The Tumor Suppressor PTEN Is Phosphorylated by the Protein Kinase CK2 at Its C Terminus IMPLICATIONS FOR PTEN STABILITY TO PROTEASOME-MEDIATED DEGRADATION

The tumor suppressor phosphatase PTEN regulates cell migration, growth, and survival by dephosphorylat-ing phosphatidylinositol second messengers and signaling phosphoproteins. PTEN possesses a C-terminal non-catalytic regulatory domain that contains multiple putative phosphorylation sites, which could play an important role in the control of its biological activity. The protein kinase CK2 phosphorylated, in a constitutive manner, a cluster of Ser/Thr residues located at the PTEN C terminus. PTEN-phosphorylated defective mutants showed decreased stability in comparison with wild type PTEN and were more rapidly degraded by the proteasome. Inhibition of PTEN phosphorylation by the CK2 inhibitor 5,6-dichloro-1- b - D -ribofuranosyl-benzim-idazole also diminished the PTEN protein content. Our results support the notion that proper phosphorylation of PTEN by CK2 is important for PTEN protein stability to proteasome-mediated degradation. The tumor suppressor gene PTEN

The tumor suppressor gene PTEN (also named as MMAC1 or TEP-1) (1-3) encodes a phosphatase with enzymatic activity toward both protein substrates and the lipid second messenger, phosphatidylinositol-3,4,5-triphosphate (4 -6). PTEN regulates distinct signal transduction pathways, including the phosphatidylinositol 3-kinase/ protein kinase B cell survival-and integrin-triggered signaling pathways (for recent reviews, see Ref. [7][8][9][10]. Structurally, PTEN protein is composed of an Nterminal dual specificity phosphatase-like enzyme domain and a C-terminal regulatory domain, which binds to phospholipid membranes (11). Mutations in the PTEN gene are present in a great number of tumors, as well as in the germ line cells of patients with several inherited cancer syndromes (reviewed in Refs. 12 and 13). The importance of PTEN catalytic activity in its tumor suppressor function is underscored by the fact that the majority of PTEN missense mutations detected in tumor specimens target the phosphatase domain and cause a loss in PTEN phosphatase activity. In addition, a large number of PTEN nonsense or frame-shift mutations found in tumors are targeted to the C-terminal domain of the protein, suggesting an important role for this domain in the regulation of the PTEN tumor suppressor activity. In this regard, the C-terminal re-gion of PTEN has been shown to be important in the regulation of the stability and half-life of the molecule (14,15). Also, the C-terminal PTEN amino acid sequence possesses a putative PDZ binding motif, which has been proposed to modulate PTEN functions by association to PDZ domain-containing proteins (16 -19). Finally, the C-terminal PTEN domain is rich in putative phosphorylation sites, and phosphorylation of the PTEN C terminus has been recently reported to affect PTEN protein stability and function (20); however, the kinase responsible for such phosphorylation remains unidentified.
In this report, we demonstrate that the C terminus of the tumor suppressor PTEN is constitutively phosphorylated by CK2 at several residues in intact cells. An insight is provided on the putative role of phosphorylation as a regulatory mechanism of PTEN biological activity by modulating its stability to proteasome-mediated degradation.

MATERIALS AND METHODS
Plasmids, Antibodies, and Reagents-The cDNA encoding full-length human PTEN was obtained by reverse transcriptase-polymerase chain reaction amplification of poly(A)ϩ mRNA from MCF-7 cells using primers flanking the human PTEN coding region (1)(2)(3), and the PTEN nucleotide sequence was confirmed by DNA sequencing. pRK5 mammalian expression vectors were made by polymerase chain reaction amplification of PTEN or GST cDNAs and subcloning. For the construction of bacteria expression plasmids encoding GST-PTEN fusion proteins, PTEN cDNAs were cloned into the pGEX-4T1 or pGEX-5X1 expression vectors. PTEN C terminus truncation mutants and amino acid substitution mutants were made by polymerase chain reaction oligonucleotide site-directed mutagenesis, and mutations were confirmed by DNA sequencing. The 12CA5 anti-HA and the 4G10 antiphosphotyrosine mAb have been described previously (34). The rabbit polyclonal anti-PTEN CS486 and BT166 antibodies were from R. Parsons and C. Eng, respectively (35,36). The anti-GST mAb (clone 22.A) * This work was supported by grants from the Ministerio de Educación y Cultura (PB96-0278) and the Generalitat Valenciana (GV-C-VS-20-047-96). 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  was obtained in our laboratory and is directed against Schistosoma japonicum GST. Horseradish peroxidase-conjugated goat anti-mouse secondary antibody was from Promega (Madison, WI). The following activating agents were used: phorbol 12-myristate 13-acetate (Sigma Chemical), 10 ng/ml, 30 min; EGF (Life Technologies, Inc.), 50 ng/ml, 10 min; insulin (Sigma), 1 g/ml, 10 min; okadaic acid (Roche Molecular Biochemicals), 1 M, 30 min; sodium pervanadate (Sigma), 10 M, 30 min; A23187 calcium ionophore (Roche Molecular Biochemicals), 10 M, 30 min. All incubations were made at 37°C. Human recombinant CK2 holoenzime was from Roche Molecular Biochemicals. The following CK2 modulators were used: spermine (Sigma), 1 mM; heparine (Sigma), 1 g/l; 5,6-dichloro-1-␤-D-ribofuranosyl-benzimidazole (DRB) (BIOMOL Research Laboratories Inc., Plymouth Meeting, PA), 100 M. The MEK1/2 inhibitor PD98059 (New England Biolabs Inc., Beverly, MA) was used at 10 M, and dibutyryl cAMP (Roche Molecular Biochemicals) was used at 50 M. The proteasome inhibitor MG132 (Peptide Institute Inc., Osaka, Japan) was used at 50 M, leupeptin (Peptide Institute Inc.) was used at 0.1 mM, and NH 4 Cl was used at 20 mM. For the experiments of inhibition of degradation, chases were performed in the continuous presence of the inhibitors.
Cell Culture, Transfections, and Isotope Cell Labeling-Human breast carcinoma MCF-7 cells were grown in RPMI medium supplemented with 10% heat-inactivated FCS (Life Technologies, Inc.). COS-7 cells were grown in Dulbecco's minimal essential medium containing high glucose (4.5 g/l) supplemented with 5% heat-inactivated FCS. COS-7 cells were transfected with the DEAE-dextran method and processed 48 -72 h after transfection. For [ 35 S]methionine labeling, cells were incubated for 1 h in methionine/cysteine-free Dulbecco's minimal essential medium containing 2% dialyzed FCS and then labeled with [ 35 S]methionine/cysteine (50 Ci/ml) for 2 h (pulse). Chases were performed by substituting the [ 35 S]methionine/cysteine-containing medium for Dulbecco's minimal essential medium, 2% FCS containing an excess of cold methionine (1 mM). For 32 P labeling, cells were incubated in phosphate-free medium with 2% FCS for 1 h and then labeled with [ 32 P]orthophosphate (50 Ci/ml) for 4 h. Cell lysis, immunoprecipitation, and immunoblot were performed as described (34). For protein turnover measurements, HA-PTEN radiolabeled bands, resolved by SDS-PAGE after immunoprecipitation with the anti-HA 12CA5 mAb, were quantitatively analyzed using a phosphorImager. Results were plotted, and linear fits were performed (r Ͼ 0.97).

PTEN Is a Phosphoprotein under Standard Cell Growth
Conditions-To study the putative role of post-translational modifications in the regulation of PTEN biological functions, the possibility that PTEN could be phosphorylated in intact cells was investigated. MCF-7 cells, known to carry a wild type PTEN gene (1), were labeled with 32 P, and cell lysates were subjected to immunoprecipitation using two distinct anti-PTEN antibodies (Fig. 1A). As shown, both anti-PTEN antibod-ies immunoprecipitated a radiolabeled protein that migrated at about 55 kDa (Fig. 1A, lanes 2 and 3), as expected for PTEN (36), whereas no signal was detected using a control serum (Fig. 1A, lane 1). Next, the region of PTEN that is phoshorylated in intact cells was analyzed using recombinant forms of this molecule. 32 P labeling was carried out in COS-7 cells transfected with HA-tagged PTEN wild type (residues 1-403) or C-terminal truncated forms (residues 1-343, 1-369, or 1-386), followed by immunoprecipitation with the anti-HA To ascertain the identity of the kinase(s) involved in the phosphorylation of PTEN, 32 P-labeled COS-7 cells were transfected with HA-PTEN and incubated in the presence of stimulators of several signal transduction pathways, including okadaic acid, sodium pervanadate, A23187 calcium ionophore, insulin, phorbol 12-myristate 13-acetate, and EGF. As shown, none of these stimuli significantly modified the phosphorylation level of HA-PTEN ( Fig. 2A, lanes 1-7). We also tested the possibility that PTEN could be phosphorylated on tyrosine residues. For these experiments, a GST-PTEN fusion protein was overexpressed in COS-7 cells, precipitated in one step using glutathione-Sepharose, and subjected to immunoblot with the anti-phosphotyrosine 4G10 mAb (Fig. 2B). No reactivity of the 4G10 mAb toward GST-PTEN was detected (Fig. 2B,  upper panel, lanes 1 and 2), whereas a strong signal was observed toward GST-ERK2 (included as a positive control of tyrosine phosphorylation) upon EGF cell stimulation (Fig. 2B,   FIG. 1. PTEN is a phosphoprotein under normal Fig. 2B shows the expression of the GST fusion proteins. The phosphorylation of GST-PTEN, precipitated from transfected COS-7 cells labeled with 32 P, is also shown as a control (Fig. 2C, lane 2). Together, these results indicate that the C-terminal region of PTEN is constitutively phosphorylated on serine and/or threonine residues.
In the left panel, cells were labeled with 32 P, and cell lysates were precipitated with glutathione-Sepharose. In the right panel, 10 g of total cell lysates were loaded and subjected to immunoblot with an anti-GST mAb. All samples were resolved by 10% SDS-PAGE under reducing conditions followed by autoradiography (A and C, left panel) or immunoblot (B and C, right panel). of phosphate/mol of protein after extensive in vitro phosphorylation by CK2.
To test the involvement of CK2 in the phosphorylation of PTEN observed in intact cells, crude extracts of COS-7 cells were incubated, in the presence of inhibitors or activators of CK2, with the GST-PTEN fusion proteins purified from bacteria, and in vitro kinase assays were performed. In keeping with the results obtained after in vitro phosphorylation by CK2, GST-PTEN-(1-403) and GST-PTEN-(202-403) were specifically phosphorylated by the kinase activity present in the cell extracts, whereas GST-PTEN-(1-202) or GST alone were not (Fig. 3C, lane 1, and data not shown). Remarkably, this phosphorylation was substantially increased in the presence of spermine, an activator of CK2 (Fig. 3C). On the other hand, in the presence of any of two CK2 inhibitors, DRB or heparin, the phosphorylation of GST-PTEN was inhibited (Fig. 3C). No effect was observed in the presence of other protein kinase effectors, such as the MEK1/2 inhibitor PD98059 or the protein kinase A activator dibutyryl cAMP (data not shown). These results sustain the notion that the kinase activity that phosphorylates PTEN in vivo is attributable to CK2. Interestingly, treatment of cells with DRB resulted in a decrease in both the in vivo phospholabeling and the protein amount of PTEN, suggesting that DRB cell treatment could diminish PTEN phosphorylation and/or PTEN protein content (data not shown; and see Fig. 5C).
PTEN phosphorylation was also analyzed on 32 P-labeled COS-7 cells overexpressing the distinct HA-PTEN CK2 phosphorylation mutants (Fig. 4). A significant reduction in the in vivo PTEN phosphorylation was observed on the S370A and S385A mutants (Fig. 4, upper panel, lanes 2 and 6, respectively). Also, a consistently reduced labeling was observed on the S380A mutant (Fig. 4, upper panel, lane 3), whereas the effect on the phosphorylation of the individual T382A and T383A mutants was less manifest (Fig. 4, upper panel, lanes 4 and 5,  respectively). Because these results suggested the involvement of multiple residues on PTEN phosphorylation in intact cells, combined mutations were generated and tested for in vivo phosphorylation. The phosphorylation of the double mutant S370A/S385A (DMA) was nearly absent (Fig. 4, upper panel,  lanes 8 and 12), and the phosphorylation of the triple mutant S380A/T382A/T383A (TMA) was markedly decreased (Fig. 4,  upper panel, lane 11), indicating an additive contribution of all these residues to PTEN phosphorylation in intact cells. The phosphorylation of a double phosphomimetic S370E/S385E (DME) mutant was also analyzed. As shown, the S370E/S385E mutant was hyperphosphorylated in comparison with the S370A/S385A mutant (Fig. 4, upper panel, lane 13), suggesting that phosphorylation of these two PTEN residues might prime for additional phosphorylation of nearby residues. Finally, the putative in vivo phosphorylation of the Thr-401 residue, located within the PDZ binding consensus motif of PTEN, was also tested. As shown, the phosphorylation level of the T401A mutant was almost indistinguishable from the wild type PTEN (Fig. 4, upper panel, lane 7). Together, these results support the notion that CK2 constitutively phosphorylates in COS-7 cells a cluster of Ser/Thr residues at the C terminus of PTEN.
Phosphorylation of PTEN C Terminus Is Important for PTEN Protein Stability to Proteasome-mediated Degradation-To analyze the effect of PTEN phosphorylation on the stability of the molecule, pulse-chase experiments were performed on [ 35 S]methionine-labeled COS-7 cells transfected with HA-PTEN wild type or phosphorylation mutants. The kinetics of degradation of HA-PTEN wild type was two to three times slower than that showed by the distinct Ser/Thr to Ala phosphorylation mutants (Fig. 5A). Comparative plots of degradation of wild type PTEN with the S370A, T380A, T382A, T383A, S385A, and S370A/S385A (DMA) mutants are shown (Fig. 5A). These results suggested that the presence of the phosphorylated residues could be important for PTEN stability. To test this possibility, the degradation of the phosphomimetic S370E/S385E (DME) mutant was analyzed. Interestingly, the stability of the S370E/S385E mutant was almost identical to that obtained with the wild type PTEN (Fig. 5A,  right panel), reinforcing the hypothesis of a role of phosphorylation of PTEN in protein stabilization.
To ascertain the degradative pathway of PTEN under these conditions, transfected COS-7 cells were pulsed with [ 35 S]methionine, and chases were performed in the presence of the proteasome inhibitor MG132 or the lysosome inhibitors leupeptin ϩ NH 4 Cl. As shown, the proteasome inhibitor MG132 inhibited the degradation of both PTEN wild type and S370A/ S385A mutant (Fig. 5B, lanes 3 and 7), whereas no inhibition was observed in the presence of leupeptin ϩ NH 4 Cl (Fig. 5B,  lanes 4 and 8). The effect of MG132 on the degradation of the S370A/S385A mutant was consistently less pronounced than that observed for the wild type PTEN. Finally, experiments were carried out to elucidate whether the effect of DRB cell treatment in the decrease of 32 P-radiolabeled PTEN detected in our studies was, in fact, because of the inhibition of CK2. Cells were left untreated or were preincubated with the proteasome inhibitor MG132 to prevent PTEN degradation, were then labeled with 32 P or [ 35 S]methionine in the presence of DRB, and the amount of radiolabeled PTEN was detected as above. Upon DRB cell treatment, both the 32 P phospholabeling and the amount of [ 35 S]methionine-labeled PTEN were decreased (Fig.  5C, lane 2). Remarkably, in the presence of DRB plus MG132 the phosphorylation of PTEN was decreased without affecting the amount of PTEN protein detected by [ 35 S]methionine labeling (Fig. 5C, lane 3). Together, these results indicate that defective PTEN phosphorylation by CK2 accelerates the proteasome-mediated degradation of PTEN. DISCUSSION On the basis that many CK2 substrates are oncogene products or tumor suppressor proteins as well as key signal transduction proteins, a role for CK2 in the regulation of cell growthand cell cycle-related processes has been proposed (22,23). In this report, we provide evidence that the tumor suppressor phosphatase PTEN is phosphorylated by CK2 both in vitro and in intact cells. CK2 phosphorylation sites in PTEN were located within a C-terminal cluster of Ser/Thr residues, which were found to be important for PTEN stability to proteasome-mediated degradation. The CK2 phosphorylation sites were identified as the PTEN residues Ser-370, Ser-380, Thr-382, Thr-383, and Ser-385. Because some of these sites can be generated after previous phosphorylation of nearby residues, and because the phosphomimetic S370E/S385E mutant is phosphorylated more efficiently than the S370A/S385A mutant, hierarchic CK2 multisite phosphorylation of this region of PTEN is likely to exist (37). Several findings support the notion that CK2 is the major kinase involved in the phosphorylation of PTEN in intact cells. First, cell treatment with a panel of different stimuli did not substantially affect the phosphorylation of PTEN, suggesting that such phosphorylation is constitutive and messenger-independent, as is known for most CK2-mediated phosphorylations. Second, tyrosine phosphorylation was not detected on PTEN, suggesting that tyrosine kinases do not directly phosphorylate PTEN. Third, phosphorylation of PTEN by the kinase activity present in COS-7 cell extracts was activated or inhibited by CK2 activators or inhibitors, respectively. Fourth, phosphorylation of PTEN in intact COS-7 cells was inhibited by cell treatment with the CK2 inhibitor DRB. Fifth, mutation to Ala of the PTEN CK2-phosphorylation sites, as well as deletion of the C-terminal portion of PTEN including these sites, resulted in an almost complete lack of PTEN phosphorylation in intact cells. However, the possibility cannot be ruled out that other kinase(s) not monitored in our assays, using overexpressed HA-PTEN in COS-7 cells, may contribute to the phosphorylation of endogenous PTEN in vivo. Remarkably, the CK2 phosphorylation sites in PTEN are conserved in species from mammals to Xenopus laevis, and clusters of putative CK2 phosphorylation sites are also present at the C terminus of PTEN from Drosophila melanogaster, Caenorhabditis elegans, and Saccharomyces cerevisiae.
We found that PTEN suffers a rapid degradation in COS-7 cells, which was inhibited by the proteasome inhibitor MG132 but not by lysosome inhibitors, indicating that the turnover of PTEN in cultured COS-7 cells depends mainly on proteasomemediated degradation. The involvement of the proteasome in the degradation and regulation of the functions of short-lived proteins, including oncoproteins, tumor suppressors, and cell cycle proteins has been described extensively (for reviews, see Refs. 38 -40). Thus, regulated degradation of PTEN via the proteasome could be envisaged as a major physiological mechanism that controls the amount of PTEN in specific cell types and tissues. In this context, it has been reported recently that C-terminal truncation mutants of PTEN have shorter half-lives than the wild type molecule (14,15). In addition, we show that Ser/Thr to Ala amino acid mutations at the PTEN C terminus, leading to defective phosphorylation by CK2, accelerate the proteasome-mediated degradation of PTEN. However, a double Ser to Glu mutation (the S370E/S385E mutant), that mimicked constitutive CK2 phosphorylation of PTEN at these sites showed a degradation rate comparable with the wild type molecule. Furthermore, cell treatment with the CK2 inhibitor DRB, which inhibited PTEN constitutive phosphorylation, resulted in a diminution in PTEN protein content but only in the absence of the proteasome inhibitor MG132. While this manuscript was in preparation, Vazquez et al. (20) also showed that phosphorylation of some of the PTEN C-terminal residues an- alyzed in our study regulate protein stability as well as PTENmediated G1 cell cycle arrest. Thus, constitutive phosphorylation of PTEN by CK2 may play a key role in the regulation of PTEN biological functions upon normal cell growth conditions.
Our findings suggest that phosphorylation of the PTEN C terminus itself does not provide enough PTEN protein stability (see below) but rather could be important to acquire the proper, stable conformation of the PTEN C-terminal tail. For instance, it is well documented that phosphorylation of the C terminus of IB␣ by CK2 affects the stability and constitutive turnover of this NF-B regulator (41)(42)(43)(44). In addition, the possibility exists that phosphorylation of the C terminus of PTEN could modulate its subcellular location and function by affecting its association with regulatory and/or stabilizing molecules. In this regard, deletion of the C-terminal PDZ binding motif of PTEN results in the impairment of some PTEN cellular functions (16). Also, PTEN has been found to associate, through its PDZ binding motif, with PDZ domain-containing proteins, including hMAST205 and MAGUK scaffolding proteins (17)(18)(19). Interestingly, MAGUKs might regulate the functions of PTEN through the assembly and stabilization of multiprotein signaling complexes at specific subcellular compartments, as known for other related PDZ-containing proteins (45). It has been suggested that phosphorylation of the Thr-401 residue within the PDZ binding motif of PTEN could regulate its binding to PDZcontaining proteins (17). However, our results suggest that the PTEN Thr-401 residue is not phosphorylated in vivo. Mutations that did not abrogate PTEN phosphorylation, but are predicted to disrupt the association with PDZ-containing proteins (PTEN-(1-386) and T401A mutant), also diminished the stability of PTEN (data not shown), favoring the hypothesis of a putative role for phosphorylation by CK2 in PTEN stabilization through protein association. Frame-shift mutations at the end of the PTEN coding region have been found in tumors, indicating that such mutations could produce C-terminal truncated PTEN molecules with altered tumor suppressor function (13). Furthermore, by using anti-PTEN antibodies, variable expression levels of PTEN protein are detected in tumor specimens as compared with normal tissues (46 -49), suggesting that mechanisms other than loss of the PTEN gene may account for alterations on PTEN protein expression in some tumors. The results presented here provide evidence that the CK2 phosphorylation sites contained within the C terminus of PTEN play an important role in its stabilization and turnover, in a process mediated by the proteasome. Our findings support a model in which changes in the phosphorylation status of the C terminus of PTEN might alter the basal expression levels of this tumor suppressor. In this regard, we have not detected dephosphorylation of PTEN after incubation with crude cell extracts, and PTEN phosphorylation status did not change upon cell incubation with the PP2A inhibitor, okadaic acid (Fig.  2). 2 The possibility exists that dephosphorylation of PTEN takes place only under particular conditions of activation of an unidentified phosphatase and/or under specific subcellular location of PTEN. Further work will be necessary to ascertain whether regulated dephosphorylation events could affect PTEN phosphorylation and function in vivo.