Research ArticleCancer Biology

PTEN Deficiency in Endometrioid Endometrial Adenocarcinomas Predicts Sensitivity to PARP Inhibitors

See allHide authors and affiliations

Science Translational Medicine  13 Oct 2010:
Vol. 2, Issue 53, pp. 53ra75
DOI: 10.1126/scitranslmed.3001538

Abstract

PTEN (phosphatase and tensin homolog) loss of function is the most common genetic aberration in endometrioid endometrial carcinomas. In addition to its well-described role in cell signaling, PTEN is involved in the maintenance of genomic stability. Loss of PTEN function causes defects in repair of DNA double-strand breaks by homologous recombination and, therefore, sensitizes cells to inhibition of the poly(adenosine diphosphate ribose) polymerase (PARP). Here, we determined the PTEN status of eight endometrioid endometrial carcinoma cell lines and correlated it with in vitro sensitivity to the PARP inhibitor KU0058948. PTEN-deficient cells showed a significantly greater sensitivity to KU0058948 than the two endometrioid endometrial carcinoma cell lines with wild-type PTEN. The cell lines lacking PTEN expression were unable to elicit a homologous recombination damage response as assayed by RAD51 focus function (a marker of competent homologous recombination DNA repair) upon irradiation and treatment with PARP inhibitors. PTEN silencing in PTEN wild-type Hec-1b cells resulted in reduced RAD51 foci formation after DNA damage and increased sensitivity to PARP inhibition. PTEN reexpression in PTEN-null cell lines resulted in enhanced RAD51 foci formation and in relative resistance to KU0058948. Given that up to 80% of endometrioid endometrial cancers lack PTEN expression, our results suggest that PARP inhibitors may be therapeutically useful for a subset of endometrioid endometrial cancers.

Introduction

Endometrial cancer is the fourth most common malignancy among women in the United States with 42,160 new cases in 2009 (1). Most endometrial cancers (72%) are detected in early stage (stage I/II), whereas 20% have regional metastasis (stage III) and 8% have distant metastasis (stage IV) (2). The two most common histological types are endometrioid adenocarcinoma (75 to 80%) and serous carcinoma (<10%) (3). Endometrial clear cell and mucinous carcinomas are rare and together account for ~5% of cases. Systemic treatments for advanced or metastatic endometrial cancer yield minimal and short-lived responses, and are accompanied with high toxicity (4, 5). Targeted treatments have been tested in recent years, but have not been adopted for routine clinical practice. Therefore, effective targeted treatments with a favorable side effect profile are required.

Loss of function of the tumor suppressor gene PTEN (phosphatase and tensin homolog), located on chromosome 10q23, is the most common genetic aberration in endometrioid endometrial cancers (EECs) and is seen in up to 80% of cases (6). PTEN encodes a protein with a tensin-like domain and a catalytic domain, which is similar to that of dual-specificity protein tyrosine phosphatases. PTEN plays a pivotal role in the regulation of the phosphatidylinositol 3-kinase (PI3K)–AKT–mammalian target of rapamycin (mTOR) signaling pathway by modulating the intracellular levels of phosphatidylinositol 3,4,5-trisphosphate (68). PTEN loss of function may be caused by a variety of mechanisms, mainly gene mutation or deletion (6, 9), although gene promoter methylation is also an alternative mechanism of PTEN inactivation, mostly found in metastatic cancers (10). Additionally, several microRNAs (miR-214, miR-221, and miR-222) and a pseudogene (PTENP1) have been suggested to down-regulate PTEN in various cancer types including endometrial cancer (1113).

Apart from its functions in regulating the PI3K-AKT-mTOR pathway, PTEN loss of function causes genetic instability. Embryonic Pten−/− mouse cells show genomic instability, a phenotype attributed to either defects in homologous recombination (HR) DNA repair of double-strand breaks (DSBs) (14) or defects in cell cycle checkpoints (15).

Preclinical studies (1619) have demonstrated that cancer cells with defective HR DNA repair, caused by either BRCA1 or BRCA2 inactivating mutations, display an exquisite sensitivity to inhibitors of the poly(adenosine diphosphate ribose) polymerase (PARP) inhibitors. PARP is an enzyme that plays a pivotal role in the repair of single-strand breaks (SSBs); in the presence of a DNA SSB in normal cells, PARP works as a molecular beacon to attract the other components of the DNA repair machinery. When PARP is inhibited, SSBs are not corrected and, during S phase, these errors lead to stalling and collapse of the replication forks, resulting in the persistence of DNA DSBs (20). Results of phase I and II clinical trials with PARP inhibitors in advanced breast and ovarian cancer in BRCA1 and BRCA2 mutation carriers have been promising, with objective and sustained clinical response to monotherapy with a potent oral PARP inhibitor (2123). PARP inhibitors cause minimal toxicity in comparison to standard chemotherapy (2123). This has led to the design of several clinical trials to test PARP inhibitors for patients with cancers harboring BRCA1/BRCA2 mutations (21, 2426).

The HR DNA repair defect caused by PTEN deficiency renders cancer cells sensitive to PARP inhibitors, both in vitro and in vivo (27, 28). Given that most (~80%) EECs harbor loss of function of PTEN and somatic mutations of BRCA1 or BRCA2 are rare events in EECs (29), we investigated whether EEC cell lines are sensitive to PARP inhibition due to loss of PTEN.

Results

Molecular characterization of EEC cell lines reveals that PTEN loss of expression is a common event

Immunohistochemical analysis of EEC cell lines (Table 1) revealed that PTEN was expressed in the nucleus in Hec-1b and EFE-184 cells, whereas the remaining cell lines were devoid of PTEN expression, apart from EN, which showed cytoplasmic but not nuclear PTEN staining (Table 1). All cell lines, except EFE-184 and RL95-2, displayed loss of expression of at least one mismatch repair gene as determined by immunohistochemistry. RL95-2 displayed microsatellite instability (MSI)–high status but showed protein expression of all mismatch repair proteins tested (Table 1). Because specificity and sensitivity of immunohistochemical analysis of PTEN are controversial (30), we tested PTEN protein expression by Western blotting using an antibody reacting against a C-terminal epitope, which confirmed the observation that all cell lines, except Hec-1b and EFE-184, lacked full-length PTEN expression (Table 1 and Fig. 1A). Nonsynonymous PTEN mutations were detected in five of the six cell lines lacking PTEN expression. Nou-1 cells did not express PTEN messenger RNA (mRNA). Fluorescence in situ hybridization (FISH) demonstrated a homozygous deletion of the PTEN locus in this cell line (fig. S1).

Table 1

Characterization of EEC cell lines. Neg, negative; Pos, positive.

View this table:
Fig. 1

EEC cell lines lacking PTEN expression are sensitive to inhibitors of PARP. (A) PTEN expression in EEC cell lines detected by Western blotting (β-actin control). (B) RAD51 foci formation presented as percentage of cells with γ-H2AX foci formation in untreated cells (white bars), 6 hours after 10 Gy of irradiation (gray bars) or in the presence of 10 μM KU0058948 for 24 hours (black bars). The Capan-1 cell line harbors a BRCA2 mutation (mut) and was included as a positive control for RAD51 foci formation deficiency. A minimum of 100 cells was counted to determine RAD51 focus formation frequency in three independent experiments. Error bars, SD. P value (unpaired t test) represents significance of difference in survival fraction for cell lines with PTEN mutation relative to PTEN wild-type (wt) cell lines. (C) Fourteen-day SRB colorimetric assay done in triplicates for cytotoxicity of eight endometrial cancer cell lines treated with PARP inhibitor (KU0058948); survival was expressed relative to that of DMSO-treated cells. Error bars, SEM. Red, PTEN wild-type cell lines; blue, cell lines harboring PTEN mutations. P value (unpaired t test) represents significance of difference in survival fraction for cell lines with PTEN mutation relative to PTEN wild-type cell lines. (D) RAD51 protein expression in protein extracts of the cytoplasmic (c) and nuclear (n) fractions of endometrial cancer cell lines. β-Tubulin, p84, and histone H2A were used as loading controls for cytoplasmic, nuclear, and chromatin-bound proteins.

Of the eight cell lines tested, only Ishikawa cells expressed the estrogen receptor (ER), whereas p53 nuclear expression was observed in all cell lines. Microarray comparative genomic hybridization (aCGH) and gene expression data of the cell lines revealed genomic aberrations and transcriptomic profiles similar to those of primary human EECs (figs. S2 and S3 and table S1) (31). aCGH analysis revealed recurrent gains of chromosomes 1q, 3q, and 8, as well as recurrent losses in chromosomes 4, 13, 15, and X, which are genomic regions reported to be altered in primary EEC (31). On the basis of gene expression analysis, all eight cell lines were predicted to have an expression profile consistent with those of human EECs and significantly distinct from those of non-EECs (table S1).

Together, these results provide strong circumstantial evidence to suggest that the EEC cell lines used in this study have molecular features that recapitulate those of primary human EECs (6, 3234), and in particular, as with human EECs (6, 35, 36), most of these cell lines are deficient for PTEN (Fig. 1A).

Endometrial cancer cell lines lacking PTEN expression cannot elicit HR DNA repair responses

To test whether loss of PTEN expression in EEC cell lines is associated with the lack of ability to elicit RAD51 foci formation (that is, a surrogate of competent HR DNA repair), we measured RAD51 foci formation after 10 Gy of irradiation, using confocal microscopy as previously described (3739). Cell lines lacking PTEN (Ishikawa, Nou-1, and AN3-CA) displayed significantly fewer irradiation-induced RAD51 foci than cells with PTEN expression (Fig. 1B). Capan-1, a BRCA2-mutant pancreatic cell line, which is deficient in HR and unable to elicit RAD51 foci formation after DNA damage, was used as a negative control (fig. S4) (37).

Lack of PTEN expression is associated with sensitivity to a potent PARP inhibitor

We next determined whether lack of competent HR due to loss of PTEN expression might sensitize endometrial carcinoma cells to PARP inhibition. Upon treatment with the potent PARP inhibitor KU0058948 (16), cell lines lacking PTEN expression displayed significantly fewer RAD51 foci (Fig. 1B) and a significantly higher sensitivity to KU0058948 (Fig. 1C) than cells with wild-type PTEN expression in a 14-day colony formation assay. However, the levels of phospho-γ-H2AX foci formation after PARP inhibitor treatment did not differ between the two groups (fig. S5), indicating that the PARP inhibitor caused similar amounts of DNA damage regardless of PTEN status. RAD51 protein was primarily present in the cytoplasm rather than in the nucleus in all cell lines studied regardless of PTEN status except for Nou-1 (Fig. 1D). Protein extracts from MCF-7 before and after irradiation served as a positive control for the detection of nuclear RAD51 (fig. S6). Given that RAD51 foci formation occurs mainly during S phase of the cell cycle (16), we tested whether differences in cell cycle profile rather than in ability to form RAD51 foci might explain these observations. However, there was no correlation between the proportion of cells in S phase and that showing RAD51 foci after DNA damage (fig. S7).

To determine whether PTEN might be required for EEC cells to elicit HR responses in the presence of DSBs and to be resistant to PARP inhibitors, we silenced PTEN using validated short hairpin RNA (shRNA) constructs in EEC cells with wild-type PTEN. As expected, PTEN silencing led to increased activation of the PI3K-AKT-mTOR pathway, demonstrated by increased phosphorylation of AKT, mTOR, and S6 ribosomal protein (RPS6) (Fig. 2A), and to a significantly higher sensitivity to PARP inhibition, which was at least in part associated with the level of PTEN silencing (Fig. 2B). In addition, the ability to elicit RAD51 foci formation upon treatment with ionizing radiation or the PARP inhibitor KU0058948 was diminished in the PTEN shRNA-silenced sublines (Fig. 2, C and D).

Fig. 2

PTEN silencing leads to activation of downstream targets, sensitivity to the PARP inhibitor KU0058948, and impaired ability to elicit RAD51 foci formation in EEC cells. (A) Expression of PTEN and total and phosphorylated AKT, RPS6, and mTOR in PTEN-silenced Hec-1b cell line using two independent shRNA libraries detected by Western blotting. Proteins were extracted after 7 days of antibiotic selection after transfection. (B) Effect of shRNA-mediated PTEN silencing in Hec-1b cells on the sensitivity to the PARP inhibitor KU0058948, as defined by the 14-day SRB colorimetric assay done in triplicate. SF50, 50% survival fraction inhibitor concentrations. Red, shControl-infected cells; blue, cells infected with short hairpin libraries for PTEN. Error bars, SEM. (C) RAD51 foci formation as a percentage of cells with phospho-γ-H2AX foci formation without treatment, 6 hours after 10 Gy of irradiation, and with 10 μM KU0058948 for 24 hours. A minimum of 100 cells was counted to determine RAD51 focus formation frequency in three independent experiments. Error bars, SD. P value (unpaired t test) represents significance of difference in RAD51 focus formation for cell lines with PTEN silencing relative to the shControl cell line. (D) Immunofluorescence images of representative fields of Hec-1b cells with wild-type PTEN cells or PTEN-silenced Hec-1b cells using two independent shRNA libraries showing DAPI-stained nuclei (blue), γ-H2AX (green), and RAD51 foci (red) after 10 Gy of irradiation.

Next, two cell lines lacking PTEN function, Nou-1 and Ishikawa, were transduced with wild-type PTEN. Reexpression of full-length wild-type PTEN led to a reduction of activated AKT, mTOR, and RPS6 (Fig. 3A), increased ability to elicit RAD51 foci in the presence of DNA DSBs (Fig. 3, B and C), and significantly decreased sensitivity to KU0058948 (Fig. 3D).

Fig. 3

Reexpression of PTEN restored the ability to elicit RAD51 foci formation and causes relative resistance to the PARP inhibitor KU0058948 in EEC cells. (A) PTEN and total and phosphorylated AKT, RPS6, and mTOR expression of Nou-1 and Ishikawa after transfection with an empty pBabe-puro vector and one containing wild-type PTEN by Western blot. Proteins were extracted after 7 days of antibiotic selection after transfection. (B) RAD51 foci formation shown as a percentage of cells with phospho-γ-H2AX foci formation without treatment, 6 hours after 10 Gy of irradiation, and with 10 μM KU0058948 for 24 hours. A minimum of 100 cells was counted to determine RAD51 focus formation frequency in three independent experiments. Error bars, SD. P value (unpaired t test) represents significance of difference in RAD51 focus formation for cell lines with PTEN reexpression relative to the empty vector cell lines. (C) Immunofluorescence pictures using confocal microscopy showing DAPI, γ-H2AX, and RAD51 foci after 10 Gy of irradiation. (D) Effect of PTEN reexpression in Nou-1 and Ishikawa on the sensitivity to the PARP inhibitor KU0058948, as defined by the 14-day SRB colorimetric assay performed in triplicate. Error bars, SEM. Blue, cell lines with PTEN loss of function; red, cell lines with reexpression of wild-type PTEN.

Together, these results demonstrate that (i) EEC cells lacking PTEN function are unable to elicit competent HR DNA repair in the presence of DNA DSBs and are relatively sensitive to PARP inhibitors, (ii) PTEN silencing significantly reduces the ability of PTEN wild-type EECs to elicit RAD51 foci formation and increases the sensitivity to PARP inhibitors, and (iii) reexpression of PTEN in PTEN-mutant EEC cells leads to an increased ability to elicit RAD51 foci formation in the presence of DNA DSBs and relative resistance to PARP inhibitors (Fig. 4).

Fig. 4

Diagram illustrating the principle of synthetic lethality between PTEN loss of function and PARP inhibition in endometrial cancer. DNA damage is a common phenomenon. SSBs are normally repaired by base excision repair (BER). PARP is a key component of the BER machinery. In the presence of a single-strand break (SSB), PARP localizes to the site of breakage and acts as a molecular beacon to the other components of the BER machinery. Inhibition of PARP leads to persistence of SSBs that, when unrepaired, lead to replication fork collapse or the formation of a double-strand break (DSB) during DNA replication. In EEC cells with wild-type PTEN, these breaks are repaired by the HR DSB repair pathway. In the absence of functional PTEN, HR DNA repair is impaired, resulting in the persistence of the DNA DSB or its repair by alternative pathways that are highly error-prone [for example, nonhomologous end joining (NHEJ) or single-strand annealing (SSA)]. This results in genomic instability and cell death.

Discussion

PTEN is a tumor suppressor gene frequently inactivated in multiple different types of cancers such as glioblastoma and prostate cancer (4042). In EEC, PTEN loss of function is one of the most prevalent molecular aberrations (6). Recently, in addition to its canonical functions in cell signaling, it has been demonstrated that PTEN-null cells lack competent HR DNA repair (14, 27, 28). This type of DNA repair, which is also abrogated in tumors with BRCA1 and/or BRCA2 loss of function, is of paramount importance for the accurate repair of DNA DSBs during S phase (16). In the absence of HR DNA repair, DNA DSBs are repaired by error-prone mechanisms, such as nonhomologous end joining (NHEJ) and single-strand annealing (SSA) (24). This deficiency of HR DNA repair has been shown to sensitize cells to agents that cause DNA DSBs such as PARP inhibitors (16, 1921, 24, 43).

Here, we demonstrate that, akin to BRCA1 and/or BRCA2 loss of function, (i) PTEN loss of function in EECs is associated with lack of competent HR DNA repair and sensitivity to a PARP inhibitor, (ii) PTEN silencing leads to abrogation of RAD51 foci formation and sensitivity to a PARP inhibitor, and (iii) reexpression of PTEN in cell lines with PTEN loss of expression leads to restoration of the ability to elicit HR DNA repair and resistance to a PARP inhibitor in EEC cells. Furthermore, we demonstrate that PTEN loss of function is associated with sensitivity to a PARP inhibitor in cell lines independent of MSI.

Experimental modulation of PTEN, by either RNA interference (RNAi) or complementary DNA (cDNA) expression, causes modest changes in PARP inhibitor sensitivity (Figs. 2 and 3) when compared to differences in KU0058948 response in EEC cells with PTEN loss of function or wild-type PTEN (Fig. 1). It is conceivable that effects such as incomplete plasmid infection rate/incomplete selection of infected cells, transient and variable expression of shRNA/cDNA-coding sequences, and vector copy number effects may all affect the scale of sensitivity/resistance observed (Figs. 2 and 3). An alternative is that genetic background effects in different cell lines might also affect sensitivity. Nevertheless, the significant changes in PARP inhibitor sensitivity caused by modulating PTEN levels (Figs. 2 and 3) firmly establish the causative nature of PTEN in determining PARP inhibitor sensitivity in EEC models and agree with previously published data examining KU0058948 sensitivity in other types of cancer cells (28).

PTEN has been suggested to be in involved in the HR DNA repair pathway through transcriptional regulation of RAD51 (14, 28). Similar to the observations of others on glioblastoma cell lines (27), our results demonstrate that in type I EEC cell lines, RAD51 was not expressed at lower levels in cell lines with PTEN loss of function. To rule out differences in expression of RAD51 in the cytosol and the nucleus in the different cell lines, we separated the subcellular fractions and observed RAD51 expression primarily in the cytosol in untreated cells, as previously demonstrated (44, 45). After induction of DNA damage through irradiation or PARP inhibition, PTEN wild-type cells displayed the ability to form RAD51 foci in the nucleus, whereas PTEN-mutated cells did not. These findings suggest that in EEC cells, PTEN may play a role in the localization of RAD51 to the nucleus upon DNA damage rather than in regulation of RAD51 expression as previously suggested (14, 28); further studies are warranted to test this hypothesis. Note, however, that we have confirmed previous observations (14, 27, 28, 46) that the ability of cancer cells to elicit RAD51 foci formation upon DNA damage is a predictor of sensitivity to PARP inhibitors. RAD51 foci formation has been shown to be predictive for response to PARP inhibitors in xenografts of cancer cells with BRCA2 mutations (16) and in a recent in vivo study using PTEN-null and wild-type isogenic cells (28). Furthermore, RAD51 foci formation has also been shown by independent groups to be a functional predictor of response to PARP inhibitors in ovarian cancer cells and of response to neoadjuvant chemotherapy in breast cancer (46, 47).

The genome-wide aCGH and transcriptomic analyses and the comprehensive immunohistochemical profiling of the cell lines used in this study demonstrate that they recapitulate the cardinal molecular features of primary human EECs (31, 34). Furthermore, the frequency of MSI and PTEN mutations observed in the cell lines closely matches that of primary EECs (34). These observations provide evidence to suggest that these cell lines represent a useful model for in vitro studies testing therapeutic approaches for type I (endometrioid) endometrial cancers. In addition, there is also evidence that these cell lines recapitulate the properties of human endometrial cancers in xenograft studies (48, 49).

Current systemic therapies for patients with advanced endometrial cancer are of limited clinical benefit and, in the case of chemotherapy, are associated with significant side effects (4). This is particularly relevant given the high incidence of endometrial cancer around the seventh decade of life (3). Because most human EECs lack PTEN expression and PTEN-null endometrial adenocarcinoma cell lines are exquisitely sensitive to a potent PARP inhibitor, our results provide strong circumstantial evidence to suggest that this class of agents may constitute a useful therapeutic approach for a subgroup of EECs, and a molecular rationale to test PARP inhibitors for the management of patients with advanced EECs lacking PTEN expression.

Materials and Methods

Cancer cell lines

The EEC cell lines RL95-2, AN3-CA, Hec-1b, Ishikawa, Nou-1 (R. Zeillinger, Austria), Hec-59 (A. Fedier, Switzerland), Capan-1 (purchased from the American Type Culture Collection), EN, and EFE-184 [purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ)] were grown in Dulbecco’s modified Eagle’s or RPMI 1640 media (both Gibco, Invitrogen) supplemented with 10% (v/v) fetal bovine serum (PAA Gold, Gibco, Invitrogen), penicillin (100 U/ml), and streptomycin (60 μg/ml) and maintained at 37°C in a humidified atmosphere at 5% CO2. DNA from cell lines was extracted with the DNeasy kit (Qiagen Ltd.) according to the manufacturer’s recommendations. RNA was extracted at 80% confluency with the Trizol method (Invitrogen). In all cell lines, short tandem repeat profiling was performed to rule out misidentification of cell lines (PowerPlex 1.2 System, Promega). All profiles were matched with the DSMZ database.

Gene expression profiling and aCGH

All cell lines were subjected to mRNA gene expression profiling (Illumina Human WG6_v3.0 Bead-Array) and aCGH (Breakthrough Breast Cancer Research Centre 32K BAC array platform) as described (50, 51). Data acquisition and analysis were fully MIAME (minimum information about a microarray experiment)–compliant and are described in the Supplementary Material.

PTEN mutation analysis

Sequencing of the full-length cDNA of PTEN, derived from mRNA, was performed for all cell lines with five pairs of primers (table S2) as previously described (50). Cancer cell line cDNA (50 ng) was amplified, and sequencing reactions were carried out with the DNA Sequencing Kit BigDye Terminator v 1.1 Cycle Sequencing Ready Reaction Mix (Applied Biosystems), as previously described (50). Sequences were analyzed with Mutation Surveyor software (SoftGenetics). Mutations were confirmed by repeat polymerase chain reaction (PCR) and sequencing of forward and reverse strands.

PTEN gene copy number by FISH

Cell pellets of Nou-1 and Hec-1b were washed twice with phosphate-buffered saline and fixated with methanol/acetic acid solution (3:1) onto slides. The Vysis LSI PTEN (10q23)/chromosome 10 centromere (CEP 10) Dual Color Probes (Abbott Molecular) were hybridized to representative slides of the cell lines according to the manufacturer’s instructions. Signals were counted in 50 nonoverlapping nuclei with the Leica TCS SP2 confocal microscope.

MSI analysis

MSI-high status was determined with three microsatellite markers (BAT25, BAT26, and BAT40). Standard PCR conditions, the ABI 377 or ABI 3100 sequencer, and Genotyper software were used as previously described (52).

Analysis of RAD51 foci formation

Nuclear γ-H2AX and RAD51 foci were visualized and quantified by confocal microscopy as previously described (16) and used as surrogate markers for induction of DNA DSBs and competent HR DNA repair, respectively. In brief, cells were grown onto l-lysine–coated coverslips (BD Biosciences) and exposed to either 10 μM PARP inhibitor KU0058948 (AstraZeneca) for 24 hours or 10 Gy of ionizing radiation. After 6 hours, cells were fixed, permeabilized, and co-immunostained with primary antibodies targeting RAD51 (polyclonal rabbit antibody, H-92; Santa Cruz Biotechnology) and γ-H2AX (mouse monoclonal, #05-636; Millipore) and detected with fluorescein isothiocyanate– and Texas red–conjugated secondary antibodies (Molecular Probes, Invitrogen), respectively. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI). The presence of γ-H2AX and RAD51 foci was evaluated in a minimum of 100 cells in three independent experiments with a Leica TCS SP2 confocal microscope and Leica Confocal Software 2.61 (Leica Microsystems Heidelberg GmbH) by two observers. Cells with more than five RAD51 foci were considered positive as previously described (28). RAD51 focus formation frequency was calculated as a percentage of cells showing γ-H2AX foci.

Immunoblotting

Whole-cell protein extracts were prepared from cells lysed with an NP-40 lysis buffer supplemented with protease inhibitor cocktail tablets (Roche) and phosphatase inhibitors (Sigma-Aldrich). Nuclear and cytoplasmic proteins were extracted with the ProteoExtract subcellular proteome extract kit (Merck KGaA). Protein concentrations were measured with Bio-Rad Protein Assay Reagent. Western blotting was performed as described previously (51) with antibodies against PTEN (rabbit monoclonal antibody against the C terminus, 138G6; Cell Signaling Technology), RAD51 (rabbit polyclonal antibody, H-92; Santa Cruz Biotechnology), β-tubulin (mouse monoclonal antibody, T4026; Sigma), β-actin [rabbit monoclonal antibody conjugated with horseradish peroxidase (HRP) against the N terminus, 13E5; Cell Signaling Technology], p84 (mouse monoclonal antibody, 5E10; Abcam), histone H2A (rabbit polyclonal, ab13923; Abcam), phospho-mTOR (Ser2448) (rabbit polyclonal antibody, 2971; Cell Signaling Technology), mTOR (rabbit monoclonal antibody, 7C10; Cell Signaling Technology), phospho-AKT (Ser473) (rabbit monoclonal antibody, 193H12; Cell Signaling Technology), AKT (mouse monoclonal antibody, 2H10; Cell Signaling Technology), RPS6 (rabbit monoclonal antibody, 2217; Cell Signaling Technology), and phospho-RPS6 (Ser235/236) (rabbit monoclonal antibody, 2211; Cell Signaling Technology). Incubation with primary antibody, except for β-actin, was followed by incubation with an HRP-conjugated secondary antibody (Pierce Biotechnology) and chemiluminescent detection of proteins (Amersham Pharmacia).

Drug sensitivity assay

Long-term survival assays were performed as previously described (16, 37). For measurement of sensitivity to the PARP inhibitor KU0058948, cells were seeded in six-well plates at a concentration of 1000 to 2000 per well and treated with the inhibitor after 24 hours and then continuously exposed to the drug (10−8 to 10−4 M) dissolved in dimethyl sulfoxide (DMSO). The control was treated with DMSO. Media and drug were replaced every 3 to 5 days. After 10 to 15 days, cells were fixed and stained with sulforhodamine B (SRB) (Sigma) and a colorimetric assay was performed as described previously (37). Surviving fractions were calculated compared to the DMSO-treated control, and drug sensitivity curves were plotted as previously described (16).

Generation of stable cell lines

For PTEN reexpression, a validated full-length wild-type PTEN plasmid cloned into tag-free pBabe-puro retroviral expression vector and the empty pBabe-puro vector as a negative control were used as previously described (28). Transfections were carried out with FuGene (Roche) according to the manufacturers’ instructions. For PTEN silencing, two PTEN shRNA libraries were used. One library (shPTEN1) contained three shRNA clones and the other library (shPTEN2) contained eight clones, all cloned into pGIPZ vector (Open Biosystems; table S3). Briefly, cells were transduced with lentivirus containing either library or control vector. Twenty-four hours after transfection or transduction, cells were selected in puromycin (1 mg/ml) (InvivoGen) for 7 days to remove noninfected cells, then seeded in six-well plates (at 1500 cells per well), and treated with the PARP inhibitor the following day, as previously described (28).

Immunohistochemistry

At 80% confluency, cells were trypsinized and cell pellets were fixed in 4% buffered formalin overnight and embedded in paraffin. Each cell line was arrayed into a cell line microarray block containing two representative replicate 0.6-mm cores from each of the eight endometrial cancer cell lines. Sections (4-μm thick) of the cell line were cut and subjected to immunohistochemical analysis with antibodies against PTEN, ER, p53, MLH1, MSH2, MSH6, and PMS2. Antibodies, antigen retrieval methods, and cutoffs are described in table S4. All markers were scored by two pathologists (F.C.G. and J.S.R.-F.) blinded to the results of the Western blotting.

Supplementary Material

www.sciencetranslationalmedicine.org/cgi/content/full/2/53/53ra75/DC1

Methods

Fig. S1. Fluorescence in situ hybridization images of PTEN on Nou-1 and Hec-1b.

Fig. S2. Genome plots of eight endometrioid endometrial cancer cell lines derived from microarray-based comparative genomic hybridization.

Fig. S3. Frequency of gains and losses in eight endometrioid endometrial cancer cell lines.

Fig. S4. Immunofluorescence images of eight endometrioid endometrial cell lines showing γ-H2AX and RAD51 foci after irradiation.

Fig. S5. Phospho-γ-H2AX foci formation after irradiation and PARP treatment in eight endometrioid endometrial cancer cell lines.

Fig. S6. RAD51 protein expression in the cytoplasmic and nuclear subcellular fractions of MCF-7 before and after of ionizing irradiation.

Fig. S7. Cell cycle analysis after 10 Gy of irradiation in endometrial cancer cell lines.

Table S1. Posterior class probability from prediction analysis for microarrays (PAMs) for histological subtypes of endometrial cancers.

Table S2. Primers for PCR and sequencing of PTEN cDNA.

Table S3. Specifications of short hairpin RNA (shRNA) used for PTEN silencing.

Table S4. Summary of antibodies, clones, dilutions, and antigen retrieval methods.

References

Footnotes

  • Citation: K. J. Dedes, D. Wetterskog, A. M. Mendes-Pereira, R. Natrajan, M. B. Lambros, F. C. Geyer, R. Vatcheva, K. Savage, A. Mackay, C. J. Lord, A. Ashworth, J. S. Reis-Filho, PTEN deficiency in endometrioid endometrial adenocarcinomas predicts sensitivity to PARP inhibitors. Sci. Transl. Med. 2, 53ra75 (2010).

References and Notes

  1. Acknowledgments: We thank R. Zeillinger and A. Fedier for the endometrial cancer cell lines and B. Weigelt for her insightful comments. Funding: Supported by Breakthrough Breast Cancer and a Swiss National Science Foundation fellowship (K.J.D.). Author contributions: K.J.D., D.W., A.M.M.-P., C.J.L., A.A., and J.S.R.-F. conceived and designed the experiments. K.J.D., D.W., R.N., M.B.L., F.C.G., R.V., K.S., and A.M. performed the experiments. K.J.D., D.W., A.M.M.-P., A.A., and J.S.R.-F. analyzed the experiments. K.J.D., D.W., A.A., and J.S.R.-F. wrote the manuscript. J.S.R.-F. and A.A. devised the project and obtained funding. All authors revised and approved the final version of the manuscript. Competing interests: C.J.L. and A.A. may benefit financially from the development of PARP inhibitors through patents held jointly with KuDOS-AstraZeneca through the Institute of Cancer Research “rewards to inventors” scheme. Accession numbers: Gene expression data were deposited in ArrayExpress (accession number: E-TABM-992) and aCGH data on ROCK (http://brcabase.icr.ac.uk/browse/browse_acgh.jsp/).
View Abstract

Navigate This Article