GATA3 inhibits GCM1 activity and trophoblast cell invasion

Abstract

Development of human placenta involves the invasion of trophoblast cells from anchoring villi into the maternal decidua. Placental transcription factor GCM1 regulates trophoblast cell invasion via transcriptional activation of HtrA4 gene, which encodes a serine protease enzyme. The GATA3 transcription factor regulates trophoblast cell differentiation and is highly expressed in invasive murine trophoblast giant cells. The regulation of trophoblastic invasion by GCM1 may involve novel cellular factors. Here we show that GATA3 interacts with GCM1 and inhibits its activity to suppress trophoblastic invasion. Immunohistochemistry demonstrates that GATA3 and GCM1 are coexpressed in villous cytotrophoblast cells, syncytiotrophoblast layer, and extravillous trophoblast cells of human placenta. Interestingly, GATA3 interacts with GCM1, but not the GCM2 homologue, through the DNA-binding domain and first transcriptional activation domain in GCM1 and the transcriptional activation domains and zinc finger 1 domain in GATA3. While GATA3 did not affect DNA-binding activity of GCM1, it suppressed transcriptional activity of GCM1 and therefore HtrA4 promoter activity. Correspondingly, GATA3 knockdown elevated HtrA4 expression in BeWo and JEG-3 trophoblast cell lines and enhanced the invasion activities of both lines. This study uncovered a new GATA3 function in placenta as a negative regulator of GCM1 activity and trophoblastic invasion.

Figure 1. Expression of GATA3 and GCM1 in placenta.

First-trimester (8 weeks, (a–d)) and full-term (e–l) placental sections were subjected to immunohistochemistry with mouse anti-GATA3 (a,e,i), rabbit anti-GCM1 (b,f,j), and mouse anti-cytokeratin 7 (CK7, (c,g,k)) Abs, respectively. In the first-trimester placenta, GATA3 and GCM1 are expressed in the CK7-positive trophoblast cells of cell column and villous cytotrophoblast cells. Magnified images of the boxed regions in first-trimester placental sections are shown on the right. Both GATA3 and GCM1 are highly expressed in the CK7-positive EVT cells and the CK7-positive syncytiotrophoblast layer of full-term placenta. As negative controls, first-trimester and full-term placental sections were incubated with normal mouse IgG (d) and normal rabbit IgG (h,l), respectively. Scale, 50 μm.

Figure 2. GATA3 downregulates HtrA4 promoter activity.

293 T (A) cells were transfected with 0.05 μg of pHtrA4-1 kb, 0.1 μg of pHA-GCM1, and increasing amounts of pGATA3-FLAG (0.1 and 0.3 μg), whereas JEG-3 (B) and BeWo (C) placental cells were transfected with pHtrA4-1 kb and increasing amounts of pGATA3-FLAG (0.1, 0.2, and 0.3 μg). At 48 h post-transfection, cells were harvested for luciferase reporter assays. (D) GCM1, but not GATA3, recognizes HtrA4 promoter. BeWo cells were subjected to ChIP analysis using GCM1 and GATA3 Abs, respectively. The immunoprecipitated complexes were analyzed by PCR with specific primer sets (arrows) flanking the GCM-binding site (GBS) or the putative GATA3 sites. Mean values and the S.D. obtained from three independent experiments are presented in (A–C).

Figure 3. Physical interaction between GATA3 and GCM1.

(A) GATA3 suppresses GCM1 transcriptional activity. 293 T cells were transfected with 0.05 μg of p(GBS)₄E1bLuc, 0.1 μg of pHA-GCM1, and increasing amounts of pGATA3-FLAG (0.1 and 0.3 μg), followed by luciferase reporter assays. Mean values and the S.D. obtained from three independent experiments are presented. (B,C) GATA3 interacts with GCM1, but not GCM2. 293 T cells were transfected with 2.5 μg of pGATA3-FLAG and 2.5 μg of pHA-GCM1 or pHA-GCM2 (B). In a separate experiment, 293 T cells were transfected with 2.5 μg of pHA-GATA3 and 2.5 μg of pG1G2-FLAG or pG2G1-FLAG, which encodes a chimeric GCM1 and GCM2 protein (C). At 48 h post-transfection, cells were harvested for reciprocal coimmunoprecipitation assays with FLAG and HA Abs. (D) Direct interaction between GCM1 and GATA3. Approximately 45 ng of recombinant GATA3-FLAG protein was incubated with glutathione-conjugated agarose beads prebound with 0.5 μg of GST or 1.5 μg of GST-GCM1 in pull-down assays.

Figure 4. Interaction and co-localization of GATA3 and GCM1 in placenta.

(A) Interaction of endogenous GATA3 and GCM1 in placental cells. BeWo, JEG3, and primary human trophoblast cells were subjected to coimmunoprecipitation assays with rabbit normal IgG or rabbit anti-GCM1 Ab for immunoprecipitation and mouse anti-GATA3 Ab for immunoblotting. Input controls are provided for GCM1, GATA3, and CK7 (human trophoblast cells) in the lower panels. (B) Immunofluorescence microscopy of GATA3 and GCM1 in placental cells. BeWo (a–f) and primary human trophoblast (g–l) cells were fixed, permeabilized, and stained with mouse anti-GATA3 (a,g), rabbit anti-GCM1 (b,h), and mouse anti-CK7 (j) Abs. Cells were then incubated with secondary Abs and examined under a confocal microscope. Panels c and i are the merge images of GATA3 and GCM1 staining in BeWo (a,b) and human trophoblast (g,h) cells, respectively. Nuclei were stained with DAPI (blue, d–f,j–l). As negative controls, cells were stained with normal mouse IgG (e,k) and normal rabbit IgG (f,l). Scale, 50 μm.

Figure 5. Analysis of interaction domains in GCM1 and GATA3.

(A) Mapping of GATA3-interacting domains in GCM1. 293 T cells were transfected with 2.5 μg of pHA-GATA3 and a serial of pGal4-GCM1-FLAG constructs (2.5 μg) encoding Gal4 fusion proteins with full-length GCM1 or different GCM1 domains. (B) Mapping of GCM1-interacting domains in GATA3. 293 T cells were transfected with 2.5 μg of pHA-GCM1 and a series of pGal4-GATA3-FLAG constructs (2.5 μg) encoding Gal4 fusion proteins with full-length GATA3 or different GATA3 domains. At 48 h post-transfection, cells were harvested for coimmunoprecipitation assays with HA and FLAG Abs.

Figure 6. GATA3 blocks GCM1 transcriptional activity.

(A) GATA3 does not affect the DNA-binding activity of GCM1. Approximately 20 ng of recombinant GCM1-FLAG alone or plus increasing amounts of GATA3-FLAG (20, 40, and 60 ng) was incubated with 2 ng of radiolabeled HtrA4-GBS probe in the EMSA analysis. The asterisk indicates the GCM1-FLAG-DNA complex. Of note, 20 ng of recombinant GATA3-FLAG alone did not recognize the probe. (B) Suppression of the activity of GCM1 TAD by GATA3. 293 T cells were transfected with different combinations of 0.05 μg of pG5-Luc, 0.1 μg of pGal4-VP16, 0.1 μg of pGal4-GCM1(167–436), and 0.3 μg of pGATA3-FLAG. At 48 h post-transfection, cells were harvested for luciferase reporter assay. Mean values and the S.D. obtained from three independent experiments.

Figure 7. Regulation of placental cell invasion by GATA3.

(A,B) GATA3 knockdown elevates HtrA4 expression in BeWo and JEG-3 cells. Culture media and whole cell lysates (WCLs) were harvested from BeWo and JEG-3 cells stably expressing scramble or GATA3 shRNA (shGATA3) for immunoblotting analysis with the indicated Abs. In a separate experiment, the scramble and GATA3 shRNA-expressing JEG-3 cells were harvested for quantitative real-time PCR analysis for GCM1, GATA3, and HtrA4 transcripts, respectively. (C) GATA3 knockdown enhances placental cell invasion. BeWo and JEG-3 cells stably expressing scramble or GATA3 shRNA were plated in Matrigel-coated chambers for 24 h. The invasive BeWo and JEG-3 cells in the lower surface of the filters were fixed, stained, and counted. Mean values and the S.D. obtained from three independent experiments are presented in (B,C).