Endometrial NCOR1 deficiency contributes to implantation failure in endometriosis-associated infertility

J.-W.J. was responsible for the concept of the study. S.L.Y. and B.A.L. collected human samples. L.T.K.N., D.N.T., T.H.K., and H.R.K. carried out experiments. L.T.K.N., D.N.T., T.H.K., R.A., and J.-W.J. analyzed data. J.A. provided transgenic mice. L.T.K.N., D.N.T., J.-Y.Y., T.H.K., and J.-W.J. prepared the manuscript. All authors contributed to editing the manuscript.

We examined the expression of NCOR1 in the eutopic endometrium from fertile women and infertile women with endometriosis using immunohistochemistry ( Figures 1 A and 1B). NCOR1 was highly expressed in the endometrial stromal and epithelial cells of controls. However, the eutopic endometrium from infertile patients with endometriosis revealed significantly decreased levels of NCOR1 in epithelial (2.63 fold, p = 0.0037) and stromal (2.17 fold, p = 0.0338) cells compared to controls. These results suggest that reduced NCOR1 expression in the human endometrium is associated with endometriosis-associated infertility. Figure 1 NCOR1 is decreased in eutopic endometrium from infertile women and mice with endometriosis (A and B) Representative immunohistochemistry images and (B) H-score quantification of NCOR1 protein expression in eutopic endometrium from fertile women without endometriosis ( n = 13) and infertile women with endometriosis ( n = 18). Each n represents an independent patient sample. Images are representative of the analyzed samples. (C and D) Representative immunohistochemistry images and (D) H-score quantification of NCOR1 protein expression in eutopic endometrium of sham-operated mice and mice with induced endometriosis at 1 month and 3 months after induction, assessed at GD 3.5 ( n = 6 per group per time point). Each n represents an individual mouse. Images are representative of independent biological replicates. Data are presented as mean ± SEM. Statistical significance was determined by two-tailed unpaired Student’s t test (B) or one-way ANOVA followed by Tukey’s post hoc test (D). ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗p < 0.001. NCOR1 is decreased in eutopic endometrium from infertile women and mice with endometriosis (A and B) Representative immunohistochemistry images and (B) H-score quantification of NCOR1 protein expression in eutopic endometrium from fertile women without endometriosis ( n = 13) and infertile women with endometriosis ( n = 18). Each n represents an independent patient sample. Images are representative of the analyzed samples. (C and D) Representative immunohistochemistry images and (D) H-score quantification of NCOR1 protein expression in eutopic endometrium of sham-operated mice and mice with induced endometriosis at 1 month and 3 months after induction, assessed at GD 3.5 ( n = 6 per group per time point). Each n represents an individual mouse. Images are representative of independent biological replicates. Data are presented as mean ± SEM. Statistical significance was determined by two-tailed unpaired Student’s t test (B) or one-way ANOVA followed by Tukey’s post hoc test (D). ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗p < 0.001. To determine whether NCOR1 expression is altered following the development of endometriosis, endometriosis was surgically induced in 2-month-old female Pgr cre/+ Rosa26 mTmG/+ mice 13 using a validated autologous uterine tissue transplantation model. One month after induction, successful establishment of endometriosis was confirmed by the presence of GFP-positive ectopic lesions, and the lesions were further validated by histological analysis demonstrating endometrial glands and stroma. Immunohistochemistry was performed to assess NCOR1 expression in the eutopic endometrium of sham-operated and endometriosis-induced mice ( Figures 1 C and 1D). Interestingly, NCOR1 levels were significantly reduced in the endometrial epithelial (1.16 fold, p = 0.0072) cells of the eutopic endometrium from mice with endometriosis compared to the sham group at 3 months after endometriosis induction. Moreover, NCOR1 levels were significantly lower in the stromal cells of the eutopic endometrium from mice with endometriosis compared to the sham control group at 1 month and 3 months after endometriosis induction (1.17 fold, p = 0.0042 and 1.68 fold, p < 0.001, respectively). These results indicate that the development and progression of endometriosis are associated with reduced NCOR1 expression in the eutopic endometrium. To determine the effect of NCOR1 loss on endometriosis development and uterine function, we generated a mouse model with Ncor1 conditionally ablated in Pgr -positive cells ( Pgr cre/+ Ncor1 f/f ; Ncor1 d/d ). Our immunohistochemical analysis validated the ablation of Ncor1 in the uterus that NCOR1 levels were significantly reduced in the endometrium of Ncor1 d/d mice compared to Ncor1 f/f mice at GD 3.5 ( Figure 2 A). Next, endometriosis was surgically induced in control Pgr cre/+ Rosa26 mTmG/+ and Ncor1 d/d Rosa26 mTmG/+ mice by inoculating endometrial tissues into the peritoneum. One month after endometriosis induction, GFP-positive ectopic lesions were examined in mice with Ncor1 f/f and Ncor1 d/d ectopic lesions. We found that the number and weight of ectopic lesions from mice with Ncor1 d/d were significantly increased (1.55 fold, p = 0.0460 and 3.21 fold, p = 0.0063, respectively) compared to those in mice with control ectopic lesions ( Figure 2 B). Figure 2 Loss of Ncor1 promotes endometriosis development and causes severe subfertility in mice (A) Validation of uterine Ncor1 ablation by immunohistochemistry in the uteri of Ncor1 f/f and Ncor1 d/d mice at GD3.5 ( n = 5 mice per genotype). Each n represents an individual mouse. Images are representative of independent biological samples. (B) Representative fluorescence images and quantification of ectopic lesion number in Ncor1 f/f and Ncor1 d/d mice one month after endometriosis induction ( n = 5 mice per genotype). Each n represents an individual mouse. Ectopic lesions are identified by GFP-positive tissue outside of uterus. White arrows indicate ectopic lesions. Data are presented as Mean ± SEM. Statistical significance was determined by two-tailed unpaired Student’s t test. ∗p < 0.05 and ∗∗ p < 0.01. (C) Cumulative number of pups produced over a 6-month fertility trial in Ncor1 f/f and Ncor1 d/d female mice ( n = 7 females per genotype). Each n represents an individual female mouse continuously housed with wild-type males. Loss of Ncor1 promotes endometriosis development and causes severe subfertility in mice (A) Validation of uterine Ncor1 ablation by immunohistochemistry in the uteri of Ncor1 f/f and Ncor1 d/d mice at GD3.5 ( n = 5 mice per genotype). Each n represents an individual mouse. Images are representative of independent biological samples. (B) Representative fluorescence images and quantification of ectopic lesion number in Ncor1 f/f and Ncor1 d/d mice one month after endometriosis induction ( n = 5 mice per genotype). Each n represents an individual mouse. Ectopic lesions are identified by GFP-positive tissue outside of uterus. White arrows indicate ectopic lesions. Data are presented as Mean ± SEM. Statistical significance was determined by two-tailed unpaired Student’s t test. ∗p < 0.05 and ∗∗ p < 0.01. (C) Cumulative number of pups produced over a 6-month fertility trial in Ncor1 f/f and Ncor1 d/d female mice ( n = 7 females per genotype). Each n represents an individual female mouse continuously housed with wild-type males. Before investigating the effect of NCOR1 loss in the uterus, we examined the cell-specific and temporal expression profiles of NCOR1 in mice during early pregnancy (GD 0.5 to 7.5). Our immunohistochemical analysis showed dynamic expression of NCOR1 protein during early pregnancy ( Figure S1 ). At GD 0.5, NCOR1 protein was weakly detected in the epithelial and stromal cells of the endometrium in wild-type mice. From GD 2.5 to GD 3.5, NCOR1 protein levels significantly increased in both endometrial epithelial (2.13 fold, p < 0.0001 and 2.18 fold, p < 0.001, respectively) and stromal cells (1.95 fold, p = 0.0017 and 2.14 fold, p = 0.001, respectively) ( Figure S1 B and S1C). At the implantation stage GD 4.5, NCOR1 was detected weakly in both the luminal epithelium and decidualizing stroma cells surrounding implantation sites. At GD 5.5, NCOR1 was present in the primary decidual zone (PDZ) and secondary decidual zone (SDZ). At GD 7.5, NCOR1 protein levels decreased in the PDZ, but significantly increased in the SDZ ( Figure S1 A). To explore the impact of Ncor1 ablation on female fertility, we performed a fertility test over a 6-month period where Ncor1 f/f and Ncor1 d/d female mice were mated with wild-type male mice ( n = 7) ( Figure 2 C). Ncor1 f/f female mice produced an average of 7.28 ± 1.23 pups per litter in an average of 4.86 ± 0.34 litters per mouse. However, six of seven Ncor1 d/d mice (85.7%) were infertile, and one Ncor1 d/d mouse produced a single litter. These results suggest that uterine Ncor1 loss is associated with severe subfertility. In order to determine whether Ncor1 d/d subfertility might result from ovarian defects, we examined ovarian function by measuring levels of serum E2 and P4, as well as assessing ovarian histology of Ncor1 d/d mice ( n = 6). The serum levels of E2 and P4 showed no significant difference between Ncor1 f/f (E2 = 4.60 ± 0.30 pg/mL; P4 = 11.79 ± 1.66 ng/mL) and Ncor1 d/d (E2 = 4.94 ± 0.25 pg/mL; P4 = 9.42 ± 1.29 ng/mL) (E2, p = 0.4529; P4, p = 0 . 3278 ) mice at GD 3.5 ( Figure S2 A). We also confirmed normal ovarian morphology in Ncor1 f/f and Ncor1 d/d mice ( Figure S2 B). In addition, Ncor1 d/d mice showed normal uterine architecture and gross morphology with no significant differences in uterine weight ( p = 0.0623), uterine-to-body weight ratio ( p = 0.2452), and the ability of fertilized embryos to develop to the blastocyst stage at GD 3.5 ( p = 0.2588) ( Figures S2 C and S2D). These findings demonstrated normal ovarian function as well as uterine development in Ncor1 d/d female mice, suggesting that subfertility of Ncor1 d/d female mice is due to uterine dysfunction. To assess the effect of NCOR1 loss during early pregnancy, ultrasonography was performed at GD 5.5, GD 7.5, and GD 9.5 in Ncor1 f/f and Ncor1 d/d mice ( Figures 3 A and 3B). The number and size of embryos were analyzed on the ultrasonography results during early pregnancy. Although there was no difference in the number of embryos between Ncor1 f/f and Ncor1 d/d mice at GD 5.5, the sizes of embryos from Ncor1 d/d mice (2.24 ± 0.22 mm 3 ) were significantly smaller ( p = 0 . 0170 ), showing a 1.34-fold reduction compared to Ncor1 f/f mice (1.67 ± 0.21 mm 3 ) ( Figure 3 C). Thereafter, the sizes of embryos in Ncor1 f/f mice gradually increased at GD 7.5 (8.94 ± 0.27 mm 3 ) and GD 9.5 (22.94 ± 0.43 mm 3 ). However, no embryos were detected in Ncor1 d/d mice from GD 7.5 to GD 9.5 ( Figures 3 C and 3D). These results suggest that the deletion of Ncor1 leads to early pregnancy loss. Figure 3 Ncor1 d/d mice exhibit early pregnancy loss detected by ultrasound imaging (A) Representative two-dimensional (2D) grayscale ultrasound images of pregnant Ncor1 f/f and Ncor1 d/d mice at GD 5.5, GD 7.5, and GD 9.5 ( n = 5 per genotype per time point). Each n represents an individual pregnant mouse. Images are representative of independent biological samples. (B) Representative three-dimensional (3D) reconstructions of implantation sites generated from ultrasound imaging of Ncor1 f/f and Ncor1 d/d mice. (C) Quantification of embryo volume bases on 3D reconstruction of implantation sites from Ncor1 f/f and Ncor1 d/d mice ( n = 5 mice per genotype per time point). Each n represents an individual mouse. (D) Representative gross uterine images of Ncor1 f/f and Ncor1 d/d mice at GD 9.5. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗ p < 0 . 05 . Ncor1 d/d mice exhibit early pregnancy loss detected by ultrasound imaging (A) Representative two-dimensional (2D) grayscale ultrasound images of pregnant Ncor1 f/f and Ncor1 d/d mice at GD 5.5, GD 7.5, and GD 9.5 ( n = 5 per genotype per time point). Each n represents an individual pregnant mouse. Images are representative of independent biological samples. (B) Representative three-dimensional (3D) reconstructions of implantation sites generated from ultrasound imaging of Ncor1 f/f and Ncor1 d/d mice. (C) Quantification of embryo volume bases on 3D reconstruction of implantation sites from Ncor1 f/f and Ncor1 d/d mice ( n = 5 mice per genotype per time point). Each n represents an individual mouse. (D) Representative gross uterine images of Ncor1 f/f and Ncor1 d/d mice at GD 9.5. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗ p < 0 . 05 . To determine the cause of early pregnancy loss, we performed histological analysis at implantation sites in female Ncor1 f/f and Ncor1 d/d mice. Although Ncor1 f/f mice exhibited normal implantation sites at GD 7.5, no implantation sites were detected in Ncor1 d/d mice at GD7.5 ( Figure 4 A). At an earlier time point (GD5.5), implantation sites in Ncor1 d/d mice could be divided into two groups: group #1, which had smaller implantation sites, and group #2, which had none ( Figure 4 B). Further histological analysis was performed on uteri from Ncor1 f/f and Ncor1 d/d mice at GD 5.5. In group #1 (3/6 mice), the uterine lumen was not closed, the implantation chamber was formed, but the embryos were resorbed. In group #2 (3/6 mice), the uterine lumen remained open, and the embryos were not detected within the uterine lumen ( Figure 4 C). Figure 4 Ncor1 d/d mice exhibit implantation failure (A) Representative gross uterine morphology of Ncor1 f/f and Ncor1 d/d mice at GD 7.5 ( n = 3 mice per genotype). (B) Representative gross uterine morphology of Ncor1 f/f ( n = 5 mice) and Ncor1 d/d mice at GD 5.5 ( n = 6 mice total; three mice per subgroup). (C) Representative histological analysis of implantation sites in uteri from Ncor1 f/f ( n = 5 mice) and Ncor1 d/d ( n = 6 mice total; three mice per subgroup) mice at GD 5.5. Arrows indicate implantation sites. (D) Analysis of implantation chamber morphology in whole-mount stained uteri at GD 5.5 using Hoechst and CDH1. Control uteri exhibit V-shaped implantation chambers, whereas Ncor1 d/d uteri display U-shaped or abnormal W-shaped implantation chambers. Yellow inset shows enlarged W-shaped chamber. Yellow arrowheads indicate persistence of an open lumen at the mesometrial pole in Ncor1 d/d mice. Scale bars, 100 μm. Images are representative of independent biological replicates ( n = 3–4 mice per genotype). Ncor1 d/d mice exhibit implantation failure (A) Representative gross uterine morphology of Ncor1 f/f and Ncor1 d/d mice at GD 7.5 ( n = 3 mice per genotype). (B) Representative gross uterine morphology of Ncor1 f/f ( n = 5 mice) and Ncor1 d/d mice at GD 5.5 ( n = 6 mice total; three mice per subgroup). (C) Representative histological analysis of implantation sites in uteri from Ncor1 f/f ( n = 5 mice) and Ncor1 d/d ( n = 6 mice total; three mice per subgroup) mice at GD 5.5. Arrows indicate implantation sites. (D) Analysis of implantation chamber morphology in whole-mount stained uteri at GD 5.5 using Hoechst and CDH1. Control uteri exhibit V-shaped implantation chambers, whereas Ncor1 d/d uteri display U-shaped or abnormal W-shaped implantation chambers. Yellow inset shows enlarged W-shaped chamber. Yellow arrowheads indicate persistence of an open lumen at the mesometrial pole in Ncor1 d/d mice. Scale bars, 100 μm. Images are representative of independent biological replicates ( n = 3–4 mice per genotype). To examine the three-dimensional structure of defective implantation sites from Ncor1 d/d mice, uteri with decidual swellings from both control and group #1 of Ncor1 d/d mice were stained whole mount for implantation chamber analysis at GD5.5 ( Figure 4 D). Implantation chambers in control mice were V-shaped, and embryos were observed in all chambers (7/7 chambers, n = 3 mice). Implantation chambers in the Ncor1 d/d uteri were often devoid of embryos (9/13 chambers, n = 4 mice). The implantation chambers in Ncor1 d/d uteri were U or W shaped and the lumen at the mesometrial pole was open. Defective growth of the implantation chamber may restrict the growth of the embryo beyond this stage in the Ncor1 d/d mice. Next, implantation sites were assessed at GD 4.5 in female Ncor1 d/d mice compared to control Ncor1 f/f mice. Implantation sites were visualized as blue bands along the uterus following intravenous injection of Chicago blue in the mice ( Figure 5 A). While 50% of Ncor1 d/d mice (4/8) exhibited a normal number of implantation sites (8.24 ± 0.20; group #1), the remaining 50% of Ncor1 d/d mice (4/8) showed no Chicago blue-stained implantation sites (group #2). Next, more detailed histological analysis revealed the presence of blastocysts in all Ncor1 f/f mice. While embryos in group #1 of Ncor1 d/d mice were found closely opposed to the luminal epithelium and surrounded by decidualized cells, embryos in group #2 of Ncor1 d/d mice were not attached but were instead floating within the uterine lumen. Furthermore, the luminal epithelium in group #2 of Ncor1 d/d mice remained unenclosed. Figure 5 Ncor1 d/d mice exhibit implantation defects at GD 4.5 (A) Visualization of implantation sites following intravenous injection of Chicago Sky Blue 6B dye and representative histological analysis of implantation sites in Ncor1 f/f and Ncor1 d/d mice at GD 4.5 Ncor1 f/f mice ( n = 7), group #1 Ncor1 d/d mice ( n = 4), and group #2 Ncor1 d/d mice ( n = 4). Each n represents an individual pregnant mouse. Images are representative of independent biological samples. (B and C) Representative immunohistochemistry analyses of COX-2 and pSTAT3 in the uteri from Ncor1 f/f and Ncor1 d/d mice at GD 4.5. Arrows indicate implantation sites. Images are representative of independent biological replicates. (D) H-score quantification of pSTAT3 protein expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 5), and group #2 Ncor1 d/d mice ( n = 5). Each n represents an individual mouse. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗∗ p < 0.01, ∗∗∗ p < 0.001. Ncor1 d/d mice exhibit implantation defects at GD 4.5 (A) Visualization of implantation sites following intravenous injection of Chicago Sky Blue 6B dye and representative histological analysis of implantation sites in Ncor1 f/f and Ncor1 d/d mice at GD 4.5 Ncor1 f/f mice ( n = 7), group #1 Ncor1 d/d mice ( n = 4), and group #2 Ncor1 d/d mice ( n = 4). Each n represents an individual pregnant mouse. Images are representative of independent biological samples. (B and C) Representative immunohistochemistry analyses of COX-2 and pSTAT3 in the uteri from Ncor1 f/f and Ncor1 d/d mice at GD 4.5. Arrows indicate implantation sites. Images are representative of independent biological replicates. (D) H-score quantification of pSTAT3 protein expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 5), and group #2 Ncor1 d/d mice ( n = 5). Each n represents an individual mouse. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗∗ p < 0.01, ∗∗∗ p < 0.001. The uterine decidualization markers COX-2 32 and pSTAT3 33 , 34 were examined by immunohistochemistry at GD 4.5 to determine whether Ncor1 d/d mice altered the expression of COX-2 and pSTAT3 in decidual cells near implantation sites. COX-2 protein was strongly expressed in the decidual stromal cells surrounding the embryos in uteri of Ncor1 f/f mice and group #1 of Ncor1 d/d mice ( Figure 5 B). However, COX-2 protein was not detected in luminal epithelium and implantation sites in group #2 of Ncor1 d/d mice. The expression of pSTAT3 was not significantly disrupted in the uterine epithelium of the group #1 ( p = 0.6643) and #2 ( p = 0.9728) of Ncor1 d/d mice compared to the Ncor1 f/f mice ( Figure 5 C). Interestingly, the expression of pSTAT3 was significantly decreased in the stromal uterine cells of the group #2 of Ncor1 d/d mice compared to the Ncor1 f/f and group #1 of Ncor1 d/d mice (2.16 fold, p = 0.0009 and 2.04 fold, p = 0.0021, respectively) ( Figure 5 D). These results indicate that subfertility observed in Ncor1 d/d mice is accompanied by implantation failure. The spatiotemporal expression and localization of PGR and FOXO1 in the endometrial epithelium are important for successful implantation during pregnancy in both humans and mice. 35 Therefore, we next assessed the expression of PGR and FOXO1 at implantation sites on GD 4.5 to determine the cause of implantation failure in Ncor1 d/d mice. Immunohistochemistry revealed that PGR was absent in the luminal epithelium and strongly expressed in the stromal compartment of the uterus in Ncor1 f/f mice. In contrast, PGR remained localized in the nuclei of epithelial cells in both group #1 and group #2 of Ncor1 d/d mice ( Figure 6 A). Quantitative analysis showed significantly higher levels of PGR expression in the luminal epithelium of both group #1 and group #2 of Ncor1 d/d mice compared to Ncor1 f/f mice (13.72 fold, p = 0.0484 and 15.91 fold, p = 0.0213, respectively) ( Figure 6 B). In addition, the expression of PGR in stromal cells of the group #2 of Ncor1 d/d mice was significantly reduced compared to the Ncor1 f/f mice (1.84 fold, p = 0.0017 and 1.59 fold, p = 0.0195, respectively) ( Figure 6 B). The maintenance of PGR expression in epithelial cells at early implantation blocks embryo implantation, as embryos are unable to attach to the luminal epithelium or undergo the stromal decidual response. 36 Furthermore, the downregulation of epithelial PGR expression activates FOXO1 expression, which translocated from the cytoplasm to the nucleus of endometrial epithelium during the implantation window in mice. 35 We observed that FOXO1 was located to the nucleus of endometrial epithelium in Ncor1 f/f mice, whereas it was abnormally retained in the cytoplasm of endometrial epithelium in group #1 of Ncor1 d/d mice ( Figure 6 A). Interestingly, FOXO1 expression was significantly lower in the endometrial epithelium of group #2 of Ncor1 d/d mice compared to Ncor1 f/f (1.56 fold, p = 0.0002) and group #1 of Ncor1 d/d (1.63 fold, p = 0.0022) mice ( Figure 6 C). Additionally, the expression of FOXO1 in the stromal cells of group #1 of Ncor1 d/d mice was significantly higher than Ncor1 f/f and group #2 of Ncor1 d/d mice (6.18 fold, p = 0.0017 and 2.78 fold, p = 0.0195, respectively) ( Figure 6 C). Figure 6 Ncor1 d/d mice display dysregulation of PGR and FOXO1 at the implantation sites (A) Representative immunohistochemistry analyses for PGR and FOXO1 at the implantation sites in Ncor1 f/f and Ncor1 d/d mice at GD 4.5. Images are representative of independent biological samples ( n = 5 mice per group). (B and C) H-score quantification of PGR and FOXO1 protein expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 5), and group #2 Ncor1 d/d mice ( n = 5). Each n represents an individual pregnant mouse. (D) Representative double immunofluorescence analysis of PGR (green) and FOXO1 (red) expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 5), and group #2 Ncor1 d/d mice ( n = 5) at GD 4.5. Nuclei were counterstained with DAPI (blue). Images are representative of independent biological replicates. (E) Quantification of epithelial and stromal expression of PGR, FOXO1, and PGR/FOXO1 co-localized cells (merge). Each n represents an individual mouse ( n = 5 per group). Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. Ncor1 d/d mice display dysregulation of PGR and FOXO1 at the implantation sites (A) Representative immunohistochemistry analyses for PGR and FOXO1 at the implantation sites in Ncor1 f/f and Ncor1 d/d mice at GD 4.5. Images are representative of independent biological samples ( n = 5 mice per group). (B and C) H-score quantification of PGR and FOXO1 protein expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 5), and group #2 Ncor1 d/d mice ( n = 5). Each n represents an individual pregnant mouse. (D) Representative double immunofluorescence analysis of PGR (green) and FOXO1 (red) expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 5), and group #2 Ncor1 d/d mice ( n = 5) at GD 4.5. Nuclei were counterstained with DAPI (blue). Images are representative of independent biological replicates. (E) Quantification of epithelial and stromal expression of PGR, FOXO1, and PGR/FOXO1 co-localized cells (merge). Each n represents an individual mouse ( n = 5 per group). Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. To examine the spatial relationship between PGR and FOXO1, we performed double immunofluorescence staining in the uteri of Ncor1 f/f and Ncor1 d/d mice at GD 4.5 ( Figures 6 D and 6E). In Ncor1 f/f uteri, FOXO1 expression was predominantly detected in the epithelium (84.35 ± 6.70%), whereas epithelial PGR expression was minimal (0.23 ± 0.08%). In contrast, PGR expression was strongly detected in the stromal compartment (28.66 ± 1.91%), while stromal FOXO1 expression was low (0.14 ± 0.14%). In the epithelium of group #1 Ncor1 d/d mice, FOXO1 remained the predominant signal (71.94 ± 5.61%). However, PGR expression (12.61 ± 5.26%) and limited co-localization of PGR and FOXO1 (2.99 ± 1.16%) were detected. In the stroma of group #1 Ncor1 d/d mice, PGR and FOXO1 were detected at 22.36 ± 3.21% and 10.88 ± 3.94%, respectively, with minimal co-localization (0.76 ± 0.28%). In the epithelium of group #2 Ncor1 d/d mice, PGR and FOXO1 were detected at comparable levels (13.63 ± 1.64% and 12.61 ± 1.57%, respectively), with low co-localization (3.71 ± 0.96%). In the stroma of group #2 Ncor1 d/d mice, PGR and FOXO1 were detected at 6.80 ± 1.25% and 0.76 ± 0.28%, respectively, with 0.60 ± 0.14% co-localization of PGR and FOXO1. Overall, these expression patterns were consistent with the IHC results and reveal altered spatial distribution and coordination of PGR and FOXO1 in Ncor1 -deficient uteri. Co-expression of PGR and FOXO1 was barely detectable. Collectively, these findings support an association between NCOR1 loss and disrupted progesterone-responsive signaling during implantation. To evaluate the effect of NCOR1 loss on decidualization, we employed a mouse artificial decidualization model. 37 The stimulated uterine horn of the Ncor1 f/f mice responded robustly to the artificial decidual stimulus; however, Ncor1 d/d mice exhibited only a partial decidual response ( Figure 7 A). The weight ratio of the stimulated to control horn was significantly decreased (3.18 fold, p < 0.0001) in Ncor1 d/d mice (3.90 ± 0.62, n = 6) compared to Ncor1 f/f mice (14.37 ± 1.15, n = 7) ( Figure 7 A). While histological analysis of the stimulated uterine horn revealed enlarged, cuboidal decidual cells in both Ncor1 f/f and Ncor1 d/d mice, the number of decidualized cells was remarkably smaller in Ncor1 d/d mice compared to Ncor1 f/f mice ( Figure 7 B). These data support the conclusion that NCOR1 plays a critical role in decidualization. Figure 7 Ncor1 d/d mice exhibit impaired decidualization (A) Representative gross uterine morphology following 5 days of artificial decidualization induction in Ncor1 f/f ( n = 6) and Ncor1 d/d ( n = 7) mice. Each n represents an individual mouse subjected to unilateral uterine horn stimulation. Images are representative of independent biological samples. (B) Representative histological analysis of control (unstimulated) and mechanically stimulated uterine horns from Ncor1 f/f ( n = 6) and Ncor1 d/d ( n = 7) mice at decidualization day 5. Data are presented as mean ± SEM. Statistical significance was determined by two-tailed unpaired Student’s t test. ∗∗∗ p < 0.001. Ncor1 d/d mice exhibit impaired decidualization (A) Representative gross uterine morphology following 5 days of artificial decidualization induction in Ncor1 f/f ( n = 6) and Ncor1 d/d ( n = 7) mice. Each n represents an individual mouse subjected to unilateral uterine horn stimulation. Images are representative of independent biological samples. (B) Representative histological analysis of control (unstimulated) and mechanically stimulated uterine horns from Ncor1 f/f ( n = 6) and Ncor1 d/d ( n = 7) mice at decidualization day 5. Data are presented as mean ± SEM. Statistical significance was determined by two-tailed unpaired Student’s t test. ∗∗∗ p < 0.001. During the preimplantation period, epithelial and stromal cell proliferation is essential for establishing uterine receptivity, which is critical for successful embryo apposition, attachment, implantation, and pregnancy maintenance. 38 P4 and E2 are master regulators of these processes. In mice, E2 induces the proliferation of epithelial and stromal cells. However, P4 inhibits epithelial proliferation just before implantation to prepare the uterus for an embryo-receptive state. 38 To further explore the effects of NCOR1 loss on receptivity of Ncor1 d/d mice, we examined cell proliferation in the uterine luminal epithelium and stroma of Ncor1 d/d mice by assessing the expression of the proliferation marker Ki67. The expression of Ki67 at GD 3.5 was reduced in the luminal epithelium of Ncor1 f/f mice (95.21 ± 1.91%), while Ncor1 d/d mice revealed two distinct patterns of luminal epithelial proliferation ( Figure 8 A). In group #1 (no proliferation in the luminal epithelium, 7/15), the number of Ki67-positive cells were not detected in the luminal epithelium, as shown in the controls. In group #2 (aberrant active proliferation in luminal epithelium, 8/15), luminal epithelial proliferation was significantly increased compared to controls (10.80 fold, p < 0.0001) and group #1 of Ncor1 d/d mice (5.13 fold, p < 0.0001) ( Figure 8 A). Furthermore, stromal cell proliferation was significantly decreased in the uteri of both group #1 and #2 of Ncor1 d/d mice (1.96 fold, p = 0.0017, and 5.50 fold, p < 0.0001, respectively) compared to control mice ( Figure 8 A). Figure 8 Ncor1 d/d mice result in a non-receptive endometrium with defective progesterone signaling (A) Representative immunohistochemistry images and H-score quantification of Ki67 expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 7), and group #2 Ncor1 d/d mice ( n = 8) at GD 3.5. Each n represents an individual pregnant mouse. Images are representative of independent biological samples. (B) Representative immunohistochemistry images and H-score quantification of PGR expression in uteri from Ncor1 f/f ( n = 5) and Ncor1 d/d ( n = 15) mice at GD3.5. Each n represents an individual mouse. (C) RT-qPCR analysis of P4 target genes, Il13ra2 , Fst , Hand2 , Areg , and Lrp2 , in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 7), and group #2 Ncor1 d/d mice ( n = 8) at GD 3.5. (D) RT-qPCR analysis of E2 target genes, Muc1 , Clca3 , and Ltf , in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 7), and group #2 Ncor1 d/d mice ( n = 8) at GD 3.5. For RT-qPCR analyses, each n represents one biologically independent uterine sample. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. Ncor1 d/d mice result in a non-receptive endometrium with defective progesterone signaling (A) Representative immunohistochemistry images and H-score quantification of Ki67 expression in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 7), and group #2 Ncor1 d/d mice ( n = 8) at GD 3.5. Each n represents an individual pregnant mouse. Images are representative of independent biological samples. (B) Representative immunohistochemistry images and H-score quantification of PGR expression in uteri from Ncor1 f/f ( n = 5) and Ncor1 d/d ( n = 15) mice at GD3.5. Each n represents an individual mouse. (C) RT-qPCR analysis of P4 target genes, Il13ra2 , Fst , Hand2 , Areg , and Lrp2 , in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 7), and group #2 Ncor1 d/d mice ( n = 8) at GD 3.5. (D) RT-qPCR analysis of E2 target genes, Muc1 , Clca3 , and Ltf , in uteri from Ncor1 f/f ( n = 5), group #1 Ncor1 d/d mice ( n = 7), and group #2 Ncor1 d/d mice ( n = 8) at GD 3.5. For RT-qPCR analyses, each n represents one biologically independent uterine sample. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗ p < 0.05, ∗∗ p < 0.01, ∗∗∗ p < 0.001. Next, we assessed the expression of PGR, a primary mediator of P4 action in the uterus. We found that the expression of PGR showed no change in the epithelial and stromal cells in the uteri of Ncor1 d/d mice ( p = 0.9658 and p = 0.1052, respectively) compared to the Ncor1 f/f mice at GD 3.5 ( Figure 8 B). However, mRNA levels of P4 target genes Il13ra2 , Fst , Hand2 , Areg , and Lrp2 were significantly downregulated in both group #1 (1.76 fold, p = 0.0230; 2.25 fold, p = 0.0008; 1.97 fold, p = 0.0046; 4.38 fold, p = 0.0044, and 1.62 fold, p = 0.0231, respectively) and group #2 (6.23 fold, p < 0.0001; 13.59 fold, p < 0.0001; 2.36 fold, p = 0.0009; 6.45 fold, p = 0.0016, and 19.51 fold, p < 0.0001, respectively) of Ncor1 d/d mice compared to controls ( Figure 8 C). Interestingly, mRNA levels of P4 target genes, Il13ra2 , Fst , and Lrp2 , were significantly decreased (3.58 fold, p = 0.0146; 6.02 fold, p = 0.0086, and 12.07 fold, p = 0.0003, respectively) in the group #2 of Ncor1 d/d mice compared to group #1 of Ncor1 d/d mice ( Figure 8 C). To determine whether Ncor1 ablation caused excess E2 signaling, we assessed mRNA expression levels of E2 target genes, Muc1 , Clca3 , C3 , and Ltf , at GD3.5 ( Figure 8 D). However, only Muc1 and Ltf mRNA levels were significantly increased in group #2 of Ncor1 d/d mice compared to control mice (2.08 fold, p = 0.0415, and 5.70 fold, p = 0.0493, respectively) ( Figure 8 D). These results suggest that NCOR1 loss is associated with a non-receptive endometrium, most likely due to dysregulation of P4 signaling. To determine whether NCOR1 and HDAC3 influence each other’s expression, we assessed NCOR1 and HDAC3 protein levels by IHC in uteri from control, Hdac3 d/d , and Ncor1 d/d mice at GD3.5. In control mice, HDAC3 and NCOR1 were robustly expressed in both the stromal compartment (167.84 ± 2.84 and 173.58 ± 2.54, respectively) and the epithelial compartment (197.39 ± 1.63 and 208.42 ± 3.05, respectively). As expected, expression of the targeted protein was nearly absent in both the stroma (2.62 ± 0.52 and 7.62 ± 0.73, respectively; p < 0.001) and epithelium (5.91 ± 0.51 and 0.60 ± 0.24, respectively; p < 0.001) of Hdac3 d/d and Ncor1 d/d mice. Notably, in Ncor1 d/d mice, HDAC3 expression was significantly reduced in both the stroma (96.38 ± 3.70; p < 0.001) and epithelium (141.87 ± 11.34; p < 0.001) compared to controls ( Figures 9 A and 9B). Notably, NCOR1 expression was nearly absent in Hdac3 d/d mice (stroma, 7.62 ± 0.73; epithelium, 16.03 ± 4.24; both p < 0.001) relative to controls. These findings suggest that NCOR1 and HDAC3 are both necessary to sustain normal expression of the other in the uterus, supporting a functional association that may be important for uterine progesterone responsiveness ( Figures 9 C and 9D). These data indicate that NCOR1 and HDAC3 exhibit reciprocal expression changes consistent with interdependence, supporting a functional interrelationship between these corepressor components during early pregnancy. Figure 9 Reciprocal expression of NCOR1 and HDAC3 in the uteri of Hdac3 d/d and Ncor1 d/d mice at GD3.5 (A and B) Representative immunohistochemistry images and (B) H-score quantification of HDAC3 expression in uteri from control, Hdac3 d/d , and Ncor1 d/d mice at GD3.5 ( n = 5 mice per genotype). (C and D) Representative immunohistochemistry images and (D) H-score quantification of NCOR1 expression in uteri from control, Hdac3 d/d ,and Ncor1 d/d mice at GD3.5 ( n = 5 mice per genotype). For all images, each n represents an individual pregnant mouse. Images are representative of independent biological samples. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗∗ p < 0.01 and ∗∗∗ p < 0.001. Reciprocal expression of NCOR1 and HDAC3 in the uteri of Hdac3 d/d and Ncor1 d/d mice at GD3.5 (A and B) Representative immunohistochemistry images and (B) H-score quantification of HDAC3 expression in uteri from control, Hdac3 d/d , and Ncor1 d/d mice at GD3.5 ( n = 5 mice per genotype). (C and D) Representative immunohistochemistry images and (D) H-score quantification of NCOR1 expression in uteri from control, Hdac3 d/d ,and Ncor1 d/d mice at GD3.5 ( n = 5 mice per genotype). For all images, each n represents an individual pregnant mouse. Images are representative of independent biological samples. Data are presented as mean ± SEM. Statistical significance was determined by one-way ANOVA followed by Tukey’s post hoc test. ∗∗ p < 0.01 and ∗∗∗ p < 0.001.

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jae-Wook Jeong ( [email protected] ). This study did not generate new reagents or mouse lines for this study. • All data supporting this study are included in the article and supplemental information . • This study does not report original code. • Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request. All data supporting this study are included in the article and supplemental information . This study does not report original code. Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

We found that the expression of NCOR1 is decreased in the eutopic endometrium of infertile women with endometriosis compared to controls. Epigenetic modifications play a significant role in endometriosis-associated infertility by altering gene expression in the endometrium without changing the underlying DNA sequence. 39 Despite growing evidence linking epigenetic regulation to endometriosis-associated infertility, several critical gaps remain. These knowledge gaps hinder the development of targeted therapies and diagnostic biomarkers. Using a mouse model of endometriosis, we demonstrated that endometriosis leads to NCOR1 loss in the eutopic endometrium. Although the precise molecular mechanisms of NCOR1 loss in the etiology and pathophysiology of endometriosis require further study, NCOR1 loss affects pathways critical for endometrial receptivity, decidualization, immune tolerance, and hormone responsiveness, which are essential for implantation and early pregnancy. The spatial regulation of NCOR1 expression within the decidua is consistent with an interface between decidual remodeling and immune regulation. The PDZ is characterized by restricted immune cell infiltration and the establishment of an immunotolerant environment to support embryo survival, whereas the SDZ contains a higher density of immune cells involved in tissue remodeling and vascular adaptation. The observed reduction of NCOR1 expression in the PDZ, together with its increased expression in the SDZ, is consistent with previously described roles for NCOR1 in modulating immune activation and tolerance, including its function in dendritic cell-mediated immune regulation. These findings suggest that spatially restricted NCOR1 expression may be associated with the balance between immune tolerance and immune activation during decidualization. Importantly, fundamental differences exist between mouse and human decidualization that must be considered when interpreting these findings. 40 In humans, decidualization of endometrial stromal cells occurs spontaneously during the secretory phase in response to progesterone, independent of embryo presence, 41 whereas in mice decidualization is initiated by embryo attachment and implantation. 7 , 42 Accordingly, the implantation-associated phenotypes observed in Ncor1 -deficient mice reflect disruptions in hormone-regulated uterine receptivity and stromal differentiation within a murine context. In this study, mouse models are therefore used to interrogate conserved progesterone-dependent regulatory mechanisms of endometrial function rather than to directly model human decidualization or implantation events. Genetically engineered animal models are essential tools for studying female infertility because they allow precise investigation of the molecular, genetic, and physiological mechanisms underlying reproductive function. We used a uterine-specific conditional Ncor1 knockout mice ( Ncor1 d/d ) model to assess the effect of NCOR1 loss in uterine functions, which resulted in early pregnancy loss due to implantation failure. At implantation period, COX-2, a marker of decidualization, is strongly and specifically expressed in stromal cells surrounding the embryos that are beginning to decidualize near the embryo invasion region. 32 Ncor1 d/d mice show a significant reduction of COX2 in implantation sites. In addition, phosphorylation of STAT3 (pSTAT3) was strongly observed in the luminal epithelium and the stroma surrounding the implanting blastocyst at implantation sites at implantation stage, GD 4.5. STAT3 is activated by many cytokines, growth factors, E2, and P4 in uterus. 32 , 43 , 44 In contrast, pSTAT3 levels were decreased markedly in the recurrent implantation failure patients. 45 STAT3 is important for mouse embryo implantation and decidualization. 33 , 46 , 47 Our results showed that the expression of pSTAT3 was significantly reduced in the luminal epithelium and the stroma surrounding the implanting blastocyst in both groups of Ncor1 d/d mice. These results demonstrated early pregnancy loss was due to implantation failure in Ncor1 d/d mice. In addition to its role in epithelial and stromal compartments, NCOR1 has been characterized as a regulator of immune cell development and inflammatory signaling in multiple tissues. 27 , 28 , 31 Maternal immune cells constitute a substantial proportion of the decidua during early pregnancy and play essential roles in immune tolerance, tissue remodeling, and implantation support. 40 , 41 , 48 Although immune cell-specific analyses were not performed in this study, cytokine-mediated STAT3 signaling pathways, which were found to be altered in Ncor1 -deficient uteri, are well established as critical mediators of decidualization and embryo implantation. 33 , 34 , 45 , 46 The reduction in stromal pSTAT3 observed in Ncor1 d/d mice is therefore interpreted as being consistent with a disrupted cytokine-responsive signaling at the decidual-stromal interface, rather than direct evidence for altered immune cell composition or function. Importantly, these interpretations are inferential, as immune cell-specific analyses were not performed in the present study. Furthermore, the uteri of Ncor1 d/d mice revealed abnormal aberrant expression of PGR and dysregulation of FOXO1 in the luminal epithelium at implantation stage. FOXO1 is a critical marker of endometrial receptivity that is controlled by epithelial PGR; nuclear translocation of FOXO1 in the luminal epithelium is required for receptivity and embryo attachment. 35 The withdrawal of PGR in epithelial cells and the rapid accumulation of FOXO1 in the endometrial epithelium at GD 4.5 are necessary to start the epithelial degradation for embryo implantation success. 35 , 49 Therefore, NCOR1 is required for accurate and dynamic regulation of PGR and FOXO1 during implantation stage. Notably, implantation failure in Ncor1 d/d mice manifested as a spectrum of phenotypes, ranging from reduced or abnormal implantation sites to a complete absence of detectable implantation. While these outcomes were categorized as “small implantation site” and “no implantation site” groups for descriptive purposes, the limited number of animals within each subgroup precluded statistically powered molecular stratification analyses. Accordingly, these phenotypes should be interpreted as representing varying degrees of defective implantation rather than as distinct mechanistically defined entities. Despite this limitation, several findings from the present study suggest potential mechanisms that may contribute to the observed phenotypic variability. Disrupted progesterone-responsive gene expression, aberrant epithelial PGR and FOXO1 regulation, and altered decidual signaling pathways, including reduced stromal pSTAT3 activation, were consistently observed in Ncor1-deficient uteri. Given the established roles of progesterone signaling and cytokine-mediated pathways such as IL-6/STAT3 in coordinating decidualization and immune adaptation during implantation, variability in the extent or timing of these disruptions may underline the range of implantation phenotypes observed in Ncor1 d/d mice. Although the relationship between NCOR1 loss and endometriosis-associated infertility was not investigated in our studies, the defective decidualization in mice with endometriosis is similar to that observed in Ncor1 d/d mice. 13 Uterine decidualization process is a key event in implantation, controlled by stromal cell proliferation and differentiation. The decidualization processes are orchestrated by E2 and P4 through their cognate receptors, PGR and ESR1. 7 , 50 E2 increases proliferation and differentiation of the epithelium during the first 2 days of pregnancy, while P4 shifts proliferation from the epithelium to the stroma on GD 3.5. Increasing stromal cell proliferation is an initiator of decidualization. After embryo attachment to the luminal epithelium, the stromal cells surrounding embryos proliferate and differentiate into decidua cells to provide nutrients and support for the developing fetus before the placenta starts to fully function. 40 Eutopic endometrial stromal cells of endometriosis patients also exhibited a reduced capacity for decidualization. 51 Ablation of Ncor1 in uterine cells using Pgr-cre model results in insufficient decidualization. Therefore, NCOR1 must be tightly regulated for decidualization during early pregnancy. Attenuation of NCOR1 leads to implantation failure and ultimately results in early pregnancy loss. P4 and E2 in the uterus are crucial for conferring uterine receptivity and implantation. 8 The lack of sufficient P4 and E2 action can result in infertility and pregnancy loss in both humans and mice. 3 , 50 , 52 , 53 An imbalance P4 and E2 action is also found in the endometrium of women with endometriosis. 52 We found that P4-responsive genes were downregulated in the uteri of Ncor1 d/d mice at GD 3.5, and the stromal cell proliferation was significantly reduced. Although there was no change in the expression of PGR in the uterus of Ncor1 d/d mice at GD 3.5, the expression of PGR was reduced in the uterus of Ncor1 d/d mice at GD 4.5. Despite these alterations, there was no difference in the serum E2 and P4 levels between control and Ncor1 d/d mice, indicating normal endocrine regulation of hormone secretion. These findings suggest that loss of Ncor1 leads to unopposed epithelial E2 action due to defective P4 function. The action of NCOR1 is dependent on a large corepressor complex containing HDAC3 and NCOR1/SMRT. 54 Our previous work indicated that HDAC3 was downregulated in the endometrium of infertile women with endometriosis and that loss of HDAC3 is associated with aberrant hormonal signaling. 25 NCOR1 binds to nuclear receptors and subsequently recruits HDACs to regulate the expression of nuclear receptor target genes. 28 The interaction between NCOR1 and HDACs forms a corepressor complex to downregulate target genes. 55 Several studies have revealed that NCOR1 is required for nearly all HDAC3 enzyme activity in vivo . 56 The activation of HDAC3 by NCOR1 plays an important role in epigenetic regulation and metabolic physiology of multiple tissues. 57 , 58 Therefore, NCOR1 mediates its biological functions through the recruitment and activation of HDAC3. However, the precise molecular mechanisms causing HDAC3/NCOR1 corepressor complex dysregulation in disease remain unknown. In summary, we identified the attenuation of NCOR1 in human eutopic endometrium from infertile women with endometriosis. Using a mouse model for endometriosis, we showed down-regulation of NCOR1 expression after endometriosis induction. Loss of Ncor1 in the mouse uterus was associated with implantation and decidualization defects due to dysregulation of steroid hormone signaling. By demonstrating that loss of the epigenetic regulator NCOR1 causes endometriosis-associated infertility due to an implantation failure, this project will significantly advance understanding of idiopathic female infertility and early pregnancy loss. This study has several limitations that should be considered when interpreting the findings. First, although we analyzed NCOR1 expression in human eutopic endometrium and demonstrated its reduction in women with endometriosis, the sample size was relatively limited and did not fully capture the heterogeneity of disease stage, menstrual cycle phase, or prior treatment history. Endometriosis is a clinically and molecularly heterogeneous disorder, with variability in lesion stage, hormonal milieu, and inflammatory burden that may influence endometrial gene expression. 59 , 60 , 61 In addition, while patient age and fertility status were defined and other uterine pathologies were excluded, potential confounding factors such as body mass index, comorbid polycystic ovary syndrome, and tubal factor infertility were not systematically assessed and therefore could not be controlled for in the analyses. These factors are known to be independently associated with altered endometrial receptivity and progesterone responsiveness. 62 , 63 , 64 Larger and more diverse patient cohorts will be required to validate the clinical significance of NCOR1 attenuation in endometriosis-associated infertility. In addition, serum estradiol and progesterone levels were not available for the human cohort examined in this study. Although ovarian steroid hormone production is generally preserved in women with endometriosis as reported in prior clinical studies, 65 , 66 the absence of patient-specific hormone measurements limits direct comparison with the normal ovarian function observed in Ncor1 -deficient mice. Furthermore, although our mouse studies revealed aberrant epithelial PGR and FOXO1 expression at implantation sites, corresponding immunohistochemical validation in human endometrial samples collected during the receptive window was not performed. Nevertheless, altered epithelial PGR and FOXO1 expression has been well documented in women with non-receptive endometrium, implantation failure, and progesterone resistance, 3 , 67 , 68 supporting the relevance of our murine findings to human endometrial dysfunction. Future studies incorporating parallel human histopathological analyses will be important to more directly define conserved and species-specific aspects of NCOR1-dependent progesterone signaling. 40 , 41 , 69 Finally, our analysis primarily relied on immunohistochemistry, histology, and targeted gene expression assays. While these approaches provided evidence of dysregulated progesterone signaling and defective decidualization, and immune-related signaling pathways, unbiased high-throughput approaches such as ChIP sequencing, single-cell transcriptomics, chromatin accessibility profiling, or proteomics would offer a more comprehensive view of NCOR1-dependent regulatory networks in the endometrium. 70 , 71 Comprehensive characterization of decidual immune cell populations and activation states was beyond the scope of the current study and represents an important direction for future investigation. 48 , 72

The US Centers for Disease Control and Prevention estimates that approximately 6% of married women aged 15–44 years are infertile. 1 Furthermore, 15%–20% of all pregnancies end in miscarriage before 20 weeks of gestation, and approximately 75% of failed pregnancies are associated with implantation failure. 2 , 3 Successful implantation requires dynamic molecular and morphological changes in both epithelial and stromal cells of the endometrium. 4 , 5 These processes are primarily regulated by progesterone (P4) and estrogen (E2), which activate transcription of target genes through their cognate receptors (PGR and ESR1). 6 , 7 E2 and P4 are critical for regulating the human endometrium, including proliferation, secretion, and menstrual shedding. Dysregulation of E2 and P4 signaling can result in infertility and early pregnancy loss in both humans 3 and mice. 8 Due to the complexity and dynamic nature of implantation, the molecular processes underlying these changes are not fully understood. Improving fertility rates will require a deeper understanding of the molecular mechanisms that govern implantation. Endometriosis is one of the most common gynecological diseases, affecting around 15% of women of reproductive age and is associated with pain, infertility, or both. 9 The total economic burden of endometriosis in the United States is estimated to be $78–$119 billion annually. 10 , 11 Up to 50% of women with infertility have endometriosis, and 50% of women with endometriosis are infertile. 12 Recent animal studies highlight that endometriosis causes implantation failure and defective decidualization, as has been hypothesized in humans. 13 The precise mechanisms by which endometriosis leads to early pregnancy loss remain unclear. Epigenetics is defined as stable and heritable modification of gene expression without changing the DNA sequence. 14 Epigenetic regulation has gained significant attention for its role in implantation and decidualization during early pregnancy. 15 Epigenetic regulators can modulate uterine P4 and E2 signaling by regulating the expression of PGR, ESR1, and their downstream targets. 4 , 16 Alterations in epigenetic regulators can cause endometrial dysfunction, often resulting in infertility. 17 , 18 Aberrant DNA methylation in the promoter regions of genes involved in regulating balanced hormone responses can disrupt the effects of steroid hormones by altering the expression levels of their receptors, thereby affecting endometrial function and contributing to the development of endometriosis. 19 Epigenetic regulation is critical to normal endometrial steroid hormone responses and essential for proper endometrial functions. 20 , 21 Histone deacetylases (HDACs), including HDAC3, epigenetically regulate gene expression by modulating acetylation of histone and nonhistone substrates. 22 , 23 HDAC3 is an epigenetic regulator of diverse transcription factors such as nuclear receptors. 24 Our previous work showed that uterine-specific Hdac3 knockout mice ( Pgr cre/+ Hdac3 f/f ; Hdac3 d/d ) are sterile due to defective implantation and decidualization. 25 HDAC3 knockdown abrogates decidualization in human primary endometrial stromal cells. 25 HDAC3 and nuclear receptor corepressor 1 (NCOR1) are functionally and physically interconnected components of a key transcriptional repression complex that regulates chromatin remodeling and gene expression. NCOR1 is involved in the epigenetic regulation of transcription factors. 26 NCOR1 interacts with both PGR and ESR1 27 , 28 and plays a role in regulating PGR function in the endometrium during the estrous cycle. 29 Therefore, NCOR1 is considered a key regulator of both PGR and ESR1 transcriptional activity. 29 , 30 These studies suggest that NCOR1 may also play a critical role in implantation and decidualization. However, due to the embryonic lethality caused by NCOR1 deficiency, 31 the role of NCOR1 in early pregnancy and normal E2 or P4 signaling remains unknown. In this study, we sought to define the role of NCOR1 in endometrial function and its potential contribution to endometriosis-associated infertility. Although NCOR1 is a key epigenetic coregulator of steroid hormone receptors and has been implicated in uterine hormone responsiveness, its role in endometriosis and early pregnancy remains poorly understood. To address this gap, we combined analysis of human endometrial tissues with genetic approaches in mice to investigate how NCOR1 regulates uterine receptivity, implantation, and decidualization. By elucidating NCOR1-dependent mechanisms in the endometrium, this study aims to provide insight into the epigenetic regulation of implantation and identify pathways that may be disrupted in endometriosis-associated infertility.

The authors declare they have no actual or potential competing financial interests.

REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies NCOR1 Cell Signaling Cat# 34271; RRID: AB_2799050 COX2 Cayman Cat# aa 570-598 pSTAT3 Abcam Cat# ab76315; RRID: AB_1658549 FOXO1 Cell Signaling Cat# 2880; RRID: AB_2106495 PGR Cell Signaling Cat# 8757; RRID: AB_2797144 CDH1 Takara Biosciences Cat# M108 FOXA2 Abcam Cat# ab108422; RRID: AB_11157157 Goat anti-rabbit IgG Biotinylated Vector Laboratories Cat# BA-1000; RRID: AB_2313606 Goat Anti-Mouse IgG Biotinylated Vector Laboratories Cat# E42;BA-9200; RRID: AB_2336171 goat anti-rat 647 Invitrogen Cat# A21247 donkey anti-rabbit 555 Invitrogen Cat# A31572 Goat anti-rat 647 Invitrogen Cat# A21247 Hoechst Sigma Aldrich Cat# B2261 Chemical, peptide, and recombinant proteins 16% paraformaldehyde Fisher Scientific Cat# 50-980-487 10% buffered formalin Fisher Scientific Cat# 3191-1 Hematoxylin Fisher Scientific Cat # NC9220898 Eosin Fisher Scientific Cat #245-827 MMLV Reverse Transcriptase Invitrogen Cat# 28025021 SYBR green Applied Biosystems Cat# 4309155 DMSO Fisher Scientific Cat# D1281 Methanol Fisher Scientific Cat# 42395-0040 Triton X-100 Fisher Scientific Cat# A16046 Critical commercial assays RNeasy Total RNA Isolation Kit Qiagen Cat# 74104 Oligonucleotides Oligonucleotides sequnces are listed in Table S1 N/A N/A Experimental models: Organism/stains PR cre/+ (Progesterone receptor cre knockin mice, C57Bl6J X129Sv) Dr. DeMayo at NIEHS or Dr. Lydon at Baylor College of Medicine N/A Ncor1 f/f (Ncor1 floxed mice, C57Bl6J X129Sv) Dr. Auwerx at Fédérale de Lausanne N/A Software and algorithms GraphPad 10.2.392 GraphPad Prism Software, Inc https://www.graphpad.com/ Vevo Lab software 5.5.1 Fujifilm https://www.visualsonics.com/ Leica software LASX version 3.5.5 Leica https://www.leica-microsystems.com/ All mice were housed and bred in a designated animal care facility at the University of Missouri under controlled temperature and humidity conditions with a 12-hour light/dark cycle. All animal procedures were conducted in accordance with institutional guidelines and approved by the University of Missouri Animal Care and Use Committee (ACUC) under protocol number 65323. Ncor1 f/f mice were used as controls, and Pgr cre/+ Ncor1 f/f ( Ncor1 d/d ) mice were generated by crossing Pgr cre/+ with Ncor1 f/f mice. 73 , 74 Eight -week-old female mice were used for all other experiments. Pregnant uterine samples were obtained by mating Ncor1 f/f and Ncor1 d/d female mice with C57BL/6 male mice the morning of a vaginal plug designated as gestation day (GD) 0.5. Mice were killed at GD 3.5, 4.5, 5.5 and 9.5. For the fertility study, adult female Ncor1 f/f and Ncor1 d/d female mice were placed with wild-type C57BL/6 male mice continuously for 6 months. The number of litters and pups born by each female mouse during that period was recorded. Uterine tissues were collected at the time of dissection and then immediately either snap-frozen and stored at −80 °C for RNA extraction or fixed with 4% (vol/vol) paraformaldehyde for histology or immunohistochemistry. The levels of progesterone and estrogen in serum from Ncor1 f/f and Ncor1 d/d female mice at GD 3.5 were measured by the University of Virginia Center for Research in Reproduction Ligand Assay and Analysis Core. For experiments examining NCOR1 protein levels, endometrial samples were analyzed from 13 fertile women without endometriosis and from 18 women with endometriosis (reproductive age, 18-40). Fertility in the control group was defined as a documented history of at least one prior live birth without the use of assisted reproductive technologies. Control endometrial tissues were laparoscopically confirmed to be negative for endometriosis and were free of other uterine pathologies, including uterine fibroids and adenomyosis, and had not been exposed to any hormonal therapies for at least three months prior to surgery. Similarly, patients with endometriosis was evaluated to exclude coexisting uterine pathologies other than endometriosis. Histologic dating of endometrial samples was performed according to the criteria established by Noyes et al. 75 and confirmed by subsequent histopathological examination by an experienced fertility specialist (B.A.L.), as described in our previous study. 13 All patients with endometriosis-associated infertility underwent endometrial sampling prior to the surgical removal of endometriosis. Body mass index (BMI) data were not collected as part of this study. For immunohistochemistry, samples were fixed in 10% buffered formalin prior to paraffin embedding. This study was approved by the Institutional Review Boards of the University of Missouri, Duke University, and Atrium Health Wake Forest Baptist (IRB00057549). Human endometrial samples were obtained from Atrium Health Wake Forest Baptist and Duke University. Written informed consent was obtained from all participants. All procedures were conducted in accordance with institutional guidelines and the principles of the Declaration of Helsinki. Total RNA was extracted from frozen uterine tissues with the RNeasy Total RNA Isolation Kit (Qiagen, Valencia, CA). RNA purity and concentration were determined using NanoDrop. 1 μg of RNA was used for complementary DNA (cDNA) synthesis with MMLV Reverse Transcriptase (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. RT-qPCR was performed on cDNA to assess the expression of genes of interest with SYBR green analysis using an Applied Biosystems StepOnePlus (Applied Biosystems, Foster City, CA, USA). Experimental gene expression data were normalized against the housekeeping gene 60S ribosomal protein L7 ( Rpl7 ), which was selected based on our prior validation and published evidence demonstrating stable expression in endometrial tissues across hormonal states and inflammatory conditions. 37 In the present study, Rpl7 expression showed minimal variability across experimental groups, supporting its use as an internal normalization control. Primer sequences used in these studies are shown in Table S1 . For histological analysis, dewaxed, hydrated paraffin-embedded tissue sections were stained with Hematoxylin (Hematoxylin Stain, Mayer’s solution, VOLU-SOL) and Eosin (Eosin-Y, Fisherbrand). Immunohistochemistry analyses were performed as previously described 76 , 77 with specific commercially available primary antibodies for (anti-NCOR1 (1:500 dilution; CS- 34271; Cell Signaling), anti-COX2 (1:500 dilution; aa 570-598; Cayman), anti-pSTAT3 (1:5000 dilution; ab76315; Abcam), anti-FOXO1 (1:500 dilution; mAb 2880; Cell Signaling), anti-PGR (1:2000 dilution; CS-8757S; Cell Signaling). Goat anti-rabbit IgG Biotinylated (BA-1000) or Goat Anti-Mouse IgG Biotinylated (BA-9200, Vector Laboratories) were used after administration of primary antibodies. Immunoreactivity was detected using the Vectastain Elite DAB kit (Vector Laboratories) and analyzed using Akoya PhenoImager HT scanner from AKOYA Biosciences company (Marlborough, MA, USA). A semiquantitative grading system (H-score) was used to compare the immunohistochemical staining intensities as previously described. The overall score ranged from 0 to 300. To monitor embryo development during the early pregnancy, a non-invasive high-resolution imaging tool was used at GD 5.5, 7.5 and 9.5 using the Vevo F2 LAZR-X small animal ultrasound system (FUJIFILM VisualSonics Inc., Toronto, Canada). Mice were anesthetized using 3% isoflurane and positioned supine on a heated imaging stage with the face toward the scientist according to the manufacturer’s instructions. ECG, body temperature, and respiratory physiology were always controlled. Eye cream was used to protect and maintain the physiological ocular welfare. The hair on the abdomen was removed using commercial hair removal cream. Rinse off hair removal cream to avoid irritation or darkening the skin before scanning. Then, pre-warmed ultrasound gel was applied directly onto the depilated skin. Next, the bladder was used to be identified as reference point using the transducer 47 MHz (MS550D-0421). The transducer was attached to the motor to translate the transducer from the left to the right side of the abdomen with step sizes of scanning (10 μm). B-mode or brightness mode was used to render a two-dimensional grayscale image of anatomical structures. The acquired 2D images were processed to reconstruct and analyze the 3D volume of implantation with Vevo Lab software (Fujifilm, VisualSonics, version 5.5.1). The whole-mount immunofluorescence was performed as described previously. 78 Briefly, the dissected uteri were fixed in DMSO: Methanol (1:4) and stored at -20°C. For staining, the uteri were rehydrated in a 1:1 solution of Methanol: PBT (1% Triton X-100 in PBS) for 15 minutes, followed by a 15 minutes PBT only wash. The uteri were then incubated in a blocking solution (2% powdered milk in PBT) for 1 hour at room temperature, followed by incubation with primary antibodies diluted in blocking solution, for seven nights at 4°C. Uteri were then washed with PBT twice for 15 minutes each and four times for 45 minutes each and then incubated with secondary antibodies (1:500) at 4°C for three nights. Uteri were then washed with PBT once for 15 minutes and three times for 45 minutes each and incubated at 4°C overnight with 3% H 2 O 2 solution prepared in methanol. The uteri were then washed with 100% methanol twice for 15 minutes and once for 60 minutes. Uteri were cleared overnight using a 1:2 solution of benzyl alcohol: benzyl benzoate (Sigma-Aldrich, 108006, B6630). Primary antibodies used include rat anti-CDH1 (M108, Takara Biosciences; 1:500), rabbit anti-FOXA2 (Abcam, ab108422; 1:500), and Armenian-hamster anti-CD31 (DSHB, AB_2161039; 1:200). Alexa Fluor-conjugated secondary antibodies include goat anti-rat 647 (Invitrogen, A21247), donkey anti-rabbit 555 (Invitrogen, A31572), and Goat anti-rat 647 (Invitrogen, A21247). Hoechst (Sigma Aldrich, B2261) was used to stain the nucleus. The stained and cleared samples were imaged using a Leica TCS SP8 X Confocal Laser Scanning Microscope System with a white-light laser. A 10x air objective was used to image the whole uterine tissues with a 7.0 μm Z stack to image the samples (Madhavan et al., 2022). Images were merged using Leica software LASX version 3.5.5 and saved as .LIF files. Image analysis was performed using Imaris v9.2.1 (Bitplane). The confocal image (.LIF) files were imported into Imaris and implantation chambers and embryos were visualized with z-stacks ranging from between 7-271 μm. Endometriosis was induced following previously reported approaches. 13 Two-month-old female Pgr cre/+ Rosa26 mTmG or Pgr cre/+ Ncor1 /f Rosa26 mTmG mice received three daily doses of E2. After hormone treatment, the animals underwent autologous uterine tissue transplantation under anesthesia, during which the abdomen was opened, and one uterine horn was excised. The isolated uterine tissue was placed in phosphate-buffered saline (PBS; pH 7.5), opened longitudinally, and sectioned into small fragments (approximately 1 mm 3 in size). These tissue fragments were implanted onto the peritoneal wall of the same animal, after which the abdominal wall and skin were closed. One month after endometriosis induction, fluorescently labeled ectopic lesions were examined, collected, and quantified under dissecting fluorescence microscopy. Successful establishment of endometriosis was verified by the presence of GFP-positive ectopic lesions located outside the uterus. Ectopic lesions were further confirmed based on their gross morphology and anatomical location outside the uterus, consistent with ectopic endometrial implants. Ectopic lesions were sectioned and confirmed by H&E histology showing endometrial glands and stroma. Quantitative analyses of ectopic lesion number and weight were performed. For artificial decidualization, female Ncor1 f/f and Ncor1 d/d female mice were ovariectomized at 6 weeks of age. 37 After resting at least 2 weeks to remove endogenous ovarian hormones completely, mice were administered daily the following hormonal regimen: 100 ng of E2 per day for three days; two days of rest; then three daily injections of 1 mg of P4 plus 6.7 ng of E2. To induce artificial decidualization, a single horn of each mouse was mechanically stimulated by scratching the full length of the anti-mesometrial side with a blunted needle six hours following the third P4 and E2 injection, while the other horn was left unstimulated as a control. Daily injections of P4 (1 mg/mouse) + E2 (6.7 ng/mouse) were continued for five days to maximize the decidual response. Five days after the mechanical trauma, the uterine tissues were excised to measure the weights of the stimulated and control uterine horns of each mouse. For histological analysis, uterine tissues were fixed in 4% paraformaldehyde. For RNA extractions, uterine tissues were collected and immediately frozen in dry ice. Statistical significance analysis was performed using the Student’s t test for data with only two groups. One-way ANOVA was used, followed by Tukey’s post hoc test for multiple comparisons for data containing more than two groups. All data are presented as means ± SEM. A p < 0.05 was considered statistically significant. All statistical analyses were performed using the Prism version 10.2.392 from GraphPad (San Diego, CA, USA).

This work was supported by the 10.13039/100009633 Eunice Kennedy Shriver National Institute of Child Health & Human Development of the 10.13039/100000002 National Institutes of Health under award numbers P01HD106485 , R01HD102170 , and R01HD101243 (to J.-W.J.), R01HD112332 (to. T.H.K.), R01HD109152 (to R.A.), T32HD087166 (to H.R.K), and the 10.13039/501100001711 Swiss National Science Foundation under award numbers SNSF 31003A_179435 (to J.A.). We thank The 10.13039/100008457 University of Virginia Center for Research in Reproduction Ligand Assay and Analysis Core for their services.