Characterization of p150, an adaptor protein for the human phosphatidylinositol (PtdIns) 3-kinase. Substrate presentation by phosphatidylinositol transfer protein to the p150.Ptdins 3-kinase complex.

Genetic and biochemical studies have shown that the phosphatidylinositol (PtdIns) 3-kinase encoded by the yeast VPS34 gene is required for the efficient sorting and delivery of proteins to the vacuole. A human homologue of the yeast VPS34 gene product has recently been characterized as part of a complex with a cellular protein of 150 kDa (Volinia, S., Dhand, R., Vanhaesebroeck, B., MacDougall, L. K., Stein, R., Zvelebil, M. J., Domin, J., Panaretou, C., and Waterfield, M. D. (1995) EMBO J. 14, 3339-3348). Here, cDNA cloning is used to show that the amino acid sequence of this protein, termed p150, is 29.6% identical and 53% similar to the yeast Vps15p protein, an established regulator of Vps34p. Northern blot analysis showed a ubiquitous tissue distribution for p150 similar to that previously observed with PtdIns 3-kinase. Recombinant p150 associated with PtdIns 3-kinase in vitro in a stable manner, resulting in a 2-fold increase in lipid kinase activity. Addition of phosphatidylinositol transfer protein (PI-TP) further stimulated the lipid kinase activity of the p150·;PtdIns 3-kinase complex 3-fold. A PtdIns 3-kinase activity could also be co-immunoprecipitated from human cell lysates using anti-PI-TP antisera. This observation demonstrates that an interaction between a PtdIns 3-kinase and PI-TP occurs in vivo, which further implicates lipid transfer proteins in the regulation of PtdIns 3-kinase activity. These results suggest that the Vps15p·;Vps34p complex has been conserved from yeast to man and in both species is involved in protein trafficking.

The functional identity of organelles is maintained largely by a set of unique proteins that reside within them. As a result, the accurate and efficient sorting of proteins is crucial for the proper function of these organelles and ultimately of the cell itself. The yeast vps mutants define a set of 50 genes whose functions include those required for the sorting and delivery of soluble vacuolar hydrolases from the late trans-Golgi network (TGN) 1 to the vacuole, the yeast equivalent of the mammalian lysosome (1). Other phenotypes shown by the vps mutants involve a temperature-sensitive defect in growth and defects in osmoregulation and in vacuole segregation at mitosis. The VPS34 gene was originally identified as part of a molecular complex responsible for the control of intracellular protein trafficking in yeast (2,3). The VPS34 gene product, Vps34p, shows sequence similarity to phosphoinositide (PI) 3-kinases (2,4). Vps34p phosphorylates phosphatidylinositol (PtdIns) at the 3Ј-position of the inositol ring, but not PtdIns(4)P or PtdIns(4,5)P 2 (5); hence, Vps34p is a PtdIns-specific 3-kinase. This lipid kinase activity is of physiological relevance since point mutations in the Vps34p kinase domain result in defective vesicle-mediated transport, strongly suggesting a role for PtdIns(3)P as a signal transducer in protein trafficking (6). The severe defect in sorting displayed by another vps mutant, termed vps15, suggested that the VPS15 and VPS34 gene products act at the same step of the vacuolar protein-sorting pathway. Subsequent biochemical studies revealed that Vps34p and Vps15p, a 160-kDa Ser/Thr protein kinase, exist as a complex in vivo (7). Vps15p recruits Vps34p to the membrane of the Golgi complex and enhances Vps34p PtdIns 3-kinase activity. Mutational inactivation of Vps15p protein kinase activity stops its association with Vps34p, blocks activation of Vps34p lipid kinase activity, and prevents recruitment of Vps34p to the appropriate membrane site (5,7).
The first mammalian PI 3-kinase identified, a heterodimer of a regulatory p85 subunit and a catalytic p110␣ subunit, was purified and cloned from bovine brain (4,8,9). Subsequently, the use of PCR and degenerate oligonucleotides based on the kinase domain sequences of p110␣ and Vps34p made it possible to identify several forms of PI 3-kinase from mammals and Drosophila (10 -12). The identification of human (13), Drosophila, 2 soybean (15), Arabidopsis (16), and Dictyostelium (17) Vps34p homologues indicates that the amino acid sequence and enzymatic activity of this lipid kinase have been conserved through evolution and can be assigned to a distinct PI 3-kinase class (18). The human PtdIns 3-kinase has amino acid sequence homology (37% identity and 53% similarity) and similar biochemical characteristics to the yeast enzyme Vps34p, suggesting that it is a Vps34p homologue (13). In mammalian cells, the use of the potent PI 3-kinase inhibitors wortmannin and LY294002, which also inhibit PtdIns 3-kinase (13), causes mammalian lysosomal marker enzymes analogous to those studied in yeast to be mistargeted (19,20). These findings implicate a PI 3-kinase activity in vesicular transport from the Golgi complex to lysosomes. The identification of a human Vps34p homologue and the effect of PI 3-kinase inhibitors on protein sorting strongly suggest that a PI 3-kinase, which may be the PtdIns 3-kinase, is involved in protein trafficking.
Other genetic studies in yeast have independently implicated phosphoinositides in vesicular transport. Secretory protein transport from the Golgi complex to the plasma membrane in yeast is dependent upon the activity of a PtdIns/phosphatidylcholine phospholipid transfer protein (PI-TP) encoded by the SEC14 gene (21,22). A mammalian PI-TP was identified as a cytosolic factor required for the ATP-dependent priming of secretory granules in PC12 cells (23). PI-TP also participates in the synthesis of phosphoinositides by presenting PtdIns to PtdIns 4-kinase (24,25). Recent studies of the biogenesis of immature secretory vesicles from the TGN in cell-free systems have led to the isolation of mammalian PI-TP as a factor required for vesicle budding (26). Such studies implicate mammalian PI-TP in vesicular transport from the TGN. Furthermore, the requirement of PI-TP in Golgi complex function may reflect a general role for this protein in PtdIns substrate presentation (27) not only to PtdIns 4-kinase, but also to other PtdIns-specific lipid kinases such as the human PtdIns 3-kinase.
In this study, we describe the cloning and expression of a human 150-kDa protein (p150), which is homologous to the yeast Vps15p protein kinase. In addition, we report the association of p150 with the human Vps34p homologue, PtdIns 3-kinase, and demonstrate the activation of this p150⅐PtdIns 3-kinase complex when the substrate, PtdIns, is presented in the presence of PI-TPs.

EXPERIMENTAL PROCEDURES
Cloning and Sequencing of the p150 cDNA -The probe used for screening a U937 ZAP cDNA library was amplified by PCR. The forward and reverse primers 5Ј-CCAGATCCTTTCTGTAG and 5Ј-GAATGGACGGGTACTGATGC, respectively, were designed based on the human clone HFBEP44. To facilitate subcloning of the amplified fragment, an EcoRI restriction site was incorporated into the 5Ј-end of each primer. PCR was carried out using a Perkin-Elmer DNA thermal cycler for 30 cycles of 94°C for 1 min, 55°C for 15 s, and 72°C for 30 s. The 50-l reaction volume contained the U937 cDNA library, 10 mM Tris-HCl, pH 8.0, 50 mM KCl, 1.5 mM MgCl 2 , 200 M dNTP, 0.1 mM forward and reverse primers, and 2 units of Taq DNA polymerase (Life Technologies, Inc.). The 330-base pair PCR product was cloned into the EcoRI site of pBluescript SK. After sequencing, the PCR fragment was labeled with [␣-32 P]dCTP to a specific activity of 10 9 cpm/g of DNA using random primers (Amersham International) and was used to screen a U937 ZAP cDNA library (Stratagene). Approximately 3 ϫ 10 6 plaques of the library were plated and transferred to nylon filters (Hybond-N, Amersham International). They were hybridized with the [␣-32 P]dCTP-labeled probe for 16 h at 65°C in 0.5 M sodium phosphate, pH 7.2, 7% SDS, and 1 mM EDTA, pH 8.0. After washing with 0.5 ϫ SSC (1 ϫ SSC ϭ 0.15 M NaCl and 0.015 M sodium citrate) and 0.1% SDS twice for 20 min at 60°C, the filters were exposed to x-ray film for 48 h at Ϫ70°C. Following two rounds of screening, one positive clone was plaque-purified, and the insert (pSKp150) was rescued using the helper phage ExAssist (Stratagene). The nucleotide sequence was determined from the double-stranded pSKp150 plasmid. The nucleotide sequence of one strand was determined by Lark Sequencing Technology, Inc. using the dideoxy chain termination method (53), a modified T7 DNA polymerase, and [␣-35 S]thio-dATP. The complementary strand was sequenced on an ABI 370A sequencer using the Taq DyeDeoxy Terminator Cycle Sequencing system (ABI Advanced Biotechnologies, Inc.).
RNA Analysis-For Northern blotting, multiple human tissue RNA blots (CLONTECH) were used. The 330-base pair PCR fragment used to screen the cDNA library was labeled with [␣-32 P]dCTP by random priming as described above. Hybridization was carried out using Ex-pressHyb (CLONTECH); hybridization and washing were carried out according to the manufacturer's instructions.
Plasmid Constructs-For the expression of p150 as a GST fusion protein in insect Sf9 cells, the full-length open reading frame (ORF) was inserted into pAcGEX-2T in three steps. First, BamHI/NotI and EcoRI/ KpnI sites were introduced by PCR at the 5Ј-and 3Ј-ends, respectively, of nucleotides 1-350 of the ORF. The forward and reverse primers 5Ј-CGGATCCAGCGGCCGCATGGGAAATCAGCTTGCTGGC and 5Ј-CGGAATTCGCGGTACCGAATGGACGGGTACTGAT (restriction sites are underlined), respectively, were used in a 30-cycle PCR of 94°C for 30 s, 48°C for 15 s, and 72°C for 30 s. The final volume of 100 l contained 40 ng of pSKp150, 20 mM Tris-HCl, pH 8.8, 10 mM KCl, 10 mM (NH 4 ) 2 SO 4 , 2 mM MgSO 4 , 0.1% Triton X-100, 200 mM dNTP, each primer at 0.1 mM, and 1 unit of Vent DNA polymerase (New England Biolabs Inc.). This was subcloned into pAcGEX-2T cut with BamHI and EcoRI to create pAcGEX-2T/p150-(1-350). Second, nucleotides 141-3487 were introduced by digesting pSKp150 with StuI and KpnI and inserting this fragment into pAcGEX-2T/p150-(1-350) digested with the same enzymes, to create pAcGEX-2T/p150-(1-3487). Finally, this construct was digested with KpnI to allow insertion of nucleotides 3488 -4077, the final PCR-generated KpnI-digested 590-base pair fragment. This PCR product was created using the forward primer 5Ј-CGGTACCATGGCTTGTTGGGA (KpnI site is underlined) and the reverse primer 3Ј-CGGTACCAGCGGCCGCTTATTTCCACACCTT (KpnI/ NotI sites are underlined) using the PCR conditions described above. The introduction of NotI sites at the 5Ј-and 3Ј-ends allowed the fulllength 4077-base pair ORF from this construct to be subcloned into the NotI-digested baculovirus vector pVL1393 (Invitrogen), thereby enabling expression of p150 without the GST tag.
Cell Culture and Transfections-Jurkat 6 and U937 cells were cultured in RPMI 1640 medium supplemented with 5% fetal calf serum. Baculoviral plasmid DNA was cotransfected with Baculogold DNA (Pharmingen) into Sf9 insect cells using Lipofectin (Life Technologies, Inc.); recombinant virus was isolated and characterized as described previously (28).
Expression of Recombinant Protein-Sf9 cells were infected at a density of 1 ϫ 10 6 /ml with recombinant baculoviruses for 60 h. Cells were harvested and lysed in TBS-T (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, phenylmethylsulfonyl fluoride, and 100 units/ml aprotinin), and the extract was incubated with glutathione-Sepharose beads (Pharmacia Biotech Inc.) or antisera for 1 h at 4°C. GST-tagged proteins bound to beads or immunoprecipitates were washed four times with TBS-T, resolved by SDS-PAGE, and visualized by Coomassie Blue or silver staining. Recombinant PI-TP␣, PI-TP␤, and Sec14p were expressed and purified as described (29,30).
Immunoprecipitation and Western Blotting-Antisera against PtdIns 3-kinase were raised in rabbits and affinity-purified as described previously (13). Antibody for detecting PI-TP␣ and PI-TP␤ on immunoblots was raised in rabbits as described (31). For immunoprecipitating PI-TP, a mouse monoclonal antibody was used. For immunoprecipitations, cell lysates were clarified by centrifugation at 14,000 ϫ g for 10 min and incubated with antibody for 2 h at 4°C. Immunoprecipitates were collected on protein A-Sepharose or protein G-Sepharose beads (Pharmacia Biotech Inc.) and washed four times with TBS-T. Proteins were extracted in 2 ϫ Laemmli sample buffer, separated by SDS-PAGE, and transferred to polyvinylidene difluoride membrane (Millipore Corp.). Blots were incubated with antibody in phosphate-buffered saline containing 0.05% (w/v) Tween 20 and 5% (w/v) dried milk. Immunoreactive bands were visualized using horseradish peroxidase-conjugated anti-rabbit or anti-mouse antibody and an enhanced chemiluminescence kit (Amersham International).
Cell Labeling-Jurkat cells were spun down and resuspended at 2 ϫ 10 7 /ml in phosphate-free Dulbecco's modified Eagle's medium supplemented with 1% dialyzed fetal calf serum. Cells were then labeled with [ 32 P]orthophosphate (Amersham International) at 1 mCi/ml for 4 h at 37°C. After this time, radiolabeled cells were incubated in the presence and absence of phorbol 12-myristate 13-acetate (10 ng/ml) for 20 min, washed once with phosphate-buffered saline, and then solubilized with TBS-T. The extracts were precleared with protein A-Sepharose followed by immunoprecipitation with preimmune and PtdIns 3-kinase-specific antisera. Precipitates were washed four times with TBS-T and resolved by SDS-PAGE. Proteins were then transferred to Immobilon, which was autoradiographed and then subjected to Western blotting as described above. For analysis of p150 myristoylation, Sf9 cells (1 ϫ 10 6 ) were infected with recombinant baculovirus and labeled 24 h postinfection for 16 h with [9,10-3 H]myristic acid (Amersham International) at 100 Ci/ml. Labeled cells were lysed with TBS-T, and extracts were incubated with glutathione-Sepharose beads. Immobilized proteins were washed four times with TBS-T and resolved by SDS-PAGE. The gel was treated for fluorography using Amplify (Amersham International), dried, and autoradiographed.

Protein Kinase Assay of Recombinant PtdIns 3-Kinase and p150 -
The protein kinase activity of immobilized recombinant GST-PtdIns 3-kinase, p150, or GST-p150⅐PtdIns 3-kinase complex was determined by incubating the immobilized protein with 30 l of a phosphorylation mixture containing 0.1 Ci of [␥-32 P]ATP, 30 mM Tris-HCl, pH 8.0, 20 mM MgCl 2 or MnCl 2 , and 3.0 mg/ml protein or peptide substrate. The substrates used were as follows: protein kinase C substrate (PLSRTLS-VATAKK), myelin basic protein, Kemptide (LRRASLG), tyrosine protein kinase substrate (RRLIGDAGYAARG), p34 cdc2 protein kinase substrate (ADAQHATPPKKKRKEDPKDF), histone, and syntide 2 (PKTPKKAKKL). After 30 min of incubation at 30°C, the kinase reaction was stopped by spotting 25 l of the supernatant on phosphocellulose P-81 paper. [␥-32 P]ATP was separated from the labeled substrate by washing the P-81 filter papers five times for 5 min with 75 mM H 3 PO 4 . The papers were dried, and the radioactivity incorporated into protein and peptide substrates was determined by Cerenkov counting.
Assay of Phosphatidylinositide 3-Kinase Activity-PI 3-kinase assays were performed essentially as described (32), in a volume of 50 l containing 20 mM Tris-HCl, pH 7.5, 100 mM NaCl, 3.5 mM MnCl 2 , 40 mM ATP (0.2 mCi of [␥-32 P]ATP), and 1 mM phosphoinositide (PtdIns, PtdIns(4)P, or PtdIns(4,5)P 2 ; Sigma). After preincubating the enzyme and lipid for 20 min at 4°C, each reaction was initiated by the addition of ATP and allowed to proceed for 15 min at room temperature. To examine the effect of PI-TP on lipid kinase activity, PI-TP or Sec14p (0 -2.6 M final concentration) was preincubated with 1 mM PtdIns at 37°C for 5 min and then added to recombinant PtdIns 3-kinase for 20 min on ice before addition of ATP. Assays were terminated with acidified chloroform/methanol, and the lipids were extracted and resolved by TLC. For standard assays using PtdIns as substrate, the solvent used was chloroform, methanol, and 4 M ammonium hydroxide (45:35:10). For separation of PtdIns(3)P, PtdIns(3,4)P 2 , and PtdIns(3,4,5)P 3 , propan-1-ol and 2 M acetic acid (65:35) were used as the solvent system. For separation of PtdIns(3)P from PtdIns(4)P, a borate buffer system was used (33). PtdIns 4-kinase assays were performed on TBS-T A431 cell lysates in a volume of 50 l containing 10 mM MgCl 2 , 25 mM Tris-HCl, pH 7.5, 70 M ATP (0.2 Ci of [␥-32 P]ATP), and 1 mM PtdIns. Reactions were carried out as described above for PI 3-kinase assays. Radioactivity was analyzed by autoradiography or quantified using a Phosphor-Imager (Molecular Dynamics, Inc.).
Assay of PI-TP Association in Vitro-PI-TP (5 g) was preincubated with 1 mM PtdIns at 37°C for 5 min. Immunoprecipitated proteins or proteins immobilized on glutathione-Sepharose beads were then added in a volume of 300 l of TBS-T. Incubation was continued for 1 h at 4°C. The resulting protein complexes were washed once with TBS-T and twice with phosphate-buffered saline and resolved by SDS-PAGE. PI-TP was detected by Western blotting using specific antibody.

RESULTS
Cloning of p150 -Immunoprecipitations from human Jurkat 6 cells performed with an affinity-purified anti-PtdIns 3-kinase antibody contain a 150-kDa protein in addition to PtdIns 3-kinase. The partial amino acid sequence of the 150-kDa protein was determined, yielding five amino acid sequences: 1) kVF/ TAIQDPXLP, 2) kLLALK 3) kETFLSADEEk, 4) kVVTLLSD-PENIVFK, and 5) kXLNMENRHLN (12). No significant sequence similarities were identified when the p150 peptide sequences were compared with the protein sequences of the Swissprot data base. Using the program TFASTA (34), however, which compared the p150 peptides with the six possible translations of the data base DNA entries, the short anonymous expressed sequence tag (EST) HS8804 (human clone HFBEP44) was shown to encode a sequence identical to peptide 1. The human HFBEP44 cDNA sequence was translated, and the resulting amino acid sequence was used to rescreen the Swissprot data base using the program FASTA (34). This new search identified the N terminus of the yeast Vps15p protein kinase as having the greatest homology to the predicted prod-uct of HFBEP44. Using specific primers and reverse transcription-PCR, we amplified the HFBEP44 partial cDNA from U937 mRNA. This product was used as a probe to screen a ZAP cDNA library from the human cell line U937. One 5-kilobase clone was isolated and found to contain sequence identical to the HFBEP44 PCR product. This 5-kilobase cDNA contained a 4-kilobase ORF with a putative start codon preceded by a 5Ј-Kozak consensus site (35) and an in-frame stop codon. The ORF encodes a protein of 1358 amino acids (Fig. 1A), including all five of the sequenced peptides with a calculated molecular mass of 150,000 Da. The predicted amino acid sequence of this protein shows homology to yeast Vps15p over its entire length (29.6% identity and 53% similarity) (Fig. 1A) and is hereafter referred to as human p150.
Regions of Conserved Sequence in p150 -A comparison of the sequences of p150 and yeast Vps15p using the program COM-PARE (34) defined three main areas of colinear homology between the two proteins (Fig. 1, B and C): homology region i, amino acids 1-10; region ii, amino acids 11-262; and region iii, amino acids 400 -700.
The N-terminal region i contains a consensus site for the attachment of myristic acid. Myristic acid is cotranslationally added to an N-terminal glycine residue of specific eukaryotic proteins via an amide linkage following the removal of the initiating methionine to the adjacent glycine residue (36). Analysis of the substrate specificity of the yeast myristoyl-CoA: protein N-myristoyltransferase suggests that a consensus sequence exists for N-terminal myristoylation of G 1 X 2 Z 3 Z 4 S 5 Z 6 (where uncharged residues are indicated by X and neutral residues by Z) (37). The Vps15p sequence of (M)G 1 A 2 Q 3 L 4 S 5 L 6 fits this consensus, suggesting that Vps15p is myristoylated in vivo.
Homology region ii is conserved between p150 and Vps15p (47% identity and 67% similarity) and shows similarity to the catalytic region of Ser/Thr protein kinases. Both Vps15p and p150 contain the consensus triplets APE and DFA, which are often found within the catalytic domains of protein kinases (38,39). The central core of homology region ii (amino acids 100 -200 of Vps15p and p150) contains a stretch of amino acids that can be used as an indicator of substrate specificity. The shared sequence DIKTEN within this stretch closely resembles the consensus element DLKPEN, which is specific for Ser/Thr protein kinases (40), suggesting that Vps15p and p150 are Ser/Thr kinases. Indeed, Vps15p has been shown to be a functional Ser/Thr kinase in that it is able to autophosphorylate on a serine residue (5).
The central region of both p150 and Vps15p (homology region iii) shows similarity to a repeating unit found in a number of proteins including the 65-kDa regulatory subunit (PR65) of protein phosphatase 2A. The PR65 subunit consists of 15 leucine-rich imperfect repeating units of 38 -40 amino acids (41) that produce a rod-like structure important for intersubunit interactions in protein phosphatase 2A (42). Thus, the two PR65-related leucine-rich repeats found in Vps15p and p150 might also function as protein-protein interaction modules.
Region iv at the C terminus of p150 (amino acids 1000 -1300) contains WD repeat units, which were first found in the ␤-subunit of heterotrimeric G-proteins (43). These repeats have a region of variable length (6 -94 amino acids) followed by a core of more constant length (23-41 amino acids), which is bracketed by two characteristic pattern elements, a conserved His and WD: Northern blot analysis using poly(A) ϩ RNA isolated from human tissues revealed that the cDNA encoding p150 hybridized to a 5.5-kilobase species, which is similar in size to the isolated cDNA (Fig. 2). p150 is detectable in the majority of the human tissues analyzed, with the highest levels present in skeletal muscle, liver, ovary, testis, and thymus and the lowest levels in kidney and lung.
To initiate the biochemical characterization of p150, Sf9 cells were infected with a baculovirus encoding a GST-p150 fusion protein. Affinity purification on glutathione-Sepharose beads revealed a 180-kDa protein representing p150 tagged with the 27-kDa GST moiety. Co-infection with GST-p150 and a PtdIns 3-kinase baculovirus demonstrated that PtdIns 3-kinase associates with GST-p150 (Fig. 3A) and that the resulting complex possesses PI 3-kinase activity (Fig. 3B). Therefore, the GST tag at the N terminus of p150 did not significantly alter its ability to bind PtdIns 3-kinase or affect lipid kinase activity. Similarly, when Sf9 cells were co-infected with viruses expressing GST-PtdIns 3-kinase and p150 (Fig. 3C) or PtdIns 3-kinase and p150 (Fig. 3D), a p150⅐PtdIns 3-kinase complex could be detected. In contrast, p150 was not observed when GST-p110␣ was affinitypurified from Sf9 cells coexpressing GST-p110␣ and p150 (data not shown), indicating that p150 associates specifically with the PtdIns-specific PtdIns 3-kinase and does not associate with p110␣.
Post-translational Modification of PtdIns 3-Kinase and p150 -Consistent with the presence of an N-terminal myristoylation site, yeast Vps15p is myristoylated in vivo (45,46). To analyze whether p150 was also myristoylated in vivo, Sf9 cells were infected with baculovirus to express either GST-PtdIns 3-kinase or the GST-p150⅐PtdIns 3-kinase complex and radiolabeled with tritiated myristic acid. Subsequent autoradiography of affinity-purified GST-p150⅐PtdIns 3-kinase showed that p150 was specifically labeled, suggesting that it is modified by the addition of myristic acid in vivo (Fig. 4A).
Since both Vps15p and Vps34p exist as phosphoproteins in vivo (5), the phosphorylation state of the p150⅐PtdIns 3-kinase complex in [ 32 P]orthophosphate-radiolabeled U937 and Jurkat cells was investigated. The cells were incubated in the presence or absence of phorbol 12-myristate 13-acetate, and then lysates were prepared and immunoprecipitated with antisera specific to PtdIns 3-kinase (Fig. 4B). In U937 cells, both PtdIns 3-kinase and p150 appear to be phosphorylated. Stimulation with phorbol 12-myristate 13-acetate did not result in a significant increase in the phosphorylation of PtdIns 3-kinase or p150. In Jurkat cells, PtdIns 3-kinase and p150 were also found to be FIG. 1-continued   FIG. 2. Northern blot analysis of p150. A Northern blot (CLON-TECH) of human poly(A) ϩ RNA was hybridized with full-length p150 cDNA radiolabeled with [␣-32 P]dCTP. Lanes 1-16 contain 2 g of RNA from pancreas, kidney, skeletal muscle, liver, lung, placenta, brain, heart, peripheral blood leukocytes, colon, small intestine, ovary, testis, prostate, thymus, and spleen, respectively. Hybridization was performed at 65°C using the ExpressHyb protocol (CLONTECH). The positions of RNA molecular size markers (in kilobase pairs (kb)) are indicated. phosphorylated. Furthermore, a 120-kDa phosphoprotein was also co-immunoprecipitated from these cell lysates. We were, however, unable to observe the in vitro autophosphorylation of PtdIns 3-kinase (13) or p150 (data not shown), which has previously been reported for Vps34p and Vps15p (5). As a complementary approach, the protein kinase activities of the recombinant GST-p150⅐PtdIns 3-kinase complex toward peptide and protein substrates were investigated (Table I). These results demonstrate that the GST-p150⅐PtdIns 3-kinase complex is able to phosphorylate protein kinase C substrate and myelin basic protein. Interestingly, this protein kinase activity displays a cation preference for Mn 2ϩ , which is also observed for the Vps34p and Vps15p protein kinase activities (5).
Association of p150 Increases the PI 3-Kinase Activity of PtdIns 3-Kinase-To investigate whether p150 could regulate the enzymatic activity of PtdIns 3-kinase, insect cells were infected with baculoviruses to express PtdIns 3-kinase or both p150 and PtdIns 3-kinase. Proteins were immunoprecipitated with anti-PtdIns 3-kinase antibody, and the resulting immune complexes were assayed for lipid kinase activity. Each lipid kinase assay was performed in the presence of Mn 2ϩ using PtdIns as substrate (as described under "Experimental Procedures"). Quantitation of the phosphorylated lipid products using a PhosphorImager (Fig. 5A) demonstrated that the lipid kinase activity of PtdIns 3-kinase increases 2-fold when associated with p150. Western blotting using anti-PtdIns 3-kinase antibodies on immunoprecipitates produced in parallel to those used above (Fig. 5B) showed that equal amounts of the enzyme were used.
PI-TP Increases the PI 3-Kinase Activity of the GST-p150⅐PtdIns 3-Kinase Complex-Mammalian PI-TP has been identified as a factor required for the formation of secretory vesicles from the TGN and has been shown to play a role in lipid substrate presentation to PtdIns 4-kinase (24,26). To investigate the possible effect of PI-TP on PtdIns 3-kinase activity, the GST-p150⅐PtdIns 3-kinase complex expressed in Sf9 cells was used in lipid kinase assays in the absence or presence of various PI-TPs. Before starting the reaction, PI-TP␣, PI-TP␤, or Sec14p protein (0 -2.6 M) that had been preincubated with 1 mM PtdIns was added to the complex in a dose-dependent manner. The radiolabeled PtdIns(3)P produced was extracted and quantitated using PhosphorImager analysis. Enzymatic activity was expressed as a percentage change over that detected in reactions performed in the absence of PI-TP (Fig. 6). The results showed that a dose-dependent increase in lipid kinase activity took place upon the addition of PI-TP. At concentrations over 1 M, both PI-TP␣ and PI-TP␤ exhibited slight inhibitory effects. Sec14p gave the greatest increase in PtdIns 3-kinase activity (375%); PI-TP␤ gave a 315% increase, and PI-TP␣ increased activity by 275%. To investigate if PI-TP physically associated with the GST- GST fusion proteins were affinitypurified from labeled lysates using glutathione-Sepharose beads. Immobilized proteins were washed and resolved by SDS-PAGE. Following fluorographic processing, radioactive bands were visualized by autoradiography. B, in vivo phosphorylation of PtdIns 3-kinase (PtdIns 3-K) and p150. U937 and Jurkat cells (2 ϫ 10 7 cells/sample) were radiolabeled with [ 32 P]orthophosphate for 4 h at 37°C and then treated in the presence and absence of phorbol 12-myristate 13-acetate (PMA; 10 ng/ml) as described under "Experimental Procedures." Using antisera specific to PtdIns 3-kinase, protein complexes were immunoprecipitated from U937 cells (lanes 2 and 3) and Jurkat cells (lane 5). Controls were performed using preimmune serum on radiolabeled U937 (lane 1) and Jurkat (lane 4) cells. Immunoprecipitates were washed extensively, resolved by SDS-PAGE, and transferred to Immobilon. Radioactive proteins were visualized by autoradiography.   in the presence of Mg 2ϩ or Mn 2ϩ Sf9 cells were infected with GST-PtdIns 3-kinase and p150 baculovirus, and after 60 h, GST fusion protein was affinity-purified from cell lysates using glutathione-Sepharose beads. The immobilized GST-p150⅐PtdIns 3-kinase complex was incubated for 30 min with different protein kinase substrates and 0.1 Ci of [␥-32 P]ATP as described under "Experimental Procedures." All substrates used were at a final concentration of 3.0 mg/ml. Background phosphorylation has been subtracted to show the phosphorylation attributable to the GST-p150⅐PtdIns 3-kinase complex.  5. Effect of p150 on the lipid kinase activity of PtdIns 3-kinase. A, using anti-PtdIns 3-kinase antisera, immunoprecipitates were obtained from Sf9 cells infected with PtdIns 3-kinase (PtdIns 3-K) or with PtdIns 3-kinase and p150. These immune complexes were subjected to in vitro PI 3-kinase assays in the presence of Mn 2ϩ using PtdIns as substrate as described under "Experimental Procedures." Extracted lipids were resolved by TLC and quantitated using the Phos-phorImager system. PI(3)P, (PtdIns(3)P). B, immunoprecipitates produced in parallel to those used in A were fractionated by SDS-PAGE. Immunoprecipitated recombinant PtdIns 3-kinase (lane 1) and PtdIns 3-kinase coexpressed with p150 (lane 2) were transferred to Immobilon and incubated with anti-PtdIns 3-kinase antibody. Immunoreactive proteins were visualized by ECL. p150⅐PtdIns 3-kinase complex, recombinant PI-TP␤ was incubated with either GST protein or GST-p150⅐PtdIns 3-kinase, and the associated proteins were collected on glutathione-Sepharose beads. Western blot analysis indicated that PI-TP␤ associated with the GST-p150⅐PtdIns 3-kinase complex (Fig.  6D, lane 3

), but not with GST alone (lane 2).
A PtdIns-specific PI 3-Kinase Associates with PI-TP in Vivo-To investigate whether PI-TP associates with a PtdIns 3-kinase in vivo, PI-TP was immunoprecipitated from Jurkat cells and assayed for kinase activity toward PtdIns in the presence of Mn 2ϩ . Extracted lipids were separated using a borate solvent system (33), which separates PtdIns(4)P from PtdIns(3)P (Fig. 7A). A lipid kinase activity co-immunoprecipitating with PI-TP was detected and found to be a PI 3-kinase and not a PI 4-kinase (Fig. 7A). We were unable to detect PtdIns 3-kinase by Western blot analysis of the PI-TP immunoprecipitates; this could be due to a lack of sensitivity, which may fail to show the presence of low levels of PtdIns 3-kinase in these immunoprecipitates. To investigate the substrate specificity of this PI-TP-associated PI 3-kinase further, PI-TP was immunoprecipitated from Jurkat cells, and the immune complex was subjected to PI 3-kinase assays in the presence of Mg 2ϩ or Mn 2ϩ using PtdIns, PtdIns(4)P, or PtdIns(4,5)P 2 as substrate (Fig. 7B). The PI 3-kinase activity associated with PI-TP primarily used PtdIns as substrate and exhibited a cation preference for Mn 2ϩ . This biochemical profile parallels that of the human Vps34p homologue, PtdIns 3-kinase, which uses PtdIns as its preferred substrate and has a cation preference for Mn 2ϩ . These results demonstrate biochemically that PI-TP associates with a PtdIns-specific 3-kinase in vivo. DISCUSSION The Ser/Thr protein kinase encoded by the VPS15 gene and the PtdIns 3-kinase encoded by the VPS34 gene have previ-ously been shown to be required for the efficient sorting and delivery of proteins to the yeast vacuole (2,45,47). The yeast Vps34p lipid kinase, unlike other PI 3-kinases, can only use PtdIns as substrate. A recently characterized human PtdIns 3-kinase has been shown to have extensive sequence homology to Vps34p and to display a substrate specificity for PtdIns, suggesting that the major components of the yeast VPS intracellular trafficking complex are conserved between the species (13).
In this study, we describe the cloning and molecular characterization of human p150. This protein was initially observed in complex with human PtdIns 3-kinase when immunoprecipitated from Jurkat cells using affinity-purified anti-PtdIns 3-kinase antisera (13). Human p150 and yeast Vps15p share sequence homology (29.6% identity and 43% similarity) at the amino acid level (Fig. 1A), which can be divided into four main areas of interest (Fig. 1, B and C). The N-terminal region represents a site for the attachment of myristic acid. This modification has been demonstrated biochemically for Vps15p (46), and in the present study, we confirm this for p150 (Fig.  4A). Myristoylation of p150 might allow its targeting to the cytoplasmic face of the Golgi membrane in mammalian cells, thereby allowing PtdIns 3-kinase to gain access to PtdIns in the membrane. The second area of homology is also located at the N terminus and has sequence similarity to the catalytic region of Ser/Thr protein kinases. Both Vps15p and Vps34p have been shown to be phosphorylated in vivo and can autophosphorylate in vitro (5). Autophosphorylation activity could not be demonstrated for either the human PtdIns 3-kinase (13) or p150. PtdIns 3-kinase and p150 were shown to be phosphorylated in vivo using both U937 and Jurkat cells. In addition, a 120-kDa phosphorylated protein was also co-immunoprecipitated from Jurkat cells; this is similar in size to a protein previously FIG. 6. Effect of PI-TP on PtdIns 3-kinase activity and in vitro association of PI-TP with GST-p150⅐PtdIns 3-kinase. Sf9 cells were infected with GST-p150 and PtdIns 3-kinase. Coexpressed protein complexes were affinity-purified from cell lysates using glutathione-Sepharose beads. PI 3-kinase assays were performed on the immobilized protein in the presence of Mn 2ϩ using PtdIns as substrate. To investigate the effect of PI-TP on the lipid kinase activity of the complex, increasing amounts of Sec14p (A), PI-TP␣ (B), or PI-TP␤ (C) were preincubated with 1 mM PtdIns at 37°C for 5 min. This was then added to the GST-p150⅐PtdIns 3-kinase complex, and the reaction was initiated as described under "Experimental Procedures." Extracted lipids were resolved by TLC and quantitated using the PhosphorImager. The increase in lipid kinase activity is expressed as a percentage of that given by the GST-p150⅐PtdIns 3-kinase complex in the absence of PI-TP. To investigate the association of PI-TP with the GST-p150⅐PtdIns 3-kinase complex, PI-TP␤ (5 g) was preincubated with 1 mM PtdIns at 37°C for 5 min (D). This was added to either immobilized GST protein (lane 2) or GST-p150⅐PtdIns 3-kinase (lane 3) in 300 l of TBS-T for 1 h at 4°C. Protein complexes were washed and resolved by SDS-PAGE. PI-TP was detected by Western blotting using PI-TP-specific antibody. Recombinant PI-TP␤ protein was run as a control in lane 1. observed to be associated in PtdIns 3-kinase immunoprecipitates from Jurkat cell lysates (13). Despite the inability of p150 and PtdIns 3-kinase to mediate autophosphorylation in vitro, a complex of both enzymes was able to phosphorylate exogenous peptide and protein substrates in a Mn 2ϩ -dependent manner. The preference for Mn 2ϩ resembles the Mn 2ϩ dependence of Vps34p and Vps15p for autocatalytic activity. These findings support the hypothesis that similar regulatory mechanisms exist for the protein kinase activity of the yeast and human proteins. Homology region iii of both p150 and Vps15p is a domain that could be involved in protein-protein interactions, possibly mediating binding to PtdIns 3-kinase (Fig. 1C). Region iv of p150 contains WD motifs; these repeat elements are also thought to mediate protein-protein interactions. The absence of similar consensus WD motifs in Vps15p may impart functional or regulatory differences between the two proteins.
Co-infection experiments in Sf9 cells demonstrated that recombinant p150 protein is able to associate with PtdIns 3-kinase; the inability of p150 to interact with mammalian p110␣ shows that this binding is specific to PtdIns 3-kinase. The p150-PtdIns 3-kinase interaction results in a 2-fold stimulation of lipid kinase-specific activity. These findings are consistent with those reported from yeast studies where strains deleted for VPS15 are extremely defective in Vps34p lipid kinase activity (7). Formation of the p85/p110 PI 3-kinase heterodimer does not lead to lipid kinase activation. In contrast, when p85␣ associates with p110␣ in vitro, a decreased lipid kinase activity is observed, which correlates with serine phosphorylation of the p85␣ adaptor subunit (48).
The amino acid homology of PtdIns 3-kinase and p150 to yeast Vps34p and Vps15p, respectively, suggests that function could also be conserved through evolution. This implicates the p150⅐PtdIns 3-kinase complex in the regulation of mammalian vesicle formation and protein traffic from the TGN to organelles such as the lysosome that are functionally analogous to the yeast vacuole. The observation that wortmannin causes mistargeting of procathepsin D from the Golgi complex to the plasma membrane in mammalian cells implicates a PI 3-kinase in intracellular protein sorting (19,20). Additionally, wortmannin has been observed to affect the localization of lysosomal type I integral membrane glycoproteins in cells, suggesting a role for PI 3-kinase activity in regulating membrane traffic late in the endocytic pathway (49).
In mammalian cells, the 3Ј-phosphorylated products of the p85/p110 PI 3-kinase, such as PtdIns(3,4)P 2 and PtdIns-(3,4,5)P 3 , have been postulated to serve as second messenger molecules in regulating cell growth and proliferation (50). Recent studies suggested that PtdIns(3)P is likely to have an independent physiological role since levels of PtdIns(3,4,5)P 3 and PtdIns(3,4)P 2 , but not PtdIns(3)P, increase rapidly following agonist stimulation. As the production of PtdIns(3)P remains constant, this phospholipid may play a constitutive role in mammalian cells, facilitating the formation of vesicles (27). This would suggest that a PtdIns-specific 3-kinase is required in vesicle and protein traffic. Phospholipid metabolism has already been demonstrated to regulate protein traffic through the secretory pathway. Sec14p, which acts as a PI-TP, is required for Golgi function in yeast (21). Mammalian PI-TP isoforms have also been identified as factors required for the formation of secretory vesicles from the TGN (26). Although yeast Sec14p has little sequence homology to mammalian PI-TP at the amino acid level, it is able to substitute for mammalian PI-TP in secretory vesicle formation. This suggests that there is conservation of binding and exchange functions and perhaps three-dimensional structure (26). Furthermore, PI-TP has been shown to have a role in substrate presentation to PtdIns 4-kinase. This lipid kinase is thought to be involved in the regulation of intracellular protein traffic because it colocalizes with secretory vesicles and insulin-responsive membrane vesicles in rat adipocytes (51). Additionally, PtdIns 4-kinase activity is required for the stimulated secretion of granuleassociated catecholamine from chromaffin cells (52).
Our findings show that PI-TP may also present PtdIns to other lipid kinases. Addition of PI-TP stimulated the lipid kinase activity of the p150⅐PtdIns 3-kinase complex in a dosedependent manner. Of the two mammalian PI-TPs, PI-TP␤ provided the greater stimulation of PtdIns 3-kinase activity. Stimulation of this lipid kinase may occur as a result of a stable interaction since we show that a PI 3-kinase activity is present in PI-TP immunoprecipitates from Jurkat cells. Further characterization indicates that this PI-TP-associated PI 3-kinase could be PtdIns 3-kinase since it has a substrate preference for PtdIns and a cation preference for Mn 2ϩ , which matches the biochemical profile of the human Vps34p homologue. This interaction is consistent with immunofluorescence studies showing the PI-TP␤ isoform to be preferentially associated with the Golgi system (14).
In this study, we have described the identification of p150, the human homologue of yeast Vps15p. Its ability to act as an adaptor protein for PtdIns 3-kinase demonstrates that the p150⅐PtdIns 3-kinase complex is likely to be functionally equivalent to the yeast Vps15p/Vps34p protein-sorting system. The lipid transfer protein PI-TP associates with the p150⅐PtdIns 3-kinase complex both in vitro and possibly in vivo, resulting in the activation of lipid kinase activity. We propose that the FIG. 7. A PtdIns-specific PI 3-kinase activity associates with PI-TP in vivo. A, PI 3-kinase activity was immunoprecipitated from Jurkat cells (2 ϫ 10 7 cells) using anti-PI-TP and anti-PtdIns 3-kinase antisera. Lipid kinase assays were performed on the immune complexes in the presence of Mn 2ϩ using PtdIns as substrate as described under "Experimental Procedures." The positions of PtdIns(4)P (PI(4)P) and PtdIns(3)P (PI(3)P) standards generated using A431 lysates (lane 1) and PtdIns 3-kinase (lane 2), respectively, are indicated. Lane 3 shows the lipid kinase activity associated with anti-PI-TP immunoprecipitates from Jurkat cells, and lane 4, that associated with protein G-Sepharose alone. B, lipid kinase assays were performed on recombinant GST-p110 and immunoprecipitates (IP) obtained from Jurkat cells using either anti-PtdIns 3-kinase (PtdIns 3-K) or anti-PI-TP antisera. The assays were performed in the presence of 10 mM Mg 2ϩ or Mn 2ϩ using PtdIns (lanes 1), PtdIns(4)P (lanes 2), or PtdIns(4,5)P 2 (lanes 3) as the lipid substrate.
formation of this heterotrimeric complex plays a role in protein trafficking in mammalian cells through the generation of PtdIns(3)P.