GATA3 inhibits GCM1 activity and trophoblast cell invasion

Expression of GATA3 and GCM1 in placenta

We investigated GATA3 and GCM1 expression in placenta by immunohistochemistry. The expression of the trophoblast marker, cytokeratin 7 (CK7), in placenta was also analyzed. In the first-trimester, both GCM1 and GATA3 were expressed in the nuclei of CK7-positive villous CTB cells and trophoblast cells of cell columns (Fig. 1, panels a–c). In the final stage of pregnancy, expression of both GCM1 and GATA3 was also detected in the nuclei of CK7-positive villous STB layer and EVT cells of full-term placentas (Fig. 1, panels e–g for EVT cells and i–k for STB layer). These observations suggested that GATA3 and GCM1 may be critical players in the control of trophoblast functions during pregnancy.

GATA3 suppresses HtrA4 expression

Our recent study has indicated that the GCM1 target gene HtrA4 is expressed in EVT cells and may modulate EVT cell invasion3. Because both GATA3 and GCM1 are expressed in EVT cells, we were curious about whether GATA3 may regulate HtrA4 gene expression. We first examined the effects of GATA3 on GCM1-mediated transcriptional activation of HtrA4 promoter in transient expression experiments. As shown in Fig. 2A, GATA3-FLAG did not affect luciferase activity directed by HtrA4 promoter in the pHtrA4-1 kb reporter construct in 293 T cells, which lack endogenous GCM1 and GATA3 activity (Supplementary Figure 1). While HA-GCM1 alone stimulated the luciferase activity directed by pHtrA4-1 kb, coexpression of GATA3-FLAG counteracted the stimulatory effect of HA-GCM1 on pHtrA4-1 kb in a dose-dependent manner (Fig. 2A). We further performed similar experiments in JEG-3 and BeWo trophoblast cell lines, which exhibit endogenous GCM1 activity, and demonstrated that HtrA4 promoter activity is suppressed by increasing amounts of GATA3-FLAG (Fig. 2B,C). We also performed ChIP analyses in BeWo cells to test whether GATA3 recognizes potential GATA3 sites in the same HtrA4 promoter region assessed in pHtrA4-1 kb. Using a ChIP-grade GATA3 Ab, we failed to detect association of GATA3 with the HtrA4 promoter region harboring three putative GATA3 sites (Fig. 2D). As expected, GCM1 associated with the same assessed HtrA4 promoter region, which contains a previously-identified GCM1-binding site (GBS) (Fig. 2D). These results suggested that GATA3 may antagonize GCM1-mediated HtrA4 gene expression without binding to the promoter region of HtrA4 gene.

GATA3 physically interacts with GCM1

We speculated that GATA3 may physically interact with GCM1 to antagonize the transcriptional activity of GCM1. To test this hypothesis, we first examined the effect of GATA3 on GCM1 transcriptional activity using a GCM1-specific reporter construct, p(GBS)₄E1bLuc. As shown in Fig. 3A, GATA3-FLAG significantly suppressed the transcriptional activity of HA-GCM1 in 293 T cells cotransfected with pHA-GCM1 and increasing amounts of pGATA3-FLAG. Subsequently, we preformed reciprocal coimmunoprecipitation assays in 293 T cells cotransfected with pHA-GCM1 and pGATA3-FLAG and detected specific interaction between HA-GCM1 and GATA3-FLAG (Fig. 3B). As a control, coexpression of GATA3-FLAG with the HA-tagged GCM2 homologue revealed no interaction between GATA3 and GCM2 (Fig. 3B). By domain swapping analysis, we further demonstrated that the DBD or TAD of GCM1 confers GATA3-binding activity to DBD or TAD of GCM2 (Fig. 3C). Direct physical interaction between GCM1 and GATA3 was supported by pull-down assays in which recombinant GATA3-FLAG bound to GST-GCM1, but not GST (Fig. 3D). Specific interaction between endogenous GCM1 and GATA3 in BeWo, JEG-3, and primary human trophoblast cells was confirmed by coimmunoprecipitation assays with rabbit anti-GCM1 Ab for immunoprecipitation and mouse anti-GATA3 Ab for immunoblotting (Fig. 4A). Furthermore, immunofluorescence microscopy revealed co-localization of endogenous GCM1 and GATA3 in BeWo and primary human trophoblast cells (Fig. 4B, panels c,i). Collectively, these results suggested that GATA3 directly interacts with GCM1 and inhibits the transcriptional activity of GCM1 in placenta.

Characterization of Interaction Domains of GCM1 and GATA3

To further characterize the interaction between GCM1 and GATA3, we mapped the interaction domains of GCM1 for GATA3, and vice versa. Coimmunoprecipitation assays were conducted in 293 T cells cotransfected with pHA-GATA3 and pGal4-FLAG or pGal4-GCM1-FLAG harboring the full-length GCM1 or truncated GCM1 domains. As shown in Fig. 5A, strong interaction was detected between HA-GATA3 and full-length GCM1 and the N-terminal domain (amino acids 1–167) of GCM1. In addition, modest interaction was detected between HA-GATA3 and the region of amino acids 167–349 in GCM1, which constitutes the first TAD (TAD1) of GCM1 (Fig. 5A). To map the interaction domain of GATA3 for GCM1, 293 T cells were cotransfected with pHA-GCM1 and pGal4-FLAG or pGal4-GATA3-FLAG harboring the full-length GATA3 or truncated GATA3 domains for coimmunoprecipitation assays. Interaction was detected between HA-GCM1 and full-length GATA3 and GATA3 truncated mutants harboring the N-terminal TADs or the first zinc finger 1 (ZF1) domain (Fig. 5B). We also generated ZF1-deletion mutant GATA3 constructs, pGal4-GATA3ΔZF1-FLAG and pGATA3ΔZF1-FLAG, and demonstrated that Gal4-GATA3ΔZF1-FLAG is able to interact with HA-GCM1 and that GATA3ΔZF1-FLAG suppresses the transcriptional activity of HA-GCM1 in transient expression experiments (Supplementary Figure 2). Therefore, the N-terminal TADs in ZF1-deletion mutant GATA3 are very likely responsible for interaction with HA-GCM1.

GATA3 inhibits the activity of GCM1 TAD

The suppressive effect of GATA3 on GCM1 activity can be attributed to inhibition of the activity of GCM1 DBD and/or TAD by GATA3. To test whether GATA3 interferes with the DNA-binding activity of GCM1, EMSA analysis was performed using recombinant GCM1-FLAG and GATA3-FLAG proteins and a radiolabeled HtrA4 GBS (HtrA4-GBS) probe. As shown in Fig. 6A, the HtrA4-GBS probe was specifically recognized by GCM1-FLAG, but not GATA3-FLAG. Importantly, the binding of GCM1-FLAG to HtrA4-GBS was not affected by increasing amounts of GATA3-FLAG (Fig. 6A). To test the effect of GATA3 on the transcriptional activity of GCM1, the GCM1 TAD (amino acids 167–436) was fused with the Gal4 DBD [Gal4-GCM1(167–436)]. We cotransfected 293T cells with pG5-Luc reporter construct, which harbors five tandem copies of Gal4-binding site, and different combinations of pGATA3-FLAG, pGal4-VP16, and pGal4-GCM1(167–436). As shown in Fig. 6B, the stimulation of luciferase reporter gene expression by Gal4-GCM1(167–436) was significantly suppressed by GATA3-FLAG. Taken together, these results suggested that GATA3 suppresses GCM1 activity through inhibition of its TAD activity.

GATA3 negatively regulates HtrA4 expression and placental cell invasion

We now investigated the effects of GATA3 on GCM1-regulated cellular activity in terms of cell invasion. As HtrA4 is a GCM1 target gene modulating placental cell invasion, HtrA4 expression was analyzed in BeWo and JEG-3 cells expressing scramble or GATA3 shRNA. Knocking down GATA3 in both BeWo and JEG-3 cells increased the HtrA4 protein level in culture media and whole cell lysates compared with that of control cells expressing scramble shRNA (Fig. 7A,B). Correspondingly, the HtrA4 transcript level was upregulated by knocking down GATA3 as measured in the GATA3-knockdown JEG-3 cells (Fig. 7B). The invasion activity of GATA3-knockdown BeWo and JEG-3 cells was measured by Matrigel-coated transwell assay. Compared with the control cells expressing scramble shRNA, the invasion activities of GATA3-knockdown BeWo and JEG-3 cells were significantly enhanced (Fig. 7C).

Plasmid constructs

The HA-tagged GCM1 expression plasmid, pHA-GCM1, has been described previously19. A DNA fragment encoding human GATA3 with an N-terminal triple HA tag or a C-terminal triple FLAG tag was subcloned into a CMV or EF1 promoter-driven expression vector to generate pHA-GATA3 or pGATA3-FLAG. The HtrA4 promoter region from nt −971 to 29 (relative to the translational initiation site) was subcloned into pGL3E1B to generate the pHtrA4-1 kb reporter plasmid. The p(GBS)₄E1bLuc reporter plasmid was generated by subcloning four copies of the proximal GCM1-binding site derived from the syncytin-1 gene in the pGL3E1B luciferase reporter vector. The pG5-Luc reporter plasmid which harbors five copies of the Gal4-binding site was obtained from Promega (Madison, WI).

Cell culture, transfection, and lentivirus transduction

293 T, BeWo, and JEG-3 cells were obtained from the American Type Culture Collection (Manassas, VA) and maintained at 37 °C in minimal essential medium alpha medium (for 293 T and JEG-3) or F-12 K medium (for BeWo) supplemented with fetal bovine serum, streptomycin, and penicillin. Placental villous tissues were obtained from term placentas of healthy women undergoing elective cesarean section with informed consent according to the study protocol (CT-100058) approved by the Institutional Review Board of Cathay General Hospital of Taiwan. Primary human trophoblast cells were prepared from normal term placentas according to Kliman et al.20 and cultured at 37 °C in Iscove’s modified Dulbecco’s medium containing 10% heat-inactivated FBS and the aforementioned antibiotics. The methods were carried out in accordance with the approved guidelines. For transient expression, cells were transfected with expression plasmids using the Lipofectamine 2000 reagent (Invitrogen). For luciferase reporter assays, cells were harvested and analyzed with a commercial kit (Promega). Specific luciferase activities were normalized by protein concentration, which was measured by using the BCA protein assay kit (Pierce, Rockford, IL).

For generating stable knockdown cell lines, BeWo or JEG-3 cells were infected by lentiviral pLKO.1-Puro short-hairpin RNA (shRNA) expression plasmids harboring a scrambled sequence (5′-CCTAAGGTTAAGTCGCCCTCG-3′) or target sequences for GATA3 (5′-CATCCAGACCAGAAACCGAAA-3′) provided by the National RNAi Core Facility of Taiwan. The infected cells were subjected to antibiotic selection using 5 μg/ml of puromycin, and puromycin-resistant clones were pooled for further study.

Immunohistochemistry and immunofluorescence

To examine GATA3 and GCM1 expression in placenta, early and term human placental tissue biopsy specimens were fixed in 4% paraformaldehyde, dehydrated, embedded in paraffin, and sectioned at 5 μm. Tissue sections were deparaffinized, rehydrated, and subjected to immunostaining by incubation with GCM1, GATA3 (Santa Cruz Biotechnology, Santa Cruz, CA), and cytokeratin 7 (CK7, EMD Millipore, Billerica, MA) antibodies (Abs) overnight, respectively. The sections were then incubated sequentially with biotinylated secondary Ab (Jackson ImmunoResearch, West Grove, PA) and HRP-conjugated streptavidin (Jackson ImmunoResearch). Antigenic detection was performed using DAB peroxidase substrate kit SK-4100 (Vector Laboratories, Burlingame, CA) and the sections were further counterstained with hematoxylin. Co-localization of GCM1 and GATA3 was performed in BeWo and primary human trophoblast cells by co-staining both factors with GCM1 and GATA3 Abs overnight, and then co-incubated with rhodamine-labeled secondary Ab for GCM1 and Cy2-labeled secondary Ab (Jackson ImmunoResearch) for GATA3. Nuclei were stained by DAPI. Immunofluorescence was examined under a Zeiss laser scanning confocal microscope (LSM510).

Coimmunoprecipitation and pull-down assay

To study the interaction between GCM1 and GATA3 in vivo, 293 T cells were transfected with pHA-GCM1 and pGATA3-FLAG. At 48 h post-transfection, cells were harvested in lysis buffer containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 2 mM EDTA, 10% glycerol, 0.5% NP-40, 1 mM DTT, 5 mM NaF, 1 mM Na₃VO₄, 1 mM PMSF, and a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO), followed by coimmunoprecipitation with FLAG (Sigma-Aldrich) and hemagglutinin (HA, Sigma-Aldrich) Abs. The interaction between endogenous GCM1 and GATA3 was studied in BeWo, JEG-3, and primary human trophoblast cells subjected to coimmunoprecipitation with GCM1 and GATA3 Abs.

To study the interaction between GCM1 and GATA3 in vitro, recombinant GATA3-FLAG was first immunopurified with FLAG antibody-conjugated agarose beads (Sigma-Aldrich) from 293 T cells transiently expressing GATA3-FLAG. Recombinant GATA3-FLAG was then incubated with bacterially expressed GST or GST-GCM1, which is a GST fusion protein of GCM1, pre-bound to glutathione-conjugated agarose beads (GE Healthcare Biosciences, Pittsburgh, PA) in lysis buffer at 4 °C for 3 h. After washing three times with lysis buffer, the proteins that were pulled down were analyzed by immunoblotting with FLAG Ab.

To map the GCM1 domain that interacts with GATA3, 293 T cells were transfected with pHA-GATA3 and pGal4-FLAG or pGal4-GCM1-FLAG (the latter contained full-length GCM1 or deletion mutants of GCM1) for coimmunoprecipitation with HA and FLAG Abs. A similar experiment was performed in 293 T cells transfected with pHA-GCM1 and pGal4-FLAG or pGal4-GATA3-FLAG (the latter contained full-length GATA3 or deletion mutants of GATA3) to map the GATA3 domain that interacts with GCM1.

Chromatin immunoprecipitation (ChIP) analysis and electrophoretic mobility shift assay (EMSA)

ChIP analysis was performed to study the association of GCM1 or GATA3 with HtrA4 promoter in BeWo cells using GCM1 or GATA3 Ab. The primer sequences for ChIP analysis of the GBS in HtrA4 gene are 5′-TGGAAACTGTTACGCTTCTCA-3′ and 5′-GTCTCTAGCCCTACCCG-3′ and the putative GATA3 sites in HtrA4 gene are 5′-GTAGTTGAGGAGTGGAGTTTAGTA-3′ and 5′-GGAGGAGGTCAAACAGTATG-3′ and the promoter region of GAPDH gene are 5′-AAAAGCGGGGAGAAAGTAGG-3′ and 5′-CTAGCCTCCCGGGTTTCTCT-3′.

Direct interaction between GCM1 and the GBS of HtrA4 gene was performed by EMSA as previously described4. In brief, a radiolabeled oligonucleotide probe harboring the HtrA4 GBS (5′-CAGTCTGCCCTCATTGTCGG-3′) was incubated with recombinant GCM1-FLAG alone or GCM1-FLAG plus increasing amounts of recombinant GATA3-FLAG. The reaction mixtures were analyzed by electrophoresis on 5% nondenaturing polyacrylamide gels. Recombinant GCM1-FLAG was immunopurified with FLAG antibody-conjugated agarose beads (Sigma-Aldrich) from 293 T cells transiently expressing GCM1-FLAG.

Quantitative real-time PCR

Cells were lysed using a RealTime ready cell lysis kit (Roche Applied Science, Indianapolis, IN) and reverse transcribed, followed by analysis of GCM1, HtrA4, and GATA3 transcripts in a LightCycler^® 480 real-time PCR instrument II (Roche Applied Science). The sequences of the primer sets for PCR analysis are as follows: 5′-CTGACAAGGCTTTTTTCTTCACA-3′ and 5′-CCAGACGGGACAGGTTT-3′ for GCM1, 5′-GACCCACCACCCCATCA-3′ and 5′-GGTTCTGTCCGTTCATTTTGT-3′ for GATA3, 5′-GTCAGCACCAAACAGCG-3′ and 5′-GGAGATTCCATCAGTCACCC-3′ for HtrA4, and 5′-AACTCCATCATGAAGTGTGACG-3′ and 5′-GATCCACATCTGCTGGAAGG-3′ for β-actin.

Cell invasion assay

Cell invasion analysis of GATA3-knockdown BeWo and JEG-3 cells were performed using Matrigel invasion chambers (BD Biosciences, Bedford, MA) according to the manufacturer’s instructions. In brief, cells were plated into the chambers and incubated for 48 h. Migrated cells in the lower surface of the filters were fixed with paraformaldehyde and visualized by hematoxylin stain and counted. Four microscopic fields per sample were randomly selected for quantification in each of three independent experiments. Images were prepared for presentation using Adobe Photoshop v7.0.

Statistical analysis

Statistical analysis of the data was performed using Student’s t test. Statistical significance was classified as *P < 0.05, **P < 0.01 or ***P < 0.001. A P value of <0.05 was considered statistically significant. A P value of >0.05 was considered to represent no statistical significance and denoted as “NS”.