ORIGINAL RESEARCH ARTICLE 3749 JournalJournal ofof Cellular The High-Mobility Group Physiology A1-Estrogen b Nuclear Interaction Is Impaired in Human Testicular Seminomas FRANCESCO ESPOSITO, 1,2 FRANCESCA BOSCIA, 3 VINCENZO GIGANTINO, 4 2 2 4 1 MARA TORNINCASA, ALFREDO FUSCO, RENATO FRANCO, AND PAOLO CHIEFFI * 1Dipartimento di Medicina Sperimentale, II Universita ` di Napoli, Via Costantinopoli, Naples, Italy 2Dipartimento di Biologia e Patologia, Universita ` degli Studi di Napoli ‘‘Federico II,’’ Via Pansini, Naples, Italy 3Dipartimento di Neuroscienze, Universita ` degli Studi di Napoli ‘‘Federico II,’, Via Pansini, Naples, Italy 4Istituto Nazionale dei Tumori ‘‘Fondazione G. Pascale,’’ Via Semmola, Naples, Italy

It is well established that estrogens participate in the control of normal spermatogenesis and endogenous or environmental estrogens are involved in pathological germ cell proliferation including testicular germ cell tumors. The effects of estrogen are now known to be mediated by - a (ER a) and ER b subtypes, but only ER b has been found in human germ cells of normal testis. However, its expression was markedly diminished in human testicular seminomas. The expression and the possible interaction of ER b and HMGA1 were studied in normal germ cells and in human testicular seminomas. GC1 and TCam-2 germ cell lines, were used; in addition, a tissue micro-array (TMA) was built using the most representative areas from 35 cases of human testicular seminomas. The expression and the interaction of ER b and HMGA1 were observed by using immunoprecipitation and Western blot analyses in combination with immunocytochemistry and immunofluorescence analyses. Here, we show that ER b interacts with HMGA1 in normal germ cells, while down regulation of ER b associates with transcriptional co-regulator HMGA1 over-expression and cytoplasmic localization both in human testicular seminomas and in TCam-2 cell line. In addition, we show that 17 b-oestradiol induces an HMGA1 increased cytoplasmic expression associated to an ER b down-regulation in TCam-2 cell line. Taken together, our results suggest that exposure to estrogens or estrogen-mimics, in some as of yet undefined manner, diminishes the ER b-mediated growth restraint in human testicular seminoma, probably due to the HMGA1 cytoplasmic delocalization associated with ER b down-regulation. J. Cell. Physiol. 227: 3749–3755, 2012. ß 2012 Wiley Periodicals, Inc.

Testicular germ cell tumors (TGCTs), the most common association with a myriad of co-activators and repressors, act as malignancy in males between 15 and 34 years of age, represent a nuclear transcription factors for targeted genes (Pettersson and major cause of death attributable to cancer in this age group Gustafsson, 2001). It has been well documented in literature (Oosterhuis and Looijenga, 2005; Chieffi, 2007, 2011; Chieffi that ER b, which is expressed in normal testicular cells, is instead et al., 2009). TGCTs can be subdivided into seminoma and non- down regulated in seminomas and embryonal cell carcinomas seminoma germ cell tumors (NSGCTs), including embryonal (Pais et al., 2003; Esposito et al., 2011). cell carcinoma, choriocarcinoma, yolk sac tumor, and teratoma. The mammalian high mobility group A (HMGA) family of Neoplasms containing more than one tumor cell components, chromosomal proteins includes HMGA1a and HMGA1b, which for example, seminoma and embryonal cell carcinoma, are are encoded by the same gene, HMGA1 , through alternative referred to as mixed germ cell tumors. Seminomas and splicing (Johnson et al., 1989), and the closely related HMGA2 NSGCTs do not only present distinctive clinical features, but (Wood et al., 2000). They are small, non-, chromatin- they also show significant differences as far as therapy and associated proteins that bind DNA in AT-rich regions through prognosis are concerned (Oosterhuis and Looijenga, 2005; three basic domains called ‘‘AT-hooks’’ (Johnson et al., 1989; Chieffi, 2007, 2011; Chieffi et al., 2009). Whereas the etiology Wood et al., 2000). Members of the HMGA protein family have of TGCTs remain undefined, exposure to certain hormones (i.e., estrogen) at the time of testicular differentiation in utero has long been implicated as a risk factor for developing these neoplasmas (Henderson et al., 1979). Recently, it has been suggested that the carcinogenic effects of estrogen on testicular The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research cells may involve estrogen receptor-mediated oxidative DNA reported. damage (Wellejus and Loft, 2002). Estrogen signaling is mediated by two nuclear receptors, Contract grant sponsor: Italian Ministry of University and Research; estrogen receptor- a (ER a) and b (ER b), that are estrogen Contract grant numbers: MIUR-PRIN 2007, N. 2007KR7PRY_002. dependent transcription factors. ER a is expressed at high levels *Correspondence to: Paolo Chieffi, Dipartimento di Medicina in human epididymis and efferent ductules, but not in the testis, Sperimentale, II Universita` di Napoli, Via Costantinopoli 16, 80138 whereas ER b is expressed in spermatogonia, spermatocytes, Naples, Italy. E-mail: paolo.chieffi@unina2.it and in early round spermatids in human testis (Saunders et al., Received 9 February 2012; Accepted 24 February 2012 2002; Vicini et al., 2006). The ER b subtype is the principal Published online in Wiley Online Library mediator of estrogen action in promoting germ cell survival and (wileyonlinelibrary.com): 5 March 2012. development (Pentikainen et al., 2000; Saunders et al., 2002; DOI: 10.1002/jcp.24087 Vicini et al., 2006). After activation, these receptors, in

ß 2012 WILEY PERIODICALS, INC. 3750 ESPOSITO ET AL.

no intrinsic transcriptional activity, but they can regulate recombinant fragment of estrogen receptor a (ER a; Chieffi et al., transcription by altering the architecture of chromatin, so 2003); (f) polyclonal rabbit antibody anti-estrogen receptor- b furthering the assembly of multiprotein complexes of (ER b, #sc-8974, Santa Cruz Biotechnology, Inc., Santa Cruz, CA); transcriptional factors (Thanos and Maniatis, 1992). Both genes (g) mouse monoclonal anti-estrogen receptor- b (clone PPG5/10 are widely expressed during embryogenesis (Thanos and code #M7292 DakoCytomation, Copenaghen, Denmark); Maniatis, 1992), and are over-expressed in several (h) polyclonal mouse antibody anti-PATZ1 (#ab. 68646; Abcam, experimental and human malignancies (Zhou et al., 1995). We Cambridge, UK); (i) mouse monoclonal anti g-Tubulin (#T-5326, have shown that HMGA1 and HMGA2 are present in mitotic Sigma Chemical Corporation, St. Louis, MO); and (j) polyclonal germ cells (spermatogonia and primary spermatocytes), rabbit anti-SP1 (#sc-14027 Santa Cruz Biotechnology, Inc., Santa whereas only HMGA2 is highly expressed in meiotic and post- Cruz, CA). meiotic cells (secondary spermatocytes and spermatids; Chieffi et al., 2002; Di Agostino et al., 2004). More recently, we have TMA building shown that HMGA1 and HMGA2 represent a valuable Tissue micro-array (TMA) was built using the most representative diagnostic tool in same cases of problematic histological areas from each single case. All tumors and controls were reviewed differential diagnosis and to assess correlation with differential by two experienced pathologists. Discrepancies between two histological diagnosis (Franco et al., 2008). In fact, we found that pathologists from the same case were resolved in a joint analysis of they are differently expressed with respect to the state of the cases. Tissue cylinders with a diameter of 0.6 mm were differentiation of TGCTs, with overexpression of HMGA1 in punched from morphologically representative tissue areas of each seminomas, with overexpression of both proteins in ‘‘donor’’ tissue block and brought into one recipient paraffin block pluripotential embryonal carcinoma cells, and loss of (3 cm  2.5 cm) using a semiautomated tissue arrayer (Galileo expression of HMGA1 in yolk sac tumors and of both proteins TMA, Milan, Italy). in mature adult tissue of teratoma areas (Franco et al., 2008). Indeed, HMGA1 is able to bind proteins involved in Immunohistochemistry analysis transcriptional regulation such as RNF4 (Pero et al., 2001, 2003) and PATZ1 that it has been shown to be delocalized and over- Immunohistochemical staining was done on slides from formalin- expressed in human testicular seminomas (Fedele et al., 2008). fixed, paraffin embedded tissues, to evaluate the expression of More recently, we have shown that PATZ1 interacts with ER b HMGA1 and ER b. Antigen retrieval was performed with slides in normal germ cells, while down regulation of ER b associates heated in 0.01 M citrate buffer (pH 6.0 for HMGA1) or Tris–EDTA 8 with transcriptional co-regulator PATZ1 delocalization in (pH 9 for ER b) in a bath for 20 min at 97 C. After antigen retrieval, human testicular seminomas (Esposito et al., 2011). the slides allow to cool (Chieffi et al., 2002; Franco et al., 2008). The In the present study, our results indicate for the first time slides were rinsed with TBS and the endogenous peroxidase was that HMGA1, but not HMGA2, interacts with ER b, and also inactivated with 3% hydrogen peroxide. After protein block (BSA with PATZ1 in normal germ cells. Moreover, in human 5% in PBS 1 Â), the sections were incubated overnight with testicular seminomas, HMGA1 is over-expressed and antibodies against: (a) HMGA1 (diluted 1: 200; Chieffi et al., 2002; delocalized in the cytoplasm, and it associates with ER b down Franco et al., 2008); and (b) estrogen receptor- b (diluted 1:200, regulation. In addition, we show that 17 b-oestradiol (E2) mouse monoclonal anti-estrogen receptor- b, code #M7292 induces an HMGA1 increased cytoplasmic expression Dako). The sections were rinsed in TBS and incubated for 20 min associated with ER b down-regulation in TCam-2 seminoma with Novocastra Biotinylated Secondary Antibody (RE7103), a cells. biotin-coniugated secondary antibody formulation that recognized Taken together, these results suggest that the failed mouse and rabbit immunoglobulins. Then, the sections were rinsed association between the two proteins might have a role in the in TBS and incubated for 20 min with Novocastra Streptavidin-HRP genesis of the testicular neoplasia. (RE7104), and then peroxidase reactivity was visualized using a 3,3 0-diaminobenzideine (DAB). Finally, the sections were counterstained with hematoxylin and mounted. Results are Materials and Methods interpreted using a light microscope. Immunohistochemistry Tissue samples and cell culture expression was evaluated independently by a pathologist. Using a As source of neoplastic tissues, the tissue Bank of National Cancer semiquantitative method, percentage of positive cells for HMGA1 Institute ‘‘G. Pascale’’ provided 35 cases of seminomas. Ethical and ER b was evaluated and high expression was recorded when Committee approval was given in all instances. The GC-1 and positive cells were >30%. Moreover, for HMGA1 expression, TCam-2 cell lines were cultured as previously described (Hofmann cytoplasmic delocalization was considered. et al., 1992; Mizuno et al., 1993). As source of normal germ cells, Immunofluorescence confocal microscopy 5 CD1 adult mice (Charles River, Italy) were killed and testes removed and processed as previously described (Fedele et al., Confocal immunofluorescence procedures were performed as 2008). TCam-2 cells were also treated with 17 b-oestradiol (E2, described (Boscia et al., 2009a,b). Four micrometer sections of 100 mM; Sigma, Milan, Italy), and with 8-Bromoadenosine-3 0,5 0- cryopreserved seminomas were used. GC1 and TCam-2 cells cyclic monophosphate (8Br-cAMP, 100 mM; Sigma, Milan, Italy). cultured on glass coverslips coated with 30 mg/ml poly- L-lysine (Sigma, St. Louis, MO) were rinsed in 50 mM Tris-buffered Antibodies saline pH 7.4 (TBS) and fixed in 4% paraformaldehyde (Sigma, Milan, Italy) for 30 min. The coverslips were first blocked in 3% bovine Antibodies were purchased from the following sources: (a) serum albumin (BSA; Sigma, Milan, Italy) and 0.05% Triton X-100 polyclonal rabbit antibody anti-HMGA1 raised against a synthetic (Bio-Rad, Milan, Italy) for 1 h and then incubated with the following peptide located in the NH2-terminal region (Chieffi et al., 2002; primary antibodies: rabbit polyclonal anti-HMGA1 (1:500), mouse Franco et al., 2008); (b) polyclonal rabbit antibody anti-HMGA1 monoclonal anti-estrogen receptor- b (diluted 1:300), at 4 8C (#Ab. 4078, Abcam, Cambridge, UK); (c) polyclonal rabbit antibody overnight. Next, the cells were incubated in a mixtures of anti-HMGA2 raised against a synthetic peptide located in the NH2- fluorescent-labeled secondary antibodies Alexa 488-conjugated terminal region (Chieffi et al., 2002; Franco et al., 2008); (d) anti-mouse and Alexa 594-conjugated anti-rabbit IgGs (diluted polyclonal rabbit antibody anti-HMGA2 (#Ab. 41878, Abcam, 1:200; Molecular Probes, Eugene, OR) for 1 h. Cell nuclei were Cambridge, UK); (e) a mixture of two monoclonal antibodies stained with Hoechst-33258 (Sigma, Milan, Italy). (#1062 and #1063) raised against a C-terminal c-DNA Immunofluorescence images were observed using a Zeiss LSM 510

JOURNAL OF CELLULAR PHYSIOLOGY HMGA1 AND ER b INHUMANTESTICULARSEMINOMAS 3751

META Laser-Scanning Confocal Microscope (Carl Zeiss, Thornwood, NY). Single images, taken with an optical thickness of 0.7 mm and a resolution of 1,024 Â 1,024 pixels, were used for illustrations.

Protein extraction, immunoprecipitation, and Western blot analysis Total cell extracts (TCE) and mouse germ cells were prepared as previously described (Fedele et al., 2008). For immunoprecipitations, protein A-Sepharose (3 mg  sample) was incubated with rabbit anti-HMGA1, and anti-HMGA2 antibodies (5 ml  sample) for 1 h at 4 8C. Five hundred microgram of total protein was added to the anti-HMGA1, and anti HMGA2 antibodies-Protein A mix for 2 h at 4 8C. Immunoprecipitates were subjected to SDS–PAGE (10% polyacrylamide) under reducing Fig. 1. Immunoprecipitation and Western blot analyses with anti-HMGA1, anti-HMGA2, anti-ER b, and anti-PATZ1 antibodies in condition. GC1 cells. Cell lysates were immunoprecipitated with anti (A) Differential extraction of nuclear or cytoplasmic proteins were HMGA1, HMGA2 (B), resolved by SDS–PAGE, transferred to obtained as previously described (Chieffi et al., 2000a,b, 2004; nitrocellulose membranes and challenged with anti-ER b or anti- Esposito et al., 2009). Membranes were blocked with 5% non-fat HMGA1 or anti-HMGA2, or anti-PATZ1 antibodies. C: Immunoprecipitation and Western blot detection in adult mice germ milk proteins and incubated with the following primary antibody: cells. Cell lysates were immunoprecipitated with HMGA1, resolved (a) anti-HMGA1 (diluted 1:500); (b) anti-HMGA2 (diluted 1:500); by SDS–PAGE, transferred to nitrocellulose membranes and (c) anti-ER b (diluted 1:1,000); (d) anti-PATZ1 (diluted 1:1,000); (e) challenged with anti-ER b, anti-HMGA1, and anti-PATZ1. TCE, total anti g-tubulin (diluted 1:1,000); and (f) anti SP1 (diluted 1:1,000). cell extracts. The blots are representative of three separate experiments. For Western blot analyses additional antibodies were used for HMGA1, HMGA2, and ER b which gave the same results (not shown). Bound antibodies were detected by the horseradish peroxidase-conjugated secondary antibodies followed by enhanced chemiluminescence (Amersham, Life Science, Bucks, HMGA1 and ER b protein expression in germ cell lines UK). SP1 and g-tubulin were used as controls of fractionated The expression of HMGA1 and ER b protein expression have proteins. been analyzed in GC1 and TCam-2 cell lines, respectively, derived from immortalized type B murine spermatogonia and Statistical analysis human seminoma (Hofmann et al., 1992; Mizuno et al., 1993). Confocal double immunofluorescence staining indicated that Data are expressed as mean Æ SEM of values obtained in three HMGA1 and ER b immunoreactivities were co-expressed in separate experiments. Statistical comparison between GC1 cells nuclei of GC1 cells (Fig. 2A,H). Quantitative analysis showed and TCam-2 cells were performed using Student’s t- test. P < 0.05 that 50% of GC1 nuclei expressed HMGA1 or ER b proteins, was considered significant. and among them 90% of nuclei co-localized (Fig. 2A–H). By contrast, in the human derived seminoma TCam-2 cells, less than 10% of nuclei express either HMGA1 or ER b proteins. In Results particular, in these cells (90%) HMGA1 expression was HMGA1, but not HMGA2, interacts with ER b confined to the cytoplasmic compartment (Fig. 2I–P). Our data indicate that HMGA1, but not HMGA2, associates Interestingly, in these cells, showing HMGA1 delocalization, no with ER b (Fig. 1A,B) in GC1 cell line; in fact, total protein GC1 ER b immunosignal was observed (Fig. 2I–P). cell lysates were immunoprecipitated with anti-HMGA1, In agreement with the immunofluorescence results, submitted to SDS–PAGE, and then immunoblotted against ER b, Western blot analysis of the cytoplasmic and nuclear protein and HMGA1 antibodies. Co-immunoprecipitation of fractions from either GC1 or TCam-2 cells showed that both endogenous ER b and HMGA1 was detected (Fig. 1A). Since we, HMGA1 and ER b protein expression were abundantly recently, demonstrated the interaction between ER b and observed both in the nuclear and cytoplasmic fractions of GC1 PATZ1 (Esposito et al., 2011), we explored if PATZ1 also co- cells. By contrast, in TCam-2 cells, HMGA1 protein was largely immunoprecipitated with both HMGA1 and ER b; as shown we detected in the cytoplasmic fraction. In this latter fraction, ER b detected co-immunoprecipitation of endogenous PATZ1, expression was undetectable (Fig. 3). In the nuclear fraction of HMGA1, and ER b (Fig. 1A). By contrast, when total protein TCam-2 cells, both HMGA1 and ER b expression were barely GC1 cell lysates were immunoprecipitated with anti-HMGA2, detectable (Fig. 3). HMGA2 was detected in the nuclear submitted to SDS–PAGE, and then immunoblotted against ER b, compartment of GC1 cells, while we did not observe HMGA2 HMGA2, and PATZ1 antibodies no co-immunoprecipitation of expression in TCam-2 cell line derived from human seminoma; endogenous HMGA2, ER b, and PATZ1 was detected (Fig. 1B). these findings are in accordance with our previously published Next, to further confirm the relationship between HMGA1 data in which we never observed the expression of HMGA2 in and ER b, we investigated the HMGA1 and ER b protein human testicular seminomas (Franco et al., 2008). interaction in adult mice germ cells (Chieffi et al., 2002; Vicini ER b down-regulation associates with HMGA1 et al., 2006). Immunoprecipitation analysis clearly indicated that delocalization in human seminomas HMGA1 and ER b interacted; in fact, total protein mice germ cells lysates were immunoprecipitated with anti-HMGA1 Recently, we have described the increased expression of antibody, submitted to SDS–PAGE, and then immunoblotted HMGA1 in human testicular seminomas (Franco et al., 2008); against ER b antibody; co-immunoprecipitation of endogenous here, we wanted to observe the correlation of HMGA1 and ER b and HMGA1 was detected (Fig. 1C). Furthermore, PATZ1 ER b expression by using a tissue microarray assay (TMA). We co-immunoprecipitated with endogenous ER b and HMGA1 have examined samples from 35 patients with seminomas by in normal germ cells obtained from adult mice (Fig. 1C). immunohistochemistry using different antibodies raised against

JOURNAL OF CELLULAR PHYSIOLOGY 3752 ESPOSITO ET AL.

Fig. 2. Co-localizationofHMGA1proteinandER binGC1andTCam-2cells.LowmagnificationimagesshowingGC1(A–D)orTCam-2(I–L)cells double-labelled for HMGA1 and ER b. Higher magnification images showing GC-1 (E–H) or TCam-2 (M–P) cells double-labelled for HMGA1 and R R R ER b. Quantification of HMGA1 cells, ER b cells, HMGA1/ER b double-labelled cells counted in 10 non-overlapping microscope fields from M five coverslips of GC-1 and TCam-2 cells. Data are expressed as percentages of total Hoechst-positive nuclei (Q–S); P < 0.05 versus TCam-2 cells (Q–S). Scale bars: A–D 20 mm; E–H 10 mm; I–L 50 mm; M–P 20 mm. [Color figure can be seen in the online version of this article, available at http://wileyonlinelibrary.com/journal/jcp]

human HMGA1 and ER b. As expected, in our cases the subunit of PKA, and the translocation of PATZ1 from the HMGA1 protein was highly expressed in seminomas (Table 1). cytoplasm into the nucleus is regulated by cAMP in the In particular, HMGA1 protein was highly expressed in 33 of TCam-2 human seminoma cell line (Esposito et al., 2011). 35 seminomas examined (Figs. 4 and 5, Table 1). These findings To test whether cAMP might be able to promote relocalization are in line with our previous published results (Franco et al., of HMGA1 into the nucleus in a human seminoma cell 2008). More interestingly, we found that when the HMGA1 line, TCam-2 cells were exposed to the second messenger was expressed in the cytoplasmic compartment it was analogue 8Br-cAMP (100 mM), and HMGA1 protein associated with ER b down regulation (Fig. 4A–D), conversely, expression was analyzed by Western blot analysis in the HMGA1 nuclear localization associated with high levels of ER b nuclear and cytoplasmic fractions of TCam-2 cells. We (Fig. 5A–D). did not observe any modification of HMGA1 expression after 0.5, 2, 4, or 6 h of 8Br-cAMP treatment (data not shown). E2 treatment induces an increased HMGA1 cytoplasmic When cells were instead stimulated with E2 (100 mM) for 0.5, 2, localization in TCam-2 seminoma cells 4 h an increased of cytoplasmic HMGA1 expression was observed, when compared to unstimulated cells, associated Recently, we have shown that the ER b interacts with BTB-POZ with ER b and ER a down-regulation, while it decreased at 6 h domain zinc-finger of PATZ1 that interacts also with the RIa (Fig. 6).

JOURNAL OF CELLULAR PHYSIOLOGY HMGA1 AND ER b INHUMANTESTICULARSEMINOMAS 3753

TABLE 1. Immunohistochemical analysis of ER b and HMGA1 in human testicular seminomas by using tissue micro-array (TMA)

Cases Age (years) ER b HMGA1-n HMGA1-c

1 27 þþþ þþþ À 2 32 þ þ þþþ 3 44 þ þ þþþ 4 34 þþþ þþþ À 5 33 À þ þþþ 6 44 À À þþþ 7 33 þ þ þþþ 8 34 þ þ þþþ 9 25 þ À þþ 10 19 þþþ þþþ À 11 35 þ À þþþ 12 37 þ þ þþ 13 27 þ À þþþ 14 27 þþþ þþþ À 15 33 þ þ þþþ 16 31 þþþ þþþ À 17 27 þ þ þþþ 18 25 þ þ þþþ 19 26 þþþ þþþ À 20 35 þ þ þþþ 21 32 À þ þþþ Fig. 3. Western blot analysis of HMGA1, HMGA2, and ER b on 22 53 þ þ þþþ cytoplasmic and nuclear extracts of GC1 and TCam2 cells; 20 mg 23 31 þþþ þþþ À of proteins were resolved on 10% SDS–PAGE, transferred onto 24 29 þ þ þþ nitrocellulose filters and Western blotted with anti-HMGA1, 25 33 þþ þþþ À anti-HMGA2, and anti-ER b rabbit polyclonal antibodies; SP1 and 26 38 þþ þ þ g-Tubulin were used as controls of fractionated proteins. T, total 27 41 À þ þþþ extracts; N, nuclear extracts; C, cytoplasmic extracts. The blots are 28 31 þþþ þ þ representative of three separate experiments. 29 37 þþ þþ À 30 34 þþþ þþþ þ 31 43 þ À þþþ 32 30 þ þ þþþ 33 30 þþ þ þ 34 22 þ þ þþþ 35 24 þþ þþ þ Discussion HMGA1-n, nuclear expression of HMGA1; HMGA1-c, cytoplasmic expression of HMGA1; The down-regulation of ER b observed in seminomas, in À, 0–10% of neoplastic expressing cells; þ, 11–30% of expressing neoplastic cells; þþ , 31– accordance with data from animal models, human cell culture 70% of expressing neoplastic cells; þþþ , >70% of expressing neoplastic cells. studies, and with our previously published results (Esposito et al., 2011), suggests that ER b may control and limit cell proliferation during testicular (seminoma), breast, prostate, ovarian, and colon cancer progression (Pasquali et al., 2001; Roger et al., 2001; Weyant et al., 2001; Staibano et al., 2003). This consideration has led us to hypothesize that exposure to estrogens or estrogen-mimics, in some as of yet undefined manner, diminishes the ER b-mediated growth restraint in spermatagonia, which favors unscheduled cell proliferation. The affected spermatogonia or their descendants may then be able to escape normal cell cycle regulation and be at a higher risk of undergoing malignant transformation. In the current study, we consistently detected that a decreased expression of ER b protein associates with a delocalization of HMGA1 in human testicular seminomas, when compared with normal testes. A hypothesis is that the aberrant expression of HMGA1 suggests an important role of HMGA1 in tumorigenesis (Franco et al., 2008). Moreover, it has been shown that HMGA1, throught PATZ1 interaction, is an (AR) co-regulator that acts by modulating the effect of the AR co-activator RNF4, a protein expressed in normal germ cells but not in human testicular tumors (Pero et al., 2001; Chieffi, 2011). In addition, it is interesting to note that SNURF/RNF4 is a co-regulator of ER b and are both severely repressed in seminomas (Pais et al., 2003; Chieffi, 2011; Esposito et al., 2011). Fig. 4. ImmunohistochemistryanalysisofER b and HMGA1 proteins expression in human testicular seminomas by using anti-HMGA1 and For the first time in this article, we describe that HMGA1 ER b antibodies. Seminoma (case #2) in which an absence of interacts and may function as a novel ER b co-regulator and immunopositivity of ER b (A; B high magnification) was observed in associates also with PATZ1. For several years, the mechanism association with an intense and diffuse HMGA1 cytoplasmic of steroid-receptor action has been considered very simple immunosignal (C; D high magnification) by immunohistochemistry (magnification: A, C 20 T; B, D 40 T). [Color figure can be seen in the consisting in hormone binding followed by nuclear import, online version of this article, available at http:// recognition of specific DNA motifs, and direct transcriptional wileyonlinelibrary.com/journal/jcp] activation (or repression). During the last few years, several

JOURNAL OF CELLULAR PHYSIOLOGY 3754 ESPOSITO ET AL.

indicates that the activation process is complex and is controlled, at least in part, by the cellular ratio of co-regulatory proteins. In addition, it appears always more evident that steroid-receptor co-factors are important in regulating the receptors function through activation or repression of various target genes in endocrine cancers (Barnes et al., 2004). Here, we show that E2 increased the cytoplasmic expression of HMGA1 in TCam-2 cells in which the ER b expression is normally down-regulated. Our observation suggests a novel signaling mechanism in which the sequestration of HMGA1 in the cytoplasm is regulated by E2 in absence of ER b expression in TCam-2 cell line, thus enabling HMGA1 to translocate into nucleus and transactivate its target genes upon activation of the E2 pathway. These observations led to the hypothesis that GPR30 activation could represent an alternative pathway for estrogen-mediated activity in various epithelial tumors that are more often ER negative. Recently, we have published that GPR30 is over-expressed in human testicular seminomas, that are more often ER a/b negative (Franco et al., 2011). In addition, it is well documented that GPR30 is an alternative estrogen receptor that is overexpressed in high-risk endometrial carcinomas where expression of ER and PR are down-regulated (Prossnitz and Barton, 2011). For example, it has been shown in Fig. 5. ImmunohistochemistryanalysisofER b and HMGA1 proteins expression in human testicular seminomas by using anti-HMGA1 ovarian cancer cell lines that GPR30 is able to mediate gene and ER b antibodies. Seminoma (case #4) in which a ER b (A; B high expression changes and proliferation in response to 17 b- magnifiction) and HMGA1 (C; D high magnification) intense nuclear oestradiol and the selective GPR30 agonist G-1 (Prossnitz and positivity was observed by immunohistochemistry (magnification: A, Barton, 2011). Therefore, further understanding of the C 20 T; B, D 40 T).[Color figurecanbeseenintheonlineversionofthis article, available at http://wileyonlinelibrary.com/journal/jcp] interaction of HMGA1 and their effects in growth regulation may offer insights on the mechanisms of E2 mediated growth control. Taken together, our results demonstrate for the first time the interaction between HMGA1, PATZ1, and ER b in normal putative co-regulatory proteins have been identified for nuclear germ cells; in addition, we show that the HMGA1 increased receptors. Although the physiological role of cytoplasmic expression is mediated by E2 in absence of ER b in co-activators remains to be elucidated, it is becoming clear that TCam-2 cells probably through the GPR30 activation. these ligand-inducible transcription factors are capable of Consistently, the HMGA1 cytoplasmic delocalization associating, probably simultaneously, with multiple target associated with ER b down-regulation in human testicular proteins. Identification of an increasing number of proteins with seminomas, could be ascribed to an impaired E2 mediated the potential to enhance or repress steroid-receptor action signaling to generate the testicular neoplasia.

Fig. 6. EffectoncytoplasmicandnuclearexpressionofHMGA1 inTCam-2cells following 0.5, 2, 4,6hofE2 (100 mM) exposure; 20 mg of proteins wereresolvedon10%SDS–PAGE,transferredontonitrocellulosefiltersandWesternblottedwithanti-HMGA1,ER b,andER aantibodies;SP1and g-Tubulin were used as controls of fractionated proteins. N, nuclear extracts; C, cytoplasmic extracts. The blots are representative of three separate experiments.

JOURNAL OF CELLULAR PHYSIOLOGY HMGA1 AND ER b INHUMANTESTICULARSEMINOMAS 3755

Acknowledgments Franco R, Boscia F, Gigantino V, Marra L, Esposito F, Ferrara D, Pariante P, Caraglia M, Minucci S, Chieffi P. 2011. GPR30 is over-expressed in post puberal testicular germ cell This work was supported by grant to Paolo Chieffi from tumors. Cancer Biol Ther 11:609–613. HendersonBE, Benton B, Jing J, Yu MC, Pike MC. 1979.Risk factors for cancer of the testis in Italian Ministry of University and Research (MIUR-PRIN 2007, young men. Int J Cancer 23:598–602. N. 2007KR7PRY_002). Hofmann MC, Narisava S, Hess RA, Milla´n JL. 1992. Immortalization of germ cells and somatic testicular cells using the SV40 large T antigen. Exp Cell Res 201:417–435. Johnson KR, Lehn DA, Reeves R. 1989. Alternative processing of mRNAs coding for human HMGI and HMGY proteins. Mol Cell Biol 9:2114–2123. Mizuno Y, Gotoh A, Kamidono S, Kitazawa S. 1993. Establishment and characterization of a Literature Cited new human testicular germ cell tumor cell line TCam-2. Nippon Hinyokika Gakkai Zasshi 84:1211–1218. Barnes CJ, Vadlamudi RK, Kumar R. 2004. Novel estrogen receptor coregulators and Oosterhuis JW, Looijenga LH. 2005. Testicular germ-cell tumors in a broader perspective. signaling molecules in human diseases. Cell Mol Life Sci 61:281–291. Nature Rev Cancer 5:210–222. Boscia F, Gala R, Pannaccione A, Secondo A, Scorziello A, Di Renzo G, Annunziato L. 2009a. Pais V, Leav I, Lau KM, Jiang Z, Ho SM. 2003. Estrogen receptor-b expression in human NCX1 expression and functional activity increase in microglia invading the infarct core. testicular germ cell tumours. Clin Cancer Res 9:4475–4482. Stroke 40:3608–3617. Pasquali D, Rossi V, Esposito D, Abbondanza C, Puca GA, Bellastella A, Sinisi AA. 2001. Loss Boscia F, Esposito CL, Di Crisci A, de Franciscis V, Annunziato L, Cerchia L. 2009b. GDNF of expressionin malignant humanprostate cells in primary cultures selectively induces microglial activation and neuronal survival in CA1/CA3 hippocampal and in prostate cancer tissues. J Clin Endocrinol Metab 86:2051–2055. regions exposed to NMDA insult through Ret/ERK signalling. PLoS ONE 4:e6486. Pentikainen V, Erkkila K, Suomalainen L, Parvinen M, Dunkel L. 2000. Estradiol acts as a germ Chieffi P. 2007. Molecular targets for the treatment of testicular germ cell tumors. Mini-Rev cell survival factor in the human testis in vitro. J Clin Endocrinol Metab 85:2057–2067. Med Chem 7:755–759. Pero R, Lembo F, Di Vizio D, Boccia A, Chieffi P, Fedele M, Pierantoni GM, Rossi P, Iuliano R, Chieffi P. 2011. New prognostic markers and potential therapeutic targets in human Santoro M, Viglietto G, Bruni CB, Fusco A, Chiariotti L. 2001. RNF4 is a growth inhibitor testicular germ cell tumors. Curr Med Chem 18:5033–5040. expressed in germ cells and lost in human testicular tumours. Am J Pathol 159:1225–1230. Chieffi P, Colucci D’Amato GL, Staibano S, Franco R, Tramontano D. 2000a. Estradiol- Pero R, Lembo F, Chieffi P, Del Pozzo G, Fedele M, Fusco A, Bruni CB, Chiariotti L. 2003. induced mitogen-activated protein kinase (extracellular signal-regulated kinase 1 and 2) Translationalregulationof a novel testis-specific RNF4transcript. Mol Reprod Dev 66:1–7. activity in the frog (Rana esculenta) testis. J Endocrinol 16:77–84. Pettersson K, Gustafsson JA ˚ . 2001. Role of estrogen receptor beta in estrogen action. Ann Chieffi P, Franco R, Fulgione D, Staibano S. 2000b. PCNA in the testis of the frog, Rana Rev Physiol 63:165–192. esculenta: A molecular marker of the mitotic testicular epithelium proliferation. Gen Prossnitz ER, Barton M. 2011. The G-protein-coupled estrogen receptor GPER in health and Comp Endocrinol 119:11–16. disease. Nature Rev Endocrinol 7:715–726. Chieffi P, Battista S, Barchi M, Di Agostino S, Pierantoni G, Fedele M, Chiariotti L, Roger P, Sahla ME, Makela S, Gustafsson JA, Baldet P, Rochefort H. 2001. Decreased Tramontano D, Fusco A. 2002. HMGA1 and HMGA2 protein expression in mouse expression of estrogen receptor beta protein in proliferative preinvasive mammary spermatogenesis. Oncogene 21:3544–3550. tumors. Cancer Res 61:2537–2541. Chieffi P, Kisslinger A, Sinisi A, Abbondanza C, Tramontano D. 2003. 17(-estradiol-induced Saunders PT, Millar MR, Macpherson S, Irvine DS, Groome NP, Evans LR, Sharpe RM, Scobie activation of ERK1/2 through endogenous androgen receptor–estradiol receptor GA. 2002. ER b1 and the ER b2 splice variant (ER bcx/ b2) are expressed in distinct cell (-Src complex in human prostate cells. Int J Oncol 23:797–801. populations in the adult human testis. J Clin Endocrinol Metab 87:2706–2715. Chieffi P, Troncone G, Caleo A, Libertini S, Linardopoulos S, Tramontano D, Portella G. Staibano S, Franco R, Mezza E, Chieffi P, Sinisi A, Pasquali D, Errico ME, Nappi C, 2004. Aurora B expression in normal testis and seminomas. J Endocrinol 181:263–270. Tremolaterra F, Somma P, Mansueto G, De Rosa G. 2003. Loss of oestrogen receptor b, Chieffi P, Franco R, Portella G. 2009. Molecular and cell biology of testicular tumors. Int Rev high PCNA and expression and aneuploidy as markers of worse prognosis in ovarian Cell Mol Biol 278:277–308. granulosa cell tumours. Histopathology 43:254–262. Di Agostino S, Fedele M, Chieffi P, Fusco A, Rossi P, Geremia R, Sette C. 2004. Thanos D, Maniatis T. 1992. The high mobility group protein HMGI(Y) is required for NF- kB Phosphorylation of high-mobility group protein A2 by Nek2 kinase during the first meiotic dependent virus induction of human- b gene. Cell 71:777–789. division in mouse spermatocytes. Mol Biol Cell 15:1224–1232. Vicini E, Loiarro M, Di Agostino S, Corallini S, Capolunghi F, Carsetti R, Chieffi P, Geremia R, Esposito F, Libertini S, Franco R, Abagnale A, Marra L, Portella G, Chieffi P. 2009. Aurora B Stefanini M, Sette C. 2006. 17- b-estradiol elicits genomic and non-genomic responses in expression in post-puberal testicular germ cell tumours. J Cell Physiol 221:435–439. mouse male germ cells. J Cell Physiol 206:238–245. Esposito F, Boscia F, Franco R, Tornincasa M, Fusco A, Kitazawa S, Looijenga LH, Chieffi P. Wellejus A, Loft S. 2002. Receptor-mediated ethinylestradiolinduced oxidative DNA damage 2011. Down-regulation of estrogen receptor- b associates with transcriptional coregulator in rat testicular cells. FASEB J 16:195–201. PATZ1 delocalization in human testicular seminomas. J Pathol 224:110–120. Weyant MJ, Carothers AM, Mahmoud NN, Bradlow HL, Remotti H, Bilinski RT, Bertagnolli Fedele M, Franco R, Salvatore G, Paronetto MP, Barbagallo F, Pero R, Chiariotti L, Sette C, MM. 2001. Reciprocal expression of ER a and ER b is associated with estrogen-mediated Tramontano D, Chieffi G, Fusco A, Chieffi P. 2008. PATZ1 gene has a critical role in the modulation of intestinal tumorigenesis. Cancer Res 61:2547–2551. spermatogenesis and testicular tumours. J Pathol 215:39–47. WoodLJ,MaherJF,BuntonTE,ResarLM.2000.Theoncogenicproperties of the HMG-I gene Franco R, Esposito F, Fedele M, Liguori G, Pierantoni G, Botti G, Tramontano D, Fusco A, family. Cancer Res 60:4256–4261. Chieffi P. 2008. Detection of high mobility group proteins A1 and A2 represents a valid Zhou X, Benson KF, Ashar HR, Chada K. 1995. Mutation responsible for the mouse pygmy diagnostic marker in post-puberal testicular germ cell tumours. J Pathol 214:58–64. phenotype in developmentally regulated factor HMGI-C. Nature 376:771–774.

JOURNAL OF CELLULAR PHYSIOLOGY