Uterine glands coordinate on-time embryo implantation and impact endometrial decidualization for pregnancy success

Abstract

Uterine glands are essential for pregnancy establishment. By employing forkhead box A2 (FOXA2)-deficient mouse models coupled with leukemia inhibitory factor (LIF) repletion, we reveal definitive roles of uterine glands in embryo implantation and stromal cell decidualization. Here we report that LIF from the uterine glands initiates embryo-uterine communication, leading to embryo attachment and stromal cell decidualization. Detailed histological and molecular analyses discovered that implantation crypt formation does not involve uterine glands, but removal of the luminal epithelium is delayed and subsequent decidualization fails in LIF-replaced glandless but not gland-containing FOXA2-deficient mice. Adverse ripple effects of those dysregulated events in the glandless uterus result in embryo resorption and pregnancy failure. These studies provide evidence that uterine glands synchronize embryo-endometrial interactions, coordinate on-time embryo implantation, and impact stromal cell decidualization, thereby ensuring embryo viability, placental growth, and pregnancy success.

Fig. 1

The uterine transcriptome is dysregulated in mice that lack glands. RNA-sequencing was performed using uteri from gland-containing Ltf^iCre/+Foxa2^f/f glandless Pgr^Cre/+Foxa2^f/f, and control mice on GD 4. a Heatmap of the top 150 differentially expressed genes (log2 FPKM values) in FOXA2-deficient mice compared to controls. b Venn diagram comparing unique or common transcripts between the uterus of gland-containing Ltf^iCre/+Foxa2^f/f, glandless Pgr^Cre/+Foxa2^f/f, and control mice on GD 4. The LE vs GE represents genes that were enriched in the GE compared to the LE on GD 4. Superscripts denote genes combined for analysis. c Receptor and ligand interactions between genes uniquely differentially expressed in the uteri of glandless mice that are also enriched in GE on GD 4. Black boxes indicate genes uniquely differentially expressed in uteri of glandless Pgr^Cre/+Foxa2^f/f compared to control mice. Circles represent ligands, and half circles are receptors. All analysis was conducted using four biological replicates

Fig. 2

LIF initiates implantation in FOXA2-deficient mice. a In situ localization of Hbegf mRNA in the uterus of control and FOXA2-deficient mice without LIF repletion on GD 4 at 2200 hours. Uterine sections were counterstained with hematoxylin after chromogenic detection of Hbegf mRNA (red). Top panel—Scale bar: 500 μm; Bottom panel—Scale bars: 25 μm. b Gross morphology of uteri on GD 5 at 0800 hours. FOXA2-deficient mice received intraperitoneal (i.p.) injections of saline or recombinant mouse LIF on GD 4. Implantation sites accumulate Evans Blue Dye. White arrowheads point to individual implantation sites. Scale bar: 1 cm. c In situ localization of Hbegf mRNA in the uterus on GD 5 at 0800 hours. FOXA2-deficient mice received intraperitoneal (i.p.) injections of saline or recombinant mouse LIF on GD 4. Scale bar: 50 μm. AM: antimesometrial, M: mesometrial, Em: embryo, LE: luminal epithelium, S: stroma. All images are representative of three independent experiments

Fig. 3

Stromal cell decidualization and epithelial remodeling in LIF-replaced FOXA2-deficient mice. Control, gland-containing Ltf^iCre/+Foxa2^f/f and glandless Pgr^Cre/+Foxa2^f/f mice received i.p. injections of recombinant mouse LIF on GD 4 and were evaluated on GD 5 at 0800 hours. a Immunofluorescence analysis of PTGS2 and CK8 in implantation sites. PTGS2 is present in the stroma around the implanting embryo on the antimesometrial side of the uterus. Scale bar: 100 μm. b Immunofluorescence analysis of CDH1 in implantation sites. Scale bar: 100 μm. c Immunofluorescence analysis of CLDN1 in implantation sites. Scale bar: 100 μm. AM: antimesometrial, M: mesometrial, Em: embryo, LE: luminal epithelium, S: stroma. All images are representative of three independent experiments

Fig. 4

On-time removal of the LE is defective in LIF-replaced glandless Pgr^Cre/+Foxa2^f/f mice. Control, gland-containing Ltf^iCre/+Foxa2^f/f and glandless Pgr^Cre/+Foxa2^f/f mice received i.p. injections of recombinant mouse LIF on GD 4 and were evaluated at 2000 hours on GD 5. a Section of implantation sites on GD 5 were stained with hematoxylin and eosin. Representative images are shown for two independent mice representing observed variation. Columns 1 and 3—Scale bar: 250 μm; Columns 2 and 4—Scale bar, 50 μm. b Immunofluorescence analysis of CDH1 in implantation sites. Areas of breached epithelium are indicted with white arrowheads. Row 1—Scale bar: 100 μm; Row 2—Scale bar: 50 μm; Row 3—Scale bar: 10 μm. c Immunofluorescence analysis of CLDN1 in implantation sites. Areas of breached epithelium are indicted with white arrowheads. Scale bar: 50 μm. AM: antimesometrial, M: mesometrial, Em: embryo, LE: luminal epithelium. All images are representative of three independent experiments

Fig. 5

Uterine glands are essential for decidual progression and pregnancy establishment in mice. Control, gland-containing Ltf^iCre/+Foxa2^f/f and glandless Pgr^Cre/+Foxa2^f/f mice received i.p. injections of recombinant mouse LIF on GD 4 and were evaluated on GDs 6, 7, and 8. a Sections of implantation sites were stained with hematoxylin and eosin. Note the variation in embryo and decidua development on GDs 7 and 8 in the glandless Pgr^Cre/+Foxa2^f/f uterus. Row 1—Scale bar: 250 μm; Row 2—Scale bar: 150 μm; Row 3—Scale bar: 150 μm. b Immunofluorescence analysis of PTGS2 and Ki67 (cell proliferation marker) in implantation sites on GD 6. Scale bar: 100 μm. c Immunofluorescence analysis of PTGS2 and Ki67 (cell proliferation marker) in implantation sites on GD 7. Scale bar: 250 μm. d Gross morphology of the uterus on GD 8 (top panel). White arrowheads indicate individual implantation sites. Scale bar: 1 cm. TUNEL staining to detect apoptosis in implantation sites on GD 8 (bottom panel). Scale bar: 250 μm. Em: embryo, Dec: decidua, Pdz: primary decidual zone, Sdz: secondary decidual zone. All images are representative of three independent experiments

Fig. 6

The uterine transcriptome is dysregulated in the implantation sites of GD 6 mice that lack endometrial glands. RNA-sequencing analysis was conducted using implantation sites from LIF-replaced glandless and control mice on GD 6 (n = 4 per genotype). a Volcano plot of all genes detected in transcriptome analysis. All data points above the red horizontal line are significant. Known gland-specific genes have red points and are labeled with gene names. b Heatmap for 53 genes (log2 FPKM values) known to be involved in decidualization that were different in GD 6 implantation sites in LIF-replaced glandless Pgr^Cre/+Foxa2^f/f mice. c Visualization of biological process GO terms associated with genes differentially expressed in LIF-replaced glandless Pgr^Cre/+Foxa2^f/f and control mice on GD 6. Color indicates the FDR P value and size indicates the frequency of the GO term in the underlying annotation database. Darker circles with a larger size represent highly significant terms that are more general