S100A8 regulated by estrogen improves injured endometrial epithelium reconstruction by promoting tight junction formation and stromal cell transformation

Abstract

Estrogen is used for endometrial repair; however, it has limited effectiveness. Cytokine S100A8 expression is affected by estrogen and is involved in regulating the damage repair. We aimed to confirm the effects of S100A8 and explore its mechanisms during the reconstruction of the injured endometrium. We investigated the effects of estrogen on S100A8 expression in healthy endometrium. A rat model of endometrial injury and cultured primary endometrial cells were used to determine the pleiotropic effects of S100A8 on endometrial epithelial repair. Estrogen regulated the recruitment of S100A8-positive immune cells (S100A8-PICs) and the release of S100A8 in the healthy endometrium, but estrogen plays a limited role in regulating seriously injured endometrium via self-S100A8. S100A8 administration to the uterine cavity significantly improved the morphology of injured epithelium; influenced the reverse migration of S100A8-PICs; and promoted epithelial localization of proliferating cells, cell junction formation, and transformation of stromal cells to epithelial cells. S100A8 in the uterine cavity has pleiotropic effects that improve endometrial epithelial reconstruction and can compensate for the deficiency in estrogen repair regulation.

Keywords: Cell junction; Endometrial injury; Epithelium reconstruction; Estrogen; S100A8; Transformation.

Fig. 1

Effects of estrogen treatment on S100A8 expression in endometrium and epithelium morphology. (a) S100A8 immunohistochemical staining before and after estrogen treatment (1d/7d); black arrows indicate S100A8-positive immune cells (S100A8-PICs).The blue arrow indicates the release of S100A8 from S100A8-PICs. Scale bar is 100 or 400 μm. (b) Rat endometrial hematoxylin–eosin staining before and after estrogen treatment (1d/7d); red arrows show disrupted cell junctions. Scale bar is 100 μm. (c) S100A8-PICs number in the endometrium (1/mm²) before and 1 d/7d after estrogen treatment (n = 18, from six rats, the number was counted from three immunohistochemical images of each sample , mean ± SEM). (d) S100A8 expression in the epithelium before and 1 d/7d after estrogen treatment (n = 18, from six rats, the expression level was measured from three immunohistochemical images of each sample, mean ± SEM). * Results with P values < 0.05 were considered significant. ** Results with P values < 0.01 were considered very significant.

Fig. 2

Effects of uterine cavity administration of S100A8 on the distribution of S100A8-PICs and endometrial morphology. (a) Distribution of S100A8-positive cells in the endometrium of four different groups with estrogen treatment (control, injury, P407, and S100A8 + P407 groups). Black arrows show S100A8-PICs. S100A8-PICs were distributed around the glands and blood vessels in the control group. S100A8-PICs in the site of injury or adhesion in the injury group and in the P407 group. S100A8-PICs reverse migrated to the glands or blood vessels in the S100A8 + P407 group. Scale bar is 100 μm or 1 mm. (b) Endometrial thickness in the four groups (n = 15, from 5 rats; 3 position was measured randomly in an image of each rat, mean ± SEM). (c) The number of endometrial glands in each group (n = 10, from 10 rats, glands were counted from a uterine horn cross-section image at 40 × of each rat, mean ± SEM). (d) The interior wall integrity of uterine cavity in each group (n = 10, from 10 rats, data was measured from a uterine horn cross-section image at 40 × of each rat, mean ± SEM). (e) The number of endometrial crypts in each group (n = 10, from 10 rats, crypts were counted from a uterine horn cross-section image at 40 × of each rat, mean ± SEM). * Results with P values < 0.05 were considered significant. ** Results with P values < 0.01 were considered very significant.

Fig. 3

Effects of S100A8 on the distribution of Ki-67-positive proliferating cells in the uteri and the proliferation of endometrial cells. (a) Ki-67-positive proliferating cells were distributed in the endometrium or uterine interior wall in all groups (control, injury, P407, and S100A8 + P407 groups). Scale bar is 100 or 400 μm. (b) Number of Ki-67-positive cells in the endometrium of each group (n = 20, from ten rats, the number was counted from two immunohistochemical images of each sample, mean ± SEM). (c) Number of Ki-67-positive cells in the uterine interior wall of each group (n = 20, from ten rats, the number was counted from two immunohistochemical images of each sample, mean ± SEM). (d) Effects of S100A8 and its receptor inhibitor (FPS-ZM1) on the proliferation of cultured endometrial cells in vitro (n = 5, mean ± SEM). Scale bar is 100 μm. * Results with P values < 0.05 were considered significant. ** Results with P values < 0.01 were considered very significant.

Fig. 4

Effects of S100A8 on tight junctions of endometrial epithelium or endometrial cells. (a) Immunohistochemical staining showing Claudin-1 distribution in the endometrium of different groups (control, injury, P407, and P407 + S100A8 groups); black arrows show Claudin-1 accumulated below the cell membrane of the lateral contact zone between adjacent cells and red arrows show cells expressing Claudin-1 in the stroma. Scale bar is 100 μm. (b) Claudin-1 expression in the endometrium of different groups (n = 10, mean ± SEM). (c) Western blot showing ZO-1 expression in endometrial cells of three different groups (control, S100A8, and S100A8 + FPS-ZM1 groups, n = 3, mean ± SEM). Actin was used as the loading control. The full-length bands can be found in supplementary Fig. 1. The bands of ZO-1 and actin cropped from different parts of the same gel. (d) Immunofluorescence showing the co-expression of Claudin-1 (red) and ZO-1 (green) in the endometrial cells of the three groups (n = 5, mean ± SEM). The nucleus is stained blue. The scale bar is 200 μm. * Results with P values < 0.05 were considered significant. ** Results with P values < 0.01 were considered very significant.

Fig. 5

Effects of S100A8 on endometrial “double-featured” cells. (a) Vimentin (red) and ZO-1 (green) immunofluorescent double staining of cultured endometrial cells. Scale bar is 200 μm. (b) Western blot showing CK-18 and vimentin expression in endometrial cells from each group (control, S100A8, and FPS-ZM1 groups; n = 3, mean ± SEM). The full-length bands can be found in supplementary Fig. 1. Actin was used as the loading control. As the target proteins have a similar molecular weight to that of the loading control, the bands of CK18/vimentin and actin were cropped from different gels. (c) Immunofluorescence showing vimentin expression in endometrial cells of each group (n = 5, mean ± SEM). Scale bar is 200 μm. * Results with P values < 0.05 were considered significant. ** Results with P values < 0.01 were considered very significant.