A procedure for the rapid purification in high yield of human glucocerebrosidase using immunoaffinity chromatography with monoclonal antibodies

ANALYTICAL

BIGCHEMISTRY

154,655-663

(1986)

A Procedure for the Rapid Purification in High Yield of Human Glucocerebrosidase Using lmmunoaffinity Chromatography with Monoclonal Antibodies JOHANNES M.F.G. AERTS,* JOHN A. BARRANGER,?

WILMA E.DONKITR-KOOPMAN,* GAR~J. MURRAY,? JOSEPH M. TAGER.* AND ANDRE?W. SCHRAM*

Received

November

18. 1985

A novel chromatographic immunoafinity procedure is described for the puritication of Form I glucocerebrosidase (see J. M. F. G. Aerts, W. E. Danker-Koopman. M. K. Van der Vliet. L. M. V. Jonsson. E, I. Ginns, G. J. Murray% J, A. Barranger, J. M. Tager, and A. W. Schram. 1985. Eur. .f. Bimhwx 150, 565-574) from extracts of human tissues. The affinity support consists of two monoclonal anti-(glucocerebrosidase) antibodies immobilized by covalent coupling to CNBr-activated Sepharose 49, After adsorption of the enzyme from a crude detergent extract? the column is washed successively with 3Oq’ ethylene glycol in citrate b&er (pH 6). lcrj Triton X-100 in citrate phosphate buffer (pH :52)? and SO%* ethylene glycol in citrate buffer. The enzyme is eiuted with 90% ethylene glycol in citrate buffer. After dilution to 3OQc ethylene glycol. the immunoaffinity purification is repeated. The procedure can be completed within less than IX h. The fmal preparations have a high specific activity (50 U/mg protein (n = 4) for the placental enzyme) and contain no detectable impurities after polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. The yield is high (8 I 3~ 8% for the placental enzyme). The immunoaffinity column has a high capacity. can be regenerated easily. and can be utilized repeatedly without loss of activity. !c lY86 Acadcmx Press Inc. KEY WORDS: glucocerebrosidase; $glucosidase; Gaucher disease: immunoaffinity chromatography: monoclonal antibodies; sphingolipidoscs

Glucocerebrosidase’ (N-acylsphingosyl- l0-/j-D-glucoside:glucohydrolase, EC 3.2. I .45) is a lysosomal enzyme that catalyzes the hydrolysis of the glycosphingolipid glucocerebroside to glucose and ceramide. The enzyme is deficient in Gaucher disease (1.2). which leads to accumulation of glucocerebroside in lysosomes in cells of the reticuloendothelial system (3). Multiple molecular forms of glucocerebrosidase have been identified in extracts of tissues or cells by immunoblotting ’ This enzyme was formerly known as [Sglucocerebrosidase and is often referred to as &glucosidase or acid 6.glucosidase. The IUB Commission on Enzyme Nomenclature recommends glucocerebrosidase as the trivial name.

(4-6) or isoelectric focusing (7-10). The heterogeneity in pZ and I\/~ reflects the presence of differently glycosylated forms of glucocerebrosidase. as indicated by the fact that deglycosylation results in the appearance of one main .&lr and pZ form (8.1 1). We have recently shown ( IO) that spleen contains two forms of glucocerebrosidase that can be distinguished immunologically. Form I, the major form (80-90s of the total glucocerebrosidase activity in spleen). is precipitated by rabbit antiserum (4) or monocional antibodies ( 12) against purified placental glucocerebrosidase, whereas Form Il. the minor form, is not ( 10). Form II glucocerebrosidase is also found in liver and kidney but is very

656

.AERTS

low (less than 5%) of the total glucocerebrosidase activity) in placenta and fibroblasts (J. M. F. G. Aerts, unpublished observations). Several procedures have been described for the purihcation of glucocerebrosidase ( 13- 16) and the steps employed have made USCof the fact that all forms of glucocerebrosidase are hydrophobic glycoproteins that require negatively charged lipids or bile salts for maxima1 rates of hydrolysis of glucocerebroside or synthetic substrates. The purification steps employed have included lectin affinity chromatography using Concanavalin A-Sepharose ( 13): hydrophobic affinity chromatography using phosphatidylserine bound to Sepharose ( 13), decyl-Agarose ( 14). or octyl-Sepharose (14); and substrate affinity chromatography using the substrate analogs glucosylsphingosine ( 15) 6’-aminohexanoyl-(2-AT-sphingosyll-@%D-glucoside) (16). or 6’-aminohexyldodecanedioyl-(2-N-sphingosyl1-o-@-D-gjucoside) (16) covalently bound to Sepharose. The procedures discussed above have been developed for the purification of glucocerebrosidase from placenta. We have recently produced monoclonal antibodies against glucocerebrosidase ( 12) some of which bind the enzyme with a high affinity. In this paper we describe a rapid procedure for the purihcation of Form I glucocerebrosidase from several sources with high yields. making use of two immobilized monoclonal antibodies. MATERIALS

AND METHODS

Materiak Glucocerebroside isolated from spleen was labeled in the glycosyl moiety with “C as described in (4). Conduritol B-epoxide was kindly supplied by Dr. A. Gal. CNBr-activated Sepharose 4B was from Pharmacia (Uppsala. Sweden), 4-methylumbelliferyl-/3-Dglucopyranoside and the p-nitrophenylglycosides were from Koch-Light (Combrook, U.K.), and sodium taurocholate (grade A) was from Calbiochem (San Diego, Calif.). All other reagents were of the purest grade available. An antiserum against purified placental glucoce-

ET

Al

rebrosidase was raised in a rabbit as described in (4). Monoclonal antibodies against placental glucocerebrosidase were obtained from mouse ascites fluid as described previously ( 12). ~/~:~Y~z~ u.s~\~.t~.s. Glucocerebrosidase activity was measured at 37OC as described in Ref. (I 7). The reaction mixture (final vol 200 ~1) contained 60 rnM potassium phosphate buffer (pH 5.8). 0.1%’ (v/v) Triton X-100. 0.25%~ (w/v) taurocholate, 5 ~1 ,6-D-[ 1-‘4C]glucocerebroside (7.5 mg/ml: sp act. 450 cpm/ mmol) and enzyme preparation. For the estimation of &glucosidase activity. the reaction mixture (linal vol 0.5 ml) contained 5 mM 4-methylumbelliferyl-$-D-glucopyranoside, 0.1% (v/v) Triton X- 100. O.?Z (w/v) sodium taurocholate, and, where indicated, conduritol B-epoxide. After 20- 180 min at 37’C, the reaction was stopped by adding 2.0 ml of0.3 M glycine-NaOH (pH 10.6). The 4-methylumbelliferone formed was determined fluorometrically. The contribution of glucocerebrosidase to total &glucosidase activity was calculated as the conduritol B-epoxide inhibitable hydrolysis of the synthetic substrate (10). Enzyme preparations were preincubated with 5 rnM conduritol B-epoxide for IO min. The concentration ofthe inhibitor in the reaction mixture was 1 mivt. Units are defined as micromoles substrate hydrolyzed per minute. Prcjparutim o/’ tix\w mttacts. All procedures were performed at 4OC. Tissues, which had been stored at -7O’C. were homogenized with the aid of an Ultraturrax in 4 vol 50 rnM potassium phosphate buffer (pH 6.5) containing 0.25% (w/v) Triton X- 100. The suspension was centrifuged at lOO.OOOgfor 1 h, and the supernatant was used for further studies.

Preparation oj’ in~tnobilized uwtmlonal antihdie~~ (8~4/~C7-Sep}lur(~st~ 4B). The monoclonal antibodies used were those produced by hybridoma clones 8E4 and 2C7: they bind to different epitopes on glucocerebrosidase ( 12). The hybridoma clones were grown as ascites in BALB/c mice. The monoclonal antibodies were immobilized according to the

PURIFICATION

OF

GLUCOCEREBROYDASE

procedure described by Pharmacia, by coupling 200 ~1 ascites fluid (about 4 mg protein) per milliliter CNBr-activated Sepharose. Itntnmoq#it~ity chrotnuto~ruph~~. A 4-ml column containing immunoaffinity resin was equilibrated with 0.1 M citrate-O.5 M NaCl (pH 6.5) (buffer A). Extracts were applied to the column at a flow rate of 0.2-0.5 ml/min. The column was washed with 20 vol buffer A. IO vol 30%>ethylene glycol in buffer A, 10 vol 1% Triton X-100 in 10 rnM sodium citrate20 rnM sodium phosphate (pH 5.2). and 10 vol 50%1ethylene glycol in buffer A at a flow rate of 2 ml/min. The column was eluted with 30 ml 90%’ ethylene glycol in buffer A al a flow rate of 0.2-0.5 ml/min. All steps were performed at room temperature and fractions were collected on ice. The column was regenerdted by washing with 50 vol 90%) ethylene glycol-1% Triton X-100 in buffer A and 10 vol buffer A. The immunobeads were stored at 4’C in buffer A containing 1 rnM sodium azide; their capacity to bind glucocerebiosidase remained unchanged after repeated use and regeneration for at least 6 months.

657

Isof~lcUric jixzsitzg. Preparative isoelectric focusing, ( 19) in Ultrodex was performed using a LKB 21 17 Multiphor apparatus according to the manufacturer’s instructions. After completion of electrophoresis the gel was divided into fractions and the fractions were suspended in water. After centrifugation aliquots of the supernatant were taken for analysis.

RESULTS

lmmunoprecipitation experiments were carried out to determine the extent of binding of glucocerebrosidase to 8E4/2C7-Sepharose. A Triton X-100 extract of placenta was incubated with different amounts of 8E4/2C7Sepharose or, as a control. immobilized anti(cy-glucosidase) monoclonal antibodies. After 1 h at room temperature, the beads were removed by centrifugation and glucocerebrosilttltt~~~tlo~~r~~~~i~~it~~t;ot~ (?~‘~~lll~~(~~~~r~~br(~~~i~~~l.~~~dase activity was measured as the conduritol B-epoxide inhibitable 4-methylumbelhferyl-fi\t,ith t&hi1 ut~ii-[~l~i~~o~~r~~~r~~.si~~~.s~~) mtitwdi0.s. Rabbit anti-(placental glucocerebrosidase) glucoside hydrolyzing activity in the superantibodies were immobilized by adsorption to natant. As shown in Fig. I approximately 95%1 Protein A-Sepharose 4B as described in ( 10). of the activity was removed by incubation of Enzyme preparations were incubated with difthe extract with 8E4/2C7-Sepharose, whereas ferent amounts of immobilized antibodies for none was removed by incubation with im1 h at room temperature with rotation. The mobilized anti-( cy-glucosidase) antibodies. The activity remaining in the supernatant after inincubation mixture (final volume 0.25 ml) contained 0.1% (w/v) Triton X-100-0.05 M cubation with 8E4/2C7-Sepharose could not potassium phosphate buffer (pH 6.5). After be precipitated by incubation with an excess centrifugation for 5 min at 3OOOgthe enzyme of immobilized rabbit anti-(glucocerebrosiactivity remaining in the supernatant ‘was dedase) antibodies (not shown) and therefore represents Form II glucocerebrosidase (10). termined. ~~~l~~~~,r~lluttli~~l gci t~le~~tr~tp~~ore.~i.~ Also in other tissues examined, including (PAGE)’ in tk presetzce 01’ .sodimi dodeqd spleen. liver. brain, and kidney. and in fibrosri&tc~ (SDS). Enzyme preparations were blasts, Form I glucocerebrosidase. defined as subjected to SDS-PAGE according to Lathe activity precipitablc by rabbit anti-(giuemmli ( 18). cocerebrosidase) antibodies ( 10). was completely precipitated by incubation with SE4/ 2C7-Sepharose (not shown). ’ Ahhreviations used: PAGE, polyacrylamide gel electrophoresis: SDS sodium dodecyl sulfate. It can be concluded that the distinction in

658

AERTS

pt

SEPHAROSE BEAOS

FIG. 1. Immunotitration with immobilized monoclonal antibodies ofglucocerebrosidase in a Triton X- 100 extract of placenta. A mixture of the monoclonal anti-(glucocerebrosidase) antibodies 8E4 and 2C7 was coupled to CNBractivated Sepharose 4B as described under Materials and Methods. As a control. monoclonal anti-(cv-glucosidase) antibodies were coupled in a similar manner. Samples of a Triton X-100 extract of placenta were incubated with different amounts of the immobilized antibodies for I h at room temperature. After ccntrifugation to remove the beads, glucocerebrosidase activity, measured as the conduritol B-epoxide inhibitable hydrolysis of 4-methylumbelliferyl-/J-giucoside. was determined in the supernatant. (A) SE4/2C7-Sepharose: (A) anti-(a-giucosidase)-Scpharose. The results of a typical experiment are shown.

total glucocerebrosidase that is made by 8E4/ 2C7-Sepharose is identical to that made by polyclonal rabbit antibodies. The capacity of the SE4/2C7-Sepharose beads prepared as described under Materials and Methods was ~2.0 U/ml packed volume of beads. The binding of glucocerebrosidase to 8E4/ 2C7-Sepharose was not affected by increasing the Triton X-100 concentration (up to 1.O% v/v). by high salt concentrations (up to 2 M NaCl), by the presence of the substrate analogs gluconolactone and octyl-&glucoside, or by varying the pH in the range 5-7 (not shown). The recovery of glucocerebrosidase activity in the immunoprecipitation was complete, the sum of enzyme activity in the supematant and that bound to the beads equaling the original input of enzyme activity. ~?tdtw.st~ ~~f’Bowd

~~ll~,~)~~~~~~~b~~~~i~~~.~l~

Perhaps as a consequence of the high binding affinity and capacity of 8E4/2C7-Sephar-

ET

Al

osc, the release of bound glucocerebrosidase proved to be difficult when using conventional procedures. General methods for the elution of bound antigen from immobilized antibodies-for instance, variation of the pH or ionic strength or the use of potassium thiocyanateeither failed to bring about a detectable release of glucocerebrosidase activity or resulted in rapid inactivation of the enzyme (Table I). Ethylene glycol has been reported to be an effective eluting agent for glucocerebrosidase in hydrophobic affinity chromatography (14) and in substrate affinity chromatography (16). To examine the usefulness of ethylene glycol as an eluting agent in immunoaffinity chromatography. aliquots of SE4/I?C7-Sepharose with bound placental glucocerebrosidase were incubated for 15 min with increasing concentrations ofethylene glycol. After the beads were pelleted by centrifugation. enzyme activity in the supernatant was determined. Table I shows that ethylene glycol concentrations above 80%) (v/v) released all bound glucocerebrosidase.

TABLE ELUTION

~~GI.LJ~.~EREBROSIDASE ~MMUNOAFFINITY

Eluting

I

agent

Glycine-HCI. pH 3 IiCNS Triton X- IO0 Ethylene glycol

Concentration 0.2 M 2.0 M IY 40’$ 50?;* 60% 1ov X0% 90?

FROM

SUPPORT

Activity in eluate (% of initial activity bound) <5 4 <5 <5 <5 I0 28 96 97

IVO/
PURIFICATION

OF

659

GLUCOCEREBROSIDASE

When 10 U of purified placental ghrcocerebrosidase (20) in 2 ml buffer A containing 0, 1% bovine serum albumin was applied at a flow rate of 0.2 ml/min to an 8E4/2C7-Sepharose column (packed vol 4 ml), about 96%) of the activity was bound to the column. The column was washed with 40 ml 30% and 40 ml 50%~ ethylene glycol without elution of activity. The column was next eluted with IO ml 90% ethylene glycol at a flow rate of 0.2-0.4 ml/min. All steps were performed at room temperature and the fractions were collected in an ice-water bath. More than 95% of the activity was eluted in the first 5 ml 907~ ethylene glycol fractions and no detectable activity was associat’ed with the beads after elution. The eluted glucocerebrosidase was stable for at least several weeks when stored in 90%’ ethylene glycol at O’C.

A Triton X- 100 extract of 25 g placenta was prepared as described under Materials and Methods and the extract was applied to the 8E4/2C7-Sepharose column at a flow rate of 0.5 ml/min. Figure 2A shows the elution profiles for protein and enzyme activity obtained after washing the column successively with buffer A. 30% ethylene glycol, I%# Triton X- 100. and 50%) ethylene glycol, and finally eluting with 90% ethylene glycol. The bulk of the protein was ehtted after washing with buffer A. A small amount of activity was eluted with 50% ethylene glycol. The bulk of the activity was eluted in the first 5 ml of 90% ethylene glycol. These fractions were collected, diluted to 30% ethylene glycol, and applied once more to the immunoaffinity column. The elution profiles obtained in this second immunoaffinity purification step are shown in Fig. 2B. Again, the bulk of the enzyme activity. about 90% of that applied to the column, was eluted in the first 5 ml of 90%’ ethylene glycol.

I-

O

10 20 30 FRACTION NUMBER

&O

50

FIG. 2. Elution protiles of enzyme activity and protein during immunoaffinity chromatography puriticdtion of glucoceretxosidase from placenta. A Triton X-100 extract of placenta was subjected to two successive immunoaffinity purification steps (A and B%respectively) as described under Materials and Methods. In each step the column was washed with butler A and then eluted successively with buffer A containing the following components: t I) 3O’V ethylene gtycol; (2) 50’% ethylene glycol (first step only): (3) I % Triton X- IO0 in citrate-phosphate buffer: (4) 90% ethylene glycol.

The results of a typical purification of placental glucocerebrosidase are shown in Table 2. The enzyme had a specific activity of 50 U/ mg protein which represents a lOO.OOO-fold purification. The overall yield was 8OYc.Similar results were obtained in four separate experiments. The procedure was found to be equally applicable to extracts of spleen, liver. and brain (Table 2) to fibroblast extracts (not shown). and to soluble gtucocerebrosidase in urine (not shown). In each case. only Form I glucocerebrosidase was purified. The procedure is rapid, the total time involvedl being less than 18 h. The procedure can be scaled up by using a series of columns in parallel. The use of a discontinuous gradient of ethylene glycol results in elution of a highly concentrated solution of glucocerebrosidase so that concentration procedures which lead to losses of enzyme activity (not shown) can be avoided.

AERTS

660

El

TABLE PURIFKATI~N~FGLIJCOCEREBROSIDASEFROM Source of enzyme Placenta

Liver

Brain

Spleen

Puritication

step

Total protein Cm3

Triton X-100 extract 1st affinity step 2nd affinity step

2096

Triton X-100 extract 1st affinity step 2nd affinity step

I498

Ttiton X-100 extract 1st affinity step 2nd affinity step Triton X- IO0 extract 1 st affinity step 2nd affinity step

Total activity (mu)

AL. 2 DIFFERENI

Specific (mU/mg

TISSUES

activity protein)

Purification (-fold)

Yield YCJ

x.0 0.01

775 663 649

0.37 791 50616

2.13x 133.600

100 x2 83

I.6 0.06

319 I50 I69

0.2 I 94 >2863

44x >13.633

100 47 53

882 33.0 0.03

200 I56 I44

0.23 I83 >5708

796 >24,817

IO0 78 76

319 240 223

0.30 217 >7020

723 >23.400

IO0 75 70

I059 I.1 0.03

Nurc Triton X-100 extracts were prepared from IO0 g placenta. 25 g liver, 25 g brain, and 25 g spleen as described under Materials and Methods. The eluate obtained after the first immunoaffinity purification step was subjected to a second immunoaffinity purification step carried out as described under Materials and Methods. The results of a typical purification are given.

Figure 3 shows that after SDS-PAGE placental glucocerebrosidase purified by immunoaffinity chromatography contained a single protein band at 1!4~approximately 67*000, exactly as seen in placental glucocerebrosidase isolated by a modification (20) of the method of Furbish ef ul. ( 14). Isoelectric Focwing Placenta (and other tissues) contain different pZ forms of glucocerebrosidase (7-10). Figure 4 shows that glucocerebrosidase isolated from brain and placenta by immunoaffinity chromatography show isoelectric forms similar to those seen in the original Triton X- 100 extracts. Thus the immunoaffinity procedure does not lead to preferential purification of one particular p1 form. DISCUSSION

Procedures developed for the purification of glucocerebrosidase have made use of the

1 2

-67

-43

-30 FIG. 3. SDS-PAGE ofglucocerebrosidase purified from placenta. The enzyme was purified either by immunoafhnity chromatography (lane 1) or by using a modification (20) of the procedure of Furbish IZI u/. (14) (lane 2). Stained with Coomassie blue.

PURIFICATION

OF

GLUCOCEREBROSIDASE

FIG. 4. Isoelectric focusing of glucocerebrosidase in a Triton X-100 extract of placenta and brain and after purification of the enzyme by immunoaliinity chromatography. For conditions see Materials and Methods. (A) Triton X- IO0 extract of placenta (input 10 mU glucocerebrosidase). (B) Purified placental ghtcocerebrosidase (input IO mU glucocerebrosidase). (C) Triton X-100 extract of brain (input 3 mU glucocerebrosidase). (D) Purified brain glucocerebrosidase (input 2 mU glucocerebrosidase),

fact that the enzyme is a hydrophobic glycoprotein catalyzing the hydrolysis of o-D-glucosidic bonds. The isolation of glucocerebrosidase to apparent homogeneity from placental membranes is possible through procedures based on hydrophobic affinity chromatography ( 14,20) and substrate analog affinity chromatography ( 15.16). However, the general applicability ofthese procedures for the isolation of glucocerebrosidasc from other sources is not known. Heterogeneity in glucocerebrosidase has been observed in several tissues and cell types. The occurrence of soluble and tightly membrane-associated forms has been reported ( 10). as well as forms differing in apparent molecular mass and isoelectric point (4-10). The aim of this study was to develop a rapid. general procedure for the purification of enzymatically active Form I glucocerebrosidase from various sources without selection of particular molecular forms. The immunoaffinity chromatographic procedure using immobi-

661

lized monoclonal antibodies 8E4 and 2C7 as affinity supports and ethylene glycol as eluting agent fulfills these requirements. The 8E4/ 2C7-Sepharose column shows a high alfmity and capacity for binding Form I glucocerebrosidase and continued to do so after repeated use and storage over a prolonged period of time. Elution of SE4/2C7-Sepharose resulted in a concentrated stable glucocerebrosidase preparation of high purity. The isoelectric point profile of Form I glucocerebrosidase in a homogenate did not change upon immunoaffinity chromatographic purification. The procedure is applicable to both soluble and membrane-associated glucocerebrosidase: it works equally well with Triton X- 100 extracts oftotal tissue (Table 2). with urine or aqueous extracts of placenta and spleen (not shown) and with Triton X-100 extracts of placental or splenic membranes (not shown). WC have found that mutant Form I glucocerebrosidase from the spleens of patients with different phenotypes of Gaucher disease is completely, precipitated by SE4/2C7-Sepharose and have used the immunoathnity procedure to isolate the enzyme from spleen of a Type 1 patient (J. M. F. G. Aerts and W. E. Donker-Koopman, unpublished observations). We have also tested the efficiency of the purification procedure for the placental enzyme using only one monoclonal antibody. The affinity of the binding of glucocerebrosidase to immobilized 2C7 was lower than that of immobilized 8E4 or 2C7/8E4 (not shown). Although the use of SE4 alone results in efficient purification of glucocerebrosidase from placenta. there are indications that the epitope recognized by 8E4 may be modihed in frbroblasts from patients with Type 2 Gaucher disease (6). We thus used an immunoaffinity support containing two antibodies directed against different epitopes on glucocerebrosidase ( 12). The fact that the procedure can also be used for purification of glucocerebrosidase from urine indicates that urine from patients might

662

AERTS

ET

TABLE

AL.. 3

SPECIFIC ACTIVITY AND YIELD ~FGI LJCOCEREBROSIDASE PURIFIED PLACENTA ANDSPLEEN BY DIFFEREN~-A~J~~~~~R~

FROM

Yield

source of enzyme Placenta

Spleen

Authors

Ref.

Furbish cf al. Grabowski el Aerts ~1 al. Pentchev el Aerts el ul.

Specific (U/mg

activity protein)

This

paper

I5 60 50

This

(17) paper

12 >I

(14)

(16)

al.

al.

Percentage of activity in tissue homogenate

be an easily available source of pure mutant enzyme for comparative studies. Form II glucocerebrosidase, which constitutes about 10% of the total glucocerebrosidase activity in spleen? liver, and kidney and less in other tissues (lo), is by definition not purified by the immunoaffinity procedure so that an immunoaffinity chromatographic step could be used to deplete a preparation completely of Form I glucocerebrosidase, allowing a better characterization of Form II glucocerebrosidase. In conclusion. the procedure described in this paper provides a method for the purification of gIucocerebrosidase with high yields from various tissues to give preparations with specific activities as high as the purest preparations reported in the literature (Table 3). Essential features of the method are that the enzyme is tightly bound to the immunoafinity column, that impurities are removed by washing the column with relatively high concentrations of ethylene glycol and T&on X100, and that the enzyme, which is known to be a highly hydrophobic protein (14,16). is finally eluted with 907% ethylene glycol. The procedure may therefore be of general applicabiIity for the purification of hydrophobic proteins, provided that suitable antibodies are available. This possibiIity is being investigated.

expressed

as

Percentage of activity in extract of membranes 28 50

Units/kg tissue

SO

1 s.4 29

2 70

I.4 8.9

ACKNOWLEDGMENTS The authors thank Ronald Oude Elferink, Peter Plomp? Theunis Egelte, and Ed Ginns for stimulating discussions and helpful suggestions. Martin van der Vliet for his help in some of the preliminary experiments. and Wendy van Noppen for her help in the preparation of the manuscript. This study was supported by a grant from the Netherlands Organization for Pure Scientific Research (ZWO) under the auspices of the Netherlands Foundation for Fundamental Medical Research (FUNGO) and by grants from the National Gaucher Foundation. USA.

REFERENCES I. Brady.

R. 0..

Kanfer,

J. N.. and Shapiro.

D. ( 1965)

Biochem Bioph~x. Rtx Cotmmn. 18, 22 l-225. 2. Patrick, A, D. ( 1965) Bimhettz. J. 97. l7c- l7d. 3. Brady. R. 0.. and Barranger, J. A. ( 1983) itz The Met-

4,

5.

6.

7.

abolic Basis of Inherited Diseases (Stanbury, J. B., Wijngaarden. J. B.. Frederickson, D. S.. Goldstein, J. L.. and Brown? M. S.. eds.). pp. 842-856, McGraw-Hill. New York. Ginns. E. I.. Brady, R. 0.. Pirucello, S.. Moore. C.. Sorrell. S., Furbish. F. S., Murray. G. J., Tager, J. M., and Barranger, J. A. (1982) Pr~c,. N&/. , lead Ski. US.4 79, 5607-56 IO. Pirucello, S., Barranger. J. A., Barton, N. W., Brady, R. 0.. and Ginns, E. I. ( 1984) Bidwtn. Med. 31, 73-79. Ginns. E. I.. Tegelaers. F. P. W., Barneveld. R.? Galjaard, H.. Reuser. A. J. J.? Tager. J. M., and Barranger. J, A, (1983) C/in. Chittz. .4cfa 131, 283287. Maret, A., Salvayre. R., N?gre, A.. and Doust&Blazy. L. (1981) Ezlr, J. Bidrem. llS, 455-461.

PURIFICATION

OF

8. Ginns, E. I.. Brady, R. 0.. Stowens. D. W.. Furbish? F. S.. and Barranger, J. A. ( 1982) itz Gaucher Disease: A Century of Delineation and Research (Desnick. R. J,, Gatt. S.? and Grabowski, G. A.. eds.), pp. 405-414. Liss, New York. 9. Ginns, E. l., Brady. R. 0.. Stowens D. W.. Furbish? F. S.. and Barranger, J. A. ( 1980) Biochm?. Biop/t),.T Rm. ~‘~Hwz~I. 97, 1103-I 107. IO. Acrts. J. M. F. G.. Danker-Koopman W. F.., van der Vliet, M. K.. Jonsson. L. M. V,? Murray, G. J.. Ginns. E. I., Barranger. J. A.. Tager, J. M.. and Schram. A. W. ( 1985) &fr. ./. ~;&zc~~rr ISO, 565574.

1 I. Erickson. A. H.. Ginns. E. I.. and Barranger. J. A. (1985) .I ho/ Chwz. 260, 14319-14324. 12. Bameveld. R., Tegelaers. F. P. W.. Ginns, E, 1.. V&r, P.. Laanen. E. A.. Brady? R. 0.. Galjaard, H.. Barranger. J. A.. Reuser, A. J. J.. and Tager, J. M. (1983) &r. J. E;o&r~ 134, 585-589.

663

GLUCOCEREBROSIDASE

13, Dale, G. L., and Beutler. E. (1976) Proc,. Nu//. .4cud. Sci, C1S.d 13, 4672-4674. 14. Furbtsh. F. S.. Blair, H. E., Shiloach, J., Pentchev. P. G.. and Brady% R. 0. (1977) Proc N&l. .kad Sci. [:.%I 74, 3560-3563. IS, Strasberg, P. M.. Lowden, J. A.. and Mahuran. D. (1’982) Cut&. J. Bim-hem 60, 1025-103 I. 16. Grabowski. G. A.. and Dagan, A. (1984) .4tru/ Bmhun. 141, 267-219. 17. Pentchev, P. G.. Brady- R. 0.. Blair. H. E.> Britton. D E.? and Sorrell. S. ( 1978) I’roc. Nuf/ .4cud. SC? L:*T.4 74, 3970-3973. 18. Laemmli. U. K. (1970) Nufrlre /I,o~I&] 227, 680685.

19. Radola,

B. J. ( 1969) Biwhim

Bicydiyv.

.ICTU 194, 335-

3513.

20.

Murray. G. J., Youle. G. C.. and Barranger, 147.301-310.

R. J., Candy? S E.. Zirzow? J. A. (1985) .4&. Bicdww.