A Novel Phosphatidylinositol-5-phosphate 4-Kinase (Phosphatidylinositol-phosphate Kinase IIγ) Is Phosphorylated in the Endoplasmic Reticulum in Response to Mitogenic Signals*

Here, we identify a novel rat phosphatidylinositol-5-phosphate 4-kinase, phosphatidylinositol-phosphate kinase IIγ (PIPKIIγ). PIPKIIγ comprises 420 amino acids with a molecular mass of 47,048 Da, showing greater homology to the type IIα and IIβ isoforms (61.1 and 63.7% amino acid identities, respectively) of phosphatidylinositol-phosphate kinase than to the type I isoforms. It is predominantly expressed in kidney, with low expression in almost all other tissues. PIPKIIγ was found to have phosphatidylinositol-5-phosphate 4-kinase activity as demonstrated in other type II kinases such as PIPKIIα. The PIPKIIγ that is present endogenously in rat fibroblasts, PC12 cells, and rat whole brain lysate or that is exogenously overexpressed in COS-7 cells shows a doublet migrating pattern on SDS-polyacrylamide gel electrophoresis. Alkaline phosphatase treatment and metabolic labeling in [32P]orthophosphate experiments revealed that PIPKIIγ is phosphorylated in vivo, resulting in a shift in its electrophoretic mobility. Phosphorylation is induced by treatment of mitogens such as serum and epidermal growth factor. Immunostaining experiments and subcellular fractionation revealed that PIPKIIγ localizes dominantly in the endoplasmic reticulum (ER). Phosphorylation also occurs in the ER. Thus, PIPKIIγ may have an important role in the synthesis of phosphatidylinositol bisphosphate in the ER.

Phosphatidylinositol 4,5-bisphosphate (PI(4,5)P 2 ) 1 is a phospholipid with a variety of functions in vivo including not only the production of second messengers such as diacylglycerol and inositol 1,4,5-trisphosphate, but also the regulation of actin regulatory proteins and the activation of phospholipase D and ADP-ribosylation factor. It has also been reported that PI(4,5)P 2 synthesis is potentiated by various stimuli including GTP␥S (1-3), phorbol esters (4), tyrosine kinases (5), and integrins (6). The variations in its function and the regulation of its synthesis indicate that enzymes responsible for the production of PIP 2 , such as PI kinase and PIPK, also show large diversities. Among PIPKs, two major subtypes (types I and II), each comprising two isoforms (I␣, I␤, II␣, and II␤), have been identified to date (13, 16 -18), and it is thought that the role for each subtype in vivo is different. The type I isozyme has been reported to be activated by phosphatidic acid (7), to bind physically to the small GTPases Rho (8) and Rac (9), and to be involved in Ca 2ϩ -dependent exocytosis in PC12 cells (10). Human PIPKI␤ has been shown to be identical to the STM7 gene, the putative gene responsible for Friedreich's ataxia, suggesting that this isozyme plays roles in vesicular trafficking such as neurotransmitter release (11). On the other hand, type II isozymes have also been reported to have several functions in vivo. In platelets, PIPKII␣ was shown to translocate to the cytoskeletal fraction after stimulation by thrombin (12). PIP-KII␤ was identified by its specific interaction with a cytoplasmic region of the p55 tumor necrosis factor-␣ receptor, and a role for PIPK in tumor necrosis factor-␣ signaling has been suggested (13).
Here, we identify a novel PIPKII isozyme (PIPKII␥) by a reverse transcription-PCR method using degenerate primers designed from highly conserved primary sequences in PIPK family members. PIPKII␥ is phosphorylated on serine residues in vivo, resulting in a mobility shift on SDS-polyacrylamide gel electrophoresis. Mitogenic stimulation, such as by serum, EGF, or PDGF treatment, results in phosphorylation of PIPKII␥. The results of immunofluorescence experiments and subcellular fractionation suggest that PIPKII␥ has important roles in the production of PIP 2 in the ER.

EXPERIMENTAL PROCEDURES
Materials-PIPs were purified by neomycin column chromatography from crude phospholipids extracted from bovine spinal cord as described (14). [␣-32 P]dCTP, [␥-32 P]ATP, [ 32 P]orthophosphate, and [ 3 H]PI(4,5)P 2 were from NEN Life Science Products. The Colony/ PlaqueScreen used to screen the cDNA library was from NEN Life Science Products. The polyvinylidene difluoride membranes used for Western blot analysis were from Nihon Eido (Tokyo, Japan). Ni 2ϩnitrilotriacetic acid-agarose was from QIAGEN Inc. (Chatsworth, CA). The Partisphere SAX column was from Whatman International Ltd. (Maidstone, United Kingdom). The thin-layer chromatography silica plates and the cellulose plate used to separate phospholipids and phosphoamino acids, respectively, were from Merck (Darmstadt, Germany). Monoclonal anti-Myc antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal anti-BiP antibody was from Stressgen Biotech Corp. Monoclonal anti-␤-tubulin antibody was from Chemicon International, Inc. (Temecula, CA). Rhodamine-and fluorescein-conjugated anti-rabbit IgG antibodies and fluorescein-conjugated anti-mouse IgG antibody were from Organon Teknika Corp. (West Chester, PA). Rhodamine-conjugated wheat germ agglutinin was from Molecular Probes, Inc. (Eugene, OR). * 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 nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF030558 and AF033355.
Cell Culture-COS-7 and 3Y1 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. PC12 cells were grown in Dulbecco's modified Eagle's medium containing 10% horse serum and 5% fetal bovine serum.
Reverse Transcription-Polymerase Chain Reaction-Total RNA isolated from rat brain was reverse-transcribed into cDNA by murine leukemia virus reverse transcriptase and used as a template for PCR using degenerate primers (5Ј-GAITAYTGYCCIRWIGTITTYMG-3Ј, 5Ј-ATICYIABIAIIARRCTRTARTCCAT-3Ј, and 5Ј-ATICYIABIAIIARI-GARTARTCCAT-3Ј) corresponding to two highly conserved sequences in mammalian and yeast PIPKs ((D/E)YCPXVFR and MDYSLLLG(I/ M)). The polymerase chain reaction was carried out as follows: 95°C for 1 min, 43°C for 1 min, and 72°C for 2 min, for 40 cycles. The PCR product, ϳ500 base pairs long, was subcloned into the SmaI site of the pBluescript SK(Ϫ) vector and sequenced.
cDNA Cloning of PIPKII␥-The PCR product encoding a novel sequence was cut out from the vector with EcoRI and BamHI, labeled with [␣-32 P]dCTP, and used as a probe for screening a rat brain cDNA library. The longest clone obtained (ϳ2.4 kilobases) encoded an open reading frame as long as ϳ400 amino acids, but did not include a potent start codon. On the other hand, another partial clone was obtained that included a potent start codon preceded by a sequence consistent with a Kozak consensus sequence (15), but did not include a stop codon. From the sequences of these two clones, we could determine the complete sequence for this novel PIPK.
Northern Blot Analysis-A partial fragment corresponding to 418 -1500 base pairs of cDNA was labeled and used as a probe for Northern blot analysis. Hybridization was carried out on mouse multiple tissue Northern blot membrane (MTN TM , CLONTECH).
Production of Polyclonal Antibody-A partial fragment encoding amino acids 130 -420 was ligated into the PstI-HindIII site of a pQE32 His tag expression vector (QIAGEN Inc.). The His-tagged protein was expressed in Escherichia coli and purified on Ni 2ϩ -nitrilotriacetic acidagarose as described by the manufacturer. The purified protein was injected as an antigen into rabbits to raise polyclonal antiserum. The resulting antibody was affinity-purified with the antigen protein transferred onto a polyvinylidene difluoride membrane or immobilized on a Hi-Trap NHS-activated column (Amersham Pharmacia Biotech).
Transfection into COS-7 Cells-The full-length cDNAs of mouse PIPKI␤ and rat PIPKII␤ and PIPKII␥ were ligated into the SalI-BamHI site of pCMV-Myc or the XhoI-BamHI site of pSR␣XEBNeo mammalian expression vectors. Twenty micrograms of each plasmid was mixed with 1 ϫ 10 7 cells, and the mixtures were subjected to electroporation with a Gene Pulser (Bio-Rad). The cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum.
Measurement of PIPK Activity-Forty-eight hours after electroporation, the expression vector-transfected COS-7 cells were lysed with lysis buffer (20 mM Hepes, pH 7.2, 50 mM NaCl, 30 mM sodium pyrophosphate, 1% Nonidet P-40, 1 mM EGTA, 25 mM NaF, 0.1 mM sodium vanadate, and 1 mM phenylmethylsulfonyl fluoride). The expressed enzyme was immunoprecipitated with monoclonal anti-Myc antibody and washed three times with lysis buffer and once with reaction buffer (50 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , and 1 mM EGTA). The reaction was started by adding 50 M PIP, 50 M ATP, and 10 Ci of [␥-32 P]ATP in 50 l. After incubating for 10 min at room temperature, the lipids were extracted with 1 N HCl and chloroform/methanol (2:1, by volume) and spotted on TLC plates. The plates were developed in chloroform/ methanol/ammonia/water (14:20:3:5, by volume), and the products were observed by autoradiography or quantified by a Fuji BAS2000 image analyzer.
Analysis of Phosphoinositides by SAX HPLC-Phosphoinositides separated by TLC were scraped out, deacylated, and analyzed by SAX HPLC as described (19).
Dephosphorylation of Phosphoinositides by SHIP-A partial fragment corresponding to 1084 -3947 base pairs of the cDNA of human Src homology domain-containing inositol-polyphosphate phosphatase (SHIP) was cut out with SalI and BamHI and ligated into the SalI-BamHI site of pCMV-Myc. The resulting expression vector was transfected into COS-7 cells as described above. Myc-SHIP was immunoprecipitated, and the dephosphorylation of lipids was carried out in 50 mM Tris-HCl, pH 7.5, and 10 mM MgCl 2 at 37°C for 60 min. The lipids were extracted and separated by TLC (chloroform/methanol/acetic acid/water, 43:38:5:7, by volume).
Metabolic 32 P Labeling of PC12 Cells and Phosphoamino Acid Analysis-The culture medium was changed to phosphate-free Dulbecco's modified Eagle's medium, and the PC12 cells were cultured for 30 min.
[ 32 P]Orthophosphate (0.2 mCi/ml) was then added, and the cells were incubated for 24 h. Labeled cells were lysed in lysis buffer, and PIPKII␥ was immunoprecipitated with anti-PIPKII␥ antibody and transferred to a polyvinylidene difluoride membrane. The band corresponding to PIPKII␥ was cut out and hydrolyzed in 6 N HCl for 1 h at 110°C. The resulting amino acids, together with standard phosphoamino acids, were spotted on TLC plates and separated by electrophoresis in pH 3.5 buffer (5% acetic acid and 0.5% pyridine). The labeled phosphoamino acids were detected by autoradiography. The positions of the standard phosphoamino acids were detected by ninhydrin staining.
Immunofluorescence of PIPKII␥-Cells growing on glass coverslips were fixed with 3.7% formaldehyde in phosphate-buffered saline for 15 min and permeabilized with 0.2% Triton X-100 in phosphate-buffered saline for 5 min. Incubation with the first antibody (polyclonal anti-PIPKII␥ and monoclonal anti-BiP) was carried out for 1 h, and incubation with the second antibody or rhodamine-conjugated wheat germ agglutinin for 30 min. The cells were observed with a confocal fluorescence microscope (Bio-Rad).
Subcellular Fractionation-The subcellular fractionation was performed as described (20) with some modifications. Rat liver or 3Y1 fibroblasts were homogenized in 0.25 M sucrose, 50 mM triethanolamine HCl, pH 7.5, 50 mM potassium acetate, 6 mM magnesium acetate, 1 mM EDTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, and 10 g/ml aprotinin. After centrifugation at 800 ϫ g for 10 min and 10,000 ϫ g for 10 min to devoid nuclei and mitochondria, respectively, the "post-mitochondrial" supernatant was obtained. The supernatant was layered over a cushion of 1.3 M sucrose in the same buffer and centrifuged at 202,000 ϫ g for 2.5 h to yield three distinct fractions: the "post-microsomal" supernatant (representing the cytosol), interfacial "smooth microsomes" (representing the smooth ER and the Golgi apparatus), and the "rough microsomal" pellet (representing the rough ER).

RESULTS
Identification of a Novel PIPK-To identify novel PIPKs, we applied a reverse transcription-PCR method using degenerate primers corresponding to amino acid sequences highly conserved among mammalian PIPKs and their putative yeast homologs, Mss4p and Fab1p. Total RNA isolated from rat brain was reverse-transcribed and used as a template for further PCR. The PCR product was subcloned into the pBluescript vector and sequenced. Among several sequences corresponding to known PIPKs, one novel sequence homologous to the type II isoform of PIPK was obtained. We then tried to isolate a fulllength cDNA using this fragment as a probe. By screening a rat brain cDNA library including random-primed clones, we obtained two clones encoding overlapping sequences. Between these clones, we found an open reading frame of 1260 base pairs encoding 420 amino acids (Fig. 1A). The calculated molecular mass is 47,048 Da.
The Novel PIPK Belongs to the Type II Subfamily-The whole amino acid sequence of the novel PIPK was revealed to be homologous to the type II␣ and II␤ isoforms of PIPK rather than to the type I isoforms (61.1 and 63.7% identities to types II␣ and II␤, respectively, for the entire amino acid sequence, compared with 33.0% identity to the type I isoforms for the kinase domain), indicating that this PIPK is a third member of the type II isoform subgroup (Fig. 1B). Thus, we propose that this novel PIPK be called type II␥ (PIPKII␥). Alignment of the amino acid sequences between type II PIPKs revealed that PIPKII␥ has a highly conserved kinase homology domain separated by an insert domain showing no similarity to other PIPK family members (18).
Tissue Distribution of PIPKII␥-To study the tissue distribution of PIPKII␥, Northern hybridization was carried out on mRNA from various mouse tissues. An mRNA of ϳ3.5 kilobases was detected in almost all tissues, with the most abundant expression in kidney (Fig. 2). The pattern of distribution is different from that of any type I isoform or any other type II isoform (13, 16 -18), suggesting specific functions for this isoform.
PIPKII␥ Is a Phosphatidylinositol-5-phosphate 4-Kinase-We transfected a Myc-tagged version of the full-length cDNA of PIPKII␥ (Myc-PIPKII␥) into COS-7 cells. The protein expressed in the whole cell lysate and the anti-Myc immunoprecipitate was detected in a doublet form by Western blotting with anti-Myc antibody (Fig. 3A and discussed further below). Myc-PIPKII␥ was immunoprecipitated with anti-Myc antibody, and the PIPK activity was measured. The immunoprecipitate phosphorylated PIP purified from bovine spinal cord (see "Experimental Procedures"), whereas anti-Myc immunoprecipitates from cells transfected with vector alone failed to do so (Fig. 3B). By using SAX HPLC, the resulting PIP 2 was confirmed to be PI(4,5)P 2 (Fig. 3C).
PIPKII␥ Is a Phosphoprotein-A polyclonal antibody was produced with a partial His-tagged protein expressed in E. coli as an antigen. With this polyclonal antibody, endogenous PIP-KII␥ was detected as doublet bands at 47 kDa in lysates from rat brain, PC12 cells, and 3Y1 fibroblasts by Western blotting (Fig. 4A). When the full-length cDNA (without the Myc tag) was transfected into COS-7 cells, the same doublet band was detected (Fig. 4A), suggesting that this doublet corresponds to some modification of PIPKII␥ such as proteolysis or phosphorylation and is not due to cross-reactivity of the antibody to another protein. To examine the possibility that these doublet bands correspond to phosphorylated PIPKII␥, we treated the Myc-tagged version of PIPKII␥ with alkaline phosphatase. The Myc-PIPKII␥ that was immunoprecipitated from overexpressing COS-7 cells showed a doublet banding pattern with the upper band predominant. When the immunoprecipitated Myc-PIPKII␥ was incubated with calf intestine alkaline phosphatase, the upper band disappeared completely (Fig. 4B), whereas the lower band increased in intensity. This indicates that the doublet migrating pattern is due to the phosphorylation of PIPKII␥. To confirm this conclusion, we next labeled PC12 cells metabolically with [ 32 P]orthophosphate. After labeling, the ells were lysed, and PIPKII␥ was immunoprecipitated, showing that PIPKII␥ was phosphorylated (Fig. 4C). Together with the results of Western blotting, it was confirmed that this phos-phorylated protein corresponds to the upper band of PIPKII␥ (Fig. 4C). Next, the phosphorylated band was cut out from membrane, and phosphoamino acid analysis was carried out. The results show that PIPKII␥ phosphorylation occurs predominantly on serine residues (Fig. 4D). To determine whether the enzymatic activity of PIPKII␥ is affected by its phosphorylation, we measured the activity of Myc-PIPKII␥ after alkaline phosphatase treatment. Myc-PIPKII␥ retained considerable activity even after alkaline phosphatase treatment (Fig. 4E), indicating that the phosphorylation of PIPKII␥ does not affect its enzymatic activity.
PIPKII␥ Is Phosphorylated in Response to Extracellular Stimuli-In response to extracellular stimuli such as growth factors or hormones, intracellular protein kinases are activated and phosphorylate their physiological substrates. Since PIP-KII␥ was found to be phosphorylated on serine residues in vivo, we examined whether the level of PIPKII␥ phosphorylation is potentiated by extracellular stimuli. First, we treated rat 3Y1 fibroblasts with 10% serum for various periods. The upper band of PIPKII␥ increased in a time-dependent manner, suggesting that PIPKII␥ is phosphorylated in response to serum (Fig. 5A). We then examined other extracellular stimuli including EGF, PDGF, bradykinin, and lysophosphatidic acid for their abilities to induce the phosphorylation of PIPKII␥. Among them, EGF and PDGF enhanced the phosphorylation as well as serum (Fig. 5, B and C). Lysophosphatidic acid and bradykinin also induced phosphorylation to a lesser extent. Fig. 4D clearly shows that PIPKII␥ phosphorylation does not take place on tyrosine residues. Moreover, PIPKII␥ was not recognized by an anti-phosphotyrosine antibody, PY20 (data not shown). Therefore, it seems likely that the phosphorylation is mediated by a serine/threonine kinase downstream of mitogenic signals mediated by receptor tyrosine kinases. Protein kinase C does not seem to be involved since phorbol 12-myristate 13-acetate did not potentiate phosphorylation (Fig. 5, B and C). In addition, a specific protein kinase C inhibitor, H-7, did not suppress phosphorylation in 3Y1 cells (data not shown).
Intracellular Localization of PIPKII␥-Using a polyclonal antibody, we next examined the intracellular localization of PIPKII␥. The polyclonal antibody used was confirmed to recognize specifically the doublet band corresponding to PIPKII␥ in 3Y1 cell lysates by Western blotting (Fig. 4A). When rat 3Y1 fibroblasts were stained, PIPKII␥ was seen to predominate in the perinuclear regions, suggesting that it is localized in microsomal organelles such as the ER. To confirm this possibility, we doublestained rat 3Y1 fibroblasts with anti-PIPKII␥ antibody and with anti-BiP antibody, an ER-retaining protein. Both staining patterns (Fig. 6A) clearly indicate the localization of this enzyme in the ER. This staining pattern does not overlap with that of wheat germ agglutinin, a trans-Golgi staining reagent (Fig. 6A).
The intracellular localization of PIPKII␥ was studied further by a subcellular fractionation method using a 1.3 M sucrose cushion (see "Experimental Procedures"). PIPKII␥ was detected in smooth and rough microsomal fractions of rat liver the same as the ER marker BiP, whereas a cytosolic protein, ␤-tubulin, was detected only in the top of the gradient (Fig. 6B). The phosphorylated form of PIPKII␥ was also detected in these fractions, suggesting that phosphorylation of PIPKII␥ occurred in microsomes. To confirm this possibility, post-mitochondrial supernatants of 3Y1 fibroblasts were subjected to the same subcellular fractionation after mitogenic stimulations. PIPKII␥, together with BiP, was detected predominantly in the smooth microsomal fraction (Fig. 6C). On stimulation by EGF, the phosphorylated form increased in this fraction, indicating that the phosphorylation of PIP-KII␥ occurred within the ER (Fig. 6C). In addition, immuno- fluorescence staining also showed that the localization of PIPKII␥ in the ER was not affected by stimulation of the cells with serum or EGF (data not shown). These results indicate that PIPKII␥ is phosphorylated in the ER in response to mitogenic signals, thus suggesting that it has important roles in the synthesis of PIP 2 in the ER. DISCUSSION Purification and cDNA cloning of the 53-kDa PIPKII␣ from erythrocytes revealed that the lipid kinase belongs to a distinct kinase family different from those of PI 3-and PI 4-kinases and protein kinases (16). This family also seems to include yeast homologs such as Mss4p and Fab1p. Furthermore, cDNA cloning of types I␣ and I␤, members of another subtype of mammalian PIPK, also showed them to belong to this same distinct lipid kinase family (17,18). Members of this novel lipid kinase family have several conserved regions within their primary sequences. Using a reverse transcription-PCR method involving degenerate primers corresponding to these highly conserved sequences, we succeeded in identifying a novel PIPK isoform and named it PIPKII␥.
Although PIPKII␥ seems to belong to the type II subtype, the similarity between PIPKII␥ and other members of the type II PIPK family is not very high (61.1% for II␣ and 63.3% for II␤) compared with the homology between PIPKII␣ and PIPKII␤ (76.7%). This, together with the difference in its expression pattern from that of other PIPKs, suggests that PIPKII␥ has some distinct functions in vivo.
PIPKII␥ was detected as a doublet migrating protein by Western blotting with a specific polyclonal antibody not only in rat brain lysates, but also in 3Y1 fibroblasts and PC12 cells. The same doublet patterns were also observed when PIPKII␥ was overexpressed in COS-7 cells. The evidence presented in this study shows that PIPKII␥ is phosphorylated in vivo and that the upper band represents the phosphorylated form. Furthermore, phosphoamino acid analysis revealed that phosphorylation occurs predominantly on serine residues. We also observed that mitogens such as serum and growth factors immediately induced phosphorylation of PIPKII␥. The total cellular amount of PIP 2 and the PIPK activity have been reported to increase in response to various extracellular stimuli, including EGF (5), formyl-methionyl-leucyl-phenylalanine, platelet-activating factor (1), thrombin (22), phorbol ester (4), FIG. 4. PIPKII␥ is a phosphoprotein. A, Western blot analysis of lysates from rat brain, PC12 cells, 3Y1 cells, and PIPKII␥-overexpressing COS-7 cells. The PIPKII␥ overexpressed in COS-7 cells used here was not the tagged form. B, alkaline phosphatase treatment of immunoprecipitated Myc-tagged PIPKII␥. Immunoprecipitated Myc-PIPKII␥ was incubated with 2 units of calf intestine alkaline phosphatase (CIAP) at 30°C for 60 min. The resulting products were detected by Western blotting with anti-Myc antibody. C, 32 P labeling experiment. PC12 cells were metabolically labeled with [ 32 P]orthophosphate. PIP-KII␥ was immunoprecipitated (IP) from the labeled cell lysates and detected by autoradiography or Western blotting with anti-PIPKII␥ antibody. h. c., heavy chain; IB, immunoblotting. D, phosphoamino acid analysis of PIPKII␥. The immunoprecipitated PIPKII␥ used in B was hydrolyzed with 6 N HCl for 1 h at 110°C. The resulting amino acids were separated by TLC electrophoresis and detected by autoradiography. The positions of standard phosphoamino acids and free orthophosphate are also indicated. E, PIPK activity of Myc-PIPKII␥ after calf intestine alkaline phosphatase treatment. Immunoprecipitates treated with calf intestine alkaline phosphatase as described for B were washed with reaction buffer for PIPK, and the reaction was carried out. Results are representative of three independent experiments.

FIG. 5. Phosphorylation of PIPKII␥ induced by mitogenic stimulation of 3Y1 fibroblasts.
A, PIPKII␥ is phosphorylated in response to stimulation by serum. Rat 3Y1 fibroblasts were serumstarved for 48 h before treatment with 10% fetal bovine serum for the indicated times. Cells were lysed with SDS sample buffer (125 mM Tris-HCl, pH 6.2, 2% SDS, and 0.2% 2-mercaptoethanol) and immunoblotted with anti-PIPKII␥ antibody. B, Rat 3Y1 fibroblasts were serumstarved (control) and then stimulated with 10% fetal bovine serum (serum), 2 M lysophosphatidic acid (LPA), 1 M bradykinin, 100 ng/ml EGF, 20 units/ml PDGF, and 1 M phorbol 12-myristate 13-acetate (PMA) for 10 min, respectively. The phosphorylation level of PIPKII␥ was evaluated by Western blotting as described for A. C, quantitative representation of B. Each band corresponding to the phosphorylated or unphosphorylated form was quantified by densitometry. The ratio of the phosphorylated form to the unphosphorylated form was calculated. Results are representative of three independent experiments. and adhesion to fibronectin (6). Some of these extracellular stimuli have been reported to increase PIPK activity, especially in the cytoskeleton. In addition, the involvement of G-proteins, including small GTPases such as Rac and Rho, has also been suggested by data showing that the PIPK activity is potentiated by non-hydrolyzable GTP or is associated with recombinant Rho and Rac proteins. Despite the above observations, the exact molecular mechanism by which PIPK is regulated has not been made clear. Here, we provide evidence for the phosphorylation of PIPKII␥. It is possible that PIPK is regulated by a protein kinase downstream of extracellular stimuli. At present, we do not know what kinase is responsible for the phosphorylation. The phosphorylation was found to be enhanced by tyrosine kinase activators such as EGF and PDGF rather than activators related to triplet G-protein-coupled signalings, such as bradykinin and lysophosphatidic acid. Moreover, dibutyryl cAMP (data not shown) and phorbol 12-myristate 13-acetate did not increase the phosphorylation markedly. Considering that phosphorylation occurs on serine residues rather than on tyrosine residues, a serine kinase, other than protein kinase A or C, downstream of a tyrosine kinase must phosphorylate PIPKII␥. Although the exact roles of the phosphorylation re-main unclear, it is possible that the phosphorylation of PIPKII␥ regulates its localization. Hinchliffe et al. (12) reported that the translocation of PIPKII␣ to the cytoskeletal fraction of platelets in response to thrombin is inhibited by okadaic acid treatment, suggesting the importance of dephosphorylation for translocation. Although they also showed that the activity of PIPKII␣ is regulated by its phosphorylation state (23), we did not observe any change in the activity of PIPKII␥ after phosphorylation by mitogenic stimulation or dephosphorylation by alkaline phosphatase ( Fig. 4E and data not shown).
In this study, we demonstrated that PIPKII␥ is specifically localized in the ER in rat 3Y1 fibroblasts. Although most PIPK activity is found in the plasma membrane and cytosol, Helms et al. (24) reported that PI(4,5)P 2 synthesis occurs in the ER. Several phosphoinositide-metabolizing enzymes have been reported to be localized in the microsomal fraction. Wong et al. (25) reported that PI 4-kinase ␣ is localized in the ER, whereas PI 4-kinase ␤ is localized in the Golgi apparatus in HeLa cells. Most PI synthase activity is also detected in the ER (24,26,27). It is conceivable that PI(4,5)P 2 synthesis occurs efficiently in microsomes because of the relay of substrates between PI synthase, PI kinase, and PIPK. In addition, PI5P, the preferential substrate for type II isozymes in PI(4,5)P 2 synthesis, is rare in NIH3T3 cells (21) compared with PI4P, which exits abundantly in the cell. It may be important for this minor phosphoinositide to be localized in a restricted area such as in microsomes with its metabolizing enzyme, PI5P 4-kinase, for efficient PI(4,5)P 2 synthesis. Many of the characteristics of PI5P have yet to be elucidated, including its synthetic pathway as well as the identity of PI 5-kinase and its exact intracellular localization. However, together with the observation that PIPKII␥ is localized in the ER after phosphorylation by mitogenic signals, our results suggest that PIPKII␥ is involved in the synthesis of PI(4,5)P 2 in the ER.
Shibasaki et al. (28) reported that the type I PI4P 5-kinase overexpressed in COS-7 cells by an adenovirus expression system is localized mainly in plasma membranes and the cytosol. They further reported that type I PIPKs induce a pine needlelike structure of the actin cytoskeleton downstream from Rho. In contrast, we observed no change in the actin cytoskeleton when type II␤ and II␥ isozymes were transiently overexpressed in COS-7 cells (data not shown). From these results, it is possible to conclude that each subfamily of PIPK has a distinct localization and function and is also responsible for the synthesis of distinct intracellular PIP 2 sources.