Original Paper

Oncogene (2005) 24, 1882–1894. doi:10.1038/sj.onc.1208368 Published online 24 January 2005

Loss of the tumor suppressor gene PTEN marks the transition from intratubular germ cell neoplasias (ITGCN) to invasive germ cell tumors

Dolores Di Vizio1,5, Letizia Cito2,5, Angelo Boccia2, Paolo Chieffi3, Luigi Insabato1, Guido Pettinato1, Maria Letizia Motti2, Filippo Schepis4, Wanda D'Amico4, Fernanda Fabiani4, Barbara Tavernise4, Salvatore Venuta4, Alfredo Fusco2 and Giuseppe Viglietto2,6

  1. 1Dipartimento di Scienze Biomorfologiche e Funzionali, Facoltà di Medicina e Chirurgia, Università di Napoli 'Federico II', via S. Pansini, 5, 80131 Napoli, Italy
  2. 2Istituto di Endocrinologia ed Oncologia Sperimentale & Dipartimento di Biologia e Patologia Cellulare e Molecolare, Facoltà di Medicina e Chirurgia, Università 'Federico II', via S. Pansini 5, 80131 Naples, Italy
  3. 3Dipartimento di Medicina Sperimentale, Seconda Università di Napoli, via Costantinopoli 16, 80138 Naples, Italy
  4. 4Dipartimento di Medicina Sperimentale e Clinica, Viale Europa-Germaneto, 88100 Università 'Magna Graecia' Catanzaro, Italy

Correspondence: G Viglietto, E-mail: viglietto@sun.ceos.na.cnr.it

5These authors equally contributed to this work

6Current address: Laboratorio Oncologia Molecolare III, Dipartimento di Medicina Sperimentale e Clinica, Campus Universitario di Germaneto-Viale Europa, 88100 Università 'Magna Graecia' Catanzaro, Italy

Received 15 November 2004; Accepted 16 November 2004; Published online 24 January 2005.



PTEN/MMAC1/TEP1 (hereafter PTEN) is a tumor suppressor gene (located at 10q23) that is frequently mutated or deleted in sporadic human tumors. PTEN encodes a multifunctional phosphatase, which negatively regulates cell growth, migration and survival via the phosphatidylinositol 3'-kinase/AKT signalling pathway. Accordingly, Pten+/- mice develop various types of tumors including teratocarcinomas and teratomas. We have investigated PTEN expression in 60 bioptic specimens of germ cell tumors (32 seminomas, 22 embryonal carcinomas and six teratomas) and 22 intratubular germ cell neoplasias (ITGCN) adjacent to the tumors for PTEN protein and mRNA expression. In total, 10 testicular biopsies were used as controls. In the testis, PTEN was abundantly expressed in germ cells whereas it was virtually absent from 56% of seminomas as well as from 86% of embryonal carcinomas and virtually all teratomas. On the contrary, ITGCN intensely expressed PTEN, indicating that loss of PTEN expression is not an early event in testicular tumor development. The loss of PTEN expression occurs mainly at the RNA level as determined by in situ hybridization of cellular mRNA (17/22) but also it may involve some kind of post-transcriptional mechanisms in the remaining 25% of cases. Analysis of microsatellites D10S551, D10S541 and D10S1765 in GCTs (n=22) showed LOH at the PTEN locus at 10q23 in at least 36% of GCTs (three embryonal carcinoma, three seminoma, two teratoma); one seminoma and one embryonal (9%) carcinoma presented an inactivating mutation in the PTEN gene (2/22). Finally, we demonstrated that the phosphatidylinositol 3'-kinase/AKT pathway, which is regulated by the PTEN phosphatase, is crucial in regulating the proliferation of the NT2/D1 embryonal carcinoma cells, and that the cyclin-dependent kinase inhibitor p27kip1 is a key downstream target of this pathway.


PTEN, germ cell tumors, ITGCN, p27kip1



The tumor suppressor gene PTEN encodes a multifunctional phosphatase of 403 amino acids that plays a role in the pathogenesis of different human cancers (Li et al., 1997; Li and Sun, 1997; Steck et al., 1997). Germ-line mutations in the PTEN gene cause Cowden's disease, Lhermitte-Duclos and Bannayan-Zonana syndromes, three related hamartoma disorders that predispose to an elevated risk of various cancers, including breast and thyroid cancer (Liaw et al., 1997; Marsh et al., 1997; Nelen et al., 1997). Somatic alterations of PTEN are frequently found in cell lines and such primary tumors as glioblastoma (Rasheed et al., 1997; Steck et al., 1997; Wang et al., 1997), endometrioid endometrial carcinoma (Risinger et al., 1997; Tashiro et al., 1997; Mutter et al., 2000) and prostate cancer (Cairns et al., 1997).

PTEN acts both as a protein phosphatase with dual specificity (it dephosphorylates threonine and tyrosine residues) and as a lipid phosphatase that dephosphorylates the D3 position of a special class of lipids: phosphatidylinositol-phosphate PtdIns(3,4,5)P3 and PtdIns(3,4,)P2 (Haas-Kogan et al., 1998; Maehama and Dixon, 1998; Stambolic et al., 1998). PtdIns(3,4,5)P3 and PtdIns(3,4,)P2 result from phosphatidylinositol-3-kinase (PI3K) activity (Myers et al., 1997). The generation of PtdIns(3,4,5)P3 by PI3K activates a class of signal transducing proteins, including the protein kinase B/AKT, that carry the so-called pleckstrin homology (PH) domain (Marte and Downward, 1997). Binding of the PH domain to PI3K products results in translocation of AKT to the plasma membrane, where it is activated through phosphorylation (Marte and Downward, 1997). Once activated, AKT transduces signals that regulate protein synthesis, glucose utilization, apoptosis and proliferation (Datta et al., 1999).

It appears that, thanks to its phosphoinositide 3-phosphatase-like activity, PTEN negatively regulates cell growth and survival by blocking the PI3K/AKT pathway (Haas-Kogan et al., 1998; Stambolic et al., 1998). PTEN inhibits cell growth by inducing G1 arrest in gliomas and thyroid cancer cells (Furnari et al., 1997; Lu et al., 1999; Bruni et al., 2000). Growth arrest is accompanied by an increase in the cell cycle inhibitor p27Kip1, and decreased CDK activity (Da-Ming and Hong, 1998; Bruni et al., 2000). P27Kip1 is a key target of PTEN-dependent growth-suppressing activity (Bruni et al., 2000). PTEN inhibits cell growth also by inducing apoptosis in breast cancer cells (Lu et al., 1999).

Mice heterozygous for one null Pten allele (Pten+/-) are prone to develop different types of tumors, including teratocarcinomas (Di Cristofano et al., 1998; Suzuki et al., 1998; Podsypanina et al., 1999). Furthermore, the conditional knockout of the pten gene in primordial germ cells causes the development of bilateral testicular teratomas, which resulted from impaired mitotic arrest and outgrowth of cells with immature characteristics (Kimura et al., 2003).

However, the question as to whether PTEN is involved in human germ cell tumors has not yet been addressed. Germ cell tumors of the testis (GCT) are a heterogeneous group of neoplasms seen mainly in young men (Chaganti and Houldsworth, 2000). They are classified as seminomatous (SE-GCT) and nonseminomatous (NSE-GCT) tumors, both of which appear to arise from intratubular germ cell neoplasias (ITGCN) (Ulbright, 1993; Chaganti and Houldsworth, 2000). The former is constituted by neoplastic germ cells that retain the morphology of spermatogonial germ cells, whereas NSE-GCT display primitive zygotic (embryonal carcinomas), embryonal-like somatically differentiated (teratomas) and extra-embryonally differentiated (choriocarcinomas, yolk sac tumors) patterns (Ulbright, 1993; Chaganti and Houldsworth, 2000). GCTs are frequently associated with ITGCN that, often, progresses to invasive cancer (Vos et al., 1990; Houldsworth et al., 1997).

The molecular basis of germ cell malignant transformation is poorly understood. The most common genetic alterations detected in GCT and ITGCN are a triploid/tetraploid chromosomal complement and an increased copy number of 12p, which results in the hyper-expression of the product of the CCND2 gene, that is, G1 cyclin D2 (Houldsworth et al., 1997). On the other hand, GCTs are often accompanied by hyper-expression of autocrine and/or paracrine growth and angiogenic factors (Viglietto et al., 1996; Baldassarre et al., 1997).

In this study, we sought to determine whether the tumor suppressor lipid and protein phosphatase PTEN plays a role in the pathogenesis of germ cell tumors. We investigated: (1) PTEN expression in 60 male germ cell tumors (32 seminomas and 22 embryonal carcinomas and six teratomas); (2) PTEN expression in intratubular germ cell neoplasia; and (3) the effects of PTEN re-expression in an embryonal carcinoma cell line. The findings reported herein indicate that loss of PTEN expression may play a role in the development of testicular germ cell tumors and that the cyclin-dependent kinase inhibitor p27kip1 is a key PTEN target in embryonal carcinoma cells.



PTEN expression in normal testis

Normal germ cell epithelium showed positive cytoplasmic staining for PTEN, as observed in prostatic and endometrial epithelium (McMenamin et al., 1999; Mutter et al., 2000). Intense nuclear staining also occurred in several cells, as reported for thyroid and endocrine pancreatic tumor cells (Gimm et al., 2000; Perren et al., 2000), though the functional meaning of nuclear PTEN staining remains unclear. See Figure 1a for a representative experiment. In normal testis, PTEN expression was heterogeneous: the outer layer of cells (spermatogonia, spg) stained irregularly, with several cells showing positivity for PTEN expression; spermatocytes (spc) and spermatids (spt) also stained positive for PTEN antibodies (Figure 1a). Endothelial cells and Sertoli cells stained positive for PTEN (Figure 1a).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Immunoistochemical analysis of PTEN expression in normal testis, in situ neoplasia and germ cell tumors. (a) PTEN expression in human normal testis, magnification times 400. spg, spermatogonia; spc, spermatocytes; spt, spermatids; ser, Sertoli cells. (b) PTEN expression in intratubular neoplasia, magnification times 400. (c) Seminoma with rare focal and faint positivity for PTEN, magnification times 400. (d) Embryonal carcinoma negative for PTEN staining, magnification times 400. (e) Teratoma negative for PTEN staining, magnification times 400. (f) Peptide neutralization assay. Serial sections derived from the same biopsy were incubated with monoclonal anti-PTEN antibody with and without (inset) a molar excess peptide antigen, magnification times 150

Full figure and legend (472K)

PTEN expression in mouse testicular cells

The presence of few germ cells in normal human tissues hampered a detailed analysis of PTEN expression in the testis. To better define the cells in which PTEN is expressed in normal testis, immunohistochemical analysis was performed on serial sections of mouse testis using antibody against PTEN protein. PTEN protein was widely expressed in the germinal epithelium (spermatogonia, spermatocytes, spermatids) and Sertoli cells, while it was not detectable in spermatozoa (Figure 2a). The antiserum used in this study fulfils the criteria of specificity. In particular, immunoadsorption tests revealed that the labeling was totally blocked by preincubation of the antibody with 10-6 M of the cognate peptide (data not shown).

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

PTEN expression in the mouse testicular cells. (a) Localization of the PTEN protein in sections of adult mouse testis by immunocytochemistry. A representative seminiferous tubule showing staining in spermatogonia (spg), spermatocytes (spc), spermatids (spt), and Sertoli cells (ser); magnification times 400. (b) Western blot analysis of PTEN protein in mouse adult testis (lane 1), interstitium (lane 2), Sertoli cells (lane 3), and in normal mouse testis germ cells (lane 4–7) (50 mug/lane). Whole lysates were detected by anti-PTEN monoclonal serum or with anti-ERK antibodies used as internal standard. ERK antibodies recognize both ERK1 and ERK2, which are expressed at similar levels in all cell types with the exception of spermatozoa (Sette et al., 1997)

Full figure and legend (128K)

We confirmed the differential expression of PTEN in the different cell types in the mouse testis, by Western blot analysis of cell extracts from adult mouse testis fractionated in interstitial, Sertoli, spermatogonia, spermatocytes, spermatids and spermatozoa. Immunoblot analysis performed on cell types enriched in the different types of germ cells showed a single product migrating as a 55 kDa protein (Figure 2b). Among germ cells, PTEN was abundant in spermatogonia, present in spermatocytes and spermatids, absent in spermatozoa in agreement with immunohistochemical results. PTEN protein was also present in the interstitial and Sertoli extract cells (Figure 2b).

PTEN expression in germ cell tumors


ITGCN was present in 22 tumor samples. In all cases, the neoplastic cells present in ITGCN showed strong PTEN staining. As with normal germ cells, PTEN occurred both in the nuclear and in the cytoplasmic compartment of precancerous cells (Figure 1b). Interestingly, in cases in which tubules with ITGCN were entrapped inside a fully malignant tumor, strong PTEN expression was observed in the cells from ITGCN but not in the adjacent area (Figure 1b): in fact, PTEN staining was weak in the nuclei and cytoplasms of cancer cells. Conversely, endothelial cells showed moderate to strong PTEN expression, and thus served as internal positive controls.

PTEN protein expression was reduced in tumors as witnessed by the low signal obtained per single cell and by the decreased number of cells/field stained with the anti-PTEN antibody (Table 1). In particular, most embryonal carcinomas (19/22), approximately 60% of seminomas (18/32) and virtually all teratomas (6/6) showed no staining with anti-PTEN antibody. Moreover, the remaining tumors showed weak and focal PTEN staining. A representative immunodetection experiment of PTEN expression is reported in Figure 1, where PTEN-negative seminoma, embryonal carcinoma and teratoma are shown (Figure 1c, d and e, respectively).

To verify that the monoclonal anti-PTEN antibody (clone A2B1) was suitable for immunostaining experiments (Figure 1f), serial 5-mum sections of the same samples were incubated with anti-PTEN antibody with and without a 10-fold excess of a competing peptide. As shown in the inset of Figure 1f, peptide competition almost completely abolished the signal induced by the anti-PTEN antibody, demonstrating the specificity of the reaction. Similar results were obtained with another monoclonal anti-PTEN antibody (#26H9) from Cell Signaling (not shown). Recently, a testis-specific PTEN homologue, denoted PTEN2, has been described (Wu et al., 2001). However, since the C-terminal peptides used for the generation of antibodies used in this study are present in PTEN but absent in PTEN2, it is highly unlikely that the antibodies recognized PTEN2 in immunostaining.

Western blot

We next compared the immunoistochemical PTEN expression with immunoblot data. First the specificity of the monoclonal PTEN antibody to be used in immunoblot experiments was tested. As positive control, we used the breast cancer cell line MCF7, which is known to express PTEN, and as negative control we used the breast cancer cell line MDA-MB-468 that bears a hemizygous deletion of PTEN and a truncating mutation in exon 2 of the remaining allele, which results in the loss of PTEN expression (Li et al., 1997). The anti-PTEN antibody recognized a single band of 55–60 kDa only in the MCF-7 cells but not in the MDA-MB-468 cells (Figure 3a).

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Western blot analysis of PTEN expression in normal and neoplastic testis (a) The monoclonal anti-PTEN antibody recognized PTEN protein in MCF7 cells but not in MDA-MB-468 cells used as positive and negative controls, respectively. (b) PTEN expression and AKT phosphorylation in germ cell. In total, 40 mug of total proteins were resolved on 10% SDS–PAGE, transferred onto nitrocellulose filters and Western blotted with anti-PTEN monoclonal antibody, anti-phospho-Ser473 AKT and anti-total AKT. Antibodies to beta-tubulin served as loading control. Lane NT: normal testis; seminomas: lanes 1–4, 9–12; embryonal carcinomas: lanes 5–8, 13–16. Films were scanned and the intensity of bands was quantified by the NIH Image 1.57 program

Full figure and legend (54K)

We next determined PTEN expression in 16 primary germ cell tumors (eight seminomas and eight embryonal carcinomas) using immunoblotting, selecting them on the basis of PTEN expression, among the samples undergone immunohistochemical analysis. Proteins from four non-neoplastic testes served as controls (NT, normal testis). The amount of PTEN protein was high in normal testis (Figure 3b, lane 1) and low in several tumors (6/8 of seminomas and 6/8 embryonal carcinomas presented low PTEN expression, respectively); see for an example Figure 3b. A good correlation between the immunostaining and immunoblot data was observed.

As PTEN activity prevents AKT activation in a variety of human tumors and cell lines (Haas-Kogan et al., 1998; Bruni et al., 2000), we investigated whether the downregulation of PTEN observed in testicular tumors resulted in AKT activation, measured as increased phosphorylation at specific serine (ser473) and threonine (thr308) residues. To this end, we determined the expression and the phosphorylation status of AKT in the same representative set of tumors (Figure 3b). As expected, AKT phosphorylation on Ser473 was higher in some tumors (Figure 3b, lanes 1, 3, 4, 6, 7, 8, 10, 11, 14, 15 and 16) than in normal testis. Out of 11 tumors with low PTEN expression, 10 had high levels of phosphorylated Akt (Figure 3b, lanes: 1, 4, 6, 7, 8, 10, 11, 12, 14, 15 and 16). Anti-beta-tubulin antibody was used as a loading control. In general, PTEN expression inversely correlated with the level of phosphorylated Akt. Exceptions were tumor SE2 in lane 2, which showed low Akt phosphorylation in the presence of low PTEN expression and tumor SE3 in lane 3, which showed high Akt phosphorylation in the presence of high levels of PTEN protein. While we do not have any reasonable explanation for tumor SE2, Akt hyperexpression or activating mutations in the PI3K catalytic subunit may account for increased Akt activity in the case of tumor SE3 as recently reported (Samuels et al., 2004).

In situ hybridization

In order to assess whether loss of PTEN protein, demonstrated by immunohistochemistry and confirmed by immunoblot, was a consequence of reduced mRNA expression, we performed mRNA in situ hybridization (ISH) on a subgroup of germ cell tumors selected for being negative for the expression of PTEN protein (12 seminomas, six embryonal carcinomas, four teratomas). Results are reported in Table 2. We observed a direct correlation between the amount of PTEN protein and PTEN-specific mRNA in 75% of cases. The majority of seminomas (9/12, 75%), embryonal carcinomas (5/6, 83%) and teratomas (3/4, 100%) analysed showed reduced or no mRNA in tumor cells (Figure 4b and c). Staining for PTEN mRNA was observed instead in the samples positive for PTEN protein (not shown). In all cases, non-neoplastic atrophic tubules, adjacent to the tumor, showed nuclear and cytoplasmic staining of germinal cells, thus functioning as internal control (Figure 4a). Conversely, five out of 22 cases (23%) analysed retained PTEN mRNA despite the absence of protein expression (Table 2).

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

In situ hybridization analysis of PTEN expression in normal testis and germ cell tumors. (a) Expression of PTEN mRNA in human normal testis, magnification times 400. (b) Expression of PTEN mRNA in a PTEN deficient seminoma, magnification times 400. (c) Expression of PTEN mRNA in an embryonal carcinoma negative for PTEN staining, magnification times 400

Full figure and legend (451K)

Genetic analysis of PTEN in germ cell tumors

Loss of heterozigosity analysis

PCR-based analysis to determine LOH of markers spanning the PTEN locus was performed on a series of 22 germ cell tumor samples (12 seminomas, six embryonal carcinomas and four teratomas) and the corresponding adjacent normal tissues. Results are reported in Table 2. All samples were analysed for microsatellite markers surrounding the PTEN locus at 10q23 (D10S551, D10S1765 and D10S541). In particular, these markers present a centromere-to-telomere orientation, covering 5 MB of chromosome 10 that includes the PTEN locus. The 5' end of the PTEN gene is approximately 20 kb downstream D10S1765 and the 3' end 270 kb upstream of D10S541. Three samples were noninformative (NI) for D10S551 (see Table 2), three were non informative for D10S1765 and four for D10S541. Overall, the LOH frequency in germ cell tumors was 41% (nine of 22). Four of 19 informative samples were homozygous for D10S551, seven samples exhibited apparent LOH for D10S1765 and three for D10S541. Five samples showed LOH at two loci while the majority of samples showed LOH for just one marker. LOH was slightly more frequent in embryonal carcinomas and teratomas (3/6 and 2/4, respectively) than in seminomas (3/12).

Mutation analysis

Subsequently, we analysed the same 22 germ cell tumors for the presence of mutations in the coding region of the PTEN gene by PCR amplifying all the nine exons of the PTEN gene and subsequent direct automated DNA sequencing of the PCR products. Genomic DNA extracted from paraffin-embedded samples was amplified using intron-specific primers that flanked exons 1–9 as previously described (Bruni et al., 2000). Samples from the corresponding adjacent normal tissues were included as controls. Results are reported in Table 2. DNA sequencing of exons 1–9 of PTEN gene demonstrated the existence of a pathogenetic mutation in at least two samples (one embryonal carcinomas and one seminomas): a TAT right arrow TAG transversion at the codon 138 in exon 5 that caused the formation of a premature termination codon (Y138 right arrow term) in a patient affected by an embryonal carcinoma (#13); a transition CGC right arrow CAC at the codon 233 in exon 7 that caused a R233 right arrow H missense mutation in a patient affected by an embryonal carcinoma (#19). The two mutations found in GCTs likely impair PTEN function: the missense mutation (R233 right arrow H) hits a residue that has been reported to be germline mutated in a family affected by Cowden Disease (Liaw et al., 1997). On the other hand, the mutation Y138term generates a truncated protein whose function is impaired. Accordingly, a mutation that hits the residue 139 has been found in a patient affected by Cowden Disease. Thus, our analysis demonstrated the presence of somatic mutations (at a frequency of about 9%) in sporadic germ cell tumors.

Regulation of PTEN expression in embryonal carcinoma cells

The finding that post-transcriptional mechanisms are involved in the loss of PTEN expression in at least a quarter of GCTs, made us investigate whether protein degradation was involved in the loss of PTEN expression by using the embryonal carcinoma NT2/D1 cell line as a model system.

To determine the molecular mechanisms whereby PTEN expression is lost in neoplastic germ cells, we used a well-known model of human embryonal carcinoma cells: the NTERA-2 cell line (NT2/D1) (Andrews, 1984). Though this cell line derives from a late stage lesion (embryonal carcinoma cell line), and does not allow to reproduce the transition from early-lesion (ITGCN) to late lesion (full blown cancer), it still represents a good model because is amenable to manipulation in vitro, allows to study the mechanisms whereby PTEN expression is regulated in EC cells and finally allows to pinpoint the relevant pathways downstream PTEN.

Treatment of NT2/D1 cells with two highly specific proteasome inhibitors (the peptide aldheyde N-acetyl-leucyl-leucine norleucinal or LLnL and the inhibitor MG132) increased the level of PTEN. Treatment of NT2/D1 cells for 2, 8 or 12 h with 20 muM of MG132 (Figure 5a, lanes MG), or 50 muM of LLnL (not shown) resulted in two- to 4.5-fold increase in the level of PTEN expression compared to DMSO-treated cells (Figure 5a, lane C), suggesting that in embryonal carcinoma cells the proteasome-dependent pathway takes part in PTEN turnover. Conversely, treatment of embryonal carcinoma cells with 5 muM of azaC for 2–5 days did not result in a significant increase in the level of PTEN RNA and protein (Figure 5b), suggesting that PTEN promoter methylation is not implicated in the downregulation of PTEN expression in NT2/D1 cells.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Regulation of PTEN expression in embryonal carcinoma cells. (a) Treatment of NT2/D1 cells with with 20 muM DMSO (C) or MG132 (MG) for the indicated times (2, 8, 12 h). (b) Northern blot analysis of PTEN expression in NT2/D1 cells treated with solvent alone or 5 muM 5-azacytidine for 2 days

Full figure and legend (54K)

Adoptive expression of PTEN into embryonal carcinoma cells induces growth arrest

Subsequently, we used the NT2/D1 cells also as a model system to determine the effects exerted by PTEN in neoplastic germ cells. NT2/D1 cells were plated in 10-mm dishes and transfected with wild type or mutant (C124S, G129E) FLAG-tagged PTEN constructs (FLAG-PTEN, FLAG-PTEN/C124S or FLAG-PTEN/G129E) or with the control empty vector. At 48 h after transfection, cells were collected and analysed by FACS. Enforced PTEN expression in NT2/D1 cells resulted in G1 arrest but not apoptosis at 24–48 h (Figure 6a). In fact, 50.3% of NT2/D1 cells transfected with wild-type FLAG-PTEN were in G1 phase versus 29.4% of vector-transfected cells. Neither C124S nor G129E PTEN mutants suppressed growth (30 and 30.5% of cells were in the G1 compartment, respectively). As the G129E PTEN mutant has lost the lipid phosphatase activity but not the protein phosphatase activity, these findings demonstrate that the growth suppression induced by PTEN in NT2/D1 cells requires the ability to dephosphorylate lipid but not protein substrates. Statistical analysis was performed using the one-way ANOVA with post hoc multiple comparisons assessed with the two-tailed Dunnett's t-test, and the differences resulted significative (P<0.05).

Figure 6.
Figure 6 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

PTEN-induced growth suppression in NT2/D1 cells. (a) Flow cytometry of NT2/D1 cells transfected with wild type and mutant PTEN constructs. Values are meansplusminuss.d. of three experiments. Statistical analysis was performed using the one-way ANOVA with post hoc multiple comparisons assessed with the two-tailed Dunnett's t-test. *P<0.05 vs Control. (b) Flow cytometry analysis of NT2/D1 cells treated with DMSO alone or with the PI3K inhibitor LY294002. Values are meansplusminuss.d. of three experiments. Data are mean valueplusminuss.d. Statistical analysis was performed using the unpaired two-tailed Student's t-test. *P<0.05, P<0.01, P<0.005 vs Control

Full figure and legend (88K)

Treatment with pharmacological PI3K inhibitors LY294002 or wortmannin for 24 h decreased the proliferation of breast cancer cell lines (Lu et al., 1999). To determine the relevance of the PI3K/PTEN/AKT pathway in embryonal carcinoma cells, we investigated the effects exerted by PI3K inhibitors (LY294002 and wortmannin) on NT2/D1 cells. As with PTEN, treatment of NT2/D1 cells with 20 muM LY294002 (or 25 muM wortmannin, not shown) greatly reduced S phase entry as determined by flow cytometry (Figure 6b) and BrdU incorporation (Figure 9a). Therefore, inhibition of the PI3K pathway induces G1 arrest in NT2/D1 cells.

Figure 9.
Figure 9 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

PTEN exerts its growth suppression activity through p27kip1 (a). BrdU incorporation assay of NT2/D1 cells transfected with PTEN constructs in the presence or absence of p27kip1 antisense oligonucleotides. First column: green FLAG-transfected cells incorporate BrdU (yellow arrow). Second column: green FLAG-PTEN-expressing cells do not incorporate BrdU (white arrows). Third column: green FLAG-PTEN-transfected NT2/D1 cells incorporate BrdU in the presence of p27kip1 antisense oligonucleotides (yellow arrow). A times 100 Neo-Achromat Zeiss lens was used. Data are meanplusminuss.d. of three experiments; P<0.01. (b) BrdU incorporation assay of NT2/D1 cells treated with DMSO alone (column 1), LY294002 (column 2) or with LY294002 in the presence of excess of p27kip1 antisense oligonucleotides (column 3). First row: transfected cells are identified by green fluorescence of EGFP; second row: cells that incorporate BrdU are stained with Texas Red-conjugated secondary antibodies (red); third row: cell nuclei stained with Hoechst (blue). A times 100 Neo-Achromat Zeiss lens was used. Data are meanplusminuss.d. ofthree experiments; P<0.01

Full figure and legend (152K)

PTEN-dependent growth arrest in embryonal carcinoma cells requires p27kip1

Previously, we have demonstrated that p27kip1 is a key regulator of the growth and differentiation of NT2/D1 cells (Baldassarre et al., 1999, 2000). The function of p27kip1 is regulated by the activity of the PTEN/PI3K/Akt pathway through different strategies. In different cell lines the PTEN/PI3K/Akt pathway regulates both expression and localization of p27kip1 (Da-Ming and Hong, 1998; Bruni et al., 2000; Viglietto et al., 2002). Therefore, we investigated: (1) whether inhibition of PI3K signalling either by PTEN or by change of localization and (2) whether p27kip1 upregulation was required for the growth-inhibitory effects exerted by blocking the PI3K pathway.

FLAG-PTEN expression in NT2/D1 cells or treatment with the PI3K inhibitor LY294002 reduced AKT phosphorylation (Figure 7a, lanes 2 and 3, and Figure 7b, lanes 2–4, respectively), and induced a two-fold increase in the levels of p27kip1 (Figure 7a, lane 2, and Figure 7b, lanes 2 and 3, respectively).

Figure 7.
Figure 7 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Upregulation of p27kip1 in NT2/D1 cells by blockage of the PI3K pathway. (a) Immunoblot analysis of PTEN, AKT and p27kip1 expression in NT2/D1 cells transfected with PTEN constructs. Lane 1, FLAG-transfected cells; lane 2, FLAG-PTEN-transfected cells; lane 3, FLAG-PTEN-transfected in the presence of p27kip1 antisense oligonucleotides. (b) Immunoblot analysis of analysis of PTEN, AKT and p27kip1 expression in LY290042-treated NT2/D1 cells, in the presence or in absence of anti-p27kip1 antisense oligonucleotides. Lane 1, proliferating NT2/D1 cells; lane 2, LY290042-treated NT2/D1 cells; lane 3, LY290042-treated NT2/D1 cells in the presence of control oligonucleotides; lane 4, LY290042-treated NT2/D1 cells in the presence of p27kip1 antisense oligonucleotides

Full figure and legend (36K)

Furthermore, adoptive expression of PTEN and/or treatment of NT2/D1 cells with LY294002 induced cytoplamic re-localization of p27kip1 (Figure 8a). The effects exerted by PTEN or by LY294002 on p27kip1 were similar, in agreement with the concept that wild-type PTEN as well as LY294002 block PI3K-dependent activation of Akt; conversely, the mutant PTEN allele G129E has no effect on the localization and the phosphorylation of p27kip1.

Figure 8.
Figure 8 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Nuclear accumulation of p27kip1 in NT2/D1 cells by blockage of the PI3K pathway. Immunoblot analysis of p27kip1 localization on cytoplasmic and nuclear extracts of NT2/D1 cells treated with LY294002 or transfected with PTEN constructs. Lane 1, untreated mock-transfected cells; lane 2, mock-transfected cells treated with 10 muM LY294002; lane 3, wild-type FLAG-PTEN-transfected cells; lane 4, G129E FLAG-PTEN-transfected cells. beta-Tubulin and SP1 were used as controls of fractioned proteins

Full figure and legend (35K)

To determine whether p27kip1 was necessary for the growth arrest induced by wild-type PTEN and LY294002 in NT2/D1 cells, we suppressed p27kip1 expression by using antisense oligonucleotides, and measured S phase entry by determining the rate of BrdU incorporation. Antisense oligonucleotides (1 muM) spanning the ATG initiation codon of p27kip1 efficiently blocked the increase in p27kip1 expression induced by PTEN or LY294002 in NT2/D1 cells (Figure 7a, lane 3, and Figure 7b, lane 4) and almost completely rescued the growth arrest induced by PTEN or by PI3K inhibitors (Figure 9a and b, respectively).

Transfected cells were identified by cotransfection of relevant plasmids with pEGFP, a plasmid encoding the eukaryotic green autofluorescent protein (EGFP). pFLAG-transfected cells incorporated BrdU (yellow arrows); whereas FLAG-PTEN-transfected cells did not (white arrows). However, when NT2/D1 cells were transfected with FLAG-PTEN in the presence of 1 muM of p27kip1 antisense oligonucleotides, PTEN-transfected cells incorporated BrdU (yellow arrow). Analogous results were obtained when cells were treated with LY294002.

Taken together, these results indicate that the cyclin-dependent inhibitor p27kip1 is required for PTEN growth-suppressing activity in embryonal carcinoma cells and that this effect is mediated by AKT.



Inactivation of the tumor suppressor gene PTEN leads to the development of testicular germ cell cancer in heterozygous mice (Di Cristofano et al., 1998; Suzuki et al., 1998; Podsypanina et al., 1999; Kimura et al., 2003). In this study we address whether PTEN is also implicated in the development of human GCTs (seminomas, embryonal carcinomas and teratomas). Our results clearly demonstrate that the loss of PTEN expression marks the transition from noninvasive ITGCN to invasive cancer, being PTEN expression retained in ITGCN, the presumed precursor lesion of germ cell tumors, and lost in tumors. Since ITGCN frequently progresses to invasive cancer, the findings reported in this study, suggest that PTEN loss is required at later stages of cancer development to facilitate the emergence of a more aggressive phenotype.

This conclusion is in agreement with the concept that PTEN may be inactivated at different stages of tumor development (initiation and/or progression) in different tissues, and thus serves for different purposes depending on cell type (Iqbal, 2000): In endometrial cancer PTEN expression/activity is already absent in early, precancerous lesions (complex atypical hyperplasia) (Mutter et al., 2000); conversely, PTEN loss is associated with a high Gleason score in prostate cancer (McMenamin et al., 1999), with advanced pathological stage in high-grade glioblastomas (Rasheed et al., 1997; Wang et al., 1997), and late stage disease (metastatic) in melanomas (Zhou et al., 2000).

In situ hybridization analysis of GCT samples that had lost PTEN protein supported the hypothesis that loss of PTEN protein reflect the reduction of PTEN mRNA levels. Moreover, the genetic analysis of GCTs performed in this study with microsatellites spanning about 5 Mb around the PTEN locus at 10q23 (D10S551, D10S1765 and D10S541) clearly implicated PTEN loss in the development of a subset of GCTs (approximatively 35%). Consistent with the idea that PTEN is the major target of deletion at 10q23 in GCTs, LOH was most frequent for D10S1765, which is closest to PTEN (Table 2). In GCTs, loss of genetic material associated with chromosome 10q23 is observed in seminomas, embryonal carcinomas and teratomas (Skotheim and Lothe, 2003). Moreover, DNA sequence analysis of exons 1–9 of the PTEN gene uncovered the presence of mutations in the PTEN gene in two GCTs (9%); in both cases, the inactivating mutation was found in a sample that retained a certain degree of PTEN expression (#13, #19) and was not accompanied by LOH. Overall, our results demonstrate that one copy of PTEN is lost in 50% of GCTs. These results are consistent with previous studies on the cytogenetic profile of human tumors, which have shown a range of 10–15% loss of chromosome 10q in GCTs (Mertens et al., 1997), and with the report of 60% LOH and 33% mutations in cultured testicular cancer cell lines (Teng et al., 1997). The observation that about 25% of GCTs retain PTEN mRNA expression despite decreased PTEN protein levels, along with the finding that, in NT2/D1 cells, PTEN expression is upregulated by pharmacological inhibition of the proteasome, indicate that increased turnover of PTEN protein may account for the loss of PTEN expression in an additional 25% of GCTs.

These results are in agreement with the recent finding that the regulation of PTEN expression may occur through the control of the stability of the protein (Torres and Pulido, 2001; Wu et al., 2003; Okahara et al., 2004). However, we cannot rule out the existence of other mechanisms, such as promoter methylation, that contribute to inactivate PTEN gene in GCTs, especially in those cases that did not apparently show LOH, mutations or retainment of PTEN mRNA.

Previous works have failed to detect the presence of PTEN protein in the seminiferous tubule of the 17-day embryo in the human (Gimm et al., 2000) and PTEN mRNA in the mouse embryo (Luukko et al., 1999). However, PTEN mRNA is easily detected by Northern blot in the whole testis (Suzuki et al., 1998) and by immunoblot and immunostaining in maturating germ cells in adult testis (this work). Furthermore, targeted inactivation of Pten in mouse predisposes for development of teratocarcinomas and teratomas (Di Cristofano et al., 1998; Suzuki et al., 1998; Podsypanina et al., 1999; Kimura et al., 2003).

In the testis, germ cells undergo a complex program of proliferation and differentiation to form mature sperms (Chaganti and Houldsworth, 2000). Correct proliferation and apoptosis is required to regulate the size of cell lineages and the timing of differentiation (Matsui, 1998; Chaganti and Houldsworth, 2000). Therefore, the loss of PTEN expression observed in GCTs may serve multiple purposes in germ cell transformation. The window in which PTEN is expressed in mouse and human testis, as observed in this work (spermatogonia right arrow spermatocytes right arrow spermatids), overlaps the timing of massive apoptosis that occurs during maturation of germ cells in the seminiferous tubule after birth (Chaganti and Houldsworth, 2000). By preventing apoptosis-driven germ cell selection and maturation, PTEN loss would allow clones of germ cells to elude programmed cell death and undergo malignant transformation. In agreement with this thought, the targeted disruption of Akt1 in the mouse, as well as the 'knockin' mice of Stem Cell Factor/Kit receptor mutated in the docking site for the regulatory subunit of the PI3K, attenuates spermatogenesis and induces testicular atrophy, due to increased apoptosis restricted to the germ cell compartment (Chen et al., 2001).

PTEN loss may also result in unrestrained cell cycle progression and prevention of terminal differentiation. Accordingly, a recent paper has suggested that estrogen-mediated PTEN downregulation markedly increases the growth of primordial germ cells in culture and that PTEN-deficient germ cells are much more sensitive to tumorigenic transformation induced by proliferative stimuli (Moe-Behrens et al., 2003). Indeed, primordial germ cells from pten-/- mice exhibit an increased proliferative capacity (Kimura et al., 2003).

The serine/threonine kinase PKB/Akt is an important cellular target downstream PTEN that transmits proliferative and antiapoptotic signals (Datta et al., 1999). Accordingly, the loss of PTEN in GCTs inversely correlated with Akt activation. Moreover, the adoptive expression of PTEN in embryonal carcinoma NT2/D1 cells and pharmacological inhibition of the PI3K pathway induced a reduction in the level of Akt activation.

Blockage of the PI3K/PTEN/Akt pathway arrests the growth of embryonal carcinoma cells. PTEN-transfected or LY294002-treated NT2/D1 cells accumulate in G1 phase but show no sign of apoptosis, at least after 24–48 h. Defective regulation of cell cycle progression in PTEN-deficient germ cells may depend either on increased expression of cyclins or on decreased expression of CDK inhibitors. In fact, Akt increases the stability of cyclin D1 by suppressing glycogen synthase kinase-3 (GSK-3) activity, which targets cyclin D1 to phosphorylation-mediated degradation (Diehl et al., 1998). As the threonine residue phosphorylated by GSK-3 is highly conserved in all D-type cyclins, it is likely that Akt regulates also the levels of cyclin D2 and D3. Thus, the loss of PTEN function may contribute to the overexpression of cyclin D2 frequently observed in germ cell tumors (Chaganti and Houldsworth, 2000).

On the other hand, the cyclin-dependent kinase inhibitor p27kip1 is a key target downstream the PI3K/Akt signalling pathway (Bruni et al., 2000). Also, in embryonal carcinoma cells the effects exerted by inhibition of the PI3K/Akt pathway on cell cycle progression are dependent on p27kip1. In fact, our results demonstrate that the adoptive expression of PTEN and the pharmacological inhibition of PI3K activity with LY294002 moderately upregulates p27kip1 in NT2/D1 cells and that suppression of p27kip1 synthesis by antisense oligonucleotides prevents growth arrest induced either by PTEN or by LY294002. It is noteworthy that the PTEN mutant, which lacks lipid phosphatase activity but retains protein phosphatase activity (i.e. G129E), neither induces p27kip1 expression nor blocks S phase entry.

The PI3K pathway also regulates subcellular localization of p27kip1 through Akt-dependent phosphorylation of p27kip1 (Viglietto et al., 2002). Accordingly, PTEN-dependent inactivation of Akt in NT2/D1 cells results in the accumulation of p27kip1 in the nuclear compartment. This suggests that regulation of p27kip1 localization may contribute, along with regulation of p27kip1 expression, to the proliferative arrest induced either by PTEN or by LY294002 in NT2/D1 cells.

In conclusion, inactivation of PTEN is a critical step in the progression of germ cell cancer, and the cyclin-dependent kinase inhibitor p27kip1 is a key target of PTEN signalling pathway. Further studies are necessary to identify the molecular targets that act downstream PTEN in the transformation of the germ cell.


Materials and methods

Preparation of mouse testicular cells

Testicular cells were prepared from testes of adult CD1 mice (Charles River Italia). Testes were freed from the albuginea membrane, and digested for 15 min in 0.25% (w/v) collagenase (type IX, Sigma) at room temperature under constant shaking. Seminiferous tubules were cut into pieces, with a sterile blade and further digested in minimum essential medium containing 1 mg/ml trypsin for 30 min at 30°C. Digestion was stopped by adding 10% fetal calf serum; released germ cells were collected after sedimentation (10 min at room temperature) of tissue debris. Germ cells were centrifuged for 13 min at 1500 r.p.m. at 4°C and the pellet resuspended in 20 ml of elutriation medium (120.1 mM NaCl, 4.8 mM KCl, 25.2 mM NaHCO3, 1.2 mM MgS4 (7H2O), 1.3 mM CaCl2, 11 mM glucose, 1 times essential amino acid (Life Technologies, Inc.), penicillin, streptomycin, 0.5% bovine serum albumin. Pachytene spermatocyte and spermatid germ cells were obtained by elutriation of the unfractionated single-cell suspension as described elsewhere (Meistrich, 1977). Homogeneity of cell populations ranged between 80 and 85% (pachytene spermatocytes) and 95% (spermatids), was routinely monitored morphologically. Mature spermatozoa were obtained from the cauda of the epididymus of mature mice as described previously (Sette et al., 1997). Spermatogonia and Sertoli cells were obtained from prepuberal mice as previously described (Grimaldi et al., 1993; Rossi et al., 1993).

Cell lines and reagents

The embryonal carcinoma NT2/D1 and the breast MCF7 and MDA-MB-468 tumor cell lines that have been used in this study are described elsewhere (Andrews, 1984; Lu et al., 1999). Cells were grown in Dulbecco's modified Eagle's Medium (DMEM) containing 10% foetal calf serum (FCS) (Invitrogen). MG132 and 5-aza-cytidine were from Sigma-Aldrich (St Louis, MO, USA).

Tissue samples and immunohistochemistry

Paraffin-embedded specimens were obtained from the di Scienze Biomorfologiche e Funzionali, Università Federico II (Naples, Italy). For PTEN detection, sections were dewaxed and incubated with primary antibody for 1 h at room temperature. The conventional avidin-biotin complex procedure was used according to the manufacturer's protocol (LSAB Plus DAKO, Carpinteria, CA, USA). Monoclonal anti-PTEN antibodies were purchased from Santa Cruz Biotechnology Inc. (clone A2B1) and from Cell Signaling (#26H9). Positive signal was revealed by DAB chromogen, according to the supplier's conditions. Nuclei were counterstained with Mayer hematoxylin. For peptide neutralization control, the reaction with anti-PTEN antibody was preceded by overnight incubation with a 10-fold excess of the corresponding peptide antigen (Santa Cruz Biotechnology, Inc.).

In situ hybridization

In situ hybridization was performed using biotin-labelled probes at 5'OH, which were obtained tailing reaction using biotin-dUTP as marker. Hybrid detection was achieved by amplification using biotinylated tyramide (Gen Point K620 Kit, DAKO, Carpinteria, CA, USA). Sections were prepared from each sample and assayed according to the instructions of the Dako Gen Point K620 Kit. Briefly, sections were deparaffinized, re-hydrated, treated with proteinase K (6 mug/ml) in a buffer of Tris-HCl 0.05 M, pH 7.6, and then incubated in 0.3% H2O2, at RT, for 20 min, to quench endogenous peroxidase. Optimal hybridization and stringent wash temperatures were determined and slides were rinsed in the stringent solution provided with the kit. Amplified detection was performed using an antidigoxigenin antibody coupled to a peroxidase (HRP), which precipitated biotinylated tyramide. The precipitated biotin bound to streptavidin-linked HRP, which in turn precipitated the dimethylaminobenzidine (DAB) chromogen provided with the kit. Nuclei were counterstained with Mayer's hematoxylin. Negative controls were obtained using an antisense probe.

Protein extraction and immunoblotting

Total proteins were prepared as described (Baldassarre et al., 1999). Differential extraction of nuclear or cytoplasmic proteins was obtained by lysing cells in ice-cold Nonidet-P40 (NP-40) lysis buffer (0.2% NP-40, 10 mM HEPES pH 7.9, 1 mM EDTA, 60 mM KCl) supplemented with protease and phosphatase inhibitors (aprotinine, leupeptine, PMSF, and okadaic acid) and incubated on ice for 5 min. The cytosolic fraction was collected by centrifugation. Nuclei were separated through a 30% sucrose cushion and lysed by resuspension in ice-cold hypertonic buffer (250 mM Tris-HCl pH 7.8, 60 mM HCl supplemented with phosphatase and protease inhibitors) followed by repeated cycles of rapid freeze and thaw. Proteins were separated by electrophoresis in SDS-containing polyacrylamide gels, transferred onto nitrocellulose membranes (Hybond C, Amersham Pharmacia Biotech, Inc.), blocked in 5% non-fat dry milk, incubated with primary and secondary antibodies for 2 and 1 h, respectively, and revealed by enhanced chemiluminescence (ECL, Amersham Pharmacia Biotech, Inc.). Polyclonal antibodies to phospho-AKT-Ser473 and AKT were purchased from New England Biolabs (Lake Placid, NY, USA); monoclonal anti-p27kip1 and anti-beta-Tubulin were acquired, respectively, from Transduction Laboratories and NeoMarkers. Two PTEN antibodies were used in this study both for immunoblot and immunostaining: a monoclonal antibody elicited to a C-terminal peptide (clone A2B1) from Santa Cruz Inc. and a monoclonal antibody elicited to a C-terminal peptide (PTEN 26H9) from Cell Signaling. The anti phospho-Akt motif antibody was from Cell Signaling (#9611).

Vectors and transfections

The PTEN constructs are described elsewhere (Bruni et al., 2000). Transfection experiments were performed as described (Baldassarre et al., 1999). NT2/D1 cells were seeded at a density of 2 times 106 cells per 100-mm dish. The next day, cells were transiently transfected by the lipofectamine 2000 procedure (Invitrogen). At 48 h post-transfection, cells were scraped into ice-cold PBS and lysed in NP-40 lysis buffer. Where needed, the p27kip1 antisense oligonucleotides (5'-GTCTCTCGCACGTTTGACAT-3') were used at a concentration of 1 muM.

DNA preparation and mutation analysis by direct DNA sequencing

Paraffin-embedded germ cell tumors and the corresponding adjacent normal tissue samples were selected from the pathology files of Dipartimento di Anatomia Patologica, Università Federico II (Naples, Italy). Genomic DNA from 22 testes (normal or cancer tissues) was isolated with a High Pure polymerase chain reaction (PCR) Template Preparation Kit (Roche Molecular Biochemicals, Mannheim, Germany) and the PTEN mutation status was determined. Briefly, DNA from tumor samples and from the corresponding normal tissues was extracted from 3–4 8-mum-thick serial sections and subjected to PCR amplification for exons 1–9 as previously described (Scala et al., 1998; Bruni et al., 2000). PCR amplification of each single PTEN exon was performed by use of intronic primers designed at the 5' and 3' ends of each exon, followed by reamplification with nested primer pairs. Primer sequences for PCR amplification of each PTEN exon were previously reported (Steck et al., 1997). Amplified DNA was purified using Microspin S300HR Columns (Pharmacia Biotech) and sequenced using the Big Dye Terminator cycle sequencing kit (ABI PRISM, Applied Biosystems, CA, USA) and the ABI 3100 PRISM DNA sequencer (Applied Biosystems).

LOH analysis at the PTEN locus

LOH on chromosome 10 was studied by PCR-based microsatellite analysis as previously described (Mutter et al., 2000). Three polymorphic markers spanning the PTEN gene (D10S551, D10S1765, D10S541) were selected to cover deletions at the whole PTEN locus on chromosome 10q23. DNA from normal testis adjacent to tumors on histological sections from the same patient was used as reference. LOH was calculated according to the following formula: (peak height of normal allele 2)/(peak height of normal allele 1) divided by (peak height of tumor allele 2)/(peak height of tumor allele 1). LOH at a single locus was considered present when the signal corresponding to one allele showed at least a 45% reduction of intensity.

Immunofluorescence analysis

5-Bromo-2'deoxyuridine-5'-monophosphate (BrdU) incorporation assay was performed as described previously (Baldassarre et al., 1999). Briefly, 5 times 105 cells were transfected with 6 mug each of control empty vector or of wild type or mutant PTEN constructs, respectively, together with 3 mug of a vector encoding green fluorescent protein (Clontech). Labelling was carried out as recommended by the manufacturer (Roche). Fluorescence was visualized with Zeiss 140 epifluorescent microscope equipped with filters that discriminated between Texas Red and fluorescein. All assays were performed three times in duplicate.

Fluorescence-activated cell sorter (FACS) analysis

Cells were washed into ice-cold PBS and fixed by adding dropwise ice-cold 70% ethanol. Fixed cells were washed with cold PBS, labelled with 10 mug/ml propidium iodide (Sigma) and 5 mug/ml Rnase A (New England Biolabs) and analysed with a FACScan flow cytometer (Becton Dickinson, San Jose, CA, USA) interfaced with a Hewlett Packard computer (Palo Alto, CA, USA). Cell cycle analysis was performed with the CELL-FIT programme (Becton Dickinson). All FACS were performed in triplicate.

Northern blot analysis

Northern blot analysis was performed according to a standard procedure. In brief, equal amounts of total RNA (20 mug/lane) were denatured and resolved electrophoretically through formaldehyde–agarose gels. The RNA was transferred onto a nylon membrane and crosslinked by UV irradiation, Human PTEN cDNA was labeled with 32P-dCTP using a random primer labeling kit (Amersham Pharmacia Biotech), and hybridization was performed at 42°C in the presence of 50% formamide.



  1. Andrews PW. (1984) Dev. Biol. 103: 285−293. | Article | PubMed | ChemPort |
  2. Baldassarre G, Barone MV, Belletti B, Sandomenico C, Bruni P, Spiezia S, Boccia A, Vento MT, Romano A, Pepe S, Fusco A & Viglietto G. (1999) Oncogene 18: 6241−6251. | Article | PubMed | ChemPort |
  3. Baldassarre G, Boccia A, Bruni P, Sandomenico C, Barone MV, Pepe S, Angrisano T, Belletti B, Motti ML, Fusco A & Viglietto G. (2000) Cell Growth Differ 11: 517−526. | PubMed | ISI | ChemPort |
  4. Baldassarre G, Romano A, Armenante F, Rambaldi M, Paoletti I, Sandomenico C, Pepe S, Staibano S, Salvatore G, De Rosa G, Persico MG & Viglietto G. (1997) Oncogene 15: 927−936. | Article | PubMed | ChemPort |
  5. Bruni P, Boccia A, Baldassarre G, Trapasso F, Santoro M, Chiappetta G, Fusco A & Viglietto G. (2000) Oncogene 19: 3146−3155. | Article | PubMed | ChemPort |
  6. Cairns P, Okami K, Halachmi S, Halachmi N, Esteller M, Herman JG, Jen J, Isaacs WB, Bova GS & Sidransky D. (1997) Cancer Res. 57: 4997−5000. | PubMed | ISI | ChemPort |
  7. Chaganti RSK & Houldsworth J. (2000) Cancer Res. 60: 1475−1482. | PubMed | ISI | ChemPort |
  8. Chen WS, Xu PZ, Gottlob K, Chen ML, Sokol K, Shiyanova T, Roninson I, Weng W, Suzuki R, Tobe K, Kadowaki T & Hay N. (2001) Genes Dev. 15: 2203−2208. | Article | PubMed | ISI | ChemPort |
  9. Da-Ming L & Hong S. (1998) Proc. Natl. Acad. Sci. USA 95: 15406−15411. | Article | PubMed | ChemPort |
  10. Datta SR, Brunet A & Greemberg ME. (1999) Genes Dev. 13: 2905−2927. | Article | PubMed | ISI | ChemPort |
  11. Di Cristofano A, Pesce B, Cordon-Cardo C & Pandolfi PP. (1998) Nat. Genet. 19: 348−355. | Article | PubMed | ISI | ChemPort |
  12. Diehl JA, Cheng M, Roussel MF & Sherr CJ. (1998) Genes Dev. 12: 3499−3511. | PubMed | ISI | ChemPort |
  13. Furnari FB, Lin H, Huang HS & Cavenee WK. (1997) Proc. Natl. Acad. Sci. USA 94: 12479−12484. | Article | PubMed | ChemPort |
  14. Gimm O, Perren A, Weng LP, Marsh DJ, Yeh JJ, Ziebold U, Gil E, Hinze R, Delbridge L, Lees JA, Mutter GL, Robinson BG, Komminoth P, Dralle H & Eng C. (2000) Am. J. Pathol. 156: 1693−1700. | PubMed | ISI | ChemPort |
  15. Grimaldi P, Piscitelli D, Albanesi C, Blasi F, Geremia R & Rossi P. (1993) Mol. Endocrinol. 7: 1217−1225. | Article | PubMed | ChemPort |
  16. Haas-Kogan D, Shalev N, Wong M, Mills G, Yount G & Stokoe D. (1998) Curr. Biol. 8: 1195−1198. | Article | PubMed | ISI | ChemPort |
  17. Houldsworth J, Reuter V, Bosl GJ & Chaganti RS. (1997) Cell Growth Differ. 8: 293−299. | PubMed | ChemPort |
  18. Iqbal UA. (2000) J. Natl. Cancer Inst. 92: 861−862. | Article | PubMed | ChemPort |
  19. Kimura T, Suzuki A, Fujita Y, Yomogida K, Lomeli H, Asada N, Ikeuchi M, Nagy A, Mak TW & Nakano T. (2003) Development 130: 1691−1700. | Article | PubMed | ISI | ChemPort |
  20. Li DM & Sun H. (1997) Cancer Res. 57: 2124−2129. | PubMed | ISI | ChemPort |
  21. Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, Puc J, Miliaresis C, Rodgers L, McCombie R, Bigner SH, Giovanella BC, Ittmann M, Tycko B, Hibshoosh H, Wigler MH & Parsons R. (1997) Science 275: 1943−1947. | Article | PubMed | ISI | ChemPort |
  22. Liaw D, Marsh DJ, Li J, Dahia PL, Wang SI, Zheng Z, Bose S, Call KM, Tsou HC, Peacocke M, Eng C & Parsons R. (1997) Nat. Genet. 16: 64−67. | Article | PubMed | ISI | ChemPort |
  23. Lu Y, Lin YZ, LaPushin R, Cuevas B, Fang X, Yu SX, Davies MA, Khan H, Furui T, Mao M, Zinner R, Hung MC, Steck P, Siminovitch K & Mills GB. (1999) Oncogene 18: 7034−7045. | Article | PubMed | ChemPort |
  24. Luukko K, Ylikorkala A, Tiainen M & Makela TP. (1999) Mech. Dev. 83: 187−190. | Article | PubMed | ISI | ChemPort |
  25. Maehama T & Dixon JE. (1998) J. Biol. Chem. 273: 13375−13378. | Article | PubMed | ISI | ChemPort |
  26. Marsh DJ, Dahia PL, Zheng Z, Liaw D, Parsons R, Gorlin RJ & Eng C. (1997) Nat. Genet. 16: 333−334. | Article | PubMed | ISI | ChemPort |
  27. Marte BM & Downward J. (1997) Trends Biochem. Sci. 22: 355−358. | Article | PubMed | ISI | ChemPort |
  28. Matsui Y. (1998) Acta. Pathol. Microbiol. Immunol. Scand. 106: 142−148. | ChemPort |
  29. McMenamin ME, Soung P, Perera S, Kaplan I, Loda M & Sellers WR. (1999) Cancer Res. 59: 4291−4296. | PubMed | ISI | ChemPort |
  30. Meistrich ML. (1977) Methods Cell Biol. 15: 15−54. | PubMed | ChemPort |
  31. Mertens F, Johansson B, Hoglund M & Mitelman F. (1997) Cancer Res. 57: 2765−2780. | PubMed | ISI | ChemPort |
  32. Moe-Behrens GH, Klinger FG, Eskild W, Grotmol T, Haugen TB & De Felici M. (2003) Mol. Endocrinol. Dec. 17: 2630−2638. | Article | ChemPort |
  33. Mutter GL, Lin MC, Fitzgerald JT, Kum JB, Baak JP, Lees JA, Weng LP & Eng CJ. (2000) Natl. Cancer Inst. 92: 924−930. | Article | ChemPort |
  34. Myers MP, Pass I, Batty IH, Van der Kaay J, Stolarov JP, Hemmings BA, Wigler MH, Downes CP & Tonks NK. (1997) Proc. Natl. Acad. Sci. USA 94: 9052−9057. | Article | PubMed | ChemPort |
  35. Nelen MR, van Staveren WC, Peeters EA, Hassel MB, Gorlin RJ, Hamm H, Lindboe CF, Fryns JP, Sijmons RH, Woods DG, Mariman EC, Padberg GW & Kremer H. (1997) Mol. Genet. 6: 1383−1387. | Article | ChemPort |
  36. Okahara F, Ikawa H, Kanaho Y & Maehama T. (2004) J. Biol. Chem. 279: 45300−45303. | Article | PubMed | ChemPort |
  37. Perren A, Komminoth P, Saremasiani P, Matter C, Feurer S, Lees JA, Heitz PU & Eng C. (2000) Am. J. Pathol. 157: 1097−1103. | PubMed | ChemPort |
  38. Podsypanina K, Ellenson LH, Nemes A, Gu J, Tamura M, Yamada KM, Cordon-Cardo C, Catoretti G, Fisher PE & Parsons R. (1999) Proc. Natl. Acad. Sci. USA 96: 1563−1568. | Article | PubMed | ChemPort |
  39. Rasheed BK, Stenzel TT, McLendon RE, Parsons R, Friedman AH, Friedman HS, Bigner DD & Bigner SH. (1997) Cancer Res. 57: 4187−4190. | PubMed | ISI | ChemPort |
  40. Risinger JI, Hayes AK, Berchuck A & Barrett JC. (1997) Cancer Res. 57: 4736−4738. | PubMed | ISI | ChemPort |
  41. Rossi P, Dolci S, Albanesi C, Grimaldi P, Ricca R & Geremia R. (1993) Dev. Biol. 155: 68−74. | Article | PubMed | ChemPort |
  42. Samuels Y, Wang Z, Bardelli A, Silliman N, Ptak J, Szabo S, Yan H, Gazdar A, Powell SM, Riggins GJ, Willson JK, Markowitz S, Kinzler KW, Vogelstein B & Velculescu VE. (2004) Science 304: 554−558. | Article | PubMed | ISI | ChemPort |
  43. Scala S, Bruni P, Lo Muzio L, Mignogna M, Viglietto G & Fusco A. (1998) Int. J. Oncol. 13: 665−668. | PubMed | ChemPort |
  44. Sette C, Bevilacqua A, Bianchini A, Mangia F, Geremia R & Rossi P. (1997) Development 124: 2267−2274. | PubMed | ISI | ChemPort |
  45. Skotheim RI & Lothe RA. (2003) APMIS 111: 136−150. | Article | PubMed | ChemPort |
  46. Stambolic V, Suzuki A, de la Pompa JL, Brothers GM, Mirtsos C, Sasaki T, Ruland J, Penninger JM, Siderovski DP & Mak TW. (1998) Cell 95: 29−39. | Article | PubMed | ChemPort |
  47. Steck PA, Pershouse MA, Jasser SA, Yung WK, Lin H, Ligon AH, Langford LA, Baumgard ML, Hattier T, Davis T, Frye C, Hu R, Swedlund B, Teng DH & Tavtigian SV. (1997) Nat. Genet. 15: 356−362. | Article | PubMed | ISI | ChemPort |
  48. Suzuki A, de la Pompa JL, Stambolic V, Elia AJ, Sasaki T, del Barco Barrantes I, Ho A, Wakeham A, Itie A, Khoo W, Fukumoto M & Mak TW. (1998) Curr. Biol. 8: 1169−1178. | Article | PubMed | ISI | ChemPort |
  49. Tashiro H, Blazes MS, Wu R, Cho KR, Bose S, Wang SI, Li J, Parsons R & Ellenson LH. (1997) Cancer Res. 57: 3935−3940. | PubMed | ISI | ChemPort |
  50. Teng DH, Hu R, Lin H, Davis T, Iliev D, Frye C, Swedlund B, Hansen KL, Vinson VL, Gumpper KL, Ellis L, El-Naggar A, Frazier M, Jasser S, Langford LA, Lee J, Mills GB, Pershouse MA, Pollack RE, Tornos C, Troncoso P, Yung WK, Fujii G, Berson A & Steck PA. (1997) Cancer Res. 57: 5221−5225. | PubMed | ISI | ChemPort |
  51. Torres J & Pulido R. (2001) J. Biol. Chem. 276: 993−998. | Article | PubMed | ISI | ChemPort |
  52. Ulbright TM. (1993) Am. J. Surg. Pathol. 17: 1075−1091. | PubMed | ISI | ChemPort |
  53. Viglietto G, Motti ML, Bruni P, Melillo RM, D'Alessio A, Califano D, Vinci F, Chiappetta G, Tsichlis P, Bellacosa A, Fusco A & Santoro M. (2002) Nat. Med. 8: 1136−1144. | Article | PubMed | ISI | ChemPort |
  54. Viglietto G, Romano A, Maglione D, Rambaldi M, Paoletti I, Lago C, Califano D, Monaco C, Mineo A, Santelli G, Manzo G, Botti G, Chiappetta G & Persico MG. (1996) Oncogene 13: 577−587. | PubMed | ISI | ChemPort |
  55. Vos A, Oosterhuis JW, de Jong B, Buist J & Schraffordt Koops H. (1990) Cancer Genet. Cytogenet. 46: 75−81. | Article | PubMed | ChemPort |
  56. Wang SI, Puc J, Li J, Bruce JN, Cairns P, Sidransky D & Parsons R. (1997) Cancer Res. 57: 4183−4186. | PubMed | ISI | ChemPort |
  57. Wu W, Wang X, Zhang W, Reed W, Samet JM, Whang YE & Ghio AJ. (2003) J. Biol. Chem. 278: 28258−28263. | Article | PubMed | ChemPort |
  58. Wu Y, Dowbenko D, Pisabarro MT, Dillard-Telm L, Koeppen H & Lasky LA. (2001) J. Biol. Chem. 276: 21745−21753. | Article | PubMed | ISI | ChemPort |
  59. Zhou XP, Gimm O, Hampel H, Niemann T, Walker MJ & Eng V. (2000) Am. J. Pathol. 157: 1123−1128. | PubMed | ISI | ChemPort |


This work was supported by grants from the Associazione Italiana Ricerca sul Cancro (AIRC) to GV. We are indebted to Jean Ann Gilder for revising and editing the text.