Abstract
Mouse studies support a role for endometrial early growth response 1 (EGR1) in uterine receptivity and decidualization, which are processes controlled by estrogen and progesterone. However, the importance of this transcription factor in similar cellular processes in human is unclear. Analysis of clinical samples indicate that endometrial EGR1 levels are decreased in the endometrium of women with recurrent implantation failure, suggesting tight control of EGR1 levels are necessary for normal endometrial function. Therefore, we used siRNA-mediated knockdown of EGR1 expression in cultured human endometrial stromal cells (hESCs) to assess the functional role of EGR1 in hESC decidualization. Protein expression studies revealed that EGR1 is highly expressed in pre-decidual hESCs. However, EGR1 protein levels rapidly decrease following administration of an established deciduogenic hormone stimulus containing estradiol, medroxyprogesterone acetate, and cyclic adenosine monophosphate. Intriguingly, EGR1 knockdown in pre-decidual hESCs blocks the ability of these cells to decidualize later, indicating that EGR1 is required to transcriptionally program pre-decidual hESCs for decidualization. Support for this proposal comes from the analysis of transcriptome and cistrome datasets, which shows that EGR1 target genes are primarily involved in transcriptional regulation, cell signaling, and proliferation. Collectively, our studies provide translational support for an evolutionary conserved role for human endometrial stromal EGR1 in the early events of pregnancy establishment.
Keywords: Early growth response 1, human endometrial stromal cells, decidualization, RNA-sequencing, ChIP-sequencing, pregnancy
1. Introduction
As an immediate early response transcription factor to a myriad of external stimuli, early growth response 1 (EGR1) modulates transcriptional programs obligate for proliferation, differentiation, and programmed cell death within a broad spectrum of physiological and pathophysiological contexts [1]. A member of the evolutionary conserved EGR family of Cys2-His2-type zinc finger transcription factors (containing also EGR2, EGR3, and EGR4), EGR1 directly controls gene transcription of target genes through binding GC-rich DNA motifs via its three zinc finger DNA binding domains. Because EGR1 is required for the expression of the luteinizing hormone beta subunit in the gonadotropes of the anterior pituitary gland, ablation of Egr1 in the female mouse causes infertility primarily due to impairments in ovulation and luteinization [2, 3]. In the case of the uterus, recent studies revealed that Egr1 is induced by estrogen in the endometrial luminal epithelium and stroma during early pregnancy [4–7]. Mouse studies also showed that luminal epithelial Egr1 is essential for uterine receptivity [7] and that Egr1 in pre-decidual stromal cells in the subluminal region is necessary for complete decidualization in response to an artificial deciduogenic stimulus [4, 5, 7].
At the RNA level, we recently demonstrated that EGR1 is expressed in pre-decidual human endometrial stromal cells (hESCs) in culture; however, EGR1 transcript levels significantly decline in response to an established hormone deciduogenic stimulus [8]. While these findings suggest that downregulation of EGR1 expression is a prerequisite for decidualization, the role of EGR1 in pre-decidual hESCs remained an open question. Intriguingly, we also found in a previously published dataset that EGR1 transcript levels are decreased in the endometrium of women with recurrent implantation failure [9]. Therefore, we used RNA interference of EGR1 in cultured hESCs and found that abrogation of EGR1 in pre-decidual hESCs impairs their ability to decidualize when exposed to a hormone deciduogenic stimulus. Furthermore, application of RNA-sequencing (RNA-seq) and chromatin immunoprecipitation followed by sequencing (ChIP-seq) revealed that EGR1 in pre-decidual hESCs regulates the expression of a large number of genes involved in transcriptional regulation, cell signaling, and proliferation, cellular processes predicted to prime these cells for immediate proliferation and subsequent differentiation in response to the deciduogenic signal.
2. Materials and methods
2.1. Cell culture
Following a human protocol prospectively approved by the Institutional Review Board at Baylor College of Medicine and in accordance with the guidelines of the declaration of Helsinki [10], primary hESCs were isolated and maintained as described [11]. Unless otherwise stated, cell culture reagents were purchased from Thermo Fisher Scientific (Waltham, MA). Primary hESCs were used for initial validation of gene function. The telomerase-immortalized T-HESC cell line was purchased from the American Type Culture Collection (Manassas, VA (CRL-4003)) and cultured in DMEM/F12 without phenol red with 10% fetal bovine serum (FBS; Sigma-Aldrich, St. Louis, MO), 100 units/ml penicillin, 0.1mg/ml streptomycin, 0.5 µg/ml of puromycin (Sigma-Aldrich), and Insulin-Transferrin-Selenium-Sodium Pyruvate (ITS-A). The identity of this cell line has been confirmed by short tandem repeat (STR) profiling by the Baylor College of Medicine Tissue Culture Core. Because of the number of cells required, the T-HESC cell line was used for ChIP-seq experiments and parallel RNA-seq. Importantly, the T-HESC cell line has been shown to decidualize similarly to primary hESCs in vitro [12].
2.2. In vitro decidualization
Primary hESCs and T-HESCs were plated into 6-well plates and transfected with small interfering (si)RNAs for 48 hours in OptiMEM with 2% charcoal-stripped FBS (sFBS) [11]. If cells were not transfected with siRNAs, they were incubated in OptiMEM with 2% sFBS for 48 hours prior to hormone treatment. For in vitro decidualization, cells were treated with a combination of 17β-estradiol (1×10−8 M), medroxyprogesterone acetate (1×10−6 M (primary hESCs) and 5×10−6 M (T-HESC cells)), and N6,2-O-dibutyryladenosine 3,5-cyclic monophosphate sodium salt (5×10−5 M), (EPC) as described in [11]. Non-targeting (NT) and EGR1 siRNAs were purchased from Dharmacon (Lafayette, CO (D-001810–10-05)) and Qiagen Inc. (Germantown, MD (GS1958)) respectively. For short-term (hours) EPC treatments, EPC was diluted in DMEM without serum and without phenol red. For long-term (days) EPC treatments, EPC was diluted in Opti-MEM without phenol red supplemented with 2% sFBS.
2.3. Isolation of RNA and quantitative reverse transcriptase (RT)-PCR
Total RNA was isolated with the RNeasy Plus kit with DNA removal columns (Qiagen Inc.) according to the manufacturer’s instructions. Complementary (c)DNA was prepared with SuperScript VILO Master Mix (Thermo Fisher Scientific); cDNA of interest was quantified by Taqman real-time PCR as described in [11]. Taqman assays used in this study are listed in Supplemental Table ST1
2.4. RNA sequencing
Before RNA-seq, the purity and integrity of RNA was evaluated with a NanoDrop spectrophotometer (Thermo Fisher Scientific) and a 2100 Bioanalyzer with RNA chips (Agilent Technologies, Santa Clara, CA) respectively. For each experimental group, RNA samples from three independent experiments were used. Sequencing libraries were prepared with the TruSeq Stranded mRNA kit (Illumina Inc., San Diego, CA) from 250 ng of RNA. Quality analysis of libraries was performed on 4200 TapeStation with D1000 ScreenTape assays (Illumina Inc.). Libraries were quantified with KAPA Library Quantification Kit (KAPA Biosystems, Wilmington, MA). Following equimolar pooling, libraries were quantified on the 2100 Bioanalyzer (using the High Sensitivity DNA Kit and DNA chips) and KAPA Library Quantification Kit for Illumina platforms. Sequencing of libraries was performed on the NextSeq 500 platform (Ilumina Inc.). Paired-end 75 base pair (bp) sequencing reads were generated at mid-output and mapped to the human genome (UCSC hg19) using STAR [13] with NCBI RefSeq genes as the reference. The Bioconductor package containing EdgeR was used to analyze the read counts to detect differentially expressed genes between EGR1 and NT siRNA groups [14]. The false discovery rate (FDR) of differentially expressed genes was estimated using the Benjamini and Hochberg method [15]. Differentially expressed genes were gated for further analysis using absolute fold change (IFCI) > 1.5 and FDR < 0.05. Gene ontology enrichment analysis was performed using DAVID (Database for Annotation, Visualization, and Integrated Discovery) functional annotation clustering tool [16].
2.5. Chromatin immunoprecipitation-sequencing and quantitative (q)PCR
Chromatin was isolated from 0.8–2×107 T-HESC cells fixed in formaldehyde. Chromatin was immunoprecipitated by Active Motif (Carlsbad, CA) using ChIP grade antibodies against either EGR1 (Cell Signaling Technology (Danvers, MA (4154BF)) or acetylated lysine 27 of histone H3 ((H3K27ac) (Active Motif, 39133)). For ChIP sequencing, input and immunoprecipitated DNA was sequenced on a NextSeq 500 platform with single-end 75 bp read. Sequencing reads were aligned to GRCh37/hg19. Following normalization by down-sampling, peaks were called with the MACS 2.1.0 algorithm with p-value cutoffs of 10−7 for narrow peaks and 10−1 for broad peaks. Primers used for ChIP-qPCR are listed in Supplemental Table ST2.
2.6. Western blot analysis
Protein was extracted from cells with NP40 buffer with proteinase inhibitors (cOmplete Mini, Roche, Basel, Switzerland) and phosphatase inhibitors (Sigma Aldrich) and quantified with Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Protein (15μg) was resolved by SDS-PAGE in 4–15% polyacrylamide gel and transferred onto polyvinylidene fluoride (PVDF) membranes. Following incubation of PVDF membranes in 5% milk in Tris-buffered saline with 0.1% Tween 20 (TBST), the following antibodies were used for antigen detection: anti-EGR1 (Cell Signaling Technology (4154); diluted 1:2000) and anti-β-actin (Sigma-Aldrich (A1978); diluted 1:50000). Horseradish peroxidase (HRP)-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology (Dallas, TX). Peroxidase activity was detected with Pierce ECL Western Blotting Substrate (Thermo Fisher Scientific).
2.7. Mouse husbandry
Care and handling of mice occurred according to an animal protocol previously approved by the Institutional Animal Care and Use Committee (IACUC) at Baylor College of Medicine (BCM) and to guidelines described in the Guide for the Care and Use of Laboratory Animals (National Research Council (Eighth Edition 2011)). Mice were housed at an AAALAC-accredited vivarium with a 12h light-12h dark cycle at the Center for Comparative Medicine at BCM and were fed and watered ad libitum. For timed pregnancy experiments using CD1 proven stud males, the morning of detecting a vaginal plug in CD1 females was designated as gestational day 1. Following euthanasia by approved methods, uterine tissue was dissected and processed for histology at a predetermined gestation day. Two hours prior to euthanasia, mice received a dose of 1 mg per 20 g of body weight of 5-bromo-2’-deoxyuridine (BrdU; BD Biosciences, San Jose, CA) via intraperitoneal injection.
2.8. Immunohistochemical analysis
Mouse uterine tissues were fixed overnight in 4% paraformaldehyde in phosphate buffered saline (PBS) before paraffin embedding and sectioning onto slides for immunohistochemistry [17]. The EGR1 transcription factor was detected with a primary rabbit monoclonal antibody (Cell Signaling Technology Inc. (4153); diluted 1:100) followed by a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (Vector Laboratories, Burlingame, CA (P-1000); diluted 1:200). The peroxidase activity was detected with VECTASTAIN Elite ABC-HRP Kit (Vector Laboratories, Burlingame, CA). Following immunostaining, tissue sections were counterstained with hematoxylin prior to applying permount and affixing coverslips.
2.9. Immunofluorescent analysis
Cells were plated on coverslips (acid-etched and uncoated) in 24-well plates at 1.25×105 cells per well. Following hormone treatment, cells were washed with ice-cold Dulbecco’s PBS containing Ca2+ and Mg2+ and fixed with ice-cold 4% paraformaldehyde in Pipes-EGTA-MgCl2 (PEM) buffer. Autofluorescence was quenched by incubating in 0.1 M ammonium chloride in PEM for 10 min. Cells were permeabilized with 0.5% Triton X-100 in PEM for 30 min. Following incubation in 5% milk in TBST, EGR1 was detected with a rabbit monoclonal primary antibody (Cell Signaling Technology (4153); diluted 1:500) followed by secondary antibody, a goat anti-rabbit Alexa Fluor 594-conjugated antibody (ThermoFisher Scientific (R37117)). Following another round of fixation and autofluorescence quenching, cells were stained with DAPI and coverslips were mounted with Prolong Gold Antifade Mountant (ThermoFisher Scientific). Cells were imaged using GE Healthcare Inverted Deconvolution/Image Restoration Microscope System at the Baylor College of Medicine Integrated Microscopy Core. Intensity of EGR1 staining was measured using ImageJ’s (https://imagej.nih.gov/ij/) Analyze Particles tool on 16-bit images and expressed as the nuclear mean grey value using DAPI images for nuclear demarcation. Samples without the inclusion of primary EGR1 antibody were used for background subtraction in the analysis.
2.10. Statistical analyses
Results are presented as averages ± standard deviation. Analyses were performed using R (version 3.3.1) in R Studio (R Studio Inc., Boston, MA). Quantile comparison plots were generated to inspect normality of data. Statistical analyses were performed with two sample t-test or Wilcoxon rank sum test; or with ANOVA or Kruskal-Wallis rank sum test with post hoc analyses performed with either Tukey’s range test or Wilcoxon rank sum test. Differences with p-values < 0.05 were considered significant with the following designations: p < 0.05 (*), p ≤0.01 (**), and p ≤0.001 (***). Specific tests used were annotated in the Figure Legends.
2.11. Analysis of EGR1 expression in the endometrium of women with recurrent implantation failure
Data to evaluate the levels of EGR1 transcript in the endometrium of women with recurrent implantation failure and in control endometrium were exported from a published and publicly available dataset in the Gene Expression Omnibus (GEO) database: GSE58144 [9]. These data were extracted using GEO’s GEO2R portal. The full accession ID number of the EGR1 expression data is GSE58144/GPL15789/17940. Wilcoxon rank sum test was used to compare the control and the recurrent implantation failure groups.
3. Results
3.1. Perturbation of EGR1 levels is associated with human endometrial disorders linked to implantation failure
Our recent hESC studies indicate that tight control of EGR1 expression levels is required for these cells to decidualize [8], suggesting that perturbation of EGR1 levels may be associated with endometrial disorders linked to early embryo implantation failure. Support for this supposition comes from our analysis of a recently published transcriptome dataset from van Koot et al. [9]. The dataset comprised microarray-profiled transcriptome information (GEO accession: GSE58144) from human endometrial tissues biopsied during the mid-luteal stage of the ovarian cycle (i.e. the receptivity (or implantation) window at days 5 to 8 after luteinizing hormone (LH) surge). Biopsies were obtained from two patient groups: (1) women who successfully became pregnant within two cycles after in vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI) and therefore their infertility was inferred to be of non-endometrial origin; and (2) women with recurrent implantation failure (RIF) despite undergoing IVF/ICSI. Our analysis revealed that relative EGR1 mRNA levels are significantly lower in the endometrium of women with RIF as compared with women who were able to become pregnant (Fig. 1). This is the first result to reveal a connection between deregulation of normal EGR1 levels and an endometrial disorder linked to human implantation failure. This finding also underscores the importance of delineating the normal functional contribution of EGR1 in the pre-decidual stromal cell to human endometrial decidualization, an essential cellular process in uterine periimplantation biology.
Figure 1. Transcript levels of EGR1 are significantly decreased in the endometrium of women with recurrent implantation failure.
Expression data was retrieved from the Gene Expression Omnibus dataset GSE58144 [9]. Human endometrial tissue, which was biopsied during the mid-luteal phase of the menstrual cycle, was the RNA source. Control samples were derived from women who successfully conceived following IVF/ICSI whereas recurrent implantation failure samples were obtained from subjects who failed to get pregnant. Statistical analysis performed with Wilcoxon rank sum test.
3.2. Expression of EGR1 protein in pre-decidual and decidualized hESCs
In agreement with our previous findings at the mRNA level [8], EGR1 protein levels significantly decrease in hESCs upon EPC exposure within 3 days (Fig. 2A). The decrease in EGR1 protein levels is rapid, with most of the decrease occurring within 4–5 hours of EPC exposure (Fig. 2B; note: EGR1 mRNA levels decline even more rapidly during this period (Supplemental Fig. 1A); siRNA mediated knockdown of EGR1 confirms specificity of the EGR1 antibody used (Supplemental Fig. 1B)). Before EPC treatment, EGR1 is predominantly located in the nuclei and exhibits a broad spectrum of expression levels per cell (Fig. 2C and Supplemental Fig. 1C); ~80% of cells are positive for EGR1 expression at this time-point. The EGR1 expression levels decrease globally within 6 hours of EPC treatment. A minor sub-population of cells re-express at high levels EGR1 at a later time-point (24 h) despite prolonged EPC exposure (cells with a nuclear mean grey value (NMGV) > 1000; Fig. 2C and Supplemental Fig. 1C).
Figure 2. Early growth response 1 protein expression is rapidly downregulated with EPC administration.
Western analysis of EGR1 protein expression in T-HESC cells under EPC exposure for: (A) 0 to 6 days (shown in quadruplicate) and (B) 0 to 360 minutes; for each β-actin is a loading control. (C) Boxplots representing the nuclear mean grey value (NMGV) of EGR1 immunofluorescent staining at 0, 240, and 360 min and 24 h of EPC treatment. Percentage of cells with NMGV > 1000 for each time-point is noted above plot. *** denotes p-values calculated for comparisons with the 0 min of EPC treatment data point. Examples images of EGR1 immunofluorescent detection are shown in Supplemental Fig. S1B. Statistical analysis performed with Kruskal-Wallis rank sum test with post hoc pairwise Wilcoxon rank sum tests.
Through ChIP-qPCR analysis, we demonstrated that the rapid reduction of EGR1 levels with EPC treatment is not accompanied by a decrease in acetylation of the lysine 27 of histone H3 (H3K27Ac) in the promoter region upstream of the annotated EGR1 gene (Supplemental Fig. 2A B). These results show that the EGR1 gene in hESCs is not regulated by changes in this epigenetic mark and suggest that this gene may be maintained in a “poised state” [18] under EPC treatment. In contrast, the EGR2 gene in hESCs exhibits significant downregulation of H3K27Ac at a putative intronic enhancer [18] within 3 days of EPC treatment (Supplemental Fig. 2A B), which is coincident with the decrease in EGR2 expression levels during hESC decidualization [11].
The striking decrease in EGR1 levels during hESC decidualization is also observed in the early pregnant mouse (Supplemental Fig. S3). Many murine endometrial stromal cells clearly express EGR1 protein during gestational days 4–5 (the window of implantation); however, the majority of these cells no longer express EGR1 at gestational day 6. Only a few stromal cells remain that express EGR1, which are found closely surrounding the implanting embryo (Supplemental Fig. S3).
3.3. Early growth response 1 in pre-decidual hESCs is required for decidualization
To address whether EGR1 expressed in pre-decidual hESCs is required for these cells to decidualize in response to EPC, siRNA-mediated knockdown of EGR1 was performed on hESCs forty-eight hours prior to administration of EPC cocktail. Following effective EGR1 knockdown (Fig. 3A), hESCs failed to decidualize as evidenced by significant attenuation in the induction of the decidual markers, insulin-like growth factor binding protein 1 (IGFBP1) and prolactin (PRL) [19] (Fig. 3A). This decidual molecular defect was reflected at the cellular level by an inability of hESCs with diminished EGR1 levels to transform from a fibroblastic spindle to a polygonal epithelioid morphology when exposed to EPC for six days (Fig. 3B–E). Together, our results strongly support a critical role for EGR1 in pre-decidual hESCs in the decidualization of these cells when exposed to a deciduogenic stimulus.
Figure 3. Maintenance of EGR1 levels in pre-decidual primary hESCs is required for decidualization.
(A) Expression of EGR1, IGFBP1, and PRL mRNA measured by quantitative RT-PCR following non-targeting (NT) or EGR1 siRNA transfection and EPC exposure for 0, 3, or 6 days in primary hESCs. Statistical analysis performed with ANOVA with post hoc Tukey’s range test. Cell morphology of hESCs at day 0 of EPC treatment after (B) NT siRNA or (C) EGR1 siRNA transfection and at day 6 of EPC treatment (D and E).
3.4. Early growth response 1 transcriptionally primes pre-decidual hESCs for decidualization
Because EGR1 is a transcription factor, the EGR1-regulated transcriptome was profiled in pre-decidual T-HESC cells by ChIP-seq and RNA-seq. Analysis of the ChIP-seq dataset with the Galaxy Project’s Cis-regulatory Element Annotation System Tool revealed that EGR1 primarily binds within the 5’ untranslated region (UTR) or promoter of target genes (Supplemental Fig. S4A). The Galaxy/Cistrome SeqPos Motif Tool uncovered that the top enriched cluster of motifs found in the regions bound by EGR1 was comprised of EGR family G-C-rich motifs, including the EGR1 binding motif (motif ID: MC00320; Supplemental Fig. S4B). Integration of the EGR1 cistrome and transcriptome datasets with the list of genes whose expression is significantly changed during EPC-induced decidualization [11] revealed that EGR1 controls a remarkable number of genes that are involved in decidualization (Fig. 4A). Significantly, 40% of genes that are transcriptionally altered during EPC treatment (the decidual transcriptome) are also bound by EGR1 within 10 kb upstream to 10 kb downstream of their annotated gene location. Moreover, 19% of EPC-regulated genes are bound by EGR1 within 1 kb upstream to 1 kb downstream of their annotated gene location (Supplemental Fig. S4C). Remarkably, the number of genes bound by EGR1 at day 0 of EPC treatment is on par with the number of genes bound by the progesterone receptor (PGR, the master-regulator of decidualization) at day 3 of EPC treatment [20] (Supplemental Fig. S4D). Although the number of genes that are regulated by EGR1 at the transcriptional level (either directly or indirectly) may seem small compared to the decidual transcriptome, it still amounts to 651 genes. Out of these, 47% of genes are bound by EGR1 (Fig. 4A). The majority of the EGR1 target genes are positively regulated by EGR1 (i.e. knockdown of EGR1 leads to their downregulation) while genes negatively regulated by EGR1 are primarily indirect secondary targets of EGR1 (Fig. 4B). Not surprisingly, the direct targets of EGR1 have a more robust response to EGR1 knockdown as emphasized by higher enrichment of gene expression changes with lower false discovery rates (FDR) (Fig. 4C and Supplemental Fig. S5). When performing Gene Ontology analysis of direct and indirect targets of EGR1 that are also involved in decidualization (the 140 and 169 genes depicted in the Venn diagram in Fig. 4A), it is clear that direct and indirect target genes execute markedly different biological programs (Fig. 4D). While both direct and indirect targets are involved in cell cycle regulation and cell signaling, direct targets exhibit a strong enrichment for genes encoding transcriptional regulators.
Figure 4. The transcriptome regulated by EGR1 in T-HESC cells.
(A) Number of genes within and overlaps between the following datasets: 1) Genes bound by EGR1 (within 10 kb upstream to 10 kb downstream of annotated gene), 2) transcriptome controlled by EGR1 at day 0 of EPC treatment, and 3) transcriptome of genes altered by 3 days of EPC treatment [11]. (B) Percentage of all genes up- and downregulated by EGR1 knockdown which are either direct or indirect targets of EGR1. (C) Distribution of –log10(FDR) of all genes altered by EGR1 knockdown directly or indirectly. (D) Summaries of DAVID Functional Annotation Clustering analysis of genes which are altered during decidualization (3 days of EPC treatment) [11] and which are also either direct or indirect targets of EGR1 (the 140 and 169 genes shown in (4A)).
Because our results above underscored an important functional role for EGR1 in priming pre-decidual hESCs for later decidualization (Fig. 3), we next focused on all genes regulated by EGR1, irrespective of whether the expression of these genes changed with decidualization. Using the aforementioned approach, we again found direct and indirect genes of EGR1 controlled strikingly different biological programs (Fig. 5A and B). Specifically, genes directly regulated by EGR1 were primarily involved in regulation of transcriptional processes while indirect EGR1-target genes were involved in cell signaling. Both direct and indirect targets of EGR1 were enriched for genes involved in cell cycle regulation. The expression changes induced by EGR1 knockdown of some of the targets involved in these processes were confirmed by quantitative RT-PCR (Fig. 6). Collectively, our global cistromic and transcriptomic analyses indicate that EGR1 is essential for transcriptionally priming the pre-decidual hESC for imminent decidualization. Our analyses also reveal that many of the genes directly controlled by EGR1 encode proteins involved in harnessing the transcriptional machinery for cell division while others are key to cell signaling, cellular properties predicted to be acquired by a hESC that is primed for decidualization.
Figure 5. EGR1 controls transcriptional and signaling networks in pre-decidual cells.
(A) DAVID Functional Annotation Clustering analysis of direct EGR1 target genes and (B) indirect EGR1 target genes.
Figure 6. Validation of gene expression changes induced by EGR1 knockdown.
Quantitative RT-PCR of direct and indirect targets of EGR1 following siRNA-mediated knockdown of EGR1 before EPC treatment. CD163L1 – CD163 molecule like 1; FOXL1 – forkhead box L1; GFRA1 – GDNF family receptor alpha 1; GREB1L – GREB1 like retinoic acid receptor coactivator; ID1 – inhibitor of DNA binding 1, HLH protein; ID4 – inhibitor of DNA binding 4, HLH protein; POLQ – DNA polymerase theta; SLC16A6 – solute carrier family 16 member 6; SYT7 – synaptotagmin 7; UHRF1 – ubiquitin like with PHD and ring finger domains 1. Statistical analysis performed with two-sample t-test.
4. Discussion
When embryonic aneuploidy, chromosomal abnormalities, maternal thrombophilic disorders, and obvious anomalies in uterine anatomy are excluded as etiologic factors, implantation failure intrinsic to the uterus is commonly implicated in recurrent pregnancy loss [21]. Endometrial dysfunction also compromises the full potential of assisted reproductive technologies [22–24]. Superimposed on the above, approximately a third of pregnancies in healthy women – detected early by the human chorionic gonadotropin (hCG) assay – fail soon after hCG detection, highlighting failure of implantation in the uterus as a significant causal factor for preclinical pregnancy loss even for healthy couples [25–27]. That said, the progression of a programmed sequence of uterine cellular changes is known to underpin the establishment of pregnancy [28], with endometrial decidualization following embryo implantation representing a crucial cellular transformation step that is essential not only for the formation but also the progressive development of the fetomaternal interface [29]. Therefore, expanding our understanding of decidualization at the molecular level is crucial if we are to gain new mechanistic perspectives on this cellular process. Such insights may also uncover novel molecular targets that may contribute to new clinical solutions to address adverse fertility issues arising from a dysfunctional uterus.
Extending from our previous investigations [8], we report in this study that EGR1 protein is significantly expressed in pre-decidual hESCs; however, the levels of this protein are rapidly reduced with the onset of the decidual response. These results provide translational support for observations made here and those of others [5, 7] that show EGR1 is highly expressed in the subluminal stromal compartment of the murine endometrium during the window of receptivity; however, this expression significantly declines following embryo implantation. Results from these studies raised the obvious question: Do pre-decidual hESCs require EGR1 to decidualize? With the use of cultured hESCs, together with a siRNA-mediated knockdown approach, we clearly show that EGR1 protein expression in pre-decidual hESCs is essential for their subsequent decidualization. These findings support functional studies in the murine uterus, which show a compromised decidual response in the absence of EGR1 [7]. In addition, our results offer functional support for the proposal that perturbation of uterine EGR1 levels may be linked to endometrial pathologies associated with recurrent implantation failure.
The inability of the pre-decidual hESC to respond to the deciduogenic signal without the EGR1 transcription factor was pretext to molecularly phenotype this response impairment at the cistrome and transcriptome level. Analysis of the EGR1 cistrome dataset in pre-decidual hESCs disclosed that EGR1 binds within 10 kb upstream to 10 kb downstream of over 8000 genes; approximately 50% of these genes are bound within 1 kb upstream to 1 kb downstream. Integration with our previously published hESC decidual transcriptome dataset [11] showed that over 40% of the 8000 EGR1-bound genes exhibit changes in expression during hESC decidualization. This result may explain why abnormal persistent EGR1 expression blocks hESC decidualization following administration of EPC [8]. Normal expression of these genes during decidualization may be perturbed due to inappropriate binding of EGR1 to the regulatory elements of these genes. Interestingly, examination of the EGR1 transcriptomic data from pre-decidual hESCs revealed that EGR1 directly or indirectly regulates the expression of 65 genes prior to exposure of the deciduogenic signal. The biological processes primarily governed by these genes include transcriptional regulation, cell proliferation and signaling. In particular, controlling a wide array of other transcriptional regulators suggests that EGR1 can essentially prime the entire transcriptional milieu of the pre-decidual hESC prior to decidualization.
Collectively, our studies address an understudied area of hESC peri-implantation biology. We demonstrate that the EGR1 transcription factor is indispensable for the pre-decidual hESC to decidualize in response to requisite hormone signals. Genome-wide analysis supports that EGR1 transcriptionally primes the pre-decidual hESC for decidualization by controlling the expression of genes primarily involved in transcriptional regulation, cellular proliferation, and signaling. Future studies will establish the functional importance of these genes to hESC decidualization as well as identify the transcriptional and/or post-transcriptional and post-translational mechanisms that control EGR1 levels in hESCs following the onset of the decidual response.
Supplementary Material
(A) Rapid downregulation of EGR1 mRNA levels in T-HESC cells following 0-240 min of EPC exposure measured with quantitative RT-PCR. Statistical analysis performed with ANOVA with post hoc Tukey’s range test. (B) Validation of EGR1 antibod specificity confirmed siRNA-mediated knockdown of EGR1; NT – non-targeting (C) Examples of immunofluorescent detection of EGR1 expression levels in T-HESC cells at 0, 240, and 360 minutes and 24 hours of EPC treatment. Nuclei are stained with DAPI; anti-EGR1 antibody was detected with Alexa Fluor 594-conjugated antibody.
(A) Measurement of acetylation levels of histone H3 lysine 27 (H3K27Ac) by ChIP-qPCR in the EGR1 and EGR2 promoter and enhancer regions at day 0 and day 3 of EPC treatment. (B) Location of qPCR amplicons (marked with pink arrowheads) used for quantification shown in (A) in relation to the EGR1 and EGR2 genes and H3K27Ac enrichment (ENCODE consortium; Bernstein Lab dataset from 7 cell lines (GM12878, H1-hESC, HSMM, HUVEC, K562, NHEK, NHLF)). Binding of PLZF [8] and EGR1 from ChIP-seq datasets are also depicted. EGR1 and EGR2 gene directionality indicated with navy blue arrowheads.
(A) Immunohistochemical detection of incorporated BrdU in the luminal epithelium of the mouse uterus at GD1 (white arrowhead); LE and S denote luminal epithelium and stroma respectively. (B) Low levels of EGR1 expression in the glandular epithelium (GE) and stroma (S) of the murine endometrium at (GD1 (white arrowheads)). (C) Strong immunopositivity for BrdU incorporation in the subluminal stroma (white arrowhead) at GD4 (note: the switch from luminal to stromal proliferation is indicative of the receptive endometrium [29]. (D) Numerous stromal cells express EGR1 in the stromal compartment of the murine endometrium at GD4 (white arrowhead). However, note the absence of EGR1 expression in the luminal and glandular epithelial compartments at this time. (E) Low magnification image of the implanting embryo (E) at GD6. Panels (F) and (G) represent progressively higher magnification images of (E). Note: only few stromal cells surrounding the embryo are positive for EGR1 expression (white arrowhead). (H) Absence of EGR1 expression in decidualized stromal cells within the decidua (distant from the embryo) of the murine endometrium at GD6 (white arrowhead). All scale bars denote 100 μm: scale bar in (A) also applies to (B-D); scale bar in (G) applies to (H).
(A) Enrichment distribution for genomic annotations of the EGR1 cistrome presented as log(2) ratios to expected genomic distributions. (B) EGR1-binding motif found in the top cluster of enriched motifs in the EGR1 cistrome dataset. (C) Numbers of genes in and overlaps between datasets: 1) genes bound by EGR1 within 1 kb upstream to 1 kb downstream of annotated genes; within the extended promoter region (EPR; 7.5 kb upstream to 2.5 kb downstream from the TSS); and within 10 kb upstream to 10 kb downstream of annotated genes; and 2) genes transcriptionally altered by EPC exposure for 3 days [11]. (D) Numbers of genes in and overlaps between datasets: 1) genes bound by EGR1 within 10 kb upstream to 10 kb downstream of annotated genes; 2) genes bound by PGR within 25 kb/10 kb upstream to 25 kb/10 kb downstream of annotated genes [20]; and 3) genes transcriptionally altered by EPC exposure for 3 days [11].
Volcano plot of gene expression changes caused by siRNA-mediated knockdown of EGR1 of direct and indirect EGR1-target genes.
Highlights.
Decidualization of the endometrium is critical for pregnancy establishment
Endometrial EGR1 levels are decreased in women with recurrent implantation failure
Priming of endometrial cells for decidualization requires EGR1 function
The EGR1 transcriptome is predicted to prime endometrial cells to decidualize
5. Acknowledgments
The authors thank Jie Li, Yan Ying, and Rong Zhao for their essential technical expertise. We thank the Genomic & RNA Profiling Core Facility at Baylor College of Medicine (supported by NIH NCI grant (P30CA125123)) for performing the RNA-seq. We also acknowledge the expertise of the Integrated Microscopy Shared Resources (supported by P30 Cancer Center Support Grant (NCI-CA125123); P30 Digestive Disease Center (NIDDK-56338–13/15); CPRIT (RP150578); John S. Dunn Gulf Coast Consortium for Chemical Genomics). Studies described herein were funded in part by the Intramural Research Program of the National Institutes of Health (NIH)/National Institute of Environmental Health Sciences (NIEHS) Project Z1AES103311 to FJD and the NIH/National Institute of Child Health and Human Development (NICHD) grant HD-042311 to JPL.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declarations of interest: none
6. References
- [1].O’Donovan KJ, Tourtellotte WG, Millbrandt J, Baraban JM, The EGR family of transcription-regulatory factors: progress at the interface of molecular and systems neuroscience, Trends Neurosci, 22 (1999) 167–173. [DOI] [PubMed] [Google Scholar]
- [2].Lee SL, Sadovsky Y, Swirnoff AH, Polish JA, Goda P, Gavrilina G, Milbrandt J, Luteinizing hormone deficiency and female infertility in mice lacking the transcription factor NGFI-A (Egr-1), Science, 273 (1996) 1219–1221. [DOI] [PubMed] [Google Scholar]
- [3].Topilko P, Schneider-Maunoury S, Levi G, Trembleau A, Gourdji D, Driancourt MA, Rao CV, Charnay P, Multiple pituitary and ovarian defects in Krox-24 (NGFI-A, Egr-1)-targeted mice, Mol Endocrinol, 12 (1998) 107–122. [DOI] [PubMed] [Google Scholar]
- [4].Guo B, Tian XC, Li DD, Yang ZQ, Cao H, Zhang QL, Liu JX, Yue ZP, Expression, regulation and function of Egr1 during implantation and decidualization in mice, Cell Cycle, 13 (2014) 2626–2640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Liang XH, Deng WB, Li M, Zhao ZA, Wang TS, Feng XH, Cao YJ, Duan EK, Yang ZM, Egr1 protein acts downstream of estrogen-leukemia inhibitory factor (LIF)-STAT3 pathway and plays a role during implantation through targeting Wnt4, J Biol Chem, 289 (2014) 23534–23545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Kim HR, Kim YS, Yoon JA, Lyu SW, Shin H, Lim HJ, Hong SH, Lee DR, Song H, Egr1 is rapidly and transiently induced by estrogen and bisphenol A via activation of nuclear estrogen receptor-dependent ERK½ pathway in the uterus, Reprod Toxicol, 50 (2014) 60–67. [DOI] [PubMed] [Google Scholar]
- [7].Kim HR, Kim YS, Yoon JA, Yang SC, Park M, Seol DW, Lyu SW, Jun JH, Lim HJ, Lee DR, Song H, Estrogen induces EGR1 to fine-tune its actions on uterine epithelium by controlling PR signaling for successful embryo implantation, FASEB J, 32 (2018) 1184–1195. [DOI] [PubMed] [Google Scholar]
- [8].Kommagani R, Szwarc MM, Vasquez YM, Peavey MC, Mazur EC, Gibbons WE, Lanz RB, DeMayo FJ, Lydon JP, The Promyelocytic Leukemia Zinc Finger Transcription Factor Is Critical for Human Endometrial Stromal Cell Decidualization, PLoS Genet, 12 (2016) e1005937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Koot YE, van Hooff SR, Boomsma CM, van Leenen D, Groot Koerkamp MJ, Goddijn M, Eijkemans MJ, Fauser BC, Holstege FC, Macklon NS, An endometrial gene expression signature accurately predicts recurrent implantation failure after IVF, Sci Rep, 6 (2016) 19411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].WMA, WMA Declaration of Helsinki Serves as Guide to Physicians, Calif Med, 105 (1966) 149–150. [PMC free article] [PubMed] [Google Scholar]
- [11].Szwarc MM, Hai L, Gibbons WE, Peavey MC, White LD, Mo Q, Lonard DM, Kommagani R, Lanz RB, DeMayo FJ, Lydon JP, Human endometrial stromal cell decidualization requires transcriptional reprogramming by PLZF, Biol Reprod, 98 (2018) 15–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Krikun G, Mor G, Alvero A, Guller S, Schatz F, Sapi E, Rahman M, Caze R, Qumsiyeh M, Lockwood CJ, A novel immortalized human endometrial stromal cell line with normal progestational response, Endocrinology, 145 (2004) 2291–2296. [DOI] [PubMed] [Google Scholar]
- [13].Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR, STAR: ultrafast universal RNA-seq aligner, Bioinformatics, 29 (2013) 15–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Robinson MD, McCarthy DJ, Smyth GK, edgeR: a Bioconductor package for differential expression analysis of digital gene expression data, Bioinformatics, 26 (2010) 139–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Benjamini Y, Hochberg Y, Controlling the false discovery rate: a practical and powerful approach to multiple testing, Journal of the Royal Society, Series B (Methodological), 57 (1995) 289–300. [Google Scholar]
- [16].Huang da W, Sherman BT, Lempicki RA, Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources, Nat Protoc, 4 (2009) 44–57. [DOI] [PubMed] [Google Scholar]
- [17].Hai L, Szwarc MM, He B, Lonard DM, Kommagani R, DeMayo FJ, Lydon JP, Uterine function in the mouse requires speckle-type poz protein, Biol Reprod, 98 (2018) 856–869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Creyghton MP, Cheng AW, Welstead GG, Kooistra T, Carey BW, Steine EJ, Hanna J, Lodato MA, Frampton GM, Sharp PA, Boyer LA, Young RA, Jaenisch R, Histone H3K27ac separates active from poised enhancers and predicts developmental state, Proc Natl Acad Sci U S A, 107 (2010) 21931–21936. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Brosens JJ, Hayashi N, White JO, Progesterone receptor regulates decidual prolactin expression in differentiating human endometrial stromal cells, Endocrinology, 140 (1999) 4809–4820. [DOI] [PubMed] [Google Scholar]
- [20].Mazur EC, Vasquez YM, Li X, Kommagani R, Jiang L, Chen R, Lanz RB, Kovanci E, Gibbons WE, DeMayo FJ, Progesterone receptor transcriptome and cistrome in decidualized human endometrial stromal cells, Endocrinology, 156 (2015) 2239–2253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Revel A, Defective endometrial receptivity, Fertil Steril, 97 (2012) 1028–1032. [DOI] [PubMed] [Google Scholar]
- [22].Penzias AS, Recurrent IVF failure: other factors, Fertil Steril, 97 (2012) 1033–1038. [DOI] [PubMed] [Google Scholar]
- [23].Ruiz-Alonso M, Galindo N, Pellicer A, Simon C, What a difference two days make: “personalized” embryo transfer (pET) paradigm: a case report and pilot study, Hum Reprod, 29 (2014) 1244–1247. [DOI] [PubMed] [Google Scholar]
- [24].Koler M, Achache H, Tsafrir A, Smith Y, Revel A, Reich R, Disrupted gene pattern in patients with repeated in vitro fertilization (IVF) failure, Hum Reprod, 24 (2009) 2541–2548. [DOI] [PubMed] [Google Scholar]
- [25].Wilcox AJ, Weinberg CR, O’Connor JF, Baird DD, Schlatterer JP, Canfield RE, Armstrong EG, Nisula BC, Incidence of early loss of pregnancy, N Engl J Med, 319 (1988) 189–194. [DOI] [PubMed] [Google Scholar]
- [26].Norwitz ER, Schust DJ, Fisher SJ, Implantation and the survival of early pregnancy, N Engl J Med, 345 (2001) 1400–1408. [DOI] [PubMed] [Google Scholar]
- [27].Zinaman MJ, Clegg ED, Brown CC, O’Connor J, Selevan SG, Estimates of human fertility and pregnancy loss, Fertil Steril, 65 (1996) 503–509. [PubMed] [Google Scholar]
- [28].Gellersen B, Brosens JJ, Cyclic decidualization of the human endometrium in reproductive health and failure, Endocrine reviews, 35 (2014) 851–905. [DOI] [PubMed] [Google Scholar]
- [29].Cha J, Sun X, Dey SK, Mechanisms of implantation: strategies for successful pregnancy, Nat Med, 18 (2012) 1754–1767. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(A) Rapid downregulation of EGR1 mRNA levels in T-HESC cells following 0-240 min of EPC exposure measured with quantitative RT-PCR. Statistical analysis performed with ANOVA with post hoc Tukey’s range test. (B) Validation of EGR1 antibod specificity confirmed siRNA-mediated knockdown of EGR1; NT – non-targeting (C) Examples of immunofluorescent detection of EGR1 expression levels in T-HESC cells at 0, 240, and 360 minutes and 24 hours of EPC treatment. Nuclei are stained with DAPI; anti-EGR1 antibody was detected with Alexa Fluor 594-conjugated antibody.
(A) Measurement of acetylation levels of histone H3 lysine 27 (H3K27Ac) by ChIP-qPCR in the EGR1 and EGR2 promoter and enhancer regions at day 0 and day 3 of EPC treatment. (B) Location of qPCR amplicons (marked with pink arrowheads) used for quantification shown in (A) in relation to the EGR1 and EGR2 genes and H3K27Ac enrichment (ENCODE consortium; Bernstein Lab dataset from 7 cell lines (GM12878, H1-hESC, HSMM, HUVEC, K562, NHEK, NHLF)). Binding of PLZF [8] and EGR1 from ChIP-seq datasets are also depicted. EGR1 and EGR2 gene directionality indicated with navy blue arrowheads.
(A) Immunohistochemical detection of incorporated BrdU in the luminal epithelium of the mouse uterus at GD1 (white arrowhead); LE and S denote luminal epithelium and stroma respectively. (B) Low levels of EGR1 expression in the glandular epithelium (GE) and stroma (S) of the murine endometrium at (GD1 (white arrowheads)). (C) Strong immunopositivity for BrdU incorporation in the subluminal stroma (white arrowhead) at GD4 (note: the switch from luminal to stromal proliferation is indicative of the receptive endometrium [29]. (D) Numerous stromal cells express EGR1 in the stromal compartment of the murine endometrium at GD4 (white arrowhead). However, note the absence of EGR1 expression in the luminal and glandular epithelial compartments at this time. (E) Low magnification image of the implanting embryo (E) at GD6. Panels (F) and (G) represent progressively higher magnification images of (E). Note: only few stromal cells surrounding the embryo are positive for EGR1 expression (white arrowhead). (H) Absence of EGR1 expression in decidualized stromal cells within the decidua (distant from the embryo) of the murine endometrium at GD6 (white arrowhead). All scale bars denote 100 μm: scale bar in (A) also applies to (B-D); scale bar in (G) applies to (H).
(A) Enrichment distribution for genomic annotations of the EGR1 cistrome presented as log(2) ratios to expected genomic distributions. (B) EGR1-binding motif found in the top cluster of enriched motifs in the EGR1 cistrome dataset. (C) Numbers of genes in and overlaps between datasets: 1) genes bound by EGR1 within 1 kb upstream to 1 kb downstream of annotated genes; within the extended promoter region (EPR; 7.5 kb upstream to 2.5 kb downstream from the TSS); and within 10 kb upstream to 10 kb downstream of annotated genes; and 2) genes transcriptionally altered by EPC exposure for 3 days [11]. (D) Numbers of genes in and overlaps between datasets: 1) genes bound by EGR1 within 10 kb upstream to 10 kb downstream of annotated genes; 2) genes bound by PGR within 25 kb/10 kb upstream to 25 kb/10 kb downstream of annotated genes [20]; and 3) genes transcriptionally altered by EPC exposure for 3 days [11].
Volcano plot of gene expression changes caused by siRNA-mediated knockdown of EGR1 of direct and indirect EGR1-target genes.