A Gata2-Dependent Transcription Network Regulates Uterine Progesterone Responsiveness and Endometrial Function

Abstract

Altered progesterone responsiveness leads to female infertility and cancer, but underlying mechanisms remain unclear. Mice with uterine-specific ablation of GATA binding protein 2 (Gata2) are infertile, showing failures in embryo implantation, endometrial decidualization, and uninhibited estrogen signaling. Gata2 deficiency results in reduced progesterone receptor (PGR) expression and attenuated progesterone signaling, as evidenced by genome-wide expression profiling and chromatin immunoprecipitation. GATA2 not only occupies at and promotes expression of the Pgr gene but also regulates downstream progesterone responsive genes in conjunction with the PGR. Additionally, Gata2 knockout uteri exhibit abnormal luminal epithelia with ectopic TRP63 expressing squamous cells and a cancer-related molecular profile in a progesterone-independent manner. Lastly, we found a conserved GATA2-PGR regulatory network in both human and mice based on gene signature and path analyses using gene expression profiles of human endometrial tissues. In conclusion, uterine Gata2 regulates a key regulatory network of gene expression for progesterone signaling at the early pregnancy stage.

Keywords: GATA2; TRP63; endometrium; infertility; path analysis; pregnancy; progesterone; progesterone receptor; structural equation modeling; uterus.

Figure 1. Regulation of PGR expression by GATA2

(A-B) Correlation between GATA2 and PGR mRNA levels in human tissues. N, number of samples; r, Pearson's correlation coefficient. (C) mRNA levels of Gata factors in wild type and Gata2^dd mouse uteri by qRT-PCR. N=3 for each group. (D & E) Uteri, pregnancy day 5.5 mice of denoted genotypes. Arrows mark the embryo implantation sites. Arrowheads denote ovaries. (F-I) Morphology of the decidual response in Day 2 and Day 5 post decidual stimulation. Right horn was stimulated in all groups. (J-N) PGR levels in uteri of OVX 6 wks old mice treated with 1 mg P4 or vehicle for 6 hrs. (J-M) Immunohistochemical analysis of PGR protein. LE, luminal epithelium; GE, glandular epithelium. (N) Protein levels of both isoforms of the PGR by western blot. Error bars represent standard error of the mean. ***, p < 0.001, ns, p > 0.05 by student's t-test. See also Figure S1.

Figure 2. Impaired P4 signaling in Gata2 deficient uterus

OVX mice were treated with vehicle (oil) or 1 mg P4 for 6 hrs. (A-C) mRNA levels of PGR target genes assayed by qRT-PCR. N = 3 for each group. (D-E) Gene expression profiling by microarray assays on denoted treatments and genotypes. (D) Venn diagram of numbers of P4 responsive genes in denoted genotypes. (E) Magnitude of P4 responsiveness on expression of the common P4 responsive genes between denoted genotypes in (D). (F-G) Genome-wide expression profiles of Gata2 dependent genes in the presence of P4. (F) Heatmap of hierarchically clustered profiles. Low expression values are blue and high expression values are yellow. N=3 for each group. (G) Predicted regulators of Gata2 dependent genes by the Ingenuity Pathway Analysis. Error bars represent standard error of the mean. *, p < 0.05; **, p< 0.01; ***, p < 0.001 by student's t-test. See also Figure S2 and Tables S1, S2, S3 and S7.

Figure 3. Binding and regulation of Pgr by GATA2

(A) Pgr mRNA levels by qRT-PCR in uteri of OVX mice acutely treated with vehicle for 6 hrs. (B) Gata2 binding sites in the Pgr promoter. Red denotes GATA2 binding motifs. PCR primers used for ChIP-qPCR analysis are listed here. (C) ChIP-qPCR analysis of GATA2 occupancy on Pgr proximal promoter in uteri 1 hr after P4 treatment in OVX mice. (D) Luciferase reporter analysis of Pgr promoter activities in response to various GATA2 expression vectors. GATA2, full length GATA2; ΔNT, GATA2 N-terminal zinc-finger deleted mutant; ΔCT, GATA2 C-terminal zinc-finger deleted mutant. Error bars represent standard error of the mean. *, p < 0.05; **, p< 0.01; ***, p < 0.001 by student's t-test.

Figure 4. Binding of GATA2 and PGR at loci of uterine P4 responsive genes

(A, B) Enriched GATA2 and PGR binding motifs at GATA2 binding intervals that were identified by ChIP-Seq in OVX mouse uteri 6 hrs after P4 treatment. (C) Venn diagrams of P4 regulated genes that contain binding sites for PGR or GATA2 within +/− 25 kb of gene boundaries. (D) Occupancy of PGR and GATA2 in putative enhancers for Sox17 and Ihh genes in P4 treated uteri. Blue indicates PGR while red indicates GATA2 occupancy. Brackets mark the regions used for luciferase reporter analysis. (E, F) Luciferase reporter analyses of cotransfection analysis of denoted cis-acting elements in HEC-1A cells. Cotransfection with empty vector, PGR-A, PGR-B or GATA2 expression vectors treated with vehicle or 10⁻⁸ M R5020 for 24 hrs. Experiments were performed in triplicate. Error bars represent standard error of the mean. *, p < 0.05; ***, p < 0.001 by student's t-test. See also Figures S3 and S4 and Table S4.

Figure 5. P4-independent functions of Gata2

(A) Enrichment of functional annotation in differentially expressed genes between Gata2^f/f and Gata2^d/d uteri of OVX, vehicle-treated mice by IPA analysis. (B) mRNA levels of TRP63 in uterus of OVX mice. (C, D) Immunostaining of TRP63 in uteri of OVX mice (control horn of decidualization treatment). Scale bar denotes 100 uM. (E, F) Immunostaining of TRP63 in uteri of OVX mice treated with E2 for 2 wks. (G) qRT-PCR of uterine tissues of OVX mice treated with E2. N = 6 (Gata2^f/f, vehicle, 2 wks), 6 (Gata2^d/d, vehicle, 2 wks), 6 (Gata2^f/f, E2, 2 wks), 6 (Gata2^d/d, E2, 2 wks), 6 (Gata2^f/f, vehicle, 4 wks), 6 (Gata2^d/d, vehicle, 4 wks), 4 (Gata2^f/f, E2, 4 wks), 6 (Gata2^d/d, E2, 4 wks), 6 (Gata2^f/f, vehicle, 2 mo.), 4 (Gata2^d/d, vehicle, 2 mo.), 6 (Gata2^f/f, E2, 2 mo.), 5 (Gata2^d/d, E2, 2 mo.). (H, I) qRT-PCR of uterine tissues of OVX mice treated with E2 for 2 mo. Error bars represent standard error of the mean. *, p < 0.05; **, p< 0.01; ***, p < 0.001 by student's t-test. See also Figure S5 and Table S5.

Figure 6. A GATA2-dependent regulatory network in human endometrium

(A) (Top panel) Expression array data of 115 human endometrial tissue samples from GSE58144, for genes in a transcriptional signature of GATA2, with samples ordered based on manifestation of the signature. (Bottom panel) Heatmap shows levels of genes of interest in individual samples corresponding to the order in the top panel. (B) Network of transcription factors for regulation of female fertility. (C, D) Path analysis of the genetic network by mRNA expression levels and gene signature in GSE58144 through structural equation modeling. GATA2 and PGR signatures were derived from (A) and GSE39920, respectively. (E) Model fitting for regulation of P4 signaling by GATA2 and PGR. The P4 signature is based on a gene list described in the green box of Fig. 2D. Brackets denote mRNA levels. P values are derived by path analysis based on linear statistic models. See also Figure S6 and Table S6.