Neurochemical Research, Vol. 23, No. 6, 1998, pp. 899-905

Enzymatic Release of Zn2+-GlycerophosphocholineCholinephosphodiesterase from Brain Membranes byGlycosylphosphatidylinositol-Specific Phospholipases and itsRegulation

Ji-Yeon Lee,1 Mee Ree Kim,2 and Dai-Eun Sok3

(Accepted October 27, 1997)

Enzymatic conversion of glycosylphosphatidylinositol (GPI)-linked Zn2+-glycerophosphocholinephosphodiesterase was investigated. The activity of glycosylphosphatidylinositol-specific phospho-lipase-D (GPI-PLD), based on the conversion of amphiphilic form of phosphodiesterase into hy-drophilic form, showing an optimum pH of about pH 6.6, increased continuously until 60 min.The activity of membrane-bound GPI-PL, based on the formation of hydrophilic form of phos-phodiesterase, exhibiting an optimum pH of 7.4, increased up to 30 min, and reached a plateau.Inhibition studies indicate that while GPI-PLD activity was generally sensitive to ionic bio-deter-gents, it was not inhibited by myristoyl glycerol, a neutal detergent. Meanwhile, the membrane-bound GPI-PL was not affected remarkably by these detergents except that myristoyl glycerolexpressed a modest increase of activity of membrane bound GPI-PL. In addition, the membrane-bound GPI-PL appeared to be enhanced by by suramin or oleic acid, which strongly inhibited GPI-PLD. From this results, it is suggested that in brain there may be two phospholipases responsiblefor the conversion of membrane-bound GPI-anchors to hydrophilic forms, and that this conversionmight be regulated by endogenous lipids.

KEY WORDS: GPI; phospholipase; glycerophosphocholine; phosphodiesterase; inhibition.

INTRODUCTION

In eucaryotic cells, a number of proteins includingalkaline phosphatase are attached to plasma membranesby a glycosylphosphatidylinositol (GPI)-linked anchor(1,2). Most proteins with GPI-anchors can be releasedfrom cell surfaces by bacterial GPI-phospholipase C(2,3). Anchor cleavage by this PLC also results in the

1 College of Pharmacy.2 Department of Food And Nutrition, Chungnam National University,

Taejeon 305-764, Korea.3 To whom to address reprint requests: Dai-Eun Sok, Ph.D., College

of Pharmacy, Chungnam National University, Yuseong-Ku, Taejeon305-764, Korea. Tel) 042-821-5940.

8990364-3190/98/0600-0899$15.00/0 C 1998 Plenum Publishing Corporation

conversion of the intact amphiphilic proteins to a hydro-philic proteins. The conversion can be monitored by Tri-ton X-114 phase partitioning (1,2,4). Recently, aGPI-specific phospholipase D (PLD) was identified inmammalian plasma or tissues, and purified from serumor brain tissue (5,6). The GPI-PLD was worthy of atten-tion, since one of its functions may be to release thepolar proteins from GPI-anchored proteins in mem-branes either at the cell surface or in an intracellularcompartment. However, its in vivo function is still ques-tioned because the purified PLD, which can not degradeGPI anchor, was active only in the presence of non-ionicdetergent such as triton X-100 (4,7). Moreover, there isalso an uncertainty regarding the mechanism of regula-

900 Lee, Kim, and Sok

tion of this enzyme and which of the numerous GPI-anchored proteins found on mammalian cell surfaces areits substrates. Despite these facts, there are reports onthe presence of hydrophilic alkaline phosphatase pro-duced presumably from the enzymatic conversion of am-phiphilic form of alkaline phosphatase (8). In thisrelation, it was recently reported that the hydrolysis ofmembrane-bound alkaline phosphatase by GPI-PLD re-quired bile salts (9).

In a previous study (10), we reported that the cy-tosol fraction of brain homogenate contained bothamphiphilic and hydrophilic forms of Zn2+-glycerophos-phocholine (GPC) cholinephosphodiesterase. Interest-ingly, the exposure of myelin membranes to detergentssuch as lysolecithin, Triton X-100 or deoxycholate gen-erated the hydrophilic form of enzyme. However, thestudy on the GPI-specific phospholipase responsible forthe conversion of membrane-bound proteins into hydro-philic form was not further investigated. In this study,we demonstrate that membrane (Mb)-phosphodiesteraseis converted to hydrophilic (Hyd)-phosphodiesterase bymembrane-bound GPI-phospholipase (GPI-PL) and GPI-PLD in the presence of endogenous detergents, and thatthe regulation of brain GPI-anchors degrading activitydepends on the type of detergents and substrates.

EXPERIMENTAL PROCEDURE

Materials. Concanavalin A sepharose, microbial phospholipase C(Bacillus cereus, type III), DEAE cellulose, sodium oleate and deter-gents including Triton X-100 were obtained from Sigma chemical Co..Triton X-114 and suramin were from Fluka Co. and RBI Co., respec-tively. Other reagents used were of analytical grade.

Preparation of Membrane-Bound GPI-PL and Soluble GPI-PLDfrom Bovine Brain. A fresh bovine brain, obtained from the localslaughter house, was extensively washed with 20 mM Tris buffer, pH7.4. The brain tissue (200 g) was homogenized in 1,800 ml of cold20 mM Tris buffer (pH 7.0), and then centrifuged (25,000 g, 30 min)at 4°C. The precipitate, after rehomogenization, was recentrifuged(25,000 g, 30 min). The final precipitate was suspended (150 ug pro-tein/ml) in 10 mM Tris buffer (pH 7.0), and the suspension was usedas the source of membrane-bound GPI-PL. For the preparation of sol-uble GPI-PLD, the supernatant from centrifugation of brain homogen-ate was passed through DEAE cellulose column (3.7 x 20 cm),equilibrated with 20 mM Tris buffer (pH 7.0). The column was firstwashed with 200 ml of equilibrium buffer, then with 200 ml of 20mM Tris buffer containing 50 mM NaCl, and finally, the GPI-PLDactivity was eluted with the same buffer containing 400 mM NaCl asdescribed before (6). The eluate, after centrifugation (25,000 g, 30min), was passed through concanavalin A sepharose column (0.7 X 3cm), and the final eluate was used as soluble GPI-PLD.

Preparation of Mb (Membrane-Bound)-Phosphodiesterase andAmp (Amphiphilic Form)-Phosphodiesterase. Mb-phosphodiesterasewas prepared by repeated homogenization and centrifugation of brainhomogenate as described above; the final membraneous precipitate,

suspended in 10 mM Tris buffer (pH 7.0), was used as Mb-phospho-diesterase. Meanwhile, Amp-phosphodiesterase was isolated from cy-tosol fraction of brain homogenate; the supernatant from thecentrifugation (25,000 g, 30 min) of brain homogenate was appliedonto DEAE cellulose column (3.7 X 20 cm), which was washed with20 mM Tris buffer (pH 7.0), and then the same buffer (200 ml) con-taining 50 mM NaCl. The Amp-phosphodiesterase activity was de-sorbed with the same buffer (100 ml) containing 400 mM NaCl. Thephosphodiesterase fraction was loaded onto concanavalin A-sepharosecolumn (0.7 X 3 cm) in the same way as reported (10) for Hyd-phosphodiesterase (hydrophilic form of phosphodiesterase). Finally,the bound phosphodiesterase was eluted with 20 mM Tris buffer (pH7.0) containing 0.5 M NaCl and 0.2 M a-methyl mannoside, and usedas a source of Amp-phosphodiesterase.

Assay of GPI-PLD Activity. The activity of soluble GPI-PLD,responsible for the conversion of Amp-phosphodiesterase to Hyd-phosphodiesterase, was determined by incubating soluble GPI-PLD(1.1 ug protein) and Amp-phosphodiesterase (1 milli unit) in 200 u1of 20 mM Tris buffer (pH 7.0) containing 0.015% Triton X-100 at38°C. One hour later, the reaction was stopped by addition of ice-cold2% Triton X-l 14 (800 u1), and the mixture was subjected to the assayof Hyd-phosphodiesterase activity according to Triton X-114 phaseseparation method (2,4); the mixture was incubated at 38°C for 5 min,and then centrifuged (14,000 rpm, 5 min). Aliquots (300 u1) of theaqueous phase was taken for the measurement of Hyd-phosphodies-terase. Separately, aliquots (100 ul) of detergent phase was subjectedto the assay of Amp-phosphodiesterase activity. The activity of solubleGPI-PLD was expressed as the amount (milli units) of Hyd-phospho-diesterase released under conditions used.

Assay of Membrane-Bound GPI-PL Activity. The activity of mem-brane-bound GPI-PL, responsible for the conversion of Mb-phospho-diesterase to Hyd-phosphodiesterase was measured by placing thebrain membrane (75 ug protein) in 500 ul of 10 mM Tris buffer (pH7.0) containing 0.3% deoxycholate at 38°C. After the designated time,the mixture was centrifuged (12,000 g, 10 min), and the aliquot (200u1) of supernatant was taken for the assay of the activity of Hyd-phosphodiesterase released according to Triton X-114 phase separationmethod. Separately, the conversion of Mb-phosphodiesterase intoHyd-phosphodiesterase was performed in the presence of various de-tergents as described in Figure legends. The activity ofmembrane-bound GPI-PL was expressed as the amount (milli units)of Hyd-phosphodiesterase released under conditions used.

Assay of Zn2+-GPC Cholinephosphodiesterase. The activity of thephosphodiesterase was determined by measuring the amount of p-ni-trophenol released from the hydrolysis of p-NPPC as described before(11); assays were conducted in 1.4 ml of 0.1 M bicarbonate buffer(pH 10) containing 100 uM p-NPPC. One unit of enzyme activity wasexpressed in umol of p-nitrophenol produced per min.

Measurement of Membrane-Bound GPI-PL or Soluble GPI-PLDActivity in the Presence of Detergents. The activity of membrane-bound GPI-PL was measured by placing the brain membrane (75 ugprotein) in 500 u1 of 10 mM Tris buffer (pH 7.0) containing variousconcentrations of detergents such as lysolecithin, monocaproyl glyc-erol or bile salts at 38°C. The release of Hyd-phosphodiesterase wasdetermined as described above. Separately, the activity of soluble GPI-PLD was determined in 200 u1 of 20 mM Tris buffer (pH 7.0) con-taining Triton X-100 of various concentrations (0-0.15%) at 38°C.

Inhibition of Membrane-Bound GPI-PL or Soluble GPI-PLD. Theactivity of membrane-bound GPI-PL or soluble GPI-PLD was deter-mined in the presence or absence of the respective candidate inhibitoras described in Figure legends. The activity of bound GPI-PL or sol-uble GPI-PLD was expressed as the amount (milli units) of Hyd-phos-

Enzymatic Release of Zn2+-Glycerophosphocholine Cholinephosphodiesterase 901

Fig. 1. Release of Hyd-phosphodiesterase from brain membranes bybio-detergent.Brain membrane (75 ug protein) was suspended in 500 ul of 10 mMTris, pH 7.0 containing 35 uM Zn2+ and each detergent (0.003-0.1%),and the mixture was incubated for 30 min at 38°C. The formation ofHyd-phosphodiesterase was measured according to phase separationassay, and Zn2+-GPC phosphodiesterase activity was determined asdescribed in Experimental Procedure.

phodiesterase released under conditions used. Inhibitors used weresodium oleate (0-12 mM), monomyristoyl glycerol (0-0.3 mM), laurylcholine (0-5 mM) and suramin (0-3 mM). IC50 value was indicatedas the concentration of each inhibitor to reduce the activity of control(no inhibitor) by 50%. Data, presented as milli units of Hyd-phospho-diesterase released, are the mean ± SD of two or three sets.

RESULTS

Previously, it had been reported that Zn2+-GPCCholinephosphodiesterase was solubilized from myelinmembranes during incubation with detergents. Longerincubation generated higher conversion of myelin mem-brane-bound phosphodiesterase, suggestive of involve-ment of myelin membrane-bound GPI-PL (10).

In an attempt to establish the release of Hyd-phos-phodiesterase from brain membranes, various detergentsat different concentrations (0.01-0.10%) were includedin the incubation mixture containing brain membranes,which were prepared from two successive centrifuga-tions of brain homogenate. As observed in Fig. 1,detergents used enhanced the release of the phospho-diesterase activity beyond the basal level in a concentra-tion dependent manner. Deoxycholate was more effectivethan lysolecithin or monocaproyl glycerol in releasing

Fig. 2. Release of Hyd-phosphodiesterase from brain membranes bybile acids.Brain membrane (150 ug protein / ml) was suspended in the incubationmixture including each bile acid (0.01-0.3%) at 38°C, and after 30min incubation the formation of Hyd-phosphodiesterase was deter-mined as described in Fig. 1.

Hyd-phosphodiesterase from brain membranes. How-ever, at lower concentrations (0.01-0.03%) lysolecithinseemed to be comparable to deoxycholate. Thus, the de-tergents used appeared to facilitate the conversion catay-zed by membrane-bound GPI-PL. Since, of three dif-ferent detergents, deoxycholate was the most effectivein maximizing the release, various types of bile saltswere examined for the ability to convert Mb-phospho-diesterase into Hyd-phosphodiesterase. As exhibited inFig. 2, dihydroxy acid was more effective than trihy-droxy acids (deoxycholic acid vs. cholic acid), and non-conjugated acid more efficient than conjugated acid(cholic acid vs. glycocholic acid). Noteworthy, glycoch-olate at 0.3% exhibited an inhibitory effect on the con-version. In further study using conjugate forms ofdeoxycholate, non-conjugated forms (deoxycholate andchenodeoxycholate) were more efficient than conjugatedforms (glycodeoxycholate and taurodeoxycholate).

Deoxycholate was chosen for further studies on theenzymatic conversion of Mb-phosphodiesterase intoHyd-phosphodiesterase. As demonstrated in Fig. 3,deoxycholate enhanced the conversion of Mb-phospho-diesterase into Hyd-phosphodiesterase in a dose-depen-dent fashion, but over 0.4%, the enhancement of activitywas not remarkable. Moreover, the enhancement ofdeoxycholate concentration over 0.5% interfered withthe phase separation assay. In an independent experi-

902 Lee, Kim, and Sok

Fig. 3. Conversion of Mb-phosphodiesterase to Hyd-phosphodiesterasein the presence of deoxycholate.Activity of membrane-bound GPI-PL responsible for the conversionof Mb-phosphodiesterase to Hyd-phosphodiesterase was measured byplacing brain membrane (75 ug protein) in 500 ul of 10 mM Trisbuffer (pH 7.0) containing deoxycholate of various concentrations(0.1-0.5%) at 38°C, and 30 min later, the formation of Hyd-phospho-diesterase was determined as described in Fig. 1.

Fig. 4. Conversion of Amp-phosphodiesterase into Hyd-phosphodies-terase in the presence of Triton X-100.Amp-phosphodiesterase(l milli unit) was incubated at 38°C with brainGPI-PLD(1.1 ug protein) in 200 ul of 20 mM Tris buffer, pH 7.0containing Triton X-100 of various concentrations (0-0.16%). The for-mation of Hyd-phosphodiesterase was measured as described in Fig.1.

ment, the soluble GPI-PLD-catalyzed conversion ofAmp-phosphodiesterase into Hyd-phosphodiesterasewas determined, and compared to the bound GPI-PL-catalyzed conversion. To facilitate the conversion ofAmp-phosphodiesterase into Hyd-phosphodiesterase,Triton X-100 was included in the incubation mixture asreported before (6). Fig. 4 demonstrates the effect ofTriton X-100 on the soluble GPI-PLD-catalyzed con-version of Amp-phosphodiesterase into Hyd-phospho-diesterase. Compared with the unincubated control,which contained approximately 6% Hyd-phosphodies-terase, a remarkable conversion was observed in thepresence of Triton X-100. Triton X-100 enhanced theconversion in a concentration-dependent manner up to0.04%, above which it exerted an inhibitory action.Next, the pH dependence and the time dependence forenzymatic conversion was examined. The conversion bysoluble GPI-PLD in the presence of 0.015% Triton X-100 was maximal around pH 6.6, while the bound GPI-PL-catalyzed conversion in the presence of either 0.3%deoxycholate or 0.1% lysolecithin was maximal aroundpH 7.4. Thus, the optimum pH for the enzymatic con-version differed between two enzyme systems. The time-dependence indicates that while the solubleGPI-PLD-catalyzed conversion increased continuously

up to 60 min, the bound GPI-PLD-catalyzed one in-creased with a linearity until 30 min, but over 30 minthe activity was not enhanced further.

In an attempt to find the possible regulation of theenzymatic conversion, endogenous lipid detergents suchas cholates, palmitoyl carnitine or lysolecithin weretested for the inhibition of the respective enzyme. WhenAmp-phosphodiesterase was incubated with solubleGPI-PLD in the presence of cholate, deoxycholate orglycocholate, the enzymatic conversion of Amp-phos-phodiesterase into Hyd-phosphodiesterase was inhibitedremarkably as exhibited in Fig. 5. The inhibitory potencywas in the following order; deoxycholate (IC50, 0.55mM) > cholate (IC50, 1.66 mM) > glycocholate (IC50,2.21 mM). Similarly (Fig. 6), the conversion of Amp-phosphodiesterase into Hyd-phosphodiesterase was alsoinhibited by lysolecithin (IC50, 0.15 mM) or palmitoylcarnitine (IC50, 0.1 mM). In addition, the activity wasalso inhibited by galactocerebroside (IC50, 0.7 mM).Thus, it is noteworthy that the soluble GPI-PLD-cata-lyzed conversion was inhibited by all detergents used.Of biodetergents used, palmitoyl carnitine was the mostinhibitory in the conversion of Amp-phosphodiesteraseinto Hyd-phosphodiesterase. Despite the remarkable in-hibition of soluble GPI-PLD action, all these endoge-nous detergents expressed no significant inhibition of

Enzymatic Release of Zn2+-Glycerophosphocholine Cholinephosphodiesterase 903

Fig. 5. Inhibitory effect of cholates on GPI-PLD activity.Assay of GPI-PLD activity was performed in the presence of the re-spective cholate derivative of various concentrations (0-0.1%) as de-scribed in Fig. 4.

Fig. 6. Inhibitory effect of palmitoyl carnitine or lysolecithin on GPI-PLD activity.Assay of GPI-PLD activity was carried out in the presence of eachdetergent of various concentrations (0.07—2.0 mM).

membrane-bound GPI-PL action (data not shown). Infurther studies, the inhibitory effect of other detergentsor suramin was examined (Table I). Although oleate, ananionic detergent, also expressed an inhibition of solubleGPI-PLD with an IC50 value of 0.6 mM, the conversion

Table I. Effect of Detergents or Suramin on GPI-PLD or MembraneBound GPI-PL Activity

Compound

Sodium oleateMonomyristoyl glycerolLaurylcholineSuramin

IC50, mM

Soluble GPI-PLD

0.6 ± 0.05activatory

1.5 ± 0.10.25 ± 0.03

Membrane boundGPI-PL

—activatory

—activatory

The effect of detergents or suramin on GPI-PLD or membrane-boundGPI-PL activity was determined as described in Fig. 5. — , no signif-icant inhibition at < 5 mM.

by membrane-bound GPI-PL was not prohibited remark-ably in the presence of the acid. Similarly, laurylcholineinhibited soluble GPI-PLD with an IC50% value of 1.5mM, but not membrane-bound GPI-PL. Next, the effectof monoacylglycerol, a neutral biodetergent, was ex-amined. Interestingly, monomyristoyl glycerol enhancedthe soluble GPI-PLD-catalyzed conversion slightly (10-19%) at the concentration range between 0.1 and 0.3mM. Meanwhile, a more remarkable enhancement (20-50%) by the neutral detergent at the same concentrationrange was observed with membrane-bound GPI-PL sys-tem. Additionally, suramin was tested for the inhibitionof brain GPI-PLD. Although a remarkable inhibition ofsoluble GPI-PLD was expressed by suramin with an IC50

value of 250 |oM, the compound rather enhanced theaction of the membrane-bound GPI-PL by approxi-mately 1.53-folds and 2-folds at 1 mM and 3 mM, re-spectively.

DISCUSSION

Present data indicate that in addition to the solubleGPI-PLD, there is another type of GPI-anchor degradingactivity, membrane-bound, responsible for the conver-sion of Mb-phosphodiesterase into Hyd-phosphodiester-ase, confirming the previous observation (10) thatGPI-anchor degrading activity bound to myelin mem-branes released the GPC-anchored Zn2+-GPC choline-phosphodiesterase from myelin membranes in thepresence of endogenous detergents. These may be con-sistent with earlier observations (8,9) with liver mem-branes ; a GPI-PLD activity was tightly bound to livermembranes, and the treatment of liver membranes withbile salts converted membrane-associated alkaline phos-phatase to the hydrophilic form of alkaline phosphatase.

Although at high concentrations (0.3%) deoxycho-late was much more effective than lysolecithin or mon-

904 Lee, Kim, and Sok

omyristoyl glycerol in facilitating the release ofHyd-phosphodiesterase from brain membranes by mem-brane-bound GPI-PL, the ability of release at lowerdoses (0.01-0.03%) was similar among detergents. Thehigher efficacy of deoxycholate , compared to other de-tergents, might be due to the higher capability to dis-sociate the amphiphilic form of proteins frommembranes. A small but significant release of Hyd-phos-phodiesterase by lysolecithin at lower concentrationsmight be ascribed to its ability to form micelles at lowerconcentrations. Such a detergent effect might be in-volved in the in vivo release from intact cell membranes.However, the physiological concentration of lysolecithinat normal condition might limit its role as a possibleenhancer of membrane-bound GPI-PL activity.

Most of the studies on the action of GPI-PLD em-phasized the release of hydrophilic form from amphi-philic form of protein rather than from membrane-boundprotein, since the conversion rate was much greater withthe amphiphilic form. Moreover, in these studies, syn-thetic non-ionic detergents such as Triton X-100 or Non-idet p-40 were routinely used for the facilitation of theconversion. A recent study with liver alkaline phospha-tase indicates that the conversion of liver membrane -bound alkaline phosphatase to hydrophilic form waspossible at a small range (2 mM) of bile salt concentra-tion (9). This may be somewhat different from our resultthat the effect of deoxycholate on the conversion ofMem-phosphodiesterase into Hyd-phosphodiesterase in-creased up to 10 mM. This might be ascribed to thedifference in the property of substrates or the type ofbile salts used. Alternatively, the difference of propertiesbetween liver membrane-bound GPI-PLD and brainmembrane-bound GPI-PL might be another reason. Theability of detergents to release the Hyd-phosphodiester-ase from membranes differed according to the type ofcholates; in general, cholates with smaller CMC valuesshowed higher release effect. However, despite theirsmaller CMC values, conjugated forms of cholates wereless effective than non-conjugated cholates. Moreover,glycine conjugate of cholates seemed to show an inhib-itory action at a higher concentration. This led to thenotion that the phospholipase responsible for the con-version of Mb-phosphodiesterase into Hyd-phosphodies-terase might be inhibited by conjugated forms ofcholates. Earlier observers reported that cholates inhib-ited serum GPI-PLD (12). Similarly, cholates also inhib-ited brain GPI-PLD in our study, although the inhibitorypotency of cholates was relatively smaller for brain GPI-PLD, compared to serum GPI-PLD. However, it is notself-explanatory to observe that the order for the inhib-itory potency of cholate derivatives was in parallel with

the order for their ability to convert Mb-phosphodiester-ase to Hyd-phosphodiesterase. Therefore, the identitybetween soluble GPI-PLD and membrane-bound GPI-PL was questioned.

Inhibition studies indicate that brain GPI-PLD wasinhibited by either anionic detergents such as cholatesand oleate or zwitter ion detergents such as lysolecithinor palmitoyl carnitine. Even laurylcholine, a cationic de-tergent, inhibited the enzyme at millimolar concentra-tions. In addition, suramin also prohibited the brainGPI-PLD activity. These are somewhat consistent withearlier observations that serum GPI-PLD was inhibitedby ionic detergents (12) or suramin derivatives (13).Noteworthy, oleate or palmitoyl carnitine expressed aninhibitory role at possible physiological concentrations.Meanwhile, monomyristoyl glycerol expressed an en-hancing effect at concentrations used. Thus, the role ofdetergents differed according to their chemical property.The living system contains various lipid detergents suchas lysolecithin, palmitoyl carnitine, oleate, etc.. Since to-tal concentration of endogenous lipid detergents couldbe close to an inhibitory concentration, it is supposedthat GPI-PLD might be regulated in in vivo system.

Contrary to GPI-PLD, the membrane-bound GPI-PL was not affected remarkably by endogenous deter-gents. There was a difference in the optimum pH andthe time-dependence for activity between GPI-PLD andmembrane-bound GPI-PL. Moreover, the addition ofGPI-PLD to the system for the assay of membrane-bound GPI-PL failed to further enhance the conversionby membrane-bound GPI-PL. Additionally, the inclusionof Triton X-100 did not enhance the bound GPI-PL—catalyzed release of Hyd-phosphodiesterase from mem-branes in the presence of deoxycholate (unpublisheddata). These data suggest that membrane-bound GPI-PLmay be a GPI-specific phospholipase different from asoluble GPI-PLD (6). Although the membrane-boundGPI-PL was not purified here, it is possible to supposethat the enzyme may correspond to a new type of GPI-PLD, since the HYD-phosphodiesterases produced bysoluble GPI-PLD and membrane bound GPI-PL showedthe same chromatographic behavior on CM sephadexcolumn. Taken together, it is possible to assume that theGPI-anchored Amp-phosphodiesterase can be releasedfrom membranes exposed to endogenous detergents, andthen the amphiphilic form can be further transformedinto hydrophilic form in the presence of endogenousGPI-PL or GPI-PLD as proposed in Scheme 1. There isa difference of properties between Amp-phosphodiester-ase and Hyd-phosphodiesterase; whereas the former en-zyme can associate with an anionic exchanger resin inthe medium containing 50 mM NaCl, the latter form

Enzymatic Release of Zn2+-Glycerophosphocholine Cholinephosphodiesterase 905

Scheme 1. Proposed model for the interaction of GPI-specific phos-pholipases with Mb-phosphodiesterase and Amp-phosphodiesterase.

binds to a cationic exchanger resin (10,14). Moreover,membrane-bound enzyme was observed to be more sta-ble than hydrophilic form (14). Thus, it is suggested thatthe enzymatic conversion of membrane-bound phospho-diesterase to hydrophilic form may lead to an alterationof Zn2+-GPC Cholinephosphodiesterase activity.

ACKNOWLEDGMENT

This work was financially supported by research grants from theKorea Science and Engineering Foundation (9610719111-2), KOREA.

REFERENCES

1. Low, M. G. 1989. The glycosylphosphatidylinositol anchor ofmembrane proteins. Biochem. J. 250:865-869.

2. Low, M. G., and Prasad, A. R. S. 1988. A phospholipase D spe-cific for the phosphatidylinositol anchor of cell-surface proteins isabundant in plasma. Proc. Natl. Acad. Sci. USA. 85:980-984.

3. Toutant, J-P., Roberts, W. L., Murray, N. R., and Rosenberry, T.L. 1989. Conversion of human erythrocyte acetylcholinesterasefrom an amphiphilic to a hydrophilic form by phosphatidylinosi-tol-specific phospholipase C and serum phospholipase D. Euro. J.Biochem. 180:503-508.

4. Low, M. G., and Huang, K. S. 1991. Factors affecting the abilityof glycosylphosphatidylinositol-specific phospholipase D to de-grade the membrane anchors of cell surface proteins. Biochem. J.279:483-493.

5. Davitz, M. A., Horn, J., and Schenkman, S. 1989. Purification ofa glycosylphosphatidylinositol-specific phospholipase D from hu-man plasma. J. Biol. Chem. 264:13760-13764.

6. Hoener, M. C., Stieger, S., and Brodbeck, U. 1990. Isolation andcharacterization of a phosphatidylinositol-giycan-anchor-specificphospholipase D form bovine brain. Eur. J. Biochem. 190:593-601.

7. Rhode, H., Eva, H. B., Schiling, K., Gehrhardt, S., Gohlert, A.,Buttner, A., Bublitz, R., Cumme, G. A., and Horn, A. 1995. Gly-cosylphosphatidylinositol-alkaline phosphatase from calf intestineas substrate for glycosylphosphatidylinositol-specific phospholi-pases-microassay using hydrophobic chromatography in pipet tips.Anal. Biochem. 231:99-108.

8. Kominami, T., Oda, K., and Ikehara, Y. 1984. Induction of rathepatic alkaline phosphatase and its appearance in serum : elec-trophoretic characterization of liver-membranous and serum-sol-uble forms. J. Biochem. 96:901-911.

9. Deng, J. T., Hoylaerts, M. F., De Broe, M. E., and Van Hoof, V.0. 1996. Hydrolysis of membrane-bound liver alkaline phospha-tase by GPI-PLD requires bile salts. Am. J. Physiol. 271 : 655-663.

10. Sok, D. E., and Kim, M. R. 1993. Brain myelin-bound Zn2+-gly-cero—phosphocholine Cholinephosphodiesterase is a glycosyl-phosphatidylinositol anchored enzyme of two different molecularforms. Neurochem. Res. 19:97-103.

11. Sok, D. E., and Kim, M. R. 1992. A spectrophotometric assay ofZn2+-glycerylphosphorylcholine phosphocholine phosphodiester-ase using p-nitrophenylphosphocholine. Anal. Biochem. 203:201-205.

12. Low, M. G., and Huang, K. S. 1993. Phosphatidic acid, lyso-phosphatidic acid, and lipid A are inhibitors of glycosylphospha-tidylinositol-specific phospholipase D. J. Biol. Chem. 268:8480-8490.

13. Burruner, G., Zalikow L., Burgess, E., Rifkin, D. B., Wilson E.L., Gruzeckakowalik, E., and Powis G. 1996. Inhibition of gly-cosylphosphatidylinositol (GPI) phospholipase D by suramin—like compounds. Anticancer Res. 16(5A):2513-2516.

14. Sok, D.-E. 1996. Properties of a Zn2+- glycerophosphocholineCholinephosphodiesterase from bovine brain membranes. Neuro-chem. Res. 21(10):1193-1199.