help button home button Endocrine Society JCEM JCEM Call for Nominations for EIC
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Journal of Clinical Endocrinology & Metabolism, doi:10.1210/jc.2005-0921
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
90/12/6529    most recent
Author Manuscript (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Anttonen, M.
Right arrow Articles by Heikinheimo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anttonen, M.
Right arrow Articles by Heikinheimo, M.
Related Collections
Right arrow Endocrine Oncology
Right arrow Female Endocrinology
The Journal of Clinical Endocrinology & Metabolism Vol. 90, No. 12 6529-6535
Copyright © 2005 by The Endocrine Society

High GATA-4 Expression Associates with Aggressive Behavior, whereas Low Anti-Müllerian Hormone Expression Associates with Growth Potential of Ovarian Granulosa Cell Tumors

Mikko Anttonen, Leila Unkila-Kallio, Arto Leminen, Ralf Butzow and Markku Heikinheimo

Children’s Hospital and Program for Developmental and Reproductive Biology (M.A., M.H.), Biomedicum Helsinki, and Departments of Obstetrics and Gynecology (L.U.-K., A.L., R.B.) and Pathology (R.B.), University of Helsinki, 00014 Helsinki, Finland; and Department of Pediatrics (M.H.), Washington University School of Medicine, St. Louis, Missouri 63110

Address all correspondence and requests for reprints to: Mikko Anttonen, M.B., Ph.D., Program for Developmental and Reproductive Biology, Biomedicum Helsinki, University of Helsinki, P.O. Box 63 (Haartmaninkatu 8), 00014 Helsinki, Finland. E-mail: mikko.anttonen{at}helsinki.fi.


   Abstract
Top
 Abstract
Introduction
 Patients and Methods
 Results
 Discussion
 References
 
Context: Granulosa cell tumors (GCTs) are ovarian malignancies that produce estrogens, inhibins, and anti-Müllerian hormone (AMH). The molecular pathogenesis of GCTs is likely to involve defects in the genes regulating normal granulosa cell proliferation during folliculogenesis.

Objective: The objective of this study was to test the role of factors regulating the normal granulosa cell function, i.e. AMH, inhibin-{alpha}, SF-1 (steroidogenic factor-1), and GATA transcription factors in the pathobiology and clinical behavior of GCTs.

Design: We selected randomly a cohort of 80 GCT patients treated at our university hospital during 1971–2003, analyzed protein expression in the tumor samples embedded on a tissue microarray by immunohistochemistry, and correlated the data to clinical and histopathological parameters.

Results: We found no significant differences in the immunoreactivity levels of inhibin-{alpha}, GATA-6, FOG-2 (friend of GATA-2), or SF-1 in GCTs compared with normal granulosa cells. AMH expression was, however, low (i.e. reduced) in 69% of GCTs and correlated inversely with tumor size (P = 0.0025). In contrast, GATA-4 expression was high (i.e. resembled normal granulosa cells) in 44% of GCTs and correlated positively with clinical stage and recurrence (P = 0.0232 and P = 0.0038, respectively). Fifty of the 80 patients had a follow-up for at least 10 yr, and 13 of them had recurrence(s). In multivariate analysis of recurrence, the high GATA-4 expression remained the only independent factor (risk ratio, 9.2; 95% confidence interval, 2.0–43.3; P = 0.0048).

Conclusions: The more aggressive GCTs retain a high GATA-4 expression, whereas the larger tumors lose the proliferation-suppressing AMH expression. The high GATA-4 expression in GCTs may serve as a marker of poor prognosis.


   Introduction
Top
 Abstract
 Introduction
Patients and Methods
 Results
 Discussion
 References
 
OVARIAN GRANULOSA CELL tumors (GCTs) are endocrine malignancies (1, 2) that may produce excessive amounts of estrogens and inhibins, typically causing precocious puberty, disturbances in menstrual cycle, infertility, and endometrial hyperplasia or cancer (1, 2, 3). GCTs account for 2–3% of all ovarian cancers and can occur at any age. The juvenile type is mainly detected in children and adolescents, whereas the adult type has a peak incidence at 50–55 yr of age (2, 4). At presentation, GCTs vary from microscopic to large abdominal masses and are often highly vascularized and hemorrhagic, giving rise to acute abdominal pain.

In contrast to ovarian epithelial cancers, GCTs are considered to be of low malignant potential. The reported 10-yr survival varies from 60 to 95% and tends to be lower in a longer follow-up (1). Moreover, GCTs have characteristically a high risk of recurrence, which can occur as late as 30 yr after the primary diagnosis (5). Predicting the clinical behavior of GCTs is, however, difficult. For instance, histological pattern, nuclear atypia, mitotic activity, and tumor size have been shown to have prognostic importance in some but not all studies. Staging remains the most valuable tool to guide treatment and follow-up of GCT patients (1, 2, 6).

Gene expression profile of human GCTs has similarities with that of granulosa cells of preantral and small/medium antral follicles (7, 8). Although the definitive molecular pathogenesis of GCTs remains unknown, defects in endocrine, paracrine, and autocrine signaling cascades regulating granulosa cell proliferation have been implicated. FSH is essential for granulosa cell proliferation and maturation, and it upregulates a variety of granulosa cell genes, such as cyclin D2 and FSH receptor (FSHR). The importance of cyclin D2 and FSHR for granulosa cells is obvious, given that inactivation of either gene impairs granulosa cell proliferation and folliculogenesis (9, 10). Both of these genes are overexpressed in human GCTs (7, 9), and the FSH signaling cascade is considered to be overactivated along with GCT pathogenesis (7, 8, 9). Activating mutations in FSHR or in the downstream components of the cascade in GCTs are, however, not common or hitherto not found (11, 12). Interestingly, ERK, which is also used in the FSH cascade, was recently found to be active in a human GCT cell line and in aggressive GCTs, suggesting its involvement in malignant granulosa cell proliferation (13).

Mice null for inhibin-{alpha} have gonadal tumors, suggesting that this factor acts as a tumor suppressor gene in this context (14). Inactivating mutations in inhibin-{alpha} remain, however, unreported in human GCTs. Indeed, most GCTs express this marker (8), and it is used as a diagnostic tool in surgical pathology. Inhibin-{alpha} has also been reported to have prognostic significance; in one study, the inhibin-{alpha}-negative GCTs were of advanced stage and had dismal prognosis (15). In addition to inhibin-{alpha}, anti-Müllerian hormone (AMH) [also known as Müllerian inhibiting substance or MIS] has been indicated as a GCT marker (16). AMH has an essential role in the sexual differentiation of the mammalian reproductive tract as well as in the ovarian function (17, 18, 19). Normal granulosa cells and GCTs secrete AMH to the circulation, and it is evolving as a powerful serum marker for follow-up of GCTs (17, 20). The role of AMH in the GCT pathobiology remains, however, poorly understood.

Transcription factors GATA-4 and GATA-6 regulate expression of a number of essential gonadal genes (21). They belong to a six-member family of zinc-finger transcription factors recognizing a consensus GATA motif (A/T-GATA-A/G) in the target gene promoters and enhancers. They act in cooperation with various other transcription factors in the gonads, and synergistic action of GATA-4 with, for example, SF-1 (steroidogenic factor-1) is critical for AMH expression in the gonads (22). GATA-4 acts also as an effector in the FSH signaling cascade (23, 24). We have previously characterized expression of GATA-4, GATA-6, and the GATA cofactor FOG-2 (friend of GATA-2) in fetal and adult granulosa cells in the murine and human ovary (23, 25, 26, 27). Interestingly, GATA-4 expression is activated in primary follicles along with activation of granulosa cell proliferation (25, 27).

Given that at least a subset of GCTs express GATA-4, GATA-6, and FOG-2 (25), we hypothesized that they play a role in GCT pathobiology. We therefore analyzed in detail the expression profiles of GATA-4, GATA-6, FOG-2, and SF-1, as well as the GATA-target genes and granulosa cell markers AMH and inhibin-{alpha} in a GCT tissue microarray comprising 80 primary tumors. Subsequently, we correlated the expression data with detailed clinical and histopathological parameters to gain insight into the impact of these factors in granulosa cell tumorigenesis and into their value in evaluating prognosis of the patients.


   Patients and Methods
Top
 Abstract
 Introduction
 Patients and Methods
Results
 Discussion
 References
 
Patients

The study was approved by the hospital ethics committee and the National Authority for Medicolegal Affairs in Finland. Female patients diagnosed as having GCT at Helsinki University Central Hospital between 1971 and 2003 were identified in the hospital archives. Available paraffin-embedded tissue samples were critically reevaluated by a gynecologic pathologist (R.B.) with the help of immunohistochemical markers to fulfill diagnostic criteria (28). Eighty GCT patients were included in the study. Three normal ovaries removed because of cervical cancer from women less than 35 yr old were used as normal reference tissue. The clinicopathological data of the tumor patients was retrospectively scrutinized from patient files (3). Mean age of the patients was 52 yr (range from 19 to 87 yr), and 54% were postmenopausal. None of the patients received neoadjuvant treatment; after the primary surgery, 12 patients (stage Ia in five, stage Ic or higher in seven) received adjuvant chemotherapy, two patients (stage Ic) received radiotherapy, and two patients (stage II) received both chemotherapy and radiotherapy. Treatment and follow-up of GCT patients in the district is centralized to our tertiary university hospital, thus facilitating the detection of late recurrences; survival was additionally confirmed with the Central Statistical Office of Finland.

Tumor tissue evaluation and tissue microarray construction

Subtyping and evaluation of the degree of nuclear atypia of GCTs was done according to established criteria (28). Tumors having microfollicular or macrofollicular, tubular, trabecular, and insular characteristics were grouped together and considered as differentiated subtype, separate from diffuse (sarcomatoid) subtype. Immunohistochemical cytoceratin (AE1/AE3) staining showed dot-like or perinuclear positivity in GCTs. To establish the mitotic index (MI), four areas with highest mitotic activity were sought from each tumor, and the number of mitotic figures in 10 high-power fields (HPFs) was counted in each. MI represents an average of these counts: less than 5 figures/10 HPF is low; more than 5 figures/10 HPF is high. Tissue microarrays were constructed as described previously (29). To achieve concordant immunophenotyping in relation to full sections, four core tissue biopsies were obtained from representative areas of the histological specimen (30). The resulting quadruple samples from 80 primary GCTs were arrayed in one recipient paraffin block.

Immunohistochemistry

Sections of three paraffin-embedded normal ovaries and of GCT tissue microarray were subjected to immunoperoxidase stain as described previously (27). Antigen retrieval was performed with 10 mmol/liter citric acid in a microwave oven for 10–20 min, and endogenous peroxidase was blocked with 3% hydrogen peroxide for 5 min at room temperature. Primary antibodies were incubated for 1 h at +37 C: goat antimouse GATA-4 IgG at 1:400 dilution (sc-1237; Santa Cruz Biotechnology, Santa Cruz, CA), rabbit antihuman GATA-6 IgG at 1:50 (sc-9055; Santa Cruz Biotechnology), rabbit antimouse FOG-2 IgG at 1:100 (sc-10744; Santa Cruz Biotechnology), rabbit antimouse SF-1 IgG at 1:800 (06-431; Upstate Biotechnology, Lake Placid, NY), and goat antihuman AMH IgG at 1:100 (sc-6886; Santa Cruz Biotechnology). The antimouse IgGs cross-react with the corresponding human antigens. An avidin-biotin immunoperoxidase system (Vectastain Elite ABC kit; Vector Laboratories, Burlingame, CA) and DAB (Sigma, St. Louis, MO) was used to visualize the bound antibody, and sections were counterstained with Harris hematoxylin. Immunostaining of Ki67 (A0047; Dako, Glostrup, Denmark) and inhibin-{alpha} (MCA951S; Serotec, Oxford, UK) was performed in a Dako TechMate 500 automated staining machine. In control experiments, nonimmune serum or PBS replaced the primary antibody.

Scoring of the results

The immunohistochemically stained tissue sections were scored separately by two researchers (M.A. and R.B.) blinded to the clinical parameters; in case of disparate scoring, a consensus was achieved by using a multiheaded microscope. In cases of heterogeneous stainings, detected in 13 tumors for AMH, 12 tumors for FOG-2, and less than five tumors for other antigens, the strongest and weakest of the four arrayed tissue cores of individual tumors were omitted from the analysis. Granulosa cells from normal ovaries were scored to establish the normal immunoreactivity level in growing follicles of primary/preantral stage to small/medium antral stage. Scoring of nuclear antigens represents proportion of positive nuclei: for Ki67 categories, 1–5, 5–25, and more than 25%; for GATA-4, GATA-6, FOG-2, and SF-1, negative/weak for 0–20% positive nuclei, intermediate for 20–80% positive nuclei, and high for 80–100% positive nuclei. Scoring of cytoplasmic antigens AMH and inhibin-{alpha} represents negative for 0% positive cells, weak for 1–20% positive cells, or 1–80% positive cells with weak immunoreactivity, intermediate for 20–100% positive cells with at least moderate immunoreactivity, and high for 80–100% positive cells.

Statistical analysis

For correlations, the three or four categorical tumor scorings were combined into two: 1) expression reflecting the immunoreactivity level in normal granulosa cells, and 2) expression that is reduced from the normal immunoreactivity level (see Results). Contingency tabling (2 x 2) was performed for the resulting nominal variables, followed by {chi}2 or Fisher’s exact tests if {chi}2 was not appropriate, using Statview 5.0.1 software for Macintosh. In addition, a logistic regression model was used for multivariate analysis, each of the clinical measures being the dependent variable; P < 0.05 was considered significant.


   Results
Top
 Abstract
 Introduction
 Patients and Methods
 Results
Discussion
 References
 
Expression profiles in normal proliferating granulosa cells

We first analyzed the granulosa cell expression of the studied antigens in normal ovarian tissue samples (Fig. 1Go). GATA-4 expression was overall high in granulosa cells of normal human ovaries; the expression first appeared in the primary stage follicles when granulosa cells start to proliferate and remained high through the antral stage (Fig. 1Go, A and B). Immunoreactivity for GATA-6 and FOG-2 was lower than that for GATA-4 in preantral and antral stage follicles (Fig. 1Go, C–F). Similarly to GATA-4, SF-1 expression was first observed in the primary follicles and was intermediate or high through the antral stage (Fig. 1Go, G and H). AMH and inhibin-{alpha} immunoreactivities were intermediate or high in the normal granulosa cells and somewhat diminished in the large antral follicles compared with earlier stages of folliculogenesis (Fig. 1Go, I–L). Based on these findings, the cutoff level for GATA-4 expression, when analyzing GCTs, was set between high and intermediate expression (i.e. high vs. reduced) and for the rest of the antigens between intermediate and negative/weak expression (i.e. high/intermediate vs. reduced).



View larger version (77K):
[in this window]
[in a new window]
  		FIG. 1. Immunohistochemical staining for GATA-4 (A, B), GATA-6 (C, D), FOG-2 (E, F), SF-1 (G, H), AMH (I, J), and inhibin-{alpha} (K, L) in normal granulosa cells. Panels at left present expression in primary follicles, and panels at right expression in a small antral follicle (diameter, 1 mm). Brown indicates positive staining in the cytoplasm for AMH and inhibin-{alpha} and in the nuclei for the other antigens; sections were counterstained with hematoxylin. Scale bar, 50 µm.

 
Clinicopathological features and expression profiles of GCTs

The clinical and histopathological parameters of the GCT cases are presented in Table 1Go. Nuclear atypia and MI were positively correlated (P < 0.0001), showing high MI in 67% of the tumors with moderate or higher atypia. Given the characteristically low Ki67 proliferative index in GCTs, an elevated level was set at more than 5% positive nuclei; this was detected in 17% of the tumors. Ki67 index correlated positively with MI (P < 0.0001).


View this table:
[in this window]
[in a new window]
  		TABLE 1. Characteristics of ovarian GCTs in 80 women

 
Immunoprofiling of GCTs revealed that 44% had retained high GATA-4 expression (Fig. 2AGo), whereas 56% exhibited reduced expression (Fig. 2BGo); only 6% exhibited negative GATA-4 expression. The majority of the studied tumors exhibited high or intermediate (i.e. normal granulosa cell resembling) immunoreactivity for GATA-6, FOG-2, and SF-1 (Fig. 2Go, C–H). FOG-2 and GATA-6 expressions were associated with each other (P = 0.0261); FOG-2 expression was high/intermediate in 81% of the tumors with high/intermediate GATA-6 expression.



View larger version (113K):
[in this window]
[in a new window]
  		FIG. 2. Immunohistochemical staining for GATA-4 (A, B), GATA-6 (C, D), FOG-2 (E, F), SF-1 (G, H), AMH (I, J), and inhibin-{alpha} (K-L) in representative GCTs. Panels at left present expression pattern that resembles normal granulosa cells (for comparison, see Fig. 1Go), i.e. high expression of GATA-4 and high or intermediate expression of the rest of the antigens; panels at right present reduced or low expression pattern. The number of tumors in the given category, as well as percentage, of the 80 tumors analyzed is indicated in the top right corners. Brown indicates positive staining in the cytoplasm for AMH and inhibin-{alpha} and in the nuclei for the other antigens; sections were counterstained with hematoxylin. Scale bar, 50 µm.

 
In contrast to the other antigens studied, AMH immunoreactivity was mostly reduced in GCTs (Fig. 2Go, I–J); it was weak in 40% and negative in 29%. GATA-6, rather than GATA-4, was associated with AMH (P = 0.0256); AMH expression was reduced in 87% of the tumors with reduced GATA-6 expression. Immunoreactivity for inhibin-{alpha} was intermediate or high in most GCTs (Fig. 2Go, K and L), as expected; only three tumors were totally inhibin-{alpha} negative. It is noteworthy that, during the initial evaluation of the archival GCT samples, four inhibin-{alpha}-negative tumors were excluded from the study and reclassified as undifferentiated or endometrioid carcinomas. High inhibin-{alpha} expression, detected in 14% of GCTs, was associated with high GATA-4 expression (P = 0.0215).

Reduced AMH and GATA-6 expressions correlate with larger tumor size

Next, we correlated the expression data of the GCTs to clinical stage, tumor size, ascites, tumor subtype, MI, Ki67, and nuclear atypia; only the significant correlations are presented herein (Fig. 3Go). First, GATA-6 and AMH expression correlated inversely with tumor size. GATA-6 and AMH were reduced in 44 and 87%, respectively, of tumors more than 10 cm in size (P = 0.0181 and P = 0.0025, respectively) (Fig. 3AGo). When the 23 AMH-negative tumors were compared with AMH-expressing tumors (having at least weak expression), the correlation with tumor size was highly significant (P < 0.0001); only 11% of the tumors less than or equal to 10 cm in size were AMH negative (Fig. 3AGo). Multivariate analysis showed that GATA-6 and AMH were independently associated with tumor size (data not shown). Presence of ascites at primary surgery correlated positively with tumor size more than 10 cm (P = 0.0019). The studied antigens failed to correlate with ascites, whereas elevated MI as well as elevated Ki67 were associated with the presence of ascites (P = 0.0003 and P = 0.0049, respectively).



View larger version (25K):
[in this window]
[in a new window]
  		FIG. 3. Associations of gene expression patterns in primary GCTs with the clinical parameters in 80 patients. A, Association with tumor size less than or equal to 10 cm vs. more than 10 cm. P = 0.0181 for high/intermediate vs. weak/negative GATA-6 expression; P < 0.0001 for high/intermediate/weak vs. negative AMH expression; and P = 0.0025 for high/intermediate vs. weak/negative (i.e. reduced) AMH expression. Data on tumor size were not available for one case. B, Association with clinical stage Ia vs. stage Ic, II, or III; P = 0.0232 for high vs. reduced GATA-4 expression, and P = 0.0164 for high vs. low MI. C, Association with recurrence; P = 0.0038 for high vs. reduced GATA-4 expression in all 80 patients (left panel), and P = 0.0016 for high vs. reduced GATA-4 expression in the 50 patients having at least 10-yr follow-up. Contingency tabling coupled with {chi}2 test was used to test the statistical significances; P < 0.05 considered significant.

 
High GATA-4 expression and elevated MI correlate with stage Ic or higher

High GATA-4 expression in GCTs was associated with clinical stage (P = 0.0232). In this analysis, stage was divided into two categories, i.e. stage Ia vs. stage Ic, II, or III; 60% of the more advanced stage tumors exhibited high GATA-4 expression (Fig. 3BGo). Of the stage Ia tumors, 66% exhibited reduced GATA-4 expression; 84% of them also exhibited low MI (P = 0.0164) (Fig. 3BGo). GATA-4 expression and MI did not, however, correlate with each other. Moreover, both of these factors remained independently associated with stage when multivariate analysis was applied (data not shown).

High GATA-4 expression in the primary tumor correlates with recurrence

Finally, we tested whether clinicopathological parameters and immunoreactivity levels for the studied antigens associate with recurrence. Of the 80 patients, 14 (17.5%) had a recurrence, and nine of these 14 patients had multiple recurrences (Table 2Go). Eight of the recurred tumors were local (stage Ia) (Table 2Go) at primary diagnosis. In addition to stage and MI, also the granulosa cell markers inhibin-{alpha} and AMH failed to correlate with recurrence risk. Instead, high-retained GATA-4 expression in the primary tumor correlated with the overall recurrence risk (P = 0.0038) (Fig. 3CGo, left). Furthermore, among the 50 patients being followed up for at least 10 yr, 13 (26%) had one or multiple recurrences. High GATA-4 expression correlated with recurrence risk also in these patients (P = 0.0016); 10 patients with recurrence (77%) exhibited high GATA-4 expression in the primary tumor, whereas 27 patients free of recurrence (73%) exhibited reduced (i.e. intermediate or negative/weak) GATA-4 expression (Fig. 3CGo, right). Last, we assessed the recurrence risk in multivariate analysis with GATA-4, clinical stage, tumor size, and MI. Only the high-retained GATA-4 expression remained as an independent factor for recurrence risk in the patients with a follow-up for at least 10 yr (risk ratio, 9.24; 95% confidence interval, 1.97–43.30; P = 0.0048).


View this table:
[in this window]
[in a new window]
  		TABLE 2. Key features of the primary GCTs and survival of the 14 patients with recurrence(s)

 

   Discussion
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Discussion
References
 
In this study, we characterized the expression of granulosa cell markers AMH and inhibin-{alpha}, as well as gonadal transcription factors GATA-4, GATA-6, FOG-2, and SF-1 in a series of 80 primary GCTs. Moreover, we correlated key clinicopathological measures (clinical stage, tumor size, MI, nuclear atypia, and risk of tumor recurrence) with the protein expression data from the tissue microarray. We find that GATA-4 expression was retained high in 44%, and it was intermediate or high in 94% of GCTs, indicating that GATA-4 expression is characteristic of GCTs. This differs from ovarian epithelial carcinomas, because we have shown previously that the majority of serous carcinomas and of the advanced stage mucinous carcinomas have lost GATA-4 expression (31). In GCTs, the high-retained GATA-4 expression was associated with stage Ic or higher as well as with recurrence risk. Thus, high GATA-4 expression marks a subgroup of GCTs with a more aggressive behavior.

The clinical outcome of GCTs is more favorable compared with that of ovarian carcinomas, but their unpredictable behavior requires careful evaluation at diagnosis and follow-up. Clinical stage has been universally accepted as one of the important prognostic tools, whereas the value of MI, nuclear atypia, and tumor size remains somewhat controversial (1, 6). Using the clinical and pathological parameters, it is difficult to predict particularly the late recurrences of stage I tumors accounting for up to 90% of primary GCTs (1, 6 and this study). In the present patient series, every fourth patient had a recurrence during a follow-up for at least 10 yr; the latest was detected 27.5 yr from the diagnosis, and 61% of these patients subsequently died of GCT. The prognostic value of stage remained insignificant, because most of the recurred tumors were of stage I (even stage Ia) at primary diagnosis. Of the molecular markers, we were unable to confirm the previously indicated association of inhibin-{alpha} immunoreactivity with stage or prognosis (15). Instead, we found that GATA-4 has potential as a prognostic tool. The usefulness of GATA-4 as a molecular marker, however, requires additional analyses, such as confirming the cutoff level for immunoreactivity with a computer-based scoring in different patient series.

AMH has been highlighted as a growth-inhibiting factor of ovarian, breast, and prostate cancer cells (17). AMH also inhibits the initial granulosa cell proliferation and the FSH sensitivity of the follicles (19). A previous study concluded that only a small proportion of GCT cells have conserved their AMH expression (16). We now find that the low AMH expression was associated with an increase in the size of GCTs. Considering that human GCTs express functional AMH receptors (32), this kind of reduction in the proliferation-suppressing AMH expression may promote a granulosa cell population to escape from the proliferation control. With respect to utilization of AMH as a serum marker in GCT patient follow-up (17, 20), an upcoming recurrent tumor presumably secretes AMH at levels detectable in circulation, despite the locally reduced AMH expression in the primary tumor (16 and this study). Because of the retrospective nature of our study, dating back to the 1970s, we were unfortunately unable to assess the putative correlation of the tissue expression and serum levels of AMH at diagnosis.

Findings by us and others suggest that GATA-4 plays a role in the granulosa cell proliferation. Most interestingly, GATA-4 expression is upregulated in mice and humans along with activation of granulosa cell proliferation in the primary follicle stage (23, 25, 26) and also along with adrenal tumorigenesis (33, 34). In other cellular contexts, GATA-4 has been shown to regulate expression of cyclin D2 (35), which is essential for granulosa cell proliferation, as well as of Bcl-X, an antiapoptotic factor (36). In addition, GATA-4 has been linked to signaling pathways putatively involved in the pathogenesis of GCTs, e.g. FSH and MAPK signaling (7, 9, 13). In brief, FSH upregulates GATA-4 expression (23), and protein kinase A and ERK again activate GATA-4 by phosphorylation (24, 35). Although all of this data intriguingly support the hypothesis, additional studies are required to link GATA-4 to granulosa cell proliferation.

GATA-6 has been implicated recently in the regulation of steroidogenic enzyme expression in endocrine organs (21, 37). Accordingly, GATA-6, but not GATA-4, is strongly expressed in the postnatal adrenal, the organ with the most active steroidogenesis (37, 38). Although not directly assessed here, GATA-6 may be involved in the steroidogenic function of GCTs. In contrast to a previous report demonstrating negligible SF-1 in GCTs (39), we found that GCTs express SF-1 in amounts comparable with normal granulosa cells. This finding is in accordance with the steroidogenic activity of GCTs, given that SF-1, acting in concert with GATA transcription factors, is essential for the expression of several steroidogenic enzymes.

In conclusion, our results suggest previously unknown roles for GATA-4 and AMH in normal and malignant human granulosa cells. The more aggressive GCTs retain high GATA-4 expression, characteristic of the highly proliferating normal granulosa cells, whereas large GCTs lose their AMH expression. These expression patterns also reflect the follicular stage and the type of proliferating granulosa cells from which the individual tumors arise; whether the larger GCTs lose their AMH expression at some point or exhibit only low levels of AMH at the very beginning remains unknown. Last, GATA-4 immunostaining can serve as a prognostic tool given that a subgroup of GCTs that later recur exhibit high and uniform GATA-4 expression. Our detailed clinical data with a long-term and ongoing follow-up of the 80 patients coupled to data obtained by the tumor tissue microarray highlights novel aspects in GCT biology and helps for better understanding of the peculiar clinical behavior of this ovarian cancer.


   Acknowledgments
 
We thank Ms. Gynel Arifdshan and Ms. Taru Jokinen for excellent technical assistance and Drs. Juha Tapanainen and David B. Wilson for critical reading of this manuscript.


   Footnotes
 
This work was supported by the Emil Aaltonen Foundation (to M.A.), the Maud Kuistila Memorial Foundation (to M.A.), the Jalmari and Rauha Ahokas Foundation (to M.A. and M.H.), the Finnish Cancer Organizations (to R.B.), the Helsinki University Central Hospital Research Funds (to L.U.-K. and R.B.), the Finnish Pediatric Research Foundation (to M.H.), and the Sigrid Juselius Foundation (to M.H.).

First Published Online September 13, 2005

Abbreviations: AMH, Anti-Müllerian hormone; FSHR, FSH receptor; GCT, granulosa cell tumors; HPF, high-power field; MI, mitotic index; SF-1, steroidogenic factor-1.

Received April 27, 2005.

Accepted September 2, 2005.


   References
Top
 Abstract
 Introduction
 Patients and Methods
 Results
 Discussion
 References
 

  1. Schumer ST, Cannistra SA 2003 Granulosa cell tumor of the ovary. J Clin Oncol 21:1180–1189[Abstract/Free Full Text]
  2. Stuart GC, Dawson LM 2003 Update on granulosa cell tumours of the ovary. Curr Opin Obstet Gynecol 15:33–37[Medline]
  3. Unkila-Kallio L, Tiitinen A, Wahlstrom T, Lehtovirta P, Leminen A 2000 Reproductive features in women developing ovarian granulosa cell tumour at a fertile age. Hum Reprod 15:589–593[Abstract/Free Full Text]
  4. Merras-Salmio L, Vettenranta K, Mottonen M, Heikinheimo M 2002 Ovarian granulosa cell tumors in childhood. Pediatr Hematol Oncol 19:145–156[CrossRef][Medline]
  5. Hines JF, Khalifa MA, Moore JL, Fine KP, Lage JM, Barnes WA 1996 Recurrent granulosa cell tumor of the ovary 37 years after initial diagnosis: a case report and review of the literature. Gynecol Oncol 60:484–488[CrossRef][Medline]
  6. Miller BE, Barron BA, Wan JY, Delmore JE, Silva EG, Gershenson DM 1997 Prognostic factors in adult granulosa cell tumor of the ovary. Cancer 79:1951–1955[CrossRef][Medline]
  7. Chu S, Rushdi S, Zumpe ET, Mamers P, Healy DL, Jobling T, Burger HG, Fuller PJ 2002 FSH-regulated gene expression profiles in ovarian tumours and normal ovaries. Mol Hum Reprod 8:426–433[Abstract/Free Full Text]
  8. Fuller PJ, Chu S 2004 Signalling pathways in the molecular pathogenesis of ovarian granulosa cell tumours. Trends Endocrinol Metab 15:122–128[Medline]
  9. Sicinski P, Donaher JL, Geng Y, Parker SB, Gardner H, Park MY, Robker RL, Richards JS, McGinnis LK, Biggers JD, Eppig JJ, Bronson RT, Elledge SJ, Weinberg RA 1996 Cyclin D2 is an FSH-responsive gene involved in gonadal cell proliferation and oncogenesis. Nature 384:470–474[CrossRef][Medline]
  10. Aittomaki K, Lucena JL, Pakarinen P, Sistonen P, Tapanainen J, Gromoll J, Kaskikari R, Sankila EM, Lehvaslaiho H, Engel AR, Nieschlag E, Huhtaniemi I, de la Chapelle A 1995 Mutation in the follicle-stimulating hormone receptor gene causes hereditary hypergonadotropic ovarian failure. Cell 82:959–968[CrossRef][Medline]
  11. Ligtenberg MJ, Siers M, Themmen AP, Hanselaar TG, Willemsen W, Brunner HG 1999 Analysis of mutations in genes of the follicle-stimulating hormone receptor signaling pathway in ovarian granulosa cell tumors. J Clin Endocrinol Metab 84:2233–2234[Abstract/Free Full Text]
  12. Fuller PJ, Verity K, Shen Y, Mamers P, Jobling T, Burger HG 1998 No evidence of a role for mutations or polymorphisms of the follicle-stimulating hormone receptor in ovarian granulosa cell tumors. J Clin Endocrinol Metab 83:274–279[Abstract/Free Full Text]
  13. Steinmetz R, Wagoner HA, Zeng P, Hammond JR, Hannon TS, Meyers JL, Pescovitz OH 2004 Mechanisms regulating the constitutive activation of the extracellular signal-regulated kinase (ERK) signaling pathway in ovarian cancer and the effect of RNAi for ERK1/2 on cancer cell proliferation. Mol Endocrinol 18:2570–2582[Abstract/Free Full Text]
  14. Matzuk MM, Finegold MJ, Su JG, Hsueh AJ, Bradley A 1992 {alpha}-Inhibin is a tumour-suppressor gene with gonadal specificity in mice. Nature 360:313–319[CrossRef][Medline]
  15. Ala-Fossi SL, Aine R, Punnonen R, Maenpaa J 2000 Is potential to produce inhibins related to prognosis in ovarian granulosa cell tumors? Eur J Gynaecol Oncol 21:187–189[Medline]
  16. Rey R, Sabourin JC, Venara M, Long WQ, Jaubert F, Zeller WP, Duvillard P, Chemes H, Bidart JM 2000 Anti-Mullerian hormone is a specific marker of sertoli- and granulosa-cell origin in gonadal tumors. Hum Pathol 31:1202–1208[CrossRef][Medline]
  17. Teixeira J, Maheswaran S, Donahoe PK 2001 Mullerian inhibiting substance: an instructive developmental hormone with diagnostic and possible therapeutic applications. Endocr Rev 22:657–674[Abstract/Free Full Text]
  18. Weenen C, Laven JS, Von Bergh AR, Cranfield M, Groome NP, Visser JA, Kramer P, Fauser BC, Themmen AP 2004 Anti-Mullerian hormone expression pattern in the human ovary: potential implications for initial and cyclic follicle recruitment. Mol Hum Reprod 10:77–83[Abstract/Free Full Text]
  19. Visser JA, Themmen AP 2005 Anti-Mullerian hormone and folliculogenesis. Mol Cell Endocrinol 234:81–86[CrossRef][Medline]
  20. Rey RA, Lhomme C, Marcillac I, Lahlou N, Duvillard P, Josso N, Bidart JM 1996 Antimullerian hormone as a serum marker of granulosa cell tumors of the ovary: comparative study with serum {alpha}-inhibin and estradiol. Am J Obstet Gynecol 174:958–965[CrossRef][Medline]
  21. Tremblay JJ, Viger RS 2003 Novel roles for GATA transcription factors in the regulation of steroidogenesis. J Steroid Biochem Mol Biol 85:291–298[CrossRef][Medline]
  22. Tremblay JJ, Viger RS 1999 Transcription factor GATA-4 enhances Mullerian inhibiting substance gene transcription through a direct interaction with the nuclear receptor SF-1. Mol Endocrinol 13:1388–1401[Abstract/Free Full Text]
  23. Heikinheimo M, Ermolaeva M, Bielinska M, Rahman NA, Narita N, Huhtaniemi IT, Tapanainen JS, Wilson DB 1997 Expression and hormonal regulation of transcription factors GATA-4 and GATA-6 in the mouse ovary. Endocrinology 138:3505–3514[Abstract/Free Full Text]
  24. Tremblay JJ, Viger RS 2003 Transcription factor GATA-4 is activated by phosphorylation of serine 261 via the cAMP/protein kinase a signaling pathway in gonadal cells. J Biol Chem 278:22128–22135[Abstract/Free Full Text]
  25. Laitinen MP, Anttonen M, Ketola I, Wilson DB, Ritvos O, Butzow R, Heikinheimo M 2000 Transcription factors GATA-4 and GATA-6 and a GATA family cofactor, FOG-2, are expressed in human ovary and sex cord-derived ovarian tumors. J Clin Endocrinol Metab 85:3476–3483[Abstract/Free Full Text]
  26. Vaskivuo TE, Anttonen M, Herva R, Billig H, Dorland M, te Velde ER, Stenback F, Heikinheimo M, Tapanainen JS 2001 Survival of human ovarian follicles from fetal to adult life: apoptosis, apoptosis-related proteins, and transcription factor GATA-4. J Clin Endocrinol Metab 86:3421–3429[Abstract/Free Full Text]
  27. Anttonen M, Ketola I, Parviainen H, Pusa AK, Heikinheimo M 2003 FOG-2 and GATA-4 are coexpressed in the mouse ovary and can modulate mullerian-inhibiting substance expression. Biol Reprod 68:1333–1340[Abstract/Free Full Text]
  28. 2003 WHO classification of tumors: pathology and genetics of tumors of the breast and female genital organs. Lyon, France: IARC Press
  29. Kononen J, Bubendorf L, Kallioniemi A, Barlund M, Schraml P, Leighton S, Torhorst J, Mihatsch MJ, Sauter G, Kallioniemi OP 1998 Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med 4:844–847[CrossRef][Medline]
  30. Hoos A, Urist MJ, Stojadinovic A, Mastorides S, Dudas ME, Leung DH, Kuo D, Brennan MF, Lewis JJ, Cordon-Cardo C 2001 Validation of tissue microarrays for immunohistochemical profiling of cancer specimens using the example of human fibroblastic tumors. Am J Pathol 158:1245–1251[Abstract/Free Full Text]
  31. Lassus H, Laitinen MP, Anttonen M, Heikinheimo M, Aaltonen LA, Ritvos O, Butzow R 2001 Comparison of serous and mucinous ovarian carcinomas: distinct pattern of allelic loss at distal 8p and expression of transcription factor GATA-4. Lab Invest 81:517–526[Medline]
  32. Salhi I, Cambon-Roques S, Lamarre I, Laune D, Molina F, Pugniere M, Pourquier D, Gutowski M, Picard JY, Xavier F, Pelegrin A, Navarro-Teulon I 2004 The anti-Mullerian hormone type II receptor: insights into the binding domains recognized by a monoclonal antibody and the natural ligand. Biochem J 379:785–793[CrossRef][Medline]
  33. Bielinska M, Parviainen H, Porter-Tinge SB, Kiiveri S, Genova E, Rahman N, Huhtaniemi IT, Muglia LJ, Heikinheimo M, Wilson DB 2003 Mouse strain susceptibility to gonadectomy-induced adrenocortical tumor formation correlates with the expression of GATA-4 and luteinizing hormone receptor. Endocrinology 144:4123–4133[Abstract/Free Full Text]
  34. Kiiveri S, Siltanen S, Rahman N, Bielinska M, Lehto VP, Huhtaniemi IT, Muglia LJ, Wilson DB, Heikinheimo M 1999 Reciprocal changes in the expression of transcription factors GATA-4 and GATA-6 accompany adrenocortical tumorigenesis in mice and humans. Mol Med 5:490–501[Medline]
  35. Suzuki YJ, Day RM, Tan CC, Sandven TH, Liang Q, Molkentin JD, Fanburg BL 2003 Activation of GATA-4 by serotonin in pulmonary artery smooth muscle cells. J Biol Chem 278:17525–17531[Abstract/Free Full Text]
  36. Aries A, Paradis P, Lefebvre C, Schwartz RJ, Nemer M 2004 Essential role of GATA-4 in cell survival and drug-induced cardiotoxicity. Proc Natl Acad Sci USA 101:6975–6980[Abstract/Free Full Text]
  37. Jimenez P, Saner K, Mayhew B, Rainey WE 2003 GATA-6 is expressed in the human adrenal and regulates transcription of genes required for adrenal androgen biosynthesis. Endocrinology 144:4285–4288[Abstract/Free Full Text]
  38. Kiiveri S, Liu J, Westerholm-Ormio M, Narita N, Wilson DB, Voutilainen R, Heikinheimo M 2002 Differential expression of GATA-4 and GATA-6 in fetal and adult mouse and human adrenal tissue. Endocrinology 143:3136–3143[Abstract/Free Full Text]
  39. Takayama K, Sasano H, Fukaya T, Morohashi K, Suzuki T, Tamura M, Costa MJ, Yajima A 1995 Immunohistochemical localization of Ad4-binding protein with correlation to steroidogenic enzyme expression in cycling human ovaries and sex cord stromal tumors. J Clin Endocrinol Metab 80:2815–2821[Abstract]



This article has been cited by other articles:


Home page
 Mol. Endocrinol.Home page
R. S. Viger, S. M. Guittot, M. Anttonen, D. B. Wilson, and M. Heikinheimo
Role of the GATA Family of Transcription Factors in Endocrine Development, Function, and Disease
Mol. Endocrinol., April 1, 2008; 22(4): 781 - 798.
[Abstract] [Full Text] [PDF]

Home page
 J Mol EndocrinolHome page
M Anttonen, H Parviainen, A Kyronlahti, M Bielinska, D B Wilson, O Ritvos, and M Heikinheimo
GATA-4 is a granulosa cell factor employed in inhibin-{alpha} activation by the TGF-{beta} pathway.
J. Mol. Endocrinol., June 1, 2006; 36(3): 557 - 568.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
90/12/6529    most recent
Author Manuscript (PDF)
Right arrow Submit a related Letter to the Editor
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Anttonen, M.
Right arrow Articles by Heikinheimo, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Anttonen, M.
Right arrow Articles by Heikinheimo, M.
Related Collections
Right arrow Endocrine Oncology
Right arrow Female Endocrinology

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals