Novel Inositol Polyphosphate 5-Phosphatase Localizes at Membrane Ruffles*

We have cloned a novel inositol polyphosphate 5-phosphatase from the rat brain cDNA library. It contains two highly conserved 5-phosphatase motifs, both of which are essential for its enzymatic activity. Interestingly, the proline content of this protein is high and concentrated in its N- and C-terminal regions. One putative SH3-binding motif and six 14–3-3 ζ-binding motifs were found in the amino acid sequence. This enzyme hydrolyzed phosphate at the D-5 position of inositol 1,4,5-trisphosphate, inositol 1,3,4,5-tetrakisphosphate, and phosphatidylinositol 4,5-bisphosphate, consistent with the substrate specificity of type II 5-phosphatase, OCRL, synaptojanin and synaptojanin 2, already characterized 5-phosphatases. When the Myc-epitope-tagged enzyme was expressed in COS-7 cells and stained with anti-Myc polyclonal antibody, a signal was observed at ruffling membranes and in the cytoplasm. We prepared several deletion mutants and demonstrated that the 123 N-terminal amino acids (311–433) and a C-terminal proline-rich region containing 277 amino acids (725–1001) were essential for its localization to ruffling membranes. This enzyme might regulate the level of inositol and phosphatidylinositol polyphosphates at membrane ruffles.

Here, we report the cloning and characterization of a novel 5-phosphatase. This enzyme dephosphorylates the D-5 position of inositol and phosphatidylinositol polyphosphates at ruffling membranes.

EXPERIMENTAL PROCEDURES
Molecular Cloning of the Rat Novel Inositol Polyphosphate 5-Phosphatase-A partial human cDNA clone (GenBank TM accession number H14886) was obtained from Genome System, Inc. (St. Louis, MO). H14886 is a 468-base pair human cDNA fragment, and its deduced amino acid sequence contains a 5-phosphatase catalytic motif2-like sequence. The fragments were labeled by random hexamer priming and used to screen a ZAPII rat brain cDNA library (Stratagene, La Jolla, CA). Positive clones were subcloned into pBluescript SK(Ϫ) by an in vivo excision method and sequenced. Complete sequence data was obtained from both strands using a conventional dideoxy-termination method.
Northern Blot Analysis-Membranes containing mRNA (2 g of poly(A) RNA was contained in each lane) were purchased from OriGene Technology, Inc. (Rockville, MD). Total RNA was isolated from cultured cells using MagExtractor-RNA-and MagExtractor System (TOYOBO Co., Ltd., Osaka, Japan), and approximately 10 g of the RNA was * This work was supported in part by a grant-in-aid for Research for the Future Program from the Japan Society for the Promotion of Science. 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.
blotted on a nylon membrane. The sequence of rat novel 5-phosphatase specific region, nucleotides 2362-3000, was amplified by polymerase chain reaction and used as an [␣-32 P]dCTP-labeled 0.7-kb cDNA probe. Hybridization was performed following the protocol of OriGene Technology, Inc., and the probe was hybridized at 42°C.
Cell Culture and Expression of Recombinant Proteins in COS-7 Cells-COS-7 cells were cultured in Dulbecco's modified Eagle's medium (Nissui, Tokyo, Japan) containing 10% fetal calf serum and 60 g/ml kanamycin, and kept at 37°C in a humidified atmosphere of 95% air and 5% CO 2 . Rat novel 5-phosphatase cDNA, its deletion mutants and partial human SHIP 1 cDNA (nucleotides 1461-4079 of human SHIP 1 cDNA; GenBank TM accession number U57650) were subcloned into the eukaryotic pCMV6-Myc expression vector. Constitutively active Rac1, Rac1G12V construct, was ligated into the eukaryotic pEF-BOS-FLAG expression vector. All constructs were transfected into COS-7 cells by a conventional electroporation method, and cells were harvested or fixed after 48 h (22). The expression and the size of expressed proteins were checked by immunoblotting.

Dephosphorylation of Novel Inositol Polyphosphate 5-Phosphatase
Protein-Recombinant novel 5-phosphatase expressing cells (1 ϫ 10 6 cells/60-mm tissue culture dish (Falcon)) were washed once with PBS(Ϫ) and harvested with 200 l of the reaction buffer for alkaline phosphatase (50 mM Tris-HCl, pH 7.5, 1 mM MgCl 2 , 10 g/ml leupeptin, and 10 g/ml aprotinin). After a brief sonication, the cell lysate was incubated for 10 min at 30°C. Then 100 l of the lysate with or without 30 units of calf intestine alkaline phosphatase (Takara Shuzo Co., Ltd., Biomedical Group, Shiga, Japan) was further incubated for 15 min at the same temperature. The mixture was analyzed by immunoblotting. Enzyme Assay-Cells (7 ϫ 10 6 cells/150-mm tissue culture dish (Falcon)) transfected with Myc-epitope-tagged novel 5-phosphatase construct or empty vector were cultured for 2 days and harvested with 1 ml of cold lysis buffer (40 mM Tris-HCl, pH 7.6, 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 10 g/ml leupeptin, and 10 g/ml aprotinin). The cells were briefly sonicated and centrifuged at 10,000 ϫ g for 20 min at 4°C. Supernatant was collected and rotated with 25 l of anti-Myc monoclonal antibody (Santa Cruz Biotechnology, Inc.) for 1 h at 4°C. Then protein A beads (Pierce) (50 l) were added, and the solution was rotated for 1 h. After that, the beads were washed with lysis buffer five times and finally suspended in buffer for inositol polyphosphates or PtdIns(4,5)P 2 phosphatase assay.
Assay of inositol polyphosphate 5-phosphatase activity was carried out as described by Connolly et al. (36) (37). The flow rate was 0.6 ml/min, and each fraction was collected for 30 s. Collected samples were diluted to 1/20 volume with distilled water and quantitated by liquid scintillation counting. For the assay of Ins(1,3,4)P 3 hydrolyzing activity, the D-5 position phosphate of Ins(1,3,4,5)P 4 was hydrolyzed by recombinant SHIP 1 protein and used as a substrate. The SHIP 1 protein was prepared the same way as the novel 5-phosphatase.
The PtdIns(4,5)P 2 phosphatase activity was determined (38). The spots on the thin layer chromatography plate were visualized by exposing the plate to x-ray film (Eastman Kodak Scientific Co., Rochester, NY) for 6 days at Ϫ80°C. Before the exposure, the intensity of the radioactivity was enhanced by EN 3 HANCE spray (NEN Life Science Products).
Cell Staining-Transfected cells cultured on glass coverslips were fixed with 10% formaldehyde in PBS(Ϫ) for 10 min at room temperature and washed three times with PBS(Ϫ). Then they were incubated for 5 min in 0.2% Triton X-100 in PBS(Ϫ) at room temperature and washed with PBS(Ϫ) five times. The cells were further incubated with anti-Myc polyclonal antibody (Santa Cruz Biotechnology, Inc.) for 1 h. After five washes with PBS(Ϫ), the cells were incubated with fluorescein isothiocyanate-conjugated anti-rabbit IgG antibody (ICN Pharmaceuticals, Inc., Costa Mesa, CA) and rhodamine-phalloidin (Molecular Probes, Eugene, Oregon) for 30 min. When active Rac was co-transfected, anti-FLAG monoclonal antibody (Eastman Kodak Scientific Co.) and anti-Myc polyclonal antibody were used as primary antibodies and rhodamine-conjugated anti-mouse IgG antibody (ICN Pharmaceuticals, Inc.) was substituted with rhodamine-phalloidin. The cells were mounted in glycerin-PBS(Ϫ) and observed with a fluorescence microscope.

RESULTS AND DISCUSSION
cDNA Cloning of the Novel Inositol Polyphosphate 5-Phosphatase-The predicted amino acid sequence of human EST clone H14886 is very similar to OCRL and synaptojanin, and contains the 5-phosphatase catalytic motif2-like sequence (2,6). Therefore, we regarded H14886 as a partial cDNA fragment of a novel 5-phosphatase. Using H14886 as a probe, we obtained several cDNA clones from a rat brain cDNA library and isolated a 3,322-base pair rat putative 5-phosphatase cDNA (Fig. 1). Our rat cDNA clone contained the polyadenylation signal AATTAAA (nucleotides 3304 -3310) (Fig. 1). The same polyadenylation signal and poly(A) sequence are found in another 1496-base pair human EST clone, U45975. There was no in-frame upstream stop codon in our cDNA clone. However, we concluded that the ATG underlined in Fig. 1 is the initiation codon for rat putative 5-phosphatase mRNA because 1) the sequence around the first putative ATG codon agrees with Kozak's consensus rule (GCAGACATGG versus GCC(G/A)C-CATGG) (39), and 2) Northern blot analysis indicated that the rat putative 5-phosphatase mRNA is 3.4-kb long, which is nearly identical to the length of our cDNA clone (Fig. 4). Searches of the current GenBank TM data base with BLASTN algorithm revealed that part of synaptojanin and synaptojanin 2 are homologous to the 1900 -2040-base pair region of the rat putative 5-phosphatase cDNA sequence (60 and 69%, respectively), which includes the 5-phosphatase catalytic motif2.
Our results suggest that recombinant PIPP is highly phosphorylated in the cytoplasm. But Western blot analysis revealed that anti-phosphotyrosine antibody (PY20) did not stain phosphorylated PIPP (data not shown). Therefore, PIPP is probably phosphorylated at serine/threonine residues, which might regulate the activity.
Northern Blot Analysis-The results of Northern blot analysis are shown in Fig. 4. PIPP was expressed in brain, heart, kidney, stomach, small intestine, and lung. The size of the mRNA was estimated at 3.4 kb in those tissues. In spleen, thymus, skeletal muscle, testis, and skin, no signal was observed. All cultured cells examined expressed PIPP, but Jurkat and HL-60 cells showed an especially high PIPP mRNA content. Only liver showed the 2.4-kb band. This signal might indicate the existence of a splicing isoform of the PIPP or another undiscovered 5-phosphatase in liver.
Overexpression of PIPP and Its Deletion Mutants in COS-7 Cells-Myc-epitope-tagged PIPP was expressed in COS-7 cells, and its cellular localization was revealed by anti-Myc poly-clonal antibody. As indicated in Fig. 7A, some PIPP was clearly condensed at the cell periphery and some dispersed in the cytoplasm. The same cells were stained with rhodamine-phalloidin to visualize the actin cytoskeleton (Fig. 7B). Some PIPP was localized in cortical areas with the actin filaments locating at ruffling membranes. However, alterations in the actin cytoskeleton were not observed, in contrast to reports on overexpression of OCRL and synaptojanin in cells (22,47).
We prepared several deletion mutants of PIPP to determine the region essential for its localization to ruffling membranes and co-expressed it with constitutively active Rac1 mutant in COS-7 cells (Fig. 8, panel I). Constitutively active Rac1 mutant co-localizes with actin filaments at ruffling membranes and accelerates membrane ruffling (48,49). Apparently, PIPP colocalized with the active Rac1 mutant (Fig. 8, panel II, A and B) confirming that PIPP localizes at ruffling membranes. Mut2 and Mut3 exhibited the same localization pattern as the fulllength of PIPP (data not shown). Truncated PIPP without the N-terminal proline-rich region (Mut4) located to membrane ruffles at a much lower intensity than wild type or mutants with partial proline-rich regions (Mut2 and Mut3, Fig. 8, panel II, C and D). When the C-terminal proline-rich region containing 277 amino acids (725-1001) was deleted (Mut6), the localization of the enzyme to the ruffling membranes was abolished (Fig. 8, panel II, E and F). Mut5 did not contain either the Nor C-terminal proline-rich regions and did not localize to ruffling membranes (Fig. 8, panel II, G and H). These results indicate that the C-terminal proline-rich region from residues 725 to 1001 is essential for the localization of PIPP to ruffling membranes and a part of the N-terminal proline-rich region from residues 311 to 433 contributes to the localization. PIPP contains one putative SH3-binding motif and five 14 -3-3 -binding motifs in the proline-rich regions that are involved in its cellular distribution. The identification of PIPP-binding proteins should deepen the understanding of the function of PIPP.
Mut1 did not contain the two catalytic motifs of 5-phosphatase localized to ruffling membranes (Fig. 8, panel II, I and J). Membrane ruffling, induced by constitutively active Rac1 mutant, was not affected by the co-expression of any deletion mutants of PIPP (Fig. 8, panel II, B, D, F, H, and J). These results indicate that PIPP does not participate in the re-organization of the actin cytoskeleton but may be involved in modulation of the function of inositol and phosphatidylinositol polyphosphate-binding proteins that are present at membrane ruffles.
In summary, we have cloned a novel 5-phosphatase from a rat brain cDNA library. It is a proline-rich protein and hydro-lyzed the D-5 position of phosphate in Ins(1,4,5)P 3 , Ins(1,3,4,5)P 4 and PtdIns(4,5)P 2 . Therefore, we designated this novel 5-phosphatase as PIPP. PIPP is localized at membrane ruffles and may be involved in the modulation of the function of proteins that are present at membrane ruffles. FIG. 6. Lipid phosphatase activity of PIPP. [ 3 H]PtdIns(4,5)P 2 was incubated with PIPP or mock, and the reaction product was analyzed by thin layer chromatography. In the PIPP lane, PtdIns(4,5)P 2 hydrolysis was observed, but no hydrolysis was observed when mock was incubated with PtdIns(4,5)P 2 .