GATA-4 is required for sex steroidogenic cell development in the fetal mouse


  • Malgorzata Bielinska,

    1. Department of Pediatrics, Washington University School of Medicine, St. Louis Children's Hospital, St. Louis, Missouri
    Search for more papers by this author
  • Amrita Seehra,

    1. Department of Pediatrics, Washington University School of Medicine, St. Louis Children's Hospital, St. Louis, Missouri
    Search for more papers by this author
  • Jorma Toppari,

    1. Departments of Physiology and Pediatrics, University of Turku, Turku, Finland
    Search for more papers by this author
  • Markku Heikinheimo,

    1. Department of Pediatrics, Washington University School of Medicine, St. Louis Children's Hospital, St. Louis, Missouri
    2. Children's Hospital, Program for Developmental and Reproductive Biology, Biomedicum Helsinki, University of Helsinki, Helsinki, Finland
    3. University Central Hospital and University of Tampere, Tampere, Finland
    Search for more papers by this author
  • David B. Wilson

    Corresponding author
    1. Department of Pediatrics, Washington University School of Medicine, St. Louis Children's Hospital, St. Louis, Missouri
    2. Department of Molecular Biology and Pharmacology, Washington University School of Medicine, St. Louis Children's Hospital, St. Louis, Missouri
    • Box 8208, Washington Univ. School of Medicine, 660 S. Euclid Ave, St. Louis, MO 63110
    Search for more papers by this author


The transcription factor GATA-4 is expressed in Sertoli cells, steroidogenic Leydig cells, and other testicular somatic cells. Previous studies have established that interaction between GATA-4 and its cofactor FOG-2 is necessary for proper Sry expression and all subsequent steps in testicular organogenesis, including testis cord formation and differentiation of both Sertoli and fetal Leydig cells. Since fetal Leydig cell differentiation depends on Sertoli cell–derived factors, it has remained unclear whether GATA-4 has a cell autonomous role in Leydig cell development. We used two experimental systems to explore the role of GATA-4 in the ontogeny of testicular steroidogenic cells. First, chimeric mice were generated by injection of Gata4−/− ES cells into Rosa26 blastocysts. Analysis of the resultant chimeras showed that in developing testis Gata4−/− cells can contribute to fetal germ cells and interstitial fibroblasts but not fetal Leydig cells. Second, wild-type or Gata4−/− ES cells were injected into the flanks of intact or gonadectomized nude mice and the resultant teratomas examined for expression of steroidogenic markers. Wild-type but not Gata4−/− ES cells were capable of differentiating into gonadal-type steroidogenic lineages in teratomas grown in gonadectomized mice. In chimeric teratomas derived from mixtures of GFP-tagged Gata4+/+ ES cells and unlabeled Gata4−/− ES cells, sex steroidogenic cell differentiation was restricted to GFP-expressing cells. Collectively these data suggest that GATA-4 plays an integral role in the development of testicular steroidogenic cells. Developmental Dynamics 236:203–213, 2007. © 2006 Wiley-Liss, Inc.


GATA-4 is one of the transcription factors implicated in the regulation of gonadal development and function. In the male gonad, GATA-4 is expressed in somatic cells, including Sertoli and Leydig cells (Viger et al., 1998; Ketola et al., 1999, 2002). Gata4 knockout mice die by 9.5 dpc secondary to defects in ventral morphogenesis and heart development (Kuo et al., 1997; Molkentin et al., 1997; Narita et al., 1997a, b; Watt et al., 2004), so the role of this transcription factor in reproduction cannot be ascertained from these animals. Analysis of two other genetically-engineered mouse strains, however, has shown that interactions between GATA-4 and its cofactor, FOG-2, are necessary for normal testis development (Tevosian et al., 2002). Fog2−/− mice and Gata4ki/ki mice, which bear a knock-in mutation that abrogates the interaction of GATA-4 with FOG cofactors (Crispino et al., 2001), exhibit identical testicular phenotypes that include decreased Sry expression, impaired cell proliferation, and aberrant differentiation of Sertoli and fetal Leydig cells, manifested as diminished expression Sox-9, MIS, and cholesterol side chain cleavage cytochrome P450 (P450scc, Cyp11a), an enzyme required for testosterone synthesis (Tevosian et al., 2002). Since Sertoli cell–secreted factors influence Leydig cell differentiation and survival, the aforementioned experiments did not answer whether GATA-4 acts cell-autonomously in the development of fetal or adult Leydig cells.

In the male fetus, Leydig cells arise from multiple sources including coelomic epithelium, gonadal ridge mesenchyme, and migrating mesonephric cells (Merchant-Larios and Moreno-Mendoza, 1998; Habert et al., 2001; Yao et al., 2002; O'Shaughnessy et al., 2006). The appearance of fetal Leydig cells is preceded by Sry expression in pre-Sertoli cells, and fetal Leydig cell differentiation depends on Sertoli cell–derived factors such as desert hedgehog (DHH) (Clark et al., 2000; Yao et al., 2002) and platelet-derived growth factor-A (PDGF-A) (Gnessi et al., 2000; Brennan et al., 2003). After birth, fetal Leydig cells are replaced by adult Leydig cells, which originate from undifferentiated mesenchymal stem cells (Mendis-Handagama and Ariyaratne, 2001) or from vascular smooth muscle cells and pericytes (Davidoff et al., 2004). Both fetal and adult type Leydig cells produce the androgens required for masculinization of the male during embryogenesis and for spermatogenesis in the adult animal (Sriraman et al., 2005). Functional maturation of adult Leydig cells is dependent on stimulation by luteinizing hormone (LH). In mice lacking either LH or its cognate receptor, LHR, Leydig cell number is reduced postnatally and circulating androgen levels are low (O'Shaughnessy et al., 1998; Lei et al., 2001; Zhang et al., 2001; Baker and O'Shaughnessy, 2001). In contrast, mouse fetal Leydig cells do not require LH for either their specification or constitutive secretion of androgens, although these cells express LHR and retain the capacity to proliferate and increase sex steroid production in response to stimulation by gonadotropins (Kuopio et al., 1989; O'Shaughnessy et al., 1998; Lei et al., 2001; Zhang et al., 2001; Baker and O'Shaughnessy, 2001).

To probe the significance of GATA-4 in testicular steroidogenic cell development, we used two complementary experimental approaches. First, we examined the ability of XYmath image ES cells to contribute to fetal Leydig cells in chimeric mouse embryos. Second, an ES cell-derived teratoma model was used to investigate the impact of GATA-4 deficiency on steroidogenic cell differentiation in response to endocrine signals. Based on these models, we conclude that GATA-4 is required cell autonomously for proper differentiation of Leydig cells.


GATA-4 Is Required for the Cell Autonomous Differentiation of Fetal Leydig Cells in Chimeric Mice

We examined the pattern of expression of GATA-4 during the fetal mouse testis using immunoperoxidase staining and in situ hybridization. Consistent with published reports (Viger et al., 1998; Ketola et al., 1999, 2002), we detected GATA-4 protein and mRNA in the testis at 13.5 days postcoitum (dpc; Fig. 1A–C) and at subsequent stages of fetal and postnatal development. GATA-4 expression was evident in the interstitial compartment (fetal Leydig cells, fibroblast-like interstitial cells, and peritubular myoid cells) and in Sertoli cells but not in germ cells (Fig. 1A–C).

Figure 1.

Variability in the expression of lacZ among cell types in the fetal testis of Rosa26 mice. A,B: GATA-4 immunoperoxidase staining of paraffin-embedded sections of 13.5-dpc testis. B is a high magnification view of the boxed area in A. Note that GATA-4 is expressed in Sertoli cells (sc), interstitial cells (ic), and peritubular myoid cells (pm) but not in germ cells (gc), mesonephros (me), or blood vessels (v). C,D: Adjacent cryosections of 13.5-dpc testis subjected to GATA-4 in situ hybridization (C) or X-gal staining (D). GATA-4 mRNA is evident in interstitial cells and in cells within the testicular cords, presumed to be Sertoli cells. At 13.5 dpc, cells in the interstitial compartment stain intensely with X-gal, whereas cells in the testicular cords (Sertoli and germ cells) do not. E: X-gal stained cryosection of 17.5-dpc testis. F: Plastic section of X-gal stained 17.5-dpc testis (two fields are shown). In late gestation, germ cells and interstitial cells, including fetal Leydig cells (flc), exhibit intense X-gal staining, whereas Sertoli cells stain only weakly with X-gal. Scale bars = 50 μm (A–E), 5 μm (F).

To monitor the fate of Gata4-deficient cells in the developing testis, XYGata4−/− or wild-type XY ES cells were injected into Rosa26 embryos, which bear a “ubiquitously” expressed β-galactosidase (lacZ) transgene that facilitates lineage tracing. In Rosa26 fetal testis, we observed intense X-gal staining in Gata4-expressing interstitial cells, including fetal Leydig cells (Fig. 1C–F). In agreement with a prior report (Adams and McLaren, 2002), Rosa26 germ cells exhibited weak X-gal staining at 13.5 dpc; the most consistent finding was a perinuclear cyan dot in prospermatogonia (Fig. 1D). By 17.5 dpc, more intense and diffuse cytoplasmic staining was evident in Rosa26 germ cells (Fig. 1E,F). Fetal Sertoli cells stained weakly with X-gal, and eosinophilic, basally-located Sertoli cell nuclei were conspicuous in X-gal stained tissue sections of Rosa26 testis (Fig. 1D–F). This weak expression of lacZ in Sertoli cells precluded an assessment of the contribution of Gata4 null ES cells to this lineage.

X-gal staining of XYGata4−/− ↔ XYRosa26 14.5-18.5 dpc chimeras (n = 7, 25–60% chimerism by GPI isozyme analysis), showed that Gata4-null progenitors retained the capacity to contribute to testicular germ cells (Fig. 2A,B) and to interstitial cells (Fig. 2C–F). In situ hybridization of adjacent tissue sections verified that lacZ-negative interstitial cells lacked expression of Gata4 mRNA and were juxtaposed to testicular cords that contained Gata4-expressing cells, presumably wild-type Sertoli cells. This juxtaposition ensures that fetal Leydig cell progenitors in the interstitium are exposed to Sertoli-derived growth factors (Fig. 2E). Despite proximity to host Sertoli cells, there was little or no expression of the Leydig cell marker P450c17 in the Gata4−/− interstitial cells (Fig. 2F). Imaging software was used to quantify the relative expression of this steroidogenic differentiation marker in ES cell- vs. host-derived testicular tissue in 18.5-dpc chimeras. The ratio of P450c17 mRNA expression in ES cell- vs. host-derived testicular tissue approached unity in XYGata4+/+ ↔ XYRosa26 chimeras (0.90 ± 0.17, n = 3) but was significantly reduced in XYGata4−/− ↔ XYRosa26 chimeras (0.01 ± 0.003, n = 3, P < 0.05, Student's t-test). Reinforcing the premise that GATA-4 is required for proper differentiation of fetal Leydig cells, multi-label immunofluorescence microscopy of testes from late-gestation, highly-chimeric XYmath image ↔ XYRosa26 mice demonstrated that 132/138 (96%) of cells that expressed the Leydig cell marker P450scc also expressed GATA-4 (Fig. 3). Immunofluorescence analysis of non-chimeric, age-matched mice yielded a similar result [105/111 (95%) of cells with P450scc immunoreactivity co-expressed GATA-4].

Figure 2.

Analysis of XYGata4−/− ↔ XYRosa26 chimeric mice. A,B: Gata4−/− cells contribute to germ cells in the fetal testis of 17.5-dpc XYGata4−/− ↔ XYRosa26 chimeric mice. Adjacent cryosections were stained with H&E (A) or X-gal (B). Note the presence of both host-derived (white arrows) and null ES cell–derived (black arrow) germ cells (gc) in the testicular cords. CF: Reduced expression of a fetal Leydig cell differentiation marker in Gata4−/− testicular interstitium of 18.5-dpc XYGata4−/− ↔ XYRosa26 chimeras. Adjacent tissue sections were stained with X-gal (C,D) or subjected to in situ hybridization for GATA-4 (E) or P450c17 (F). Note the uniform expression of GATA-4 mRNA in Sertoli cells in the testicular cords (t), suggesting that these cells are exclusively host-derived. The arrows highlight Gata4−/− interstitial cells (ic) adjacent to host Sertoli cells. P450c17 transcripts are largely restricted to the Gata4+/+ interstitial tissue. Scale bars = 50 μm.

Figure 3.

Multi-label immunofluorescence microscopy of testis from a 17.5-dpc XYGata4−/− ↔ XYRosa26 chimera (60% chimerism by GPI isozyme analysis). A: GATA-4 immunoreactivity (red) is evident in Sertoli cells (sc) and fetal Leydig cells (lc). P450scc immunoreactivity (green) is present in Leydig cells. B: DAPI counterstaining to demonstrate nuclei. The arrow highlights a cell that could be either a null ES cell–derived interstitial cell or a host-derived interstitial cell that does not express GATA-4. Examination of testicular sections from multiple chimeras (n = 4) established that the vast majority of P450scc-positive cells also express GATA-4, supporting the premise that this transcription factor is essential for Leydig cell differentiation or function. Scale bar = 25 μm.

We used transmission electron microscopy (EM) to extend the lineage tracing analysis. Fetal Leydig cells can be readily distinguished from other interstitial cells types by virtue of their distinctive ultrastructural features, including round osmophilic lipid droplets and numerous mitochondria (Fig. 4A) (Merchant-Larios and Moreno-Mendoza, 1998). When Rosa26 testes were stained with X-gal, electron-dense crystalloids accumulated in the cytoplasm of fetal Leydig cells (Fig. 4A), fibroblast-like mesenchymal cells (Fig. 4B), and other interstitial cell types (Merchant-Larios and Moreno-Mendoza, 1998). Ultrastructural analysis of control Rosa26 testes demonstrated that, under the staining conditions used, the vast majority of fetal Leydig cells and interstitial fibroblasts contained electron dense crystalloids (Fig. 4C). Using the presence of crystalloids as a means of lineage tracing, we examined the ability of GATA-4 deficient cells to contribute to fetal Leydig cells and interstitial fibroblasts in testes from XYGata4−/− ↔ XYRosa26 chimeric mice (n = 4, 30–60% chimerism, 18.5 dpc). Approximately half of the interstitial fibroblasts in these chimeras lacked crystalloids, confirming that XYGata4−/− cells can differentiate into this lineage (Fig. 4C). In contrast, morphologically recognizable fetal Leydig cells were derived exclusively from host (crystalloid-positive) cells (Fig. 4C). On the basis of these light and electron microscopic studies, we conclude that XYGata4−/− ES cells exhibit a cell-autonomous defect in fetal Leydig cell differentiation.

Figure 4.

EM-based lineage tracing in X-gal stained testis from chimeric mice. Electron-dense crystalloids (arrows), products of the hydrolysis of X-gal by β-galactosidase, were detected in Rosa26 fetal Leydig cells (A) and interstitial fibroblasts (B). No crystalloids were evident in interstitial cells of X-gal stained wild-type testis (data not shown). The bar graphs show the percentage of Leydig cells and interstitial fibroblasts containing at least one electron-dense crystalloid in 18.5-dpc Rosa26 mice (C) or 18.5-dpc XYGata4−/− ↔ XYRosa26 mice (35–50% chimerism by GPI isozyme analysis). Each bar represents results mean ± SEM from a total of 100–200 cells from 4 testes. The asterisk indicates a statistically significant difference (P < 0.05, Student's t-test). Note that Gata4−/− cells can contribute to fibroblasts, as evidenced by the decrease in the percentage of crystalloid-positive cells. Scale bars = 1 μm.

GATA-4 Participates in the Gonadectomy-Induced Differentiation of Sex Steroidogenic Cells in ES Cell-Derived Teratomas

Gonadectomy can alter the fate of steroidogenic progenitors in the adrenals of certain inbred strains of laboratory mice, including NU/J nude mice. In response to elevated serum gonadotropin levels that accompany ovariectomy or orchiectomy, adrenocortical cells undergo metaplasia into sex steroid-producing gonadal-like stroma. Steroidogenic and non-steroidogenic cells in these metaplastic lesions express GATA-4 (Bielinska et al., 2003, 2005, 2006; Johnsen et al., 2006), which is normally expressed only in fetal and not adult adrenal (Kiiveri et al., 2002). We postulated that the hormonal changes associated with gonadectomy might induce sex steroidogenic lineage differentiation in ES cell-derived teratomas grown in NU/J nude mice and that this process would be dependent on GATA-4. To test this hypothesis, we injected XYGata4+/+ or XYGata4−/− ES cells into the flanks of intact or gonadectomized nude mice. Both male and female mice were used as hosts to track changes that might result from gender-dependent differences in the post-gonadectomy hormonal environment (e.g., elevated LH levels) (Bielinska et al., 2003, 2005). All of the inoculated mice developed flank tumors, which were harvested after 17–21 days when 0.5–1 cm in diameter. We used semi-quantitative RT-PCR to determine whether teratomas derived from wild-type or Gata4−/− ES cells expressed gonadal steroidogenic markers. In addition to steroidogenic enzymes, we examined expression of transcription factors and signaling molecules known from previous knockout studies to be crucial for the differentiation and development of Sertoli or Leydig cells.

As expected, GATA-4 mRNA was expressed in teratomas derived from Gata4+/+ ES cells (Fig. 5, lanes 3, 5, 8, 10) but not Gata4−/− ES cells (Fig. 5, lanes 2, 4, 7, 9), and sex steroidogenic markers were detected in Gata4-positive teratomas grown in gonadectomized but not intact mice. Among the markers expressed, irrespective of host gender, were the transcription factors SF-1 and WT-1, the hormone receptors LHR and MISRII (Müllerian inhibitory substance receptor type II), and the steroidogenic enzymes P450scc, P450c17, and HSD3β1 (3β-hydroxysteroid dehydrogenase/Δ54-isomerase type I) (Fig. 5, lanes 5, 10). Each of these factors is known to be present in Leydig cells, suggesting that gonadectomy induces differentiation of this lineage in a Gata4-dependent but host environment–independent fashion.

Figure 5.

RT-PCR analysis of steroidogenic cell markers in teratomas. Intact or gonadectomized (GDX) nude mice were inoculated subcutaneously with Gata4+/+ or Gata4−/− ES cells. After 3 weeks, teratomas were harvested and subjected to semi-quantitative RT-PCR analysis for a series of gonadal markers. Ribosomal protein L19 was included as a loading control. “Blank” indicates no input RNA. Note that steroidogenic markers are co-expressed with GATA-4 in Gata4+/+ teratomas grown in gonadectomized mice. The variability of steroidogenic marker expression among tumor specimens is discussed in the text.

We also examined a series of markers known to be expressed in both gonadal and extragonadal lineages (FOG-2, SOX8, and DHH). mRNA for the GATA-4 cofactor, FOG-2, was detected in teratomas derived from Gata4+/+ and Gata4−/− ES cells (Fig. 5, lanes 2–5, 7–10). Gonadectomy appeared to down-regulate levels of this message in both male and female hosts (Fig. 5, lanes 4, 5, 9, 10). mRNA for SOX8 also was detected in teratomas derived from both Gata4+/+ and Gata4−/− ES cells (Fig. 5, lanes 2–5, 7–10). Gonadectomy reduced the amount of SOX8 mRNA in Gata4+/+ teratomas in male hosts (Fig. 5, lane 5). DHH, another Sertoli cell marker indispensable for the induction of fetal Leydig cells, was expressed equally in all Gata4+/+ and Gata4−/− teratomas, in a gonadectomy- and host-independent manner (Fig. 5, lanes 2–5, 7–10).

Expression of the Leydig cell marker MISRII was Gata4- and gonadectomy-dependent in both male and female hosts (Fig. 5, lane 5). On the other hand, host gender did appear to influence the expression of its ligand MIS (Fig. 5, lanes 7–10). MIS mRNA was observed in Gata4+/+ and Gata4−/− teratomas grown in female hosts (Fig. 5, lanes 7, 8), and gonadectomy seemed to upregulate expression of this transcript (Fig. 5, lanes 9, 10). It is unclear whether MIS in the teratomas was confined to Sertoli-like cells or was expressed in extragonadal cell types.

Gata4+/+ but not Gata4−/− teratomas expressed vanin, a membrane-associated pantheinase, which has an important antioxidant function in the developing fetal gonads, especially the testis (Wilson et al., 2004; Berruyer et al., 2004). In Gata4+/+ teratomas grown in male hosts, the appearance of vanin was gonadectomy-dependent (Fig. 5, lane 5), whereas in female hosts expression of this transcript was gonadectomy-independent (Fig. 5, lanes 8, 10).

There was variability in the extent of steroidogenic marker expression among the Gata4-expressing tumor specimens grown in gonadectomized nude mice, which may reflect tumor heterogeneity. To determine the reproducibility of gonadal steroidogenic cell differentiation in this model, we performed semi-quantitative RT-PCR for P450c17, P450scc, HSD3β1, and WT-1 transcripts in a series of teratomas harvested from gonadectomized female mice. P450c17 mRNA was detected in 5 of 8 (63%) Gata4+/+ teratomas and in 0 of 4 Gata4−/− teratomas (P < 0.05, two populations proportion test). P450scc mRNA was detected in 7 of 8 (87%) Gata4+/+ teratomas and in 0 of 4 Gata4−/− teratomas (P < 0.05). HSD3β1 mRNA was seen in 8 of 8 (100%) Gata4+/+ teratomas and in 0 of 4 Gata4−/− teratomas (P < 0.05). WT-1 mRNA was seen in 7 of 8 (87%) of Gata4+/+ teratomas and in 0 of 4 Gata4−/− teratomas (P < 0.05).

Immunohistochemical analysis of Gata4+/+ teratomas showed that GATA-4 co-localized with inhibin-α, a protein normally produced in gonadal somatic cells (Fig. 6A,B). Additionally, GATA-4 immunoreactivity overlapped with that of the sex steroidogenic marker, LHR (Fig. 6C,D). Neither inhibin-α nor LHR immunoreactivity was observed in Gata4−/− teratomas (data not shown). Multi-label immunofluorescence staining demonstrated co-expression of GATA-4 and P450scc in a subset of cells in the Gata4+/+ teratomas (Fig. 7A,B). In addition, we generated “chimeric” teratomas derived from mixtures of GFP-tagged XYGata4+/+ and unlabeled XYmath image ES cells. Double-label immunofluorescence showed that P450scc was expressed exclusively in GFP-positive cells (Fig. 7C–E), supporting a cell-autonomous role for GATA-4 in the differentiation of steroidogenic lineages.

Figure 6.

Immunoperoxidase staining of teratomas harvested from gonadectomized mice. Teratomas, prepared by subcutaneous injection of wild-type ES cells into the flank of gonadectomized mice, were harvested and then fixed; paraffin-embedded tissue sections were subjected to immunoperoxidase staining for GATA-4 (A,C), inhibin-α (B), and LHR (D). Adjacent tissue sections are shown in A,B and C,D. Note that GATA-4 expression overlaps with inhibin-α and LHR expression. Scale bars = 50 μm.

Figure 7.

Multi-label fluorescence microscopy of teratomas harvested from gonadectomized mice. A,B: Teratomas, prepared by subcutaneous injection of wild-type ES cells into the flank of gonadectomized mice, were harvested, cryosectioned, fixed, and then stained with goat anti-GATA-4 and rabbit anti-P450scc primary antibodies, followed by anti-goat Cy3 and anti-rabbit FITC secondary antibodies. Nuclei were visualized with DAPI. The white arrowheads designate a presumptive Leydig cell (lc) that co-expresses GATA-4 and P450scc. Adjacent cells that express only GATA-4 may represent Sertoli-like cells. The white arrows highlight a cell that is negative for both GATA-4 and P450scc. CE: Teratomas, prepared by subcutaneous injection of a 2:1 mixture of GFP-labeled Gata4+/+ ES cells and unlabeled Gata4−/− ES cells, were harvested, cryosectioned, fixed, and then stained with rabbit anti-P450scc followed by anti-rabbit Cy3 secondary and DAPI. Yellow arrows highlight a group of Gata4+/+ cells that co-express GFP and the steroidogenic marker P450scc. Yellow arrowheads show a group of Gata4+/+ cells that do not express this steroidogenic marker. The asterisk highlights Gata4−/− cells, which lack expression of P450scc. Examination of sections (n = 4) from “chimeric” teratomas documented that P450scc-positive cells were always derived from Gata4+/+ (GFP-positive) and not Gata4−/− cells. Scale bars = 25 μm.

To distinguish the impact of gonadotropin elevation from other gonadectomy-induced hormonal changes, a subset of the gonadectomized female mice inoculated with Gata4+/+ teratomas ES cells were infused continuously with the luteinizing hormone releasing hormone (LHRH) agonist [D-Trp6] LHRH, which suppressed LH secretion by the pituitary to baseline (non-gonadectomized) levels (Fig. 8, left panel). The LHRH agonist altered the expression profile of steroidogenic markers in the teratomas. While GATA-4 expression remained unchanged, the expression of LHR, WT-1, and to a lesser extent SF-1 was attenuated in teratomas grown in the presence of [D-Trp6] LHRH (Fig. 8, right panel), suggesting that gonadotropin elevation is required for the full expression of these transcripts. In contrast, the steady-state levels of transcripts encoding P450scc, P450c17, and HSD3β1 were not altered by LHRH agonist treatment, implying that steroidogenic enzyme expression in this model does not require continuous LH signaling. This lack of LH-dependence is reminiscent of steroidogenesis in fetal Leydig cells. Interestingly, [D-Trp6] LHRH treatment induced expression of first exon variant of P450c19 that is specific for gonadal steroidogenic cells (Honda et al., 1996). This might reflect de-repression of the P450c19 gene in response to the decrease in WT-1 expression (Gurates et al., 2003). The FSH receptor (FSHR), which is normally expressed in Sertoli cells, was also induced in teratomas in response to [D-Trp6] LHRH treatment. It has been proposed that, in the absence of LH, FSH can stimulate steroidogenesis in Leydig cells (Baker et al., 2003). Thus, the induction of P450c19 and FSHR expression may reflect a compensatory response to low circulating LH levels.

Figure 8.

Left: Effect of gonadectomy and LHRH analog administration on serum LH levels in nude mice bearing ES cell–derived teratomas. XYGata4+/+ or XYGata4−/− ES cells were injected subcutaneously into the flanks of intact or gonadectomized NU/J nude mice. LHRH analog was administered continuously via osmotic pump to a subset of the female gonadectomized animals. Serum was collected after 4 weeks. Values are the mean ± SEM of 3 to 5 mice. ND, not done. Right: RT-PCR analysis of steroidogenic markers in Gata4+/+ teratomas grown in ovariectomized nude mice in the absence or presence of LHRH analog. Results are representative of two experiments.

In summary, the hormonal changes elicited by gonadectomy can promote differentiation of sex steroidogenic cells from ES cell-derived precursors, and have shown that GATA-4 is a key participant in the developmental cascade that guides this differentiation.


Studies of Gata4 knock-in (Gata4ki/ki) mice have established that GATA-4 is required for Sertoli cell development and the subsequent steps in testicular organogenesis (Tevosian et al., 2002). Since Leydig cell differentiation depends on Sertoli cell–derived factors, it has remained unclear whether GATA-4 has a cell-autonomous role in fetal Leydig cell development. In vitro studies support the premise that GATA-4 plays a role in the differentiation and/or steroidogenic function of gonadal somatic cells, including fetal and adult Leydig cells. Co-transfection experiments have shown that GATA-4, working alone on in concert with other transcriptional co-activators, can drive expression of numerous genes involved in gonadal somatic cell function, including MIS (Tremblay and Viger, 1999, 2003), StAR (Hiroi et al., 2004), P450c17 (Fluck and Miller, 2004), aromatase (Tremblay and Viger, 2001), inhibin-α and -β subunits genes (Feng et al., 2000; Tremblay and Viger, 2001), LHR (Rahman et al., 2004), and HSD3β2 (Martin et al., 2005). To explore the role of GATA-4 in testicular steroidogenic cell development, we used two complementary models: chimeric mice and ES cell–derived teratomas.

A distinctive feature of chimeras derived from null ES cells is that the mutant cells are provided with all possible lineage options normally available during development; consequently, the cells are subject to a test of the full range of lineage potency (Tam and Rossant, 2003). Such an approach is advantageous in the study of fetal Leydig cells, because this lineage could potentially arise from multiple sources including coelomic epithelial cells, migrating mesonephric cells, or genital ridge mesenchyme (Merchant-Larios and Moreno-Mendoza, 1998; Mendis-Handagama and Ariyaratne, 2001; Yao et al., 2002). In contrast, a conditional knockout strategy relying on stage- and lineage-specific Cre expression might not be effective at disrupting differentiation from these diverse progenitor types. Using chimeric mouse analysis, we found that XYGata4−/− cells retain the capacity to differentiate into testicular interstitial fibroblasts but exhibit a cell autonomous defect in fetal Leydig cell differentiation. Due to technical limitations, it was not possible to apply the chimeric mouse model to the study of postnatal Leydig cell differentiation.

We also developed a new teratoma model to independently compare the capacity of Gata4+/+ and Gata4−/− ES cells to differentiate into sex steroidogenic lineages in response to endocrine signals. To provide a hormonal milieu conducive to sex steroidogenic cell differentiation (i.e., elevated gonadotropin levels), Gata4+/+ and Gata4−/− teratomas were grown in gonadectomized nude mice. Previous studies have shown that in NU/J nude mice, gonadectomy causes adrenocortical cells to transform into sex steroidogenic cells (Bielinska et al., 2006). Similarly, teratomas grown in gonadectomized hosts exhibited sex steroidogenic differentiation. Specifically, gonadal-like stroma expressing SF-1, WT-1, LHR, vanin, MISIIR, and the steroidogenic enzymes P450scc, P450c17, and HSD3β1 was evident in teratomas derived from wild-type but not Gata4−/− ES cells, and gonadectomy at the time of injection was a prerequisite for expression of these sex steroidogenic differentiation markers. Gonadectomy did not induce expression of gonadal lineage markers in Gata4−/− teratomas. This block in steroidogenic cell differentiation was cell autonomous, as shown by analysis of “chimeric” teratomas derived from a mixture of Gata4−/− and GFP-labeled Gata4+/+ ES cells. Thus, analysis of gonadal markers in teratomas derived from Gata4+/+ vs. Gata4−/− ES cells demonstrates that the presence of GATA-4 is necessary for the process of sex steroidogenic cell differentiation. Our studies underscore the inherent strengths of the teratoma model: it is genetically tractable, and steroidogenic differentiation can be assessed in hormonal milieus representative of different physiologic and pathologic states.

Mouse fetal Leydig cells do not require LH for either their specification or constitutive secretion of androgens (Kuopio et al., 1989; O'Shaughnessy et al., 1998; Lei et al., 2001; Zhang et al., 2001; Baker and O'Shaughnessy, 2001). The persistent expression of steroidogenic enzymes in teratomas from gonadectomized mice following [D-Trp6] LHRH administration suggests that this experimental system is more reflective of fetal than adult-type Leydig cell differentiation. Reinforcing the importance of GATA-4 in sex steroidogenic cell differentiation, it has recently been shown that rat Leydig stem cells expressing GATA-4 but not LHR or steroidogenic enzymes were able to colonize, propagate, and differentiate when transplanted into the interstitium of rats (Ge et al., 2006).

Although GATA-4 appears to play an integral role in the ontogeny of testicular steroidogenic cells in the fetal testis, this transcription factor is not required for adrenal steroidogenesis. GATA-4 is not expressed in adrenocortical cells of adult mice, and Gata4−/− cells can contribute to the adrenal cortex of chimeric mice (Kiiveri et al., 2002). The adult mouse adrenal gland lacks expression of P450c17, a potential GATA-4 target gene, so this tissue cannot synthesize androgenic steroids (Keeney et al., 1995). Interestingly, gonadectomy-induced metaplasia of adrenocortical cells into sex steroid-producing gonadal stroma, a phenomenon observed in certain inbred mouse strains, is accompanied by the induction of GATA-4 expression, suggesting that this transcription factor, through its effect on target genes, plays a key role in the trans-differentiation (Bielinska et al., 2006).

Expression of genes required for steroid biosynthesis is reduced in the testes but not adrenal glands of phthalate-treated fetal rats, a model of human testicular dysgenesis syndrome (TDS) (Thompson et al., 2004, 2005). The mechanisms underlying this selective repression of testicular steroidogenesis are unknown. Since GATA-4 is required for steroidogenesis in the testis but not adrenal gland of rodents, dysregulation of this transcription factor could in theory account for the selective repression of gonadal steroidogenesis that accompanies TDS. Delineating the molecular pathways by which GATA-4 and related factors regulate fetal and adult sex steroidogenic cell differentiation may shed light on the pathogenesis of TDS and other gonadal developmental defects.


Experimental Animals and Cell Lines

All of the experimental procedures were approved by institutional committees for laboratory animal care and were conducted in accordance with NIH Guidelines for the Care and Use of Laboratory Animals and with the European Union Normative for the care and use of experimental animals. NU/J nude, C57BL/6J, and Rosa26 (C57BL/6J, Gpi-1b) (Friedrich and Soriano, 1991) were purchased from Jackson Laboratory (Bar Harbor, ME). The XYGata4−/− ES cells (2 independently selected clones) and the wild-type parental line (CCE, 129/Sv//Ev, Gpi-1c) have been described previously (Soudais et al., 1995; Kuo et al., 1997). GFP-tagged XY ES cells were generously provided by Drs. Tim Ley and Tim Graubert (Washington University, St. Louis, MO). This cell line was derived from transgenic strain C57BL/6-Tg(Actb-EGFP)OsbY01 (Okabe et al., 1997), which bears an enhanced green fluorescent protein cDNA under the control of a chicken β-actin promoter and cytomegalovirus enhancer.

Generation of Chimeric Mice

Male C57BL/6J mice homozygous for the Rosa26 transgene (Friedrich and Soriano, 1991) were mated to supraovulated C57BL/6J females. Rosa26 embryos were harvested at 2.5 dpc and injected with ES cells as described (Narita et al., 1997a). Injected embryos were transferred to pseudopregnant Swiss-Webster females, and the resultant chimeric embryos were harvested between 12.5 dpc and 18.5 dpc. Embryo morphology served as a reference for staging the embryos. Chimeras were initially identified by GPI-1 isoenzyme analysis of tail tissue (Narita et al., 1997a). XIST RT-PCR analysis of tail or hindlimb tissue was used to distinguish chimeras derived from XX versus XY host blastocysts (Natoli et al., 2004).

X-Gal Staining

Frozen sections (10 μm) were prepared by embedding mouse embryos in OCT (Tissue-Tek, Torrance, CA). Sections were then fixed with 0.2% glutaraldehyde for 10 min, permeabilized with 100 mM potassium phosphate, pH 7.4, 0.02% NP-40 and 0.01% sodium deoxycholate for 5–15 min, and then incubated in 0.5 mg/ml X-gal (Promega, Madison, WI) with 10 mM K3[Fe(CN)6], 10 mM K4[Fe(CN)6], 100 mM potassium phosphate, pH 7.4, 0.02% NP-40 and 0.01% sodium deoxycholate at 30°C overnight. X-gal stained sections were counterstained with eosin (Narita et al., 1997a). Alternatively, late gestation mouse embryos were harvested and fixed by intracardiac perfusion with 4% paraformaldehyde and 0.2% glutaraldehyde in PBS; testes were then harvested, transected, and incubated in 4% paraformaldehyde and 0.2% glutaraldehyde in PBS for an additional 15 min at room temperature. The tissue was then rinsed with PBS, and stained with X-gal for 36 hr at 30°C, using the permeabilization and staining solutions described above. After staining, the organs were rinsed with PBS, post-fixed with 2% paraformaldehyde and 2% glutaraldehyde, and processed for transmission EM as described (Soudais et al., 1995).

ES Cell-Derived Teratomas

The flanks of weanling female or male NU/J nude mice were injected subcutaneously with 1.5 × 106 wild-type or XYGata4−/− ES cells. Alternatively, mice were injected with a mixture of 1 × 106 GFP-tagged wild-type XY ES cells and 0.5 × 106 XYGata4−/− ES cells. In some cases, mice were gonadectomized (Bielinska et al., 2005) at the time of ES cell inoculation. Where indicated, osmotic pumps (Alzet, Cupertino, CA) releasing [D-Trp6] LHRH (Sigma, St. Louis, MO) at a rate of 25 μg/day were implanted subcutaneously in some of the female mice at the time of gonadectomy. After 3 weeks, the mice were killed by CO2 asphyxiation and blood, adrenals, and teratomas were harvested (Bielinska et al., 2005). Tumors were either frozen in OCT for cryosectioning, fixed with 4% paraformaldehyde in PBS for paraffin tissue sections, or used to isolate mRNA. Serum LH concentrations were measured as described previously (Bielinska et al., 2003).

Semi-Quantitative RT-PCR

Tumor fragments were homogenized in TRIzol (Invitrogen, Carlsbad, CA). Purified RNA (200 ng) was subjected to RT-PCR using a TITANIUM™ one-step kit (BD Biosciences, Palo Alto, CA), oligo(dT) primers for the reverse transcriptase reaction, and the PCR primers listed in Table 1. Agarose gel electrophoresis (1.2%) in the presence of ethidium bromide demonstrated a single band of the expected size for each of the PCR primer pairs.

Table 1. PCR Primers for Transcripts Measured by Semi-quantitative RT-PCR
GeneFunctionPrimers (5′ → 3′)Size (bp)Reference
L19HousekeepingF: gaaatcgccaatgccaactc405(Kero et al., 2000)
  R: tcttagacctgcgagcctca  
MISSertoli cell functionF: gcagttgctagtcctacatc352(Salmon et al., 2005)
  R: tcatccgcgtgaaacagcg  
VaninSertoli cell functionF: cggtgcaggagagactcagc381(Berruyer et al., 2004)
  R: gccaaatgaggaaggacgtc  
DHHSertoli cell functionF: ccatcgcggtgatgaacatg237(Takabatake et al., 1997)
  R: ttatcagctttgaccgatac  
FOG-2Sertoli cell differentiationF: gcggatcctaactcgcatcaggtttccagcct1136(Svensson et al., 2000)
  R: ccgctcgagaacaggctttccgtttattttgtc  
SOX8Sertoli cell differentiationF: tcctactcgcactccggg357(Salmon et al., 2005)
  R: gccgacgggatgaatgga  
FSHRSertoli cell differentiationF: aaggtctattccctgcccaaccat274(Looyenga and Hammer, 2006)
R: ctgggttcatcatctacgagagag
SF-1Steroidogenesis and Sertoli cell functionF: acaagcattacacgtgcacc478(Babu et al., 2002)
R: gctggcatagggctctggatac
WT-1Steroidogenesis and Sertoli cell functionF: cagagagcaaggcaccag221(Salmon et al., 2005)
R: taagagcccagtgctagtg
MISRIISteroidogenesis and Sertoli cell functionF: tcatgcagtcgtctgggccg400(Mishina et al., 1996)
P450sccSteroidogenesisF: agaagctgggcactttggagtcag545(Arensburg et al., 1999)
  R: tcacatcccaggcagctgcatggt  
LHRSteroidogenesisF: ctctcacctatctccctgtc702(Rahman et al., 2004)
  R: tctttcttcggcaaattcctg  
HSD3β1SteroidogenesisF ccaaggacagttctactacatc276(Bain et al., 1991)
  R ctcatagcccagatctcgct  
P450c17SteroidogenesisF: ccagatggtgactctaggcctcttgtc300(Bielinska et al., 2003)
  R: ggtctgtatggtagtcagtatcg  
P450c19SteroidogenesisF: acagcattgtgattgtccctct344(Honda et al., 1996)
  R: catcttgcgctatttggcctc  

Single and Multi-Label Immunostaining

Tissue sections were processed for immunoperoxidase staining or immunofluorescence microscopy as described elsewhere (Jacobsen et al., 2002; Bielinska et al., 2005). The following primary antibodies were employed: (1) goat anti-mouse GATA-4 IgG (sc-1237, Santa Cruz Biotechnology, Inc., Santa Cruz, CA), 1:200 dilution; (2) mouse anti-human LHR (CRL-2685) hybridoma conditioned media (CRL-2685, ATCC, Manassas, VA), 1:100 dilution; (3) rabbit P450scc, Research Diagnostics, Concord, MA) 1:200 dilution; (4) rabbit anti-SF1 (Affinity Bioreagents Inc, Golden, CO), 1:1,000 dilution. Secondary antibodies employed for immunoperoxidase staining were: donkey anti-goat biotinylated IgG (Jackson Immunoresearch, West Grove, PA) 1:1,000 dilution; donkey anti-mouse biotinylated IgG (Jackson Immunoresearch), 1:2,000 dilution; goat anti-rabbit biotinylated IgG (NEF-813, NEN Life Science, Boston, MA), 1:2,000 dilution. The avidin-biotin immunoperoxidase system (Vectastain Elite ABC Kit, Vector Laboratories, Inc., Burlingame, CA) and diaminobenzidine (Sigma-Aldrich Corp., St. Louis, MO) were used to visualize the bound antibody; slides were then counterstained with 100% hematoxylin. Secondary antibodies used for immunofluorescence microscopy were goat anti-rabbit CY3 (Jackson Immunoresearch Lab), 1:800 dilution and rabbit anti-goat FITC (Jackson Immunoresearch Lab.), 1:200. The slides were then mounted with DAPI fluorescent mounting media (Vector Laboratories, Burlingame, CA).


We thank K.C. Choi, T. Ley, and T. Graubert for providing the GFP-tagged ES cells. We thank Karen Hutton in the DDRCC Morphology Core and Mike Veith in the EM facility for their assistance.