Characterization and Purification of Membrane

1 downloads 0 Views 3MB Size Report
Mar 8, 1988 - red blood cell membranes, it was solubilized by high .... (second halo ... ml) or less of fresh red blood cells drawn into citrate phosphate.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 264, No. 9, Issue of March 25, pp. 5080-5088,1989 Printed in U.S.A.

Characterization andPurification of Membrane-associated Phosphatidylinositol-4-phosphateKinase fromHuman Red Blood Cells* (Received for publication, March 8, 1988)

Leona E. Ling, John T. .Schulz,and Lewis C. Cantley From the Department of Physiolugy, Tufts University Schoolof Medicine, Boston, Massachusetts 021 11

The membrane-bound form of phosphatidylinositol- from hydrolysis of PtdInsP21 by an agonist-activated, phos4-phosphate (PtdInsP) kinasewas purified 4,300-fold phoinositide-specific phospholipase C, PtdInsP kinase, the from human red blood cells to aspecific activity of 117 enzyme which phosphorylates PtdInsP to PtdInsP', also apnmol min" mg". Although this enzyme copurified with pears to be important to theproduction of second messengers. red blood cell membranes, it was solubilized by high Increased synthesis of PtdInsP' is necessary during second salt extraction in the absence of detergent indicating messenger generation because only a minute amount of horthat it is a peripheral membrane protein. The major mone-sensitive PtdInsP' is presentin cells, and upon agonistprotein seen in the most purified preparation migrated activated phosphoinositide turnover thispopulation is quickly at 53,000daltons on sodium dodecyl sulfate-polyacryl- depleted (6,7). Quantitativemeasurements show that inositol amide gel electrophoresis (SDS-PAGE). The major trisphosphate produced during activated phosphoinositide PtdInsP kinase activity in this preparation was also turnover greatly exceeds that which is available from PtdInsP' coincident with this 53,000-dalton band upon renatur- present prior to stimulation (8-10). Taylor et al. (lo), for ation of activity from SDS-PAGE.To test further example, determined thattherate of PtdInsP'synthesis whether the 53,000-dalton proteincontained PtdInsP during concanavalin A activation of thymocytes must be kinase activity, antibodies were prepared against the stimulated 10-fold to account for the amount of inositol gel-purified 53,000-dalton protein. Thisantiserum trisphosphate produced. was able to precipitate both the 53,000-daltonpeptide Although the PtdInsPkinase is responsible for synthesis of and PtdInsP kinase activityfrom red blood cell mem- PtdInsP', a rate limiting component in second messenger branes. The apparent size of the native enzyme in the production, few studies have focused directly on this enzyme most purifiedpreparationwas determined to be andits role in phosphoinositide turnover. Early work on 150,000 f 25,000 daltons by gel filtration. This Ptd- PtdInsP kinase in rat brain homogenates defined an ATPInsP kinase activity wasat least 100-fold more active and divalent cation-dependent enzyme activity capable of in phosphorylating PtdInsP than phosphatidylinositol phosphorylating PtdInsP toPtdInsP' (11).Structural studies and was easily separated from the red cell membrane of PtdInsPz and PtdInsP formed in rat brain tissue showed phosphatidylinositol kinase by salt extraction. Analy- that PtdInsP' was phosphorylated at the 4' and 5' positions sis of the reaction product, phosphatidylinositol 4,5- of the inositol ring, whereas PtdInsP was phosphorylated only bisphosphate, indicates that the enzyme phosphoryl- at the4' position (12). Thus, itwas hypothesized that PtdIns ates phosphatidylinositol 4-phosphate specifically at is phosphorylated first by PtdIns kinase at the 4' position the 5'-hydroxyl of the inositol ring. The apparent K , followed by phosphorylation at the 5' position by PtdInsP for ATP was 2 NM, and the concentrationsof M 8 + and kinase (13-15).However, recent findings with the PtdIns at two types of this enzyme Mn2+giving half-maximal activity were 2 and 0.2 mM, kinase suggest that there are least with different positional specificities. Purified PtdIns kinase respectively. Mg2+ supported %fold higheractivity than Mn2+at optimal concentrations. The enzymatic from bovine uterus was shown to phosphorylate at the exactivity was inhibited by its product, phosphatidyli- pected 4' position (16),buta new class of PtdIns kinase nositol 4,5-bisphosphate and enhanced by phosphati- activity which associates with the middleT/pp60""" transforming protein and activated platelet-derived growth factor dylserine. receptor was discovered to phosphorylate PtdIns at the 3' position (17, 18).Furthermore, two types of PtdIns 4'-kinase activity have been defined in bovine brain (19).These findings indicate that there are at least three distinct PtdIns kinase activities which can be divided into two classes by their ability Agonist-activated phosphoinositide turnover generates two second messengers, inositol trisphosphate which elicits a rise in intracellular calcium and diacylglycerolwhich activates protein kinase C (Ca'+/phospholipid-dependent enzyme) (15 ) . Although it is clear that these second messengers arise * This work wassupported by National Institutes of Health Grants GM36624 and DK07542 (Pathobiology and Digestive Diseases). 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 18 U.S.C. Section 1734 solely to indicate this fact.

The abbreviations used are: PtdInsP2, phosphatidylinositol bisphosphate; PtdIns, phosphatidylinositol; PtdInsP, phosphatidylinositol phosphate; PtdSer, phosphatidylserine; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; MES, (2-[N-morpbolino] ethanesulfonic acid); HEPES, N-2-hydroxyethylpiperazine-N'-2ethanesulfonic acid; DTT, dithiothreitol; PMSF, phenylmethylsulfonyl fluoride; PBS, phosphate-buffered saline; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; HPLC, high pressure liquid chromatography; PtdIns(4)P, phosphatidylinositol 4phosphate; GroPtdIns(4,5)P2, glycerophosphorylinositol4,5-bisphosphate; Ins(1,4,5)P3,inositol 1,4,5-trisphosphate.

5080

Characterization Purification andMembrane-associated of

PtdImP

5081

TABLE I Summary of purification data PtdInsP kinase activity was assayed using method A except where indicated. Membranes extract NaCl

Cell lysate

Activity (nmol/min) % Yield Specific activity (nmol/min/mg) Specific activity (nmol/min/mg) -Fold purification Protein (mg) Protein concentration (mg/ml) 3 Number of experiments a

37.4 f 11.5 44 3.6 0.346 -C 0.127

34.7 f 18.2 41 0.038 & 0.012

64.6 100 0.0035

SP peak

6.0 f 0.5 16 1.02 f 0.33

2.9 f 0.3 4.5 8.75 f 0.60

SP peak (second halo

2.3 f 0.8 14.9 f 3.7

117 f 13 4,300 10 1,432 f 460 4.2 f 0.2

24,171 7.8 3

2,500 100 118.5 f 38.6 0.13 f 0.02

3

1

290 7.0 f 1.2 0.3 f 0.1

0.336 f 0.016 0.032 f 0.000 2

0.155 k 0.024 0.007 & 0.002 4,

2O

Assayed using method B. DEAE-SephacelColumn r 2

51

r 0.4

z

- 0.2 = uz

- 0

0

DEAE peak

10

20

30

Fraction Number

FIG. 1. DEAE-Sephacel separation of membrane extract. The membrane extract was loaded onto DEAE-Sephacel and eluted with a linear NaCl gradient from 0 to 0.4 M (solid line). 4-ml fractions were collected and assayed for PtdInsP (PIP)kinase activity using method A (open squares).Proteincontent in each fraction was determined by absorbance a t 280 nm (solid diamonds).

to phosphorylate at the4' or 3' position. It is not known whether different types of PtdInsP kinase exist. PtdInsP kinase is found both in soluble and particulate fractions of cell homogenates (20-24), but the relationship between the soluble and particulate enzymes is unclear. Currently, only the soluble enzyme from rat brainhas been purified and characterized (22-24). Further studies of PtdInsP kinase including purification, characterization, and development of immunological or pharmacological techniques are needed to resolve this issue and to elucidate the function and regulation of PtdInsP kinase in uiuo. As a preliminary step toward understanding the function and regulation of this important enzyme of the phosphoinositide turnover pathway, we purified the PtdInsP kinase from human red blood cell membranes 4300-fold and did an initial characterization of this enzyme. EXPERIMENTALPROCEDURES

Materials-[r-32P]ATP (10 mCiml-'), [3H]Ins(1,4,5)P3,[3H]inositol 1,3,4,5-tetrakisphosphate, and [3H]phosphatidylinositol 4,5-blSphosphate were purchased from Du Pont-New England Nuclear. PtdIns, PtdSer, and phosphatidylcholine (crude egg lecithin) were purchased from Avanti Polar Lipids. PtdInsP, PtdInsP2,ATP, Triton X-100, Nonidet P-40, protein A-Sepharose CL-4B, DEAE-Sephacel, SP-Sephadex, PMSF,DTT, and other chemicals were obtained from Sigma. Electrophoresis grade SDS and nitrocellulose (0.22 pm) were purchased from Bio-Rad. Aquacide I1 was from Calbiochem. Molecular weight markers for SDS-PAGE and gel filtration columns were

obtained from Bethesda Research Laboratories and Sigma, respectively. Preparation of Human Red Blood Cell Membranes and NaCl Extract-Membranes were prepared from 2 units (approximately 900 ml) or less of fresh red blood cells drawn into citrate phosphate dextrose or acid citrate dextrose. Briefly, red blood cells were suspended in twice their volume of phosphate-buffered saline (PBS) (150 mM NaCl, 3 mM KCl, 8 mM Na2HP04, 1.4 mM KHZPO., pH 7.2) and centrifuged at 200 X g. White cells, platelets, and residual serum were removed by aspiration. The remaining red cells were washed 2 or 3 times by resuspension in PBS and centrifugation at 1,000 X g. The packed red blood cells were then diluted 16-fold in ice-cold lysis buffer (5 mM sodium phosphate, 1 mM EDTA, 25 p~ ammonium vanadate, 1mM DTT, pH7.2). All subsequent purification steps were carried out at 4 "C unless otherwise indicated. Lysed red blood cells were centrifuged at 27,000 X g for 20 min and resuspended in lysis buffer. This step was repeated until membranes free of hemoglobin were obtained. The NaCl extract of membranes was prepared by resuspending the pellet in lysis buffer with 1M NaCl to a volume 45 timesthe original volume of packed red blood cells. Within 10 min of resuspension, the membranes were centrifuged at 27,000 X g for 20 min. The supernatant (i.e. NaCl extract) was placed in dialysis tubing (14,000-15,000-dalton cutoff) and concentrated about 3-fold with Aquacide 11. The concentrated extract was then dialyzed exhaustively into lysis buffer and finally into DEAE buffer (25 mM Tris, 5 mM MgC12,l mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.1 mM PMSF, 5% glycerol, pH 7.5). DEAE-Sephacel Chromatography of Membrane Extract-The salt extract from membranes was applied to a 25-ml DEAE-Sephacel column. After collection of the flow-through material, the column was washedwith 60 ml of DEAE buffer and eluted with 120 ml of the same buffer with a linear gradient of NaCl from 0 to 0.4 M. 4-ml fractions were collected and assayed under standard conditions as described below. SP-Sephadex Chromatography of DEAE-Sephacel Peak-The pooled peak of PtdInsP kinase activity from DEAE-Sephacel chromatography was dialyzed into S P buffer (20 mM MES, 5 mM MgC12, 1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.1 mM PMSF,15% glycerol, pH 5.5) at 4 "C to remove salt and equilibrate the pH to5.5. The dialyzed material was loaded onto a 6-ml SP-Sephadex column. After collection of the flow-through, the column was washed with 20 ml of S P buffer and eluted with a 40-ml linear gradient of NaCl from 0 to 0.8 M in S P buffer. 1-ml fractions were collected and assayed under standard conditions, except that Tris was present at 100 mM in order to adjust the pH of the reaction to 7.5. PtdInsP Kinase Assays and Thin Layer Separation of Phospholipids-Standard PtdInsP kinase assays were performed as follows: 20 pl of PtdInsP kinase was mixed with 20 pl of a sonicated solution of phospholipid (f0.176 Triton X-100). The reaction was initiated with 10 pl of 0.5 mM [T-~'P]ATP(2000 Ci mol"). The assay conditions employed depended upon the purpose of the assay. Method A was used for assays during purification and measured both PtdInsP and PtdInskinase activities. These reactions contained 0.04% Triton X-100,25 mM Tris, 5 mMMgC12, 1 mM EDTA, 0.1 mM EGTA, 1 mM DTT, 0.1 mM PMSF, 80 p M PtdInsP, 80 PM PtdIns, 40 p~ PtdSer, 0.1 mM M e - A T P (0.2 mCi/ml), pH 7.4, and were incubated for 5 min at room temperature (23-25 "C). Under these conditions, the

Characterization and Purification of Membrane-associated PtdInsP

5082

SP-Sephadex Column

0

s

10

1s

A

20

Fraction Number

B M A

B C I 2 3 1 5 6 7 8 9 IO I I 12 13 11 IS 16 17 I8 1920 M

incorporation of 32Pinto PtdInsP, was linear for 7 min (data not shown). Optimal conditions for PtdInsP kinase activity were employed in method B. The conditions were essentially the same as for method A except that the assays were done with 1 mg ml" PtdInsP alone and were incubated for 2.5 min a t 30 "C. Reactions were stopped by adding 0.5 ml of methanokl N HCI (1:l) and extracted with 0.3 ml of chloroform. The organic layer was dried under nitrogen, resuspended in ch1oroform:methanol:water(2:1:0.01), and spottedonto 0.2-mm oxalate-pretreated Silica Gel 60 plates (Merck). Plates were developed in chloroform:methanokwater:37% ammonium hydroxide (45:35:8.5:1.5), and phospholipids were visualized with iodine vapor after chromatography. The reaction products were visualized by autoradiography and identified by comparison with unlabeled standard phospholipids. Radioactivity in phospholipids was determined by excising the appropriate piece of TLC plate and counting it in liquid scintillation mixture. SDS-Polyacrylamide Gel Electrophoresis-Electrophoresis samples were either added directly to Laemmli sample buffer or precipitated with trichloroacetic acid and resuspended in sample buffer before loading. Electrophoresis was performed on SDS-polyacrylamide gels with an 8.5 or 10% separating gel and 4% stacking gel by . the method of Laemmli (25). Renaturation of PtdInsP Kinase Activitv from SDS-PAGE-Samples containing ?5 pl of SP-Sephadex-pukked PtdInsP kinase (approximately 20,000 fmol/min) or 10 p1 of SP-Sephadex-purified PtdInsP kinase (25-50 ng of 53,000-dalton protein) were incubated in SDS sample buffer (1% SDS (recrystallized), 10 mM DTT, 10% glycerol, 20 mM Tris, pH 6.8) for 5-10 min a t room temperature. The samples were subjected to SDS-PAGE on8%gels as described above except that the gels were polymerized with 2.5 pg/ml riboflavin in place of ammonium persulfate and electrophoresed a t 5-10 "C. The lanes containing 10pl of purified PtdInsP kinase were silver-stained and scanned by laser densitometry. The lanes containing 75 pl of the purified material were transferred to nitrocellulose. After transfer (23 h, 4 "C),the nitrocellulose strip was washed briefly in 0.1% Triton X-100, 10% glycerol, 1 mM EGTA, 1 mM DTT, 10 pg/ml aprotinin and leupeptin, 50 mM Tris, pH 7.4. The paper was then divided into 16 pieces and incubated for 15-24 h in buffer containing 350 pg/ml PtdInsP, 800 pg/ml crude phosphatidylcholine, 0.2% Triton X-100, 10% glycerol, 1 mM EGTA, 1 mM DTT, 10 pg/ml aprotinin and leupeptin, 50 mM HEPES,pH 7.5. PtdInsP kinase activity was assayed by adding 10 pl of [T-~'P]ATP to0.15 mM (2 mCi/ml) and incubating a t room temperature for 90 min. The reactions were terminated, extracted, and analyzed as described above. Prestained molecular mass markers were used to delineate lane borders and to measure efficiency of transfer tonitrocellulose. Recovery of PtdInsP kinase activity was 0.01%. Control assays of gel sections containing no protein showed no activity (data notshown). Identification of Phosphorylntion Position on the Inositol Ring of PtdInsP2 Product-The SP-Sephadex purified PtdInsP kinase was used to catalyze the phosphorylation of PtdIns(4)P in the presence of 32P-labeled ATP as described above except that PtdInsP was present a t 400 pg ml". Analysis of the products by thin layer chromatography as described above showed the presence of 3ZP-labeled PtdInsP2alone. The labeled PtdInsP2was treated with 5 p1of purified rat liver cytosolic phospholipase C (a gift from Dr. John Lowenstein, Brandeis University) for 30 min a t 37 "C in buffer containing 50 mM Tris, 100 p~ EGTA, 100 p~ CaC12, 180 mM NaCI, and 0.1% deoxycholate, pH 7.5. The reaction was terminated and extracted with 0.4 ml of 1:l 1 N HC1:methanol and 0.4 ml of chloroform. The aqueous phase containing inositol phosphates was lyophilized and analyzed by HPLC on a Whatman 5-pm partisphere SAX column along with 3H-labeled InsP3 standards (33). Ins(1,3,4)P2 standard was prepared by incubation of [3H]inositol 1,3,4,5-tetrakisphosphatewith a mouse fibroblast cell lysate for 10 min a t room temperaturein 50 mM HEPES, 1 mM EGTA, 4 p~ CaC12, 5 mMMgC12, 120 mM KCI, 30 mM NaCI, 1 mM DTT, pH 7.5. The PtdInsP2 product was also analyzed by deacylation and HPLC. The labeled PtdInsP2 was treated for 50 min a t 53 "C with 1.8 ml of methylamine reagent containing 5.77 ml of 25% methylaminein water, 6.16 ml of methanol, and 1.54 ml of 1-butanol. After lyophilization, the sample was resuspended in 1.8 ml of water. To separate the water-soluble GroPtdInsP2products from unreacted PtdInsP2 the resuspended sample was extracted twice with 2 ml of a mixture of 1butanoklight petroleum (boiling point 40-60 "C):ethyl formate (204:l). The extracted aqueous phase was lyophilized and analyzed by HPLC as described above (17). [3H]GroPtdIns(4,5)P2 preparedby "

WKd WKd IUM

43Kd

31Kd

C 1

2

'W

uly

4200Kd

FIG.2. SP-Sephadex separation of DEAE peak. A, the peak fractions from DEAE-Sephacel chromatography were pooled, dialyzed into pH 5.5 SP-Sephadex buffer to remove salt, and loaded onto SP-Sephadex. The column was eluted with a linear NaCl gradient from 0 to 0.8 M NaCl (solid line). 1-ml fractions were collected and assayed for PtdInsP kinase activity using method A (open squares). The protein content (expressed as relative units; solid diamonds) of each fraction was determined by laser densitometry of the silver-stained gel presented in B. B, 25 pl of each SP-Sephadex fraction was analyzed on 10% SDS-PAGE. Proteins were visualized by silver stain. The 53,000-dalton band is indicated by the large arrow. Smull arrows show molecular masses. Lanes were loaded as follows: lane A, pooled DEAE peak fractions; lane B, flow-through eluate after loading pooled DEAEpeak fractions; lane C, flow-through eluate after wash with SP-Sephadex loading buffer; lanes 1-20, fractionated NaCl gradient eluate; lane M, molecular mass markers. C, fractions 11-20 were pooledand 10% of the total volume wasanalyzed by 8.5% SDS-PAGE (lane 1). Lane 2 shows molecular mass markers. The proteins were stained with Coomassie Blue.

"

Characterization and Purification of Membrane-associated PtdInsP 53K

67K ‘I

94 K

2%

5083

‘I

43K

FIG. 3. Renaturation of PtdInsP kinase activity from SDS-PAGE. SP-Sephadex-purified PtdInsP kinase was separated bv SDS-PAGE and either silver-stained f i r protein (A) or trans- E. ferred to nitrocellulose, renatured, and assayed for PtdInsP kinase activity ( B ) . a The location of molecular mass markers andthe 53,000-dalton proteinareindi3 cated by arrowheads. ;

i Zf E

1

I 100

B

2 0

1

2

3

4

5

8

Gel Secllonr

deacylation of [3H]phosphatidylinositol4,5-bisphosphate was added as an internal standard. Preparation of Antibodies against the 53,000-Dalton Protein-The 53,000-dalton protein was excised from an SDS-PAGE of pooled fractions from the second half of the SP-Sephadex peak of activity (see Fig. 2C). The protein was electroeluted from the gel, precipitated with 5 volumes of acetone, and resuspended in PBS. Approximately 50-100pgof purified 53,000-dalton protein in 1:l PBSFreund’s complete adjuvant was injected into the popliteal node of a New Zealand White rabbit. The rabbit was given booster injections subcutaneously of 20-100 pg of the purified protein in PBS and incomplete adjuvant (1:l)every 1-2 weeks. Serum was obtained by bleeding from the ear vein 1-2 weeks after each injection. Prior toimmunization, preimmune serum was obtained from the same rabbit. Immunoprecipitation of PtdZnsP Kinase and 53,000-Dalton Protein-Indicated amounts of preimmune or immune serum were added to solubilized red blood cell membranes in 1% Nonidet P-40 and phosphate-buffered saline (with or without added tracer amounts of the same material labeled with lZ6I-Bolton-Hunterreagent) (see Fig. 4). The mixtures were brought to a total volume of 50pl with phosphate-buffered saline, and incubated for 18 h at 4 “C. Protein A linked to Sepharose CL-4B was then added and the mixture was incubated for 1-3 h at 4 “C with occasional mixing. Beads were pelleted by brief centrifugation, and the supernatant was removedfor later analysis. After 2 washes with 1%Nonidet P-40 inPBS followed by 3 washes in PBS, the beads were either assayed directly for PtdInsP kinase activity or incubated with SDS sample buffer and analyzed bygel electrophoresis and autoradiography. 20 pl of the remaining supernatants were also assayed for PtdInsP kinase activity. The 53,000-dalton band was excised and quantitated by scintillation counting. Protein Determinution-Protein was quantitated by the method of Lowry (26). Relative protein content of SDS-PAGE separated proteins was determined by absorbance at 280 nm or by absorbance on an LKB 2202 Ultroscan laser densitometer of CoomassieBlue-stained proteins. Data Analysis-In figures where more than two data sets are presented, the data is expressed as the mean f S.E. When two experiments are presented, the bars indicate the range. The apparent K,,, for ATP and apparentrC, values for divalent cations were determined by nonlinear least squares fit of a simple hyperbolic model to the data. RESULTS

Purification of PtdInsP Kinase from RedBloodCellsPtdInsP kinase activity was found in both the soluble and particulate fractions of lysed fresh human red blood cells. Approximately 40% of the total PtdInsP kinase activity in these lysates copurified with membranes. Between 80 and 100% of the membrane-bound activity was extracted by incubating the membranes briefly with 1 M NaCl in lysis buffer (Table I). The activity of PtdIns kinase was also monitored

7

8

9

1 0 1 1

1 2 1 3 1 4 1 5 1 8

(lop to Boltom)

during purification procedures. In contrast to PtdInsPkinase, essentially all of the PtdIns kinase present in cell lysates remained with the membranes. Only 5% of the PtdInskinase activity partitioned intothe supernatantupon extraction with 1 M NaCl. Thus, PtdInsP kinase was easily separated from the majority of the PtdIns kinase in this extraction step. Preparation of red blood cell membranes and extraction with high salt afforded a 100-fold purification of PtdInsP kinase with 44% recovery from the cell lysate (Table I). The PtdInsP kinase was further purified by ion exchange chromatography. The fractions obtained from these columns were assayed routinely by method A (“Experimental Procedures”) which detects PtdInsP kinase activity as well as any possible contaminating PtdIns kinase activity. The high salt extract was first chromatographed on a DEAE-Sephacel column as shown in Fig. 1. At pH 7.5, PtdInsP kinase activity binds quantitatively to DEAE-Sephacel. The activity eluted with a linear gradient of NaCl as a single peak at approximately 250 mM NaCl and was purified %fold from the extract (Table I).The peak of activity from the DEAE-Sephacel was further purified 8-9 fold on an SP-Sephadex column (Fig. 2). All of the activity was bound to the column at pH 5.5 and eluted as a broad peak at 400 mM NaCl in a linear salt gradient. The bulk of protein eluted before the peak of activity (Fig. 2, A and B). Specific activities were determined for both the entire pooled peak represented by fractions 9-20 and the more purified second half of the peak represented by fractions 11-20. When assayed using method A, these two pools exhibited specific activities of 9 and 15 nmol/min/mg, respectively, which correspond to 2500- and 4300-fold purifications from the cell lysate (Table I). These fractions contained virtually no PtdIns kinase activity. The purified PtdInsP kinase peak was at least 100-foldmore active in phosphorylating PtdInsP than PtdIns.Purified PtdInsP kinase activity in fractions 1120 was also assayed using method B which has conditions optimized for PtdInsP kinase activity and does not measure PtdIns kinase activity. The specific activity obtained with method B was 117 nmol/min/mg. SDS-PAGE analysis of the column fractions shows that a 53,000-dalton protein coeluted with PtdInsP kinase activity (Fig. 2B). The gel in Fig. 2B,which was overdeveloped for the 53,000-dalton protein, shows a number of bands of higher and lower molecular mass which do not coelute with activity and are thus likely to be contaminants. When the pooled fractions 11-20 which gave the highest specific activity were stained with Coomassie Blue in the linear range for the

Characterization and Purification of Membrane-associated PtdInsP

5084 1400 -I

y = 51 .1564+ 42.2433X R = 0.99

A

k

k -m n .-c

0

0

v

v

1000-

L Q)

.-C .-m e n

C

.2 e n

C

L

w

500-

w

In

Ln

o f . , . , . 0

1

I

2

.

,

3

.

,

a

4

i

5

6

30

20

Serum (ul)

10

0

PtdlnsP KinaseActivity in Pellet (fmollmin)

35 1 30

-

25

-

B

E

200 Kd *

.

o !

. , .

I

0

1

1

2

8

.

3

.

.

I

4

I

5

97 Kd=

6

Serum (ul)

68 Kd* 4 5 3 Kd 4 45 Kd

43Kd*

U 1 2 3 4 5 FIG. 4-continued

"

,

0

.

l

1

.

l

2

'

I

.

3

I

4

.

I

5

.

. ...

l

n

6

Serum (ul)

FIG. 4. Immunoprecipitation of PtdInsP kinase and 53,000-dalton protein. Trace amounts of lz5I-1abeledmembranes (lo6 cpm) were mixed with 12 pg of unlabeled membranes and the mixture was immunoprecipitated and assayed as described under "Experimental Procedures." A, the amount of 'Z51-labeled53,000dalton protein precipitated by immune (solid diamonds) and preimmune (open squares) serum. B , PtdInsP kinase activity precipitated by immune (solid diamonds) and preimmune (open squares) serum; C, PtdInsP kinase activity remaining in the supernatantafter precipitation with immune (solid diamonds)and preimmune (open squares) serum. D,the amount of 53,000-dalton protein found in the immunoprecipitates from A plotted against the activity in immunoprecipitates from B.E , 0.2 pg ofa 1:1 mixture of NaCl extract of membranes and the PtdInsP kinase activity peak from Fig. 1 were mixed with trace amounts of the same material labeled with and then incubated with the indicated amounts of serum followed byprecipitation with protein A-Sepharose (see "Experimental Procedures"). The precipitates were analyzed by SDS-PAGE. Lanes 1-5 correspond to precipitates with 0.2, 0.5, 1.5,3, and 5 pl of immune serum.

53,000-dalton protein the higher and lower molecular mass bands were not apparent, suggesting that the contaminant proteins are present only in trace amounts (Fig. 2C). Thus, the 53,000-daltonprotein was the major protein in these peak fractions and was purified to near homogeneity. PtdInsP kinase activity in this most purified preparation was stable upon storage in concentrated form at -70 "C. Storage at 4 "C

9

R

,

20

30

40

50

80

F m t h Nu-

FIG. 5. Gel filtration and sizing of purified PtdInsP kinase. Pooled peak fractions from the SP-Sephadex column were loaded onto an S-400 Sephacryl column. 1-ml fractions were collected and assayed for PtdInsP kinase activityas described under "Experimental Procedures." Molecular mass markers used were apoferritin, 443,000 daltons; B-amylase, 200,000 daltons; alcohol dehydrogenase, 150,000 daltons; bovine serum albumin, 67,000 daltons; carbonic anhydrase, 29,000 daltons. The elution peaks of each molecular mass markerare indicated by diamonds above the plot of PtdInsP kinase activity. Inset shows a logarithmic curve fit to molecular mass Versus the elution volume ( V,)/void volume ( VO).

caused the 53,000-dalton band to beslowlydegraded to a 45,000-dalton peptide. 53,000-Dalton Protein Contains PtdInsP Kinase Activity

Characterization Purification and m n

--100

Il

of Membrane-associated PtdInsP

5085

a

I

.1

R

a

lflS(1,3,4)P3

0.4

ap 0.2

, 0.-0# 0

0.0 5

15

10

20

0

5

70

Elution Time (min)

---

2000

BI

1

1

2

1/IATP]

4

I \

0 54

52

200 400 600 8 0 0 1000

B

0

500

30

[ATP] uM

85

80

25

56 0

58

62

64

Elution Time (min)

FIG. 6. Identification of 5’ phosphorylation on the inositol ring. A , PtdInsPP was produced by reacting PtdIns(4)P and [y3’P] ATP using SP-Sephadex-purified PtdInsP kinase and theconditions of method A except that 400 pg ml” PtdInsP alone was present. The 32P-labeledPtdInsPp was treated with phosphoinositide-specificphospholipase C. [3H]Ins(1,4,5)P3 and [3H]Ins(1,3,4)P3 (open squares) standards were added to the 3ZP-labeledaqueous products of the reaction (solid squares), and the mixture was analyzed by HPLC. B, 32P-labeledPtdInsP2 was prepared as in A . After deacylation of the PtdInsPz product, a [3H]GroPtdIns(4,5)P2 standard (open squares) was added to the 32P-labeledPtdInsP2 sample (solid diamonds) and both were analyzed by HPLC separation. An arrow indicates the elution time of the GroPtdInsPz derived from PtdInsPz produced by the middleT-pp6OC’”“-PtdIns kinase complex.

3

4

5

(1IW

FIG. 7. ATP dependence. SP-Sephadex-purified PtdInsP kinase was assayed using conditions described under “Experimental Procedures” except for the absence of Triton X-100 and PtdIns and the presence of 50 pg ml” PtdInsP and 150 pg ml” PtdSer (the optimal PtdInsP toPtdSer ratio inthe absence of Triton X-100). The apparent ICmcaTp, was 2 F M as determined by nonlinear least squares fit of the Michaelis-Menton equation to the data. The plot was prepared with normalized data from four experiments each represented by a different symbol. Inset shows a double reciprocal plot of the data.

Immunoprecipitation of a PtdInsP Kinase Activitywith Antibodies against the 53,000-DaltonProtein-To further verify that the 53,000-dalton protein contained PtdInsP kinase activity, a rabbit was immunized with SDS-PAGE-purified 53,000-dalton peptide. Immune serum obtained from this rabbit precipitated in a specific and dose-dependent manner both the 53,000-dalton protein (Fig. 4, A and E) and PtdInsP kinase activity (Fig. 4, B and C) from membranes as well as from NaCl extracts of membranes and from more purified fractions. Precipitation of the 53,000-dalton protein correlated very well with the precipitation of PtdInsP kinase activity (Fig. 4 0 ) . Preimmune serum precipitated neither the after Renaturation from SDS-PAGE-To determine the mi- 53,000-dalton protein nor PtdInsP kinase activity (Fig. 4, A gration position of the PtdInsP kinase on SDS-gel electro- and B ) . Up to 60% of the total activity was removed from phoresis. A renaturation procedure was utilized. The SPghosts by the immune serum (Fig. 4C). This activity removed Sephadex-purified preparation was run on SDS-PAGE, and by immunoprecipitation from ghosts was not reflected in the the proteins were renatured by transferring to nitrocellulose immune pellets, suggesting that activity may be inhibited in and incubating overnight in phospholipid and detergent con- these immune complexes (Fig. 4, B and C). In some preparataining buffer. Fig. 3 shows the profile of protein (A) versus tions, a45,000-dalton protein was detectable by protein stainPtdInsP kinase activity ( B ) .The peak of activity is coincident ing; the immune serum also precipitated this peptide (Fig. with the 53,000-dalton protein. No activity was obtained in 4E), consistent with the possibility that it is a cleavage the absence of renaturation. A small amount of activity was product of the 53,000-dalton protein. This cleavage product also seen in the gel sections between 60,000 and 67,000 dal- is absent from fresh preparations, but is formed during lZ5I tons. However, appearance of this more slowly migrating labeling and storage at 4 “C (see “Purification of PtdInsP activity was variable and may have been due to the 53,000- Kinase from Red Blood Cells” under “Results”). dalton proteinwhich had not completely denatured (the samGel Filtration and Size Determination of Purified PtdInsP ple was not boiled in SDS sample buffer prior to electropho- Kinme-The size of the native PtdInsP kinase in the SPresis). These results demonstrate that the 53,000-dalton pro- Sephadex-purified preparation was determined by gel filtratein contains PtdInsP kinase activity but do not rule out the tion on an S-400 Sephacryl column. Fig. 5 shows that the possibility that other impurities also migrate at this position. peak of PtdInsP kinase activity elutes near the150,000-dalton

Characterization Purification and

5086 25

* z .-c5.

20

-

of Membrane-associated PtdInsP

-

-

A

30

.-* e

I

.-5

-

a

e

u

-

A

.p-

5 20-

S$i Ee

-B2

El

8 .

n o 10-

n 1

0

5

mM

10

.

1

I

15

20

I

0.0

.

0.5 2.5

,

.

1.0 2.0

[Mg2+]

lo

. , . , .

#

.

3.0

0.5

I

3.5

[PIP] rng/rnl

B

1

0 0.0

#

1.5

6

0.2

0.0

1.5 2.52.0 [Mn2+] rnM 1.0

0.4

0.6

0.8

1.0

1.2

1.4

[PIP] mg/rnl

FIG. 9. PtdInsP dependence. PtdInsP kinase activity was asFIG. 8. Divalent cation dependence. SP-Sephadex-purified material was assayed for PtdInsP kinase activity in the absence of sayed at 25 "C with SP-Sephadex-purified PtdInsP kinase in increasTriton X-100 and PtdIns as described in Fig. 7. Assays were per- ing amounts of PtdInsP alone in the presence ( A ) or absence ( B ) of formed with varying concentrations of free M$+ ( A ) or free Mn2+ 0.04% Triton X-100. Data in A are from four experiments and in B ( B ) . The pooled peak fractions were dialyzed to remove MgCl and from a representative experiment. Curves were estimated and drawn EDTA and assayed in the absence of EDTA. Dataare from a by hand. representative experiment.

marker. A logarithmic plot of molecular mass versw elution volume/voidvolume revealed an apparent molecular mass between 125,000 and 175,000 daltons for the native, SPSephadex-purified PtdInsP kinase particle. PtdZnsP Kinase Actiuity Phosphorylates PtdZns(4)P at the 5' Position-The characterization of PtdInsP kinase by Brockerhoff and Ballou (14) showed that PtdInsP kinase activity in brain tissue phosphorylated the inositol ring at the 5'-hydroxyl. Analysis of the PtdInsPn formed by the red blood cell PtdInsP kinase is also consistent with phosphorylation of PtdIns(4)P at thisposition. Phospholipase C treatment of the PtdInsPn product yielded inositol trisphosphate which was then analyzed by HPLC. The analysis showed a single peak comigrating with Ins( 1,4,5)P3 and distinct from inositol 1,3,4-trisphosphate (Fig. 6A). The PtdInsPzproduct was also analyzed by deacylation toGroPtdInsPz.The deacylation product comigrated exactly with GroPtdIns(4,5)PZ(Fig. 6B). This method of analysis has been used to resolve the 4,5isomer from another PtdInsPz isomer produced in vitro using immunoprecipitates of polyoma middleT.' Both of these results suggest that thepurified PtdInsP kinase phosphorylates PtdIns(4)P specifically at the 5'position of the inositol ring. ATP Dependence and Divalent Cation Dependence-The

* L. A. Serunian, L. E. Ling, and L. C. Cantley, unpublished results.

Effect of Phosphatidylserine

Percent PtdlnsP 100

80

60

40

20

o

Percent PtdSer

FIG. 10. Effect of PtdSer. The pooled peak fractions from the SP-Sephadex column were assayed for PtdInsP kinase activity using method A except that PtdIns was omitted. The ratio of PtdInsP to PtdSer was varied while total phospholipid concentration was kept constant at 200 pg ml-'. Relative PtdInsP kinase activity at various PtdInsP to PtdSer ratios were assayed in the absence of Triton X100 (open squares) and the presence of Triton X-100 (solid diamonds). Activity at the optimal PtdInsP to PtdSer ratios for assays under either condition were 2 pmol min". Data in the absence of Triton X-100 are an average of two experiments; in the presence of Triton X-100, an average of three experiments.

Characterization and Purification Membrane-associated of 140 1

Ptdlns

20

-

0

PtdlnsP2 1

0

1

.

1

2

-

1

3

-

1

-

4

PtdlnsP2/PtdlnsP or PtdlnslPtdlnsP

i

5

PtdInsP

5087

fold higher than with 100% PtdInsP. In thepresence of 0.04% Triton X-100, maximal activity was obtained at a higher PtdInsP to PtdSer ratio, 70:30 (Fig. 10B). Although Triton X-100 decreases the PtdSerrequired for maximal activity, the activities at theoptimal PtdSer concentrations were approximately the same with and without detergent. Under these conditions, PtdInsPz, the product of PtdInsP phosphorylation, was found to inhibit PtdInsP kinase activity both in the presence and absence of detergent. Fig. 11 demonstrates that addition of increasing amounts of PtdInsPz to the optimal PtdInsPPtdSer concentrationsinhibited PtdInsP kinase activity to a greater extent than the addition of PtdIns. The slight inhibition seen with PtdIns athigh PtdIns to PtdInsPratios probably reflects a dilution of the substrate PtdInsP. PtdInsP:! clearly shows an additional inhibitory effect. Half-maximal inhibition by PtdInsPz occurred at PtdInsPz to PtdInsPratios of 1.5:l in the absence of Triton X-100 and 0.4:l in the presence of Triton X-100. DISCUSSION

This characterization of the PtdInsP kinase from human red blood cellmembranes elucidates several biochemical properties of the enzyme. The most purified preparation had a specific activity of 117 nmol min"mg" andcontaineda major 53,000-dalton protein. Antibodies made against the highly purified 53,000-dalton protein specifically precipitated both a 53,000-dalton protein and PtdInsP kinase activity. Finally, renaturation of activity from SDS-PAGE shows the PtdlnsP2 peak of activity to be at 53,000 daltons. These results indicate that themajor PtdInsP kinase in human red blood cell membranes has a peptide molecular mass of 53,000 daltons. The 0 1 2 3 4 5 results of renaturation also show some activity at higher molecular masses, but renaturation of activity at these molecPtdlnsPZPtdlnsP or PtdlWPtdlnsP ular masses was notconsistent in several experiments. If FIG. 11. Inhibition by PtdInsPz. PooledSP-Sephadex peak fractionswere assayed for PtdInsP kinase activity as described under higher molecular mass forms of PtdInsP kinase do exist in "Experimental Procedures" except that the assays contained 50 pg this preparation, they may be unrelated to the 53,000-dalton ml" PtdInsP and 150 pg ml" PtdSer and increasing amounts of form since they are not recognized by the antiserum against PtdInsPn(solid diamonds) or PtdIns (open squares) in the presence the 53,000-dalton protein. Another caveat which can not be ( B ) and absence ( A ) of Triton X-100. Data are from representative ruled out at present is that the53,000-dalton protein may not experiments. consist entirely of PtdInsP kinase. PtdInsP kinase activity in the most purified preparations ATP concentration dependence and M F and Mn2+concen- specifically phosphorylated PtdIns(4)P at the 5' position. tration dependence of the SP-Sephadex-purified PtdInsP ki- These results are consistent with the conclusions of Brocknase are presented in Figs. 7 and 8. The apparent K,,, for ATP erhoff and Ballou (14). Unlike the recently purified 45,000was 2 p ~ and , optimal activity was observed at 0.1 mM (Fig. dalton brainmyelin protein which contained bothPtdIns and 7). Mg2f was three times more effective than Mn2+ in sup- PtdInsP kinase activities (28), the red blood cell enzyme was porting PtdInsP kinase activity (Fig. 8). However, the appar- specific for PtdInsP and exhibited less than 1%PtdIns kinase ent K , for Mg2f was approximately 2 mM, whereas that for activity. These results supportthe hypothesis that PtdIns and Mn2+was approximately 0.2 mM. PtdInsP kinase are distinctenzymes which sequentially phosPtdImP Dependence and Phospholipid Effectors-The SP- phorylate PtdIns at the 4' position andthe 5' position, Sephadex-purified PtdInsP kinase showed different PtdInsP respectively, to produce PtdInsPz. dependences in the presence and absence of 0.04% Triton XThe membrane-associated erythrocyte PtdInsP kinase de100 (Fig. 9). Half-maximal activity in the presence of deter- scribed here showed a number of similarities to a soluble gent occurs at 0.7mgml" (Fig. 9A) and in the absence of PtdInsP kinase purified from rat brain (22, 23). The enzyme detergent a t 0.25mgml" (Fig. 9B). The optimal activity from red blood cells had a molecular mass of 53,000 daltons, obtained in Triton X-100 was 6-foldhigher than thatobtained and thenative purified enzyme migrated between 125,000and in the absence of Triton X-100. 175,000 daltons. The ratbrain PtdInsP kinase was identified Previous results with PtdInsP kinase from Friend murine as a45,000-dalton protein which migrated as a110,000-dalton erythroleukemia cells (27)and rat brain(22) suggest that particle in the nondenatured form. Both enzymes required PtdSerenhances activity. Fig. 10 illustrates the effect of Mg2' as a cofactor with an optimum concentration of 10-20 PtdSer on the purified red blood cell PtdInsP kinase. This mM (22-24). Mn2+was able to substitute for M$+ but supexperiment was done both with and without0.04% Triton X- ported a much lower level of activity (22, 23). Red blood cell 100 with the total phospholipid concentration held constant PtdInsP kinase, like the rat brain (22) and Friend murine at 200 pg ml-'. In the absence of detergent, optimal PtdInsP erythroleukemia cell (27) enzymes, showed enhanced activity kinase activity was obtained at a Ptd1nsP:PtdSer ratio of in the presence of PtdSer. PtdSer increased activity 3-4-fold 2575 (Fig. 10A). PtdInsP kinase activity at this ratio was 2- both with and without detergent. Although the same maximal

5088

Characterization and

Purification of Membrane-associated PtdInsP

4. Moolenaar, W. H., Tertoolen, L. G. J., and de Laat, S. W. (1984) Nature 3 1 2 , 371-374 5. Rosoff, P. M., and Cantley, L.C. (1985) J. Biol.Chem. 2 6 0 , 14053-14059 6. Monaco, M. E. (1982) J . Biol. Chem. 267,2137-2139 7. Fain, J. N., and Berridge, M. J. (1979) Biochem. J. 180,655-661 8. Chahwala, S. B., Fleischman, L. F., and Cantley, L. (1987) Biochemistry 26,612-622 9. Pike, L. J., and Eakes, A. T. (1987) J. Biol. Chem. 2 6 2 , 16441651 10. Taylor, M. V., Metcalfe, J. C., Hesketh, T. R., Smith, G. A., and Moore, J. P. (1984) Nature 3 1 2 , 462-465 11. Kai, M., Salway, J. G., Michell, R.H., and Hawthorne, J. N. (1966) Bwchem. Biophys Res. Commun. 22,370-375 12. Hawthorne, J. N. (1964) Vitamins Hormones 22,57-79 13. Wagner, H., Lissau, A., Holzl, J., and Horhammer, L. (1962) J . Lipid Res. 3, 177 14. Brockerhoff, H., and Ballou, C. E. (1962) J. Biol.Chem. 237, 49-52 15. Ellis, R. B., and Hawthorne, J. N. (1962) Biochem. J. 84, 19P20P 16. Porter, F. D., Li., Y.-S., and Deuel, T. F. (1988) J. Biol. Chem. 263,8989-8995 17. Whitman, M., Downes, C. P., Keeler, M., Keller, T., and Cantley, L. (1988) Nature 3 3 2 , 644-646 18. Endeman, G., Dunn, S. N., and Cantley, L. C. (1987) Biochemistry 26,6845-6852 19. Whitman, M.,Kaplan, D., Roberts, T. M., and Cantley, L.C. (1987) Bwchem. J. 2 4 7 , 165-174 20. Van Rooijen, L. A.A., Rossowska, M., and Bazan, N. G. (1985) Biochem. Biophys. Res. Commun. 126,150-155 21. Imai, A., Rebecchi, M. J., and Gershengorn, M. C. (1986) Biochem. J. 240,341-348 22. Cochet, C., and Chambaz, E. M.(1986) Biochem. J. 237, 25-31 23. Van Dongen, C. J., Zwiers, H.,and Gispen, W. H. (1984) Biochem. J. 223,197-203 24. Kai, M., Salway, J. G., and Hawthorne, J. N.(1968) Biochem. J. 106,791-797 Acknowledgments-We are grateful to Andrea Graziani for his 25. Laemmli, U. K. (1970) Nature 227,680-685 invaluable advice on renaturation of activity from SDS-PAGE and 26. Lowry, 0.H.,Rosebrough, N. J., Farr, A. L., and Randall, R. J. to John Lowenstein for his generous gift of phosphoinositide-specific (1951) J. Biol. Chem. 1 9 3 , 265-275 phospholipase C. Helpful suggestions contributed by Leslie Serunian, 27. Schulz, J. T. (1987) Mechanisms of Ouabain Resistance. Ph.D. Kurt Auger, and Stephen Soltoff were greatly appreciated. We also Thesis, Department of Biochemistry and Molecular Biology, Harvard University thank Marilyn Keeler for her expertise in HPLC analysis. 28. Saltiel, A. R., Fox, J. A., Sherline, P., Sahyoun, N., and Cuatrecasas, P. (1987) Biochem. J. 241,759-763 REFERENCES 29. Ling, L. E. (1988) Characterization of Two Membranebound Proteins: (Na,K)ATPuse and Phosphatidylinositol-4-phos1. Berridge, M. J. (1984) Biochem. J. 2 2 0 , 345-360 phate Kinuse. Ph.D. Thesis, Department of Biochemistry and 2. Nishizuka, Y.(1984) Nature 3 0 8 , 693-698 Molecular Biology, Harvard University 3. Michell, R. H. (1982) Cell Calcium 3,429-444

activity was reached with or without detergent, the optimal ratio of PtdInsP to PtdSer was increased in detergent, a result which is most likely due to dilution of PtdInsP by detergent. The red blood cell enzyme was inhibited by its product, PtdInsPz. Inhibition of PtdInsP kinase activity byexogenously added PtdInsP, has been observed previously in soluble fractions of retinal (20) and pituitary cell (21) homogenates. In contrast, PtdInsP kinase activity inparticulate fractions of pituitary cell homogenates has been found to be activated by PtdInsP, (21). The significance of these varied results is difficult to assess since the phospholipid environment (membranes uersus vesicles and micelles) has not been characterized in these systems. However, these results are useful for optimizing activity assays under given conditions. It has not been determined if there is heterogeneity in the enzyme itself which may have led to differences in the effect of PtdInsP,. As discussed previously, PtdInsP kinase has been found in both soluble andparticulate fractions in several different tissues (20-24). Hypotonically lysed red blood cells contained a roughly equal distribution of activity between the membrane and cytosolic fractions which is similar to the distribution found in rat brain (22). The membrane-associated form of PtdInsP kinase in red blood cells appeared to be a peripheral membrane protein since it was easily solubilized in the absence of detergent by high ionic strength. Given the loose association of red blood cell PtdInsP kinase with the membrane and its similarities to thesoluble enzyme from rat brain, it will be of interest to determine if the soluble and membrane-bound forms are composed of two distinct isozymes or a single enzyme whichdistributes between these two locations.

Suggest Documents