Post-translational modification of human brain type I inositol-1,4,5-trisphosphate 5-phosphatase by farnesylation.

In brain, type I inositol-1,4,5-trisphosphate 5-phosphatase (InsP3 5-phosphatase) is the major isoenzyme hydrolyzing the calcium-mobilizing second messenger InsP3. Activity of this enzyme could be measured in both soluble and particulate fractions of tissue homogenates. The protein sequence showed a putative C-terminal isoprenylation site (CVVQ). In this study, two mutants have been generated. The first mutant (C409S) has a serine replacing a cysteine at position 409 of the wild-type enzyme. The second mutant (K407D1) is a deletion mutant that lacks the last five C-terminal amino acids. These constructs were individually expressed by transfection in COS-7 cells. Western blot analysis of wild-type transfected cells indicated that both soluble and particulate fractions had a 43-kDa immunoreactive band, with a higher proportion of the original homogenate associated with the particulate part. On the contrary, when the two mutated constructs were transfected in COS-7 cells, the phosphatase was predominantly soluble. Confocal immunofluorescence studies showed the wild-type enzyme to be present on the cell surface of transfected COS-7 cells and in subcellular compartments around the nucleus. This was not observed for the two mutants, where uniform immunofluorescence labeling was observed throughout the cytosol. Recombinant type I InsP3 5-phosphatase expressed in Escherichia coli was a substrate of purified farnesyltransferase. Altogether, the data therefore suggest a direct participation of Cys-409 in a C-terminally anchored InsP3 5-phosphatase by farnesylation.

In response to several extracellular signals, two second messengers, inositol 1,4,5-trisphosphate (InsP 3 ) 1 and diacylglyc-erol, are produced (1). InsP 3 mobilizes intracellular Ca 2ϩ (2), while diacylglycerol is the specific activator of C-type protein kinases (3). InsP 3 concentration is dependent upon the relative activities of phospholipase C, together with several enzymes that either phosphorylate or dephosphorylate this molecule: InsP 3 5-phosphatase is responsible for dephosphorylation into inositol 1,4-bisphosphate (4), while InsP 3 3-kinase leads to the synthesis of inositol 1,3,4,5-tetrakisphosphate (InsP 4 ) (5). The InsP 4 molecule itself has been shown to interact with several candidate receptors. It is involved, at least in some cells, in the mechanism of Ca 2ϩ entry and/or in the regulation of neurotransmitter release (6 -10). The recent cloning of a cDNA encoding an InsP 4 -binding protein in porcine platelets shows that it is a member of the GTPase-activating protein family (11).
Three isoenzymes of InsP 3 5-phosphatase have been described: (a) a type I 43-kDa protein initially reported in human erythrocyte membranes (4) and later purified from many sources (12)(13)(14)(15)(16)(17), (b) a 75-kDa protein originally isolated from human platelets (18), and (c) a protein identified due to its deficiency in the Lowe or oculocerebrorenal syndrome (19). A truncated mutant of this protein was recently shown to possess InsP 3 , InsP 4 , and phosphatidylinositol-4,5-bisphosphate 5-phosphatase activities (20). Type I InsP 3 5-phosphatase hydrolyzes both InsP 3 and InsP 4 , with higher affinity for InsP 4 (K m ϭ 1 M versus 10 M for InsP 3 ), but lower velocity (ratio of V max ϭ 11 in favor of InsP 3 ) (14). Phosphatidylinositol 4,5-bisphosphate is not a substrate (21). In many tissues, the activity appears to be associated with the particulate fraction of cell homogenates; in rat brain, for example, only 5% of the total activity of homogenates could be measured in the soluble part of the tissue (22). In intestinal epithelial cells, the activity was much greater in the basolateral region of the cell in comparison with the brush border, suggesting that the sites of InsP 3 generation and inactivation are in close proximity (23). This is opposite to the polarized distribution of particulate 5-phosphatase in hepatocytes (24). In these studies, however, cellular distribution was determined by enzymatic assay following cell fractionation without any distinction between possible isoenzymes. cDNAs encoding type I InsP 3 5-phosphatase have been isolated (25)(26)(27). All studies have demonstrated a putative Cterminal isoprenylation site (CVVQ). Insertion of type I InsP 3 5-phosphatase into membranes could therefore be assured by post-translational modification of this motif (28). This hypothesis was tested in the present study by comparison for the first time of biochemical data and immunofluorescent localization of the phosphatase in transfected cells. Labeling of COS-7 cells at the cell surface and in subcellular compartments surrounding the nucleus could be shown for intact InsP 3 5-phosphatase, but not for two mutants of the isoprenylation motif.

Materials
[ 3 H]InsP 3 , farnesyl pyrophosphate (FPP), geranylgeranyl pyrophosphate (GGPP), and ENHANCE solution for fluorography were from DuPont NEN. Alkaline phosphatase-coupled anti-rabbit IgG was from Sigma. Geneticin (G418), Taq DNA polymerase, and 5-bromo-4-chloro-3-indolyl phosphate and nitro blue tetrazolium reagents were from Life Technologies, Inc. pMal-cRI expression vector was from New England Biolabs Inc. pTrcHis expression vector was from Invitrogen. pcDNA3 expression vector was from Stratagene. Rainbow colored protein molecular mass markers, 125 I-labeled protein A, and Hyperfilm-MP were from Amersham Corp. Protan BA79 cellulose nitrate membranes were from Schleicher & Schuell, Zwittergent 3-14 was from Calbiochem, and Pefabloc was from Pentapharm. The Sequenase Version 2.0 DNA sequencing kit was purchased from U. S. Biochemical Corp. Fluoresceinlabeled goat secondary antibody and Texas Red-labeled swine secondary antibody were from Prosan. Insta-Gel was from Packard Instrument Co. Glass-fiber filters (2.5 cm) were from Whatman. Soluble and particulate fractions from rat cerebellum homogenates were prepared as described previously (22). Wild-type and mutant CVLL-type Ha-Ras proteins were from BIOMOL Research Labs Inc. Purified preparations of FTase and geranylgeranyltransferase-1 (GGTase-1) were prepared as described (29,30) and kindly provided by Dr. Patrick Casey (Duke University).

Methods
Generation of C-terminal Mutants-Two mutants were prepared. For the first (C409S), the cysteine at position 409 (26) was replaced by a serine molecule. The second mutant (K407D1) is a deletion mutant that lacks the last five C-terminal amino acids (i.e. CCVVQ). Site-directed mutagenesis was performed using polymerase chain reaction amplification with plasmid DNA (clone D1 in Ref. 26) using the 5Ј-primer containing an AcyI restriction site (underlined), 5Ј-TACCCGTACAGT-GAGGACGCC-3Ј, and the 3Ј-primer containing the mutated codon (in boldface for C409S) and an EcoRI restriction site (underlined), 5Ј-TCGGAATTCTCACTGCACGACACTACACTTGTG-3Ј. The polymerase chain reaction product (250 base pairs) was isolated as an AcyI/EcoRI fragment and directly subcloned into pBluescript containing the 5Ј-end of the D1 plasmid. K407D1 was prepared by polymerase chain reaction using the same 5Ј-primer as before and the 3Ј-primer with an EcoRV site (underlined): 5Ј-GATCGATATCTCACTTGTGCACATGGGCATGA-GG-3Ј. As before, the polymerase chain reaction product was introduced into pBluescript and sequenced to verify the absence of unwanted mutations. A BamHI/EcoRV fragment of each construct in pBluescript containing the coding sequence and expected mutation was subcloned in pcDNA3 eukaryotic expression vector. The same restriction sites were used for the wild-type plasmid for subcloning in pcDNA3. Alternatively, a BamHI/SalI fragment was subcloned in pMal expression vector digested with the same restriction enzymes for expression in Escherichia coli.
Transient Expression in COS-7 Cells and Preparation of Cell Extracts-Twenty-four hours after seeding (1 ϫ 10 6 cells/6-cm dish), COS-7 cells were transfected with the DNA of each construct using the DEAE-dextran method (29). COS-7 cells were grown in a humidified atmosphere of 5% CO 2 at 37°C and maintained in Dulbecco's modified Eagle's medium supplemented with 1% sodium pyruvate, 1% Fungizone, 2% penicillin/streptomycin, and 10% fetal calf serum. After 48 h, the transfected cells were harvested, pelleted, and resuspended in 500 l of homogenization buffer A (15 mM Tris-HCl pH 7.5, 2 mM MgCl 2 , 0.3 mM EDTA, pH 7.5, 1 mM EGTA, 2.5 M leupeptin, and 0.4 mM Pefabloc). The cells were homogenized by passage through a 26-gauge needle (ϳ10 times). The cell lysate was centrifuged at 80,000 ϫ g for 30 min at 4°C. The resulting supernatant was transferred to a new tube, and the pellet was resuspended in the same volume of buffer B (50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 0.5 mM EDTA, and protease inhibitors) and, in some experiments (indicated in the figure legends), recentrifuged under the same conditions. The resulting supernatant was combined with the first supernatant (total soluble fraction), and the pellet (washed particulate fraction) was resuspended in buffer B. The volume of the washed particulate fraction was the same as the total volume of the soluble fraction.
Stable Expression in CHO cells-The DNA encoding type I InsP 3 5-phosphatase was transfected in CHO cells using the calcium phosphate precipitation method (32). Cells were transfected in 6-cm diameter culture dishes. Two days after transfection, selection for transfected cells started by the addition of fresh complete medium (Ham's F-12 medium supplemented with 10% fetal calf serum, 1% Fungizone, and 2% penicillin/streptomycin) containing 400 g/ml Geneticin (G418). After death of all untransfected cells, 16 G418-resistant clones were isolated and transferred into 9-cm diameter culture dishes. The medium was changed every 48 h. The CHO cells were maintained in complete medium containing Geneticin in an atmosphere of 5% CO 2 at 37°C. Out of 16 clones tested, four were positive, with InsP 3 5-phosphatase activity ranging from 10-to 30-fold over untransfected cells.
Confocal Immunofluorescence Microscopy-Immunofluoresence using antiserum raised against type I InsP 3 5-phosphatase was performed on both transfected COS-7 and CHO cells and on cerebellum sections of Wistar rats. Cerebellum was rapidly fixed in freshly prepared 4% paraformaldehyde solution in 0.1 M phosphate-buffered saline at pH 7.4 for 24 h at 4°C. The tissue was cryoprotected in graded solutions of sucrose in phosphate-buffered saline (10, 20, and 30%; 1 day each). Coronal sections of 15 m were prepared using a Leitz cryostat, mounted onto poly-L-lysine-coated slides, and stored at Ϫ20°C. Transfected COS-7 and CHO cells were grown on uncoated glass coverslips. Coverslips (cryostat sections) were rinsed with Tris-buffered saline (TBS) and then fixed in 4% paraformaldehyde in TBS solution for 10 min. The cells were washed three times for 10 min with TBS and then permeabilized with 0.15% Triton X-100 in TBS for 5 min and washed again with TBS (three times). The fixed cells were incubated for 1 h at room temperature with 1:20 normal serum in TBS (swine, goat, or sheep serum depending on the origin of the secondary antiserum). Incubation with immune serum was performed overnight in the presence of blocking serum diluted 1:20 in TBS. The antiserum to InsP 3 5-phosphatase was used at a 1:250 dilution. Controls were performed using preimmune serum at the same dilution. In some experiments, serum was purified as previously reported (33). After being rinsed with TBS, cells were incubated for 60 min in the dark with either a fluorescein-labeled goat secondary antibody or a Texas Red-labeled swine secondary antibody at a 1:30 dilution. The cells were then washed three times with TBS for 10 min, rinsed with water, and mounted with a drop of Gelvatol solution containing 100 mg/ml Dabco reagent.
It was verified that labeling was observed neither on cells transfected with the pcDNA3 vector alone and using anti-InsP 3 5-phosphatase antiserum nor on transfected cells using the preimmune serum. Labeling using the purified antibodies was qualitatively identical to that obtained with the total antiserum. Although these controls were all consistent with the detection of the genuine InsP 3 5-phosphatase, we cannot exclude cross-reactions with unknown related proteins, and all results should therefore be read as InsP 3 5-phosphatase-like immunoreactivity.
Cells were observed under a Nikon Optiphot fluorescence microscope, and images were obtained using a laser-scanning confocal microscope (MRC 1000, Bio-Rad) equipped with an argon-krypton laser and COMOS software (Bio-Rad). Images were further analyzed using Imagespace software (Molecular Dynamics, Inc.).
Protein Prenyltransferase Assay-Protein prenyltransferase activities were determined by quantitating the amount of 3  with GGTase-1. Incubation was for 30 min at 37°C before acid precipitation (4% SDS, 30% trichloroacetic acid). The tubes were vortexed and left on ice for 1 h, after which 2 ml of 2% SDS, 6% trichloroacetic acid solution was added. Each mixture was filtered on a 2.5-cm glass-fiber filter. The tubes were washed with 10 ml of 6% trichloroacetic acid, dried, and counted with 5 ml of Insta-Gel in a scintillation counter. Alternatively, after the incubation at 37°C, 5 l of denaturing buffer was added, and the samples were separated by SDS-polyacrylamide gel electrophoresis (35). After migration, the gel was fixed overnight in 30% methanol, 10% acetic acid, and 10% glycerol. After fluorography in ENHANCE solution for 1 h and a 30-min incubation in 30% glycerol, the gel was dried for autoradiography.
Production of Antiserum against InsP 3 5-Phosphatase-Antiserum against recombinant type I InsP 3 5-phosphatase was produced in New Zealand male rabbits (36). Western blotting analysis was carried out with the polyclonal antiserum diluted 1000-fold. The immunocomplexes were visualized by colorimetric reagents (37) or with 125 I-labeled protein A followed by autoradiography.
Preparation of Plasmids for Expression of InsP 3 5-Phosphatase-Human type I InsP 3 5-phosphatase was subcloned in two bacterial expression vectors: pMal-cRI and pTrcHis. The first construction encodes maltose binding protein and resulted in the expression of a 85-kDa maltose-binding protein-InsP 3 5-phosphatase fusion protein (36). The two mutated recombinant maltose-binding protein fusion proteins were prepared under the same conditions. The second construction (pTrcHis) was expressed as described (36) and applied to 50 ml of Ni 2ϩ -nitrilotriacetic acid-agarose resin. The 43-kDa InsP 3 5-phosphatase was eluted in 100 mM imidazole. Activities of 150 -250 mol/ min/mg of purified enzyme were obtained with both expression vectors. InsP 3 5-phosphatase assay was as described previously (14).

Intracellular Distribution of Type I InsP 3 5-Phosphatase in
Rat Cerebellar Purkinje Neurons-InsP 3 5-phosphatase antibodies recognized a 43-kDa protein band in Western blots of rat cerebellum homogenates (Fig. 1). The presence of an excess (competing) concentration of enzyme used to raise the antiserum blocked such recognition, and no signal could be detected with preimmune serum. The phosphatase was shown to be present in rat cerebellar Purkinje neurons (Fig. 2). As shown by immunofluorescence at low magnification, the Purkinje cells appeared strongly labeled both in their somas and dendritic trees; the nuclei remained unlabeled ( Fig. 2A). At higher magnification, it could be seen that labeling was not uniform since some reticular-like structures were more heavily labeled in the soma (Fig. 2B). In addition, the proximal segments of the apical dendrites were underlined by vesicle-like patches containing high levels of immunoreactivity. This is consistent with in situ hybridization data showing the expression of 5-phosphatase mRNA in the same cells (26).
InsP 3 5-Phosphatase Activity Expressed in COS-7 and CHO Cells-For expression in COS-7 and CHO cells, the InsP 3 5-phosphatase clone (26) was subcloned in pcDNA3 expression vector. When measured in crude COS-7 cell homogenates, activity increased up to 100-fold in InsP 3 5-phosphatase-transfected cells as compared with cells transfected with vector alone (specific activity data of a typical experiment showed an increase from 0.3 to 22 nmol/mg/min). In transfected CHO cells, activity increased up to 50-fold compared with untransfected CHO cells. Crude homogenates were subjected to SDSpolyacrylamide gel electrophoresis with subsequent Western blotting using antibodies to InsP 3 5-phosphatase. A single immunoreactive 43-kDa band could be identified in transfected COS-7 or CHO cells (Fig. 1). No signal could be detected in COS-7 cells transfected with vector alone or in untransfected CHO cells (Fig. 1).
When COS-7 cells were transfected with increasing amounts of DNA, the activity increased markedly until a plateau was reached at 1 g of DNA (data not shown). 5-Phosphatase activity was distributed in soluble and particulate fractions of the homogenates in a 1:2 ratio. The addition of Triton X-100 to the incubation mixture was shown to increase activity in the particulate fraction; this effect was not observed in the soluble fraction. A maximal effect was shown at 0.1% (v/v) detergent with either rat brain or transfected COS-7 cells (data not shown). Western blot analysis indicated that both soluble and particulate fractions of transfected COS-7 cells had a 43-kDa immunoreactive band, with a higher proportion associated with the particulate fraction (Fig. 3A).
Using antibodies able to interact with both the soluble and particulate InsP 3 5-phosphatases, we were able to determine the relative amounts of these enzymes in transfected COS cells. This was done on Western blots by measuring the radioactivity of 125 I-labeled protein A associated with the 43-kDa protein bands. Taking into account the InsP 3 5-phosphatase activity applied to the gel, we have estimated the specific activities of the two fractions of transfected cells. The specific activity of the particulate enzyme was 6-fold lower than that of the soluble enzyme (Table I). Similar results have been obtained by comparing soluble and particulate fractions of rat cerebellum (data not shown). peared to be predominantly soluble as shown by both activity assay and immunoblot analysis (Fig. 3, A and B). The data therefore suggested a direct participation of Cys-409 in a Cterminally anchored InsP 3 5-phosphatase.

C-terminal Mutants and Their Expression in COS-7 Cells-To
Immunofluorescence Microscopy of COS Cells Transfected with DNA of Wild-type and Mutant Enzymes and of Transfected CHO Cells-The cellular distributions of wild-type InsP 3 5-phosphatase and the two mutants were compared using confocal microscopy. Each dish of cells subjected to the transfection protocol contained a mixture of transfected (Ϯ30%) and untransfected cells (Fig. 4D). Immunofluorescence revealed that wild-type InsP 3 5-phosphatase was found along the plasmalemma membrane either as a continuous line (Fig. 4, A and  B) or as discrete patches or vesicle-like structures (Fig. 4C). Within the cell, labeling appeared with a reticular and granular aspect throughout the cell and was prominent around the nucleus (Fig. 4C), therefore resembling the endoplasmic reticulum and Golgi apparatus. In some cells, the immunofluorescence was concentrated on the cell surface, while in others, it was largely found in the perinuclear region.
Both InsP 3 5-phosphatase mutants showed distinct distributions in transfected COS-7 cells: uniform and diffuse labeling was localized throughout the cytosol (Fig. 5). The plasma membrane appeared unlabeled, and no reticular, granular, vesiclelike structures or perinuclear labeling reminiscent of the Golgi apparatus could be recognized.
The immunofluorescence pattern given by anti-5-phosphatase antibodies in COS-7 cells was not confined to this cell line. A similar staining pattern was also seen in transfected CHO cells: a clear and strong staining at the cell plasma membrane, which in most cells appeared as a continuous line with few membrane vesicle-like patches (data not shown).

In Vitro Isoprenylation of Recombinant InsP 3 5-Phosphatase-Purified FTase transferred [ 3 H]farnesyl from [ 3 H]FPP to
Ha-Ras with its normal CAAX box (CVLS). The enzyme also transferred [ 3 H]farnesyl to wild-type InsP 3 5-phosphatase (Fig. 6A). The incorporated radioactivity could be visualized as a band of the expected size (i.e. 43 kDa) on SDS-polyacrylamide gel electrophoresis (data not shown). In our assay, the maximal velocity for the transfer of [ 3 H]farnesyl was ϳ3-fold reduced for InsP 3 5-phosphatase as acceptor compared with Ha-Ras. No transfer could be shown to a Ras mutant in which the last amino acid had been changed to leucine (Ha-Ras-CVLL). There was no appreciable transfer of [ 3 H]geranylgeranyl to any of the protein substrates (Fig. 6B). GGTase-1 transferred [ 3 H]geranylgeranyl to the leucine-terminated mutant form of Ha-Ras, but had only a slight ability to geranylgeranylate InsP 3 5-phosphatase or Ha-Ras-CVLS (Fig. 7B). With GGTase-1, there was no appreciable transfer of [ 3 H]farnesyl to any of the protein substrates (Fig. 7A). DISCUSSION The cloning of cDNAs encoding type I InsP 3 5-phosphatase has provided molecular tools to study the levels of enzyme expression in various cells and its subcellular distribution. Protein kinase C might activate the enzyme, a feedback mechanism that had been suggested in human platelets (38,39). Data obtained with human recombinant type I InsP 3 5-phosphatase (the isoform suspected to be phosphorylated in platelets) have shown that the purified enzyme was not a substrate of protein kinase C (36). Another mechanism was suggested from cDNA analysis: the C-terminal end of type I InsP 3 5-phosphatase shows an isoprenylation site (CVVQ) that may be critically important for subcellular localization and biological activity as well as for interaction with regulatory and effector proteins. The two valine residues are also present in the Cterminal end of one of the Ras proteins (Ni-Ras), and as for Ras (40), positively charged residues are also adjacent to the modified cysteine.
Previous studies in other cells/systems have suggested that InsP 3 5-phosphatase (presumably type I) was preferentially associated with plasma membranes (23,24). The data have been obtained by the use of differential centrifugation of cell homogenates. However, such strategies are subjected to poten-tial problems arising from cellular disruption and subcellular fractionation, where redistribution of proteins may occur. In the present study, the subcellular distribution of type I InsP 3 5-phosphatase enzyme was performed by indirect immunofluorescence. In both transfected COS-7 and CHO cells, anti-InsP 3 5-phosphatase immunofluorescence displayed a reticular pattern of staining throughout the whole cell. Furthermore, a signal at the cell membrane was clearly observed. A crescentshaped staining around the nucleus suggestive of the Golgi apparatus was also observed. All these subcellular localizations (membrane, Golgi apparatus, endoplasmic reticulum, and vesicle-like structures) were clearly superimposable to previously reported localizations of other membrane proteins such as opiate and glutamate metabotropic receptors expressed in COS-7 cells and analyzed using similar techniques (41,42).
Data obtained in the present studies support the post-translational modification of type I InsP 3 5-phosphatase by farnesylation. First, most of the activity present in crude homogenates of rat cerebellum or transfected COS-7 or CHO cells was found in the particulate fraction (illustrated by Western blotting). Deletion of the CAAX motif or the construction of a 5-phosphatase mutant in which the cysteine has been replaced by serine resulted in proteins that were localized largely in the cytosol. This was visualized by activity assay, by Western blotting after disrupting transfected COS-7 cells, or directly by immunofluorescence (where uniform labeling was observed throughout the cytosol). Finally, our data indicated that recombinant type I InsP 3 5-phosphatase was a substrate of purified FTase.
A C-terminal CCVVQ sequence of the 5-phosphatase would be predicted to be perhaps dually prenylated, as in the case of RhoB that ended with CCKVL and was shown to contain both farnesyl as well as geranylgeranyl groups (43). RhoB was farnesylated as well as geranylgeranylated by GGTase-1 (44). Our data indicated that InsP 3 5-phosphatase was not a substrate of GGTase-1, but was farnesylated by FTase. Consistent with this, the carboxyl-terminal residue of the CAAX motif in general determines which isoprenoid will be added. When X is serine, methionine, or glutamine (as in the case of type I InsP 3 5-phosphatase), proteins are recognized by FTase (45). The hepatitis D virus large antigen is a virally encoded protein also containing a C-terminal CAAX motif where X is glutamine. Recent data indicated that this protein is farnesylated as well by the same FTase. 2 The post-translational modification of proteins by isoprenoids facilitates protein-membrane and protein-protein interactions. This was shown, for example, for the ␥-subunit of 2 P. Casey, personal communication. transducin (46). It may also affect the enzymatic activity by a conformational change of the protein. This hypothesis has now been tested for type I InsP 3 5-phosphatase. Using anti-43-kDa protein antisera to titrate the relative amounts of enzyme, we could demonstrate that particulate InsP 3 5-phosphatase had a lower specific activity compared with the soluble enzyme.
Recent data indicated that type II InsP 3 5-phosphatase also shares an isoprenylation motif (CNPL) at the C-terminal end of the protein, although a proline residue at this position is unusual in mammalian systems (47). Immunofluorescent localization of this enzyme in transfected COS-7 cells was not reported. Furthermore, Jefferson and Majerus (47) reported that the cysteine C-terminal mutant was less active relatively than the wild-type enzyme toward the lipid phosphatidylinositol 4,5-bisphosphate as substrate. No effect on activity could be seen when InsP 3 was used as substrate. The data indicated that membrane association was not required for activity. Whether isoprenylation could modify the specific activity of this phosphatase when it is anchored as compared with the soluble enzyme is not known.
Isoprenylation could target the InsP 3 -metabolizing enzyme to specific intracellular distributions, and further studies should establish possible colocalization with the InsP 3 receptor and InsP 3 3-kinase. In the soma of Purkinje cells, the reticularlike pattern is reminiscent of both InsP 3 and ryanodine receptors that have been localized to the endoplasmic reticulum using electron microscopy (48).
Perhaps inhibition of the specific activity of particulate type I InsP 3 5-phosphatase (compared with the soluble enzyme) may be regarded as a mechanism of regulation of InsP 3 5-phosphatase activity. In this context, the myristoylated alanine-rich protein kinase C substrate and Ki-Ras interact with the cytoplasmic surface of the plasma membrane by both hydrophobic and electrostatic interactions (49). Phosphorylation of the myristoylated alanine-rich protein kinase C substrate reduces the electrostatic interaction, and the protein is released to the cytosol. It is not yet known whether membrane binding of Ki-Ras, and perhaps InsP 3 5-phosphatase, is regulated.