Role of the GATA Family of Transcription Factors in Endocrine Development, Function, and Disease

Ontogeny-Reproduction Research Unit (R.S.V., S.M.G.), CHUQ Research Centre, Quebec City, Canada G1V 4G2; Centre de Recherche en Biologie de la Reproduction (R.S.V.), Department of Obstetrics and Gynecology, Faculty of Medicine, Université Laval, Quebec City, Canada G1V 0A6; Department of Obstetrics and Gynecology and Women's Health Research Program (M.A.) and Children's Hospital and Women's Health Research Program (M.H.), Biomedicum Helsinki, University of Helsinki, FI-00014 Helsinki, Finland; and Department of Pediatrics (D.B.W., M.H.), Washington University and St. Louis Children's Hospital, St. Louis, Missouri 63110

Address all correspondence and requests for reprints to: Robert S. Viger, Ph.D., Ontogeny-Reproduction, Room T1-49, CHUQ Research Centre, 2705 Laurier Boulevard, Quebec City, Quebec, Canada G1V 4G2. E-mail: ac.lavalu.luhcrc@regiv.trebor; or Markku Heikinheimo, M.D., Ph.D., Children's Hospital, P.O. Box 22, 00014, University of Helsinki, Helsinki, Finland. E-mail: if.iknisleh@omiehnikieh.ukkram.

Received 2007 Nov 16; Accepted 2007 Dec 21. Copyright © 2008 by The Endocrine Society

Abstract

The WGATAR motif is a common nucleotide sequence found in the transcriptional regulatory regions of numerous genes. In vertebrates, these motifs are bound by one of six factors (GATA1 to GATA6) that constitute the GATA family of transcriptional regulatory proteins. Although originally considered for their roles in hematopoietic cells and the heart, GATA factors are now known to be expressed in a wide variety of tissues where they act as critical regulators of cell-specific gene expression. This includes multiple endocrine organs such as the pituitary, pancreas, adrenals, and especially the gonads. Insights into the functional roles played by GATA factors in adult organ systems have been hampered by the early embryonic lethality associated with the different Gata-null mice. This is now being overcome with the generation of tissue-specific knockout models and other knockdown strategies. These approaches, together with the increasing number of human GATA-related pathologies have greatly broadened the scope of GATA-dependent genes and, importantly, have shown that GATA action is not necessarily limited to early development. This has been particularly evident in endocrine organs where GATA factors appear to contribute to the transcription of multiple hormone-encoding genes. This review provides an overview of the GATA family of transcription factors as they relate to endocrine function and disease.

GATA FACTORS ARE a group of evolutionarily conserved transcriptional regulators that play crucial roles in the development and differentiation of all eukaryotic organisms. In vertebrates, the GATA family comprises six members (named GATA1 to -6) that can be separated into two subgroups based on spatial and temporal expression patterns. GATA1/2/3 are expressed in hematopoietic cell lineages and are essential for erythroid and megakaryocyte differentiation, the proliferation of hematopoietic stem cells, and the development of T lymphocytes (1). Their expression, however, is not limited to hematopoietic cells but is also present in the brain, spinal cord, and inner ear where they play important roles in the development of these structures (2,3,4). In contrast, the GATA4/5/6 proteins are mainly found in tissues of mesodermal and endodermal origin such as the heart, gut, and gonads (5). Disruption of the Gata4 and Gata6 genes in mice results in early embryo lethality due to defects in heart tube formation and extraembryonic endoderm development, respectively (6,7,8,9). Although loss of GATA5 function is not lethal, female Gata5-null mice exhibit genitourinary abnormalities (10). The physiological requirement for GATA function in humans is further supported by the more recent identification of GATA gene mutations associated with human diseases, such as dyserythropoietic anemia and thrombocytopenia (GATA1 mutation); hypoparathyroidism, deafness, and renal dysplasia (HDR syndrome; GATA3 mutation); and congenital heart defects (GATA4 mutation) (11,12,13).

All vertebrate GATA proteins contain a conserved DNA-binding domain composed of two multifunctional zinc fingers involved in DNA-binding and protein-protein interaction with other transcriptional partners and/or cofactors (Fig. 1 ​ 1). ). Because members of the GATA family share a highly conserved DNA-binding domain, there is an apparent redundancy in DNA binding properties to target GATA sequences (14,15). These in vitro properties contrast, however, with their often nonredundant roles in vivo (6,7,9,16,17,18). The specificity of GATA action is governed, in part, through protein-protein interactions with other transcriptional partners. Indeed, there is now an extensive list of ubiquitously expressed or cell-restricted factors that are known to cooperate with GATA factors to control tissue-specific transcription in the hematopoietic system, the heart, and many other tissues (19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45). The activity of GATA factors has also been shown to be modulated by posttranslational modifications such as sumoylation, acetylation, and phosphorylation. The effect of these modifications most often involves enhanced transcriptional activity due to changes in GATA nuclear localization, DNA-binding, protein stability, and/or cofactor recruitment (see Table 1 ​ 1 for references). With respect to endocrine cells, GATA4 phosphorylation has been the best studied posttranslational modification to date (described further below).

An external file that holds a picture, illustration, etc. Object name is zmg0040842240001.jpg

Structure and Homology of the Vertebrate GATA Proteins

All GATA factors share a similar zinc finger DNA-binding domain, a feature that defines this family of transcription factors. The zinc finger (ZnF) region is also involved in protein interactions with cofactors and/or other transcriptional partners. Transactivation domains are located in the N-terminal (N-term) and C-terminal (C-term) regions. The percent homology among the different GATA proteins (as deduced from the mouse sequences) in the N-terminal, C-terminal, and zinc finger domains is indicated. NLS, Nuclear localization signal.

Table 1

Posttranslational Modifications of GATA Factors

ModificationGATA ProteinSpeciesAmino Acid(s)Proposed Role(s)Refs.
SumoylationGATA1MouseK137Transcriptional repression (sumoylation independent)200,201
GATA2MouseNot definedTranscriptional repression (sumoylation independent)202
GATA4MouseK366Transcriptional activation, nuclear localization203,204
AcetylationGATA1Mouse/chickenMultiple: K246, K252, K312 (major sites)Transcriptional activation, protein stability, DNA-binding, increased functional activity205,206
GATA2HumanMultipleTranscriptional activation, DNA-binding207
GATA3HumanK305Transcriptional activation, increase functional activity208
PhosphorylationGATA1MouseS26, S49, S72, S162, S178, S187, S310Protein stability209,210
GATA2MouseMAPK-mediated (Amino acid undefined)Undefined211,212,213
MouseCdk-mediated (S227, T457)Protein stability
HumanPI-3K/Akt-dependent (S401)Nuclear localization, DNA-binding
GATA3MousePKA-mediated (S308)Transcriptional activation214,215,216
HumanMAPK-mediated (undefined)Nuclear localization
GATA4MouseMAPK-mediated (S105) PKA-mediated (S261)Transcriptional activation, DNA-binding, recruitment of transcriptional partners112,114,118

Although originally classified as markers of hematopoietic and cardiac cells, the expression pattern and functions attributed to GATA factors extend well beyond these cell lineages. Interestingly, GATA factors have been described in most, if not all, of the endocrine organs where they are emerging as important regulators of numerous genes coding for hormones or proteins/enzymes involved in hormone biosynthesis. A comprehensive list of putative GATA-regulated genes in endocrine tissues is presented in Table 2 ​ 2. . In the sections that follow, our current understanding of the expression and role of GATA factors in different endocrine organs is presented.

Table 2

Putative GATA Target Genes in Endocrine Tissues

OrganGene NameSymbolGATA ProteinsSupporting EvidenceRefs.
HypothalamusGnRHGnrh1GATA4e46
PituitaryGlycoprotein hormones, α-subunitCgaGATA2/4e, t48
TSH, β-subunitTshbGATA2e, t, c,g21,52,217
GnRH receptorGnrhrUncharacterizede, t135
TestisAnti-Müllerian hormoneAmhGATA1/4/6e, t,g83,88,140,141
Inhibin β-BInhbbGATA1/4e, t218
Doublesex and mab-3-related transcription factor 1Dmrt1GATA4e, t219
FSH receptorFshrGATA1/4e, t, c220,221
Claudin 11Cldn11GATA1/4/6e, t, c,s94
Testis-derived transcriptTesUncharacterizede, t222
Coxsackie- and adenovirus receptor-like membrane proteinClmpGATA1/6e, t, c93
Sex-determining region Y chromosomeSryGATA4g88
SRY-box containing gene 9Sox9GATA4g89,90
Inhibin αInhaGATA1/4e, t, s36,78,105,223
Steroidogenic acute regulatory proteinStarGATA1/4/6e, t, c34,105,134,224
P450 aromatase (via PII promoter)Cyp19a1GATA1/4/6e, t105
17α-HydroxylaseCyp17a1GATA4g88
P450 side chain cleavageCyp11a1GATA4g88
LH receptorLhcgrGATA4e, t196
3β-Hydroxysteroid dehydrogenase type 2HSD3B2GATA4/6e, t, s35
OvarySteroidogenic acute regulatory proteinStarGATA4e, t, c132,133,134
GnRH receptorGNRHRUncharacterizede, t135
AMHAmhGATA4t76
Inhibin αInhaGATA4e, t131
P450 aromataseCyp19a1GATA4e, t, c, s120,136,137
17β-Hydroxysteroid dehydrogenase type 1(HORSE) HSD17B1Uncharacterizede138
P450 side chain cleavageCyp11a1GATA4e, t139
Placenta17β-Hydroxysteroid dehydrogenase type 1HSD17B1GATA3e, t150
3β-Hydroxysteroid dehydrogenase type 1HSD3B1Uncharacterizede, t151
P450 side chain cleavageCyp11a1GATA4e, t139
17α-HydroxylaseCYP17A1GATA4/6e, t38
Glycoprotein hormones, α-subunitCGAGATA2/3e, t48
GnRH receptorGNRHRGATA2/3e, t149
Adenosine deaminaseAdaUncharacterizedt147
Leukemia inhibitory factor receptorLIFRUncharacterizede, t148
Placental lactogen IPrl3d1GATA2/3e, t, g143,144,145
ProliferinPrl2c2GATA2g145
Thyroid and ParathyroidNK2 homeobox 1Nkx2–1Uncharacterizedt65
PTHPthUncharacterizedt56
AdrenalP450 side chain cleavageCYP11A1GATA6t156
17α-hydroxylaseCYP17A1GATA4/6e, t, c38,156
Dehydroepiandrosterone-sulfotransferaseSULT2A1GATA6e, t156,160
3β-Hydroxysteroid dehydrogenaseHSD3B2GATA4/6e, t35
Cytochrome b5CYB5GATA6e, t159
PancreasPaired box gene 4Pax4Uncharacterizede, t225
GlucagonGcgGATA4/6e, t70

e, EMSA and/or DNA footprinting; t, transfection data; c, ChIP; g, genetic (mouse) data; s, small interfering RNA-mediated knockdown data.

GATA FACTORS IN ENDOCRINE TISSUES

Hypothalamus and Pituitary

The expression of Gata4 in migrating GnRH neurons at embryonic d 13.5 (e13.5) but not in mature GnRH-secreting neurons in mouse has led to the hypothesis that GATA factors could play a role in the differentiation of the small population of neurosecretory cells that mediate central nervous system control of reproduction (46). In hypothalamic neuronal GT1-7 cells, GATA4 activates GnRH (Gnrh1) transcription through binding to two GATA motifs arranged in tandem repeat within the GnRH neuron-specific enhancer (46). Because the GATA-binding site is also involved in Gnrh1 gene regulation in the mature hypothalamus, the role for GATA in Gnrh1 expression would not appear to be limited to cell differentiation. The absence of GATA4 binding activity in adult hypothalamus nuclear extracts does not exclude the possibility of the binding of another, yet to be identified, GATA family member or some other transcriptional regulator (46).

The mature pituitary gland orchestrates the homeostasis of key endocrine organs by regulating their function in response to signals coming from brain and the periphery. Among the six adult differentiated hormone-secreting cell types, both thyrotrope and gonadotrope cells express GATA2 from e10.5 onward, whereas GATA3 is transiently expressed during development (47). The secretory products of thyrotrope and gonadotrope cells are heterodimers that share a common α-glycoprotein subunit (αGSU) and a specific β-subunit (FSHβ, LHβ, and thyrotropin-β). Interestingly, genes encoding αGSU (Cga) and thyrotropin-β (Tshb) are GATA2 target genes (21,48,49). Moreover, although a functional GATA binding site has been described in the GnRH receptor promoter in pituitary cells, the GATA factor that binds has not yet been identified (50).

GATA2 is involved in both gonadotrope and thyrotrope terminal cell specification. The detailed analysis of several transgenic mouse models such as expression of a dominant-negative form of GATA2 under the control of Cga (αGSU) regulatory sequences and overexpression of GATA2 directed by the Pou1f1 (Pit1) promoter, have highlighted the complex mechanisms whereby interactions between Pit1 and GATA2 contribute to pituitary cell fate determination (47,51). In gonadotropes where GATA2 is expressed in the absence of Pit1, GATA2 promotes the expression of gonadotrope-specific genes. In thyrotropes, where GATA2 and Pit1 are coexpressed, thyrotrope-specific genes are up-regulated by the binding of both factors to adjacent DNA cis-elements, whereas GATA motif-containing gonadotrope-specific genes are down-regulated. This discrepancy comes from the ability of Pit1 to interact with a critical DNA-binding zinc finger of GATA2, via its homeodomain, thus preventing GATA2 DNA-binding and target gene transactivation (47). Because of the early embryonic lethality of Gata2-null mice resulting from severe anemia consistent with the critical role of GATA2 in both primitive hematopoiesis of the yolk sac and definitive hematopoiesis of the liver (18), the role of GATA2 in later development was obscured. However, a recent study reporting pituitary-specific Gata2 knockout mice has shed further light into its role in pituitary function (52). Loss of GATA2 in the anterior pituitary results in a decrease in gonadotrope and thyrotrope cell numbers at birth and a marked decrease in secretory capacity of these cells in the adult (52). Although Gata2 pituitary knockout mice have defects in the same pituitary cell populations as the dominant-negative GATA2 transgenic mice (47), the extent of the defect is much less severe (52). This difference has been proposed to be due to compensation by GATA3 that is up-regulated in the Gata2 pituitary knockout mice and that would normally be targeted by the GATA2 dominant-negative. Thus, GATA2 has been proposed to have a dual role in pituitary function: the initial specification and/or proliferation of thyrotropes and the later maintenance of hormone expression in mature thyrotropes and gonadotropes (52). Consistent with this hypothesis, GATA2 is found in most αGSU-positive and thyrotropin-secreting human pituitary adenomas (53).

Thyroid and Parathyroid

The parathyroid glands regulate calcium balance in the body through the secretion and action of PTH. Developmentally, parathyroid glands, together with the thyroid, thymus, and ultimobranchial bodies, originate from the pharyngeal region by the complex interactions between endoderm and neural crest derivatives (54). GATA3 expression has been reported in human embryos in the second and third branchial arches, from which the parathyroids derive (55). Although a consensus GATA binding site has been identified in the proximal promoter of the murine PTH (Pth) gene (56), its functional relevance remains unclear, and to date, no other potential GATA3 target genes have been identified in the developing parathyroid. Homozygous Gata3 knockout mice display mid-gestation lethality (16), in part due to a noradrenaline deficiency of the sympathetic nervous system, which can be rescued by pharmacological treatment with catechol intermediates (57). Because of this early embryonic lethality, the generation of parathyroid-conditional Gata3 transgenic mice would therefore be of particular interest to gain insights into the role of GATA3 in adult parathyroid gland physiology. However, the importance of GATA3 in parathyroid function has not come from transgenic mice but rather from study of patients suffering hypoparathyroidism, sensorineural deafness, and renal anomaly (HDR) syndrome. HDR syndrome has been linked to deletions or mutations of GATA3 that affect the integrity of its two zinc fingers (13,58,59,60,61,62,63). Whereas mutations involving the first zinc finger result in either a loss of interaction with Friend of GATA (FOG) transcriptional partners or altered DNA-binding affinity, those involving the GATA3 second zinc finger or adjacent basic amino acids result in a loss of DNA-binding (58). Although no GATA factor has yet been reported to regulate thyroid development, at least two thyroid-expressed genes, Nkx2–5 and Nkx2–1, have been shown to be under GATA control in the heart and lung, respectively (64,65). Therefore, a similar regulation in the thyroid would not be unlikely.

Pancreas

The mammalian pancreas is a mixed exocrine and endocrine gland that is formed from the fusion of dorsal and ventral buds of the foregut endoderm. In recent years, our understanding of pancreatic development has benefited from the identification of essential transcription factors implicated in this process such as pancreatic and duodenal homeobox 1 (PDX1) and neurogenin 3 (NEUROG3) (66,67). The cardiac subfamily of GATA factors are well-known markers of several endodermal derivatives including yolk sac, lung, stomach, gut, and pancreas (5). GATA4 and GATA6 have overlapping expression profiles in the early pancreas (68,69). GATA4 then becomes limited to exocrine tissue and GATA6 to a subpopulation of endocrine cells (68). GATA6, but not GATA4, has also been shown to interact with Nkx2.2 (68), a critical islet transcription factor. In the same study, transgenic mice expressing a GATA4- or GATA6-Engrailed dominant repressor fusion protein in the pancreatic epithelium and in the islet revealed an important role for GATA6, but not GATA4, in early pancreas specification (68). Expression of the GATA6-Engrailed transgene had two distinct phenotypes: agenesis of the pancreas or a reduction in pancreatic tissue. In the latter, all pancreatic cell types were significantly reduced, having few differentiated endocrine cells within enlarged ductal structures (68). In contrast to these findings, others have reported that GATA4 colocalizes with early glucagon-positive cells as early as e12.5 in the mouse and that GATA4 is able to transactivate the glucagon (Gcg) gene promoter in vitro (70). The same study also demonstrated that mutation of the GATA binding element in the Gcg promoter reduces its basal promoter activity in glucagon producing cells by 55%. In another study, the development of the ventral but not dorsal pancreas is completely absent in Gata4 −/− embryos, and Gata6 −/− embryos display a similar but less dramatic phenotype (71). GATA4 has been found to be frequently overexpressed and infrequently methylated in human pancreatic cancers, whereas GATA5 is likely hypermethylated during pancreatic cancer development (72).

Testis

Reproductive tissues are prominent sites of GATA expression. In the gonads, GATA-like proteins are evolutionarily conserved because they are found in a variety species ranging from worms to humans (73,74). In mammals, four GATA factors exhibit gonadal expression (GATA1/2/4/6), and with the exception of GATA2 (75), this expression is exclusive to the somatic cell population (73,74) (Fig. 2 ​ 2). ). The testis contains two major somatic cell types that facilitate spermatogenesis: Sertoli cells, which are the supporting cell lineage, and Leydig cells, the steroidogenic cell lineage. GATA factors have been described in both cell types in several species starting from the early stages of testis differentiation to adult cells (74). Sertoli cells express GATA4 from the onset of their differentiation through to adulthood where its level of expression is considered to be constant regardless of the stage of germ cell maturation (76,77,78,79,80,81,82,83). GATA6 is also present early and is maintained in adult Sertoli cells (78,80,84). Finally, a third GATA factor, GATA1, is expressed by Sertoli cells, but the temporal expression of GATA1 in this lineage differs from that of other GATA members. GATA1 is first detected in the immature animal in Sertoli cells of all the tubules, concomitantly with the first wave of spermatogenesis (83,85). In the adult animal, GATA1 expression in Sertoli cells in stages VI–IX of the seminiferous epithelium cycle coincides with the androgen-sensitive VII–VIII stages and appears to be dependent on the presence of maturing germ cells (79,85). Although both GATA4 and GATA6 have been described in Leydig cells (78,83,84), GATA4 is by far the predominant GATA factor of this cell type, especially in rodents. GATA4 has also been reported in Sertoli and Leydig cell tumors (86).

An external file that holds a picture, illustration, etc. Object name is zmg0040842240002.jpg

Expression of GATA Factors and Their FOG Cofactor Proteins during Gonadal and Adrenal Development

Contribution of cells from adrenogenital primordium and neural crest cells ultimately give rise to the cortex and medulla of the mature adrenal gland. The testis and ovary develop from a bipotential gonad common to both sexes. GATA factors and FOG2 exhibit overlapping expression patterns throughout development of these three endocrine organs (see text for references). #, Expression specific to cells of the adrenal medulla; *, expression in germ cells.

Because of the critical role of GATA factors in hematopoietic and cardiac development and the corresponding early embryonic lethality associated with Gata-null mice, alternative strategies have had to be developed to unravel their function in testis differentiation and physiology. Conditional knockout of GATA1 in Sertoli cells failed to yield a significant phenotype, most likely due to the redundancy of multiple GATA factors in this cell type (87). Other approaches, however, have begun to give clues into the roles of GATA4 in testis cell differentiation. For example, mice homozygous for a GATA4 variant that cannot interact with its cofactor FOG2 (Gata4 ki/ki ) or homozygous for a FOG2-null mutant (Fog2 −/− ) exhibit abnormal testis morphogenesis resulting from a failure of Sertoli cells to differentiate (88). On a molecular level, this appears to be due to a significant reduction in transcription of the testis-determining gene Sry (88), as well as a block in Sox9 expression (89,90). Whether Sry and Sox9 are direct targets of a GATA4/FOG2 complex, however, remains to be demonstrated. The Gata4 ki/ki and Fog2 −/− mice also fail to express anti-Müllerian hormone (Amh) as well as several steroidogenic markers [cytochrome P450 side chain cleavage (Cyp11a1), cytochrome P450 17α-hydroxylase/17,20 lyase (Cyp17a1), and 3β-hydroxysteroid dehydrogenase type 1 (Hsd3b1)], suggesting that steroidogenic cell differentiation and/or gene expression is equally under GATA4/FOG2 control (88). Indeed, a recent study using chimeric mice has shown that that GATA4 is directly required for Leydig cell differentiation in the fetal mouse testis (91). Additional indirect support for a role for GATA factors in Leydig cells has come from the characterization of GATA-like protein 1 [GLP1 (Glp1)] knockout mice (92). Glp1 is expressed in Leydig cells where it appears to act as a repressor of GATA action (92). In males, loss of GLP1 function correlates with an early impairment of germ cell differentiation, suggesting that a complex regulation of GATA proteins in Leydig cells by GLP1, FOG2, and perhaps other modulatory factors, is crucial for their function. Although these mouse studies have illustrated the involvement of GATA factors for early testis cell differentiation, other strategies are still needed for the direct assessment of their function in the mature testis.

Important insights into GATA function in the testis have also come from promoter studies of putative target genes in cell lines or primary cultures of Sertoli and Leydig cells. Most of these GATA gene targets are either correlated with early differentiation (Dmrt1, Sry, and Sox9), or hormone synthesis such as the inhibin subunits (Inha and Inhb), Amh, and multiple proteins and enzymes of the steroidogenic pathway in Leydig cells (see Table 2 ​ 2). ). Interestingly, recent studies have identified genes encoding junctional proteins such as claudin 11 and Coxsackie- and adenovirus receptor-like membrane protein (CLMP) as novel GATA targets in the testis (93,94). Although a scaffolding protein of the blood-testis barrier, Cldn11 expression has been shown to be under germ cell and androgen regulation (95,96), with the androgen regulation apparently mediated by GATA and NFYA proteins (94).

Despite the long list of target genes for GATA factors in testicular cells, we have only begun to scratch the surface as to how these factors are themselves regulated. At the transcriptional level, it is now evident that Gata1 expression in mouse Sertoli cells is dependent on a cell-specific promoter (97,98,99) and is negatively regulated by FSH and cAMP (100). In contrast to GATA1, gonadotropins and/or cAMP agonists have been shown to induce Gata4 and/or Gata6 expression in several gonadal cell lines, including MSC-1 Sertoli cells and mLTC-1 Leydig cells (78,101). The dissection of the Gata4 promoter has begun to reveal some of its transcriptional control mechanisms (102,103). In transgenic mice, the first 5 kb of the 5′-flanking region of the Gata4 gene is sufficient to recapitulate endogenous Gata4 expression in Sertoli cells throughout development (102). Subsequent in vitro studies identified two GC-box motifs and especially one E-box element within this −118-bp region that are crucial for Gata4 transcriptional activity in both testicular and cardiac cells (102,103). In Sertoli cells, members of the upstream transcription factor (USF) family of factors, especially USF2, bind to and activate the Gata4 promoter via this critical E-box motif (102). Gata4 regulatory sequences driving expression in steroidogenic (Leydig) cells of the testis and ovarian cells have yet to be identified.

In both Sertoli and Leydig cells, GATA4 activity is modulated via cooperative interactions with other transcriptional regulators and/or cofactors (73,74,104). These include the nuclear receptors steroidogenic factor 1 (SF-1/NR5A1) and liver receptor homolog 1 (LRH-1/NR5A2), which cooperate with GATA4 to enhance transcription from the Inha, Amh, P450 aromatase (Cyp19a1), and 3β-hydroxysteroid dehydrogenase type 2 (HSD3B2) promoters (27,35,36,105). In testicular cells, GATA1 and GATA4 also interact with FOG1 and FOG2 (25,26,106). Although the FOG proteins do not bind directly to DNA, they function as either enhancers or repressors of GATA transcriptional activity depending on the cell and promoter context being studied (25,26,28,106,107,108,109). On GATA-dependent gonadal promoters, however, in vitro data suggest that FOG proteins play a strictly repressive role (110). In the mouse testis, FOG1 and FOG2 are coexpressed with GATA1/4/6 in Sertoli and Leydig cells (79,110). FOG1 is first detected in Sertoli cells at e15.5; expression is maintained throughout development but eventually becomes restricted to stage VII–XII tubules in the adult (79). FOG2 is expressed earlier than FOG1 and can be detected as early as the undifferentiated gonad stage in both sexes (76). After sex determination, testicular FOG2 expression is rapidly down-regulated by e15.5, leaving only the testis capsule and Leydig cells as significant sites of expression (79). FOG2 then reappears after birth and, like FOG1, eventually becomes restricted to stage VII–XI tubules in the adult testis (79). Different mouse models have clearly identified a role for a GATA4/FOG2 complex in early testis differentiation, acting upstream of Sry and Sox9 (88,89,90). However, it remains to be shown whether these genes and other GATA targets in the testis are directly modulated by either FOG-mediated coactivation or corepression.

GATA4 is also a target for posttranslational modifications such as phosphorylation (34,111,112,113,114,115,116,117). In gonadal cells, activation of the cAMP/protein kinase A (PKA) signaling pathway is a major mechanism for conveying the hormone responsiveness of many genes. In this classic pathway, PKA phosphorylates cAMP response element (CRE)-binding protein (CREB), which is then recruited along with the CREB-binding protein (CBP) coactivator to activate genes possessing CRE in their promoters. In MA-10 Leydig cells, GATA4 is directly phosphorylated by PKA on a specific serine residue (Ser261) (114). Phosphorylation of GATA4 allows for synergistic interactions with different transcription factors including CCAAT/enhancer-binding protein (C/EBPβ) and NR5A2 (34,35) and permits recruitment of the CBP transcriptional coactivator much like CREB (114). This leads to robust stimulation of transcription from gonadal promoters such as Star, Inha, and HSD3B2 (34,35,36,114). PKA is not the sole kinase known to target the GATA4 protein. Indeed, in the heart, phosphorylation of GATA4 at Ser105 via the MAPK signaling pathway plays a crucial role in modulating its GATA4 activity in cardiomyocytes (111,112,118). Although GATA4 Ser105 has been shown to be constitutively phosphorylated in porcine granulosa cells (119), and appears to contribute to the FSH-induced up-regulation of Cyp19a1 expression in these same cells in the rat (120), its significance in testicular cells remains to be demonstrated.

Ovary

The fetal ovary has been shown to express three GATA factors: GATA2, -4, and -6 (Fig. 2 ​ 2). ). In the mouse, GATA2 is specific for germ cells but only for a brief period during fetal development (e11.5–e15.5) (75). GATA4 is first detected in the mouse coelomic epithelium and porcine urogenital ridge of both XX and XY embryos (76,79,81,83). At this time, GATA4 and its cofactor FOG2 label the somatic cell population of the developing gonad (76,79). Subsequently, GATA4 and FOG2 continue to be expressed at high levels in ovarian somatic cells over the entire fetal period (76). GATA4 expression patterns are similar in the fetal rat and porcine ovary; the somatic, pregranulosa or follicular cells are GATA4 positive, whereas germ cells do not express this factor (80,81). In the fetal rat ovary, GATA6 is present in both postmitotic germ cells and the ovarian epithelium (80). In the human fetal ovary, GATA4 is strongly expressed in pregranulosa/stromal cells as early as gestational wk 13 and 14, when primordial follicles are not yet evident (121). By wk 16–33, GATA4 localizes to granulosa cells (121). Like GATA4, GATA6 and FOG2 are expressed in human fetal somatic cells but can also be detected in occasional germ cells (unpublished data). Thus, much like the testis, the abundant expression of GATA proteins, particularly GATA4, in the fetal ovary suggest that they play key roles in the early steps of ovarian differentiation.

In the postnatal ovary, GATA4 and GATA6 are the predominant GATA factors. GATA4 expression is negligible in primordial follicles of both the mouse and human ovary (76,80,101,122,123). GATA4 expression initiates in granulosa cells at the onset of follicle growth (i.e. their initial recruitment) and their transformation into primary follicles (76,101,122,123). As follicular growth proceeds, granulosa cells of healthy preantral and antral follicles express GATA4, where it localizes to cumulus and mural granulosa cells, theca cells, and stromal cells (76,80,101,122,123,124). In luteal glands, GATA4 expression is low or negligible in mice and humans (76,101,123). However, in the rat and pig, GATA4 is expressed in peripheral cells of new luteal glands; expression is down-regulated as these structures regress (80,124). Thus, GATA4 expression is clearly linked to both structural and functional changes in the postnatal ovary of different mammalian species. GATA6 is also expressed in multiple cell types of the mouse, pig, and human postnatal ovary. These include oocytes, luteal glands, and granulosa and theca cells of larger follicles (80,101,122,123,124). Whereas GATA4 expression in granulosa cells is usually down-regulated in follicles undergoing atresia, GATA6 expression in rodents is retained at this stage (80,101). In humans, GATA6 and FOG2 expression patterns are similar; primary follicles have negligible or weak immunoreactivity, but in later follicular stages, GATA6 and FOG2 proteins are detected in most granulosa cells, occasional oocytes, and luteal glands (122,123). In the mouse, only a few granulosa cells of primordial follicles readily express FOG2 protein. In antral follicles, FOG2 expression is often down-regulated, whereas GATA4 expression is retained (76).

Of the endocrine and paracrine regulators of folliculogenesis, gonadotropins as well as TGFβ signaling have been linked to the regulation of ovarian GATA function. Gonadotropins have a fundamental impact on ovarian folliculogenesis, especially FSH, which is required for the development from preantral to antral follicles (125). In FSH-unresponsive ovaries of women with an inactivating mutation in the FSH receptor gene (FSHR), GATA4 is absent in the large pool of primordial follicles as compared with healthy ovaries (126). However, granulosa cells in the few primary follicles seen in these patients do express GATA4 protein, suggesting that either FSH or FSH receptor is not required for the initial activation of GATA4 expression. In Fshr-null mice bearing follicular arrest at the preantral stage, Gata4 expression is, however, impaired in preantral follicles compared with age-matched wild-type animals (127). In vitro studies have further established the link between gonadotropins and the regulation of GATA4 and/or GATA6 action in the ovary. Like testicular cells (114), the transcriptional activity of GATA proteins is likely induced by FSH/cAMP/PKA-mediated phosphorylation in granulosa cells (124). In murine granulosa tumor cells, which resemble normal granulosa cells, FSH and forskolin up-regulate Gata4 mRNA levels (101). Moreover, treatment of immature 3-wk-old mice with pregnant mare serum gonadotropin (PMSG) enhances follicular expression of Gata4 and Gata6 transcripts (101). Similarly, in primary rat granulosa cells, FSH up-regulates GATA4 mRNA and protein levels (120), whereas in immature human granulosa cells, obtained from in vitro maturation of oocytes during fertility treatments, FSH was shown to up-regulate GATA6 mRNA levels (128). In cultured preovulatory human granulosa cells, which luteinize during in vitro culture, treatments with recombinant human chorionic gonadotropin or 8-Br-cAMP had no effect on GATA4 and FOG2 steady-state mRNA levels, whereas GATA6 mRNA levels were modestly but significantly up-regulated (123). The effects on GATA6 rather than GATA4 likely reflect the luteinized phenotype of human granulosa cells.

TGFβ regulation of GATA target genes has been previously demonstrated in both cardiac and T helper cells (129,130). Recent results also support a role for GATA4 in TGFβ signaling in the ovary. In mouse granulosa tumor cells TGFβ is able to up-regulate GATA4 protein levels (131). Furthermore, TGFβ activation of the Inha promoter, a GATA target in both the testis and ovary (Table 2 ​ 2), ), is blocked when the critical GATA binding sites are mutated (131). GATA4 interacts with Smad3 in granulosa tumor cells, suggesting a role for GATA4 in the transcriptional complex activated by the TGFβ-signaling pathway in the ovary. The proposed role for TGFβ and other signaling mechanisms in the regulation of GATA4 activity in gonadal cells is presented in Fig. 3 ​ 3 .

An external file that holds a picture, illustration, etc. Object name is zmg0040842240003.jpg

Involvement of Intracellular Signaling Mechanisms in the Regulation of GATA4 Activity in Gonadal Cells

In response to hormonal signals and/or growth factors, activation of the cAMP/PKA and MAPK signaling pathways leads to phosphorylation of GATA4 on specific serine residues: Ser261, a PKA target, and Ser105, a MAPK target. GATA4 phosphorylation increases its DNA-binding and transactivation properties on different gonadal promoters and enhances the ability of GATA4 to recruit and cooperate with different transcriptional partners such as SF-1 (NR5A1), liver receptor homolog 1 (LRH-1/NR5A2), CCAAT/enhancer-binding protein (C/EBPβ), and CBP. In murine and rat granulosa cells, signaling elicited by gonadotropins or TGFβ family members increases Gata4 transcription. TGFβ-mediated activation of Smad3 and its cooperation with GATA4 also leads to robust transcription of gonadal genes such as Inha (see text for references).

Much like the testis, important insights into the role of GATA factors in ovarian function have come from an analysis of potential target gene promoters in ovarian cells. Because somatic cells of both the ovary and testis express similar GATA proteins, it is not surprising that GATA factors have similar target genes in these two tissues (Table 2 ​ 2). ). To date, proposed GATA-dependent genes in the ovary include Star (132,133,134), GNRHR (135), Inha (131), Cyp19a1 (120,136,137), 17β-hydroxysteroid dehydrogenase type 1 (HSD17B1) (138), Cyp11a1 (139), and Bcl2 (unpublished data). One of the first GATA target genes to be described in gonadal cells was AMH (83). In vitro and in vivo studies have provided strong evidence that the mouse and human AMH promoter, at least in the testis, is an important target for GATA4 (140,141). Although exclusive to the fetal testis, the anti-Müllerian hormone (AMH) is nonetheless produced by granulosa cells of the postnatal ovary where GATA4 is also found (142). Therefore, a similar role for GATA4 in AMH gene regulation in the ovary is likely.

Placenta

GATA factors in placental cells were initially described in studies of αGSU (CGA) transcription in the human trophoblast-derived JEG-3 choriocarcinoma cell line (48). Nuclear extracts of JEG-3 cells were shown to contain GATA2 and GATA3 protein capable of binding to a consensus GATA motif in the proximal CGA promoter (48). Moreover, mutation of the αGSU GATA element dramatically reduced CGA promoter activity in JEG-3 cells, suggesting that GATA2 and/or GATA3 is an important contributor to CGA transcription in the placenta (48). Subsequent Northern blot and in situ hybridization studies revealed that the Gata2 and Gata3 transcripts are abundantly expressed in mouse placenta, localizing predominantly to trophoblast giant cells (143). In these cells, the murine or ovine genes encoding for the placental hormones placental lactogen 1 (PRL3D1) and proliferin (PRL2C2) have been proposed to be regulated by GATA2 and/or GATA3 (143,144,145). Indeed, although mice lacking either the Gata2 or Gata3 gene develop trophoblast giant cells, placental Prl3d1 expression is nonetheless reduced by 50% in these animals (145). For the Prl2c2 gene, a similar 50% reduction in expression was observed in Gata3-null placentas, and this reduction was even more pronounced (5- to 6-fold) in Gata2-null placentas (145). Proliferin is an angiogenic hormone known to stimulate endothelial cell migration and neovascularization of the placenta (146). Consistent with a role for GATA2 in Prl2c2 expression, decidual tissue adjacent to Gata2-null placentas has been reported to exhibit less neovascularization as compared with wild-type tissue (145).

In vitro studies have also identified other placenta-expressed gene promoters as potential targets for regulation by GATA or GATA-like factors. These include the mouse promoters for adenosine deaminase (Ada) (147) and Cyp11a1 (139) and the human promoters for leukemia inhibitory factor receptor (LIFR) (148), GnRH receptor (GNRHR) (149), (HSD17B1) (150), (HSD3B1) (151), and (CYP17A1) (38). Whether these genes are actual targets in vivo, however, awaits further genetic studies.

Adrenal

The adrenal cortex arises from mesodermal precursors that also give rise to gonadal steroidogenic cells (152). The human fetal adrenal cortex consists of a large inner zone, termed the fetal zone, and a thin outer rim of more immature cells, known as the definitive zone (152). After birth, the fetal zone regresses. In late gestation, the definitive zone begins to partition into anatomically and functionally distinct compartments: the zona glomerulosa, zona fasiculata, and zona reticularis (152). In the mouse adrenal, the zona glomerulosa and zona fasciculata are well defined, but there is no discernable zona reticularis. The adrenal cortex of the young mouse contains an additional layer, termed the x-zone, which is adjacent to the medulla and is analogous to the fetal zone of the human adrenal cortex (153).

GATA6 is abundantly expressed in the adrenal cortex throughout development (38,154,155). During human fetal development, GATA6 is evident in both the fetal and definitive zones of the adrenal cortex (154). In the adult, GATA6 is expressed in the zona reticularis and to a lesser extent the zona fasciculata (154,156,157), but its pattern of expression during early childhood and adrenarche has not been established. Based on this distribution and promoter analyses described below, GATA6 is postulated to play a key role in the production of adrenal androgens and possibly also glucocorticoids (156). In the mouse, GATA6 is expressed in both the fetal and adult adrenal cortex (154), suggesting a role beyond merely adrenal androgen biosynthesis. Although evidence for the involvement of GATA6 in adrenal steroidogenesis is accumulating, genetic proof that GATA6 is required for adrenocortical function is lacking.

By comparison, GATA4 has a more restricted pattern of expression during adrenocortical development and therefore is presumed to have a limited role in the function of this tissue (154). During human development, GATA4 mRNA is evident in the fetal zone of the adrenal, but there is only weak expression of this transcript in the adrenal cortex postnatally (154). Similarly, Gata4 is transiently expressed in the mouse adrenal cortex during fetal but not postnatal development (154), a pattern reminiscent of Cyp17a1 in the mouse (158). Given their overlapping patterns of expression, it is possible that GATA4 enhances expression of Cyp17a1 in the fetal adrenal, although there is no direct proof of this. Studies of mice harboring Gata4 loss-of-function mutations suggest that this transcription factor is not required for the differentiation of fetal adrenocortical cells (88,154).

Consistent with its proposed role in the biosynthesis of adrenal androgens and other corticoids, GATA6 has been shown to act in synergy with SF-1 (NR5A1) and other factors to enhance the transcription of CYP11A1 (156), CYP17A1 (38), cytochrome b5 (CYB5) (159), hydroxysteroid sulfotransferase (SULT2A1) (156,160,161), and HSD3B2 (35) in cell lines (see Table 2 ​ 2). ). In the case of the CYP17A1 promoter, GATA6 does not act through a GATA response element but instead through interactions with another transcription factor, specificity protein 1 (SP1) (38). In the case of CYB5, GATA6 enhances promoter activity through interactions with specificity protein 3 (SP3) (159). GATA4 can substitute for GATA6 in transactivation studies of the CYP17A1 promoter (38), suggesting that GATA4 may serve to augment CYP17A1 expression during fetal development. In addition to normal adrenal cortex, GATA4 may regulate gene expression in disease states (described below).

The differentiation, growth, and survival of steroidogenic cells in the adrenal gland are controlled by a diverse array of hormones. Endocrine hormones and paracrine factors traditionally associated with the function of gonadal steroidogenic cells, such as LH, inhibins, and activins, also influence the differentiation, proliferation and function of adrenocortical cells, both in physiological and pathophysiological states. There is no direct evidence that GATA6 is hormonally regulated in the normal adrenal. Steady-state levels of Gata6 mRNA in the adrenal gland do not change markedly in mice treated with ACTH or dexamethasone (157), nor does there appear to be a diurnal variation in GATA6 expression in the adrenal. Administration of dibutyryl cAMP to NCI-H295A human adrenocortical cells has been shown to increase GATA6 mRNA levels, raising the possibility of hormonal regulation of GATA6 in the human adrenal cortex, but the physiological relevance of this finding in vivo remains to be established. GATA6 protein is down-regulated in response to mitogenic signals in some cell types (162,163), but there is no evidence supporting a role for GATA6 during proliferation in the adrenal cortex. Human chorionic gonadotropin up-regulates Gata4 and LH receptor (Lhcgr) expression in a dose-dependent manner (164). GATA4 has a role in the induced, but not basal, regulation of the LH receptor (Lhcgr) promoter in adrenocortical tumor cells. Ectopic expression of GATA4 in the adrenal heralds the appearance of neoplastic cells (164), and this factor may serve to integrate intracellular signals evoked by different hormones.

GATA FACTORS AND ENDOCRINE DISEASE

Gonadal Disorders and Tumorigenesis

To date, reports of gonadal disorders involving GATA4 or its associated cofactors have been scarce. Human mutations in the NR5A1 gene have been identified in individuals with 46,XY sex reversal with gonadal dysgenesis and adrenal insufficiency (165,166,167). Some of the cases involved retained Müllerian duct derivatives, indicating insufficient AMH production during fetal life. One of the heterozygous NR5A1 mutations (G35E) was found to act as a dominant-negative competitor of the synergism between GATA4 and the wild-type NR5A1 protein at the level of the AMH promoter (168). This suggested that proper GATA4/SF-1 is essential for normal AMH transcription in humans (168). Although not targeting the GATA4 gene itself, a recently described de novo t(8;10) chromosomal translocation associated with heart defects and hypergonadotropic hypogonadism has been linked to the gene encoding the GATA4 cofactor FOG2 (169). The resulting truncated FOG2 protein, however, did not appear to impair AMH expression because the patient had no signs of Müllerian duct retention. Thus, unlike the mouse (88), FOG2 might be dispensable for the initiation of testicular development in humans.

In vitro studies have suggested roles for GATA factors in the regulation of steroidogenesis in both gonadal and adrenal cells (104,156). Patients with polycystic ovary syndrome (PCOS) have hyperandrogenism of both ovarian and adrenal origin. The gene expression pattern as well as the response to hormonal stimulation of PCOS theca cells differs significantly from their normal counterparts (170,171). Interestingly, steady-state GATA6 mRNA levels are augmented in PCOS theca cells due to increased gene transcription and stability of the GATA6 transcripts (172). No sequence variations at the GATA6 locus, however, were associated with PCOS patients (172), suggesting that regulatory factors acting at the level of the GATA6 promoter might be differentially expressed in normal vs. PCOS theca cells.

Ovarian carcinomas account for up to 90% of ovarian cancer and typically have a poor prognosis related to a delay in detection. They originate from the surface epithelium and occur in various forms, the most common being serous carcinoma. The serous carcinomas commonly exhibit loss of heterozygosity (LOH) in the locus 8p21–23, which bears the GATA4 gene. In mucinous carcinomas, this LOH is less common (173). Of interest, GATA4 protein expression is lost in practically all serous carcinomas (516 of 528 samples studied) but in only some mucinous carcinomas (26 of 75 samples studied) (173). Moreover, in the mucinous carcinomas, the loss of GATA4 expression correlates inversely with the grade and clinical stage of the tumors (173). In general, LOH correlates with loss of GATA4 expression in the tumors, suggesting that loss of the other allele is enough to completely silence GATA4 expression. More recent findings indicate that the GATA4 gene is often silenced in ovarian carcinomas by histone modifications. In various ovarian cancer cell lines, GATA4 expression is usually negative due to methylation of its gene promoter (174,175). However, none of the analyzed serous carcinoma samples (n =7) exhibited a GATA4 methylated state (175), indicating that LOH is the only reason for the absence of GATA4 expression in the serous subtype. In one study, GATA4 was found to be expressed in most of the studied ovarian carcinomas of different subtypes (43 of 50 samples); in contrast to previous findings (173), 30 of these tumors were of serous type, and only six were GATA4 negative (176). Instead, the authors found a frequent loss of GATA6 protein correlated with a loss of several markers of normal epithelium such as disabled-2, collagen IV, and laminin (176). Taken together, loss of GATA4 and/or GATA6 expression, due to LOH or histone modifications, is likely involved in the dedifferentiation of the ovarian epithelium and ovarian carcinogenesis.

Granulosa cell tumors (GCTs) are rare, accounting for 3–5% of all ovarian cancers (176,177). They are steroidogenically active, typically causing precocious puberty, disturbances in menstrual cycle, and endometrial hyperplasia or cancer (176,177). Although poorly understood, the pathogenesis of GCTs involves defects in the regulation of granulosa cell proliferation (178). Recent findings suggest that GATA4 and GATA6 play a role in granulosa cell tumorigenesis (122). The majority of the analyzed 80 GCTs expressed GATA4 and GATA6 as well as NR5A1 and FOG2 protein at levels comparable with those in normal granulosa cells (122). However, negative or weak GATA6 protein expression is associated with a large tumor size, suggesting that GATA6 has a role in suppressing proliferation (122). This hypothesis is supported by findings in both glomerular and vascular smooth muscle cells, in which GATA6 is able to induce cell cycle arrest by activating and/or stabilizing p21 cip1 (Cdkn1a) expression (163,179).

Thus, GATA4 likely has an opposing role to GATA6 in granulosa cell tumorigenesis. GATA4 expression associates with aggressive behavior in a subgroup of GCTs, suggesting that it may facilitate proliferation and suppress apoptosis in granulosa cells (122). The vast majority of the analyzed GCTs expressed GATA4 at an intermediate or high level corresponding to its expression in normal granulosa cells (122). GATA4 expression associates with clinical stage (122); GATA4 expression is high in 60% of tumors spread outside the ovary (stage Ic and beyond). Furthermore, high GATA4 expression correlates with recurrence risk of the GCT patients, suggesting that GATA4 immunostaining may have value as a prognostic tool in human GCTs. The reasons for high GATA4 expression in aggressive GCTs, however, remain unknown and therefore need further study.

GCTs and adrenal tumors (described below) of both mice and humans exhibit similarities in their GATA expression patterns. Given the common embryonic origin of adrenocortical cells and ovarian granulosa cells, the GATA4/GATA6 ratio may serve as an essential element in the balanced regulation of proliferation vs. apoptosis. Potential target genes involved in these processes in the ovary include cyclin D2 (Ccnd2) and the antiapoptotic genes Bcl2 and Bcl2l1, which have been shown to be under GATA control in other organs (180,181,182). Cyclin D2 is essential for granulosa cell proliferation and plays a role in GCT development (183,184). Our unpublished results show that GATA4 expression is associated with CCND2 and BCL2 expression in GCTs and that the proximal promoters of both genes are activated by GATA4 in granulosa tumor cells.

Adrenal Disorders and Tumorigenesis

Although not normally expressed in the postnatal adrenal cortex, ectopic GATA4 production has been linked to disease states in this tissue (e.g. adrenocortical tumors; see below). Given their expression patterns and target genes, it is reasonable to postulate that GATA factors may be relevant to adrenal insufficiency or other diseases characterized by aberrant steroidogenic gene expression alone or together with other factors. For example, a de novo heterozygous mutation (G35E) in the NR5A1 gene has been shown to be responsible for adrenal insufficiency and pseudohermaphroditism (185). This mutant specifically fails to synergize with GATA factors on the HSD3B2 promoter (35). GATA6 immunoreactivity is significantly diminished in most human adrenocortical adenomas and carcinomas (157,186). This decrease in GATA6 expression may be an important event for the escape of tumor cells from normal control mechanisms, although there is no direct proof of this premise. Consistent with its proposed role in adrenal androgen biosynthesis in the zona reticularis, GATA6 expression is retained in a significant fraction of virilizing adrenocortical tumors, especially carcinomas (187,188). GATA4 mRNA is expressed in only a small proportion of human adenomas and carcinomas (157,187,188). Larger amounts of GATA4 mRNA have been found among those presenting aggressive clinical behavior, although its value as a prognostic marker remains to be established (189).

Gonadectomy causes cells in the mouse adrenal cortex to transform into sex steroid-producing cells that are histologically and functionally similar to gonadal tissue (190). This phenomenon is strain dependent; susceptible strains include CE, DBA/2J, and NU/J. When these mice are gonadectomized, small, basophilic cells, known as A cells, accumulate in the subcapsular region of the adrenal cortex (191). Type A cells express GATA4 and AMH type 2 receptor (AMHR2) but lack expression of steroidogenic markers (164,192). Functionally, A cells resemble stromal cells of the postmenopausal ovary, which can metabolize cholesterol into oxysterols but have limited capacity for the synthesis of sex steroids (193). Later, foci of large, sex steroid-producing cells, termed B cells, appear in the adrenal cortex (194). B cells express Gata4, Nr5a1, Lhcgr, Inha, Amh, estrogen receptor α (Esr1), Cyp17a1, and an ovary-specific alternative splice variant of Cyp19a1 (164,192). Type B cells resemble follicular theca cells from women with PCOS, a condition marked by chronically elevated LH levels (195). Traditional adrenocortical cell markers, such as the melanocortin 2 receptor (MC2R) or enzymes for the synthesis of corticosterone or aldosterone [e.g. steroid 21-hydroxylase (Cyp21a1), P450 11β-hydroxylase (Cyp11b1), and aldosterone synthase (Cyp11b2)] are not expressed in A or B cells (164). Given that many of the genes expressed in A and B cells, such as Lhcgr, Inha, Amh, and Cyp17a1, harbor GATA-binding sites in their promoters, increased GATA4 levels may be postulated to lead to their enhanced expression in these cells.

When Inha promoter-SV40 T-antigen transgenic mice are subject to prepubertal gonadectomy, they develop sex steroid-producing adrenocortical carcinomas by 5–7 months of age (196,197). These adrenocortical carcinomas generally arise in the X-zone, but subcapsular pates of type A and B cells can accompany the X-zone lesions. The molecular markers of adrenocortical neoplasia in these mice, Lhcgr, Inha, Gata4, and Cyp17a1, are similar to those described in gonadectomized DBA/2J mice. GATA6 is also down-regulated during adrenocortical tumorigenesis in these transgenic mice. The phenomenon of gonadectomy-induced adrenocortical neoplasia has also been observed in other small animals, including rats, guinea pigs, hamsters, and ferrets (198). In the ferret adrenal, GATA4 immunoreactivity can be seen in both small and large neoplastic cells and seems to correlate with the degree of nuclear and cytoplasmic atypia (199). GATA4 is particularly abundant in anaplastic myxoid adrenocortical carcinomas but is absent from the spindle cell component of adrenocortical tumors (199). Thus, taken together, it seems likely that GATA4 serves an essential function in gonadectomy-induced adrenocortical tumorigenesis, given its established role in differentiation of gonadal stroma (88).

SUMMARY

Over the past decade, it has become plainly evident that the critical nature of GATA regulatory proteins is not simply a matter of the cardiac and hematopoietic systems. Indeed, the abundant expression of GATA proteins in multiple cell types of various endocrine organs along with their ever-expanding list of target genes is a strong indication that these factors are essential regulators of cell-specific gene expression required for the development, differentiation, and function of endocrine cells. One must be wary, however, that most studies to date have relied heavily on in vitro transactivation assays, DNA-binding analyses, and to a lesser extent chromatin immunoprecipitation. Although steadily increasing, the number of reports relying on genetics and other in vivo models remain limited as are cases of human endocrine-related disorders associated with GATA mutations and/or deregulated GATA function. Although several studies have implicated GATA factors in endocrine tumors, the underlying mechanisms remain to be elucidated. The future challenge is then to continue applying genetic models to validate the bulk of in vitro data that is currently available. For some endocrine organs, this appears to be a somewhat daunting task. This is particularly true of the gonads and adrenal where the overlapping expression of different GATA proteins in various cell types (see Fig. 2 ​ 2) ) and the potential for compensatory mechanisms render individual Gata gene knockout approaches in mice difficult to interpret. A case in point is the Sertoli cell-specific knockout of the Gata1 gene in the adult testis, which produced no overt testicular phenotype (87). In these particular cases, more sophisticated approaches will undoubtedly be required.

Acknowledgments

Drs. Jorma Toppari (University of Turku, Finland) and Timo Otonkoski (University of Helsinki, Finland) are thanked for critical reading of the manuscript.

Footnotes

This work was supported in part by a grant from the Canadian Institutes of Health Research (CIHR) to R.S.V. and by a grant from the Sigrid Juselius Foundation to M.H. and M.A. R.S.V. is titleholder of the Canada Research Chair in Reproduction and Sex Development. S.M.G. was a recipient of a postdoctoral fellowship from the CIHR Institute of Gender and Health. D.B.W. is supported by National Institutes of Health Grant DK075618.

Disclosure Statement: The authors have nothing to disclose.

First Published Online January 3, 2008

Abbreviations: AMH, Anti-Müllerian hormone; CBP, CREB-binding protein; CRE, cAMP response element; CREB, CRE-binding protein; e13.5, embryonic d 13.5; FOG, Friend of GATA; GCT, granulosa cell tumor; GLP1, GATA-like protein 1; αGSU, α-glycoprotein subunit; LOH, loss of heterozygosity; PCOS, polycystic ovary syndrome; PKA, protein kinase A; SF-1, steroidogenic factor 1.

References