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: 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 PgrCre/+Foxa2f/f mice. c Visualization of biological process GO terms associated with genes differentially expressed in LIF-replaced glandless PgrCre/+Foxa2f/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

Fig 4: Stromal cell decidualization and epithelial remodeling in LIF-replaced FOXA2-deficient mice. Control, gland-containing LtfiCre/+Foxa2f/f and glandless PgrCre/+Foxa2f/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 5: Pre-pubertal, unbranched glands fail to produce Lif even when ESR1 signaling is intact.(A) Uterine sections of P21 control mice injected with either oil, E2, or E2+P4 and stained for Hoechst (nuclei, white), FOXA2 (gland marker, green), ESR1 (a, b, c, red), and Lif mRNA (d, e, f, red). Scale bar, 40μm (a, b, c). Scale bar, 80μm (d, e, f). n=9 regions each from 3 mice per treatment. Quantitative analysis of (B) the amount of Lif normalized for gland volume and (C) the amount of Lif normalized for lumen volume. n=9 regions each from 3 mice per treatment. (B-C) Data analyzed using Mann-Whitney test. (*) = p<0.05. Yellow arrowheads point towards the luminal epithelial cells and white arrows points towards the glandular epithelial cells (D) Representative 3D reconstructions of P21 control uterine glands with various treatments. n=3 mice per treatment group. Scale bar, 100μm. Quantitative analyses of (E) average gland length measurements (each dot represents one mouse) and (F) percentage of glands with 0, 1-3, and >3 branches in P21 control mice with various treatments. Hormonal treatment does not affect gland length or gland branching. 85-450 glands analyzed per mouse. (E) Data analyzed using Mann-Whitney test. (*) = p<0.05. Two-proportion Z-test determined that the differences in percentage of gland branches between oil, E2, and E2+P4 treatment is not statistically significant.