Reciprocal Changes in the Expression of Transcription Factors GATA-4 and GATA-6 Accompany Adrenocortical Tumorigenesis in Mice and Humans
Sanne Kiiveri,1 Susanna Siltanen,1 Nafis Rahman,2 Malgorzata Bielinska,3 Veli-Pekka Lehto,4 Ilpo T. Huhtaniemi,2 Louis J. Muglia,3,5 David B. Wilson,3,5 and Markku Heikinheimo1,3
1Children’s Hospital, University of Helsinki, Helsinki, Finland
2Department of Physiology, University of Turku, Turku, Finland
3Department of Pediatrics, Washington University, St. Louis, Missouri, U.S.A.
4Department of Pathology, University of Oulu, Oulu, Finland
5Department of Molecular Biology and Pharmacology, Washington University, St. Louis, Missouri, U.S.A. Accepted June 10, 1999.
Abstract
While certain genetic changes are frequently found in adrenocortical carcinoma cells, the molecular basis of adrenocortical tumorigenesis remains poorly under- stood. Given that the transcription factors GATA-4 and GATA-6 have been implicated in gene expression and cellular differentiation in a variety of tissues, including endocrine organs such as testis, we have now examined their expression in the developing adrenal gland, as well as in adrenocortical cell lines and tumors from mice and humans. Northern blot analysis and in situ hybridization revealed abundant GATA-6 mRNA in the fetal and post- natal adrenal cortex of the mouse. In contrast, little or no GATA-4 expression was detected in adrenal tissue dur- ing normal development. In vivo stimulation with ACTH or suppression with dexamethasone did not affect the expression of GATA-4 or GATA-6 in the murine adrenal gland. To assess whether changes in the expression of GATA-4 or GATA-6 accompany adrenocortical tumori-
genesis, we employed an established mouse model. When gonadectomized, inhibin «/SV40 T-antigen trans- genic mice develop adrenocortical tumors in a gonado- tropin-dependent fashion. In striking contrast to the nor- mal adrenal glands, GATA-6 mRNA was absent from adrenocortical tumors or tumor-derived cell lines, while GATA-4 mRNA and protein were abundantly expressed in the tumors and tumor cell lines. Analogous results were obtained with human tissue samples; GATA-4 ex- pression was detected in human adrenocortical carcino- mas but not in normal tissue, adenomas, or pheochro- mocytomas. Taken together these results suggest different roles for GATA-4 and GATA-6 in the adrenal gland, and implicate GATA-4 in adrenal tumorigenesis. Immunohistochemical detection of GATA-4 may serve as a useful marker in the differential diagnosis of human adrenal tumors.
Introduction
Adrenocortical carcinoma (ACC) is a rare tumor affecting both children and adults. In the United
States the incidence is approximately 3 per mil- lion children less than 16 years of age (1) and 0.5-2 per million adults (2), although a higher
incidence has been reported in some countries (3). The majority of pediatric patients (>95%) exhibit virilization (3), and Cushing’s syndrome is common (>70%) in pediatric and adult pa- tients (3,4). Overall ACC represents 0.2% of the causes of death from cancer (2). The mainstay of treatment for ACC is surgery (3,5,6); when total resection is possible, the one year survival rate in pediatric patients is approximately 65% (3). An- tineoplastic agents have been used with limited success in cases of advanced disease (5,7,8).
The molecular mechanisms underlying ad- renocortical tumorigenesis are not known. The clonality of adrenocortical tumors has been ex- amined using X-chromosome inactivation analysis on human pathologic samples; most adrenal adenomas and carcinomas are mono- clonal in nature, whereas diffuse and nodular adrenal hyperplasias are polyclonal (9). ACC is found in association with the Li-Fraumeni (10) and Beckwith-Wiedemann (11) syndromes, suggesting that malignant transformation may be influenced by the p53 gene (1,12,13) or imprinted genes at the 11p15 locus such as p57KIP2 (14). In addition, allelic loss of the RB gene has been noted in ACC samples, indicat- ing that this gene may be involved in tumori- genesis (13). Multiple endocrine neoplasia type 1 (MEN-1) can involve the adrenal cortex (15), and ACC cells express markers typical of neuroendocrine cells, such as synaptophysin, neuron specific enolase, and vimentin, imply- ing that ACC may originate from neuroendo- crine foci within the adrenal cortex (16).
A transgenic mouse model of adrenal tumor- igenesis has been recently described; mice bear- ing an inhibin «-subunit promoter/Simian Virus 40 T-antigen fusion gene develop gonadal tu- mors originating from granulosa or Leydig cells (17-19). When these transgenic mice are gonad- ectomized before puberty, they develop malig- nant adenocorticortical carcinomas (20). These adrenal tumors do not appear in non-gonadec- tomized animals. If functional gonadectomy is induced by administration of a GnRH antagonist or by cross-breeding the transgenic mice into the
hypogonadotropic hpg genetic background, nei- ther gonadal nor adrenal tumors appear (21). This led to the hypothesis that tumor develop- ment is related to elevated gonadotropin secre- tion, which is the most obvious difference be- tween the surgical and functional gonadectomy models (22). The adrenal tumors and a cell line (Cal) derived from them were found to express functional LH receptors, and hCG treatment of the cells stimulated their proliferation and ste- roid production (22). On the basis of this and other data it has been proposed that expression of the potent oncogene, T-antigen, in adrenocor- tical cells allows a trophic hormone to function as a tumor promoter. While obvious differences exist between this mouse model and ACC in humans, certain similarities are noteworthy. Ad- renal carcinomas in humans and the transgenic mice express a number of common markers, in- cluding neuroectodermal markers and steroido- genic enzymes (16,20,22). Moreover, human ad- renocortical tumors responsive to gonadotropins have been described (23,24). These conserved features make this transgenic model potentially useful for identifying factors critical for adreno- cortical tumorigenesis in humans.
GATA-4 and GATA-6 belong to a family of zinc finger transcription factors termed the GATA-binding proteins, which regulate gene ex- pression, differentiation, and cell proliferation in a variety of tissues (25,26). Within the endocrine system, GATA-4 and/or GATA-6 are expressed in the female and male gonad (27-33), and devel- oping hypothalamus (34,35). In ovary, GATA-4 expression localizes to granulosa cells of primary and early antral follicles, whereas GATA-6 is present in late antral follicular cells and luteal glands (31). In addition, the expression of GATA-4, but not GATA-6, is stimulated by exog- enous gonadotropins in cultured gonadal cell lines (31,33). Furthermore, downregulation of GATA-4 expression accompanies follicular atre- sia caused by the mechanism of programmed cell death. In testis, GATA-4 and GATA-6 expression are most abundant at the proliferative stage of Sertoli cells, suggesting a role in the expansion of these cells (32,33). Taken together, these studies imply overlapping but distinct functions for GATA-4 and GATA-6 in the hypothalamic-pitu- itary-gonadal axis, and suggest that the relative expression of these factors within these tissues may regulate hormonal signaling or cell prolifer- ation.
Given the expression and regulation of GATA-4 and GATA-6 in the gonads, it was of
This work was performed at Children’s Hospital, University of Helsinki and Department of Pediatrics, Washington Uni- versity of St. Louis. S. Kiiveri and S. Siltanen contributed equally to this work.
interest to explore their expression in the adre- nals, which share expression of many develop- mentally and functionally important genes, in- cluding inhibin, SF-1 and DAX-1 (36 and references therein) with the gonads. We have also utilized the transgenic mouse model de- scribed above, and human pathological sam- ples to explore the role of GATA-4 and GATA-6, in adrenocortical tumorigenesis. Abundant GATA-6 mRNA was found in adre- nal cortex during mouse development, whereas GATA-4 mRNA was undetectable. In striking contrast to the findings in normal ad- renal gland, GATA-6 was absent from murine adrenocortical tumors, while GATA-4 mRNA expression was dramatically upregulated in these mouse tumors as well as in human adre- nocortical carcinomas.
Materials and Methods
Mouse Stocks
Normal adrenal glands were obtained from 7-week-old male C57BL/6J mice (Jackson Labs, Bar Harbor, ME). Mouse embryos were obtained by mating male and female B6DJLF1/J mice. For estimating embryonal age, noon of the day on which the copulation plug was found was con- sidered as 0.5 days p.c. Precise staging of dis- sected embryos was performed using The Atlas of Mouse Development (37). The inhibin @-subunit promoter/SV40 T-antigen transgenic mice have been described previously (17,18,20).
In Vivo Manipulations of Adrenal Function
To study the hormonal regulation of GATA-4 and GATA-6 in the adrenal gland we treated mice (three animals in each group) with intra- peritoneal injection of the following agents: ACTH 10 µg/kg body weight in one dose 2 hours prior to organ harvest, and dexamethasone 0.03 mg/kg body weight daily for either 2 days (acute suppression) or 5 days (chronic suppression) starting 3 or 6 days prior to organ harvest, re- spectively. Adrenal glands were harvested also from a group of animals at evening to assess possible diurnal variation in GATA-4/-6 expres- sion. The RNA isolated from the adrenals of these animals was studied for GATA-4 and GATA-6 expression by RNase protection.
Human Samples
Paraffin-embedded tissue samples originally col- lected for diagnostic purposes were analyzed for their expression of GATA-4 protein. These sam- ples originated from 3 cases of ACC (2 females, 1 male, aged 35-78 yr), 5 cases of benign adrenal adenomas (3 females, 2 males, aged 48-73 yr) and 4 cases of pheochromocytomas (3 females, 1 male, aged 12-68 yr).
Adrenal Cell Cultures and Hormonal Stimulation of the Cell Lines
An immortalized cell line, termed Cal, derived from an adrenocortical tumor in a transgenic mouse line bearing the mouse inhibin @-subunit promoter/Simian virus 40 T-antigen fusion gene, has been described elsewhere (20). The cells were cultured on plastic dishes in HEPES (20 mmol/l) buffered Dulbecco’s Modified Essential Media (DMEM) with GlutaMAX®/F12 1:1 (Gibco BRL, Paisley, Scotland) supplemented with 10% heat-inactivated fetal calf serum (FCS, Biochlear, Berks, U.K.), glucose (4.5 g/l) and gentamicin (100 mg/l). Cells were used in im- munohistochemistry after 2-3 days in culture. Untreated Cal cells or cells treated with ACTH (50 ng/ml) were collected for RNA isolation and Northern hybridization performed as described later.
In Situ Hybridization
Mouse embryos or adrenal glands were washed briefly in PBS and then frozen in O.C.T. cryo- preservation solution (Tissue Tek®, Miles Inc., Elkhart, IN). Mouse adrenal tumors were fixed with 4% paraformaldehyde and embedded in paraffin. Frozen sections (10 um) were fixed in 4% paraformaldehyde in PBS, or sections (10 um) from paraffin embedded tissues were sub- jected to in situ hybridization as described (38). Tissue sections were incubated with 1 × 106 CPM of [33P]-labeled (1000-3000 Ci/mmol, Am- ersham, Life Technologies, Arlington Heights, IL) antisense or sense riboprobe in a total volume of 80 ul. Antisense and sense riboprobes for GATA-4 and GATA-6 were prepared as described elsewhere (27,31,39).
Northern Hybridization
Total RNA was isolated using guanidinium thio- cyanate-phenol-chloroform extraction and ana- lyzed for expression of GATA-4 or GATA-6 mes-
sage using Northern hybridization (27). Twenty micrograms of denaturated total RNA was sub- jected to electrophoresis on 1% denaturing aga- rose gel and then transferred onto nylon mem- branes (Hybond N, Amersham, Arlington Heights, IL). The membranes were hybridized with [32P]labeled RNA probes for GATA-4 (27) or GATA-6 (29). Hybridization and washing of the membranes were performed as previously described (27).
Hybridization signals were detected by auto- radiography using Kodak X-Omat AR Diagnostic film XAR5. Autoradiograms were scanned by the Microcomputer Imaging device (MCID, version 1.2, from Imaging Research, Inc., St. Catherines, Ontario, Canada) to quantify messenger RNA (mRNA) species. The intensities of specific bands were quantified using Tina software (Raytest, Straubenhardt, Germany). The results were ex- pressed in arbitrary densitometric units (percent of the control value) corrected according to the intensity of 28S ribosomal subunit in the gel stained with ethidium bromide. Statistical anal- yses were performed on the basis of three inde- pendent experiments. The data were analyzed by one-way ANOVA, using a Macintosh version of the SuperANOVA program (Abacus Concepts, Inc., Berkeley, CA), followed by Duncan’s new multiple range and Fisher’s protected last signif- icant differences post-hoc tests. P < 0.05 was considered statistically significant. The results shown in the figures represent the mean ± SEM.
RNase Protection Assay
RNA from adrenal and ovarian tissue was used for RNase protection assays performed with a commercially available kit (Ambion, Austin, TX) using 10 µg of total RNA. [32P]labeled (800 Ci/ mmol, Amersham Life Technologies, Arlington Heights, IL) antisense riboprobe recognizing transcripts arising from exons II and III of the Gata4 gene was prepared by in vitro transcription of NotI linearized G14A plasmid (27) using T7 RNA polymerase and [a-32P]CTP (650 Ci/mmol, ICN Pharmaceuticals Inc., Costa Mesa, CA). An- tisense GATA-6 riboprobe was prepared from PCR II mGATA-6 plasmid (29) linearizing with PstI and using SP6 RNA polymerase. Antisense ß-actin probe was prepared according to the manufacturer’s recommendations (Ambion, Austin, TX). The sizes of full-length and pro- tected RNA probes were as follows: GATA-4, 491 and 430 nucleotides, GATA-6 140 and 100 nu-
cleotides, and ß-actin 300 and 250 nucleotides, respectively.
Immunohistochemistry and Immunocytochemistry
Frozen tissue sections from mouse adrenals and adrenal tumors, cultured mouse adrenocortical tumor cells (line Cal), and paraffin-embedded human tissue samples were fixed in 4% parafor- maldehyde and subjected to immunohistochem- istry using either affinity purified polyclonal rab- bit anti-mouse GATA-4 (27,39) or commercial goat polyclonal anti-mouse GATA-4 IgG (Santa Cruz Biotechnology, Santa Cruz, CA) or non- immune IgG as the primary antibody. A com- mercially available avidin-biotin immunoperox- idase system was used to visualize bound antibody (Vectastain Elite ABC Kit, Vector Lab- oratories, Burlingame, CA). 3-amino-9-ethylcar- bazole (Sigma Chemicals, St. Louis, MO) was used as the chromogen and the development reaction occurred in the presence of 0.03% H2O2. T-antigen was stained with rabbit poly- clonal anti-Tag antibody, kindly provided by DR. D. Hanahan, University of California, San Francisco, CA (40), as described (22).
Results
GATA-6, but Not GATA-4, Transcripts Are Expressed in Normal Mouse Adrenal Gland
We first wanted to address the normal devel- opmental pattern of GATA-4 and GATA-6 ex- pression in the adrenal gland. In initial exper- iments, we examined the expression of GATA-4 and GATA-6 in normal mouse adrenal gland using RNA analysis. In situ hybridization revealed abundant expression of GATA-6 mes- sage in postnatal adrenal cortex (Fig. 1A, B). GATA-6 mRNA was uniformly distributed throughout the zones of the adrenal cortex, but no GATA-6 message was evident in the adrenal medulla. When in situ hybridization for GATA-4 was performed on adjacent tissue sections through normal adrenal gland, no GATA-4 message was detected (Fig. 1C, D). Similar results were obtained in in situ hybrid- izations performed on fetal (day 18 p.c.) adre- nals; abundant GATA-6 message was present in the cortex, but it was devoid of GATA-4 mRNA (data not shown). RNase protection as- says confirmed these findings; abundant GATA-6 but no GATA-4 mRNA was detected in normal postnatal adrenal gland (Fig. 1E).
A
B
C
C
m
m
C
D
C
C
m
m
Effect of In Vivo Manipulation of Adrenal Function on the Expression of GATA-4 and GATA-6 mRNA In vivo stimulation with ACTH for 2 hours (10 ug/kg) or suppression with dexamethasone (0.03 mg/kg/day for 2 or 5 days) did not result in any marked change in the steady state levels of GATA-6 mRNA in the adrenal gland, as mea- sured by RNase protection assay (Fig. 1E). Nei- ther ACTH nor dexamethasone had any effect on GATA-4 mRNA levels, which remained unde- tectable (Fig. 1E). No diurnal variation was noted in the expression of GATA-6 in the adrenals as adrenals harvested in the evening were studied (data not shown); GATA-4 mRNA remained un- detected.
GATA-4 is Upregulated and GATA-6 Downregulated in Mouse Adrenocortical Tumors To assess whether changes in GATA-4 or GATA-6 expression accompany adrenocortical tumorigenesis, we analyzed inhibin « promoter/
50 um. (E) RNase protection assay showing expres- sion of GATA-6 mRNA in normal adrenal cortex (Co). No change in the expression level was noted in animals treated with ACTH or short or long treat- ment with dexamethasone (short- and long dexam., respectively). No GATA-4 mRNA could be detected in the adrenals of control or treated animals. Mouse testis RNA was used as a positive control.
SV40 Tag transgenic mice, developing adreno- cortical tumors after gonadectomy (20). In intact and gonadectomized control mice, and in intact transgenic mice, there was no GATA-4 mRNA expression in the adrenal glands (Fig. 2C). With subsequent tumor formation of the gonadecto- mized transgenic mice there was abundant GATA-4 mRNA in the adrenal tumors (Fig. 2A). In contrast, GATA-6 mRNA was present in the normal and non-tumorous transgenic adrenal glands, but totally absent in the adrenal tumors (Fig. 2B, C).
In addition to the RNA analysis, we per- formed immunohistochemistry on adrenal tu- mors from the transgenic mice. The immuno- peroxidase staining revealed uniform staining for GATA-4 protein in the tumor cells, but not in the adjacent normal adrenal tissue (Fig. 2E, F). In keeping with these results, in situ hy- bridization demonstrates GATA-4, but not GATA-6, mRNA in adrenal tumor tissue (Fig. 2H, I).
E
GATA-4 …
Actin
GATA-6
Co
ACTH
Short dexam.
Long dexam.
Testis
GATA-4, but Not GATA-6, Is Expressed in Mouse Adrenocortical Tumor Cell Lines
In addition to examining adrenal tumors of the transgenic mice, we analyzed the adrenocortical cell line Cal derived from one of the tumors. Consistent with the observations made on pri- mary tumors, the tumor cell line expressed abundant amounts of GATA-4 mRNA (Fig. 2A) and protein (Fig. 3), but not the GATA-6 mes- sage (Fig. 2B). In keeping with the in vivo find- ings, no change in GATA-4 or GATA-6 mRNA levels was observed when the tumor cell lines were treated with ACTH (data not shown).
GATA-4 Is Present in Human Adrenocortical Carcinomas
To assess whether the observations made on nor- mal and transgenic mice were relevant to human adrenocortical carcinogenesis, we performed im- munohistochemistry for GATA-4 in ACC, adre- nal adenomas and pheochromocytomas. We de- tected nuclear staining for GATA-4 in the tumor tissue from all the three ACC samples examined (Fig. 4A, B), but not in normal adrenal cortex (Fig. 4D, E) nor five adenomas (Fig. 4F, G), nor four pheochromocytomas (Fig. 4H, I). We were unable to assess the presence of GATA-6 protein
in these samples, since no suitable anti-GATA-6 antibodies were available.
Discussion
This study demonstrates that GATA-4 and GATA-6 are differentially regulated in normal versus malignant adrenal glands. Whereas GATA-6 expression is related to normal adrenal organogenesis, GATA-4 is expressed in this organ only during tumorigenesis implicating a possible role for this transcription factor in adrenocortical carcinogenesis. This is the first evidence of a link between any GATA binding protein and solid tumors in mammals. Observations of the trans- genic adrenals at different stages of the transfor- mation process indicate that GATA-4 expression is specific for the malignant adrenocortical cells, but not for the transgenic T-antigen expression in normal tissue.
Earlier studies on gonadal GATA-4 expres- sion have revealed that abundant expression of this transcription factor is associated with the proliferative phase of ovarian granulosa and tes- ticular Sertoli cells (31-33). The high level of expression in the tumor cells with high prolifer- ative capacity is consistent with this finding. Moreover, studies on ovarian follicular atresia,
A
EtBr gel
GATA-4
28S
GATA-4 mRNA Corrected for 28S Ribosome (% of Control)
500
**
**
T
1
400
L
*
T
300
*
I
200
100
0
Sp
I
Te
Adr. Co.
Adr. Tumor I
Calpha1 cells
M
M
F
B
EtBr gel
-28S
GATA-6
GATA-6 mRNA Corrected for 28S Ribosome (% of Control)
500
**
T
-
400
**
王
300
200
100
0
Sp
1
Te
Adr. Co.
1
Adr. Tumor
1
Calpha1 cells
M
M
I
F
C
EtBr gel
GATA-4
-28S
GATA-4 & GATA-6 mRNA Corrected for 28S Ribosome (% of Control)
GATA-6
-28S
400
300
I
1
1
L
200
100
0
Control Adrenals
TG Adrenals
sex
F
F
M
F
M
F
Gonadectomy
+
+
-
-
+
D
G
E
H
1
tu
nl
F
I
tu
1
nl
occurring through apoptosis, demonstrated downregulation of GATA-4 mRNA in association with programmed cell death (31). It is tempting to speculate that GATA-4 could act as a cell sur- vival factor during the transformation so that the maintenance of expression of this transcription factor could enhance tumor progression and growth.
To fully understand the roles of GATA-4 and GATA-6 in normal and malignant adrenal devel- opment, it will be necessary to define the regu- lation and target genes for these transcription factors in this organ. The combined evidence from this study and earlier observations in the gonad shows that these transcription factors are present at the sites of steroidogenesis, and more importantly also at the time of active steroido- genesis in these organs, in particular in fetal and adult testis and adrenal. These findings suggest a role for GATA-4 and/or GATA-6 in steroidogen- esis. Some genes for the steroidogenic enzymes, for example that of aromatase, harbor putative GATA-binding sites in the regulatory elements of their genes (41), and are thus candidate targets for GATA proteins expressed in endocrine or- gans. Several other genes, such as anti-Müllerian hormone, inhibin a, and brain type natriuretic peptide have been proposed as potential target genes for GATA-4 or GATA-6 in ovary and testes (32,33,42,43). In hypothalamus, GATA-4 has been implicated in the regulation of GnRH gene (34,35,44). Rigorous genetic studies would, however, be required to establish the in vivo
target genes for GATA-4 and GATA-6 in the endocrine system. Unfortunately, the published models of homozygous null mutations in GATA4 (45,46) and GATA6 (47) gene cause early embry- onic lethality, which precludes analysis of the role of these transcription factors in adrenal de- velopment.
The regulatory mechanisms in the adrenal tumorigenesis of the transgenic mouse model are still unknown, but recent evidence shows that suppression of gonadotropins in these animals inhibits tumor formation (21). Although gonad- otropin receptors are not present in normal ad- renal gland, LH receptors were expressed in the adrenal tumors (22). Of interest, our results in gonadal tumor cell lines showed that gonadotro- pins upregulate GATA-4 mRNA (31,33). Specif- ically these studies showed upregulation of GATA-4, but not GATA-6, in Sertoli and granu- losa cell lines by FSH and in Leydig cell lines by hCG. It is thus possible that the signal triggered by LH in the LHR-bearing adrenal tumor cells acts through GATA-4, which may then partici- pate in the events leading to malignant pheno- type in these cells. Our preliminary data supports the association of LH receptor and GATA-4 ex- pression in the tumorous adrenals.
The reciprocal changes in GATA-4 and GATA-6 expression associated with adrenocorti- cal tumorigenesis are consistent with earlier studies showing that GATA-4 can influence ex- pression of GATA-6 and vice versa. In GATA4-/- embryos (45,46) and cell lines (48) there is up- regulation of GATA-6 mRNA expression, while in GATA6-/- embryos (47) there is decreased ex- pression of GATA-4 mRNA. The molecular mechanisms for this cross-regulation are not known.
The current findings, in particular those on the human tumors, may also offer new tools for the diagnosis of adrenal tumors. There are no reliable cytologic criteria for distinguishing adrenocortical carcinomas from adenomas or renal cell carcinomas by light microscopy (49), although certain morphological features have been suggested to be typical for ACC (50). In addition, adrenocortical carcinoma can mimic benign lesions histologically (51). Immunocy- tochemistry of 3-beta-hydroxysteroid dehy- drogenase and adrenal-4-binding protein (SF- 1), a transcription factor implicated in the regulation of steroidogenesis, may be useful in the differential diagnosis between adrenocorti- cal and renal cell carcinomas by identifying adrenocortical parenchymal cells (49). Nuclear
A
D
E
B
F
G
C
H
I
staining with the D11 monoclonal antibody has also been shown to be useful for discrimi- nating ACC from intraadrenal metastases and adrenomedullary tumors (52). We are now in the process of further evaluating the roles of GATA transcription factors in human adrenal neoplasia. Based on the present findings in the normal and malignant human adrenal tissue, we propose that immunohistochemical detec- tion of GATA-4 might serve as a new nuclear marker to discriminate between malignant and benign adrenal tumors.
nonimmune serum (ACC). GATA-4 protein is abun- dantly present in the nuclei of carcinoma cells (A-C, arrowheads), but absent from normal cortex, adeno- mas, and pheochromocytomas. Original magnifica- tions (×200) for (A-C), bar = 50 um; (×100) for (D-I), bar = 100 um.
Acknowledgments
This work was supported by Juselius Foundation (ITH, DBW, and MH), the Finnish Cancer Foun- dation (ITH), the Finnish Pediatric Foundation (MH), the University Central Hospital in Helsinki (SK, SS, and MH), the NIH (LJM), a Burroughs Wellcome Fund Career Development Award (LJM), Pediatric Cardiology SCOR HL61006 (DBW), the Washington University Monsanto- Searle Agreement (DBW), and an Established Investigator Award from the AHA (DBW).
References
1. Wagner J, Portwine C, Rabin K, Leclerc JM, Narod SA, Malkin D. (1994) High frequency of germline p53 mutations in childhood adrenocortical cancer. J. Natl. Cancer Inst. 86: 1707-1710.
2. Cutler SJ, Young JL (eds). (1975) Biometrics Branch, National Cancer Institute, Third National Cancer Survey: incidence data. National Cancer Institute Monograph 41. Government Printing Of- fice, Washington, D.C.
3. Sabbaga CC, Avilla SG, Schulz C, Garbers JC, Blucher D. (1994) Adrenocortical carcinoma in children: clinical aspects and prognosis. J. Pediatr. Surg. 28: 841-843.
4. Boscaro M, Fallo F, Barzon L, Daniele O, Sonino N. (1995) Adrenocortical carcinoma: epidemiol- ogy and natural history. Minerva Endocrinol. 20: 89-94.
5. Haak HR, Hermans J, van de Velde CJ, et al. (1994) Optimal treatment of adrenocortical carci- noma with mitotane: results in a consecutive se- ries of 96 patients. Brit. J. Cancer 69: 947-951.
6. Favia G, Lumachi F, Carraro P, D’Amico DF. (1995) Adrenocortical carcinoma. Our experi- ence. Minerva Endocrinol. 20: 95-99.
7. Dogliotti L, Berruti A, Pia A, Paccotti P, Ali A, Angeli A. (1995) Cytotoxic chemotherapy for ad- renocortical carcinoma. Minerva Endocrinol. 20: 105-109.
8. Kasperlik-Zaluska AA, Migdalska BM, Zgliczynski S, Makowska AM. (1995) Adrenocortical carci- noma. A clinical study and treatment results of 52 patients. Cancer 75: 2587-2591.
9. Beuschlein F, Reincke M, Karl M, et al. (1994) Clonal composition of human adrenocortical neo- plasms. Cancer Res. 54: 4927-4932.
10. Li FP, Fraumeni Jr JF, Mulvihill JJ, et al. (1988) A cancer family syndrome in twenty-four kindreds. Cancer Res. 48: 5358-5362.
11. Wiedemann HR. (1983) Tumours and hemihy- pertrophy associated with Wiedeman-Beckwith syndrome. Eur. J. Pediatr. 141: 129.
12. Birch JM, Hartley AL, Tricker KJ, et al. (1994) Prevalence and diversity of constitutional muta- tions in the p53 gene among 21 Li-Fraumeni fam- ilies. Cancer Res. 54: 1298-1304.
13. Miyamoto H, Kubota Y, Shuin T, Shiozaki H. (1996) Bilateral adrenocortical carcinoma show- ing loss of heterozygosity at the p53 and RB gene loci. Cancer Genet. Cytogenet. 88: 181-183.
14. Hatada I, Ohashi H, Fukushima Y, et al. (1996) An imprinted gene p57KIP2 is mutated in Beckwith- Wiedemann syndrome. Nat. Genet. 14: 171-173.
15. Skogseid B, Larsson C, Lindgren P-G, et al. (1992) Clinical and genetic features of adrenocortical le- sions in multiple endocrine neoplasia type 1. J. Clin. Endocrinol. Metab. 75: 76-81.
16. Haak HR, Fleuren GJ. (1995) Neuroendocrine dif-
ferentiation of adrenocortical tumors. Cancer 75: 860-864.
17. Kananen K, Markkula M, Rainio E, Su J-GJ, Hsueh AJW, Huhtaniemi IT. (1995) Gonadal tu- morigenesis in transgenic mice bearing the mouse inhibin a-subunit promoter/Simian virus T-anti- gen fusion gene: characterization of ovarian tu- mors and establishment of gonadotropin-respon- sive granulosa cell lines. Mol. Endocrinol. 9: 616- 627.
18. Kananen K, Markkula M, el Hefnawy T, et al. (1996) The mouse inhibin @-subunit promoter directs SV40 T-antigen to Leydig cells in trans- genic mice. Mol. Cell. Endocrinol. 119: 135-146.
19. Rahman NA, Kananen K, Rilianawati X, et al. (1998) Transgenic mouse models for gonadal tu- morigenesis. Mol. Cell. Endocrinol. 145: 167-174.
20. Kananen K, Markkula M, Mikola M, Rainio E-M, McNeilly A, Huhtaniemi I. (1996) Gonadectomy permits adrenocortical tumorigenesis in mice transgenic for the mouse inhibin a-subunit pro- moter/Simian virus T-antigen fusion gene: evi- dence for negative autoregulation of the inhibin a-subunit gene. Mol. Endocrinol. 10: 1667-1677.
21. Kananen K, Rilianawati X, Paukku T, Markkula M, Rainio E-M, Huhtaniemi I. (1997) Suppression of gonadotropins inhibits gonadal tumorigenesis in mice transgenic for the mouse inhibin a-sub- unit promoter/Simian virus T-antigen fusion gene. Endocrinology 138: 3521-3531.
22. Rilianawati X, Paukku T, Kero J, et al. (1998) Direct luteinizing hormone action triggers adreno- cortical tumorigenesis in castrated mice transgenic for the murine inhibin @-subunit promoter/Sim- ian virus 40 T-antigen fusion gene. Mol. Endocrinol. 12: 801-809.
23. deLange WE, Pratt JJ, Doorenbos H. (1980) A gonadotropin-responsive testosterone producing adrenocortical adenoma and high gonadotropin levels in an elderly woman. Clin. Endocrinol. 12: 21-28.
24. Leinonen P, Ranta T, Siegberg R, Heikkilä P, Kahri A. (1991) Testosterone secreting virilizing adrenal adenoma with human chorionic gonatropin re- ceptors and 21-hydroxylase deficiency. Clin. Endo- crinol. 34: 31-35.
25. Orkin SH. (1992) GATA-binding transcription fac- tors in hematopoietic cells. Blood 80: 575-581.
26. Evans T. (1997) Regulation of cardiac gene ex- pression by GATA-4/5/6. Trends Card. Med. 7: 75- 83.
27. Arceci RJ, King AAJ, Simon MC, Orkin SH, Wil- son DB. (1993) Mouse GATA-4: a retinoic acid- inducible GATA-binding transcription factor ex- pressed in endodermal derivatives and heart. Mol. Cell Biol. 13: 2235-2246.
28. Laverriere AC, MacNeill C, Mueller C, Poelman RE, Burch JBE, Evans T. (1994) GATA4/5/6, a subfamily of three transcription factors tran-
scribed in developing heart and gut. J. Biol. Chem. 269: 23177-23184.
29. Narita N, Heikinheimo M, Bielinska M, White RA, Wilson DB. (1996) The gene for transcription fac- tor GATA-6 resides on mouse chromosome 18 and is expressed in myocardium and vascular smooth muscle. Genomics 36: 345-348.
30. Morrisey EE, Ip HS, Lu MM, Parmacek MS. (1996) GATA-6: a zinc finger transcription factor that is derived from lateral mesoderm. Dev. Biol. 177: 309-322.
31. Heikinheimo M, Ermolaeva M, Bielinska M, et al. (1997) Expression and hormonal regulation of transcription factors GATA-4 and GATA-6 in the mouse ovary. Endocrinology 138: 3505-3514.
32. Viger RS, Mertineit C, Trasler JM, Nemer M. (1998) Transcription factor GATA-4 is expressed in asexually dimorphic pattern during mouse go- nadal development and is a potent activator of the Müllerian inhibiting substance. Development 125: 2665-2675.
33. Ketola I, Rahman N, Toppari J, et al. (1999) Ex- pression and regulation of transcription factors GATA-4 and GATA-6 in developing mouse testis. Endocrinology 140: 1470-1480.
34. Lawson MA, Mellon PL. (1998) Expression of GATA-4 in migrating gonadotropin-releasing neurons of the developing mouse. Mol. Cell. Endo- crin. 140: 157-161.
35. Lawson MA, Buhain AR, Jovenal JC, Mellon PL. (1998) Multiple factors interacting with GATA sites of the gonadotropin releasing hormone neu- ron specific enhancer regulate gene expression. Mol. Endocrin. 12: 364-377.
36. Parker KL, Schimmer BP. (1997) Steroidogenic factor 1: a key determinant of endocrine develop- ment and function. Endocrin. Rev. 18: 361-377.
37. Kaufman MH. (1992) The Atlas of Mouse Devel- opment. London: Academic Press.
38. Wilkinson DG. (1992) Whole mount in situ hy- bridization of vertebrate embryos. In: Wilkinson DG (ed), In situ Hybridization-A Practical Ap- proach. IRL Press, Oxford, pp. 75-83.
39. Heikinheimo M, Scandrett J, Wilson DB. (1994) Localization of transcription factor GATA-4 to re- gions of mouse embryo involved in cardiac devel- opment. Dev. Biol. 164: 361-373.
40. Efrat S, Hanahan D. (1987) Bidirectional activity
of the rat insulin II 5’ flanking region in transgenic mice. Mol. Cell. Biol. 7: 192-198.
41. Richards JS. (1994) Hormonal control of gene expression in the ovary. Endocrin. Rev. 15: 725- 751.
42. Yomodiga K, Ohtani H, Harigae H, et al. (1994) Developmental stage- and spermatogenic cycle- specific expression of transcription factor GATA-1 in mouse Sertoli cells. Development 120: 1759- 1766.
43. Grépin C, Dagnino L, Robitaille L, Haberstroh L, Antakly T, Nemer M. (1994) A hormone-encod- ing gene identifies a pathway for cardiac but not skeletal muscle gene transcription. Mol. Cell. Biol. 14: 3115-3129.
44. Lawson MA, Whyte DB, Mellon PL. (1996) GATA factors are essential for activity of the neuron- specific enhancer of the gonadotropin-releasing hormone gene. Mol. Cell. Biol. 16: 3596-3605.
45. Kuo CT, Morrisey EE, Anandappa R, et al. (1997) GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11: 1048-1060.
46. Molkentin JD, Lin Q, Duncan SA, Olson EN. (1997) Requirement of the transcription factor GATA4 for heart tube formation and ventral mor- phogenesis. Genes Dev. 11: 1061-1072.
47. Morrisey EE, Tang Z, Sigrist K, et al. (1998) GATA6 regulates HNF4 and is required for differ- entiation of visceral endoderm in the mouse em- bryo. Genes Dev. 12: 3579-3590.
48. Bielinska M, Wilson DB. (1997) Introduction of yolk sac endoderm in GATA-4 deficient embryoid bodies by retinoic acid. Mech. Devel. 65: 43-54.
49. Sasano H, Shizawa S, Nagura H. (1995) Adreno- cortical cytopathology. Am. J. Clin. Pathol. 104: 161-166.
50. Sharma S, Singh R, Verma K. (1997) Cytomor- phology of adrenocortical carcinoma and compar- ison with renal cell carcinoma. Acta Cytol. 41: 385- 392.
51. Chan JK, Tsang WY. (1995) Endocrine malignan- cies that may mimic benign lesions. Semin. Diagn. Pathol. 12: 45-63.
52. Tartour E, Caillou B, Tenenbaum F, et al. (1993) Immunohistochemical study of adrenocortical carcinoma. Predictive value of the D11 monoclo- nal antibody. Cancer 72: 3296-3303.