|Home | About | Journals | Submit | Contact Us | Français|
The tumor suppressor gene PTEN (phosphatase and tensin homologue deleted on chromosome 10) is frequently mutated or deleted in various human cancers. PTEN localizes predominantly to the cytoplasm and functions as a lipid phosphatase, thereby negatively regulating the phosphatidylinositol 3-kinase-AKT signaling pathway. PTEN can also localize to the nucleus, where it binds and regulates p53 protein level and transcription activity. However, the precise function of nuclear PTEN and the factors that control PTEN nuclear localization are still largely unknown. In this study, we identified oxidative stress as one of the physiological stimuli that regulate the accumulation of nuclear PTEN. Specifically, oxidative stress inhibits PTEN nuclear export, a process depending on phosphorylation of its amino acid residue Ser-380. Nuclear PTEN, independent of its phosphatase activity, leads to p53-mediated G1 growth arrest, cell death, and reduction of reactive oxygen species production. Using xenografts propagated from human prostate cancer cell lines, we reveal that nuclear PTEN is sufficient to reduce tumor progression in vivo in a p53-dependent manner. The data outlined in this study suggest a unique role of nuclear PTEN to arrest and protect cells upon oxidative damage and to regulate tumorigenesis. Since tumor cells are constantly exposed to oxidative stress, our study elucidates the cooperative roles of nuclear PTEN with p53 in tumor suppression.
The PTEN (phosphatase and tensin homologue deleted on chromosome 10) tumor suppressor gene is mutated at high frequency in many primary human cancers and several cancer predisposition disorders (2). PTEN encodes a dually specific phosphatase that recognizes both lipid and peptide substrates (23), including phosphatidylinositol (3,4,5)-trisphosphate (PIP3), a product of phosphatidylinositol 3-kinase (PI3K). PTEN protein contains an N-terminal catalytic phosphatase domain (18, 32), a calcium-independent C2 domain (16), two PEST motifs, and a C-terminal PDZ binding domain (1). Several critical phosphorylation sites have been found in the PTEN C2 domain, including Ser-380, Thr-382, Thr-383, and Ser-385. Importantly, phosphorylation of these residues has been implicated to increase PTEN stability but decrease PTEN catalytic activity (36, 37).
Although PTEN is localized mainly to the cytoplasm, it preferentially resides in the nucleus of differentiated or resting cells (15) as exemplified in MCF-7 cells (14), in which nuclear PTEN peaks in the G1 phase and reaches a nadir in the S phase. Interestingly, changes in nuclear PTEN expression have also been observed in the endometrium during hormonal cycles (27). These data suggest that nuclear localization of PTEN is a dynamic process, associated with cell cycle, cell differentiation, and cellular functions. Decreased nuclear PTEN has been correlated with progressing thyroid carcinoma and melanoma (40), suggesting a functional role of nuclear PTEN in regulating tumorigenesis.
Several studies have shown that PTEN nuclear localization depends on the presence of noncanonical nuclear localization signals and major vault protein-mediated nuclear transport (5), whereas others have indicated that PTEN nuclear localization occurs by diffusion through the nuclear membrane (20). More recent results from several studies indicate that PTEN nuclear localization is regulated by collaborative events, including feedback regulation by PI3K-S6K signaling (21), an N-terminal nuclear localization domain that is modulated by a Ran-dependent mechanism, the presence of multiple exclusion motifs (13), mutations in the N-terminal cytoplasmic localization signal (8), or monoubiquitination at amino acid residues K289 and K13 (35).
Through its lipid phosphatase activity, PTEN controls AKT signaling and its downstream targets responsible for cell size, cell motility, cell cycle, and cell death (9, 28, 30). Mutations (C124S and G129R) in the PTEN catalytic domain lead to the loss of PTEN′s phosphatase activity as well as its tumor-suppressing ability. Moreover, it has been shown that catalytically inactive PTEN binds to and promotes stabilization, acetylation, and tetramerization of p53 in the nucleus through phosphatase-independent (12, 17) and MDM2-independent (4, 17) mechanisms. Thus, PTEN functions in the nucleus in both phosphatase-dependent and -independent manner.
The significance of nuclear PTEN in regulating tumorigenesis has recently been addressed. For example, forced PTEN nuclear expression can inhibit anchorage-independent growth, induce accumulation of the cells in G1 (22), and suppress tumor progression by inhibiting nuclear P-AKT (35). It has also been proposed that nuclear PTEN can induce cell cycle arrest, in part, by reducing cyclin D1 levels through its protein phosphatase activity (39) or through controlling mitogen-activated protein kinase signaling (6). Furthermore, nuclear PTEN has been demonstrated to control chromosome stability and DNA repair (31). However, most of these studies were based on in vitro biochemical analyses and cannot by themselves prove or disprove the significance of nuclear PTEN in maintaining normal cellular function and modulating cancer progression.
In this study, we validate that a specific pool of PTEN (P-PTENS380) accumulates in the nucleus upon oxidative stress. Nuclear P-PTEN associates with p53 to enhance cell cycle arrest and reactive oxygen species (ROS) reduction via a p53-dependent mechanism. Moreover, in the presence of p53, xenograft studies demonstrate that nuclear PTEN, independent of its phosphatase activity, is sufficient to regulate tumorigenesis in vivo. In summary, our study suggests a unique role of nuclear PTEN to protect cells against oxidative damage and to regulate tumorigenesis.
Pten WT (wild type) and PtenΔ/Δ mouse embryonic fibroblasts (MEFs), 293T cells, and p53−/−; Mdm2−/− MEFs were cultured in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal calf serum (HyClone) and 100 U/ml penicillin and streptomycin (Invitrogen). PC3 cells were cultured in Roswell Park Memorial Institute medium supplemented with 10% fetal calf serum (HyClone) and 100 U/ml penicillin and streptomycin (Invitrogen). Cell transfection was performed using Lipofectamine (Invitrogen). Cells were examined at 24 h posttransfection for Western blotting and at 48 h posttransfection for cell cycle and apoptosis assays. For oxidative stress treatment, cells were treated with 1 mM H2O2 for 1 h in serum-free medium and then subjected to analyses.
PTEN expression plasmids pSG5LPTEN-WT (WT), pSG5LPTEN-G129R (GR) tagged with nuclear localization signal/nucleus exclusion signal (NLS/NES) sequence, and 380A/380D mutants (gifts from W. R. Sellers) were constructed into a retroviral pMX-IRES (internal ribosome entry site) enhanced green fluorescent protein vector. Various cell lines were infected with the supernatant from 293T cells that had been transfected (as per Gary Nolan's protocol [http://www.stanford.edu/group/nolan/protocols/pro_helper_dep.html]) with the pMX-PTEN expression plasmids. Cells were examined at 72 h postinfection for xenograft experiments. Infection efficiency was determined by green fluorescence under microscopy.
Whole-cell extract was prepared by lysing the cells in a lysis buffer containing 50 mM Tris-HCl (pH 8.0), 120 mM NaCl, 0.5% NP-40, 0.25% sodium deoxycholate, 1 mM dithiothreitol, 2 μg/ml aprotinin, and 2 μg/ml leupeptin. Nuclear and cytoplasmic extracts were prepared with a nuclear extract kit (Active Motif). The nuclear export assay was performed as described previously (7, 39). Cell lysates from each transfection were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred onto nitrocellulose (Bio-Rad), and followed by Western blot analysis using p53 antibody (DO-1; Santa Cruz), p21, MDM2 (Ab-2; Oncogene Research), P-PTENS380 (9551; Cell Signaling), PTEN (9552; Cell Signaling), Bax (2772; Cell Signaling), histone deacetylase 1 (SC-15022; Santa Cruz), and actin (A 4700; Sigma) antibodies. Quantification was performed with ImageJ program (NIH).
The nuclear export assay was performed according to reference 7. Briefly after digitonin permeabilization, the nuclei were incubated with fractionated cytosolic protein (75 μg), an ATP-regenerating system (5 mM ATP, 5 mM creatine phosphate, 20 U/ml creatine phosphokinase), and 2 mM GTP at room temperature for up to 60 min. Nuclear export was stopped by centrifugation, and the supernatant was removed as the exported part.
Five hundred micrograms of cell lysate was incubated for 16 h at 4°C with 10 μl p53 (Santa Cruz) and PTEN antibodies (Cell Signaling) plus 50 μl protein A agarose beads (Upstate). Beads were washed three times with lysis buffer and centrifuged for 5 min at 5,000 × g between each wash. Protein was eluted from beads with 50 μl Laemmli sample buffer (Bio-Rad) and subjected to Western blotting as described above.
After treatment at the indicated time(s) or 24 h posttransfection, cells were fixed with 3% formaldehyde, permeabilized with 0.1% Triton in phosphate-buffered saline (PBS), and preblocked with 1% bovine serum albumin in PBS. Cells were then incubated with anti-PTEN antibody (Cell Signaling) or monoclonal 8-oxo-dG antibody (Trevigen) at 4°C overnight, followed by the secondary antibody conjugated with Alexa Fluor 594 or 488 (Molecular Probes). The cells were mounted with mounting medium containing 1 μg/ml DAPI (4′,6′-diamidino-2-phenylindole; Vector Labs).
Total RNAs were extracted from cells by using an RNeasy kit (Qiagen). RNAs were reverse transcribed by using a Superscript II kit (Invitrogen). Results were analyzed by the iCycler (Bio-Rad) real-time PCR and relative quantification of RNA levels normalized to actin as the difference of cycle threshold (ΔCT) = CT (target) − CT (control). Higher CT values indicate relatively lower levels of RNA expression. Primers 5′-CTCACAGCTGGTCTGTGTG-3′ (forward) and 5′-CCTCCGTGTGGCAATACC-3′ (reverse) were used to detect Sestrin mRNA.
Cells were dissociated with trypsin, washed, and resuspended in PBS as a single-cell suspension. Cells were fixed in 70% ethanol overnight, stained with propidium iodide (25 μg/ml) (Sigma), and incubated for 30 min at 37°C with RNase A (20 μg/ml). The DNA content of the cells was then evaluated by fluorescence-activated cell sorting with a FACSCalibur (BD Immunocytometry Systems). Linear red fluorescence FL2 was analyzed. For ROS production, cells were incubated with 10 μg/ml DCF reagent (Molecular Probes) for 20 min at 37°C and run on a flow cytometer with green fluorescence FL1 analyzed.
Apoptosis was determined with a terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) kit (Roche) with a modified protocol for immunofluorescence. The apoptotic index was determined by counting the total number of positive nuclei in 10 randomly selected fields at ×20 magnification.
Male SCID mice (n = 3) were bilaterally inoculated subcutaneously with 1 × 106 infected PC3 and C4-2 LNCaP cells in Matrigel (BD Bioscience) at 50:50 volume for a total volume of 100 μl/injection site. Average tumor burden was calculated in millimeters with calipers every 3 days according to the formula volume = (π/6) × length × width × height. After 4 and 8 weeks, all of the animals were sacrificed, tumors were harvested, and wet weights were determined. All surgical procedures were performed under regulations of the Division of Laboratory Animal Medicine regulations of the University of California, Los Angeles.
All data are presented as means ± standard deviation (SD). Statistical calculations were performed with Microsoft Excel analysis tools. Differences between individual groups were analyzed by paired t test. P values of <0.05 were considered statistically significant.
To identify the potential stimuli that can modulate PTEN nuclear localization, we compared the effects of several treatments, including serum starvation, serum stimulation, UV irradiation, and oxidative stress on WT MEFs (data not shown). Treatment with hydroxyurea was used as a positive control for dominant PTEN nuclear localization as reported previously (15). Interestingly, when applying oxidative stress (1 mM H2O2) to WT MEF cells, we observed an increase in total PTEN level in the nucleus (Fig. 1A and B), which was concurrent with an even more pronounced increase in nuclear phospho-PTEN at amino acid residue S380 (P-PTEN S380) (Fig. (Fig.1B).1B). Also associated with oxidative stress was increased P-PTEN-p53 complex formation (Fig. 1B and C). To further examine the interaction between p53 and PTEN upon oxidative stress, we reintroduced PTEN in Pten-null MEF cells and found that oxidative stress led to elevated P-PTEN and total PTEN levels in the nucleus, correlating to increased p53 level in the same compartment (Fig. (Fig.1D).1D). This observation is also consistent with our previous reports that PTEN complexes with and stabilizes p53 in the nucleus (12, 17). Taken together, these data suggest oxidative stress as a physiological stimulus that can modulate PTEN nuclear localization. In addition, our study indicates that, in response to oxidative stress, a specific pool of P-PTEN could accumulate in the nucleus, enhance PTEN-p53 association, and increase p53 levels (see quantifications in Fig. 1C and D).
Our data suggest that increased nuclear P-PTENS380 is associated with an increased pool of nuclear PTEN and p53 association upon exposure to oxidative stress. We next inquired whether PTEN nuclear localization is regulated by PTENS380 phosphorylation status. In order to test this hypothesis, we generated two PTEN phosphorylation mutants to mimic the unphosphorylated (380A) and constitutively phosphorylated (380D) PTEN at the residue serine 380, respectively. In the absence of H2O2, PTEN380A showed a more prominent nuclear localization than 380D (Fig. (Fig.2A),2A), consistent with our hypothesis that PTENS380 phosphorylation may negatively regulate its nuclear localization. In response to oxidative stress, the phosphorylation-mimicking PTEN380D showed a twofold increase in nuclear accumulation, as compared to PTEN380A (Fig. (Fig.2A).2A). To ascertain whether such an accumulation is due to the effects of oxidative stress on PTEN nuclear import, export, or both, we employed a nuclear export assay (see Materials and Methods). PTEN380D was found to be rapidly exported from the nucleus to the cytoplasm (Fig. (Fig.2B,2B, left), while PTEN380A showed retarded export within the same period of time (Fig. (Fig.2B,2B, right), suggesting that PTEN nuclear export could be S380 phosphorylation dependent. These data indicate the potential for oxidative stress to retain the mobile portion of nuclear P-PTEN within the nucleus. The augmented pool of nuclear PTEN under oxidative stress may lead to increased chances to interact with p53 and further regulates p53 level and activity (Fig. (Fig.11).
In order to distinguish the function of the nuclear PTEN from the cytoplasmic PTEN, we tagged an NLS or NES to a PTEN-WT expression construct to target PTEN to either the nuclear or cytoplasmic compartment, respectively (17). Also, we constructed the phosphatase-dead PTEN-GR mutant with either NLS or NES to test whether PTEN′s phosphatase activity plays a functional role in different cell compartments (Fig. (Fig.3A).3A). To further differentiate the effects of the cytoplasmic PTEN-AKT-MDM2 axis on p53 regulation (4, 25, 26), we expressed above-mentioned PTEN constructs in p53−/−; Mdm2−/− cells (17). Both NLSWT and NLSGR PTEN substantially enhanced p53 and p21 levels, whereas NESWT had more pronounced effects on inducing Bax (Fig. (Fig.3B).3B). In p53+/+ MEF cells, we showed that nuclear localization of PTEN, irrespective of its phosphatase activity, was sufficient for induction of G1 cell cycle arrest. On the other hand, PTEN′s phosphatase activity was more efficient for increasing cell apoptosis (Fig. (Fig.3C).3C). In comparison to p53+/+ MEFs, the apoptotic response was less significant in isogenic p53−/− MEF cells (Fig. (Fig.3D,3D, left panel). Furthermore, the induction of G1 growth arrest by NLSGR PTEN expression was diminished in the p53-deficient background (Fig. (Fig.3D,3D, right panel). These data suggest that p53 is an important factor for the function of nuclear PTEN, especially for PTEN lacking its phosphatase activity.
It has been proposed that p53 can decrease intracellular ROS levels (29) and therefore protect the genome from oxidative stress-induced DNA damage. Thus, we asked whether increased PTEN nuclear localization in response to oxidative stress would mediate this process. Upon H2O2 treatment, a decrease in ROS level was observed in cells with PTEN380D but not with PTEN380A expression (Fig. (Fig.4A),4A), in a manner that is p53 dependent (Fig. (Fig.4B).4B). Furthermore, NLS PTEN expression correlated with twofold increase in the expression of p53 downstream antioxidant gene Sestrin (29), as compared to NES PTEN (Fig. (Fig.4D),4D), and more significantly reduced cellular ROS (Fig. (Fig.4C).4C). Consequently, nuclear PTEN appeared to reduce oxidative stress-induced DNA damage, as indicated by a decrease in the percentage of 8-oxo-dG-positive cells (Fig. (Fig.4E4E).
Since prostate tumors are known to be constantly exposed to oxidative stress (11), we further investigated the in vivo functions of nuclear PTEN by introducing aforementioned PTEN constructs into the human prostate cancer cell line, C4-2 LNCaP (p53+/+; Pten−/−). Using C4-2 prostate cancer xenografts, we found that NLSWT and NESWT PTEN acted similarly in inhibiting tumor growth (Fig. (Fig.5A),5A), even when coexpressed with a membrane-bound constitutively activated form of AKT (Myr-AKT) (Fig. (Fig.5B).5B). Interestingly, the NLSGR mutant had substantially greater antitumor activity than its NESGR counterpart (Fig. (Fig.5A)5A) and such antitumor activity could be diminished but not abolished by Myr-AKT coexpression (Fig. (Fig.5B).5B). Consistent with the antitumor activity, NLSWT, NESWT, and NLSGR PTEN significantly enhanced p53 levels and tumor apoptosis (Fig. 5C and D). These data strongly support our in vitro observations and indicate that nuclear PTEN, independent of its phosphatase activity, is able to activate p53 and suppresses tumor growth. To further address the interplay between nuclear PTEN and p53 in tumor suppression, we employed the PC3 human prostate cancer cell line that is null for both p53 and PTEN. As shown in Fig. Fig.6A,6A, the antitumor growth effect of NLSGR PTEN observed in C4-2 was completely abolished in PC3 xenografts (Fig. (Fig.6A).6A). However, both NLSWT and NESWT PTEN still effectively suppressed tumor growth and induced apoptosis compared to NLSGR and NESGR (Fig. (Fig.6B),6B), suggesting phosphatase activity could prominently contribute to the antitumor effect of nuclear PTEN in the absence of p53.
In this study, we demonstrated that oxidative stress leads to PTEN nuclear accumulation by attenuating the nuclear export of phosphorylated PTEN. Nuclear PTEN, independent of its phosphatase activity, increases p53 levels, thereby enhancing p53 function. We propose that the increased PTEN nuclear localization in response to oxidative stress may protect the cells from DNA damage and tumorigenesis by modulating p53-dependent ROS reduction, cell cycle arrest, apoptosis, and possibly DNA damage repair (Fig. (Fig.77).
Oxidative stress can cause severe damage to DNA, lipids, and proteins (24). Under oxidative stress, p53 is able to suppress cellular ROS level and maintain genomic stability by inducing gene products responsible for cell cycle arrest (p21 and GADD45a) and antioxidation (Sestrins) (29, 38). Our study suggests that nuclear PTEN can further enhance p53-mediated cell cycle arrest and ROS reduction, at least in part, independent of its phosphatase activity.
Recent studies have explored extensively the regulations and functions of PTEN subcellular localization in different cell lines and conditions. Combined data from Gil et al. (13), Denning et al. (8), and Trotman et al. (35) suggest that mutation of the N-terminal cytoplasmic anchor (PIP2 binding/adjacent cytoplasmic localization signal) and ubiquitination at K289 and K13 contribute to PTEN nuclear import. Interestingly, we also found that constitutively phosphorylated nuclear PTEN (380D) acquired high monoubiquitination, whereas the nonphosphorylated nuclear PTEN (380A) showed diminished ubiquitination (data not shown). This may elucidate an important association between phosphorylation and ubiquitination of PTEN in the nucleus. However, the regulation of different posttranslational modifications versus stability and activity of PTEN in the nucleus requires further mechanistic investigations.
Substantial evidence also suggests the correlations among PTEN phosphorylation, subcellular localization, and potential nuclear functions. First, several kinases can phosphorylate PTEN at its C2 domain and influence PTEN stability/activity, including CK2 (33) and glycogen synthase kinase 3β (3). In addition, Rho kinase (ROCK) phosphorylates several threonine/serine residues (S229, T223, T319, T321) in the PTEN C2 domain and delocalizes PTEN from the front edge of Dictyostelium during chemotaxis (19). Of note, P-PTENS380 expresses at a high level in the nucleus of quiescent hematopoietic stem cells (41). Our study suggests that PTEN nuclear accumulation is regulated by S380 phosphorylation status: P-S380 mediates rapid exportation of nuclear PTEN to the cytosol. Moreover, upon oxidative stress, more P-PTEN can accumulate in the nucleus, bind to p53, and enhance p53-mediated functions. Recently, it was demonstrated that PTEN can undergo nuclear export through a feedback regulation of the PI3K-S6K pathway (21). However, the detailed mechanisms of how oxidative stress or other signaling pathways regulate PTEN export systems remain to be further investigated.
Our study also reveals that nuclear P-PTEN associates with and enhances p53 function upon oxidative stress. Additionally, a recent study (17) provided elaborate data showing nuclear PTEN is recruited to the p53-p300 complex and maintains high p53 acetylation and activation in response to DNA damage. Interestingly, the same study showed low levels of p53 acetylation and tetramer formation are required for PTEN to interact and function within the p53-p300 complex. It is possible that the p53-p300 complex may serve as an anchor for PTEN nuclear localization under cell stress.
Using prostate cancer xenograft models, we showed that nuclear PTEN regulates cell proliferation and tumorigenesis through various signaling pathways, either dependent on or independent of its phosphatase activity. Trotman et al. (35) showed that catalytically active nuclear PTEN is able to down-regulate nuclear P-AKT; nuclear P-AKT was previously known to inactivate FOXO3a and accelerate tumor progression (34). Our data indicate that nuclear PTEN′s lipid phosphatase activity can be dispensable, in the presence of p53, in suppressing the growth of human prostate cancer cell tumorigenesis. Since mutations in PTEN phosphatase domain occur in approximately 65% of PTEN mutations found in human predisposition cancers (10), our results underlie the possibility of treating PTEN phosphatase domain-mutated tumors via enhanced p53 expression or activity.
Nuclear PTEN has recently been demonstrated to control chromosome stability and DNA repair (31). Our in vitro and in vivo data provide novel findings indicating that, in addition to inducing growth arrest and apoptosis, nuclear PTEN is able to reduce oxidative damage and to sufficiently suppress tumor growth independent of its phosphatase activity. Thus, regulation of cell growth and survival through the canonical PI3K-AKT pathway may only endow partial antitumor function of nuclear PTEN. Therefore, further understanding PTEN nuclear localization and nuclear function by exploring the involved cellular partners and pathways will be important for future studies.
We thank members of our laboratories for helpful suggestions and insightful comments.
C.C. and B.V. are partially supported by NIH P50 CA86306 and NCI CA107166:01S1 (to H.W.). D.J.M. is supported by NIH F32 CA112988-02. This work was partially supported by grants from the NIH (P50 CA092131, RO1, and CA107166) and Department of Defense (DAMD PC031130 to H.W.).
Published ahead of print on 10 March 2008.