Nuclear Localization of PTEN Is Regulated by Ca Through a Tyrosil

Research Article

Nuclear Localization of PTEN Is Regulated by Ca2+ through a Tyrosil Phosphorylation–Independent Conformational Modification in Major Vault Protein

Takeo Minaguchi,1,2 Kristin A. Waite,1,2 and Charis Eng1,2,3,4,5

1Genomic Medicine Institute, 2Lerner Research Institute, and 3Taussig Cancer Center, Cleveland Clinic Foundation; and 4Department of Genetics and 5CASE Comprehensive Cancer Center, Case Western Reserve University School of Medicine, Cleveland, Ohio

Abstract breast, endometrium, prostate, and glioblastoma. Germ line PTEN We have recently shown in MCF-7 cells that nuclear mutations of predispose to dominantly inherited Cowden phosphatase and tensin homologue deleted on chromosome and Bannayan-Riley-Ruvalcaba syndromes, which are characterized 10 (PTEN) down-regulates phosphorylation of p44/42 and by the formation of multiple benign tumors and an increased risk of malignant and benign breast, thyroid, and endometrial tumors cyclin D1 and induces G1 cell cycle arrest, whereas cytoplasmic PTENdown-regulates phosphorylation of Akt, up-regulates (1). PTEN protein is composed of an NH2-terminal phosphatase p27, and induces apoptosis. In this manner, nucleocytoplasmic domain containing the consensus phosphatase motif, a C2 domain partitioning of PTENseems to differentially regulate the cell that binds to phospholipid membranes, and a COOH-terminal tail cycle and apoptosis. We have also reported that PTENhas containing a PDZ-binding domain (2, 3). PTEN is a dual-specificity nuclear localization signal–like sequences required for major phosphatase: PTEN’s protein phosphatase activity (PPA) dephos- vault protein (MVP)–mediated nuclear translocation. To date, phorylates tyrosine-, serine-, and threonine-phosphorylated pep- in vitro several other proteins are reported to interact with MVP, tides and dephosphorylates tyrosine in focal adhesion in vivo including extracellular signal-regulated kinases and steroid kinase ; PTEN’s lipid phosphatase activity dephosphorylates receptors, suggesting that MVP is likely to be involved in phosphatidylinositol-(3,4,5)-triphosphate, a lipid second messenger signal transduction through nucleocytoplasmic transport. and a regulator of the phosphatidylinositol 3-kinase/Akt pathway However, the exact mechanism of MVP-mediated nucleocyto- (1). PTEN’s PPA also regulates the mitogen-activated protein kinase plasmic shuttling remains elusive. PTENreportedly interacts (MAPK) pathway (4). The many important cellular functions of in vitro with the EF hand–like motif of MVP in a Ca2+- PTEN include a wide range of biological processes (i.e., G1 cell cycle dependent manner. The current study shows that small arrest, apoptosis, cell migration, spreading, polarity, chemotaxis, interfering RNA–mediated MVP silencing decreases the and focal adhesion formation; refs. 1, 5). nuclear localization of PTENand increases phosphorylation Several known tumor suppressors have been shown to undergo of nuclear p44/42. We show in situ that PTEN-MVP interaction nuclear-cytoplasmic shuttling, such as APC, BRCA1, NF2, p53, is Ca2+ dependent and is abolished by Mg2+.Nuclear and p27. In addition to these, immunohistochemical analysis of localization of PTENis decreased by increasing Ca 2+ levels in PTEN revealed nuclear PTEN expression in several tissue types, culture medium in a dose-dependent manner. Ca2+ ionophore which seemed to be highest in normal cells and diminish, with A23187 increases nuclear localization of PTENand decreases concomitant increase in cytoplasmic expression, with neoplastic phosphorylation of nuclear p44/42. Finally, we show that progression (6). Our previous study in MCF-7 cells showed that 2+ Ca -dependent PTEN-MVP interaction is not related to MVP’s nuclear PTEN protein levels peak during G0-G1 phases and nadir tyrosil phosphorylation but rather due to its conformational during S phase of the cell cycle (7). We have further shown that modification. Our observations suggest that Ca2+ regulates nuclear PTEN down-regulates phosphorylation of MAPK, down- PTEN’s nuclear entry through a tyrosil phosphorylation– regulates cyclin D1, and induces G1 cell cycle arrest, whereas independent conformational change in MVP. Collectively, cytoplasmic PTEN down-regulates phosphorylation of Akt, up- our data present evidence of a novel crosstalk between the regulates p27, and induces apoptosis (8). These observations Ca2+ signaling–mediated regulation of the cell cycle and suggest that nuclear-cytoplasmic partitioning of PTEN differen- MVP-mediated nuclear PTENlocalization and function. tially regulates the cell cycle and apoptosis and provide further (Cancer Res 2006; 66(24): 11677-82) evidence that nuclear import of PTEN should play a role in carci- nogenesis. In this context, the mechanism of nucleocytoplasmic Introduction shuttling of PTEN is thought to be crucial for the regulation of nuclear and cytoplasmic functions of PTEN. Somatic alterations of the tumor suppressor phosphatase and The vault complex is an evolutionarily conserved ribonucleo- tensin homologue deleted on chromosome 10 (PTEN) are common protein particle with a molecular mass of 13 MDa and is events in diverse human cancers, including carcinomas of the hypothesized as a general carrier molecule for nuclear-cytoplasmic transport (9). The vault is composed of multiple copies of three proteins [major vault protein (MVP), vault poly(ADP-ribose) Note: C. Eng is an honorary fellow of Cancer Research UK Human Cancer Genetics Research Group, University of Cambridge, Cambridge, United Kingdom. polymerase, and telomerase-associated protein 1] and small Requests for reprints: Charis Eng, Genomic Medicine Institute, Lerner Research untranslated RNA molecules. The main component is MVP, which Institute, Cleveland Clinic Foundation, 9500 Euclid Avenue, MailstopNE-50, Cleveland, constitutes over 70% of the total mass of the vault and determines OH 44195. Phone: 216-444-3440; Fax: 216-636-0655; E-mail: [email protected]. I2006 American Association for Cancer Research. its structure. To date, several proteins have been reported to doi:10.1158/0008-5472.CAN-06-2240 interact with MVP, including PTEN (10); estrogen, progesterone, www.aacrjournals.org 11677 Cancer Res 2006; 66: (24). December 15, 2006

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2006 American Association for Cancer Research. Cancer Research and glucocorticoid receptors (11); extracellular signal-regulated proteins were isolated with a buffer extraction system and centrifugation kinases (Erk; ref. 12); and constitutively photomorphogenic 1 according to the manufacturer’s recommendations (NE-Per, Pierce Bio- (COP1), which is an inhibitor of activator protein-1 (AP-1) technology). Proteins were separated by SDS-PAGE, transferred to transcription (13). Thus, MVP is likely to be involved in signal nitrocellulose, and subjected to immunoblotting. The monoclonal antibody 6H2.1 raised against the last 100 COOH-terminal amino acids of PTEN transduction pathways through nucleocytoplasmic transport of (Cascade Biosciences, Portland, OR) was used for PTEN immunoblotting the cargo proteins. However, the molecular mechanism of MVP- and immunoprecipitation (16). Monoclonal anti-MVP (Lab Vision, Fremont, mediated nucleocytoplasmic shuttling remains to be elucidated. CA), monoclonal anti-V5 (Invitrogen), monoclonal anti–phospho-tyrosine We have recently shown that PTEN has bipartite nuclear (Cell Signaling Technology, Beverly, MA), monoclonal anti-actin (Santa localization signal (NLS)–like sequences required for MVP- Cruz, Santa Cruz, CA), anti–phospho-p44/42, and monoclonal anti-p44/42 mediated nuclear import of PTEN (14). MVP reportedly interacts (Cell Signaling Technology, Danvers, MA) antibodies were used. with the C2 domain of PTEN through the EF hand–like motif in a Immunoprecipitation. Cells were harvested, washed with PBS, and Ca2+-dependent manner in HeLa cells (10). Phosphorylation of sonicated in lysis buffer [50 mmol/L Tris-HCl (pH 8), 150 mmol/L NaCl, MVP depends on the presence of Mg2+ in PC12 cells (15), and 1% NP40] containing inhibitors. After pretreatment with protein A/G- epidermal growth factor (EGF) stimulation increases tyrosil Sepharose beads (Santa Cruz) and normal mouse IgG (Santa Cruz), cell lysates (800–1,000 Ag protein) were incubated with the respective phosphorylation of MVP in WI38 cells (12). Therefore, we j 2+ 2+ antibodies overnight at 4 C and then incubated for 3 hours with protein hypothesized that Ca ,Mg, and tyrosil phosphorylation of A/G-Sepharose. The immune complexes were isolated via centrifugation MVP may be playing crucial roles in MVP function, particularly in and washed four times in TNN buffer [20 mmol/L Tris-HCl (pH 8), nuclear import of PTEN. To elucidate the mechanism of MVP- 100 mmol/L NaCl, 0.5% NP40] before boiling in SDS sample buffer and mediated nuclear transport of PTEN, we examined the effects of separation by 8% SDS-PAGE. Ca2+ and Mg2+ on the interaction between PTEN and MVP, Protease sensitivity assay. Protease sensitivity assay was done as subcellular localization of PTEN, and tyrosil phosphorylation and described elsewhere (17). Briefly, immunoprecipitations of recombinant potential conformational changes of MVP in the breast cancer MVP were prepared with anti-V5-tag antibody from transiently transfected 2+ cell line MCF-7. Our findings provide novel insights into the MCF-7 cells, and the beads were washed with TNN buffer containing Ca 2+ j mechanism of MVP-mediated nucleocytoplasmic trafficking and and/or Mg , split to the same volume, and incubated overnight at 37 C with 0, 5, 50, or 500 ng of sequencing grade trypsin in the presence of the also of the regulation of the nuclear and cytoplasmic functions same concentration of Ca2+ and/or Mg2+ as lysis buffer and TNN buffer. of PTEN through subcellular compartmentalization. Samples were separated by 8% SDS-PAGE and immunoblotted with anti- MVP antibody. Materials and Methods Cell lines and culture conditions. MCF-7 breast cancer cells were Results cultured in high-glucose DMEM containing 10% fetal bovine serum (FBS), siRNA-mediated silencing of MVP decreases the nuclear penicillin, and streptomycin. The MCF-7 Tet-Off cell line expressing wild- localization of PTENand increases phosphorylation of nuclear type PTEN (PTEN:WT) was generated and maintained as previously p44/42 MAPK. We have previously reported that PTEN has NLS- described (14). Vector expression was controlled with 2 Ag/mL tetracycline 2+ like sequences required for interaction with MVP and nuclear (Tet). Stock solutions of Ca ionophore A23187 (Calbiochem, San Diego, localization of PTEN (14). To further confirm the role of MVP in CA) were prepared in 99.7% DMSO and were added to the culture medium PTEN nuclear import, we examined the effect of MVP silencing as specified in the text. Small interfering RNA assay. The following small interfering RNA by siRNA on subcellular localization of PTEN. MVP-targeted (siRNA) sequences were synthesized (Integrated DNA Technologies, Coral- siRNA significantly decreased total MVP protein level compared ville, IA): MVP siRNA (plus strand, 5¶-GUUUGAGGAGGUUCUGGAUTT-3¶; with the scrambled siRNA-treated control (Fig. 1A). As expected, minus strand, 5¶-AUCCAGAACCUCCUCAAACTT-3¶) and scrambled siRNA the nuclear localization of PTEN was significantly decreased in (plus strand, 5¶-AUUAGGUGAUUUGGUGGGCTT-3¶;minusstrand,5¶- cells treated with MVP siRNA compared with the control (Fig. 1A GCCCACCAAAUCACCUAAUTT-3¶). MCF-7 cells were plated at 30% to and B). Moreover, nuclear p44/42 was significantly phosphorylat- 40% confluence in six-well plates and were transfected the following day ed by MVP siRNA, whereas cytoplasmic p44/42 was unchanged with 100 pmol/well of MVP or scrambled siRNA using Lipofectamine 2000 (Fig. 1A and C). These findings confirm that MVP is involved in (Invitrogen, San Diego, CA) in serum-free medium. Twenty-four hours after the nuclear localization of PTEN and also suggest that MVP is transfection, cells were placed in complete medium containing 10% FBS regulating downstream functions of nuclear PTEN through nuclear and incubated for additional 48 hours before harvesting. Plasmid construction and transfection. To construct pcDNA3.1D/V5- transport. in situ 2+ His/MVP, the coding region of MVP was amplified by reverse transcription- Interaction between PTENand MVP is Ca 2+ PCR using cDNA isolated from MCF-7 cells as a template with the following dependent and abolished by Mg . It has been reported that primers: 5¶-CACCATGGCAACTGAAGAGTTC-3¶ and 5¶-GCGCAGTACAGG- the interaction between PTEN and MVP depends on the presence CACCACGT-3¶. The 2.7-kb PCR product of MVP was directionally cloned of Ca2+ in glutathione S-transferase pull-down assay (10). Because into pcDNA3.1D/V5-His-TOPO vector (Invitrogen), which contains a V5 and phosphorylation of MVP is reportedly dependent on the presence poly-histidine epitope tag at the COOH terminus, according to the of Mg2+ (15), and because Ca2+ and Mg2+ are known to exert manufacturer’s protocol. The PCR product and the construct were opposing effects in many cellular functions (18, 19), we hypothe- confirmed by DNA sequencing. The generated construct was transiently sized that Ca2+ and Mg2+ might be involved in the interaction transfected into MCF-7 or PTEN:WT cells using Lipofectamine 2000 between PTEN and MVP. Immunoprecipitation showed that (Invitrogen) in antibiotics-free, serum-free medium for 24 hours as 2+ recommended by the manufacturer’s instructions and incubated in Ca -dependent interaction of PTEN with MVP was abolished by 2+ 2+ A complete medium for additional 24 hours before harvesting. addition of the same concentration of Mg as Ca (Fig. 2 ). 2+ SDS-PAGE, immunoblotting, and antibodies. Whole cell–free protein Because the effect of Ca on the interaction between PTEN and extracts were prepared according to the manufacturer’s recommendations MVP is reported to be dose dependent (10), we subsequently (M-Per, Pierce Biotechnology, Rockford, IL). Nuclear and cytoplasmic examined a dose-effect of Mg2+ on the interaction. Interestingly,

Cancer Res 2006; 66: (24). December 15, 2006 11678 www.aacrjournals.org

medium. Incubation in a dose-gradient increase of Ca2+ showed a dose-dependent decrease of nuclear localization of PTEN (Fig. 3A and B). This finding suggests that cellular Ca2+ concentration may be regulating nuclear localization of PTEN. Calcium ionophore A23187 increases nuclear localization of PTENand decreases phosphorylation of nuclear p44/42 MAPK in a dose-dependent manner. A23187 is known to increase cytoplasmic Ca2+ levels by releasing it from the endoplasmic reticulum Ca2+ store. To confirm the regulation of nuclear PTEN localization by Ca2+, we examined the effect of A23187 on PTEN’s localization. Incubation with 0.1 or 1.0 Amol/L A23187 increased nuclear localization of PTEN (Fig. 4A and B). To determine if the PTEN that was translocated to the nucleus was active, we further examined the downstream effects of this increased nuclear PTEN. A23187 incubation decreased phosphorylation of nuclear p44/42 in a dose-dependent manner (Fig. 4A and C). These data suggest that intracellular increases of Ca2+ up-regulates nuclear localization of PTEN, which consequently down-regulates phosphorylation of nuclear p44/42. Ca2+-dependent PTEN-MVP interaction is not regulated by tyrosil phosphorylation of MVP. Phosphorylation of MVP is reported to depend on the presence of Mg2+ (15), and EGF stimulation increases MVP tyrosil phosphorylation (12). Additionally, we have recently found that subcellular localization of PTEN is not regulated by PTEN phosphorylation (14). Therefore, we hypothesized that Ca2+ and Mg2+ might be regulating PTEN-MVP interaction and PTEN nuclear transport through tyrosil phosphorylation of MVP. We first verified the interaction of recombinant MVP with PTEN in situ by immunoprecipitation. Recombinant MVP interacted with PTEN in the presence of Ca2+ and addition of Mg2+ abolished the interaction (Fig. 5), which is consistent with the result on endogenous MVP (Fig. 2A). Next, we examined the effect of Ca2+ and/or Mg2+ on tyrosil phosphorylation of MVP. No

Figure 1. siRNA-mediated silencing of MVP decreases nuclear localization of PTEN and increases phosphorylation of nuclear p44/42 MAPK. A, MCF-7 cells were transfected with MVP-targeted siRNA (MVP) or scrambled siRNA (Cont) in serum-free medium for 24hours and incubated in complete medium for an additional 48 hours before harvesting. Protein levels of MVP, PTEN, phospho-p44/42 (P-p44/42), p44/42, and actin were examined by immunoblotting in whole-cell extracts and nuclear-cytoplasmic fractions. B, data in (A) were quantified by densitometric analysis, and each ratio of nuclear to cytoplasmic fraction was calculated and normalized to control. C, each ratio of phospho-p44/42 to p44/42 was calculated and normalized to control. Columns, mean (n = 3); bars, SD. as Mg2+ concentration increases, the PTEN-MVP interaction slightly increases at lower concentrations of Mg2+ and then is abolished at the half concentration of Ca2+ (Fig. 2B). These findings indicate that Mg2+ antagonizes Ca2+-dependent interaction of PTEN with MVP, and the antagonizing effect of Mg2+ is at least 2-fold stronger than Ca2+ at the same concentration. 2+ Nuclear localization of PTEN is decreased by increasing Ca Figure 2. Mg2+ abolishes Ca2+-dependent interactions between PTEN and concentrations in the culture medium in a dose-dependent MVP. PTEN:WT cells were incubated for 48 hours in Tet-free medium before harvesting, and the PTEN-MVP interaction was examined by manner. Because the interaction between PTEN and MVP is immunoprecipitation (IP). A, MVP was precipitated with anti-PTEN antibody in 2+ reportedly affected by Ca in a dose-dependent manner (10), we the presence of 10 mmol/L CaCl2 and/or 10 mmol/L MgCl2 and immunoblotted next examined a dose-effect of extracellular Ca2+ on nuclear (IB) with anti-PTEN or anti-MVP antibody. For control, immunoprecipitation was done in the absence of CaCl2 and MgCl2. B, MVP was precipitated with localization of PTEN. MCF-7 cells were subjected to artificial anti-PTEN antibody in the presence of the indicated concentrations of CaCl 2+ 2 concentrations of Ca via the addition of CaCl2 in the culture and MgCl2. www.aacrjournals.org 11679 Cancer Res 2006; 66: (24). December 15, 2006

conclusion, the present study examined the effect of siRNA- mediated silencing of MVP on PTEN subcellular localization. As expected, MVP-targeted siRNA significantly decreased total protein level of MVP and nuclear localization of PTEN compared with the control (Fig. 1A and B). Although possibilities of other pathways in PTEN nuclear import, such as passive diffusion and mediation by other transporters, still cannot be excluded, our data suggest that MVP-mediated nuclear import is a major mechanism

Figure 3. Nuclear localization of PTEN is decreased by increasing Ca2+ concentration in the culture medium in a dose-dependent manner. A, After 24hours of serum deprivation, MCF-7 cells were incubated in complete media supplemented with CaCl2 for an additional 24hours. Calculated concentrations of Ca2+ in the medium are indicated. Protein levels of PTEN and actin were examined by immunoblotting in whole-cell extracts and nuclear-cytoplasmic fractions. B, data in (A) were quantified by densitometric analysis, and ratios of PTEN’s nuclear-to-cytoplasmic fractions were calculated and normalized to the ratio of 1.6 mmol/L Ca2+. Columns, mean (n = 3); bars, SD. significant differences were found in tyrosil phosphorylation of MVP among Ca2+,Ca2+ plus Mg2+,Mg2+, and control samples (Fig. 5A and B). These observations suggest that MVP’s Ca2+- dependent interaction with PTEN is not regulated by MVP tyrosil phosphorylation. Ca2+ causes a conformational modification in MVP. It is known that Ca2+ may result in conformational changes of proteins containing the Ca2+-binding EF hand motif (20). Hence, we next examined the effect of Ca2+ and Mg2+ on the conformational change of recombinant MVP by partial tryptic digestion. Digestions with 5 and 50 ng trypsin in the presence of Ca2+ shows an increase of larger fragments from recombinant MVP compared with Ca2+/Mg2+-free control, Ca2+ plus Mg2+,orMg2+ samples, indicating that Ca2+ decreases recombinant MVP’s sensitivity to trypsin, and Mg2+ antagonizes the effect (Fig. 6). This finding indicates that Ca2+ causes MVP’s conformational modification, and Mg2+ antagonizes it. Figure 4. Calcium ionophore A23187 increases nuclear localization of PTEN and decreases phosphorylation of nuclear p44/42 MAPK in a dose-dependent manner. A, after 24hours of serum deprivation, MCF-7 cells were incubated in Discussion complete medium with 0.01, 0.1, or 1.0 Amol/L A23187 for an additional 24hours before harvesting. As a control, cells were incubated in the complete medium We have previously shown that PTEN has four putative NLS-like added with appropriate volumes of DMSO. Protein levels of PTEN, actin, p44/42, and phospho-p44/42 were examined by immunoblotting in whole-cell sequences, and that two double-site mutations cause nuclear extracts and nuclear-cytoplasmic fractions. B, data in (A) were quantified by localization defects of PTEN and abolish the interaction with MVP, densitometric analysis, and each ratio of PTEN’s nuclear to cytoplasmic fraction À concluding that MVP mediates PTEN nuclear import (14). was calculated and normalized to the ratio of DMSO 10 5%. C, each ratio of nuclear phospho-p44/42 to p44/42 of A23187 was divided by corresponding However, another grouphas recently reportedthat PTEN’s nuclear ratio of DMSO and normalized to the value of 0.01 Amol/L A23187. Columns, entry is mediated by passive diffusion (21). To further confirm our mean (n = 3); bars, SD.

Cancer Res 2006; 66: (24). December 15, 2006 11680 www.aacrjournals.org

gonistically towards each other (18, 19), we hypothesized that Ca2+ and Mg2+ might be regulating the PTEN-MVP interaction by antagonizing each other such that they regulate nuclear import of PTEN. Our immunoprecipitation results show that Mg2+ antagonizes Ca2+-dependent interaction between PTEN and MVP (Fig. 2A and B). Therefore, we speculate that nuclear localization of PTEN is regulated by subtle differences of Ca2+ and Mg2+ concentrations between the cytoplasm and nucleus by affecting the interaction with MVP. We further examined the effect of increased cytoplasmic Ca2+ levels on PTEN nuclear localization by incubation with the calcium ionophore A23187. As expected, A23187 increased nuclear localization of PTEN (Fig. 4A and B). It can be speculated that the increased PTEN-MVP interaction due to cytoplasmic Ca2+ accu- mulation results in increased import of PTEN proteins to the nucleus. However, the result of our experiment manipulating extracellular Ca2+ (i.e., in the culture medium) was apparently contradictory to the results obtained with A23187 (Fig. 3A and B). The mechanism whereby extracellular increases of Ca2+ levels decrease the nuclear localization of PTEN may be explained by PTEN being unable to dissociate from MVP even in the nucleus due to an increased interaction intensity because of the steepincrease in nuclear Ca2+. Additionally, the results of our experiments using A23187 suggest that increased nuclear localization of PTEN consequently down-regulates phosphorylation of p44/42 MAPK (Fig. 4A and C). This downstream effect is consistent with our previous observations in PTEN’s nuclear localization defect Figure 5. Ca2+- and Mg2+-dependent PTEN-MVP interaction is not related to mutants that nuclear PTEN down-regulates phosphorylation of tyrosil phosphorylation of MVP. A, to first validate the interaction of recombinant MVP with PTEN, PTEN:WT cells were transiently transfected with an MVP p44/42 and cyclin D1 and induces G0-G1 cell cycle arrest (8). construct in Tet-free medium for 48 hours before harvesting. Recombinant MVP Combined with these recent findings, our data are suggestive of the was precipitated with anti-PTEN antibody in the presence of 10 mmol/L CaCl 2 possibility that nuclear PTEN induces G0-G1 cell cycle arrest by and/or 10 mmol/L MgCl2 as indicated and immunoblotted with anti-V5-tag 2+ 2+ or anti-PTEN antibody. For the evaluation of MVP’s tyrosil phosphorylation, Ca signaling–regulated nuclear import. Ca signaling is known MCF-7 cells were transiently transfected with the MVP construct and harvested to regulate G1-S progression of the cell cycle through promoting 48 hours after transfection. Recombinant MVP was precipitated with anti-V5-tag phospholipase C/protein kinase C/MAPK and calmodulin (CaM)/ antibody in the presence of 10 mmol/L CaCl2 and/or 10 mmol/L MgCl2 as indicated and immunoblotted with anti–phospho-tyrosine (p-Tyr) or anti-MVP CaM-dependent kinase (CaM-K) kinase/CaM-K pathways (23–26). 2+ antibody. For control, immunoprecipitation was done in the absence of CaCl2 and Therefore, it is reasonable to postulate that Ca signaling is also MgCl2. B, data in (A) were quantified by densitometric analysis, and ratios of phospho-tyrosine to MVP were calculated and normalized to control. Columns, regulating nuclear entry of PTEN, resulting in modulation of G1 cell mean (n = 3); bars, SD. cycle arrest. However, it is necessary to further elucidate how calcium signaling up-regulates and down-regulates G1 cell cycle arrest through nuclear PTEN-dependent and PTEN-independent of PTEN translocation because we were able to significantly pathways. diminish PTEN nuclear localization in MVP siRNA-treated cells. Recently, it has been reported that UV irradiation enhances Additionally, we have recently reported that PTEN in both wild- tyrosil phosphorylation of MVP and causes dissociation of COP1, type PTEN-overexpressing MCF-7 cells as well as in PTEN nuclear which is an inhibitor of AP-1 transcription, from MVP (13). localization defect cells interacts equally with importin-a and importin-h (14). Therefore, it is possible that there are NLS-like sequence-independent pathways of PTEN nuclear import mediated by importins as well as passive diffusion. Although further work is required to clarify this issue, multiple pathways are likely to be involved in the nuclear import of PTEN, as analogously with Erk. Erk has been shown to be imported to the nucleus by two coexisting mechanisms: passive diffusion and Ran-dependent active transport (22). Interestingly, MVP is reported to interact with phosphorylated p44/42 (Erks) in response to EGF stimulation (12). If MVP is involved in nuclear import of phospho-p44/42 Figure 6. Ca2+ decreases recombinant MVP’s sensitivity to tryptic digestion, as well, MVP-targeted siRNA should decrease nuclear phospho- and Mg2+ antagonizes the effect. MCF-7 cells were transiently transfected with p44/42. However, our results show that nuclear localization of MVP construct and harvested 48 hours after transfection. Recombinant MVP was precipitated with anti-V5-tag antibody in the presence of CaCl2 and/or MgCl2 phospho-p44/42 is increased, most likely due to the upstream event as indicated, washed, split to the same volume, and partially digested at 37jC (i.e., decreased nuclear localization of PTEN; Fig. 1B). overnight with 0, 5, 50, or 500 ng trypsin. Samples were separated by 8% Because published findings suggest that Ca2+ and Mg2+ play SDS-PAGE and immunoblotted with anti-MVP antibody. Arrows, digestions in the presence of Ca2+ alone showed an increase of larger fragments from important roles in MVP’s function (10, 12), and because in many recombinant MVP after 5 and 50 ng tryptic digestion compared with the other cellular processes these divalent cations are known to act anta- samples. www.aacrjournals.org 11681 Cancer Res 2006; 66: (24). December 15, 2006

Additionally, the interaction between MVP and Src tyrosine kinase recombinant MVP’s conformational change. Our results here is also dependent on tyrosil phosphorylation of MVP (27). Because indicate that Ca2+ induces a conformational change within MVP, EGF stimulation increases tyrosil phosphorylation of MVP (12, 27), and Mg2+ antagonizes this effect (Fig. 6). We therefore postulate and because we previously found that PTEN’s phosphorylation that Ca2+ binding to MVP causes MVP’s conformational alteration status is not correlated with its subcellular localization (14), we so that the affinity of MVP’s EF hand–like motif for PTEN’s C2 hypothesized that Ca2+/Mg2+–dependent MVP-PTEN interactions domain is enhanced, and that Mg2+ has a competitive effect on may be mediated by tyrosil phosphorylation of MVP. However, Ca2+ binding to MVP, although our results suggest that Mg2+ has interestingly, our data suggest that the Ca2+/Mg2+–dependent agonistic effects at a lower Mg2+/Ca2+ ratio as well (Fig. 2B). We interaction of MVP with PTEN is not regulated by MVP’s tyrosil believe our postulate based on our data (i.e., Ca2+-dependent phosphorylation (Fig. 5). Nevertheless, tyrosil phosphorylation of regulation of nuclear entry by MVP) is plausible, as Ca2+ is also MVP is likely to play an important role in the growth factor/ reported to regulate opening and closing of the nuclear pore receptor tyrosine kinase/MAPK pathway because tyrosil phosphor- complex through its conformation change (29). ylated MVP reportedly forms a complex with SHP-2 and Erks in In summary, we show here that Ca2+ is regulating the nuclear response to EGF stimulation (12). Besides, it has been recently localization of PTEN and its downstream effect (i.e., phosphory- shown that EGF stimulation triggers MVP translocation from the lation of nuclear p44/42), by modulating the interaction with MVP, nucleus to the cytosol and perinuclear region where it colocalizes and that the Ca2+-dependent PTEN-MVP interaction is attributable with Src, and that the EGF-dependent MVP-Src interaction down- to MVP’s conformational change but not to MVP’s tyrosil regulates phosphorylation of MAPK (27). Hence, there seems to phosphorylation. Importantly, our findings provide a novel cross- be tyrosil phosphorylation–dependent and tyrosil phosphoryla- talk pathway between MVP-mediated nuclear transport and tion–independent mechanisms regulating MVP-mediated MAPK function of PTEN and the Ca2+ signaling–mediated regulation of pathway. the cell cycle in breast cancer. Because our data showed that tyrosil phosphorylation of MVP did not mediate ion-associated MVP-PTEN interactions, we pursued other mechanisms. Ca2+ binding to EF hand motifs is Acknowledgments known to cause conformational changes within many proteins Received 6/19/2006; revised 9/24/2006; accepted 10/13/2006. Grant support: American Cancer Society grant RSG02-151-01CCE (C. Eng) and containing EF hand motifs, including calmodulin, neuronal calcium Doris Duke Distinguished Clinical Scientist Award (C. Eng). sensor family, and S100 protein family (20). An EF hand–like motif The costs of publication of this article were defrayed in part by the payment of page of MVP interacts with C2 domain of PTEN in a Ca2+-dependent charges. This article must therefore be hereby marked advertisement in accordance 2+ with 18 U.S.C. Section 1734 solely to indicate this fact. manner (10). Because PTEN’s C2 domain has a Ca -independent T. Minaguchi thanks many members of the Eng lab for helpful discussions and 2+ 2+ motif (28), we examined the effect of the interplay of Ca /Mg on technical advice.

References of vault particles with estrogen receptor in the MCF-7 20. Ikura M, Ames JB. Genetic polymorphism and breast cancer cell. J Cell Biol 1998;141:1301–10. protein conformational plasticity in the calmodulin 1. Waite KA, Eng C. Protean PTEN: form and function. 12. Kolli S, Zito CI, Mossink MH, Wiemer EA, Bennett superfamily: two ways to promote multifunctionality. Am J Hum Genet 2002;70:829–44. AM. The major vault protein is a novel substrate for the Proc Natl Acad Sci U S A 2006;103:1159–64. 2. Li DM, Sun H. TEP1, encoded by a candidate tumor tyrosine phosphatase SHP-2 and scaffold protein in 21. Liu F, Wagner S, Campbell RB, Nickerson JA, Schiffer suppressor locus, is a novel protein tyrosine phospha- epidermal growth factor signaling. J Biol Chem 2004;279: CA, Ross AH. PTEN enters the nucleus by diffusion. J tase regulated by transforming growth factor beta. 29374–85. Cell Biochem 2005;96:221–34. Cancer Res 1997;57:2124–9. 13. Yi C, Li S, Chen X, et al. Major vault protein, in 22. Adachi M, Fukuda M, Nishida E. Two co-existing 3. Steck PA, Pershouse MA, Jasser SA, et al. Identification concert with constitutively photomorphogenic 1, nega- mechanisms for nuclear import of MAP kinase: passive of a candidate tumour suppressor gene, MMAC1, at tively regulates c-Jun-mediated activator protein 1 diffusion of a monomer and active transport of a dimer. chromosome 10q23.3 that is mutated in multiple transcription in mammalian cells. Cancer Res 2005;65: EMBO J 1999;18:5347–58. advanced cancers. Nat Genet 1997;15:356–62. 5835–40. 23. Kahl CR, Means AR. Regulation of cell cycle 4. Nakamura N, Ramaswamy S, Vazquez F, Signoretti 14. Chung JH, Ginn-Pease ME, Eng C. Phosphatase and progression by calcium/calmodulin-dependent path- S, Loda M, Sellers WR. Forkhead transcription factors tensin homologue deleted on chromosome 10 (PTEN) ways. Endocr Rev 2003;24:719–36. are critical effectors of cell death and cell cycle has nuclear localization signal-like sequences for 24. Schmitt JM, Wayman GA, Nozaki N, Soderling TR. arrest downstream of PTEN. Mol Cell Biol 2000;20: nuclear import mediated by major vault protein. Cancer Calcium activation of ERK mediated by calmodulin 8969–82. Res 2005;65:4108–16. kinase I. J Biol Chem 2004;279:24064–72. 5. Comer FI, Parent CA. PI 3-kinases and PTEN: how 15. Ehrnsperger C, Volknandt W. Major vault protein is a 25. Lin CC, Shyr MH, Chien CS, et al. Thrombin- opposites chemoattract. Cell 2002;109:541–4. substrate of endogenous protein kinases in CHO and stimulated cell proliferation mediated through activa- 6. Eng C. PTEN: one gene, many syndromes. Hum Mutat PC12 cells. Biol Chem 2001;382:1463–71. tion of Ras/Raf/MEK/MAPK pathway in canine cultured 2003;22:183–98. 16. WengLP,SmithWM,DahiaPL,etal.PTEN tracheal smooth muscle cells. Cell Signal 2002;14:265–75. 7. Ginn-Pease ME, Eng C. Increased nuclear phosphatase suppresses breast cancer cell growth by phosphatase 26. Rodriguez-Mora OG, LaHair MM, McCubrey JA, and tensin homologue deleted on chromosome 10 is function-dependent G1 arrest followed by apoptosis. Franklin RA. Calcium/calmodulin-dependent kinase I associated with G0–G1 in MCF-7 cells. Cancer Res 2003; Cancer Res 1999;59:5808–14. and calcium/calmodulin-dependent kinase kinase 63:282–6. 17. Vazquez F, Grossman SR, Takahashi Y, Rokas MV, participate in the control of cell cycle progression 8. Chung JH, Eng C. Nuclear-cytoplasmic partitioning of Nakamura N, Sellers WR. Phosphorylation of the PTEN in MCF-7 human breast cancer cells. Cancer Res phosphatase and tensin homologue deleted on chro- tail acts as an inhibitory switch by preventing its 2005;65:5408–16. mosome 10 (PTEN) differentially regulates the cell cycle recruitment into a protein complex. J Biol Chem 2001; 27. Kim E, Lee S, Mian MF, et al. Crosstalk between Src and apoptosis. Cancer Res 2005;65:8096–100. 276:48627–30. and major vault protein in epidermal growth factor- 9. Mossink MH, van Zon A, Scheper RJ, Sonneveld P, 18. Walker GM. Biotechnological implications of the dependent cell signalling. FEBS J 2006;273:793–804. Wiemer EA. Vaults: a ribonucleoprotein particle in- interactions between magnesium and calcium. Magnes 28. Murray D, Honig B. Electrostatic control of the mem- volved in drug resistance? Oncogene 2003;22:7458–67. Res 1999;12:303–9. brane targeting of C2 domains. Mol Cell 2002;9:145–54. 10. Yu Z, Fotouhi-Ardakani N, Wu L, et al. PTEN 19. Laurant P, Touyz RM. Physiological and pathophys- 29. Stoffler D, Goldie KN, Feja B, Aebi U. Calcium- associates with the vault particles in HeLa cells. J Biol iological role of magnesium in the cardiovascular mediated structural changes of native nuclear pore Chem 2002;277:40247–52. system: implications in hypertension. J Hypertens 2000; complexes monitored by time-lapse atomic force 11. Abbondanza C, Rossi V, Roscigno A, et al. Interaction 18:1177–91. microscopy. J Mol Biol 1999;287:741–52.

Cancer Res 2006; 66: (24). December 15, 2006 11682 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 2006 American Association for Cancer Research. Nuclear Localization of PTEN Is Regulated by Ca2+ through a Tyrosil Phosphorylation−Independent Conformational Modification in Major Vault Protein

Takeo Minaguchi, Kristin A. Waite and Charis Eng

Cancer Res 2006;66:11677-11682.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/66/24/11677

Cited articles This article cites 29 articles, 15 of which you can access for free at: http://cancerres.aacrjournals.org/content/66/24/11677.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/66/24/11677.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/66/24/11677. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.