PTEN Signaling Mediates Estrogen-Dependent Proliferation of

0888-8809/03/$15.00/0 Printed in U.S.A.

Molecular Endocrinology 17(12):2630–2638 Copyright © 2003 by The Endocrine Society doi: 10.1210/me.2003-0006

Akt/PTEN Signaling Mediates Estrogen-Dependent Proliferation of Primordial Germ Cells in Vitro GERD H. G. MOE-BEHRENS, FRANCESCA GIOIA KLINGER, WINNIE ESKILD, TOM GROTMOL, TRINE B. HAUGEN, AND MASSIMO DE FELICI Department of Gynecology and Obstetrics (G.H.G.M.-B., T.B.H.), Andrology Laboratory, Rikshospitalet, University Hospital, 0027 Oslo, Norway; Department of Public Health and Cell Biology (G.H.G.M.-B., F.G.K., M.D.F.), Section of Histology and Embryology, University of Rome “Tor Vergata,” 00133, Rome, Italy; Department of Biochemistry (W.E.), University of Oslo, 0316 Oslo, Norway; and Institute of Population-Based Cancer Research (T.G.), Cancer Registry of Norway, 0310 Oslo, Norway Testicular tumors in humans are reported to be significantly increasing in incidence. Embryo exposure to environmental estrogens has been proposed as one of the possible underlying causes. In mice, genetic, immunological, and experimental evidence suggest that germ cell testicular tumors may derive from primordial germ cells (PGCs), the embryonic precursors of gametes. Here we show that relatively high concentrations of estrogens stimulate mouse PGC growth in vitro through the somatic cells of the gonadal ridges. Moreover, we found that estrogens stimulate the transcription of the Steel gene and the production of c-Kit ligand in gonadal somatic cells, and that this growth factor

is likely to be responsible for the observed stimulation of PGC growth via an Akt/PTEN pathway. Finally, we show that estrogen stimulation of gonadal somatic cells in culture, in combination with PTEN down-regulation in PGCs and the presence of leukemia inhibitory factor in the culture medium, result in high frequency of PGC transformation in tumorigenic cells. Based on these results, we present a novel experimental in vitro model for tumorigenic germ cell transformation and identify molecular pathways likely involved in development of germ cell tumors after estrogen exposure. (Molecular Endocrinology 17: 2630–2638, 2003)

I

N MAMMALS, THE DEVELOPMENT of testicular germ cell tumors (teratomas and teratocarcinomas) is thought to be initiated from proliferating primordial germ cells (PGCs), the embryonic precursors of the gametes, transforming in pluripotent cells during embryonic life (for a review, see Ref. 1). Moreover, PGCs that are topographically misplaced during migration toward gonadal ridges are considered as a source of germinomas in extragonadal sites (2). In humans, germ cell tumors in all locations account for approximately 3% of malignancies in children and adolescent. Around two thirds are at extragonadal sites. The causes and molecular mechanisms of such PGC transformation are unknown. Embryo exposure to environmental estrogens during early stages of PGC development has been proposed as one of possible causes of the recent rising incidence of germ cell testicular cancer (3, 4). In line with this hypothesis, Weir et al. (5) reported a case-control study in which they found evidence that exposure to maternal hormones, particularly estrogens, is associated with testicular germ-cell cancer risk. To date, however, the lack of suitable experimental models has not allowed for verification of such a hypothesis.

In the present work, we aimed to directly verify whether estrogens affect the development of mouse PGCs and eventually cause their transformation into tumorigenic cells. This paper shows that estrogens stimulate Steel gene transcription in gonadal somatic cells increasing the production of the c-Kit ligand (KL) and consequently inducing PGC growth most likely via an Akt/PTEN-dependent pathway. In addition, we found that prolonged exposure of PGCs to estrogens in culture associated with down-regulation of the lipid phosphatase PTEN, a major inhibitor of phosphoinositol 3-OH kinase (PI3-K) activity (6) and with stimulation by leukemia inhibitor factor (LIF), resulted in a high frequency of PGC transformation into tumorigenic cells. These results highlight the risk that in certain genetic background, exposure of the embryo to estrogens may favor tumors formation from PGCs in the testis and perhaps in other extragonadal sites where PGCs can mislocalize during migration toward the gonadal ridges.

Abbreviations: AP, Activator protein; APase, alkaline phosphatase; AS, antisense PTEN oligo nucleotides; bFGF, basic fibroblast growth factor; BrdU, 5-bromo-2⬘-deoxyuridine; dpc, days post coitum; E2, 17␤-estradiol; EG, embryonal germ; ER, estrogen receptor; ERR, estrogen-related receptor; FRSK, forskolin; KL, c-Kit ligand; LIF, leukemia inhibitor factor; PGC, primordial germ cells; PI3-K, phosphoinositol 3-OH kinase; ZEA, zearalenone.

Estrogens Increase PGC Proliferation in Vitro through Stimulation of the Somatic Cells of the Gonadal Ridges

RESULTS

In initial experiments, we found that either 17␤-estradiol (E2) or the environmentally widespread mycoestrogen zearalenone (ZEA), were able to signifi2630

Moe-Behrens et al. • Akt/PTEN Signaling in PGC Proliferation

Mol Endocrinol, December 2003, 17(12):2630–2638

2631

cantly increase, in a dose-dependent manner, the number of 11.5 d post coitum (dpc) PGCs cultured for 1 and 3 d together with their own gonadal somatic cells onto STO fibroblast monolayers (Fig. 1, A and B). However, under the same culture conditions, neither compounds significantly affected the number of purified PGCs cultured onto STO cell monolayers without gonadal somatic cells (1 d of culture, control ⫽ 55% ⫾ 7%, 100 nM E2 ⫽ 45% ⫾ 5%, 10 ␮M ZEA ⫽ 57% ⫾ 11%), thus indicating that estrogens stimulate PGC growth through gonadal somatic cells. To test whether estrogens act as mitogens or cell survival factors, after 1 d of culture cells were incubated for 1 h with 5-bromo-2⬘-deoxyuridin (BrdU), and the number of alkaline phosphatase-positive cells (PGCs) that had incorporated BrdU were scored. The results showed that the number of PGCs in S phase in cultures treated with 10 ␮M ZEA was significantly higher (36.6 ⫾ 1.5%) than in control (24.1 ⫾ 1.0%). Using RT-PCR and immunoblotting, we found that 11.5 dpc gonadal somatic cells, but not PGCs, express estrogen receptor (ER) ␣ and estrogen-related receptors (ERR) ␣ (Fig. 2, A and B), whereas ER␤ and ERR␤ were not expressed by either cell types (not shown), thus confirming that PGCs are not the direct target of estrogens.

Fig. 2. The Expression of ER␣ and ERR␣ by 11.5 dpc Whole Gonadal Ridges and Purified Gonadal Somatic Cells, But Not by PGCs A, RT-PCR analysis; adult mouse uterus and 16.5 dpc fetal heart were used as positive controls for ER␣ and ERR␣, respectively. B, Western blotting showing the same ER␣ expression patterns.

The immunohistochemistry for ER␣ on tissue sections of 11.5 dpc gonads gave no conclusive results (data not shown). Estrogens Stimulate KL Production in Gonadal Somatic Cells

Fig. 1. E2 and ZEA Stimulate Gonadal Somatic Cell-Dependent PGC Growth A and B, 11.5 dpc PGCs plus gonadal somatic cells were seeded onto STO fibroblast feeder layers and cultured for 3 d in the continuous presence of indicated concentrations of E2 (A) and ZEA (B). Changes in PGC number (%) were calculated by dividing the number of PGCs at d 1 and 3 by the number of seeded PGCs. Bars, Mean ⫾ SD of three independent experiments with three replicates each (*, P ⬍ 0.05).

In vitro studies have identified several factors that can regulate PGC growth (for a review, see Ref. 7). The product of the Steel gene, c-Kit ligand (KL), also known as stem cell factor, is a major positive regulator of PGC growth. PGCs express high levels of the KL receptor c-Kit and gonadal somatic cells synthesize both soluble and membrane-bound forms of KL (8– 10). We found that the addition of E2 and ZEA to the culture medium caused, on average, a significant 4-fold (E2) and 2-fold (ZEA) increase of soluble KL production in gonadal somatic cells in culture, whereas no consistent increase of the membrane bound KL was found (Fig. 3, A and B). Further evidence that the growth promoting action of E2 and ZEA

2632 Mol Endocrinol, December 2003, 17(12):2630–2638


and increased intracellular levels of cAMP in PGCs, were not involved in the estrogen-dependent PGC growth (Fig. 3C, and data not shown). Estrogens Increase the Activity of the Steel Promoter in Gonadal Somatic Cells It is known that after estrogen binding, ER␣ binds to DNA and activates gene transcription through binding sites such as the estrogen response element, the steroidogenic factor-1 response element, or an activator protein (AP)-1 response element (11, 12). Interestingly, an AP-1 response element is located in the promoter region of the mouse Steel gene at position ⫺603 upstream of the transcriptional initiation site (13). This suggests that estrogens may stimulate Steel transcription in gonadal somatic cells by binding to an AP-1 response element of the Steel gene. To verify such a possibility, we transfected a gonadal cell line (TM4 cells) with a reporter plasmid containing the luciferase gene under transcriptional control of the Steel gene promoter AP-1 response element (see Materials and Methods). The results showed a significant, almost 3-fold increase of the reporter activity in transfected TM4 cells cultured for 48 h in the presence of 100 nM E2 or 10 ␮M ZEA (Fig. 3D). KL Increases Akt and PTEN Phosphorylation in PGCs

Fig. 3. Western Blot Analyses Showing the Increase of Soluble KL (sKL) and at Minor Not Significant Extent of Membrane-Bound KL (mKL) But Not of LIF in 11.5 dpc Somatic Cells of the Gonadal Ridges Induced by E2 (A and C) or ZEA (B) Sl4 m220 and DIA-M cells (a kind gift from Mia Buher, Edinburgh University, Edinburgh, Scotland, UK) were used as positive controls for KL and LIF production, respectively. D, Activation of the Steel gene promoter by E2 and ZEA in gonadal cells determined by the analysis of the activity of a reporter plasmid containing a luciferase gene under the transcriptional control of Steel promoter AP-1 response element (*, P ⬍ 0.05). The results reported were obtained using the mouse Sertoli cell line TM4 because sufficient numbers of 11.5 dpc gonadal somatic cells for these assays could not be readily obtained.

on PGCs is mostly due to the stimulation of the KL/cKit system activity, was from the finding that 10 ␮g/ml anti-KL antibody abolished such action (1 d of culture, control ⫽ 62 ⫾ 7%, 10 ␮M ZEA ⫽ 120 ⫾ 9%, ZEA ⫹ anti-KL antibody ⫽ 55 ⫾ 5%). In addition, we found that other factors known to favor PGC growth (for a review, see Ref. 7), such as the leukemia inhibitory factor (LIF), the basic fibroblast growth factor (bFGF)

To identify molecular pathways downstream of c-Kit activation in PGCs and to verify their involvement in the increased proliferation of PGCs caused by estrogens, we first studied whether KL stimulation of PGCs resulted in phosphorylation of the serine/threonine kinase Akt, a well-known effector of c-Kit in other cell type (for a review, see Ref. 14). Western blot analysis demonstrated that stimulation of purified PGCs, cultured in suspension in the absence of cell monolayers, for 5 min with 100 ng/ml KL, increased Akt phoshorylation (Fig. 4A). The presence in the assay of 20 ␮M SU5416, a potent inhibitor of c-Kit-dependent Akt activation (15), or of 10 ␮M LY294002, a highly specific inhibitor of the lipid kinase PI3-K, resulted in complete inhibition of Akt phosphorylation (Fig. 4A). This indicated that c-Kit and PI3-K activation in 11.5 dpc PGCs, as in other cell type, are upstream events of Akt phosphorylation. Interestingly, we also found that the lipid phosphatase PTEN, known as a major inhibitor of PI3-K activity (6), was also phosphorylated in PGCs subjected to KL stimulation (Fig. 4B). Because PTEN phosphorylation is believed to restrict PTEN activity (16), a parallel activation of Akt and inactivation of PTEN seem to occur in PGCs after KL stimulation that should contribute to sustain their proliferation. Moreover, we found that SU5416 also inhibited PTEN phosphorylation induced in PGC by KL stimulation (Fig. 4B), confirming a novel upstream c-Kit-dependent regulation of PTEN. We did not further investigate the molecular pathway(s) of the c-Kit-dependent PTEN



2633

Fig. 4. Akt and PTEN Phosphorylation in 11.5 PGC A, Western blot analysis showing increased Akt phosphorylation (Ser 473) after 5⬘ stimulation of 11.5 dpc PGCs with 100 ng/ml KL; the presence of 20 ␮M SU5416 or 10 ␮M LY294002 in the assay abolished such effect. B, Similar analysis showing increased PTEN phosphorylation (Ser 380) and the inhibitory effect of SU5416.

phosphorylation. It is unlikely, however, that they involve PI3-K activation because it is known that the PTEN tail contains phosphorylation sites for GSK3, PKA, CK1, and CK2 kinases but not for PI3-K (16). No evidence of Akt or PTEN phosphorylation in PGCs treated directly with estrogens in the absence of cell monolayers was observed (data not shown). On the other hand, it was not possible to verify whether estrogens cause Akt and PTEN phosphorylation in PGCs seeded onto cell feeder layers because PCCs could not be recovered from the culture. It was also not possible to use immunocytochemistry for such evaluation due to the high fluorescence background for the antibodies against phosphorylated Akt and PTEN in the cell monolayers (data not shown). However, we found that 20 ␮M SU5416 (Fig. 5A) or 10 ␮M LY294002 (not shown) abolished or significantly reduced, respectively, the estrogen-dependent increase of PGC growth, suggesting that estrogen action on PGCs requires c-Kit-dependent Akt and PI3-K activation. Stimulation of Cultures with Estrogens Associated with PTEN Inactivation in PGCs Result in Increased Transformation of PGCs into Tumorigenic Cells In the final series of experiments, we investigated whether inhibition of PTEN expression in PGCs affects

Fig. 5. Reduced PTEN Expression Causes Increased PGC Proliferation A, Transfection of 11.5 dpc PGCs with PTEN AS (either AS1 or AS2) (1 ␮M) results in a significant increase of their KL (100 ng/ml) and E2 (100 nM) or ZEA (10 ␮M)-dependent growth (*, P ⬍ 0.05 treatment vs. control, ⫹ P ⬍ 0.05 growth factor ⫹ AS vs. growth factor without antisense). Note that the presence of 20 ␮M SU5416 (SU) inhibits the increase of PGC growth in any group. S, Sense oligo nucleotides. B, Double staining for TG-1 and PTEN in PGCs cultured for 24 h onto STO cell monolayers after transfection; control PTEN sense oligo nucleotide-treated PGCs labeled with TG-1 (a) and PTEN antibody (b) and PTEN AS-treated PGCs labeled with SSEA-1-like TG-1 antibody (c) and PTEN antibody (d); note a marked reduction of PTEN staining in d.

their proliferation and in particular their response to estrogens in culture. Figure 5A shows that treatment of cultures with antisense PTEN oligo nucleotides (AS) results in a marked KL and estrogen-dependent increase of PGC growth after 1 d of culture. Sense oligo nucleotides did not affect PGC response to KL and estrogens. Inactivation of PTEN by oligo antisense in PGCs was confirmed by a significant decrease of PTEN immunopositivity in treated PGCs after 1 d of culture in comparison to control (Fig. 5B). We also tested if estrogen stimulation and PTEN inactivation affected the embryonal germ (EG)-forming


efficiency of PGCs in vitro. It is known that PGCs cultured in continuous presence of a combination of growth factors or compounds [for example KL, LIF, and bFGF or forskolin (FRSK)] for 7–10 d give rise to pluripotent tumorigenic cells called EG cells (17–19). We found that in cultures of no-AS-treated PGCs, estrogens can substitute for exogenous KL and partly FRSK in the cocktail of compounds necessary for EG cell formation and that the efficiency of EG cell colony formation was markedly increased in PTEN AS-treated PGC cultures. Moreover, in this latter condition ZEA or E2 plus LIF produced the most favorable compound combination for EG formation (Fig. 6). Control replicates indicated that antisense oligonucleotide treatment in the absence of growth factors and estrogens was not sufficient to induce PGC transformation in EG cells and that cell treatment with sense oligonucleotide did not produce any significant increase of EG cell formation with any compound combination in comparison to untreated cultures (data not shown). Moreover,


PGC transformation into EG cells with or without estrogen stimulation required c-Kit-dependent Akt activation because no EG colonies were formed when PGC culture was carried out in the presence of 20 ␮M SU5416 or 10 ␮g/ml anti-Kl antibody (Fig. 6). To verify the tumorigenicity of the EG cells obtained from PTEN-AS-treated PGCs cultured in the presence of E2 or ZEA plus LIF, we injected monodispersed EG cells (at passage number three on STO cells ⫹ 1000 UI LIF) into nude mice sc or under the testis capsule. Multiple teratocarcinomas developed in the injected mice with high frequency (two of three in sc-injected mice; three of three in testis-injected mice), similar to that reported for other EG cell lines (18). DISCUSSION In humans the etiology of testicular cancer is unknown, although several risk factors have been deter-

Fig. 6. The Frequency of Formation of EG Cell Colonies in Control PGCs and in PTEN AS-Treated PGCs A, PGCs obtained from 11.5 dpc embryos (129 SvSl/⫹/CD1) were seeded onto Sl4 m220 feeder cells and cultured for 7 d in the presence of a combination of factors. At the end of culture, APase-positive colonies composed of more than 20 cells were counted. The EG colony-forming efficiency (%) in untreated (white bars) and PTEN AS (1 ␮M)-treated PGCs (gray bars) was calculated by dividing the number of EG cell colonies at d 7 by the number of seeded PGCs; KL (50 ng/ml), LIF (1000 U), bFGF (10 ng/ml), FRSK (10 ␮M), ZEA (10 ␮M), SU5416 (SU) (20 ␮M), anti-Kl antibody (10 ␮g/ml). Values are mean ⫾ SEM. *, P ⬍ 0.05 vs. the corresponding not treated PGCs; ⫹, P ⬍ 0.05 vs. KL, LIF, FRSK. Similar results were obtained when ZEA was replaced by 100 nM E2 (not shown). B, EG colonies formed from AS-treated PGCs in the presence of LIF (1000 U) ⫹ ZEA (10 ␮M). C, Magnification of an APase-positive colony. D, Staining of a colony with TG-1 antibody was used as further marker of EG colonies. Bars indicate 103 ␮m (B) and 60 ␮m (C and D).


mined. These include cryptorchidism, testicular dysgenesis, carcinoma in situ (CIS), Klinefelter’s syndrome, and a familial history of testicular cancers. Embryo exposure to environmental estrogens during early stages of germ cell development has recently been proposed as one of the possible causes of the increasing incidence of germ cell testicular cancer (3–5). That proliferating PGCs, under certain genetic backgrounds, are the cells of origin for testicular teratocarcinomas in mice has been experimentally established (20, 21). Compelling evidence suggests that a variety of polypeptide growth factors and compounds control the survival and proliferation of mouse PGCs and can be involved in deregulation of PGC development (7). Once in the gonad, PGCs proliferate for a short period of time before they differentiate to form gonia. It can be hypothesized that tumors arise from PGCs failing to end proliferation and undergo proper differentiation. In support of this possibility, it has been demonstrated that PGCs incubated in vitro, in the continuous presence of certain growth factors and compounds proliferate beyond the time when they would normally stop dividing in vivo and give rise to pluripotent tumorigenic cells called embryonal germ (EG) cells (17, 18). So far, no evidence that estrogens may influence PGC proliferation and differentiation in mammals is available. Studies in mice lacking ER␣ and ER␤ did not report obvious variations in germ cell or PGC numbers (22, 23), suggesting that estrogens are dispensable for PGC proliferation. In the present paper, however, we report that relatively high concentrations of estrogens increase PGC proliferation in culture in a dose-dependent manner. In addition, we demonstrated that estrogens do not act directly on PGCs but stimulate gonadal somatic cells to synthesize PGC proliferation factor(s). In support of these observations, we found that these early stage PGCs do not express estrogen receptors, whereas gonadal somatic cells express mRNA for ER␣. The immunohistochemistry for the ER␣ on tissue sections of 11.5 dpc gonads gave no conclusive results, which might indicate a low expression of this receptor at protein level. It is known, however, that certain tissues (e.g. bone), which are highly responsive to estrogens, express a low level of ERs. In such tissues, the effect of estrogens is explained by the simultaneous expression of ERR␣ or ERR␤, which, acting via a cross talk to the ER␣ or ER␤, amplify estrogen signaling (11). This might also occur in 11.5 dpc gonadal ridges because we found that somatic cells express, in addition to ER␣, ERR␣ mRNA. At later fetal stages, both male and female mouse gonadal somatic cells and germ cells can express ERs, but their functions at these stages are unknown (24, 25). Among several compounds that can stimulate PGC proliferation (7), we obtained evidence that estrogens increase the synthesis of KL, mainly the soluble form, in gonadal somatic cells, most likely by directly increasing Steel promoter activity. In support of the possibility that increased levels of KL produced by


2635

gonadal somatic cells is responsible for the estrogendependent growth of PGC in culture, we found that this growth induction was abolished or significantly reduced when anti-KL antibodies were added to the culture medium. We attempted to identify downstream effectors of KL in PGCs, with the aim of studying their involvement in the stimulation of PGC proliferation by estrogens and the consequence of an exposure of PGCs to estrogens in culture conditions in which the action of these effectors was experimentally altered. Several gain-of-function mutations have been recently identified in the c-Kit receptor or its dependent signaling to be associated with highly malignant tumors including germ cell testicular tumors in humans (14). It is noteworthy that the c-Kit receptor is expressed at high levels in carcinoma in situ and seminoma testicular tumors (26). The lipid kinase phosphoinositol 3-OH kinase (PI3-K) and its downstream target, the proteinserine/threonine kinase Akt, are crucial effectors of tumorigenic protein-tyrosine kinases including c-Kit. Interestingly, the tumor suppressor gene PTEN encodes a phosphatase that inhibits PI3-K activity. Deletion or inactivation of PTEN results in constitutive Akt activation and is implicated in the pathogenesis of tumors of various histological origin (6). It has been suggested that PTEN may be inactivated upon growth factor stimulation and estrogens have been reported to inhibit PTEN-mediated growth suppression in MCF-7 breast cancer cells (6, 27). The stability and activity of PTEN depend on the phosphorylation of three residues (S380, T382, and T383) within its tail. In particular, PTEN tail phosphorylation is believed to restrict PTEN activity (16). Recently, Akt activation by estrogen in estrogen receptor-negative breast cancer cells has also been reported (28). Based on these reports, we decided to verify whether stimulation of PGCs by KL or estrogens resulted in Aktphosphorylation (activation) and PTEN-phosphorylation (inactivation). Western blot analysis demonstrated that stimulation of purified PGCs in suspension with 100 ng/ml KL increased both Akt and PTEN phoshorylation, thus suggesting a parallel KL-dependent activation of Akt and inactivation of PTEN. The presence in the assay of SU5416, an inhibitor of c-Kit-dependent Akt activation (15) or of LY294002, an inhibitor of PI3-K, one of the major effectors of Akt, resulted in complete inhibition of Akt phosphorylation. This indicates that c-Kit and PI3-K activation in 11.5 dpc PGCs, as in other cell types, are upstream events of Akt phosphorylation. Two studies using transgenic and knockout mice (29, 30) and one by De Miguel et al. (31), have recently reported that Akt activation in PGCs may occur independently of PI3-K action suggesting alternative downstream effectors of c-Kit in PGCs. Although we were unable to demonstrate that in our culture system estrogen action resulted in increased Akt and PTEN phosphorylation in PGCs, the findings that the estrogen-dependent increase of PGC growth


in culture was abolished or significantly decreased by anti-KL antibodies and by SU5416 or LY294002 presence in the culture medium, suggest that such effect depends on c-Kit activation and Akt/PI3-K downstream pathways. Most importantly, we show that the tumor suppressor phosphatase PTEN, a major regulator of PI3-K activity (6), plays a crucial role in regulating PGC growth and their response to prolonged estrogen stimulation. We demonstrated that the reduction of PTEN activity in PGCs markedly increased their growth in culture and that PGCs with low PTEN activity were much more sensitive to tumorigenic transformation induced by the addition to the culture medium of KL, LIF, and FRSK or of estrogens and LIF, thus suggesting that PGCs in which PTEN is inactive are more prone to tumorigenic transformation by any proliferative stimulus. Estrogen might be one of many factors that can act as such a proliferative stimulus. Finally, we found that under the culture conditions used in the present paper, PGC transformation into EG cells required a functional KL/c-Kit system. In support, our results regarding the crucial role of PTEN in germ cell tumor formation, while we were preparing the present paper, Kimura et al. (32) have reported that PGCs from PTEN-null mice exhibited an increased proliferation capacity, Akt hyper-phosphorylation and enhanced EG cell colony-formation. They have also demonstrated that the PTEN-null mice developed bilateral testicular teratomas. In conclusion, whereas studies in mice lacking ER␣ and ER␤ suggest that estrogens are not required for normal PGC development, our results indicate that exposure to estrogens during embryonic life may have a profound effect on PGC growth and differentiation. Moreover, although it remains controversial if prenatal estrogen excess exposure in itself increases the risk of testicular germ cell cancer (5, 33), in the present paper, using an in vitro culture system we show that estrogens stimulate PGC proliferation and that this can result in their tumorigenic transformation when associated with other conditions affecting the control of their proliferation and differentiation. We provide evidence that together with genetic background and conditions (i.e. KL/c-Kit mutations, PTEN inactivation) favoring increased and/or prolonged PGC proliferation and the action of factors that inhibit PGC differentiation (most importantly LIF, or members of the LIF family), exposure to high levels of estrogens, altering the normal supply of growth factors provided by estrogen-responsive somatic cells surrounding PGCs, might constitute a high risk for PGC transformation into pluripotent tumorigenic cells. These conditions may give rise to germ cell tumor formation in the fetal testis and perhaps in other extragonadal sites containing estrogen-responsive cells where PGCs can mislocalize during migration toward the gonadal ridges and find conditions favorable for their tumorigenic transformation.


MATERIALS AND METHODS Experimental Animals All animal experimentation described in the present study was conducted in accord with accepted standards of humane animal care. Mice and Cell Culture 129 SvSl/⫹ random-bred with 129 Sv⫹/⫹ mice (Jackson Laboratories, Bar Harbor, ME), and random-bred CD1 mice (Charles River, Como, MI) were used. The morning of vaginal plug was defined as 0.5 dpc. Gonadal ridges were obtained from 11.5 dpc embryos. PGC plus gonadal somatic cells and MiniMACS (Miltenyi Biotech GmbH, Bergisch Gladbach, Germany) purified PGCs and gonadal somatic cells were prepared using MiniMACS magnetic separation system as previously described (34). Culture of PGCs on mitomycin C-treated STO (ATCC, Manassas, VA) or Sl4m220 (a kind gift from Dr. Peter Donovan, T. Jefferson University, Philadelphia, PA) cell feeder layers was carried out as previously described (34) using phenol red-free DMEM with high glucose, supplemented with 15% FCS, 1 mM nonessential amino acids, 1 mM sodium pyruvate, 2 mM L-glutamine and penicillin-streptomycin. PGCs were identified by alkaline phosphatase (APase) staining (34). In some experiments, PGC identity was confirmed by TG-1 (SSEA-1-like) antibody as described (34). The number of adherent PGCs 4 h after seeding was defined as the number of seeded PGCs. Mouse recombinant KL and anti-KL antibody (MABG56) were purchased from R & D Systems (Minneapolis, MN) and mouse LIF ESGRO and human bFGF were from Life Technologies (Tåstrup, Denmark). E2, ZEA, and FRSK were from Sigma (Oslo, Norway). SU5416 was a kind gift from SUGEN, Inc. (South San Francisco, CA) and LY294002 was from ALEXIS Biochemicals (Vinci, Italy). BrdU labeling was carried out using the Amersham Pharmacia Biotech (Oslo, Norway) cell proliferation kit as reported in (35). Tumors in Nude Mice EG cells were subcultured onto STO cell feeder layers in the presence of 1000 UI LIF as described (18). Approximately 2 ⫻ 104 cells were injected sc or under the testis capsule of CD-1 nude mice (Charles River). After 3 wk, tumors were fixed in Bouin’s fixative, processed for histology and sections stained with hematoxylin and eosin. RT-PCR Total RNA was isolated from whole 11.5 dpc gonadal ridge or purified 11.5 dpc PGCs and somatic cells. RT-PCR was performed as described (36). Adult uterus and 16.5 dpc heart served as positive controls for ER␣ and ERR␣, respectively. PCR primers were designed from the murine sequence (GenBank accession no. M38651 for ER␣ and U85259 for ERR␣) and purchased from DNA Technology (Aarhus, Denmark). The sequence for ER␣ of the sense primer was 5⬘-ATTGACAAGAACCGGAG-3⬘ and that of the antisense was 5⬘-ATAGATCATGGGCGGTTCAG-3⬘. For ERR␣, the sense primer was 5⬘-GAAAGTGAATGCCCAGGTGT-3⬘ and for antisense 5⬘-GGAGATCGGATTAAGCAGCA-3⬘ was used. For ER␤ the sense primer was 5⬘-GAA GCT GGCTGACAAGGAAC-3⬘ and the antisense was 5⬘-GTGTCAGCTTCCGGCTACTC-3⬘. For ERR␤ the sense primer was 5⬘-GATGCCCTCAGCCACCAC-3⬘ and the antisense was 5⬘-CAG CCG TCGCTTGTACTTCT-3⬘. Identity of the PCR products was confirmed by direct sequencing by BMR Bio Molecular Research (Padova, Italy).



2637

Western Blot Analysis

Statistical Analysis

Western blotting analyses were performed according to standard methods. Briefly, after 10% SDS-PAGE electrophoresis, proteins were transferred to nitrocellulose, blocked with TBST/5% nonfat milk or 5%BSA and incubated with specific antibodies: anti-ER␣ (kindly provided by National Hormone & Pituitary Program, Harbor-UCLA Medical Center, Torrance, CA), anti-KL (R&D Systems, Abingdon, UK), anti-LIF (Santa Cruz Biotechnology, Santa Cruz, CA), antiactin (Sigma), antiP-PTEN-Ser380 and PTEN (Cell Signaling, Beverly, MA) or anti-P-Akt-Ser473 and anti-Akt (Cell Signaling). After exposure to the secondary antibodies (Amersham Pharmacia Biotech), blots were developed by enhanced chemiluminescence (Amersham Pharmacia Biotech). The band intensity was quantified using actin as internal quantitative control and SigmaGel software (Jandel Scientific, Chicago, IL). For each determination, at least two experiments were performed.

Data are expressed as mean ⫾ SEM of at least three independent experiments with three replicates per experimental group. Comparisons were made by one-way ANOVA, and significance was accepted at the 0.05 level of probability. P ⬍ 0.05 was considered to be significant.

Acknowledgments We are grateful to Mr. Graziano Bonelli for his expert assistance with the preparation of figures and Ms. Maddalena Vecchione for aid in the preparation of the manuscript. We thank Paula M. De Angelis, Ph.D., Robin Hobbs, Ph.D., and Maurizio Pesce, Ph.D. for critical reading of the manuscript and Prof. Steinar Tretli for advice and helpful discussion. We are in debt with Dr. L. Scaldaferri for help in the preparation of samples for immunofluorescence.

Immunohistochemistry Immunohistochemistry was performed on paraffin or cryostat sections of 11.5 dpc gonadal ridges using the antibody against ER␣ (National Hormone & Pituitary Program, HarborUCLA Medical Center, Torrance, CA) and a protocol as reported (25). For PTEN immunocytochemistry, PGCs in culture were fixed with methanol for 10 min at 4 C, washed extensively with PBS ⫹ 10% BSA and incubated overnight with 1:300 TG-1 and 1:150 PTEN (Cell Signaling, Beverly, MA) primary antibodies in PBS ⫹ BSA at 4 C. Finally, cells were labeled with secondary antibody (TRITC conjugated antimouse IgM and fluorescein isothiocyanate-conjugated antirabbit IgG, respectively) for 30 min at room temperature, washed and mounted in 90% glycerol/PBS.

Received January 8, 2003. Accepted September 10, 2003. Address all correspondence and requests for reprints to: Prof. Massimo De Felici, Dipartimento di Sanita` Pubblica e Biologia Cellulare, Universita` di Roma “Tor Vergata,” Via Montpellier 1, 00133 Roma, Italy. E-mail: defelici@ uniroma2.it. This work was supported by Ministero dell’Universita` e della Ricerca Scientifica e Tecnologica National Project 2000, by Public Health Ministry Grant 99/c/T1 (Italy), the European Community Project “GeneDisrup” (Contract No. QLK4-CT02403) and by the Norwegian Research Council (129609/310). The first two departments listed contributed equally to this work.

Reporter Gene Assay TM4 cells were transfected with 2 ␮g ␤-galactosidase plasmid and 3.5 ␮g of the pTAL-Luc (CLONTECH, Palo Alto, CA) reporter plasmid that contained a luciferase gene under transcriptional control of the thymidine kinase promoter and a Steel gene promoter AP-1 response element as described (37). Briefly, the construct 5⬘T TTA ATC CTG AGT CAC TTG TTT TC3⬘ (893–916) from Mus musculus mast cell KL gene 5⬘ flanking region U44724 (gi 1172214) was inserted in its normal orientation into a unique XhoI site of the pTAL-Luc reporter plasmid (CLONTECH). The transfected TM4 cells were treated in charcoal stripped and phenol red-free medium with 4 nM E2 or 10 ␮M ZEA. After 48 h, cells were harvested. Luciferase activity was assayed according to the manufacturer (Promega, Madison, WI), ␤-galactosidase activity was assayed as described (38). Luciferase activity was normalized to ␤-galactosidase activity, and the difference relative to the empty vector was calculated. Antisense Treatment Two PTEN AS were constructed according to the sequence obtained from the GenBank accession no. NM 008960. AS1: (5⬘-GCTCAACTCTCAAACTTCCAT-3⬘; 43% GC; corresponds to nucleotides 153–173 and AS2: (5⬘-GCCGCCGCCGTCTCTCATCTC-3⬘; 71% GC; corresponds to nucleotides 269–289). A sense oligonucleotide (S) control was (5⬘GAGATGAGAGACGGCGGCGGC-3⬘; 71% GC). The oligonucleotides were obtained from GIBCO Life Technologies. PGCs were subjected to transfection with oligonucleotides after the procedures as described (39, 40). A sense oligonucleotide control was performed for all treatments. The expression of PTEN in cultured PGCs subjected to oligo treatment was analyzed as reported above.

REFERENCES 1. Matin A, Collin GB, Varnum DS, Nadeau JH 1998 Testicular teratocarcinogenesis in mice—a review. APMIS 106:174–182 2. Sano K 1995 So-called intracranial germ cell tumours: are they really of germ cell origin? Br J Neurosurg 9:391–401 3. Toppari J, Larsen JC, Christiansen P, Giwercman A, Grandjean P, Guillette Jr LJ, Jegou B, Jensen TK, Jouannet P, Keiding N, Leffers H, McLachlan JA, Meyer O, Muller J, Rajpert-De Meyts E, Scheike T, Sharpe R, Sumpter J, Skakkebc´ k NE 1997 Testicular cancer and cryptorchidism in relation to pre-natal factors: casecontrol studies in Denmark. Cancer Causes Control 8:904–912 4. Møller H, Skakkebc´ k N 1996 Male reproductive health and environmental xenoestrogens. Environ Health Perspect 104:741–803 5. Weir H-K, Marret LD, Kreiger N, Darlington GA, Sugar L 2000 Pre-natal and peri-natal exposures and risk of testicular germ-cell cancer. Int J Cancer 87:438–443 6. Cantley LC, Neel BG 1999 New insights into tumor suppression: PTEN suppresses tumor formation by restraining the phosphoinositide 3-kinase/AKT pathway. Proc Natl Acad Sci USA 96:4240–4245 7. De Felici M 2000 Regulation of primordial germ cell development in the mouse. Int J Dev Biol 44:575–580 8. Matsui, Y, Zsebo K, Hogan BLM 1990 Embryonic expression of a haemopoietic growth factor encoded by the Sl locus and the ligand for c-kit. Nature 347:667–669 9. Manova, K, Huang, EJ, Angeles M, DeLeon V, Sanchez S, Pronovost SM, Besmer P, Bachvarova RF 1993 The ex-


10. 11.

12.

13.

14. 15.

16. 17. 18. 19.

20. 21. 22.

23. 24.

25.

26.

pression pattern of the c-kit ligand in the gonads of mice supports a role for the c-kit receptor in oocyte growth and in proliferation of spermatogenesis. Dev Biol 157: 85–99 De Felici M, Di Carlo A, Pesce M 1996 Role of stem cell factor in somatic-germ cell interactions during prenatal oogenesis. Zygote 4:349–351 Vanacker JM, Pettersson K, Gustafsson JÅ, Laudet V 1999 Transcriptional targets shared by estrogen receptor-related receptors (ERRs) and estrogen receptor (ER)␣, but not by ER␤. EMBO J 18:4270–4279 Paech K, Webb P, Kuiper GG, Nilsson S, Gustafsson J, Kushner PJ, Scanlan TS 1997 Differential ligand activation of estrogen receptors ER␣ and ER␤ at AP1 sites. Science 277:1508–1510 Bedell MA, Copeland NG, Jenkins NA 1996 Multiple pathways for Steel regulation suggested by genomic and sequence analysis of the murine Steel gene. Genetics 142:927–934 Blume-Jensen P, Hunter T 2001 Oncogenic kinase signalling. Nature 411:355–365 Smolich BD, Yuen HA, West KA, Giles FJ, Albitar M, Cherrington JM 2001 The anti-angiogenic protein kinase inhibitors SU5416 and SU6668 inhibit the KL receptor (c-Kit) in a human myeloid leukaemia cell line and in acute myeloid leukaemia blasts. Blood 97:1413–1421 Vazquez F, Ramaswamy S, Nakamura N, Sellers WR 2000 Phosphorylation of the PTEN tail regulates protein stability and function. Mol Cell Biol 20:5010–5015 Resnick JL, Bixler LS, Cheng L, Donovan PJ 1992 Longterm proliferation of mouse primordial germ cells in culture. Nature 359:550–551 Matsui Y, Zsebo K, Hogan BLM 1992 Derivation of pluripotential embryonic stem cells from murine primordial germ cells in culture. Cell 70:841–847 Kishimoto T, Nakatsuj N 1997 Functional requirement of gp-130 mediated signaling for growth and survival of mouse primordial germ cells in vitro and derivation of embryonic germ (EG) cells. Development 122:1235–1242 Stevens LC 1967 Origin of testicular teratomas from primordial germ cells in mice. J Natl Cancer Inst 38: 549–552 Stevens LC Spontaneous and experimentally induced testicular teratomas in mice. Cell Diff 15:69–74 Couse JF, Hewitt SC, Bunch DO, Sar M, Walker VR, Davis BJ, Korach KS 1999. Postnatal sex reversal of the ovaries in mice lacking estrogen receptors ␣ and ␤. Science 286:2328–2331 Couse JF, Korach KS 1999 Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 20: 358–417 Jefferson WN, Couse JF, Banks EP, Korack KS, Newbold RR 2000 Expression of estrogen receptor ␤ is developmental regulated in reproductive tissues of male and female mice. Biol Reprod 62:310–317 Greco TL, Furlow JD, Duello TM, Gorski J 1991 Immunodetection of estrogen receptors in fetal and neonatal female mouse reproductive tracts. Endocrinology 129: 1326–1332 Rajpert-De Meyts E, Skakkebc´ k NE 1994 Expression of the c-Kit protein product in carcinoma-in-situ and invasive testicular germ cell tumours. Int J Androl 17:85–92


27. Weng LP, Brown JL, Eng C 2001 PTEN coordinates G (1) arrest by down-regulation cyclin D1 via its protein phosphatase activity and up-regulation p27 via its lipid phosphatase activity in a breast cancer model. Hum Mol Genet 10:599–604 28. Tsai EM, Wang SC, Lee J N, Hung MC 2001 Akt activation by estrogens in receptor-negative breast cancer cells. Cancer Res 61:8390–8392 29. Blume-Jensen P, Jang G, Hyman R, Lee, KF, O’Gorman, S, Hunter, T 2000 Kit/stem cell factor receptor-induced activation of phosphatidylinositol 3⬘-kinase is essential for male fertility. Nat Genet 24:157–162 30. Kissel H, Timokhina L, hardy MP, Rotshild G, Tajima Y, Soares V, Angeles M, Whitow, SR, Manova K, Besmer P 2000 Point mutation in kit receptor tyrosine kinase reveals essential roles for kit signaling in spermatogenesis and oogenesis without affecting other kit responses. EMBO J 19:1312–1226 31. De Miguel PM, Cheng L, Holland EC, Federspiel MJ, Donovan P 2002 Dissection of the c-Kit signalling pathway in mouse primordial germ cells by retroviral-mediated gene transfer. Proc Natl Acad Sci USA 99: 10458–10463 32. Kimura T, Suzuki A, Fujjita Y, Yomogida K, Lomell H, Asada N, Ikeuci M, Nagy A, Mak TW, Nakano T 2003 Conditional loss of PTEN leads to testicular teratoma and enhances embryonic germ cell production. Development 130:1691–1700 33. Dieckmann KP, Endsin G, Pichlemeir U 2001 How valid is the prenatal estrogen excess hypothesis of testicular germ cell cancer? A case control study on hormonerelated factors. Eur Urol 40:677–683 34. De Felici M 1988 Isolation and culture of germ cells from the mouse embryo. In: Cell biology: a laboratory handbook. Vol I. 2nd ed. New York: Academic Press; 73–85 35. De Felici M, Dolci S, Pesce M 1993 Proliferation of mouse primordial germ cells in vitro: a key role for cAMP. Dev Biol 157:227–280 36. Pesce M, Canipari R, Ferri G-L, Siracusa G, De Felici M 1996 Pituitary adenylate cyclase-activating polypeptide (PACAP) stimulates adenylate cyclase and promotes proliferation of mouse primordial germ cells. Development 122:215–221 37. Gue´ rin SL, Pothier F, Robidoux S, Gosselin P, Parker M 1990 Identification of a DNA-binding site for the transcription factor GC2 in the promoter region of the p12 gene and repression of its positive activity by upstream negative regulatory elements. J Biol Chem 265: 22035–22043 38. Ausbel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K, Albright LM, Coen DM, Varki A, Chanda VB 1987 In: Current protocols in molecular biology. New York: John Wiley, Sons, Inc.; 13.6.2.–13.6.5 39. Lachyankar MB, Sultana N, Schonhoff CM, Mitra P, Poluha W, Lambert S, Quesenberry PJ, Litofsky NS, Recht LD, Nabi R, Miller SJ, Ohta S, Neel BG, Ross AH 2000 A role of nuclear PTEN in neuronal differenziation. J Neurosci 20:1404–1413 40. Lebedeva I, Stein CA 2001 Antisense oligonucleotides: promise and reality. Annu Rev Pharmacol Toxicol 41: 403–419