHER WENAAS, TOM VEDVICK, AND JOHN S. O'BRIEN

From the Department of Neurosciences and the Department of Pathology, School of Medicine, University of California at San Diego, La Jolla, California 92037

SUMMARY

Human liver cr-L-fucosidase has been purified 6300-fold to apparent homogeneity with 66% yield by a two-step affinity chromatographic procedure utilizing agarose e-aminocaproyl- fucosamine. Isoelectric focusing revealed that all six isoelec- tric forms of the enzyme were purified. Polyacrylamide gel electrophoresis of the purified a-L-fucosidase demonstrated the presence of six bands of protein which all contained fucosidase activity. The purified enzyme preparation was found to contain only trace amounts of seven glycosidases. Quantitative amino acid analysis was performed on the puri- fied fucosidase. Preliminary carbohydrate analysis indicated that only about 1% of the molecule is carbohydrate. Gel fil- tration on Sepharose 4B indicated an approximate molecular weight for cu-L-fucosidase of 175,000 f 18,000. High speed sedimentation equilibrium yielded a molecular weight of 230,000 f 10,000. Sodium dodecyl sulfate polyacrylamide gels indicated the presence of a single subunit of molecular weight 50,100 f 2,500. The enzyme had a pH optimum of 4.6 with a suggested second optimum of 6.5. Apparent Michaelis constants and maximal velocities were determined on the purified enzyme with respect to the 4-methylumbellif- eryl and the p-nltrophenyl substrates and were found to be 0.22 mu and 14.1 pmol/mg of protein/mm and 0.43 mM and 19.6 pmol/mg of protein/min, respectively. Several salts had little or no effect on fucosidase activity although Agf and Hg2+ completely inactivated the enzyme. Antibodies made against the purified fucosidase were found to be monospecific against crude human liver supernatant fluids and the pure antigen. No cross-reacting material was detected in the crude liver supernatant fluid from a patient who died with fucosido- sis.

a-L-Fucosidase is an important enzyme not only because its activity is deficient in the neurovisceral storage disease fucosi-

* This research was supported in part by the following grants: National Institutes of Health, Postdoctoral Fellowship GF-207A, Alfred P. Sloan Foundation grant, National Institutes of Health Grant NS08682, United States Public Health Service Program Project Grant GM17702, the National Foundation-March of Dimes Grant l-75-8, and the Gould Foundation.

.j Recipient of Research Career Development Award NSO9050- 01 from Neurological and Communicable Disorders and Stroke Institute.

dosis (1, 2) but also because the enzyme is involved in the me- tabolism of several biologically active molecules containing G fucose (3-8). A highly purified preparation of a-L-fucosidase free of contaminating glycosidases would be useful for structural studies on the many oligosaccharides, glycolipids, and glyco- proteins which contain n-fucose and for studying the nature of the molecular defect in fucosidosis.

cu-L-Fucosidase has been partially purified by conventional methods from several mammalian sources (9-13). To date, no one has reported a purification of mammalian cu-r,-fucosidase which is free of contaminating glycosidases. Carlsen and Pierce (9) have purified a-L-fucosidase from rat epididymis but their preparation contained N-acetylglucosaminidase as a 3.8 y0 (by activity) contaminant. Previously we reported a one-step affinity chromatographic procedure for the partial purification of human placental cr-n-fucosidase (14). This fucosidase preparation con- tained hexosaminidase as a 3.7 ‘% contaminant, by activity.

In the present communication we describe an improved af- finity chromatographic purification procedure for a-L-fucosidase and properties and immunochemical studies on the apparently homogeneous enzyme from human liver.

EXPERIMENTAL PROCEDURES

General-Protein was determined by the Lowry method (15) using bovine serum albumin as standard. All procedures were carried out at o-4” unless otherwise stated. Enzyme solutions were concentrated by ultrafiltration using Amicon concentrators with UM-10 Diaflo membranes at 50 to 70 p.s.i. Agarose-c-aminocap- royl-fucosamine (lot numbers AF-5, 7, 10, and 11) was purchased from Miles-Yeda Ltd.

Enzyme Assays-a-L-Fucosidase activity was assayed either by using p-nitrophenyl-cu-L-fucopyranoside (Sigma) or 4-methyl- umbelliferyla-L-fucopyranoside (Koch-Light, Ltd.) as previ- ously reported (16, 17). Enzyme assays were always performed under conditions where enzyme activity was linear with the amount of protein and time of incubation. Absorbances were read on a Gilford 2400-S spectrophotometer.

Polyacrylamide Gels-Polyacrylamide gel electrophoresis in 5yo gels was performed as previously reported (14) by a modification of the general method of Davis (18). The nresence of subunits was deter&red in 10% polyacrylamide gels b.6 X 10 cm) containing 0.1% sodium dodecyl sulfate according to the general method of Laemmli (19). Forty micrograms each of bovine serum albumin, ovalbumin, cY-chymotrypsinogen, and cytochrome c (all from Sigma) were used for molecular weight determination. Cyto- chrome c was also subjected to electrophoresis in the enzyme sam- ples (20 rg/gel) as an internal standard. Two samples of a-~- fucosidase (3O~g and 70rg) were subjected to electrophoresis after treating them with 1% sodium dodecyl sulfate, 5% P-mercapto-

7106

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7107

ethanol, and 8 M urea in 62.5 mM Tris-HCl, pH 6.8, in a boiling water bath for 5 min. Electrophoresis was at 2 ma/tube for 6 hours at room temperature. The gels were stained with 0.2yo Coomassie blue in 10% acetic acid (v/v) and 50yo methanol (v/v). The gels were destained in a transverse destainer for 6 hours in 10% (v/v) acetic acid.

Isoelectric Focusing-Isoelectric focusing was performed ac- cording to the method of Haglund (20) using an LKB 8101 (110 ml) isoelectric focusing apparatus. Two per cent ampholytes (pH 5 to 8) were used in a gradient of 0 to 67% (w/v) sucrose. The temperature was maintained at O-2” with a circulating water bath (Brinkmann Lauda K-2/R). The starting amperage was 2 ma and 520 to 530 volts. Electrofocusing was conducted for 67 to 68 hours after which 25-drop fractions (approximately 0.7 ml) were col- lected. The pH of-each fraction was measured at O-2” with a Beckman digital DH meter. and 100~rrl aliauots of each fraction - _ _ were assayed for 30 to 60 min for a-L-fucosidase with the p-nitro- phenyl substrate. Isoelectric focusing was performed on crude dialyzed human liver supernatant fluid (224 units’ and 59 mg of protein) and on the purified oc-L-fucosidase (280 units’ and 18 pg of protein). The purified fucosidase was focused in the presence of 3 mg of human serum albumin (Sigma) to stabilize the fucosi- dase (14).

Kinetic Studies-Apparent Michaelis constants (K, values) and maximal velocities (V,,,) were determined for the purified enzyme graphically by the Lineweaver-Burk method (21) using both the 4-methylumbelliferyl and the p-nitrophenyl substrates. For these kinetic studies, 0.4 pg of the purified ol-L-fucosidase was incubated in substrates of varying concentrations containing human serum albumin (3.0 mg/ml) in 100 mM citrate-sodium citrate buffer (pH 5.0). All samples were incubated in duplicate for 5 min at 37’. In one experiment with the p-nitrophenyl sub- strate, 0.3 mrvr L-fucose was present in the incubation medium of the varying substrate concentrations to determine its mode of

was determined using the meniscus depletion method of Yphantis (23). The enzyme (0.3 mg of enzyme protein/ml of 10 mM NaHtPO,, pH 5.5, containing 0.02% NaNa, w/v, and 50 mM L-fucose) was centrifuged for 17 hours at 2’ at a rotor speed of 15,220 rpm in a Spinco model E ultracentrifuge. In calculating the molecular weight, the partial specific volume of the enzyme was assumed to be 0.74 ml/g.

Amino Acid Analysis-Aliquots (300 pg) of the purified a-~- fucosidase were prepared and quantitatively analyzed in duplicate for amino acids essentially as described by Spackman et al. (24). Samples of the enzyme were put in microcarius tubes and sufficient concentrated HCl was added to give a final concentation of 6 N

HCl. The ampules were evacuated, sealed, and heated at 110” for either 24 or 70 hours. The hydrolysates were evaporated to dry- ness and dissolved in 0.2 ml of 0.2 M citrate buffer (pH 2.2). These solutions were analyzed on a single column Beckman 116 amino acid analyzer equipped with an Autolab System AA computing integrator.

Carbohydrate Analysis-Aliquots of the purified fucosidase (300 Kg) were prepared and analyzed for monosaccharides by gas liquid chromatography of trimethylsilyl derivatives of methyl glyco- sides (25).

Stability Studies-Aliquots of the purified ti-L-fucosidase stored at 04” in 10 rnM NaHtPO, (pH 5.5), 0.02% NaN3 (w/v), and 100 mM L-fucose were assayed regularly to determine the stability of the purified enzyme. Assays were carried out in duplicate using 0.2 and 0.4 pg of enzyme protein in human albumin (10 mg/ml of 100 mM citrate-sodium citrate buffer, pH 5.0) for 2 and 4 min at 37”.

Hydrolysis of Fucose-containing Oligosaccharides-Aliquots (62 wg of enzyme protein) of the purified or-L-fucosidase were in- cubated with 1.6 mg of 2’-fucosyllactose and 1.6 mg of lacto-l\r- fucopentaose I in 166 mM citrate-sodium citrate buffer (pH 5.0) for 18 hours at 24-25” followed by 235 hours at 37”. Substrate blanks and a tissue (enzyme) blank were also incubated as above.

inhibition of a-L-fucosidase. The reaction products were analyzed by thin layer chromatog- Inhibition Studies-Various salts at concentrations ranging from raphy on a 0.25 mm silica gel plate (E. Merck). The plate was de-

0.25 to 25 mM (Naf, Kf, Caa+, Mgzf, Mn2+ Zn2+, Fesf, Hgn+, Ag+, veloped for 5 hours in 1-butanol/ethyl acetate/l-propanol/acetic Fe*+, and Cu2+), 0.2 to 1.2 mM dithiothreitol, and 0.12 to 1.2 mM acid/water (7/20/12/7/6). The plate was dried at room tempera- ethylenedinitrilotetraacetic acid (EDTA) were incubated with ture, sprayed with 1% orcinol (in HtO/H2SO,, l/l), and heated at 0.2 pg of the purified fucosidase to determine their effect ona-n-fu- 105-110” for 5 min. L-Fucose was used as a standard. cosidase activity. Incubations were carried out in duplicate at 37”for 6 min. Zmmunochemical Studies

pH Optimum-The pH optimum of the purified a-L-fucosidase was determined as follows : 0.4 pg of enzyme protein (in 10 ~1 of 10 mg of human albumin/ml of HZO) was added to 90 ~1 of 100 mM buffers (citrate-sodium citrate for pH values 3.1 to 5.8; KH*PO,- NazHPO, for pH values 6.0 to 8.2) of varying pH values containing human albumin (10 mg/ml). The reaction was initiated with 200 ~1 of p-nitrophenyla-L-fucoside (3 mM) in buffers of varying pH values. All incubations were carried out in duplicate for 6 min at -̂̂

Preparation of Antiserum-Antiserum to ol-L-fucosidase (anti- a-L-fucosidase antiserum) was prepared by emulsifying the puri- fied enzyme with an equal volume of Freund’s complete adjuvant and injecting the emulsion containing 80-rg samples of the puri- fied enzyme into the foot pad of each of three rabbits. The rabbits were bled at weekly intervals from the ear artery and serum was obtained by centrifugation at 1000 X g for 30 min at room tempera- ture. The sera were heated at 56” for 30 min followed bv centrifu-

Molecular Weight Determination, by Gel Filtration-An aliquot gation at 27,700 X(g for 30 min at O-2”. The serum was stored at

of the purified fucosidase (0.35 mg of protein) was subjected to -20” and thawed before use. Double immunodiffusion experi-

gel filtration on a column (0.5 X 91 cm) of Sepharose 4B (Phar- ments were carried out using either Hyland immunoplates or

macia). The column was calibrated with the following standard plates of 1.0% Epiagar/agarose made in 30 mM sodium barbital

proteins: bovine thyroglobulin (Sigma; M, = 670,000), catalase (pH 8.6) and 0.1% (w/v) in NaN3. The plates were spotted with

(Worthington; M, = 250,000), r-globulin (Schwarz/Mann; M, = samples, placed at &2” and read after 48 hours. The rabbits were

160,000), chicken heart lactic dehydrogenase2 (M, = 135,000), allowed to produce antibodies until their antisera quantitatively

bovine serum albumin (Sigma; M, = 67,000), and ovalbumin precipitated cy-r-fucosidase from solution as determined by the

(Sigma; M, = 45,000). All standard proteins (3 to 5 mg) were ap- inhibition of a-L-fucosidase activity with the addition of antisera.

plied in and eluted with 50 mM NaHnPOc (pH 5.5) which was 0.02% The rabbits were bled out by cardiac puncture 8 weeks after in-

(w/v) in NaN3 and 50 mM in L-fucose. The elution volumes for the jection of enzyme. Their antisera were prepared as described

nonenzyme proteins were determined by peak absorbance at 280 above.

nm and those of catalase and a-r-fucosidase were determined by Preparation of ZgG Fraction (26)--A 20-ml aliquot of antiserum

assaying the fractions for maximal enzymatic activity (Refs. 22 from one rabbit was dialyzed against two changes of 200 volumes/

and 14, respectively). change of 17.5 mM NaH2PO,/Na2HPO, (pH 6.9), O.O2oJ, (w/v) in

Molecular Weight Determination by High Speed Sedimentation NaN3 at O-2”. The dialyzed antiserum was centrifuged at 20,000

EquiZibrium3-The molecular weight of the purified or-L-fucosidase X g for 30 min at 2’ and the supernatant fluid (935 mg of protein) was applied to a column (1 X 88 cm) of DEAE-cellulose (What-

37’.

___ man DE52, microgranular). The column was eluted with 17.5 i A unit of activity is the hydrolysis of 1 nmol of substrate/min mM NaH2P0,/Na2HPOd (pH 6.9) containing 0.02yo (w/v) NaNa.

at 37”. 2 The chicken heart lactic dehydrogenase was a generous gift of

Under these conditions the IgG fraction elutes near the front

Dr. Johannes Everse, Department of Chemistry, University of (150 to 324 ml) (26). The IgG fraction was concentrated from 174

California at San Diego. to 8 ml by ultrafiltration. This concentrated IgG fraction was

3 The ultracentrifugation was carried out by Dr. Johannes used for the double immunodiffusion experiments (e.g., Fig. 7).

Everse, Department of Chemistry, University of California at 4 2’-Fucosyllactose and lacto-hr-fucopentaose I were generously San Diego. supplied by Dr. David Zopf of the National Institutes of Health.

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7108

Preparation of Supernatants for Immunochemical Studies-A normal human liver supernatant (l/4, w/v) was prepared as pre- viously described (17) and concentrated B-fold by ultrafiltration. A liver supernatant (l/4, w/v) from a patient who died with fuco- sidosis (17) was prepared and concentrated 11-fold by ultrafiltra- tion. These concentrated supernatant fluids were used for the double immunodiffusion experiments.

Purification of Fucosidase

Tissue E&action-Portions of human livers (1000 g total, wet weight) from five autopsied individuals which appeared normal on gross pathological examination were washed with distilled water, cut into small pieces, and homogenized in a Waring Blendor in 4 liters of buffer containing 10 mM NaHZPO, (pH 5.5) and NaN, (0.02yo, w/v) for 1 min at low speed and 1 min at high speed. The crude homogenate was centrifuged for 30 min at 23,500 X g and the supernatant fluid was filtered through cheesecloth. The super- natant fluid contained 94% of the activity present in the crude homogenate. The supernatant fluid (4.2 liters) was dialyzed against 10 mM NaH2P0, (pH 5.5) containing NaN3 (0.02%, w/v) for 3 days using four changes of 33 liters/change. The dialyzed supernatant fluid was centrifuged for 30 min at 23,500 X g, fil- tered through cheesecloth, and then used for affinity chromatog- raphy.

Afinity Chromatography-A portion of the supernatant fluid (900 ml, 42,000 units,’ and 10.9 g of protein) was applied to a column (2.8 X 35 cm) of agarose-t-aminocaproyl-fucosamine (affinity column I) at a rate of 50 ml/hour at room temperature. The column was subsequently washed with approximately 12 liters of 10 mM NaH,PO, (pH 5.5) containing 0.02% NaNs (w/v) until the absorbance at 280 nm was 0.002. The column was then eluted sequentially at a rate of 100 ml/hour with 10 mru NaHtPOc (pH 5.5), 0.02% in NaNa (w/v) containing either 50 mM L-fucose (Sigma) or 0.25 M NaCl or 7 M urea (Fig. 1). Initially, lo-ml frac- tions were collected and 1.5-ml fractions were collected when eluting with n-fucose. After assaying for ol-n-fucosidase activity, Frart,ions 1380 to 1600 were combined, concentrated, and assayed for a-L-fucosidase in the presence of human albumin (3 mg/ml) to stabilize the enzyme (14). Polyacrylamide gel electrophoresis was carried out on an aliquot of the combined and concentrated Frac- tions 1380 to 1600 from affinity column I. The gels were either stained for protein with Coomassie blue (Canalco) or sliced and assayed for a-n-fucosidase activity as follows: each slice (0.17 cm) was placed in a test tube containing 100 ~1 of 100 mM citrate- sodium citrate (pH 5.0) and 200 ~1 of 3 mM p-nitrophenyl-ol-n- fucopyranoside. The fractions were incubated for 60 min at 37”, terminated with 0.6 ml of 250 mM glycine-carbonate (pH 9.8), and absorbances read at 420 nm.

A second portion of the concentrated Fractions 1380 to 1600 (6.8 ml) was dialyzed for 235 hours against five changes of 1 liter/ change of 10 mM NaH2P0, (pH 5.5) containing 0.02’% NaN3 (w/v) for a total of 2Js hours to remove the L-fucose. The dialyzed sample (21,800 units of ol-n-fucosidase and 1.42 mg of protein) was put on a second column (1 X 27 cm) of agarose-t-aminocaproyl- fucosamine (affinity column II) at room temperature and eluted as in the first column except that the cu-n-fucosidase was eluted from the column with 100 mM n-fucose (Fig. 2). Fractions of 3 ml were collected and assayed for a n-fucosidase in the presence of human albumin (3 mg/ml). The fractions of highest specific ac- tivity (Fractions 50 to 64) were combined, concentrated, and assayed for contaminating glycosidases by previously described methods (16, 27-29). Polyacrylamide gels were carried out using the concentrated Fractions 50 to 64 from affinity column II. These gels were either stained for protein with Coomassie blue or stained for activity for 20 min at 0” with 1 mM 4-methylumbelliferyl~- L-fucopyranoside (Koch-Light, Ltd.) and subsequently visualized with an ultraviolet lamp.

RESULTS

Purijication of cr.-L-Fucosidme

The elution profile of cY-n-fucosidase from affinity column I is depicted in Fig. 1. Eighty per cent of the enzyme was eluted with 10 mM NaH2P04 (pH 5.5) containing O.O2a/, NaNa (w/v) and 50 mM r.-fucose yielding a 3980-fold purification. As in the purifi- cation of placental cr-Icfucosidase (14), 2 to 3% of a-n-fucosidase was eluted with the buffer containing 0.25 M NaCl.

The 3980-fold purified Lu-L-fucosidase was passed through a smaller column (affinity column II) of agarose-e-aminocaproyl- fucosamine (Fig. 2). The capacity of this affinity column for the partially purified fucosidase was approximately 5 times as great as that for the ar-L-fucosidase in the crude supernatant fluid. Two per cent of the enzyme did not bind to the column and eluted near the front (Fig. 2). Eighty-three per cent of the a-n-fucosi- dase was eluted with 10 mM NaH2P04 (pH 5.5) containing 0.02% Na.N3 (w/v) and 100 mM r,-fucose. In the activity profile from this column, it appears that only a small amount of fucosidase activity was eluted by n-fucose, but this is due to the inhibition of cu-L-fucosidase by the 100 mM L-fucose (14).

Table I summarizes the data on the affinity chromatographic purification of human liver a-L-fucosidase. The two-step pro- cedure results in a 6300-fold purification with a 66% yield.

50mY 0.25Y 7Y L -fucosa NOCI UWA

~~I'lilll"'l',l'lI1' IO 30 50 500 700 900 1100 I300 I500 1700 1900 2100

FRACTION NUMBER

FIG. 1. Affinity chromatography of crude human liver super- cate where the buffers were applied. Protein (absorbance at 280 natant fluid using a column (2.8 X 35 cm) of agarose-e-amino- nm) is indicated by O-O and enzymatic activity (absorbance caproyl-fucosamine (affinity column I). The column was eluted as at 420 nm) by 0-C. Aliquots of 50 ~1 were assayed for 15 min described under “Experimental Procedures” and the arrows indi- at 37”.

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7109

IO0 m Y 7M L Iucore UREA

FIG. 2. Affinity chromatography of partially purified human liver a-~- fucosidase using a column (1 X 27 cm) of agarose-s-aminocaproyl-fucosa- mine (affinity column II). The col- umn was eluted as described under “Experimental Procedures” and the arrows indicate where the buffers were applied. Protein (absorbance at 280 nm) is indicated by O-O and en- zymatic activity (absorbance at 420 nm) by O--O. Aliquots of 50 hl were assayed for 15 min at 37”.

TABLE I PuriJication of human liver a+fucosidase by aginily chromatography

Fraction Total protein Total activity Specific activity Purification factor Over-all recovery of activity

mg unilsa

23,500 X g supernatant which was put 10,900 42,009 on affinity column I

Fractions 1380 to 1600 from affinity 2.20 33,600 column I (see Fig. 1) (80yo recovery)

Fractions 50 to 64 from affinity col- 0.75 18,100 umn II (see Fig. 2) (83% recovery*)

a A unit of activity is the hydrolysis of 1 nmol of substrate/min at 37”. * Recovery of activity put on affinity column II.

nmol/mg ~rorein/min %

3.85

15,300 3,980 80

24,200 6,300 66

FIG. 3. Polyacrylamide gel electrophoresis of human liver a- n-fucosidase. Electrophoresis was for 3 hours at 1.6 ma/tube (300 volts) in 0.5 X 9.0 cm, 50/, acrylamide gels. Gel buffer was Tris (375 mu)-HCl (60 mM), pH 8.9; running buffer was Tris (50 mM)- glycine (384 mM), pH 8.2. A, concentrated postaffinity column I Fractions 1330 to 1600 (33 rg of enzyme protein) stained with Coomassie blue. B, concentrated postafKnity column II Fractions 50 to 64 (33 cg of enzyme protein) stained with Coomassie blue.

Purity of cu-L-Fucosidase

Polyacrylamide Gels-Polyacrylamide gels of the postafhnity column I sample (Fig. 3A) indicated six bands of protein which migrated in the region of ar-cfucosidase activity (as determined by slicing and assay of the gel slices). Two contaminating bands

can be seen migrating faster toward the anode than ol-cfucosi- dase. Polyacrylamide gels of the postaffinity column II sample (Fig. 3B) demonstrated the presence of only the group of six bands of protein. Incubation of a duplicate gel with the fluo- rescent substrate 4methylumbelliferyl-o-bfucopyranoside indi- cated that all six bands of protein appear to be associated with cr-cfucosidase activity.

Sodium Do&q& Sulfate-Polyacrylamide Gels-Sodium dodecyl sulfate polyacrylamide gel electrophoresis of the postaffinity column II sample (Fig. 4 A and B; 30 pg and 70 pg of cY-cfucosi- dase, respectively) indicated the presence of a single subunit. Standard proteins were subjected to electrophoresis (Fig. 4C) on a sodium dodecyl sulfate gel and the molecular weight of the subunit was graphically determined to be 50,106 f 2,500 (Fig. 5).

Assay of a-L-Fucosidase for Contaminating Glycosidases- Table II indicates that the 6300-fold purified cu-n-fucosidase preparation had only trace amounts of other glycosidase activi- ties. Whereas in our partial purification of human placental (Y-L- fucosidase (14) hexosaminidase was a large contaminant (3.7%), it has been reduced to a trace contaminant (0.007~o) in the human liver cu-n-fucosidase preparation.

Behavior of Purijied cu-L-Fucosiduse on Sepharose .JB-Fig. 6 depicts the elution profile of the purified cu-cfucosidase during gel filtration on Sepharose 4B. Only one protein peak can be seen and this peak coincides with the one peak of fucosidase activity. Approximately 96% of the fucosidase eluted from the Sepharose 4B column over several fractions of constant specific activity.

Immunochemical Criteria of Purity-The purified ru-cfucosi-

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7110

A B C Origin+

TABLE II Activity of several glycosidases in a+fucosidase preparation

r *Bovine serum albumin

c-Ovolbumin

-a-Chymotrypsin ogen

-Cytochrome C

Glycosidase Substrate Per cent of Refer- u-L-fuco- ence for

sidase activity nZ%

a-r.-Fucosidase j3-n-Fucosidase a-Galactosidase

@-Galactosidase

p-Glucuronidase

FIG. 4. Polyacrylamide gel electrophoresis in the presence of O.l$$ sodium dodecyl sulfate and stained with Coomassie blue. See “Experimental Procedures” for details. A, purified a-L-fucosi- dase (30 ag of enzyme protein) treated for the determination of subunits (19). B, purified a-L-fucosidase (70 rg of enzyme protein) treated for the determination of subunits (19). C, protein stand- ards (bovine serum albumin, ovalbumin, a-chymotrypsinogen, and cytochrome c-40 rg of each) run for molecular weight determina- tion of a-L-fucosidase subunit.

J ,1----- 01 07. 03 04 05 06 07 08 09 IO

MIGRATION OF PROTEIN

MIGRATION OF CYTOCHROME C

Fxo. 5. Mobilities of a-n-fucosidsse treated for dissociation of subunits (19) and molecular weight standards on polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. See “Experimental Procedures” for details.

dase when injected into rabbits caused the production of mono- specific antibodies against cr-cfucosidase. This can be seen in Fig. 7 which depicts the double immunodiffusion (Ouchterlony) plate of rabbit anti-cr-l;fucosidase antiserum (IgG fraction) against purified cu-n-fucosidase (Wells A and B) and normal human liver supernatant (Wells C and D).

Chumeterization of Pur$ied cr-L-Fucosidase-Fig. 8 depicts the results of isoelectric focusing on both crude (A) and highly puri- fied (B) human liver cr-n-fucosidase. As previously reported (17), six isoelectric forms are present in the crude human liver super- natant fluid (Fig. 8A). The same isoelectric forms appear to be present, and in approximately the same proportion, in the highly purfied cr-cfucosidase (Fig. 8B).

Using the purified cr-cfucosidase, apparent K, values and V,, values were determined graphically with respect to the 4 methylumbelliferyl and p - nitrophenyl substrates. They were found to be 0.22 mM and 14.1 pmol/mg of protein/min and 0.43 mu and 19.6 pmol/mg of protein/min, respectively. Fig. 9 dem- onstrates that r.-fucose is a competitive inhibitor of cu-n-fuco- sidase and suggests a probable mechanism by which the aga- rose-e-aminocaproyl-fucosamine resin binds the enzyme.

Mn*+ (5 mu) stimulated fucosidase activity 22% and Zn* had no effect on enzymatic activity at concentrations of 0.25 to 5 mu. Hg* and Ag+ (0.1 to 0.2 mu) completely inactivated en-

a-Glucosidase

a-Mannosidase

Hexosaminidase

p-Nitrophenyl a-L-fucoside 100 p-Nitrophenyl B-n-fucoside 0.007 4Methylumbelliferyl a-n-ga- 0.002

lactopyranoside 4-Methylumbelliferyl @-n-ga- 0.001

lactopyranoside 4-Methylumbelliferyl &D-&l- 0.0007

curonide 4-Methylumbelliferyl a-n-glu- 0.001

copyranoside 4-Methylumbelliferyl a-n-man- 0.001

nopyranoside 4-Methylumbelliferyl-2-aceta- 0.007

mido-2-deoxy-j3-n-glucopy- ranoside

16 16 27

27

28

29

28

27

ELUTION VOLUME irrl)

FIG. 6. Elution profile of purified a-L-fucosidase during gel filtration on Sepharose 4B. See “Experimental Procedures” for details. Protein (absorbance at 280 nm) is indicated by 0-O and enzymatic activity (absorbance at 420 nm) by 0-O.

FIG. 7. Double immunodiffusion (Ouchterlony) plate of anti- a-n-fucosidase antiserum (IgG fraction; 0.84 mg of protein in center well) against purified a-L-fucosidase and crude liver super- natants. A and B, purified a-L-fucosdiase (0.88pg of protein/well); C and D, concentrated normal liver supernatant (1.3 mg of pro- tein/well); E and F, concentrated fucosidosis liver supernatant (1.1 mg of protein/well).

zymatic activity and dithiothreitol was mildly stimulatory (13%) at a concentration of 0.2 mru. Dithiothreitol (1.1 mn%> completely reverses the inhibition of ar-n-fucosidase activity by Ag+ and Hg*+

Quantitative amino acid analysis was performed on the puri-

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

" A. IO

0.9 E

7111

TABLE III

Amino acid composition of purified a-x,-fucosidase

OS

F 07

$ 06

9 06 z I 05

L ET 04

03

40 50 6.0 70 SO

PH

FIG. 8. Isoelectric focusing of crude human liver supernatant cr-L-fucosidase (A) and purified human liver a-L-fucosidase (B). See “Experimental Procedures” for details.

24.0 0

200 .

-8 -2 2 4 6 8 IO 12 14 16 IS 20 22 24 26 28 30 32

FIG. 9. Lineweaver-Burk plot (21) using the purified human liver a-L-fucosidaee for K, determination on the p-nitrophenyl- cr-L-fucopyranoside substrate both with (+0.1 mM L-fucose) and without L-fucose (-L-fucose).

fied fucosidase (Table III). Eight cysteine residues were present along with large amounts of aspartic and glutamic acids, serine, threonine, proline, glycine, leucine, phenylalanine, and lysine.

Initial gas-liquid chromatographic analysis (25) of fucosidase for carbohydrate indicates the presence of trace amounts of mannose and galactose.

The pH optimum curve of the purified a-L-fucosidase using the p-nitrophenyl substrate as shown in Fig. 10 indicates a major pH optimum at pH 4.6. A second optimum is suggested by a shoulder at a more neutral pH (6.5).

Fig. 11 is a plot of the results of gel filtration for several pro- teins including a-n-fucosidase on Sepharose 4B. The molecular weight of a-n-fucosidase as determined from this plot was 175,000 i 18,000, assuming an error for the method of ZIZ~O% (30). The molecular weight as determined by sedimentation equilibrium was 230,000 f 10,000.

Amino acid Residues per 230,00o molecular weight

Aspartic” 210 Threonine 102 Serine 138 Glutamica 178 Proline 172 Cysteine 8 Glycine 149 Alanine 94 Valine* 84 Methionine 21 Isoleucine 54 Leucin 167 Tyrosine” 75 Phenylalanine 120 Histidine 56 Lysine 114 Arginine 55

o No attempt was made to measure proportions which might occur as the amides.

* Calculated for 72-hour hydrolysis. c Calculated by extrapolation to zero time.

FIG. 10. pH Optimum curve for purified human liver WL- fucosidase using p-nitrophenyle-L-fucopyranoside as substrate. See “Experimental Procedures” for details.

and lacto-N-fucopentaose I as is shown by the thin layer chro- matogram in Fig. 12. From a gross inspection of the thin layer plate it appears that the enzyme was more active on 2’-fucosyl- lactose (Lane S) than on lacto-N-fucopentaose I (Lane 6).

Stability studies performed on the purified fucosidase indicated that the enzyme lost no activity for at least 163 days when stored at O-4” in 10 mM NaHzPO, (pH 5.5) containing 0.020/, (w/v) NaN3 and 100 mM n-fucose. The inhibitor n-fucose which was used to stabilize the purified enzyme was either dialyzed away or diluted to insignificant amounts when kinetic studies were performed on the purified enzyme. The purified enzyme was very thermolabile and gradually lost activity after 6 min at 37”. This is in marked contrast to the thermostability of fucosi- dase in the crude state, where 4 hour incubations at 37” did not decrease enzymatic activity.

Antibodies were made against the purified fucosidase and the IgG fraction was partially purified and concentrated by DEAE-

The purified ru-n-fucosidase hydrolyzed both 2’-fucosyllactose chromatography. This IgG fraction, as well as the antiserum,

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7112

66 r

54

62 i \

CATAL AS.! /247,OOOJ

48

t T”‘xyRO6L 06”LIN I67O,OOOJ,

2 3 4 5 6 7 8 9 10

log MOLECULAR WEIGHT

FIG. 11. Molecular weight determination by gel filtration of purified a-L-fucosidsse and several standard proteins on a column (0.5 X 91 cm) of Sepharose 4B.

1234567 FIG. 12. Thin layer chromatogram showing products of reaction

after incubation of purified liver a-L-fucosidsse with 2’-fucosyl- lactose and lacto-N-fucopentaose I. See “Experimental Pro- cedures” for details. Lanes 1 and 7, or-L-fucose standard; lane 2, tissue (enzyme) blank; lane 3, 62 pg of enzyme protein + 1.6 mg of 2’-fucosyllactose; lane 4, 2’-fucosyllactose substrate blank; lane 6,62 pg of enzyme protein + 1.6 mg of lacto-N-fucopentaose I; lane 6, lacto-N-fucopentaose I substrate blank.

was found to contain monospecific antibodies against the pure antigen (Fig. 7, A and B) and against a concentrated crude hu- man liver supernatant (Fig. 7, C and 0). Cross-reacting material was not found against a concentrated liver supernatant fluid from a patient who had fucosidosis (Fig. 7, E and F). No pre- cipitin bands were seen when controls were run against preim- munization rabbit antisera.

DISCUSSION

In thii paper we have described a two-step affinity chromato- graphic procedure for the purification of human liver cu-L-fucosi- dase to apparent homogeneity. Several criteria were used to azsess purity including polyacrylamide gel electrophoresis,

sodium dodecyl sulfate-polyacrylamide gel electrophoresis, assay of the purified fucosidase for contaminating glycosidases, be- havior of the purified cr-n-fucosidase on Sepharose 4B, and the production of monospecific antibodies against the purified fucosidase. To our knowledge this is the first report of a mam- malian cu-n-fucosidase purified to apparent homogeneity. Carlsen and Pierce (9) have described a procedure for obtaining highly purified rat epididymal cu-n-fucosidase, but their preparation contained N-acetylglucosaminidase as a 3.8% (by activity) con- taminant. Our preparation of cY-cfucosidase has only trace amounts (Table II) of other glycosidase activities. This suggests that onr preparation should be useful for structural studies of glycoproteins, glycolipids, and oligosaccharides which contain n-fucose (5-8).

Isoelectric focusing indicated that all six forms of cu-L-fucosi- dase (17) were purified by the affinity chromatographic pro- cedure. This illustrates the usefulness of the purification pro- cedure for isolating all forms of the enzyme. The structural re- lationship of the six forms of cr-n-fucosidase is not known, but the activity of at least two of the forms appears to be dependent on the presence of sialic acid residue(s) (17). Robinson and Thorpe (31) and Wiederschain et al. (13) have demonstrated the pres- ence of two forms of or-r.-fucosidase (designated I and II) in human liver by gel filtration on Sephadex G-200. The relation- ship of these two forms of a-n-fucosidase to the six isoelectric forms we have found is unclear and is currently under investiga- tion.

The purified fucosidase was characterized kinetically. The Km values of the purified fucosidase (0.22 to 0.43 mM> which we ob-

tained were about the same as those obtained for the p-nitro- phenyl substrate by Levvy and McAllan (12) (0.21 mM) for rat epididymal a-rcfucosidase and by Wiederschain and Bosenfeld (32) (0.4 mM) for pig kidney cr-n-fucosidase. Our pH optimum curve is similar to that obtained earlier for cu-n-fucosidase in crude human liver supernatant fluid with the 4-methylumbellif- eryl substrate (17). However, for the crude enzyme the curve was shifted approximately 0.8 pH unit toward a more neutral pH optimum (5.4). These results are similar to the shape of the curve and the pH 5 value obtained by Robinson and Thorpe (31) for human liver fucosidase. Human liver a-cfucosidase was found to hydrolyze the two fucose-containing oligosaccharides 2’-fucosyllactose and lacto-N-fucopentaose I. Several prepara- tions of cr-n-fucosidase have been found effective in hydrolyzing 2’-fucosyllactose and lacto-N-fucopentaose I (3, 4, 33).

The molecular weight of the purified fucosidase was deter- mined to be approximately 175,000 by gel filtration and 230,000 by sedimentation equilibrium. Since proteins partition on gel filtration according to their Stokes radii (34) and since glyco- proteins are known to give anomalous molecular weights on gel filtration (30), the molecular weight obtained by sedimentation equilibrium (230,000) is probably most accurate. Our results are similar to the value (210,000 to 220,000) found for rat epididy- ma1 cr-n-fucosidase by Carlsen and Pierce (9). Treatment of the purified fucosidase with 1% sodium dodecyl sufate, 5?6 &mer- captoethanol, and 8 M urea followed by sodium dodecyl sulfate- polyacrylamide gel electrophoresis (19) indicated the presence of a single subunit with 50,100 molecular weight. Our results, which are different from those of Carlsen and Pierce (9) who found two nonidentical subunits (49,700 and 53,700) for rat epididymal cu-n-fucosidase, suggest that human liver cu-n-fucosidaee is a tetramer of four identical subunits.

Several salts, dithiothreitol, and EDTA were tested for their effect on pure human liver cu-n-fucosidase activity. Znz+ had no

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

7113

effect on enzymatic activity unlike the large inhibition (90%) of rat epididymal a-n-fucosidase found with 5 mM Zn2+ by Carlsen and Pierce (9). Hg2+ and Ag+ were found to completely inacti- vate fucosidase activity. This inactivation was prevented in the presence of dithiothreitol. These results suggest that a-L-fucosi- dase contains a sulfhydryl group which is important for enzy- matic activity.

4. BAHL, 0. P. (1970) J. Biol. Chem. 246, 299-304 5. SPIRO, R. G. (1970) Annu. Rev. Biochem. 39, 599-638 6. WATKINS, W. &I. (i966) Science 162, 172-181 7. KABAT. E. A. (1956) Blood Grouv Substances. Academic Press.

New’York ~ ’ 8. KUHN, R., BAER, H. H., AND GAUH~, A. (1958) Chem. Ber. 91,

364-374 9.

The suggested presence of sulfhydryl-containing amino acids in fucosidase was confirmed by amino acid analysis revealing the presence of 8 cysteine residues. The amino acid composition of human liver cr-L-fucosidase (Table III) -very closely resembles that for rat epididymal cr-cfucosidase (9). Our preliminary carbohydrate analysis indicates that only a very small portion (approximately 1%) of human liver fucosidase is carbohydrate. This also is similar to the finding of Carlsen and Pierce (9) that rat epididymal cu-L-fucosidase contained 0.6% carbohydrate. Both the rat epididymal and human liver fucosidases were found to contain mannose and galactose. Large amounts of pure en- zyme will be necessary to determine the complete carbohydrate composition of human liver a-L-fucosidase.

Immunochemical studies were performed with antibodies (IgG fraction) made against a-L-fucosidase. Single precipitin lines were found in double immunodiffusion experiments when anti-cu-L-fucosidase antibody was run against either crude liver supernatant fluids or pure antigen (a-L-fucosidase). No pre- cipitin lines were found against a concentrated liver supernatant from a patient who died with fucosidosis. At least three possible explanations might account for this latter observation. The pa- tient’s liver may not be synthesizing any fucosidase (a regulatory defect), may be making small amounts of cu-L-fucosidase below the limits of detection by double immunodiffusion techniques, or may be synthesizing enzyme protein which is antigenically different from normal human liver fucosidase. Experiments using highly sensitive radioimmunochemical techniques will be carried out to distinguish between the above explanations and to characterize further the nature of the biochemical defect in fucosidosis.

CARLSEN, R. B., AND PIERCE, J. G. (1972) J. Biol. Chem. 247, 23-32

10.

11.

12.

WIEDERSCHAIN, G. Y., AND ROSENFXLD, E. L. (1971) Biochem. Biophys. Res. Commun. 44, 1008-1014

BOSMANN, H. B., AND HEMSWORTH, B. A. (1971) Biochim. Bio- phys. Acta 242, 152-171

LEVVY, G. A., AND MCALLAN, A. (1961) Biochem. J. 80, 435- 439

13.

14.

15.

WIEDERSCHAIN, G. Y., KOLIBABA, L. G., AND ROSENFELD, E. L. (1973) Clin. Chim. Acta 46, 305-310

ALHADEFF, J. A., MILLER, A. L., AND O’BRIEN, J. S. (1974) Anal. Biochem. 60, 424430

LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL, R. J. (1951) J. Biol. Chem. 193. 265-275

16.

17.

18. 19. 20. 21.

ZIELKE,‘K., ‘OKADA, S., AND O’BRIEN, J. S. (1972) J. Lab. Clin. Med. 79, 164-169

22.

ALHADEFF, J. A., MILLER, A. L., WENGER, D. A., AND O’BRIEN, J. S. (1974) Clin. Chim. Acta 67, 307-313

DAVIS, B. J. (1964) Ann. N. Y. Acad. Sci. 121, 382-392 LAEMMLI, U. K. (1970) Nature 227, 68&685 HAGLUND, H. (1967) Sci. Tools LKB Znstrum. J. 14, l-7 LINEWEAVER, H., AND BURK, D. (1934) J. Am. Chem. Sot.

66, 658-666 BEERS, R. F., JR., AND SIZER, I. W. (1952) J. Biol. Chem. 196,

133-140 23. 24.

YPHANTIS, D. A. (1964) Biochemistry 3, 297-317 SPACKMAN, D. H., STEIN, W. H., AND MOORE, S. (1958) Anal.

Chem. 30, 1190-1206 25. ETCHISON, J. R., AND HOLLAND, J. J. (1974) Virology 60, 217-

229 26. KABAT, E. A., AND MAYER, M. M. (1961) Experimental Zm-

munochemistry, 2nd Ed. Chap. 49, Charles C Thomas, Springfield, Illinois

Acknowlegments-We gratefully acknowledge the excellent technical assistance of Linda Tennant and Lois Veath.

27. Ho, M. W., BEUTLER, S., TENNANT, L., AND O’BRIEN, J. S. (1972) Am. J. Hum. Genet. 24, 265-266

28. LEROY, J. G., Ho, M. W., MACBRINN, M. C., ZIF,LKF,, K., JACOB, J., AND O’BRIEN, J. S. (1972) Pediatr. Res. 6,752-767

29. SALAFSKY, I. S., AND NADLER, H. L. (1973) J. Lab. Clin. Med. 81, 45@454

REFERENCES

30. ANDREWS, P. (1964) Biochem. J. 91, 222-223 31. ROBINSON, D., AND THORPE, R. (1973) Clin. Chim. Acta 47,

403407

1. VAN HOFF, F., AND HERS, H. G. (1968) Eur. J. B&hem. 7, 32. WIEDERSCHAIN, G. Y., AND ROSENFELD, E. L. (1969) Bull. 34-44 Sot. Chim. Biol. 61. 1075-1084

2. DURAND, P., BORRONE, C., AND DELLA CELLA, G. (1969) J. Pediatr. 76, 665-674

33. STEALEY, J. R., AND WATKINS, W. M. (1971) Biochem. J. 126, 16P-17P

3. AMINOFF, D., AND FURUKAWA, K. (1970) J. Biol. Chem. 246, 1669-1669

34. DETERMANN, H. (1969) Gel Chromatography, 2nd Ed, Springer, New York

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from

J A Alhadeff, A L Miller, H Wenaas, T Vedvick and J S O'Brienimmunochemical studies.

Human liver alpha-L-fucosidase. Purification, characterization, and

1975, 250:7106-7113.J. Biol. Chem. 

  http://www.jbc.org/content/250/18/7106Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/250/18/7106.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 31, 2020

http://ww

w.jbc.org/

Dow

nloaded from