Purification and Characterization of Two Types of Soluble Inositol ...

Vol,262,NO.36,Issue of December 25,PP. 17319-17326,198’7 Printed in V.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society for Biochemistry and Molecular Biology, Inc.

Purification and Characterizationof Two Types of Soluble Inositol Phosphate 5-Phosphomonoesterases from Rat Brain* (Received for publication, April 27, 1987)

Carl A. HansenS, Roy A. Johanson, Michael T. Williamson, and John R. Williamson$ From the Department of Biochemistry and Biophysics, University of Pennsylvania School of Medicine, .~ Philodelphiai Pennsylvania 19104

Ins(1,4,5)P3,1 the second messenger for intracellular Ca2+ mobilization (reviewed in Refs.1-3). Degradation of this second messenger was originally thought to be straightforward. Ins(1,4,5)P3 was shown to be metabolized by removal of the 5-phosphate (4) andthe Ins(1,4)P2produced was incapable of inducing release of intracellular sequestered Ca2+(1). Ins(l,4)P2 is then degraded to Ins(4)P, and subsequently to myo-inositol (5, 6). However, recent studies have demonstrated that, in addition to dephosphorylation to Ins(l,4)P2, Ins( 1,4,5)P3can also be phosphorylated by an ATP-dependent Ins(1,4,5)P3 3-kinase to Ins(1,3,4,5)P4(7, 8), which is subsequently hydrolyzed to a distinctinositol trisphosphate isomer, Ins(1,3,4)P3(8-11). The physiological significance of this metabolic branch point at the level of Ins(1,4,5)P3 is not currently known. It may simply represent amechanism for control of Ins(1,4,5)P3 concentration, or, alternatively the Ins(1,3,4,5)P4formed may also function as an intracellular messenger (7-9, 12-14). In either case, the enzymes involved at thebranch point of this pathway are likely to be highly regulated. Interestingly, both Ins(1,4,5)P3 and Ins(1,3,4,5)P4 are dephosphorylated specifically at the 5 position (8-11, 15, 16). Whether the same or separate 5-phosphomonoesterases metabolize Ins( 1,4,5)P3and Ins(1,3,4,5)P4in vivo remains to be determined. Both soluble and particulate forms of inositol phosphate 5-phosphomonoesterase have been identified and the distribution of these forms is known to vary from tissue to tissue. In erythrocytes (4), liver (5, 17, 18), macrophages (19), and brain (20), the major activity is in the particulate fraction, whereas in platelets (21) and coronary artery smooth , muscle (22), the principle activity issoluble. The soluble platelet enzyme has been purified to apparent homogeneity (21) and shown to be phosphorylated by protein kinase C, which increased the VmaXof the enzyme without affecting its K,,, (23). Recently, the soluble platelet enzyme has also been shown to hydrolyze Ins(1,3,4,5)P4 (11).The relationship of the soluble platelet enzyme to the soluble enzymes in other tissues, however, is not clear, nor is the relationship of the soluble enzymes to the membrane-bound form known. In this study, we have examined the soluble Ins(1,4,5)P3 A primary response to receptor activation by Ca2+-mobiliz- and Ins( 1,3,4,5)P45-phosphomonoesterase activity from rat ing hormones is the rapid hydrolysis of phosphatidylinositol brain. Two distinct enzymes were found and theirpurification 4,5-bisphosphate by a specific phospholipase C, producing and kinetic characterization are described.

Two soluble forms of inositol phosphate 5-phosphomonoesterase have been partially purified and characterized from rat brain and are referredto as type 1 and type 2 according to their order of elution from DEAE-Sepharose. Together, these enzymes represent 26 A 3%(mean f S.E., n = 4) of the totalinositol 1,4,5triphosphate (Ins( 1,4,5)P3) phosphatase activity assayed in crude brain homogenate and are present in approximately equal total activities in a 100,000 x g supernatant,withtheremainder being membranebound. Bothsoluble enzymes require Mg2*for activity, are moderately inhibited by Ca2+ in the micromolar range, and can be inhibited by millimolar concentrations of a variety of phosphorylated compounds. The type 1 enzyme has been purified to a specific activity of 1.06 pmol/min/mg protein. It elutes as a 60-kDa protein on Sephacryl 5-200. On sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the type 1enzyme correlates witha pair of protein bandsof 66 and 60 kDa. It has apparentK , values of 3 and 0.8 p~ for Ins(l,4,5)P3and inositol 1,3,4,5-tetrakisphosphate (Ins(1,3,4,5)P4),respectively,andhydrolyses Ins(1,4,5)P3 approximately 12 times faster than Ins(1,3,4,5)P4. The type enzyme 2 has been purified to a specific activity of 15.2 wmol/min/mg protein, elutes as a protein of 160 kDa on SephacrylS-300,and migrates as a similarly sized subunit on sodiumdodecyl sulfate-polyacrylamide gel electrophoresis. It has an apparent K , for Ins(1,4,5)P3 of 18pM. Its apparent K , for Ins(1,3,4,5)P4, however, is greater than 150 p ~ suggesting that this enzyme is primarily an Ins(l,4,5)P3 5-phosphomonoesterase. The relationship of these two enzymes to the inositol trisltetrakisphosphate pathwayis discussed.

* This work was supported by National Institutes of Health Grant DK-15120 ( t o J. R. W.) and DK-19525 ( t o C.A.H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement’’ in accordance with 18U.S.C. Section 1734 solelyto indicate this fact. 3 Recipient of a National Research Service Award HL-07146. § T o whom correspondence and reprint requests should be addressed.

’ The abbreviations used are: Ins(1,4,5)P3,inositol 1,4,5-trisphosphate; Ins(l,4)P2, inositol 1,4-bisphosphate; Ins(1,3,4,5)P4, inositol 1,3,4,5-tetrakisphosphate;Ins(1,3,4)P~,inositol 1,3,4-trisphosphate; Hepes, N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid; Mes, 2-(N-morpholino)ethanesulfonicacid; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; Mops, 3-(N-morpholino)propanesulfonic acid; Hepps, N-2-hydroxyethylpiperazine-N’-3propanesulfonic acid; HPLC, high performance liquid chromatography; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraaceticacid.

17319

17320

Soluble Inositol5-Phosphomonoesterases Phosphate

EXPERIMENTALPROCEDURES Materiafs-Ins(l,4,5)P3 (free of inositol 2,4,5-trisphosphate) was obtained from Behring Diagnostics. [3H]Ins(1,4,5)P3was from New England Nuclear. Ins(1,3,4,5)P4and [3H]Ins(1,3,4,5)P4were synthesized as described by Cerdan et al. (24). Bio-Gel HTP-hydroxyapatite, Dowex AG 1-X8 (formate form),and polyacrylamide electrophoresis materials were from Bio-Rad. ACS I1 waspurchased from Amersham Corp. All other materialswere purchased from either Sigma or Fisher Scientific. Purification of Soluble Inositol Phosphate 5-Phosphomonoesterases-Rats were killed by rapid decapitation, the brains removed, and stored frozen a t -20 "C.All purification steps were carried out below 4'C. Thirty rat brains werehomogenized (l/lO, w/v) in 25 mM Hepes, 50 mM Tris, 1 mM MgCI,, 2 mM EGTA, 1 mM dithiothreitol, pH 8.0, containing 2 pg/ml each of antipain, aprotinin, leupeptin, and pepstatin A by 15 passes of a glass Teflon homogenizer running at 625 rpm. The homogenate was centrifuged a t 12,000 X g for 10 min to remove insoluble debris and the supernatant was recentrifuged a t 100,000 x g for 45 min to obtain a clear, membrane-free supernatant. PhosphocelluloseChromatography-The clear supernatant was stirred for 2.5 h with 17 g of phosphocellulose equilibrated with 10 mM Tris-C1, 10 mM KCl, and 1 mM dithiothreitol, at pH 7.6. After the phosphocellulose had settled, the supernatant was decanted and the slurry was poured into a 5-cm diameter column and allowed to pack under gravity. The column was eluted at 250 ml/h with a 1000ml linear gradient from equilibration buffer to 500 mM of potassium phosphate, pH 6.9, and 1 mM dithiothreitol. DEAE-Sepharose GB-CL Chromatography-The peak fractions of enzyme activity from the phosphocellulose column were pooled and dialyzed overnight against 4000 ml of buffer containing 12 mM Tris, 14 mM Mes, 1 mM MgCI,, and 1 mM dithiothreitol, pH 6.8, which will subsequently be referred to ascolumn buffer. The dialysate was centrifuged at 12,000 X g for 10 min to sediment a light precipitate and the supernatant was applied a t 3 ml/min to a 2.5 X 30-cm column of DEAE-Sepharose equilibrated with column buffer. The column was eluted with a 450-ml linear gradient from 0 to 300 mMKC1 in column buffer. ATP-AgaroseAffinity Chromatography-The pooledpeak fractions from the DEAE-Sepharose column were dialyzed overnight against 2000 ml of column buffer and applied at 0.8 ml/min to a 1 X 18-cm column of ATP-agarose (attached through C-8 with a 6-carbon spacer) equilibrated with column buffer, and subsequently eluted with a 150-ml linear gradient from 0 to 400 mM KC1 in column buffer. Hydroxylapatite Chromatography-The fractions containing 5phosphomonoesterase activity from the ATP-agarose column were pooled and applied a t 0.8 ml/min to a 1 X 5-cm column of Bio-Gel HTP hydroxylapatite and subsequently eluted with a 40-ml linear gradient from 0 to 500 mM potassium phosphate in column buffer. Size ExclusionChromatography-The pooled enzyme from the hydroxylapatite column (5-6 ml) was applied to either a 1.6 X 80-cm column of Sephacryl S-200 or a 1.5 X 90-cm column of Sephacryl S300, each equilibrated with column buffer supplemented with 100 mM KC]. Flow rates were 8 ml/h. Columns were calibrated using apoferritin (443 kDa), cy-amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), carbonic anhydrase (29 kDa), and myoglobin (13.5 kDa). The fractions containing enzyme were pooled, divided into aliquots, and stored frozen a t -20 "C. Gel Electrophoresis-SDS-PAGE was carried out according to the method of Laemmli (25) employing either 7% or 10% gels. The gels were stained for protein by the silver stain method of Merril (26) using a Bio-Rad silver stain kit. Assay of Inositol Phosphate 5-PhosphomonoesteraseActiuity-The standard assay consisted of 50 mM Mops, pH 7.1,l mM MgC12,l mM dithiothreitol, 0.5% bovine serum albumin, and 30 p M of Ins(1,4,5)P3 (5000 dpm/nmol). Assays were carried out at 30 "C, initiated by the addition of enzyme and stopped at various times by the addition of formic acid (final concentration of 0.1 M). The [3H]Ins(1,4)P2formed was separated from [3H]Ins(l,4,5)P3 onminicolumns (0.5 X 0.5 cm) of fast-flow QAE-Sepharose. After sample loading, [3H]Ins(1,4)P~ was eluted with three 0.7-ml washes of 125 mM potassium phosphate, pH 3.7. [3H]Ins(1,4,5)P3was subsequently eluted with three 0.7-ml washes of 300 mM potassium phosphate, pH 3.7. Eluents were mixed with 15 ml of ACS I1 and theradioactivity was quantified by liquid scintillation counting. The assay conditions for Ins(1,3,4,5)P4 5-phosphomonoesterase were identical except for addition of 10 pM [3H]Ins(1,3,4,5)P4(15,000

dpm/nmol) as substrate. [3H]Ins(1,4,5)P3was separated from [3H] Ins(1,3,4,5)P4 on minicolumns ofDowexAG 1-X8 (0.5 X 0.8 cm). After loading the sample, [3H]Ins(1,4,5)Pswas eluted from the column with three 2-ml washes of 800 mM ammonium formate, 100 mM formic acid, followed by one 2-ml wash of 1.5 M ammonium formate, 100 mM formic acid to elute [3H]Ins(1,3,4,5)P4.Radioactivity was quantified as above. To determine the pH dependence of the enzymes, 25 mM Mes, 25 mM Mops, and 25 mM Hepps were substituted for the 50 mM Mops utilized in the standard assay buffer and the pH was adjusted between 6 and 9. The free Ca'+ in the buffer was adjusted by addition of 1 r n ~ EGTA and various amounts of CaC12 to obtain the desired free Ca2+ concentration, as determined using a Ca2' electrode (27). In studies where the free M e concentration was varied, 1 mM EDTA and various amounts of MgClz were added to the standard assay buffer and the free M e concentration was calculated using a dissociation constant of 3.3 X lo-' M for MgEDTA (28). Free ATP concentrations were calculated from the total Mg2+ and total ATP concentrations using a dissociation constant of 0.022 mM for MgATP at pH 7.1 (29). Products of Inositol Phosphate 5-Phosphomonoesterase-The product(s) of the 5-phosphomonoesterases were determined using HPLC as described in Ref. 8, and were identified by comparing the retention times of the radiolabeled products with those of known inositol phosphate isomers. Protein Concentration-Protein concentrations were determined by the method of Bradford (30), using Bio-Rad bovine y-globulin protein standard. RESULTS

Purification of Soluble Inositol Phosphate 5-Phosphomonoesterases After centrifugation of the homogenate at 100,000 X g for 45 min, 26 k 3% (mean f S.E., n = 4) of the inositol phosphate phosphatase activitywas retained in the supernatant (assayed at 100 pM Ins(1,4,5)P3), with the remainder sedimenting with the pellet. This percentage of soluble activity was approximately twice that reported by Erneuxet al. (20) for rat brain and is most likely the resultof the subsaturating concentration of Ins(1,4,5)P3 used in their study. The total enzyme activity at this stage (soluble particulate) was the sameas that assayed in the crude extract. Total enzyme activityand its distribution was similar when either fresh or frozen brains were used. The steps utilized to purify the 5-phosphomonoesterases from rat brain are summarized in Table 1. The initial step (phosphocellulose chromatography) was developed in conjection with the purification of another enzyme, Ins(1,4,5)P3 3kinase.' Subsequent steps, however, were adapted from the purification scheme reported for the soluble 5-phosphomonoesterase from platelets ( 2 1 ) . After loading the 100,000 x g supernatant onto the phosphocellulose column, Ins(1,4,5)P3 phosphatase activity was elutedas a single peak at approximately 200 mM potassium phosphate (Fig. 1). Analysis of the e n z y m er e a c t i o np r o d u c t ( s )b y HPLC s h o w e d t h a t Ins(1,4,5)P3 was converted to a singleproducthaving an identical retention timeas that of authentic Ins(l,4)Pp, indicatingtheactivitywasindeed a 5-phosphomonoesterase. Fractions containing Ins( 1,3,4,5)P4 phosphatase activity corresponded exactly with those containing the Ins(1,4,5)P3 5phosphomonoesterase activity. Again, analysis of the reaction products demonstrated the formation of only one product, which had a retention time on HPLC identical to authentic Ins(1,3,4)P3 characterized by31Pand 'H NMR? All of t h e 5phosphomonoesterase activity inthe 100,000 X g supernatant was retained bythe column, whereas Ins( 1,3,4)P3, Ins(l,4)P2, and Ins(4)P phosphatase activities did not bind. The phos-

+

R. A. Johanson, C. A. Hansen, K. E. Coll, and J . R. Williamson, submitted for publication. C. A. Hansen and J. R. Williamson, submitted for publication.

17321

Soluble Inositol Phosphate 5-Phosphomonoesterases TABLE I Purijication of soluble rat brain inositol phosphate 5-phosphomonoesterases Thirty rat brains were used. Ins(1,4,5)P3was used as substrate at a concentration of 30 and 150 p M for type 1 and type 2 enzymes, respectively. Total Volume

Step mg

136.0

ml

100,000 X g supernatant postdialysis Phosphocellulose

340.0 100

Type 1" DEAE-Sepharose ATP affinity Hydroxylapatite Sephacryl S-200

43.5 1.50 28.6 0.58 4.8 0.11 11.5

Total protein

activity

Specific activity

rmollminlmg pmollmin

2883 0.014

0.6 0.1

Purification

%

-fold

41.6

4.8 1.2

Recovery

1.01

1

23 0.314 0.840 0.970 1.060

3.6 2.4 1.4

0.3

Type 2" 8.4 10.2 1.21 24.5 59.5 DEAE-SeDharose 2.82 11 28.6 4.3 1.5 ATP affiity 0.9 2.6 2.81 6 5.8 Hydroxylapatite 15.2 3 11.5 0.1 1.3 Sephacwl S-300 "Inositol phosphate5-phosphomonoesteraseactivity was fractionatedinto twoactivitypeaksbyDEAESepharose chromatography,the first activity peak being designated type1 and the second, type 2.

0

20

40

60 69 76 86

201 201

1085

60

Fraction Number

FIG. 1. Chromatography of solubleinositol phosphate 5phosphomonoesterase on phosphocellulose. The 100,000 X g supernatant (340 ml from 30 rat brains) was applied and eluted with a 1000-ml linear gradientfrom equilibration bufferto 500 mM potassium phosphate, pH 6.9, and1 mM dithiothreitol. The flow rate was 250 ml/h; the fraction size was 13.5 ml. The assay substrate was 30 p~ Ins(1,4,5)P3.

Fraction Number FIG. 2. Chromatography of soluble inositol phosphate 6phosphomonoesterase on DEAE-Sepharose f3B-CL. The dialyzed pooled 5-phosphomonoesterasefraction elutingfrom the phosphocellulose column was applied to a DEAE-Sepharose column and eluted witha 450-ml linear gradientfrom 0 to 300 mM KC1 in column buffer. The flow rate was 3 ml/min; the fraction size was 5 ml. The assay substrate was 30 p~ Ins(1,4,5)P3. Two 5-phosphomonoesterase phocellulose column also separated the 5-phosphomonoester-activity peaks were observed and are referred to in their order of ase from Ins(1,4,5)P3 kinase,which eluted at a higher potas- elution astype 1 and type 2. sium phosphate concentration.* Fractions containing 5-phosphomonoesterase activity fromreferred to as type 1 and type2 5-phosphomonoesterases, type the phosphocellulose step were pooled. Following a n overnight 1 being the first peak to be eluted with KC1 from DEAEdialysis, the total activity increased approximately 35%, which Sepharose and type2 being the second peak. Fractions containing either the type 1or type2 5-phosphowas probably caused by the removal of inorganic phosphate, a known inhibitor of this enzyme (21). The pooled dialyzed monoesterase were pooled, dialyzed overnight againstcolumn enzyme was applied to a DEAE-Sepharose column. Elution buffer, and then chromatographed on an ATP-agarose affinity with KC1 resulted in two distinct peaks of Ins(1,4,5)P3 5- column. This step was included in the purification scheme phosphomonoesteraseactivity (Fig. 2). Theproportion of after it was found that ATP inhibited both enzymes (see Ins(1,4,5)P3 5-phosphomonoesterase in each peakvaried be- below). The elution of type 15-phosphomonoesterase is shown tween different preparations, with the first peak containing in Fig. 3. The type 2 enzymechromatographed similarly (data between 10 and 20% of the total activity. Interestingly, the not shown), but eluted at a slightly higher KC1 concentration. first peakrapidly degraded Ins( 1,3,4,5)P4,whereas, the second Fractions containing each 5-phosphomonoesterase were peak appeared t o have poor activity towards Ins(1,3,4,5)P4. pooled and directlyapplied to acolumn of Bio-Gel HTP The two 5-phosphomonoesterasepeaksfromtheDEAEhydroxylapatite. This matrix was used essentially as a conSepharose column chromatographed quite differently on size centration step prior tosize exclusion chromatography. Both exclusion chromatography (see below).Therefore, theenzyme the type 1 and type 2 enzymes eluted at approximately 300 activity in each peakwas further purified separately, and are mM potassium phosphate(data not shown),which was similar


17322

A

B

C

D

E

F

9467-0

30

IO

50

W

43-

Fraction Number FIG.3. Chromatography of soluble type 1 inositol phosphate 5-phosphomonoesterase on ATP-agarose. The dialyzed pooled 5-phosphomonoesterase fraction from the DEAE-Sepharose column was applied to the ATP-agarose column and eluted with a 150-ml linear gradient from 0 to 400 mM KC1 in column buffer. The flow rate was 0.8 ml/min; fractions 1-10 were 13 ml and subsequent fractions were 2.6 ml. The assay substrate was 30 mM Ins(1,4,5)P3. Chromatography of the type 2 enzyme on this matrix was performed identically and the activity and A m profiles were similar.

30 -

20 V "

1

I

1

1 0

m

N

0 0

-

-

-

FIG.6. Silver-stained 10% SDS gel of type 1 inositol phosphate5-phosphomonoesterasefromthe Sephacryl 5-200 chromatography step. Column lettering refers to fractions taken across the enzyme activity peak of Fig. 4.

0.04 0.03 0.02 0.01

A B C D E F

0

,O ELUTIONVOLUME

(ml)

FIG.4. Size exclusion chromatography of soluble type 1 inositol phosphate 5-phosphomonoesterase on Sephacryl S200. The pooled type 1 enzyme fraction from the hydroxylapatite column was applied to a Sephacryl S-200column and eluteda t a flow rate of 8 ml/h; fraction size was 2.3 ml. The assay substrate was 30 p~ Ins(1,4,5)P3. Protein standards were run either individually or in Dairs. Letters refer to the fractions analyzed on SDS-PAGE shown in

205 -

116 -

97 1

1 0 al N

D

0

0.03

67 -

0.02 0.01

0 D ELUTIONVOLUME

45 -

(ml)

29 -

FIG.5. Size exclusion chromatography of soluble type 2 inositol phosphate 5-phosphomonoesterase on Sephacryl S300. The pooled type 2 enzyme fraction from the hydroxylapatite FIG.7. Silver-stained 7% SDS gel of type 2 inositol phoscolumnwas applied to a Sephacryl S-300 column and eluted and phate5-phosphomonoesterasefromthe Sephacryl 5-300 assayed as in Fig. 4. Letters refer to the fractions analyzed on SDS- chromatography step. Column lettering refers to fractions taken PAGE shown in Fig. 7. across the enzyme activity peak of Fig. 5.

5-Phosphomonoesterases Phosphate Inositol Soluble

17323

I .25-

F I .oo-

0

“.

c ._

E

2 0.750 1 . c .>

a

0.25 -0.5

0.5

I

r

OO

1

I

1

2

4

6

8

IO

FIG. 8. Substrate velocity relationship of soluble type 1 inositol phosphate 5-phosphomonoesterase. The enzyme utilized was from the Sephacryl S-200 pool and the assays were carried out as described under “Experimental Procedures” except the substrate A , Ins(1,4,5)P3 as substrate. B , concentrationwasvaried. Ins(1,3,4,5)P4as substrate.

was observed on Sephacryl s200, this step was important in removing several proteins that remained following the ATPagarose column. The specific activity of the enzyme following this final purification step was 1.06 pmol/min/mg protein (Table I). The stability, yield, and finalspecific activity of the type 1 enzyme was similar to that reportedfor purification of the soluble platelet enzyme (21). The pooled type 2 enzyme eluting from the hydroxylapatitecolumn was applied to a size exclusion column of Sephacryl S-300 and subsequently eluted as a single peak with a M , of 160,000 (Fig. 5). The specific activity of type 2 5-phosphomonoesterase following this purification step was 15.2 pmol/min/mg protein (Table I). In neither case was there any indication of a second activity peak on these size exclusion columns corresponding to the other typeof 5-phosphomonoesterase. Both purified enzymes were stored at -20 “C. The type 2enzyme was relatively stable under these conditions, however, the type 1 enzyme lost approximately 10%of its activity perweek. Fig. 6 showsthe SDS-PAGE protein banding pattern across the Sephacryl S-200 activity peak of the type 1 enzyme, the lettered columns corresponding to the letters marked on Fig. 4. Type 1 activity begins in column B , peaks in column C, and decreases through columns C and D. Two protein bands, at 66 and 60 kDa, respectively, correlated with this activity profile. The relationshipof these two protein bands is not yet enzyme is likely a monomer, clear, but indicate that the 1 type since the subunit molecule weighed from SDS-PAGE agrees with the native M , of 60,000 obtained from size exclusion chromatography. SDS-PAGE across the Sephacryl S-300 activity peak of the type 2 enzyme (Fig. 5) is shown in Fig. 7. No type 2 activity was detected in column A , but rapidly increased in B, with activity similar through C and D and finally decreasing through E. The protein band at 160 kDa correlated with activity profile, indicating that the type 2 enzyme is also a monomer.

Properties of Soluble Inositol Phosphate 5-Phosphomonoesterases Substrate Velocity Relationships-The relationship of Ins(1,4,5)P3 and Ins(1,3,4,5)P4 concentrationsreaction to velocity for the type l enzyme is shown in Fig. 8. The K,,, for Ins(1,4,5)P3was 3.0 f 0.2 pM (mean f S.E., n = 3) and for Ins(1,3,4,5)P4,0.8 +. 0.1 FM (mean f S.E., n = 3). However, the apparent V,,, with Ins( 1,4,5)P3 (1.1 0.2 pmol/min/mg, mean f S.E., n = 3) was approximately 12-fold greater than that observed with Ins(1,3,4,5)P4(0.09 f 0.02 pmol/min/mg, mean f S.E., n = 3). Thetype 2enzyme had a K,,, for Ins(1,4,5)P3 of 18.3 f 2.4 ptM (mean f S.E., n = 3) (Fig. 9), but its K,,, for Ins(1,3,4,5)P4was greater than150 pM and was not further characterized. Neither enzyme hydrolyzed either ATP or para-nitrophenyl phosphate. The activity of each enzyme in the 100,000 x g supernatant Ins(1.4,5)PS(W) was estimated by comparing the dephosphorylation rates of FIG.9. Substrate velocity relationship of soluble type 2 inositol phosphate 5-phosphomonoesterase. The enzyme utilized 10 p M Ins(1,3,4,5)P4 and 150 PM Ins(1,4,5)P3. At 10 p~ was from the Sephacryl S-300 pool and the assays were carried out Ins(1,3,4,5)P4, the type 1 enzyme would be nearly saturated. as described under “Experimental Procedures” except the concentra- However, the type 2 enzyme would be at least 15-fold below tions of substrate, Ins(1,4,5)PZ,was varied. its K,,, and, assuming Michaelis-Menton kinetics, would be operating at less than 6% of its V,,,. Therefore, at this low concentration of Ins(1,3,4,5)P4, the rate of Ins(1,3,4,5)P4deto the behavior of the platelet enzyme on hydroxylapatite phosphorylation should primarily reflect type 1activity. Since (21). The type1enzyme fractions elutingfrom the hydroxylapa- the type 1 enzyme dephosphorylated Ins(1,3,4,5)P4 approxitite column were pooled and applied atosize exclusion column mately 12 times slower than Ins(1,4,5)P3, the measured rate of Sephacryl S-200. The enzyme elutedas a single peak, of Ins(1,3,4,5)P4 dephosphorylation at 10 p~ (Xl2) should corresponding to a M , of 60,000 comparedwith standard estimate total type 1 activity towardsIns(1,4,5)P3. Total type proteins run on the same column (Fig. 4). Although a large 1 and type 2 activity in the supernatant was determined by loss in activity with little apparent gain in specific activity the rateof Ins(1,4,5)P3dephosphorylation whenboth enzymes

*

17324


were saturated with 150 p~ Ins(1,4,5)P3. Total 5-phosphomonoesterase activity in the 100,000 X g supernatant was thus calculated to be 48 f 8%by type 1enzyme and 52 & 9% by type 2 enzyme. This estimate was supported by analysis of the relationship of Ins(1,4,5)P, concentration to its rate of dephosphorylation in the 100,000 X g supernatant, which resulted in a biphasic Eadie-Hofstee plot, consistentwith the presence of two enzymes having different K,,, values but approximately equal V,,, activities (data not shown). Dependence onpH-The type 1 enzyme had a fairly sharp pH profile with maximal activity at pH 7.5 using a near saturating Ins(1,4,5)P3 concentration of 30 p M (Fig. 1OA). With 1ns(1,3,4,5)P4as substrate (10 p ~ )a, similar pH profile was observed, but the optimal pH was shifted towards pH 8 FIG. 12. Inhibition of soluble type 1 and type 2 inositol (Fig. 1OA). This may represent a difference in the pK, values of the phosphates on Ins(1,4,5)P3 compared to Ins(1,3,4,5)P4. phosphate 6-phosphomonoesterases by f r e e Cas+. Enzyme acThe maximal activity of type 2 5-phosphomonoesterase at tivity was assayed in the presence of 1 mM EGTA 1 mM M P and various amounts of CaC1,. The Ins(1,4,5)P3concentration was 30 and M was between pH 6.5 and 7.0, with a rapid 150 pM, respectively, for type 1and type 2 enzymes. 150 ~ L Ins(1,4,5)P3 decrease in activity at higher or lower pH values (Fig. 10B). Effects of Mg2+and Ca2+-Fig. 11 shows the dependence of TABLE I1 Effect of phosphorylated metabolites onsoluble brain Ins(1,4,5/P3 5-phsphomonoesterases Ins(1,4,5)P3concentration was 30 pM for type 1 enzyme and 150 p~ for type 2 enzyme. Concentration of each phosphorylated metabolite was 2 mM. Values are 3c + S.E., n = 3. Percent of control activity

2,3-Diphosphoglyceric 30 acid Fructose 1,6-bisphosphate Fructose 2,6-bisphosphate Glucose 6-phosphate ATP ADP AMP

FIG. 10. Effect of p H on soluble type 1 and type 2 inositol phosphate 5-phosphomonoesterases. The enzyme utilized was from the Sephacryl S-200 and S-300 pools. The 50 mM Mops normally used in the assay buffer was replaced with 25 mM Mes, 25 mM Mops, and 25 mM Hepps, and the pH adjusted accordingly.A, type 1enzyme; Ins(1,4,5)P3 concentration was 30 p M and Ins(1,3,4,5)P4 concentration was 10 p ~B,. type 2 enzyme; Ins(1,4,5)P3concentration was 150 uM.

E

6

+2 42 + 2 46 + 5 70k 5 9+5 41 + 11 70 +. 7

3rt2 33 f 11 41 f 4 90 f 7 2f1 42 f 4 80 f 10

I-

Free ATP ( FM ) FIG. 13. Inhibition of soluble type 1 inositol phosphate 6phosphomonoesterase by freeATP. The enzyme was assayed at five different total Mg2' concentrations, with increasing total ATP concentration (0.2, 0.6, 1, 1.5, and 2 m M ) and the enzyme activity plotted as a function of the calculated free ATP concentration. 0,0.2 mM MgClz; 0, 0.6 mM MgClz; A, 1 mM MgCL; A, 1.5 mM MgCL, X, 2 mMMgC1,. A similar plot was obtained with the type 2 enzyme.

Free Mg2+ ( mM )

FIG. 11. Dependence of soluble type 1 and type 2 inositol p h o s p h a t e 5 - p h o s p h o m o n o e s t e r ~on free MI?*. Enzyme activity was assayed in the presence of 1 mM EDTA and various amounts of MgCI,. The Ins(1,4,5)P3 concentration was 30 and 150 pM, respectively, for type 1and type 2 enzymes.

type 1 and type 2 enzyme activities on the free M$+ concentration. The enzymes had an absolute requirement for free M$+ since both enzymes were completely inhibited when 1 mM excess EDTA over total M$+ was added to the assay buffer. Type 1 enzyme expressed maximal activity at 10 p~ free Mg2' and its activity remained constant at higher free M$+ concentrations. Type 2 displayed an unusual but reproducible free Mg2' activity relationship, reaching 60% of its maximal activity by 10 p t free ~ M P , followed by an increase

Soluble 5-Phosphomonoesterases Phosphate Inositol

17325

brain type 1 enzyme is 7.5, which is similar to that reported for the liver plasma membrane-bound enzyme (17, 18) and erythrocyte plasma membrane-bound enzyme (4). The physical characteristics of the brain type1enzyme alsodistinguish it from that of the purified platelet enzyme. The platelet enzyme behaves asa protein of M , = 38,000 on size exclusion chromatography (21) and has a subunit M, of 38,000-45,000 on SDS-PAGE (21, 23, 32), indicating that the native protein is a monomer. The brain type1enzyme migrates asa subunit on SDS-PAGE of M , = 66,000 or 60,000 and elutes as a protein of M , = 60,000 on size exclusion chromatography. A recent report mentions that the soluble 5-phosphomonoesterase from liver chromatographs asa protein of M, = 77,000 on Sephadex G-100 (15). Clearly, an understanding of the relationship of the brain type1enzyme to theliver- and plateletsoluble enzymes must await their further characterization. The brain type 2 enzyme is both physically and kinetically different from the type 1 enzyme and to any other inositol phosphate 5-phosphomonoesterasepreviously described. This enzyme eluted as a protein of M , = 160,000 on size exclusion chromatography and has a similar subunit molecular weight on SDS-PAGE. The K, for Ins(1,4,5)P3 is approximately5fold greater than that of the type 1 enzyme and is distin1 activity by its high K, for guished from thetype . suggests Ins(1,3,4,5)P4,which is greater than 150 p ~ This that under conditions of substrate concentrations found in the cell, this enzyme probably operates only as an Ins( 1,4,5)Ps DISCUSSION 5-phosphomonoesterase. One interesting possibility that has may have a high affinity for This study describes the isolation and characterization of not yet been tested is that it two distinct soluble inositol phosphate 5-phosphomonoester- cyclic Ins(1,2,4,5)P3. Implications for Regulation of the Inositol Phosphate Pathases from rat brain. The two enzymes represent about 25% of the total inositol phosphate 5-phosphomonoesterase activity way-There are conflicting data concerning theeffect of free inositol phosphate 5-phosphomonoesterase. assayed in the crudehomogenate with their maximal abilities Ca2+onthe t o hydrolyze Ins(1,4,5)P3 being approximately equal. Both Ins(1,4,5)P3 degradationby cytosolic fractions from macroenzymes specifically remove the5-phosphate from either phages(19) and porcine arterysmooth muscle (22) were by free Ca2+ in the physiological range Ins(1,4,5)P3 or Ins(1,3,4,5)P4, resulting in the formation of shown to be stimulated Ins(l,4)Pz andIns(1,3,4)P3, respectively. No subsequent me- of 10-7-10"j M, suggesting regulation of these soluble enzymes tabolism of either of these products occurred uponprolonged by changes of the intracellular free Ca2+ concentration. On purified soluble platelet enzyme was found incubation with thepurified enzymes. In agreement with the the other hand, the properties of 5-phosphomonoesterases characterized in other to be inhibited by free Ca2+, with a Kiof 70 p~ (21). The studies (4, 17-21), both of the soluble brain enzymes purified particulateIns(1,4,5)P35-phosphomonoesterase from RIN in the present study require free Mg2+ for the expression of m5F cells wasnot affected by free Ca2+between lo-' and M (33). while higher free Ca2+ concentrations were not exactivity. Relationship of Soluble Brain Inositol Phosphate 5-Phos- amined. The plasma membrane-bound 5-phosphomonoesterphomonoesterase to Other Inositol Phosphate 5-Phosphomon- ase from pancreatic islets was not affected by Ca2' (31),but the erythrocyte membrane-bound enzyme was found to be oesterases-The type 1 enzyme is kinetically similar to the solubleenzymepurified from platelet cytosol. The K,,, for inhibited at very high Ca2+ levels (-5 mM) (4). The data presented in this study,showing a gradual inhibition of both Ins(1,4,5)P3 of the purified platelet enzymewasoriginally revised type 1 and 2 enzymes with an increase of free Ca2+ in the reported to be 30 p~ (21), but has been subsequently to 5-10 p M (11,23). More recently, this enzyme has also been micromolar range, is similar to that reported for the soluble shown to dephosphorylate Ins(1,3,4,5)P4 to Ins(1,3,4)P3 withplatelet enzyme (211, and suggests that the soluble brain 5a K , forIns(1,3,4,5)P4 of 0.8 MM (11). Interestingly,the phosphomonoesterases are not regulated by physiological concapacity of this enzyme to degrade Ins(1,4,5)P3was over 30- centrations of free Ca2+. fold greater than its capacity to metabolizeIns(1,3,4,5)P4. Following hormone-stimulated formationof Ins(1,4,5)P3, it These kinetic characteristics are nearly identical to the brainhas been shown in a number of different cell types that a type 1 enzyme, which had a K, of 3 p~ for Ins(1,4,5)P3, a K, portion of the Ins(1,4,5)P3 is rapidly converted to of 0.8 p M forIns(1,3,4,5)P4,and a turnoverratio of Ins(1,3,4,5)P4 (8,9,33-38). Onepossible role of Ins(1,3,4,5)P4 Ins(1,4,5)P3 to Ins(1,3,4,5)P4at saturating substrate concen- generation in cells may be to prolong the Ins(1,4,5)P3 signal trations (10 X K,) of approximately 12. The K , of brain type by decreasing the rate of Ins(1,4,5)P3 degradation by the 51 enzyme for Ins(1,4,5)P3 is also similar to that reported by phosphomonoesterase (11). As the Ins(1,3,4,5)P4 concentraliver cytosol (18). tion increases, it will compete with Ins(1,4,5)P3 as substrate Despite the close kinetic similarities, other characteristics forthetype 1 5-phosphomonoesterase,since the K , for of the type 1 enzyme suggest that it may not be an identical Ins(1,3,4,5)P4of the type 1 enzyme is approximately 5-fold protein to the purified platelet enzyme. With respect to the lower than its K, for Ins(1,4,5)P3, and Ins(1,3,4,5)P4hydrolpH dependence, theoptimalpH for theplatelet enzyme ysis will effectively decrease the amount of enzyme available towards Ins(1,4,5)P3 is6.5, whereas the p H optimum for the for Ins(1,4,5)P3 hydrolysis, because the enzymes capacity to of activity as the free Mg2+ was increased to 500 p~ and a subsequent decline to 50% of the peak activityat 10 mM free M$+. The concentration of free Mg2+ required for enzyme activity in this study islower than reported in other tissues, where the concentrationof free Mg2+ required for half-maximal activity was found t o be 250-500 p M (4, 17-19). Both type 1 and type 2 enzymes responded similarly with respect to free Ca2+ concentration (Fig. 12). As the free Ca2+ , inhibition concentration was increased above 1 p ~a gradual occurred, with a 50% inhibition a t 100 pM. This is consistent with the apparent K, of 70 p~ free Ca2+ for inhibitionof the purified platelet enzyme (21). Effect of Phosphorylated Metabolites on the Inositol Phosphate 5-Phosphomonoesterases-Table I1 shows that a large variety of phosphorylated metabolites inhibited both 5-phosphomonoesterase enzymes. Interestingly, ATPwas one of the most effective inhibitors. This inhibitionwas shown to be due t o free ATP, since measurements of activity changes in buffers M$+ and total ATP concentrations showed with various total a correlation only with the calculated free ATP concentration and not with either the free Mg2+ or the calculated MgATP concentrations (Fig. 13, data for type 2 enzyme not shown). Inhibition by bisphosphate metabolites was moreeffective thanwithmonophosphates(Table 11). An inhibition of Ins(1,4,5)P3 degradation by bisphosphate glycolytic metabolites in pancreatic islets haspreviously been reported (31).

17326

Soluble Inositol Phosphate 5-Phosphomonoesterases

hydrolyze Ins(1,3,4,5)P4 is only one-twelfth that of In~(1,4,5)P,.~ This possibility is supported by the recent observation that addition of Ins(1,3,4,5)P4to permeabilized hepatocytes not only inhibits Ins(1,4,5)P3dephosphorylation, but also prolongs the time required to resequester the Ca2+released by addition of Ins(1,4,5)P3(39). The existence of multiple forms of inositol phosphate 5phosphomonoesterase raises interesting questions concerning their physiological significance. It is possible that the type 1 and type enzymes 2 are differentially distributed inthe various regions of the brain and reflect local specialization of tissue function. The characteristics of the type 2 enzyme are suggestive for a region in which little Ins(1,3,4,5)P4 is produced. Alternatively, all of the 5-phosphomonoesterases may be present within a cell and represent a mechanism for independent regulation of more than one intracellular messenger, i.e. Ins( 1,4,5)P3and Ins(1,3,4,5)P4.The soluble platelet enzymes appear to be regulated through phosphorylation by protein kinase C (11, 23, 40), representing a potential negative feedback system regulating Ins( 1,4,5)P3formation (32). Pretreatment of the brain type 1 enzyme with protein kinase C has also been observed to increase the V,,, of the enzyme (41). Whether the type 2 or the membrane-bound enzyme are both regulated by protein kinase C, or regulated in an independent manner is currently under investigation. REFERENCES 1. Berridge, M. J., and Irvine, R. F. (1984) Nature 3 1 2 , 315-321 2. Williamson, J. R., Cooper, R. H., Joseph, S. K., and Thomas, A. P. (1985) Am. J. Physiol. 248, C203-C216 3. Hokin, L. E. (1985) Annu. Reu. Biochem. 54, 205-235 4. Downes, C. P., Mussat, M. C., and Michell, R. H. (1982) Biochem. J. 203,169-177 5. Storey, D. J., Shears, S. B., Kirk, C. J., and Michell, R. H. (1984) Nature 3 1 2 , 374-376 6. Inhorn, R. C., Bansal, V. S., and Majerus, P. W. (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 2170-2174 7. Irvine, R. F., Letcher, A. J., Heslop, J. P., and Berridge, M. J. (1986) Nature 320, 631-634 8. Hansen, C. A., Mah, S., and Williamson, J. R. (1986) J. Biol. Chem. 261,8100-8103 9. Batty, I. R., Nahorski, S. R., and Irvine, R. F. (1985) Biochem. J. 232,211-215 10. Downes, C. P., Hawkins, P. T., and Irvine, R. F. (1986) Biochem. J. 238,501-506 11. Connolly, T. M., Bansel, V. S., Bross, T. E., Irvine, R. F., and Majerus, P. W. (1987) J. Biol. Chem. 262, 2146-2149 12. Mitchell, B. (1986) Nature 3 2 4 , 613 13. Houslay, M. D. (1987) Trends Bwchem. Sci. 12, 1-2

This possibility would also hold true for the membrane-bound 5phosphomonoesterase, since its preliminary characterization in isolated cortex plasma membranes indicate a K,,, for Ins(1,4,5)P~of 18 PM,a K,,, for Ins(1,3,4,5)P4of 3 p ~and , a greater capacity to hydrolyze Ins(1,4,5)P3relative to Ins(1,3,4,5)P4.

14. Irvine, R. F., and Moor, R. M. (1986) Bwchem. J. 240,917-920 15. Shears, S. B., Storey, D. J., Morris, A. J., Cubitt, A. B., Parry, J. B., Michell, R. H., and Kirk, C. J. (1987) Biochem. J. 242, 393-402 16. Williamson, J. R., Hansen, C. A., Johanson, R.A., Coll, K. E., and Williamson, M. (1988) in Regulation of Cellular Calcium Homeostasis (Pfeiffer, D. R., ed) Plenum Publishing Corp., in press 17. Seyfred, M. A., Farrell, L. E., and Wells, W. W. (1984) J. Biol. Chem. 259,13204-13208 18. Joseph, S. K., and Williams, R. J. (1985) FEBS Lett. 180, 150154 19. Kukita, M., Hirata, M., and Koga, T. (1986) Biochim. Biophys. Acta 885, 121-128 20. Erneux, C., Delvaux, A., Moreau, C., and Dumont, J. E. (1986) Biochem. Biophys. Res. Commun. 134,351-358 21. Connolly, T.M., Bross, T. E., and Majerus, P.W. (1985) J. Biol. Chem. 260, 7868-7874 22. Sasagura, T., Hirata, M., and Kuriyama, H. (1985) Biochem. J. 23 1,497-503 23. Connolly, T. M., Lawing, W. J., Jr., and Majerus, P. W. (1986) Cell 46,951-958 Johanson, R., Inubushi, T., and 24. Cerdan, S., Hansen, C.A., Williamson, J. R. (1986) J.Biol. Chem. 2 6 1 , 14676-14680 25. Laemmli, U. K. (1970) Nature 2 2 7 , 680-685 26. Merril, C. R., Goldman, D., Sedman, S. A., and Ebert, M. H. (1981) Science 2 11, 1437 27. Simon, W. D., Amman, M., Achme, M., and Morf, E. W. (1978) Ann. N. Y. Acad. Sci. 307,52-70 28. Fabiato, A., and Fabiato, F. (1979) J. Physiol. (Paris) 75, 463505 29. Philips, R. C., George, P., and Rutman, R. J. (1966) J. Am. Chem. SOC.88,2361-2366 30. Bradford, M. M. (1976) Anal. Biochem. 72,248-254 31. Rana, R. S., Sekar, M. C., Hokin, L. E., and MacDonald, M. J. (1986) J. Biol. Chem. 2 6 1 , 5237-5240 32. Majerus, P. W., Connolly, T. M., Deckmyn, H., Ross, T. E., Ishii, H., Bansal, U. S., and Wilson, D. B. (1986) Science 234,15191526 33. Biden, T. J., and Wollheim, C.B. (1986) J. Biol. Chern. 2 6 1 , 11931-11934 (1987) in Biochemical 34. Williamson, J. R., and Hansen, C.A. Actions of Hormones (Litwack, G., ed) Vol. 14, pp. 29-80, Academic Press, Orlando, FL 35. Turk, J., Wolf, B. A., and McDaniel, M. L. (1986) Biochem. J. 237,259-263 36. Hawkins, P. T., Stephens, L., and Downes, C. P. (1986) Biochem. J. 238,507-516 37. Balla, T.,Baukal, A. J., Guillemette, G., Morgan, R. O., and Catt, K. (1986) Proc. Natl. Acad. Sci. U. S. A . 83,9323 38. Morgan, R. O., Chang, J. P., and Catt,K. J. (1987) J. Biol. Chem. 262,1166-1171 39. Joseph, S. K., Hansen, C. A., and Williamson, J. R. (1987) FEBS Lett. 2 1 9 , 125-129 40. Molina y Vedia, L. M., and Lapetina, E.G. (1986) J. B i d . Chem. 2 6 1 , io493-io495 41. Williamson, J. R., Hansen, C.A., Verhoeven, A.,Coll,K. E., Johanson, R., Williamson, M. T., and Filburn, C. (1987) in Cell Calcium and the Control ofMembrane Transport (Eaton, D. C., and Mandel, L. J., e&) pp. 93-116, Rockefeller University Press, New York