Annexin VII and Annexin XI Are Tyrosine Phosphorylated in Peroxovanadate-treated Dogs and in Platelet-derived Growth Factor-treated Rat Vascular Smooth Muscle Cells*

The intraperitoneal administration of peroxovanadate results in the rapid accumulation of many tyrosine-phosphorylated proteins in the liver and kidney of treated animals. The availability of large pools of tyrosine-phosphorylated proteins derived from normal tissues facilitates the purification and identification of previously unknown targets for cellular tyrosine kinases. Using this procedure, we have thus far identified four proteins in the liver and kidney of peroxovanadate-treated dogs. Two of these, annexin VII and annexin XI, were novel and had not been previously reported to be substrates of tyrosine kinases while the remaining two, ezrin and clathrin, have been reported to be tyrosine phosphorylated in some cell culture systems. In the present study, isolated proteins were identified both by sequence analysis and immunological methods. Annexin VII and annexin XI are present in cultured rat vascular smooth muscle cells and both were tyrosine phosphorylated in response to a physiological ligand, platelet-derived growth factor-BB (PDGF-BB). Furthermore, the extent of tyrosine phosphorylation in response to PDGF-BB was augmented by the co-addition of peroxovanadate to cell cultures. In vitrophosphorylation assays showed that PDGF receptor, calcium-dependent tyrosine kinase (CADTK/Pyk-2), Src kinase, and epidermal growth factor receptor all were able to phosphorylate purified annexin VII and XI on tyrosine residues. These findings confirm the usefulness of phosphatase inhibition by peroxovanadate as a tool for identifying previously unknown physiological targets for cellular protein tyrosine kinases.

The intraperitoneal administration of peroxovanadate results in the rapid accumulation of many tyrosinephosphorylated proteins in the liver and kidney of treated animals. The availability of large pools of tyrosine-phosphorylated proteins derived from normal tissues facilitates the purification and identification of previously unknown targets for cellular tyrosine kinases. Using this procedure, we have thus far identified four proteins in the liver and kidney of peroxovanadatetreated dogs. Two of these, annexin VII and annexin XI, were novel and had not been previously reported to be substrates of tyrosine kinases while the remaining two, ezrin and clathrin, have been reported to be tyrosine phosphorylated in some cell culture systems. In the present study, isolated proteins were identified both by sequence analysis and immunological methods. Annexin VII and annexin XI are present in cultured rat vascular smooth muscle cells and both were tyrosine phosphorylated in response to a physiological ligand, platelet-derived growth factor-BB (PDGF-BB). Furthermore, the extent of tyrosine phosphorylation in response to PDGF-BB was augmented by the co-addition of peroxovanadate to cell cultures. In vitro phosphorylation assays showed that PDGF receptor, calcium-dependent tyrosine kinase (CADTK/Pyk-2), Src kinase, and epidermal growth factor receptor all were able to phosphorylate purified annexin VII and XI on tyrosine residues. These findings confirm the usefulness of phosphatase inhibition by peroxovanadate as a tool for identifying previously unknown physiological targets for cellular protein tyrosine kinases.
Although tyrosine-phosphorylated proteins constitute only a very small percentage of the total phosphoprotein content of cells, the phosphorylation and dephosphorylation of specific tyrosine residues plays an important regulatory role in signal transduction, cell cycle control, and differentiation (1). Enhancement of the tyrosine phosphorylation of specific cellular proteins may be induced by treatment with appropriate cell activating ligands (growth factors, hormones, and cytokines). Many of these extracellular ligands directly or indirectly activate specific tyrosine kinases. Alternatively, we have previ-ously reported that the simple intraperitoneal injection of a tyrosine phosphatase inhibitor (peroxovanadate) into mice, in the absence of any added ligand, results within minutes in the appearance of numerous tyrosine-phosphorylated proteins in liver and kidney. These include the EGF-R, 1 insulin receptor, hepatocyte growth factor receptor, SHC, Stat 1␣, Stat 1␤, Stat 3, Stat 5, phospholipase C␥, insulin receptor substrate-1, ␤-catenin, ␥-catenin, SHP-1, SHP-2, etc., all of which were identified using antibodies to known tyrosine-phosphorylated proteins (2). These results emphasize the importance of phosphatase activity on the steady-state levels of phosphotyrosine in cellular proteins and the extent to which global tyrosine kinase activity is always "on" in the intact animal (2).
In the present study, in situ administration of peroxovanadate to dog was used to generate a sufficiently large pool of tyrosine-phosphorylated proteins in liver and kidney to attempt the biochemical isolation and identification of proteins not previously known to be tyrosine phosphorylated. Among the dozens (hundreds?) of phosphotyrosine-containing proteins that could be induced in these organs of the intact animal, we now report the identification of four: annexin VII, annexin XI, clathrin heavy chain, and ezrin. Tyrosine phosphorylation of ezrin and clathrin heavy chain has been detected in some cell culture systems (3)(4)(5). Since tyrosine phosphorylation of annexin VII and XI has not previously been reported, we present evidence for the physiological relevance of these phosphorylations: PDGF-BB induces the tyrosine phosphorylation of both annexins in rat vascular smooth muscle cells (VSMC) and both annexins are tyrosine phosphorylated in vitro by a number of receptor and cytoplasmic tyrosine kinases.

EXPERIMENTAL PROCEDURES
Materials-Immobilon-P membranes were from Millipore (Bedford, MA). Pre-stained molecular weight standards were from Life Technologies, Inc. (Gaithersburg, MD). Polyclonal (rabbit) antibody produced against a unique peptide sequence in annexin XI was a generous gift from Dr. C. Towle (Harvard University) and was used for Western blotting (6). We also thank Dr. W. van Vernrooij (University of Nijmegen) for a gift of an antiserum containing annexin XI antibodies from a patient with an autoimmune disease (antibodies to annexin XI can be found in approximately 10% of patients with systemic autoimmune disorders (7)). The patient antibodies were used for annexin XI immunoprecipitation reactions. The following monoclonal antibodies were obtained from Transduction Laboratories (Lexington, KY): RC20H (horseradish peroxidase-conjugated anti-phosphotyrosine), annexin VII, ezrin, and clathrin. A second monoclonal clathrin antibody prepared by Dr. J. Ostermann (Vanderbilt) from X22 cell line (ATCC, Manassas, VA) was used for immunoprecipitation. Polyclonal annexin VII antibody was a generous gift from Dr. H. Pollard (Uniformed Service University, Bethesda). Horseradish peroxidase-conjugated goat anti-mouse IgG and goat anti-rabbit IgG polyclonal antibodies were from Transduction Laboratories (Lexington, KY) and Cappell (Durham, NC), respectively. Polyclonal antibody to PDGF-R was purchased from Upstate Biotechnology (Lake Placid, NY). Src kinase and JAK2 kinase (Protein A-Sepharose conjugate) were from Upstate Biotechnology (Lake Placid, NY). Polyclonal antibody to calcium-dependent tyrosine kinase (CADTK/Pyk-2) was the generous gift of Dr. Shelton Earp (University of North Carolina). Polyclonal antibody to Fak kinase was the gift of Dr. Steven Hanks (Vanderbilt). Pre-swollen DE52 (diethylaminoethyl cellulose) anion exchange resin was from Whatman (Kent, United Kingdom). Anti-phosphotyrosine-agarose was from Zymed Laboratories Inc. (South San Francisco, CA). Mouse EGF and EGF-R coupled to Affi-Gel were prepared in this laboratory. Recombinant human PDGF-BB and fibroblast growth factor-basic were purchased from Research Diagnostics, Inc. (Flanders, NJ). Angiotensin II, lysophosphatidic acid, and 12-O-tetradecanoylphorbol-13-acetate were gifts from Dr. T. Inagami (Vanderbilt). Insulin was from Squibb-Novo, Inc. (Princeton, NJ). Hepatocyte growth factor was a gift from Dr. R. Harris (Vanderbilt). Liver and kidney tissues from mouse and rat were purchased from Pel-Freez Biologicals (Rogers, AR) or obtained from stock animals in our laboratory. All other reagents were from Sigma.
Treatment of Dogs and Preparation of Tissue-A 10 mM solution of sodium vanadate in PBS was prepared by heating to boiling. Fifteen minutes prior to use, 30% H 2 O 2 was added to the vanadate solution at room temperature to a final concentration of 100 mM. The peroxovanadate solution or PBS alone was injected intraperitoneally into anesthetized adult dogs at a dose of 5 ml/kg of body weight. Twenty minutes following treatment, dogs were sacrificed and tissues removed, cut into small pieces, and immediately frozen in liquid nitrogen. Dogs (approximately 25 kg) used in these studies were obtained from the Surgery Department at Vanderbilt University and were scheduled for euthanasia following surgical procedures.
Two methods were used for tissue fractionation. In Method A, proteins associated with a membrane/particulate fraction in a Ca 2ϩ -dependent manner were isolated as follows. Ten percent homogenates (wet w/v) were prepared in a Ca 2ϩ -containing buffer (10 mM Tricine, pH 8.4, 0.1 M NaCl, 2 mM CaCl 2 ) using a Polytron and then centrifuged at 100,000 ϫ g to separate the soluble from the membrane/particulate bound fraction. The pellet was then extracted with one-fourth original volume of an EDTA-containing buffer (5 mM EDTA, 10 mM Tricine, pH 8.4) and again centrifuged to obtain an EDTA-solubilized fraction. In Method B for tissue fractionation, tissues were homogenized in an EDTA-containing buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1.5 mM EDTA) and centrifuged at 100,000 ϫ g to obtain a supernatant and a pellet. The pellet was extracted with one-half original volume of 0.1 N Na 2 CO 3 to obtain a membrane-associated protein fraction and then with one-half of the original volume of RIPA buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Triton, 0.25% deoxycholic acid, 1.5 mM EDTA) to obtain the detergent-soluble fraction (intrinsic membrane proteins). All buffers contained 1 mM sodium vanadate and 50 M sodium molybdate to inhibit protein-tyrosine phosphatases during tissue and sample manipulations. Buffers used for tissue homogenization also contained protease inhibitors (Complete TM Mini, EDTA-free protease inhibitor mixture tablets, Roche Molecular Biochemicals, Indianapolis, IN) and iodoacetic acid (5 mM) to inhibit sulfhydryl-containing proteases.
Ion Exchange Column Chromatography-A 5-ml bed volume (1 ϫ 6.5 cm) of DE52 resin in Tricine buffer (10 mM Tricine, pH 8.4) was loaded with 20 ml of EDTA-extractable proteins obtained by Method A described above. The flow-through was collected. The column was then washed with 6 ml of Tricine buffer. Two linear gradients were used to elute bound proteins. In the first 18-ml gradient, Tricine buffer containing 0 to 0.2 M ammonium acetate was used for elution. The second 18-ml gradient contained 0.2 to 1 M ammonium acetate in Tricine buffer. A total of 36 fractions of ϳ1 ml each were collected. All buffers contained 1 mM sodium vanadate and 50 M molybdate.
Anti-phosphotyrosine Affinity Chromatography-An anti-phosphotyrosine-agarose column was prepared using 50 l of packed gel. The column was equilibrated with a Hepes buffer (20 mM Hepes, pH 7.4, 50 mM NaCl) prior to sample loading. Three sets of pooled fractions (fractions 6 -13, 14 -21, 22-30) from the DE52 column ( Fig. 2) were dialyzed overnight against 20 mM Hepes buffer, pH 7.4. The samples were then concentrated to ϳ1 ml using Centricon 30 microconcentrators from Amicon (Beverly, MA). The concentrated fractions were then applied to the column and the flow-through collected and reapplied to the column 20 times. After washing the column three times with 1 ml of Hepes buffer, the phosphotyrosine proteins were eluted with 100 l of 10 mM phosphotyrosine plus 10 mM phenyl phosphate in Hepes buffer and analyzed by SDS-PAGE. All buffers contained 100 M sodium vanadate.
Phospholipid Binding-Phospholipid vesicles (7.5 mg/ml) were prepared as described previously using a phosphatidylserine:cholesterol ratio of 2.5:5 mg/ml (8,9). Five-hundred microliters of sample from DE52 column fractions (described in Fig. 2) and 15 l of the phospholipid-containing vesicle preparation were mixed and left to stand at 4°C for 15 min. After ultracentrifugation at 150,000 ϫ g for 15 min at 4°C, the supernatant was removed and mixed with a fresh 15-l aliquot of phospholipid vesicles and 5 mM Ca 2ϩ (final). The mixture was incubated for 45 min at 4°C followed by ultracentrifugation at 100,000 ϫ g for 15 min at 4°C. The resulting pellet was washed, re-centrifuged, and adsorbed proteins were analyzed by SDS-PAGE.
Western Blotting and Immunoprecipitation-For Western blotting, samples were separated by SDS-PAGE, transferred to Immobilon-P membranes, and probed with antibodies as indicated in the figure legends. Antibody binding was detected by incubation of the membranes with horseradish peroxidase-conjugated goat anti-mouse or goat anti-rabbit secondary antibody followed by ECL, except in the case of anti-phosphotyrosine blots where the primary antibody (RC20H) is conjugated directly to horseradish peroxidase.
For immunoprecipitation reactions, aliquots (200 -500 l) of tissue samples from dog kidney were prepared by Method A or Method B. Samples were incubated with 3 l of monoclonal clathrin or monoclonal annexin VII antibodies for 2 h followed by addition of 3 g of anti-mouse IgG-agarose for 30 min. For immunoprecipitation of annexin XI, the human serum containing anti-annexin XI antibodies was incubated with tissue sample for 2 h followed by addition of 50 l of protein A-agarose for 1 h. The remaining steps of the immunoprecipitation reactions were performed as described previously (2).
Amino Acid Sequence Analysis-Coomassie Blue-stained purified proteins obtained from slices of Immobilon-P membrane were submitted to the Vanderbilt Protein Chemistry Core facility for amino acid sequence analysis. Proteins were directly sequenced on a PE Applied Biosystems Procise 492 Protein Sequencer (Foster City, CA). If direct N-terminal sequencing was not successful, protein from slices of Immobilon-P was digested with endoproteinase Lys-C and/or Asp-N. Resulting polypeptide fragments were then fractionated by high performance liquid chromatography and sequenced. Sequences were then checked against peptide data bases to match experimental results with existing proteins.
Cell Culture-Rat VSMC were prepared from the thoracic aorta of 12-week-old Harlan-Sprague Dawley rats (Charles River Breeding Laboratories) by the explant method and cultured in Dulbecco's modified Eagle's medium containing 10% fetal calf serum as described previously (10). Cells (passages 3-15) at Ͼ80% confluence in 100-mm dishes were made quiescent by incubation with serum-free Dulbecco's modified Eagle's medium for 24 h prior to treatment with indicated ligands or peroxovanadate.
After treatment with the indicated stimuli, cells were washed and then lysed by passage through a 21-gauge needle 20 times. The lysis buffer (1 ml/dish) contained 20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1.5 mM EDTA, 1 mM sodium vanadate, 50 M molybdate, and protease inhibitors as described above. Following ultracentrifugation for 30 min, the supernatant from the lysate was mixed with phospholipid vesicles (30 l) and 10 mM Ca 2ϩ for 45 min at 4°C. Samples were then centrifuged again and the pellet analyzed by SDS-PAGE and Western blotting as described above.
In Vitro Phosphorylation Assays-In vitro kinase assays were performed with purified annexin VII and XI and selected kinases. Annexin VII and XI were purified together from control dog kidney by ion exchange chromatography (as described above) followed by a 28% ammonium sulfate precipitation of column fractions containing annexin VII and XI. The purified annexin VII and XI were essentially free of other proteins (as judged by Coomassie Blue staining). Kinase assays were performed with Src kinase, CADTK/Pyk-2 kinase, JAK2 kinase, Fak kinase, PDGF-R kinase, and EGF-R kinase in buffer containing 20 mM Hepes, pH 7.4, 20 mM MgCl 2 , 1 mM MnCl 2 , 100 M VO 4 , 1 mM dithiothreitol, and 5 mM ATP. Mixtures of substrate (ϳ50 ng/40 l of reaction), kinase, and buffer containing ATP were incubated at room temperature for 15 min prior to centrifugation and analysis by SDS-PAGE. For experiments with CADTK/Pyk-2 kinase, PDGF-R, and Fak kinase, the kinases were immunoprecipitated from homogenates of control dog kidney and bound to Protein A-Sepharose. After incubation with substrate and ATP, the Protein A-Sepharose and attached antibodies was removed by centrifugation to eliminate immunoglobin heavy chains from the reaction mixture prior to SDS-PAGE.

Intraperitoneal Injection of Peroxovanadate in Dog Results in the Tyrosine Phosphorylation of Multiple Proteins in Liver and
Kidney-In previous experiments, the mouse system proved quite successful for the identification of tyrosine-phosphorylated proteins in the peroxovandate-treated intact animal, using antibodies against known tyrosine kinase substrates (2). However, this system has limitations for protein purification and the isolation of novel substrates due to the small size of the animal and organs. For this reason, tyrosine-phosphorylated proteins were isolated from the liver and kidney of peroxovanadate-treated dogs, thus providing hundreds of grams of tissue for bulk isolation of phosphotyrosine-containing proteins by standard biochemical methods and identification by sequence analysis.
Adult dogs were treated by intraperitoneal injection of PBS alone or PBS containing 10 mM sodium vanadate and 100 mM H 2 O 2 as described under "Experimental Procedures." After 20 min, the liver and kidneys were excised and frozen in liquid nitrogen. Extracts were prepared and the proteins separated by SDS-PAGE and analyzed by Western blotting with antiphosphotyrosine antibodies. In both organs the administration of peroxovanadate resulted in tyrosine phosphorylation of many proteins. Similar preparations from control animals (treated with PBS alone) showed only trace levels of tyrosinephosphorylated proteins (Fig. 1) (in previous experiments with mice, treatment with vanadate or H 2 O 2 alone induced only minimal tyrosine phosphorylation in liver and kidney (2)).
Isolation and Identification of Tyrosine-phosphorylated Proteins from Liver and Kidney of Peroxovanadate-treated Dogs-Two procedures (Method A and B) were used for initial tissue fractionation. The first method (Method A) was used to obtain those proteins that reversibly associate with a membrane/particulate fraction of the tissue in a calcium-dependent manner. This procedure was previously used to identify tyrosine-phosphorylated annexin I in A431 cells (11). Method B yielded three separate fractions: cytosolic, alkali-extractable (membrane-associated proteins), and detergent-extractable (intrinsic membrane proteins).
The EDTA-extractable fraction (obtained by Method A) from both liver and kidney were analyzed by SDS-PAGE and found to contain multiple tyrosine-phosphorylated proteins as detected by Western blotting with anti-phosphotyrosine antibodies. The results with kidney extracts are shown in Fig. 2, lane labeled "orig." The proteins present in the EDTA-extractable fractions were then separated by DE52 column chromatography. Fractions were analyzed by SDS-PAGE and Western blotting using anti-phosphotyrosine antibodies (Fig. 2). Many tyrosine-phosphorylated proteins were detected throughout the gradient (Fig. 2, column fractions 5-36). Selected fractions containing tyrosine-phosphorylated proteins were further purified by two procedures: 1) adsorption on anti-phosphotyrosine-agarose and elution with phenyl phosphate and phosphotyrosine, or 2) adsorption on phospholipid-containing vesicles in the presence of calcium and elution with EDTA (see "Experimental Procedures"). The latter procedure is a general method for isolating members of the annexin family of homologous proteins that bind phospholipids in the presence of calcium (12). The samples recovered from each procedure were subjected to SDS-PAGE, transferred to Immobilon-P, and stained with Coomassie Blue. Phosphotyrosine-containing protein bands were excised from the membrane and submitted for sequence analysis in the Protein Chemistry core facility at Vanderbilt.
By these procedures, four tyrosine-phosphorylated proteins (ϳ47, ϳ56, ϳ81, and ϳ180 kDa) were identified. The N-terminal sequence of the 81-kDa protein (isolated from pooled fractions 14 -21, Fig. 2) was found to be PKPINVRVTTXD. This sequence corresponds to the N terminus of ezrin. Ezrin has been previously shown to be tyrosine phosphorylated on at least two tyrosine residues (3,13).
The sequences of a peptide derived from a Lys-C digestion of the ϳ180-kDa protein (isolated from pooled fractions 22-30, Fig.  2) was determined to be (K)ADDPSXYMEV. This peptide occurs in clathrin. Two peptides were sequenced from the 56-kDa protein (isolated from pooled fractions 6 -13, Fig. 2): an Asp-Nderived peptide with the sequence DMTLVQR(D) and a Lys-Cderived peptide with sequence (K)SLYHDIXGD. These peptides correspond to those present in annexin XI. Finally, sequences derived after Lys-C digestion of the 47-kDa protein (also isolated from pooled fractions 6 -13, Fig. 2) were (K)GFGTDEQAIV and (K)LLLAIVGQ, corresponding to sequences present internally and at the C terminus, respectively, of annexin VII.
Immunoprecipitation and Western Blotting to Confirm the Identification of the Tyrosine-phosphorylated Proteins-To confirm the protein identifications derived from sequence analysis, extracts were immunoprecipitated with specific antibodies against clathrin, annexin VII, and annexin XI and their phosphotyrosine content was detected by Western blotting with antibodies to phosphotyrosine. By these procedures we confirmed that all three proteins were tyrosine phosphorylated (Fig. 3). Furthermore, after incubation of the proteins with leukocyte antigen-related protein, protein-tyrosine phosphatase followed by SDS-PAGE and transfer of the proteins to Immobilon-P, all or most of the signal in Western blots with anti-phosphotyrosine was removed, again indicating that the proteins were tyrosine phosphorylated (data not shown).
The four identified proteins were isolated from an EDTAextractable fraction indicating that at least a portion of all four proteins were in some manner associated with the membrane/ particulate fraction in a Ca 2ϩ -dependent manner (clathrin light chain, which is associated with clathrin heavy chain, is a Ca 2ϩ -binding protein (14)), these proteins could also be detected by Western blotting in tissue fractions isolated by other procedures (Method B; see above and under "Experimental Procedures"). The soluble, alkali-extractable (Na 2 CO 3 ) and detergent-extractable (RIPA) fractions obtained using Method B each contained many tyrosine-phosphorylated proteins (Fig.   FIG. 1. Peroxovanadate induces tyrosine phosphorylation of proteins in dog liver and kidney. Solutions of PBS or PBS containing 10 mM sodium vanadate (Na 3 VO 4 ) and 100 mM hydrogen peroxide (H 2 O 2 ) were administered to adult dogs by intraperitoneal injection at a dose of 5 ml/kg body weight. The dogs were sacrificed after 20 min of treatment. The organs were excised and processed as described under "Experimental Procedures." Aliquots (10 l) of the clarified homogenate (Method B) were resolved by SDS-PAGE, transferred to Immobilon-P membranes, and Western blotted with antibodies to phosphotyrosine (RC20H). 4A). The distribution of the four identified proteins among the fractions examined was as expected. Both annexin XI and annexin VII were predominantly in the soluble fraction since the homogenization buffer contained EDTA that would chelate Ca 2ϩ and dissociate the annexins from membranes (Fig. 4, B and C). In addition, an annexin XI immunoreactive band (ϳ42-kDa) was detected in the Na 2 CO 3 fraction (Fig. 4B). However, the protein does not appear to be tyrosine phosphorylated and its identity is not known. Clathrin, as shown by Keen (15), was predominantly in the Na 2 CO 3 fraction (Fig. 4D). Although the major portion of the cytoskeletal protein ezrin was detected in the soluble fraction, some could be detected in all fractions examined (Fig. 4E).
Full-length Annexin XI and Annexin VII Are Absent in Rodent Kidney-In a preliminary experiment, prior to a search for possible physiological ligands that might induce the tyrosine phosphorylation of annexin VII and annexin XI, a comparative examination for the presence of these proteins in liver and kidney tissue from rat, mouse, dog, and human was performed. EDTA-extractable proteins (Method A) from each tissue were analyzed by SDS-PAGE and Western blotted with anti-annexin XI (Towle antibody) or anti-annexin VII (Pollard antibody).
Liver samples from all four species showed full-length annexin XI (Fig. 5A). However, the kidney samples from the four species examined showed marked differences in expression of annexin XI proteins (Fig. 5A). In the rat and mouse kidney samples no trace of full-length annexin XI could be detected (Fig. 5A). The Western blot of the human kidney yielded a weak annexin XI signal (Fig. 5A).
In our screen of kidney and liver from rat, mouse, dog, and human, the 47-kDa isoform of annexin VII was present in liver of all species (Fig. 5B). However, although the 47-kDa form was present in human and dog kidney, we could not detect annexin VII in the kidney of rat and mouse (Fig. 5B). Thus, neither the 47-kDa annexin VII nor the 56-kDa annexin XI could be detected in rodent kidney whereas both were expressed in liver of all species examined. sponse to treatment of dog with peroxovanadate, a cell culture system was sought in which the effect of known mitogenic stimuli on the tyrosine phosphorylation state of these proteins could be studied.

Annexin VII and Annexin XI Are Tyrosine Phosphorylated in
Mizutani et al. (16) previously found that annexin XI was widely distributed in rat tissues except in the kidney. In addition, they noted that annexin XI was particularly rich in homogenates of the rat aorta (16). Furthermore, we observed, in immunostained sections of dog kidney, that annexin XI was highly expressed in arterial smooth muscle cells. 2 Given the combination of these findings, we examined cultured rat VSMC by Western blotting with anti-annexin VII and XI antibodies and were able to confirm the presence of these proteins (data not shown).
Treatment of confluent rat VSMC with PDGF-BB (a major mitogenic factor for cultured rat VSMC (17,18)) or peroxovanadate resulted in an increase in tyrosine phosphorylation of annexin VII and XI as compared with control cells (Fig. 6). In addition, co-addition of peroxovanadate and PDGF-BB resulted in a synergistic increase in tyrosine phosphorylation of annexin VII and XI (Fig. 6). Treatment of rat VSMC with EGF, fibroblast growth factor-basic, hepatocyte growth factor, insulin, 12-O-tetradecanoylphorbol-13-acetate, lysophosphatidic acid, or angiotensin II did not result in detectable tyrosine phosphorylation of annexin VII or XI (data not shown).
An additional protein of molecular mass ϳ32-kDa showed a strong tyrosine phosphorylation signal in response to treatment of cells with PDGF-BB (Fig. 6). The protein was tentatively identified as annexin II by Western blotting and was expressed at levels ϳ10 -20 times greater than annexin VII and XI (as judged by Coomassie Blue staining; data not shown).

PDGR-R, CADTK/Pyk-2, Src Kinase, and EGF-R Phosphorylate Annexin VII and Annexin XI on Tyrosine Residues-
Since annexin VII and XI were tyrosine phosphorylated in rat VSMC in response to PDGF-BB, it was of interest to determine which individual protein tyrosine kinase might be responsible and to establish that the annexins are indeed substrates for tyrosine kinases. In vitro kinase assays were performed with purified annexin VII and XI and selected kinases. Incubation of purified annexin VII and XI with PDGF-R, CADTK/Pyk-2, Src kinase, or EGF-R in the presence of 5 mM ATP resulted in an increase in phosphorylation of annexin VII and XI on tyrosine residues as indicated by Western blotting with anti-phosphotyrosine antibodies (Fig. 7). Purified annexin VII and XI were shown to be free of contaminating kinase activity by incubation with 5 mM ATP followed by SDS-PAGE and Western blotting with anti-phosphotyrosine antibodies. We were unable to detect the phosphorylation of annexin VII or XI by either JAK2 or Fak kinase in vitro by similar methods (results not shown).

DISCUSSION
In the present study we have identified in the liver and kidney of peroxovanadate-treated dogs four tyrosine-phosphorylated proteins. Two of these proteins, annexin XI and annexin VII, have not previously been shown to be tyrosine phosphorylated in any model system and the other two, clathrin and ezrin, have not been shown to be tyrosine phosphorylated in the intact animal, although they have been detected in specialized cell systems (3)(4)(5)19).
Annexin VII and XI are related members of a family of homologous proteins that bind negatively charged phospholipids in the presence of calcium. The proposed physiological functions of the annexins include phospholipase A 2 inhibition, exocytosis, membrane trafficking and binding, ion channel activity, and signaling. Although the proposed functions are diverse, they are all related to the phospholipid/membrane binding properties of the proteins.
Annexins are ubiquitously expressed in a variety of tissues and cell types of eukaryotes (20,21). Family members contain conserved C-terminal domains with conserved repeats of 70 amino acids each of which forms the "endonexin-fold"-domain responsible for Ca 2ϩ and phospholipid binding. The N-terminal domains are considered unique among annexins since they vary in sequence and length and presumably determine functional diversity among members (20 -22). Like eps 8 and eps 15, annexin VII and XI are tyrosine phosphorylated, but contain neither SH2 nor phosphotyrosine-binding domains. Annexin VII and XI both have large, hydrophobic, structurally related N-terminal regions very rich in glycine, tyrosine, and proline residues (20,21).
The only annexins previously shown to be tyrosine phosphorylated are annexin I and II (20,21). Annexin I was shown to be tyrosine phosphorylated in A-431 cells by the EGF-R in this laboratory (11). Tyrosine phosphorylation of annexin II was detected in cells transformed by pp60 src (23). In addition, serine phosphorylation has also been reported for a number of annexins. Despite these findings, the physiological role of phosphorylation of annexins (as well as of annexins themselves) is still largely unknown. Tyrosine phosphorylation may modulate one or more of the proposed physiological functions for this class of proteins.
FIG. 6. Annexin VII and annexin XI are tyrosine phosphorylated in rat VSMC in response to treatment with PDGF-BB, peroxovanadate, and PDGF-BB plus peroxovanadate. Rat VSMC were treated for 5, 10, and 20 min with PDGF-BB (50 ng/ml), peroxovanadate (10 M vanadate, 50 M H 2 O 2 ), or peroxovanadate plus PDGF-BB. Lysates from control and treated rat VSMC were incubated with phospholipid vesicles and 10 mM Ca 2ϩ for 45 min prior to ultracentrifugation to pellet the phospholipid vesicles and associated proteins that include annexin VII and XI. Samples were then analyzed by SDS-PAGE followed by transfer to Immobilon-P and Western blotting with anti-phosphotyrosine (PY) antibodies. Results are representative of at least three experiments; results of all experiments were similar.
FIG. 7. PDGF-R, CADTK/Pyk-2, Src kinase, and EGF-R are able to tyrosine phosphorylate annexin VII and XI in vitro. Separately, a purified mixture of annexin VII and XI, or of PDGF-R, CADTK/ Pyk-2, EGF-R, or Src kinase, or a mixture of annexin VII and XI and individual kinases were incubated at room temperature in buffer containing 20 mM Hepes, pH 7.4, 20 mM MgCl 2 , 1 mM MnCl 2 , 100 M VO 4 , 1 mM dithiothreitol, and 5 mM ATP. After 15 min, the reaction mixtures were centrifuged and the supernatants analyzed by SDS-PAGE followed by transfer to Immobilon-P and Western blotting with antiphosphotyrosine antibodies. The positions of annexin VII and XI, as determined by Western blotting with anti-annexin VII and XI antibodies (data not shown), are indicated with an asterisk (*) to the left of each phosphotyrosine-containing band (the anti-phosphotyrosine blots and anti-protein blots coincide).
shown to occur in two isoforms: one full-length with a molecular mass of 51-kDa and one lacking exon 6 with molecular mass of 47-kDa. In our studies, only the 47-kDa isoform was observed. Annexin VII has been reported to be involved in ion channel activity, membrane fusion, aggregation, secretion, and Ca 2ϩ /GTP-regulated exocytosis (20,21,24,25).
Annexin XI was independently first identified in 1992 by Towle and Treadwell (6) from a bovine chondrocyte cDNA library and by Tukumitsu et al. (26) from rabbit lung as a calcyclin-associated protein. The N-terminal domain has been reported to be responsible for nuclear localization of the protein during rat embryonic development (16,27,28) and for antigenic stimulation in human autoimmune disease (7). Mizutani et al. (29) detected serine and threonine phosphorylation, but not tyrosine phosphorylation, of annexin XI in pp60 src -transformed cells.
We were unable to detect either annexin VII or XI in kidney of mouse and rat although both were detectable in the liver of all species examined (mouse, rat, dog, and human) and in the kidney of dog and human (Fig. 5, A and B). These observations do not appear to be an artifact due to species differences in antibody reactivity since Mizutani et al. (16) were also unable to detect annexin XI in rat kidney with a different polyclonal antibody. The reasons for these species differences are unknown.
The two other phosphotyrosine-containing proteins we isolated, clathrin and ezrin, both had previously been identified as tyrosine kinase substrates in various cell systems (3-5, 13, 19, 30 -33). The current study confirms the tyrosine phosphorylation of both proteins in liver and kidney of intact animals. Wilde and Brodsky (5) have suggested that tyrosine phosphorylation of clathrin may affect only subpopulations of clathrin such as those found on endosomes, at sites of cell adhesion, or those involved in formation of the pentagon lattice. Phosphorylation of ezrin appears to be critical for its cellular distribution by the unmasking of actin-binding sites (34 -36). Recently, tyrosine phosphorylation of ezrin on Y353 has been shown to be involved in cell survival by activation of the phosphatidylinositol 3-kinase/Akt pathway (33). Tyrosine-phosphorylated ezrin (Y353) was reported to interact with an SH2 domain on the p85 subunit of phosphatidylinositol 3-kinase and thereby protect against apoptosis by activating the phosphatidylinositol 3-kinase/Akt pathway. When Y353 in ezrin was mutated to F, cells underwent apoptosis (33).
Our ability to isolate annexin VII and XI as tyrosine-phosphorylated proteins following the administration of peroxovanadate to dog indicates that there must be a kinase(s) that phosphorylates these proteins and a phosphatase(s) that dephosphorylates them. Since annexin VII and XI had not previously been reported to be tyrosine phosphorylated in any cell system and we could detect both proteins in rat VSMC, we examined possible physiological stimuli that might induce their tyrosine phosphorylation in these cells.
Using rat VSMC we were able to show enhanced tyrosine phosphorylation of annexin VII and XI in response to a physiological stimulus, PDGF-BB (Fig. 6). Furthermore, as might be expected, peroxovanadate increased the response to PDGF (Fig. 6). Presumably, the results with the mixture of PDGF-BB and peroxovanadate duplicate those in peroxovanadate-treated dog in which endogenous ligands are constitutively present. In support of these findings, in vitro kinase assays showed that PDGF-R, CADTK/Pyk-2, Src kinase, and EGF-R were able to phosphorylate the annexins (Fig. 7). Both CADTK/Pyk-2 and Src kinases are downstream of the PDGF-R and are activated by PDGF (37-39). Furthermore, CADTK/Pyk-2 is associated with the cytoskeleton and has been shown to be important for tyrosine kinase signaling in VSMC in response to PDGF (37). Our results using a mixture of annexin VII and XI alone indicate that a trace of phosphotyrosine was present in the annexins as isolated. However, ATP was required for the appearance of phosphotyrosine following incubation of substrate with kinase (Fig. 7).
Finally, as can be concluded from data presented in Figs. 2 and 4, there are many other, as yet unidentified, phosphotyrosine proteins in our screen, a few of which may be the same as those previously identified in the mouse (2). Given the number of known substrates for tyrosine kinases, it is surprising to us that of the first four major proteins identified in our screen, two had not previously been identified as tyrosine phosphorylated, again suggesting that many more phosphotyrosine-containing proteins remain to be identified.