Steroid Receptor Coactivator-2 Drives Epithelial Reprogramming That Enables Murine Embryo Implantation

Abstract

Although we have shown that steroid receptor coactivator-2 (SRC-2), a member of the p160/SRC family of transcriptional coregulators, is essential for decidualization of both human and murine endometrial stromal cells, SRC-2’s role in the earlier stages of the implantation process have not been adequately addressed. Using a conditional SRC-2 knockout mouse (SRC-2 ^d/d) in timed natural pregnancy studies, we show that endometrial SRC-2 is required for embryo attachment and adherence to the luminal epithelium. Implantation failure is associated with the persistent expression of Mucin 1 and E-cadherin on the apical surface and basolateral adherens junctions of the SRC-2 ^d/d luminal epithelium respectively. These findings indicate that the SRC-2 ^d/d luminal epithelium fails to exhibit a plasma membrane transformation (PMT) state known to be required for the development of uterine receptivity. Transcriptomics demonstrated that the expression of genes involved in steroid hormone control of uterine receptivity were significantly disrupted in the SRC-2 ^d/d endometrium as well as genes that control epithelial tight junctional biology and the emergence of the epithelial mesenchymal transition state, with the latter sharing similar biological properties with PMT. Collectively, these findings uncover a new role for endometrial SRC-2 in the induction of the luminal epithelial PMT state, which is a prerequisite for the development of uterine receptivity and early pregnancy establishment.

1 |. INTRODUCTION

Steroid receptor coactivator-2 (SRC-2), also known as NCOA2 (or TIF2 in the human; GRIP1 in the mouse), is a member of the evolutionary conserved p160/SRC family of transcriptional coactivators, which includes SRC-1 and SRC-3 ^{1, 2}. Similar to other family members, SRC-2’s complex functional domain structure underpins its pleiotropic coregulator roles in both nuclear receptor (NR) and non-NR mediated transcriptional responses ^{2, 3}. Such transcriptional responses are critical for a broad spectrum of physiological processes, from energy metabolism ^{4, 5}, circadian rhythm homeostasis ⁶, fetal pulmonary development ^{7, 8} to reproduction ³. Concerning the latter, early studies showed that whole body SRC-2 knockout male mice display a severe subfertility phenotype, encompassing defects in spermiogenesis (teratozoospermia), incomplete adhesion of Sertoli cells to germ cells, and age-dependent testicular degeneration ⁹. These studies also demonstrated that SRC-2 knockout female mice exhibit a severe hypofertility phenotype due to placental hypoplasia arising from developmental impairments in maternal decidual cells that are juxtaposed to the placenta ⁹. These findings provided early support for conclusions drawn from observational studies on human endometrial tissue ^{10, 11} and the rat uterus ¹² that endometrial SRC-2 is critical for ensuring both uterine function and health. In the case of the human endometrium, SRC-2 levels (along with SRC-3) do not markedly change with menstrual cycle phase ¹⁰; however, both SRCs exhibit significant increases in expression levels in epithelial and stromal cells of secretory-phase endometrial tissue when biopsied from patients diagnosed with polycystic ovarian syndrome (PCOS) ¹⁰. Apart from an increased susceptibility for endometrial cancer ^{10, 13}, PCOS patients are predisposed to additional reproductive sequelae, which include low cycle fecundity and a high miscarriage rate that can reach 60–70% ^{14, 15}. Collectively, these clinical studies indicate that perturbation in the homeostatic levels of SRC-2 may be causal for endometrial pathologies that lead to subfertility or infertility.

Using advanced conditional knockout technology, we revealed that postnatal abrogation of SRC-2 in a subset of cells expressing the progesterone receptor (PGR) results in infertility (rather than subfertility) in which the endometrium fails to enable embryo implantation, a necessary early step that leads to the establishment of the fetomaternal interface ^{16, 17}. From human to mouse, the embryo implantation process advances through distinct interdependent stages: starting with embryo attachment to the luminal epithelium of the receptive endometrium, embryo invasion into the underlying endometrial stroma, and decidualization of the surrounding stromal fibroblasts into specialized epithelioid decidual cells ¹⁸. Encircling the conceptus, decidual cells control the development of the embryo and protect against cytotoxic stressors. While there is compelling support for the importance of SRC-2 in decidualization of human endometrial stromal cells and stromal cells of the murine endometrium ^{16, 19, 20}, we have only a rudimentary understanding of endometrial SRC-2’s role during the earlier stages of the implantation process ¹⁶, particularly its influence on the endometrial epithelium. This knowledge insufficiency is significant since early implantation failure is not only causal for early pregnancy loss but undercuts the full potential of assisted reproductive technologies (ARTs) that rely on the transfer of healthy embryos into a receptive endometrium ^21–24.

Using the SRC-2 conditional knockout mouse, the timed pregnancy model, and transcriptomic analysis, we demonstrate that endometrial SRC-2 is essential for the development of the receptive endometrium, and, in particular, the reprogramming of the luminal epithelium to enable attachment of the embryo and its subsequent invasion into the underlying stroma.

2 |. MATERIALS AND METHODS

2.1 |. Mouse models

The SRC-2 conditional knockout mouse (SRC-2 ^d/d) was previously described ^{16, 25}. Briefly, the SRC-2 ^d/d bigenic was generated in a C57BL6J background by crossing our progesterone receptor cre knockin (Pgr ^cre/+) mouse ²⁶ with a mouse carrying the SRC-2 floxed [L2 version] allele (SRC-2 ^f/f) ⁹. In an AAALAC (Association for Assessment and Accreditation of Laboratory Animal Care) accredited vivarium facility at Baylor College of Medicine, mice were housed in temperature-controlled rooms (22 ± 2^oC) with a programmed 12-hour light/dark daily photocycle. Mice were fed an irradiated Formulab Diet (LabDiet/Lab Supply, Fort Worth, TX (#5008)) with access to fresh water ad libitum. Mice were treated humanely and surgical procedures were performed in accordance with the guidelines described in the Guide for the Care and Use of Laboratory Animals (“The Guide”, 8^th edition, 2011), published by the National Research Council of the National Academies, Washington D.C. (www.nap.edu). All animal procedures were prospectively approved by the Institutional Animal Care and Use Committee at Baylor College of Medicine.

2.2 |. Fertility studies

For timed pregnancy experiments, SRC-2 ^f/f and SRC-2 ^d/d females (8–10 weeks old) were housed overnight with fertility proven stud males. The following morning, coitus was visually confirmed in the female by the presence of a postcoital vaginal plug; the morning of detecting the vaginal plug was designated as the morning of gestation day 1 (GD1). Pregnant mice were individually housed before euthanasia on GD 4, 5, and 5.5 for the studies described below. Using uterine horns dissected from SRC-2 ^f/f and SRC-2 ^d/d females at GD4, embryo retrieval was performed by gently flushing the uterine lumen with 100 μl of sterile phosphate buffered saline (PBS). Retrieved embryos were digitally imaged within a petri dish using a color chilled AxioCam MRc5 digital camera with a Carl Zeiss AxioImager A1 upright microscope (Zeiss Inc., Jena, Germany). To visualize incipient implantation sites along both uterine horns of mice at GD5, Chicago sky blue dye (1% in PBS; 100μl per mouse) was injected into one lateral tail vein before mice were euthanized 2–5 minutes post-injection and their reproductive tract tissue dissected for macro-level digital imaging ²⁷. The latest versions of the Photoshop and Illustrator programs in the Adobe Creative Suite software package (Adobe Systems Inc. San Jose CA) were used for general raw image processing, generation of image composites, and the insertion of annotations in the final figures. For serum isolation, whole blood was collected in microtainer tubes containing a serum separator microguard (Becton, Dickinson and Company, Franklin Lakes, NJ (#365967)) from mice at GD5 as previously described ²⁷. Microtainer tubes containing drawn blood were kept at room temperature for 30 minutes to ensure complete blood coagulation before serum was separated by centrifugation at 2000 rpm for 10 minutes at 4^oC; serum was stored at −80 ^oC until analysis. Serum 17β-estradiol and progesterone levels were measured using the ENZO ELISA kit (#ADI-900–174 and #ADI-900–011 respectively; R&D Systems Inc., Minneapolis, MN).

2.3 |. General histology and immunohistochemical analyses

Tissues were fixed in 4% paraformaldehyde, progressively dehydrated using increased concentrations of ethanol, then xylene cleared before fixed tissues were paraffin embedded. Paraffin blocks were sectioned at 5μm thickness and resultant sections stained with hematoxylin and eosin (H&E) ²⁷. Immunohistochemical detection of SRC-2; mucin 1 (MUC 1); and Snail family zinc finger 1 (SNAI1) was conducted using the following primary rabbit polyclonal antibodies: NCOA2/SRC2 (Bethyl Laboratories, Waltham, MA #A300–345A, 1:500 dilution); MUC 1 (Abcam Inc., Waltham, MA #Ab15481, 1:100 dilution); and SNAI1 (ThermoFisher Scientific Inc., Carlsbad, CA #PA5–115940, 1:100 dilution). Following incubation with their respective primary antibody, tissue sections were incubated with a goat anti-rabbit secondary antibody conjugated with horseradish peroxidase (HRP; Vector Laboratories Inc., Burlingame, CA #P-1000, 1:200 dilution). Peroxidase activity was detected with the Vectastain Elite ABC-HRP kit (Vector Laboratories Inc. #PK-6100). Proliferating cells in S-phase of the cycle were visualized in tissue sections by immunohistochemical detection of 5-bromo-2’-deoxyuridine (BrdU) incorporation using the BrdU in situ detection kit (BD Biosciences, San Jose CA #551321). Following immunostaining, tissue sections were lightly counterstained with hematoxylin before applying Permount solution for placement of coverslips. To detect E-cadherin by immunofluorescence staining, the Alexa Fluor 488 conjugated donkey anti-rabbit secondary antibody (Invitrogen Inc., Carlsbad, CA #A-21206, 1:400 dilution) was used following incubation with a rabbit primary monoclonal antibody against mouse E-cadherin (Cell Signaling Technology Inc., Danvers, MA #3195, 1:200 dilution). In the case of claudin 7 (CLDN 7) immunofluorescence staining, the Alexa Fluor 594 donkey anti-rabbit secondary antibody (Invitrogen Inc., #A-21207, 1:400 dilution) was used following incubation with a rabbit polyclonal primary antibody against human CLDN 7 (Proteintech group Inc., Rosemont, IL, #10118–1-AP, 1:200 dilution). After incubation with the primary and secondary antibodies, slides were mounted with Vectashield Antifade mounting medium containing 4, 6-diamidino-2-phenylindole (DAPI; Vector Laboratories Inc. #H-1200–10). For certain immunohistochemical studies, the staining intensity of tissues was scored (H-score ^{28, 29}) on a scale of 0= no staining, 1=faint staining, 2= moderate staining, and 3-strong staining. Tissue sections (n=5) per genotype group (n=3) were analyzed three times by two independent investigators, blinded to the identities of the tissue sections. The average values from the biological replicates are presented as histograms.

2.4 |. Immunoblotting

Protein concentration was determined using the Bradford reagent (ThermoFisher Scientific Inc. #23225) before protein lysate preparations were resolved by a 10% polyacrylamide-SDS gel and transferred to a hydrophobic polyvinylidene difluoride (PVDF) membrane. After blocking PVDF membranes with 5% nonfat milk in Tris-buffered saline containing 0.1% Tween 20 (TBST), membranes were incubated with specific primary rabbit polyclonal antibodies, depending on the experiment. The following primary antibodies were used in these studies against the estrogen receptor (ESR1; Abcam Inc., Waltham MA #ab75635, 1:1000 dilution), the progesterone receptor (PGR; ThermoFisher Scientific Inc. #MA5–14505, 1:1000 dilution) or a mouse monoclonal primary antibody against β-actin (Novus Biologicals Inc., Centennial, CO #NB100–5874, 1:10,000 dilution) overnight at 4^oC. The following morning, immunoblots were washed before incubation with an anti-rabbit HRP-conjugated secondary antibody (1:5000) or an anti-mouse HRP-conjugated secondary antibody for 1 hour at room temperature. Chemiluminescence signal detection was achieved using the SuperSignal West Pico PLUS Chemiluminescent substrate (ThermoFisher Scientific Inc. #34580). Immunoreactive protein bands were imaged using the Azure 600 Imaging Systems (Azure Biosystems, Sierra Court, Dublin CA).

2.5 |. Quantitative real-time PCR

Total RNA was prepared from tissues using the RNeasy Plus Mini kit (Qiagen Inc., Germantown, MD #74134). The NanoDrop 2000 UV/Visual spectrophotometer (ThermoFisher Scientific Inc.) was used for quantification of RNA prior to reverse transcription using the High-Capacity cDNA Reverse Transcription Kit (ThermoFisher Scientific Inc. #4368814) ³⁰. Amplified cDNA was diluted to 10ng/μl before quantitative reverse transcription-PCR (qRT-PCR) was performed with the Fast TaqMan 2X Mastermix (Applied Biosystems/Life Technologies, Grand Island, NY #4352042); the TaqMan assays used in this study are listed in Table S1. All qRT-PCR experiments were performed using the 7500 Fast Real-Time PCR system (Applied Biosystems/ Life Technologies); delta-delta cycle threshold was used to normalize expression to the 18S reference.

2.6 |. Genome-wide RNA expression profiling

Genome scale RNA-sequencing (RNA-seq) and analysis were performed as previously described ^{31, 32}. Briefly, total RNA purity and physical integrity were determined using the NanoDrop spectrophotometer (ThermoFisher Scientific Inc.) and the 2100 Bioanalyzer with RNA chips (Agilent Technologies Inc., Santa Clara CA) respectively. Total RNA preparations that scored a RNA integrity number (RIN) of 7 or greater were used in these experiments; RNA samples from three mice were used for each genotype. From 250 ng of RNA, sequencing libraries were prepared using the TruSeq Stranded mRNA kit (Illumina Inc., San Diego CA #20020594) and then PCR amplified. Quality analysis of resultant libraries was performed using the 4200 TapeStation with D1000 ScreenTape assays (Illumina Inc.). Adapter-ligated fragment concentration was estimated by the qRT-PCR assay with a KAPA Library Quantification kit (KAPA Biosystems, Wilmington MA #KR0405). After equimolar pooling, libraries were quantified with the 2100 Bioanalyzer (using the High Sensitivity DNA Kit with DNA chips) and the KAPA Library Quantification Kit. Sequencing of libraries was achieved using the Illumina NovaSeq 6000 sequencer. Raw paired-end 100 base pair (bp) sequencing reads in Illumina fastq file format were generated at mid-output and mapped to the mouse genome (Genome Reference Consortium Mouse Build 39; National Center for Biotechnology Information (NCBI)) through use of hierarchical indexing for spliced alignment of transcripts (HISAT2) software ³³. The number of reads aligned to known genes was determined by the Python-based software package HTSeq ³⁴ (http://www-huber.embl.de/users/anders/HTSeq). To reduce possible PCR bias, read duplicates were removed with Picard Tools (http://broad.institute.github.10/picard).

The Bioconductor package DESeq was applied to the gene expression dataset to detect differentially expressed genes (DEGs) between the two groups ³⁵. The false discovery rate (FDR) for DEGs was estimated using the Benjamini and Hochberg method ³⁶. Gene expression comparisons with an FDR ≤ 0.05 and an absolute fold change (IFCI) ≥ 1.5 or 2 (depending on the analysis) were considered significantly differentially expressed between the two groups. Genes with significantly altered expression were used to identify affected pathways ³⁷. Fragments per kilobase of transcript per million (FPKM) values of transcripts were used for hierarchical clustering; the pheatmap package in R was used to generate the heatmap. Using raw gene count data, principal component analysis (PCA) was performed with the R function prcomp package (https://cran.r-project.org); raw data files were deposited in the Gene Expression Omnibus (GEO) repository at the NCBI (GSE237740; www.ncbi.nilm.go/geo). Gene ontology (GO) enrichment analysis was performed using the DAVID (Database for Annotation, Visualization, and Integrated Discovery) functional annotation clustering tool (https://david.abcc.ncifcrff.gov) ³⁸. Specific genes involved in major biological and signal transduction pathways were also determined using the Kyoto Encyclopedia of Genes and Genomes software (KEGG; http://www.genome.jp/kegg/). The Ingenuity Pathway Analysis (IPA) program (www.ingenuity.com) identified both upstream regulators and canonical pathways associated with genes of interest in the DEG dataset.

2.7 |. RNA-seq data correlations

Recently used by our group ³⁹, the Illumina correlation engine (http://www.illumina.com/products/by-type/informatics-products/basespace-correlationengine.html (formerly NextBio)) is a web-based RNA sequencing and microarray database that has curated over 23,000 scientific studies to enable correlation analysis. The differentially expressed gene (DEG) dataset from the SRC-2 ^d/d vs SRC-2 ^f/f (GD5) RNA-seq study described here was uploaded to the correlation engine and compared with biosets in the database using the Running Fisher test ⁴⁰. The biosets of interest represent curated DEG datasets from a number of published mouse models exhibiting suppressed progesterone responses and/or estrogen hypersensitivity within the murine endometrium ^41–46. The Illumina correlation engine provides an assessment of the statistical significance of the correlation of the overlapping genes between a given bioset (Bioset 2) and the DEG dataset described here (Bioset 1), with a summary p-value as a major readout (a large –log (p-value) indicates a high degree of similarity).

2.8 |. Statistical analyses

Two-tailed unpaired Student t-tests were used to estimate the statistical significance of differences between the two groups, and one-way ANOVA between more than two groups. Unless otherwise stated, data were graphically presented as mean ± standard deviation (S.D.). Differences with p-values < 0.05 were considered statistically significant; asterisks represent the level of significance: * p<0.05; **p<0.001, ***p<0.001; and ****p<0.0001. Prism software version 9 (GraphPad Software Inc., San Diego CA) was used for the majority of the reported statistical analyses.

3 |. RESULTS

3.1 |. Development of the receptive endometrium requires SRC-2 for early pregnancy establishment

Our previous studies demonstrated that the SRC-2 ^d/d female mouse is infertile ^{16, 25}, and that impairment in endometrial stromal decidualization (artificial decidual response) is at least one factor contributing to the SRC-2 ^d/d infertility phenotype ^{19, 20, 25}. Because SRC-2 is expressed in both the epithelial and stromal compartments of the murine endometrium (Figure S1) and because the epithelial compartment along with the stroma plays a crucial role in the development of the receptive endometrium ⁴⁷, we examined whether endometrial SRC-2 is required for the earlier stages of the implantation process using the timed natural pregnancy model. Figure 1A shows that the SRC-2 ^d/d female produces an equivalent number of pre-implantation blastocysts at GD4; however, embryo implantation fails to occur in the SRC-2 ^d/d female (Figure 1B, C). At GD5, we confirmed that normal ovarian function and associated 17β-estradiol and progesterone serum levels are detected in the SRC-2 ^d/d female (Figure S2A, B), confirming that the infertility phenotype is intrinsic to the SRC-2 ^d/d endometrium. We also found that the expression levels of the nuclear receptor for each steroid hormone are unchanged in the SRC-2 ^d/d uterus (Figure S2C).

FIGURE 1.

Embryo implantation is blocked in the SRC-2 ^d/d mouse. (A) An equivalent number of preimplantation blastocysts are produced by the SRC-2 ^f/f and SRC-2 ^d/d at GD4 (n = 5 mice per genotype); n. s. denotes not significant. (B) Tail vein injection of Chicago sky blue dye reveals the expected number of implantation sites in the SRC-2 ^f/f at GD5 (denoted by white arrowheads (O indicates the location of the ovaries)); the scale bar applies to both panels. The similarly treated SRC-2 ^d/d female does not show blue banding along the uterine horn as seen in the SRC-2 ^f/f control. The right panel shows recovered blastocysts from flushing one uterine horn of the SRC-2 ^d/d mouse at GD5; the result is representative of five mice. (C) By GD5.5, implantation sites can be visually detected (without the need for staining) along one of the two uterine horns of the SRC-2 ^f/f mouse at GD5.5 (top panel (black arrowheads)) whereas implantation sites are not present in the uterine horn of the SRC-2 ^d/d mouse (bottom panel); scale bar in the top panel applies to both panels. The histogram on the right graphically displays the number of implantation sites for each genotype at GD5.5 (n = 7 mice per genotype).

Immunohistochemical detection of BrdU incorporation revealed that while the SRC-2 ^f/f endometrium at GD4 display the typical cell type specific proliferative pattern for a receptive endometrium (i.e. an absence of luminal epithelial proliferation with a coincident increase in the number of proliferating cells within the subluminal stroma ⁴⁷ (Figure 2A)), the SRC-2 ^d/d endometrium failed to generate this endometrial receptive response within the context of a natural pregnancy state (Figure 2B, D, and E). Cell counts of BrdU positive luminal epithelial cells revealed a nearly 10-fold increase in the average number of BrdU positive cells in the SRC-2 ^d/d luminal epithelium when compared with the SRC-2 ^f/f luminal epithelial compartment at GD4 (Figure 2E). Furthermore, while the SRC-2 ^f/f endometrium at GD5 exhibits extensive proliferation of stromal cells prior to their decidualization 24 hours later, the SRC-2 ^d/d endometrium displays a significantly lower number of proliferating stromal cells (Figure S3A-G). The results from these timed natural pregnancy experiments correlate with our previous findings using ovariectomized mice that were administered an established 17β-estradiol plus progesterone hormone treatment regimen to artificially elicit uterine receptivity ^{19, 48}.

FIGURE 2.

Retention of a subset of proliferating luminal epithelial cells in the SRC-2 ^d/d endometrium at GD4. (A) Immunohistochemical detection of cells that are positive for BrdU incorporation in the endometrium of the SRC-2 ^f/f at GD4. Note the presence of proliferating stromal cells in the endometrium of the SRC-2 ^f/f mouse at GD4 (white arrowhead). In contrast, the luminal epithelium (LE) of the SRC-2 ^f/f endometrium is negative for BrdU immunopositivity. Together, the epithelial and stromal cellular proliferative profiles are indicative of a receptive uterus at GD4 ⁴⁷. (B) A subset of luminal epithelial cells in the SRC-2 ^d/d endometrium at GD4 is consistently immunopositive for BrdU incorporation (black arrowhead); a subgroup of subluminal stromal cells is positive for BrdU incorporation (white arrowhead). Panels (C) and (D) are higher power magnifications of regions shown in panels (A) and (B) respectively. Again note the absence and presence of BrdU immunopositive luminal epithelial cells in the SRC-2 ^f/f and SRC-2 ^d/d endometrium respectively. Although qualitative, BrdU positive stromal cells in the SRC-2 ^f/f endometrium consistently register a stronger immunopositive signal than BrdU positive stromal cells in the SRC-2 ^d/d endometrium, compare the stromal (S) compartment in (C) with (D). Scale bar in (A) and (C) apply to (B) and (D) respectively. (E) Histogram graphically displays the average number of BrdU positive cells per 100 luminal epithelial cells counted from each of three separate tissue sections per mouse (4 mice per genotype).

The aberrant retention of a subset of proliferating cells in the luminal epithelial compartment of the SRC-2 ^d/d endometrium at GD4 (Figure 2B, D, and E) suggests that other cellular properties of this epithelial compartment, which are required for normal embryo implantation, may also be compromised. Immunohistochemical detection of mucin 1 (MUC 1 ^{49, 50}), the expression of which is attenuated in the endometrium of the normal murine endometrium at GD4 (Figure 3A ^{49, 50}), remains significantly elevated in the SRC-2 ^d/d endometrium at GD4 and 5 (Figure 3A-F). Furthermore, immunofluorescence investigations confirmed that E-cadherin expression is absent in the basolateral junctional regions of luminal epithelial cells of the SRC-2 ^f/f endometrium that closely contact the trophectodermal cells of the implanting embryo (Figure 3G, I). The suppression of E-cadherin expression in these cellular regions is known as one cellular alteration that signals the emergence of a plasma membrane transformation (PMT) state ^{51, 52}. Suppression of E-cadherin expression normally serves to loosen the tight intercellular adherens junctions within the epithelial compartment to facilitate embryo invasion into the underlying stroma ^52–54. However, E-cadherin expression persists in the junctional zones of the SRC-2 ^d/d epithelium at GD5 (Figure 3H, J). Together, these results suggest that absence of SRC-2 in the SRC-2 ^d/d uterus results in disruption of epithelial reprogramming that underpins the development of the PMT state, which is a prerequisite for the generation of a receptive uterus for embryo implantation.

FIGURE 3.

Luminal epithelial (LE) cell marker expression is altered in the SRC-2 ^d/d endometrium. (A) An endometrial tissue section obtained from a SRC-2 ^f/f mouse at GD4; section is immunohistochemically stained for MUC 1 expression. Note the low levels of MUC 1 expression on the apical surface of the LE (white arrowhead). (B) Expression of MUC 1 is noticeably stronger on the apical surface of the LE of the SRC-2 ^d/d endometrium at GD4 (white arrowhead). (C) By GD5, MUC 1 expression is absent in the endometrium of the SRC-2 ^f/f mouse; E indicates embryo. (D) At GD5, the SRC-2 ^d/d endometrium still retains strong MUC 1 immunopositivity on the apical surface of the LE (white arrowhead) despite the presence of an embryo. (E) and (F) show higher power magnification images shown in (C) and (D) respectively. Again note the absence of MUC 1 expression in the LE of the SRC-2 ^f/f endometrium at GD5 and the continued presence of MUC 1 expression on the apical surface of the LE compartment within the SRC-2 ^d/d endometrium (white arrowhead). (G) Immunofluorescence detection of E-cadherin in the epithelium of the SRC-2 ^f/f endometrium at GD5; E denotes embryo. Note E-cadherin expression is specific to the epithelial chamber of the SRC-2 ^f/f implantation site. (I) Higher power magnification image of region shown in (G). Note that E-cadherin expression is present in the trophectoderm of the embryo (black arrowhead) whereas there is significantly less E-cadherin expression in the basolateral regions of epithelial cells that are juxtaposed to the embryo (white arrowhead). (I) Immunofluorescence detection of the E-cadherin protein in the SRC-2 ^d/d endometrium at GD5. As in (G), E-cadherin immunopositivity is specifically localized to the epithelial compartment. (J) Higher power magnification image of a region shown in (H). Note the retention of E-cadherin expression in the apical and basolateral regions of the epithelial cells (white arrowhead); trophoblast cells of the embryo are positive for E-cadherin expression (black arrowhead). Scale bar in (G) and (I) apply to (H) and (J) respectively.

3.2 |. Abrogation of SRC-2 adversely alters the endometrial transcriptome required for embryo implantation

Because the embryo implantation process fails to advance in the SRC-2 ^d/d mouse at GD5 (Figure 1), coupled with the fact that SRC-2 is a coregulator of gene transcription ², we used RNA-seq to identify genes for which normal transcription is dysregulated in the SRC-2 ^d/d uterus at GD5 (Figure 4). Therefore, the overall experimental design entailed bulk RNA-seq profiling of uterine tissue dissected from SRC-2 ^f/f control and SRC-2 ^d/d mutant mice at GD5, with triplicate tissue samples per genotype. The complete list of genes that are differentially expressed between the two genotypes is shown in the Excel table in the supplementary section (Supplementary Folder S1). The expression of a total number of 3,685 genes (1,452 downregulated and 2,233 upregulated) was identified as significantly changed (Figure 4A). The FKPM values for all 3,685 genes were analyzed by PCA (Supplementary Folder S1 (PCA tab)). The PCA confirmed that the SRC-2 ^f/f and SRC-2 ^d/d groups were significantly segregated with respect to their group triplicates. The total number of DEGs between the two genotypes, which reached the FDR (≤ 0.05) and FC (≥ 2) threshold parameters, was 1,685 genes (497 downregulated and 1188 upregulated) (Figure 4B). Applying the agglomeration and GO applications in DAVID, genes in the DEG list were grouped according to GO terms (Figure 4C). Note the enrichment for genes associated with cellular properties linked with the apical plasma membrane. In addition, the KEGG pathway annotation software revealed that protein families involved in tight junction biology received one of the highest scores in terms of the number of genes assigned to a given biological process between the two groups (Figure 4D). Ranking based on p-values and z-scores, IPA identified an overrepresented gene group in the DEG list that is significantly associated with EMT cellular processes (Table 1).

FIGURE 4.

Genome-wide alteration of the transcriptome in the SRC-2 ^d/d uterus at GD5. (A) The volcano plot graphically displays the total number of genes differentially expressed between the SRC-2 ^f/f and SRC-2 ^d/d groups. Significance (p-value) versus fold change are plotted on the y- and x-axis respectively. The total number of genes differentially expressed between the two groups is 3,685 (1,452 and 2,233 downregulated and upregulated respectively). (B) The heatmap represents an unsupervised hierarchical clustergram of the total number of genes differentially expressed between the SRC-2 ^f/f and SRC-2 ^d/d groups that reached FDR and FC thresholds of ≤ 0.05 and ≥ 2 respectively. Each row represents a gene while each column denotes a sample replicate in the SRC-2 ^f/f or SRC-2 ^d/d group; the dendrogram on both axes show the arrangement of the clusters following analysis. The intensity of the color indicates the level of expression for each gene in the sample replicate. (C) Dot plot of enriched genes within the DEG dataset are stratified according to biological processes, the cellular component module is shown. Note apical and basolateral plasma membrane properties feature prominently in this analysis. (D) The KEGG analysis shows enrichment for tight junction biology (bold) for proteins encoded by genes for which expression is significantly changed between the SRC-2 ^f/f and SRC-2 ^d/d groups.

TABLE 1.

Ingenuity Canonical Pathways identified by IPA analysis in differentially expressed genes between SRC-2^f/f and SRC-2^d/d mouse uteri at GD5

Ingenuity Canonical Pathways	p- value	Z-Score	Molecules
Regulation Of The Epithelial Mesenchymal Transition In Development Pathway	6.46E-04	−2.529822128	BCL9,CDH1,FZD1,FZD9,GLIS1,HNF1A,LEF1,mir-8,REL,SNAI1,WNT2,WNT6,WNT7B
Antioxidant Action of Vitamin C	1.05E-04	−2.496150883	GSTO1,JAK2,MAPK11,MAPK13,NFKBIA,NOTUM,PLA2G10,PLA2G2D,PLA2G2E,PLA2G4F,PLA2G5,PLB1,PLCH1,PLCH2,REL,SLC23A3, STAT5A
Basal Cell Carcinoma Signaling	1.38E-03	−2.121320344	BMP2,BMP5,BMP8A,FZD1,FZD9,GLIS1,HNF1A,LEF1,WNT2,WNT6,WNT7B
BEX2 Signaling Pathway	1.23E-03	−2.110579412	HNF1A,ITGB6,KLK3,LEF1,MMP2,NFKBIA,PDGFC,PGF,PPP2R2B,PPP2R2C,REL,SPP1
Pulmonary Healing Signaling Pathway	1.48E-04	−2.041241452	BMPR1B,CDH1,DLK1,FGFR1,FRK,FZD1,FZD9,HNF1A,LEF1,MAPK11,MAPK13,MMP2,MMP3,NFKBIA,NOTCH4,PDGFC,PGF,PRKCG, RASD1,SMAD9, STAT5A,WNT2,WNT6,WNT7B
Colorectal Cancer Metastasis Signaling	1.29E-03	−1.963961012	ADCY2,CDH1,DCC,FZD1,FZD9,GNG12,GNG4,HNF1A,JAK2,LEF1,MMP2,MMP3,NOS2,PDGFC,PGF,PIK3C2G,PIK3R3,RAC3,RASD1,REL, RHOBTB2, RHOV, TLR1,Tlr12,WNT2,WNT6,WNT7B
S100 Family Signaling Pathway	1.29E-10	−1.843908891	ADCYAP1R1,ADGRA1,ADGRF1,ADGRG7,ADGRV1,ADORA1,ADORA2B,ADRA1A,ADRA1B,ADRA1D,ADRA2B,ANXA1,BDNF,BTC, CACNA1B, CACNA2D3,CDH1,CELSR1,CELSR2,CTSD,DRD4,EDNRB,EREG,ESR2,ETV4,F2R,F2RL3, FGFR1,FGFR3,FZD1,FZD9,GALR3, GPR160,GPRC5A,GPRC5C,HCAR2,HNF1A,HTR1D,HTR6,IFNK,Ighg2b,KLK3,LEF1,LPAR3,MAPK11,MAPK13,MMP2,MMP3,NAPSA,NMUR2, NOS2,NOTUM,NPY2R,OXTR,P2RY14, PDGFC,PGF,PIK3C2G, PIK3R3,PLA2G10,PLA2G2D,PLA2G2E,PLA2G4F,PLA2G5,PLCH1,PLCH2, PRKCG,PTGDR,PTGIR,RAC3,RXFP1,S100A14,S100A3, S100A6,S100A8,S1PR3,SMAD6,SMAD9,TACR1,TRPV5,VDR,VIPR2,WNT2,WNT6, WNT7B
Factors Promoting Cardiogenesis in Vertebrates	1.29E-03	−1.414213562	BMP2,BMP5,BMP8A,BMPR1B,FZD1,FZD9,HNF1A,LEF1,NOG,NOTUM,NPPA,PLCH1,PLCH2,PRKCG,SMAD9,WNT2,WNT6,WNT7B
Pulmonary Fibrosis Idiopathic Signaling Pathway	2.69E-05	−1.333333333	CDH1,COL13A1,COL22A1,COL25A1,COL26A1,Col6a4,COL6A5,COL6A6,EFNB2,F2R,FGF18,FGFR1,FGFR3,FZD1,FZD9,HNF1A,IL11, IL13RA1,ITGB6,JAK2, LEF1,MAPK11,MAPK13,MMP2,MMP3,NOTCH4,PDGFC,PIK3C2G,PIK3R3,PLAU, RASD1,REL,SNAI1,WNT2,WNT6, WNT7B
WNT/β-catenin Signaling	5.50E-06	−1.091089451	BCL9,CDH1,CDH5,CDKN2A,DKK3,DKKL1,FRZB,FZD1,FZD9,GJA1,HNF1A,KREMEN2,LEF1,PPP2R2B,PPP2R2C,RARB,SFRP1,SFRP2, SFRP4,SFRP5,SOX7, SOX9,WNT2,WNT6,WNT7B

3.3 |. Gene expression programs associated with murine uterine receptivity and epithelial cellular properties are significantly altered in the SRC-2 ^d/d mouse

Gene ontology enrichment scores along with focused biological, cellular, and molecular pathway analytics of DEGs between the SRC-2 ^f/f and SRC-2 ^d/d groups at GD5 revealed an overrepresentation of upregulated estrogen responsive genes in the SRC-2 ^d/d uterus (Figure 5A ⁵⁵). In parallel, a downregulation of progesterone responsive genes is predicted by this analysis to occur in the SRC-2 ^d/d uterus at GD5 (Figure 5A ⁵⁶). In addition, genes previously associated with uterine decidualization are predicted to be markedly downregulated in the SRC-2 ^d/d uterus at GD5 (Figure 5A ⁵⁶). Interestingly, genes associated with epithelial cell tight junctions and EMT are predicted to be upregulated and downregulated respectively in the SRC-2 ^d/d uterus at GD5 (Figure 5A ⁵²). In Figure 5B, results from quantitative real-time PCR analysis confirm the predicted expression changes for a number of the genes (shown in Figure 5A) in the SRC-2 ^d/d uterus at GD5. Interestingly, the expression of a subset of these genes are also altered in the SRC-2 ^d/d uterus at GD4 (Figure S4), supporting an important role for SRC-2 in the early expression changes of these genes during the development of the receptive uterus. Using the Illumina correlation engine (Figure 6 and Figure S5), we revealed that a significant number of genes in the 3111 DEG dataset (SRC-2 ^d/d vs SRC-2 ^f/f (GD5)) is present in DEGs reported for the PGR Activation Function 1 (AF1) domain mutant mouse (Pgr ^AF1/AF1, 453 genes) ⁴⁵; Indian hedgehog (Ihh ^d/d, 361 genes) ⁴²; SRY-box transcription factor 17 (Sox17 ^d/d, 978 genes) ⁴⁶; Bone Morphogenetic protein 2 (Bmp2 ^d/d, 225 genes) ⁴⁴; wingless-type MMTV integration site family, member 4 (Wnt4 ^d/d, 806 genes) ⁴³; Sirtuin 1 (Sirt1 ^d/d, 1600 genes) ⁴¹; forkhead box A2 (Foxa2 ^d/d, 617 genes) ⁵⁷; and the estradiol treated ovariectomized mouse (1317 genes (GSE 23241)). The correlative analysis revealed that the majority of overlapped genes followed a concordant (or positive) gene expression profile (both upregulated and downregulated): 92% with Pgr ^AF1/AF1; 81% with Ihh ^d/d; 63% with Sox17 ^d/d; 92% with Bmp2 ^d/d; 83% Wnt4 ^d/d; 94% Sirt1 ^d/d; 85% Foxa2 ^d/d; and 70% with the estradiol-treated ovariectomized mouse model. Together, our results indicate that SRC-2 is required for the progesterone induction of genes required to establish uterine receptivity ⁵⁵, a subset of which are known to suppress estrogen responsive gene expression ⁵⁸, a prerequisite for uterine receptivity. The results also support an important role for SRC-2 in endometrial epithelial reprogramming that normally manifests as a loosening of intercellular tight junctional connections and a triggering of a PMT cellular response ⁵², both of which are important for adherence between the apical surface of the luminal epithelium and the blastocyst and subsequent blastocyst invasion into the underlying stroma. As further support for this proposal, a number of genes—SNAI1 ^{59, 60} and CLDN7 ^{61, 62}—associated with the EMT and PMT processes were shown to be deregulated in the SRC-2 ^d/d uterus at GD 5 (Figure S6 and S7).

FIGURE 5.

Aberrant expression of genes in the SRC-2 ^d/d uterus that are involved in uterine receptivity. (A) Heat map of gene expression changes between the SRC-2 ^f/f and SRC-2 ^d/d replicates predict an increase in the expression of estrogen responsive genes (Cadherin 1 (Cdh1); Leukemia inhibitory factor (Lif); Lactoferrin (Ltf); and Mucin 1 (Muc1)), a decrease in expression of progesterone responsive genes (Cytochrome P450 family 1 subfamily B member 1 (Cyp1b1); Follistatin (Fst); Coagulation F2 receptor (F2r); Heart and neural crest derivatives expressed 2 (Hand2); Homeobox A10 (Hoxa10); Laminin subunit gamma 3 (Lamc3); Proprotein convertase subtilisin kexin type 5 (Pcsk5); Regulator of G protein signaling 2 (Rgs2)), a decrease in expression of established decidual cell markers (Bone morphogenetic protein 2 (Bmp2); Insulin-like growth factor binding protein 1 (Igfbp1); Prolactin family 8, subfamily a, member 2 (Prl8a2); and Prolactin family 3, subfamily C, member 1 (Prl3c1)), an increase in expression of tight junctional markers (Cystic fibrosis transmembrane conductance regulator (Cftr); Crumbs cell polarity complex component 3 (Crb3); Claudin 7 (Cldn7); Lorelei-like-GPI-anchored protein 2 (Llg2); Marvel domain containing 2 (Marveld2); Molecule interacting with casl-like 2 (Micall2); Protein phosphatase 2, regulatory subunit B, beta (Ppp2r2b); Solute carrier family 9, member 3, regulator 1 (Slc9a3r1); and (Tight junction protein 3 (Tjp3)), and a decrease in expression of EMT markers (Frizzled class receptor 1 (Fzd1); Frizzled class receptor 9 (Fzd9); Glis family zinc finger 1 (Glis1); Hepatocyte nuclear factor 1 alpha (Hnf1a); Lymphoid enhancer binding factor 1 (Lef1); Snail family transcriptional repressor 1 (Snai1); Wnt family member 2 (Wnt2); Wnt family member 6 (Wnt6); and Wnt family member 7b (Wnt7b)). Asterisks denote genes chosen for further expression validation by qPCR shown below. (B) Quantitative real time PCR results for a selection of genes with asterisks listed in the heat maps shown in panel (A). Important to note: the RNA samples used in these quantitative real time PCR experiments were separate from the RNA samples used for the RNA-seq study; n= 3 per genotype.

FIGURE 6.

Significant correlation in the expression of a gene subset in the SRC-2 ^d/d versus SRC-2 ^f/f DEG dataset and in uterine DEG datasets from mutant mouse exhibiting progesterone suppression and concomitant estrogen hypersensitivity. The Illumina correlation engine was used to determine the significance of positive correlation in gene expression changes (up (red histogram) and down (green histogram)) between gene subsets in the SRC-2 ^d/d vs SRC-2 ^f/f DEG dataset and in DEG datasets reported for uterine tissue from four mutant mouse models that exhibit a progesterone suppression and estrogen hypersensitivity phenotype ^{42, 44–46}.

DISCUSSION

In both the human and mouse, pregnancy establishment relies on the implantation of a high quality blastocyst into a receptive endometrium in which the latter’s luminal epithelium is the first contact site for the blastocyst ^{17, 63}. To facilitate blastocyst attachment and adhesion to the luminal epithelium, the epithelium undergoes a set of cellular and molecular changes, many of which promote a PMT state that is crucial for embryo implantation ^{51, 52}. For uterine receptivity to manifest, apical alterations of the luminal epithelium occur in which its extracellular glycoprotein composition is changed and cell-surface charge reduced to enable embryo attachment and adherence ^{49, 64}. In parallel, apical, basolateral, and planar polarity of luminal epithelial cells is lost, resulting in a weakening of lateral epithelial cell surface interactions as well as a disruption of focal adhesions between the epithelium and basal lamina. These intercellular contact alterations are essential for the penetration of embryonic trophectodermal cells through the luminal epithelium into the underlying stroma ^65–67. In the case of the mouse, one study has shown that luminal epithelial cells are lost at the contact site by entosis soon after blastocyst attachment ⁶⁸. Notwithstanding these advances in our understanding, the identification of the key regulatory signals that control the emergence of endometrial receptivity in general and in the luminal epithelial PMT state in particular remains incomplete.

Applying the timed pregnancy model to the SRC-2 ^d/d mouse, we report that SRC-2 is required for the development of endometrial receptivity. Examination of established markers of uterine receptivity revealed that signals associated with the PMT state are significantly derailed in the SRC-2 ^d/d endometrium at GD5. For example, downregulation of MUC 1 expression at the apical surface of the luminal epithelium did not occur in the SRC-2 ^d/d endometrium at GD5. A prominent constituent of the glycocalyx layer on the apical surface of the luminal epithelium of the human endometrium ⁶⁹, MUC 1 is known to exhibit anti-adhesive properties on the apical surface of the luminal epithelium ⁶⁴. In vitro studies have also shown that MUC 1 expression is suppressed in primary endometrial epithelial cells at the site of human blastocyst attachment ⁷⁰ while clinical investigations have revealed a link between abnormal endometrial MUC 1 levels and implantation failure ⁷¹. With parallels to the human, MUC 1 expression was shown to be markedly reduced in the luminal epithelium as the murine endometrium reaches the receptive state ^{49, 50}. The downregulation of E-cadherin at adherens junctions of luminal epithelial cells, another PMT signal, is also compromised in the SRC-2 ^d/d endometrium at GD5. As a result of blocking E-cadherin downregulation, embryo penetration to the stroma is prevented due to a persistently polarized SRC-2 ^d/d luminal epithelium with intact tight intercellular junctions.

Transcriptomics uncovered a broader spectrum of genes associated with steroid hormone control of uterine receptivity, decidualization, and tight junctional and EMT biology. In keeping with SRC-2’s role in PGR-mediated signaling ^{16, 19, 25, 72–76}, the induction in expression of a significant number of progesterone responsive genes is attenuated in the SRC-2 ^d/d endometrium at GD5. A number of these genes (i.e. Hand 2) have been shown to act as paracrine signals from the subepithelial stroma to the luminal epithelium to facilitate progesterone suppression of estrogen-induced signaling in the luminal epithelium ⁵⁸. The switch in luminal epithelial proliferative status accompanies the change in the structural and functional properties of this cellular compartment, which leads to PMT and uterine receptivity. The persistent expression of a subset of estrogen responsive genes in the SRC-2 ^d/d luminal epithelium indicates an unwarranted heightened estrogen response, which is known to be linked to implantation failure in the mouse ^77–79 and indirectly in the human ^{80, 81}.

In addition to the stroma, SRC-2 is expressed in the epithelia of the human ¹⁰ and murine endometrium (Figure S1), suggesting that SRC-2 directly or indirectly regulates EMT processes during the periimplantation period; our transcriptomics analysis showed EMT is a biological process significantly altered in the SRC-2 ^d/d uterus at GD5. Interestingly, EMT shares similar properties with PMT such as loss of apical-basolateral polarity, a breakdown of cell-cell junctions, and an increase in extracellular matrix degradation ⁵². Together, these findings provide compelling support for epithelial-derived SRC-2 in directly controlling the PMT state that determines progression of the epithelium to a receptive phenotype for embryo attachment and adherence. In the case of the mouse, this proposal will be addressed in the future through selective SRC-2 ablation in the endometrial epithelial compartment using cre/lox methods ⁸².

Our previous studies have established stromal-derived SRC-2 as a critical coregulator in progesterone-dependent decidualization and pregnancy establishment ^{19, 20, 25}. Using the timed natural pregnancy model, we reveal here that SRC-2 is pivotal for the PMT changes that must occur in the luminal epithelium to enable embryo attachment and adherence within the context of a receptive uterus. Given the cell-type specific roles of SRC-2 in endometrial biology and pathobiology ^{10, 19, 20, 25, 83, 84} and the knowledge-gap associated with early pregnancy establishment, our findings provide a foundation for further investigations into the specific involvement of this coregulator in the endometrial epithelium in ensuring normal endometrial function and health.

Abbreviations:

AAALAC

Association for assessment and accreditation of laboratory animal care

ANOVA

Analysis of variance

ARTs

Assisted reproductive technologies

ATCC

American type culture collection

BrdU

5-bromo-2’-deoxyuridine

CL

Corpus luteum

DAPI

4, 6-diamidino-2-phenylindole

DAVID

Database for Annotation, Visualization, and Integrated Discovery

DEGs

Differentially expressed genes

DPBS

Dulbecco’s phosphate buffered saline

EMT

Epithelial mesenchymal transition

FBS

Fetal bovine serum

FC

Fold change

FDR

False discovery rate

FPKM

Fragments per kilobase of transcript per million

GD

Gestation day

GRIP1

Glucocorticoid receptor interacting protein 1

GEO

Gene expression omnibus

GO

Gene ontology

GE

Glandular epithelium

H&E

Hematoxylin and eosin

Hand2

Heart and neural crest derivatives expressed 2

HISAT2

hierarchical indexing for spliced alignment of transcripts 2

HRP

Horse radish peroxidase

IPA

Ingenuity pathway analysis

LE

Luminal epithelium

KEGG

Kyoto Encyclopedia of Genes and Genomes

MUC 1

Mucin 1

NCBI

National center for biotechnology information

NCOA

Nuclear receptor coactivator

NR

Nuclear receptor

NT

Non-targeting

PCA

Principal component analysis

PGR

Progesterone receptor

PVDF

Polyvinylidene difluoride

PMT

Plasma membrane transformation

qRT-PCT

quantitative real-time PCR

RIN

RNA integrity number

RNA-seq

RNA sequencing

RPM

Revolutions per minute

S

Stroma

SD

Standard deviation

SDS

Sodium dodecyl sulfate

SNAI1

Snail family zinc finger 1

SRC

Steroid receptor coactivator

TBST

Tris-buffered saline with Tween

TIF2

Transcriptional intermediary factor 2

DATA AVAILABILITY STATEMENT

All data reported in the manuscript and support information is publically available. The RNA-seq data described in this manuscript has been deposited into GEO (accession number: GSE237740).