ST6GALNAC1-mediated sialylation in uterine endometrial epithelium facilitates the epithelium-embryo attachment

Abstract

Introduction: Embryo implantation requires synergistic interaction between the embryo and the receptive endometrium. Glycoproteins and glycan-binding proteins are involved in endometrium-embryo attachment. Sialyl Tn (sTn), a truncated O-glycan, is catalyzed by ST6 N-Acetylgalactosaminide Alpha-2,6-Sialyltransferase 1 (ST6GALNAC1) and can be detected by specific Sialic-acid-binding immunoglobulin-like lectins (Siglecs). Whether the sTn-Siglecs axis supports embryo implantation remains unknown.

Objectives: This paper aims to study the role of ST6GALNAC1/sTn-Siglecs axis in embryo implantation.

Methods: ST6GALNAC1 and sTn in human endometrium were analyzed by immunohistochemistry. An in vitro implantation model was conducted to evaluate the effects of ST6GALNAC1/sTn on the receptivity of human endometrial AN3CA cells to JAR spheroids. Immunoprecipitation combined with mass spectrometry analysis was carried out to identify the key proteins modified by sTn in endometrial cells. Siglec-6 in human embryos was analyzed by published single-cell RNA sequencing (scRNA-seq) datasets. Protein interaction assay was applied to verify the bond between the Siglec-6 with sTn-modified CD44. St6galnac1 siRNAs and anti-sTn antibodies were injected into the uterine horn of the mouse at the pre-implantation stage to evaluate the role of endometrial St6galnac1/sTn in embryo implantation. Siglec-G in murine embryos was analyzed by immunofluorescence staining. The function of Siglec-G is evidenced by uterine horn injection and protein interaction assay.

Results: Both human and murine endometrium at the receptive stage exhibit higher ST6GALNAC1 and sTn levels compared to the non-receptive stage. Overexpression of ST6GALNAC1 significantly enhanced the receptivity of AN3CA cells to JAR spheroids. Inhibition of endometrial ST6GALNAC1/sTn substantially impaired embryo implantation in vivo. CD44 was identified as a carrier for sTn in the endometrial cells of both species. Siglec-6 and Siglec-G, expressed in the embryonic trophectoderm, were found to promote embryo attachment, which may be achieved through binding with sTn-modified CD44.

Conclusion: ST6GALNAC1-regulated sTn in the endometrium aids in embryo attachment through interaction with trophoblastic Siglecs.

Keywords: CD44; Embryo attachment; Endometrial receptivity; ST6GALNAC1; sialyl Tn.

Graphical abstract

Fig. 1

Expression of FUTs and STs family glycogenes in human uterine endometrium across the menstrual cycle. (A) Heatmap showed the expression levels of the FUTs and STs family genes in human uterine endometrium at the proliferative phase (PE, n = 4), early-secretory phase (ESE, n = 3), mid-secretory phase (MSE, n = 8) and late-secretory phase (LSE, n = 5) in published microarray dataset (GSE4888). (B) Expression levels of eight interested genes (FUT2, FUT4, FUT8, ST3GAL1, ST3GAL6, ST6GAL1, ST6GAL2, ST6GALNAC1) were statistically analyzed and exhibited in the box plots. n.s.: not significant; *P<0.05; **P<0.01; **P<0.001.

Fig. 2

Analysis of selected glycogenes in the human uterine endometrium at single-cell resolution across the menstrual cycle. (A) Seven cell clusters from 10x Genomics scRNA-seq dataset (GSE111976) visualized by UMAP projection. (B) Expression patterns of canonical markers (CDH1, EPCAM, COL1A1, COL3A1, PECAM1 and CD34) and (C) eight glycogenes (FUT2, FUT4, FUT8, ST3GAL1, ST3GAL6, ST6GAL1, ST6GAL2 and ST6GALNAC1) of each cell were shown by UMAP plot. (D) Expression patterns of selected eight glycogenes of each cell type were shown by dot plot. Each point in UMAP plots represents a cell (A-C). The color saturation indicates the average expression level (B and C). Each dot in the dot plot represents the expression of the gene in the cell (D), The color saturation indicates the average expression level, and the dot size indicates the percentage of cells expressing the gene (D).

Fig. 3

Identification of ST6GALNAC1 expression in human endometrial tissues and cells (A) UMAP plot showed the expression pattern of ST6GALNAC1 of each cell split by different stages in endometrial tissues (GSE111976). Representative IHC images of human endometrium tissues stained with anti-ST6GALNAC1 (B) and anti-sTn (C). (D) qPCR analysis of ST6GALNAC1 transcript levels in AN3CA cells treated with 17β-estradiol (E₂) or E₂ plus progesterone (P₄). (E) qPCR analysis of ST6GALNAC1 transcript levels in AN3CA cells treated with different conditions as indicated. (F) Analysis of the ChIP-Seq dataset (GSE200802) showed binding of PR to the ST6GALNAC1 promoter in human endometrial epithelial cells cultured in organoid conditions and treated with E₂ plus medroxyprogesterone acetate (MPA). (G) AN3CA cells were treated as indicated and lysed for a ChIP assay with IgG or anti-PR antibodies. PCR was performed using primers targeting the ST6GALNAC1 promotor. Bars represent 50 μm. qPCR and ChIP were performed for three independent experiments, and representative gel images are shown. Data are presented as mean ± SD, n.s.: not significant; *** P<0.001. LE, luminal epithelium; G, gland; St, Stroma.

Fig. 4

ST6GALNAC1/sTn promotes the receptivity of endometrial cells in vitro. Lentivirus-mediated overexpression of ST6GALNAC1 in AN3CA cells was confirmed by qPCR (A) and immunoblot (B). sTn levels in LV-NC and LV-ST6 OE cells were detected by dot blot (C). The receptive ability of AN3CA cells was analyzed by an in vitro implantation model. Adhered JAR-spheroids (red) on AN3CA cell monolayer (green) were photographically recorded. Adhesion assay was performed for five independent experiments. The adhesion rate of attaching JAR-spheroids was statistically analyzed and shown in the histogram (D). Bar represents 100 μm. qPCR, immunoblot, and dot blot were performed for three independent experiments. Representative immunoblot and dot blot images were shown. Data are presented as mean ± SD, n.s.: not significant; **P<0.01; ***P<0.001.

Fig. 5

CD44 is a carrier for sTn in human endometrial cells. (A) The experimental strategy invovled immunoprecipitation followed by LC-MS/MS analysis, and eighteen identified proteins (secretory and membrane proteins) were listed. (B) Protein sequences of integrin β1 and CD44 are shown, with identified peptides were underlined. (C) qPCR analysis of ITGB1 and CD44 transcript levels in LV-NC and LV-ST6 OE AN3CA cells. (D) Immunoblots of integrin β1, CD44, and sTn in immunoprecipitated samples derived from LV-NC and LV-ST6 OE AN3CA cells. (E) Expression levels of ITGB1 and CD44 in the published microarray dataset (GSE4888) were statistically analyzed and exhibited in the box plots. qPCR (C) and IP (D) were performed for three independent experiments, and representative immunoblot images are shown. Data are presented as mean ± SD, n.s.: not significant; *P<0.05; **P<0.01; *** P<0.001.

Fig. 6

Trophoblastic Siglec-6 binds with endometrial sTn-modified CD44. (A-i) Schematic of the dynamic changes in the human endometrium throughout the menstrual cycle. (A-ii) Schematic of human pre-implantation embryo. (B) Expression patterns of canonical markers (SOX2 and GATA2), pTE markers (CCR7, CYP19A1, and DLX5), and SIGLEC6 of each cell in 1529 individual cells from 88 human preimplantation embryos (E-MTAB-3929) were shown by UMAP plot. (C-i) Immunoblot for Siglec-6 and 6*His-tag in purified samples from HTR-8/SVneo cells over-expressing Siglec-6. (C-ii) Immunoblot for 6*His-tag and CD44 in samples from purified 6*His-Siglec-6 (lane-1), purified 6*His-Siglec-6 co-incubated with protein lysates from LV-NC AN3CA cells (lane-2) and purified 6*His-Siglec-6 co-incubated with protein lysates from LV-ST6 OE AN3CA cells (lane-3). Protein interaction assays were performed independently three times, and representative immunoblot images are shown.

Fig. 7

Stn/siglec-g underlies the embryo implantationin vivo. Experimental design. (B) Representative immunofluorescent images for sTn in uterine endometrium at GD4. (C) Implanted embryos (arrows) on GD6 were visualized by blue dye. The number of implanted embryos on the two uterine sides treated as indicated (Uterine horn injection of IgG or anti-sTn antibodies on GD3) was statistically analyzed. (D) Reprehensive immunofluorescent images for Siglec-G in blastocyst flushed from the uterine lumen on GD4. (E) Implanted embryos (arrows) on GD6 were visualized by blue dye. The number of impanated embryos on the two uterine sides treated as indicated (Uterine horn injection of IgG or anti-Siglec-G antibodies on GD3) was statically analyzed. (F) Immunoblots of CD44 in input and immunoprecipitated samples derived from pooled whole cell lysates (WCL) of uteri tissues at GD4 (n = 3) with anti-sTn antibodies or IgG. (G-i) Schematic for protein interaction assay. (G-ii) Immunoblot of CD44 in samples from rmSiglec-G (lane 1), rmSiglec-G co-incubated with pooled WCL from uteri tissues at GD4 (n = 3) (lane 2), rmSiglec-G co-incubated with pooled WCL from uteri tissues at GD4 (n = 3) in the presence of IgG (lane 3) or anti-sTn antibodies (lane 4). Ponceau S staining of the NC membrane was shown to prove the equal protein loading. Bars represent 10 μm (B) and 20 μm (D). LE, luminal epithelium; TE, trophectoderm. The uterine horn injection assay was performed independently four times. IP and Protein interaction assays were performed for three independent experiments, and representative immunoblot images were shown. **P<0.01; *** P<0.001.

Fig. 8

Schematic diagram of working mechanism in the current study. (A) Schematic of the dynamic changes in human endometrium throughout the menstrual cycle. (B) After ovulation, progesterone acts on PR to up-regulate the ST6GALNAC1 (St6galnac1 in mice). ST6GALNAC1 further promotes the modification of sTn on glycoproteins in the Golgi. (C) During the “WOI”, human Siglec-6 or mouse Siglec-G recognizes the sTn-modified CD44 on the luminal epithelial cell surface to facilitate embryo-epithelium attachment.