Hydrolysis of phosphatidylinositol 3,4-bisphosphate by inositol polyphosphate 4-phosphatase isolated by affinity elution chromatography.

Inositol polyphosphate 4-phosphatase is a monomeric 110-kDa protein that hydrolyzes two substrates in the inositol phosphate pathway. Inositol 3,4-bisphosphate is converted to inositol 3-phosphate, and inositol 1,3,4-trisphosphate is converted to inositol 1,3-bisphosphate. We have exploited the fact that inositol hexasulfate inhibits the enzyme to devise an affinity elution scheme from a Mono S cation exchange column that resulted in an 11,300-fold purified preparation of rat brain 4-phosphatase. The resulting 4-phosphatase hydrolyzed phosphatidylinositol 3,4-bisphosphate to phosphatidylinositol 3-phosphate with a first order rate constant 120-fold greater than that for inositol 3,4-bisphosphate and 900-fold greater than that for inositol 1,3,4-trisphosphate. This is now the third example wherein the same enzyme hydrolyzes both an inositol lipid and its analogous inositol phosphate.

Inositol polyphosphate 4-phosphatase is a monomeric 110-kDa protein that hydrolyzes two substrates in the inositol phosphate pathway. Inositol 3,4-bisphosphate is converted to inositol 3-phosphate, and inositol 1,3,4trisphosphate is converted to inositol 1,3-bisphosphate. We have exploited the fact that inositol hexasulfate inhibits the enzyme to devise an affinity elution scheme from a Mono S cation exchange column that resulted in an 11,300-fold purified preparation of rat brain 4-phosphatase. The resulting 4-phosphatase hydrolyzed phosphatidylinositol 3,4-bisphosphate to phosphatidylinosi-to1 3-phosphate with a first order rate constant 120-fold greater than that for inositol 3,4-bisphosphate and 900fold greater than that for inositol 1,3,4-trisphosphate. This is now the third example wherein the same enzyme hyrolyzes both an inositol lipid and its analogous inositol phosphate.
The agonist-stimulated formation of inositol lipids phosphorylated at the 3-position has recently been demonstrated in many cell types (1,2). These lipids include phosphatidylinositol 3-phosphate ( P t d I n~( 3 ) P ) ,~ phosphatidylinositol 3,4-bisphosphate (PtdIns(3,4)P2), and phosphatidylinositol 3,4,5-trisphosphate (PtdIns (3,4,5)P3). Stimulation with met-Leu-Phe leads to a rapid and transient increase in the last two compounds mentioned above in neutrophils (3)(4)(5) while responses to growth factors in proliferating cells result in more sustained increases in these compounds (6). In human platelets, PtdIns (3,4)P2 and PtdIns (3,4,5)P3 are synthesized transiently in response to thrombin (4,5). The thrombin-dependent synthesis of PtdIns (3,4)P2 in platelets is blocked by the peptide RGDS that inhibits fibrinogen binding to (YIIb+3 integrin and blocks platelet aggregation. Platelets obtained from patients with thrombasthenia lack ( Y , ,~-P~ integrin and have attenuated synthesis of PtdIns (3,4)P2 upon thrombin stimulation suggest-* This research was supported by Grants HL 14147 (Specialized Center for Research in Thrombosis), HL 16634, and HG 00304 and Training Grant HL 07088 from the National Institutes of Health. 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  ing a role for PtdIns (3,4)P2 in signaling following platelet aggregation (7). Chinese hamster ovary cells expressing a mutant PDGF receptor do not synthesize 3-phosphate-containing inositol lipids in response to PDGF and do not undergo mitogenesis (8,9). The physiological target(s) for these potential second messengers is not known. However, the 6 isozyme of protein kinase C has been shown to be stimulated in vitro by PtdIns(3,4)P2 and PtdIns (3,4,5)P3 (10). The quantities of these lipids even at the point of maximal stimulation of cells is small compared with that of other phosphatidylinositols. In a recent measurement of the mass of PtdIns (3,4)P2 in thrombin-stimulated platelets it was found that the level of this lipid was about one-tenth that of PtdIns(4,5)P2 (11).
Several pathways for the metabolism of 3-phosphate-containing inositol lipids have been suggested. In human platelets and NIH 3T3 cells, labeling studies indicate that phosphorylation occurs stepwise beginning with PtdIns(3)P with subsequent phosphorylation of the 4-and 5-positions (12,13). A recent study indicates that this same pathway accounts for the production of these lipids in plants (14). A 4-kinase present in human platelets that phosphorylates PtdIns(3)P has been described (15). However, PtdIns 3-kinase can phosphorylate both PtdIns(4)P and PtdIns(4,5)P2 in vitro (16). Other labeling studies suggest that PtdIns (3,4,5)P3 is formed by the phosphorylation of the 3-position of PtdIns(4,5)P2 in neutrophils (17).
The pathways for degradation of PtdIns(3,4)Pz and PtdIns(3,4,5)P3 have not been characterized. They are not hydrolyzed by any known phospholipase C isoenzymes, and thus it appears most likely that they are degraded by phosphatases (18,19). PtdIns(3)P is degraded by a 3-phosphatase (20).

Material~-[~H]Ins(1,3,4)P~
Assay of Inositol Polyphosphate 4-Phosphatase-The assay for inositol polyphosphate 4-phosphatase using the soluble substrates was performed as previously described (25) in 20 pl of 50 mM Mops and 10 m EDTA (assay buffer 1). The enzyme fractions obtained following the affinity elution purification step were assayed using assay buffer 1 with the addition of 0.3% hexadimethrine bromide in order to reverse the In&, inhibition. First order rate constant assays for soluble inositol phosphates were performed in 20 pl of 50 mM Mops, 10 mM EDTA, 200 mM NaCl, and 0.3% n-octyl glucoside (assay buffer 2) with 3000 cpm of [3HIIns(3,4)Pz or [3HlIns(1,3,4)P3 (17 mCi/mmol) without additional unlabeled substrate. The first order rate constant assay for [32P1PtdIns (3,4)Pz was carried out in 20 pl using assay buffer 2, and the reaction was stopped by the addition of 30 pl of chlorofodmethanol (l:l), and the fraction of [32PlPtdIns (3,4)Pz converted to [32P1PtdIns(3)P was determined by TLC.
Preparation of [32P]PtdIns(3,4)P2-The procedure for preparation of ["P1PtdIns(3,4)P2 was adapted from that of Carpenter et al. (16). NIH 3T3 cells were grown to 80% confluence on T75 flasks in Dulbecco's modified Eagle's medium containing 10% calf serum and then incubated in 0.5% calf serum for 24 h after which the media were removed and the cells were incubated at 37 "C for 15 min with media containing PDGF (BB) (100 nglml). The media were decanted, and the cells were washed with 10 mM sodium phosphate, pH 7.0, containing 0.15 M NaCl (phosphate-buffered saline). Lysis buffer (1.4 ml containing 20 mM Tris, pH 7.4,50 mM NaC1, 50 mM NaF, 20 m sodium pyrophosphate, 1% Triton X-100, 200 p~ sodium orthovanadate, 1 mM PMSF, and 20 pg of leupeptidml) was added to each plate, and the cells were harvested by scraping. The lysate was incubated for 20 min on ice and then centrifuged a t 5000 x g for 10 min. Polyclonal rabbit anti-human PDGF receptor type B antibody (0.1 mVmg lysate protein) was added, and samples were rotated for 3 h at 4 "C. Protein A-Sepharose CL-4B (0.2 volume of a 1:l suspension in lysis buffer) was added and further rotated for 1.5 h a t 4 "C. The protein A-Sepharose was washed 3 times each with the following buffers: phosphate-buffered saline containing 1% Nonidet P-40 and 200 VM sodium orthovanadate; 100 m Tris, pH 7.5, containing 500 m LiCl and 200 PM sodium orthovanadate; 10 mM Tris, pH 7.5, containing 100 mM NaCl, 1 m EGTA, and 1 mM EDTA, 20 mM HEPES pH 7.4 containing 0.5 mM EDTA and 1 mM MgC1,.
PtdIns(4)P (1 mg) and phosphatidylserine (0.8 mg) in CHC1, were dried under Nz and suspended in 2 ml of 20 mM HEPES pH 7.4 containing 0.5 mM EDTA and 1 mM MgCl,. The lipid mixture was then sonicated on ice at 100 watts for 2 min with a probe sonicator, and then 2 ml of immunoprecipitate ( l : l , v/v) was added. [Y-~~PIATP (3 mCi of 3000 Ci/mmol) was then added, and the mixture was rotated for 1 h at 37 "C. The reaction mixture was extracted using chlorofodmethanol, and the organic layer was washed twice with a n equal volume of 2 M KCl. A typical preparation yielded 5-10 pCi of [32PlPtdIns (3,4)Pz. PtdIns(3,4,5)P3 and PtdIns(3)P were prepared identically except that PtdIns(4,5)P2/phosphatidylserine vesicles ( l : l , moVmo1) and PtdIns vesicles were used, respectively.
Thin Layer Chromatography-Silica Gel 60 TLC plates were treated with a solution of 1% potassium oxalate in 50% ethanol. The plates were then placed in a 90 "C oven for at least 30 min prior to use. TLC plates were developed using a solvent mixture of chlorofodacetone/ methanoVglacia1 acetic acid/water (80:30:26:24: . Phospholipid standards (PtdIns, PtdIns(4)P, and PtdIns(4,Ei)P.J were stained with iodine. Radiolabeled phospholipids were detected by autoradiography using HyperfilmTM-MP (Amersham Corp.). Radiolabeled phospholipids were scraped from the plates and mixed with 10 ml of Aquas-sureTM scintillation mixture (DuPont) and counted in a Beckman liquid scintillation counter.
Proof of Product-Deacylation and deglyceration of inositol lipids were performed as described previously (20). Water-soluble phosphoinositols were separated by HPLC on a Whatman Partisil 10 SAX column with a flow rate of 1 mVmin using the following gradients of ammonium formate, pH 3.5: a linear gradient of 40-425 m over 30 min, a step to 810 m followed by a linear gradient to 1.3 M over 33 min, a step to 1.7 M followed by a linear gradient to 3 M over 57 min.
Preparation of Rat Brain Homogenate, Phosphocellulose Batch Chromatography, a n d DEAE-HPLC-Rat brain homogenate was prepared from 3000 frozen unstripped rat brains (wet weight, 4453 g). Phospho-cellulose batch chromatography and DEAE-HPLC were performed as previously described (22) except that 0.1 rn EDTA was included in the DEAE equilibration and elution buffers. DEAE fractions were assayed for inositol polyphosphate 3-phosphatase and inositol polyphosphate 4-phosphatase, and fractions containing high 3-phosphatase activity were excluded from the fractions pooled for the further purification of 4-phosphatase. The 4-phosphatase was then precipitated by the addition of ammonium sulfate over 10 min at 4 "C with stirring to give a 35% saturated solution, and after an additional 20 min of stirring the suspension was centrifuged for 15 min at 20,000 x g. The supernatant was discarded, and the pellet was resuspended (20 mg proteidml) in 20 m Tris-C1, pH 8.0, containing 20 m 2-mercaptoethanol, 20% glycerol, 1 rn EDTA, 1 mM PMSF, and 0.5 mM benzamidine and frozen with liquid N2 and stored a t -135 "C.
Affinity Elution Chromatography Using Mono S-A Mono S 5/5 column (Pharmacia LKB Biotechnology Inc.) was equilibrated with buffer A (20 mM Tricine, pH 8.5, 2 m EDTA, 20% glycerol, 1 mM dithiothreitoll A sample of the partially purified inositol polyphosphate 4-phosphatase was thawed and then loaded onto the column, and the column was washed with 12 ml of buffer A. The column was then washed with 8 ml of buffer B (20 nm HEPES pH 7.2,2 mM EDTA, 20% glycerol, 1 nm dithiothreitol). The column was then washed with 4 ml of buffer B containing 0.1 m inositol hexaphosphate (InsP,). The 4-phosphatase was then eluted with 10 ml of buffer B containing 0.5 mM InsS,. The enzyme activity was divided into two pools. The fractions that contained the peak of protein eluting with InsS6 and those in later fractions were placed in pools 1 and 2, respectively. Pooled fractions were concentrated in a Micro-ProDiCon against buffer B containing 0.5 mM PMSF. The concentrated enzyme was stored at -135 "C.
Miscellaneous Methods-Protein was measured using a Bio-Rad protein dye reagent, and phosphate was measured by the method of Ames and Dubin (26).

RESULTS AND DISCUSSION
The inositol polyphosphate 4-phosphatase is a divalent cation-independent enzyme that hydrolyzes the 4-phosphoester of Ins(3,4)P2 and Ins(1,3,4)P3. The 4-phosphatase has previously been purified 3390-fold from calf brain and shown to be a monomeric protein of 110 kDa (25). We have now purified the rat brain enzyme using a novel affinity chromatography procedure. The initial steps in the purification are batch phosphocellulose chromatography, followed by DEAE-HPLC and precipitation by 35% saturated ammonium sulfate as described under "Experimental Procedures." In searching for further purification methods, we found that InsPs and InsSs were potent inhibitors of inositol polyphosphate 4-phosphatase (ICso of 50 and 10 PM, respectively), making them good candidates for eluants in an affinity elution scheme. Affhity elution chromatography involves the binding of a protein mixture to a nonspecific column matrix followed by elution with a specific ligand that forms a complex with the desired protein. This technique has been em-     (VM 1ns(1,3,4) ployed to purify various glycolytic enzymes and tRNA synthetases (27,28).
The partially purified enzyme was bound t o a Mono S column at pH 8.5. The column was then washed extensively and reequilibrated to pH 7.2. The major contaminating protein with a molecular mass of 160 kDa was then eluted with 0.1 mM InsP6. The 4-phosphatase was then eluted with 0.5 mM InsS6 with 60-80% recovery in several experiments (Fig. 1). The problem of assaying fractions containing inhibitor ligands was solved by the addition of the polycationic polymer, hexadimethrine bromide, to the assay buffer, which reversed both InsPs and InsS6 inhibition. Since the activity eluted as a broad peak that did not correspond to the protein profile, the active fractions were combined into two pools. Pool 1 (fractions 15 and 16) contained several contaminating proteins whereas pool 2 (fractions 17-19) had a significantly higher specific activity as shown in Table I. When analyzed by SDS-polyacrylamide gel electrophoresis, pool 2 was found to contain a 110-kDa protein, the same molecular mass as described previously for bovine brain inositol polyphosphate 4-phosphatase, with small amounts of lower molecular mass contaminants. The 11,300-fold purified pool 2 rat brain enzyme was used for further experiments. It is possible that other inositol phosphate binding proteins could be purified by similar elution chromatography schemes and that InsP&sS6 elution chromatography may be a n economical alternative to tethered InsPs affinity chromatography (29). pmol/min/mg of protein, respectively. The inhibition by InsPs was characterized as competitive as shown in Fig. 2 with an apparent Ki of 19 PM. The concentration of InsPs in HL60 cells has been reported to be 50 PM (30). InsPs may therefore inhibit cytosolic 4-phosphatase suggesting a possible role in regulating enzyme activity.
The rate of hydrolysis of PtdIns(3,4)P2 using 0.3 and 0.6 ng/ml pool 2 is shown in Fig. 3. An interesting characteristic of 4-phosphatase hydrolysis of the lipid substrate is that the amount of substrate hydrolyzed was directly proportional to the amount of enzyme added over a wide range of enzyme concentrations. The maximum fraction of PtdIns(3,4)Pz hydrolyzed was approximately 25 and 50% upon the addition of 0.3 or 0.6 ng/ml pool 2, respectively, as shown in Fig. 3. The fraction of PtdIns (3,4)Pz hydrolyzed by 0.3 ng/ml was doubled to 50% when the phospholipid concentration in the assay was halved, and when the enzyme concentration was raised to 5 ng/ml all of the substrate was hydrolyzed. Furthermore, the apparent first order rate constants for the hydrolysis of PtdIns (3,4)Pz by these two enzyme concentrations are essentially identical as shown in Fig. 5 , yielding values of 2.6 and 2.7 x s-l for 0.3 and 0.6 ng/ml, respectively. This deviates from the ideal behavior of first order enzyme kinetics in which the rate constant is directly proportional to enzyme concentration. An explanation for this behavior is that irreversible binding of the 4-phosphatase to the surface of the phospholipid vesicles prevents intervesicle enzyme movement. For the case wherein the number of enzyme molecules is less than the number of phospholipid vesicles, the fraction of PtdIns (3,4)Pz hydrolyzed is proportional to the fraction of vesicles that possess at least one bound enzyme molecule and hence proportional to the amount of enzyme added. The observed rate constant is insensitive to the concentration of enzyme until the number of enzyme molecules is greater than the number of phospholipid vesicles. Consistent with this hypothesis, the preincubation of enzyme with phosphatidylserine-PtdIns(4)P vesicles used in the assays inhibits subsequent PtdIns(3,4)P2 hydrolysis.
The first order rate constants for hydrolysis of Ins(1,3,4)P3 and Ins(3,4)Pz were determined using trace-labeled substrate. The observed rate constants for Ins(1,3,4)P3 and Ins (3,4)Pz in the presence of 0.6 ng/ml pool 2 were 3.0 x s-l and 2.2 x s-', respectively, as shown in Fig. 6. The comparison of the lipid versus the soluble substrates by first order rate constants is confounded by the fact that the observed first order rate constant for the lipid substrate is independent of the amount of enzyme added. However, the apparent first order rate constant for hydrolysis of PtdIns(3,4)P2 under these conditions was 900fold greater than that of Ins(1,3,4)P3 and 120-fold greater than that of Ins (3,4)Pz. While the V,,,,JK, values for the two watersoluble substrates under the conditions used for Michaelis kinetics were similar, the apparent first order rate constant for Ins (3,4)Pz obtained using assay buffer 2 is higher than that of Ins(1,3,4)P3 (Fig. 6) because NaCl stimulates the hydrolysis of Ins (3,4)Pz by 4-phosphatase.
In order to confirm that the activity responsible for the hydrolysis of the soluble and lipid substrates was a result of the same enzyme, a heat inactivation experiment was performed. Since the amount of [32PlPtdIns (3,4)Pz hydrolyzed is directly proportional to the amount of 4-phosphatase added, the amount of lipid phosphatase activity remaining at various time points could be determined by the maximum [32P]PtdIns(3,4)Pz hydrolyzed during the assay. This was confirmed in a preliminary experiment where it was shown that the fraction of substrate hydrolyzed was linear from 0.075 to 0.6 ng of enzyme/ml. The rate of activity loss using PtdIns (3,4)P2 (t,A = 10 min) and Ins (3,4)Pz was the same upon heating at 45 "C, indicating that the same enzyme hydrolyzes both substrates as shown in Fig.  7.