Loss of KLF15 impairs endometrial receptivity by inhibiting EMT in endometriosis

Abstract

The impaired endometrial receptivity is a major factor contributing to infertility in patients with endometriosis (EM), but the underlying mechanism remains unclear. Our study aimed to investigate the role of Kruppel-like factor 15 (KLF15) in endometrial receptivity and its regulation in EM. We observed a significant decrease in KLF15 expression in the mid-secretory epithelial endometrial cells of EM patients compared to normal females without EM. To confirm the role of KLF15 in endometrial receptivity, we found a significantly reduced KLF15 expression and a significant decrease in embryo implantation number in the rat model via uterine horn infection with siRNA. This highlights the importance of KLF15 as a regulator receptivity. Furthermore, through ChIP-qPCR, we discovered that the progesterone receptor (PR) directly binds to KLF15 promoter regions, indicating that progesterone resistance may mediate the decrease in KLF15 expression in EM patients. Additionally, we found that the mid-secretory endometrium of EM patients exhibited impaired epithelial-mesenchymal transition (EMT). Knockdown of KLF15 upregulated E-cadherin and downregulated vimentin expression, leading to inhibited invasiveness and migration of Ishikawa cells. Overexpression KLF15 promotes EMT, invasiveness, and migration ability, and increases the attachment rate of JAR cells to Ishikawa cells. Through RNA-seq analysis, we identified TWIST2 as a downstream gene of KLF15. We confirmed that KLF15 directly binds to the promoter region of TWIST2 via ChIP-qPCR, promoting epithelial cell EMT during the establishment of endometrial receptivity. Our study reveals the involvement of KLF15 in the regulation of endometrial receptivity and its downstream effects on EMT. These findings provide valuable insights into potential therapeutic approaches for treating non-receptive endometrium in patients with EM.

Keywords: KLF15; endometrial epithelial; endometriosis; endometrium receptivity; epithelial–mesenchymal transition.

Figure 1

The expression pattern of KLF15 in endometriosis. (A) Analysis of KLF15 mRNA expression in proliferative and mid-secretory endometrium of healthy women of childbearing age with or without endometriosis using GSE6364 dataset. (B) Representative photomicrographs of KLF15 protein expression in EM and control group. The left photos were shown in 200× magnification, the right photos were shown in 400× magnification. (C) Immunohistochemical H-score of KLF15 proteins in proliferative (n = 6) and mid-secretory phase (n = 6) endometrium from women with (n = 12) or without (n = 12) endometriosis. (D) RT-qPCR analysis of KLF15 expression in different phases of endometrium in patients with (n = 12) or without endometriosis (n = 12). *P < 0.05, **P < 0.01, using one-way ANOVA. Prolif, proliferative phase; Mid-Sec, mid-secretory phase; EM, endometriosis. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 2

Attenuation of KlfF15 in rat models. (A) Establishment of the KLF15 knockdown rat model. On the 4th day of pregnancy, 50 µL 10 µM Klf15 siRNA were injected into the left uterine cavity from the uterine horn near the ovary, and the same volume of 10 µM NC-siRNA (negative control) was injected into the opposite side of the uterine horn as a control. The rats were sacrificed on the 7th day of pregnancy to observe the number of embryo implantation sites. (B) The decreased number of implantation sites after KLF15 siRNA injected into the left horn of uteri (n = 8). (C) Representative photomicrographs of KLF15 protein expression in rat uterus tissues injected with NC-siRNA or KLF15-siRNA. Photos are shown in 400× magnification. (D) The protein expression of KLF15 in rat uterus tissues injected with NC-siRNA or KLF15-siRNA (n = 8). *P < 0.05, using t-test. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 3

KLF15 is regulated by progesterone receptor. (A) The protein expression of KLF15 after treatment of different doses of MPA with or without RU486 (n = 4). (B) Luciferase assays of the activation potential of the region containing pGL3-KLF15 promoter in the presence of PGR overexpression plasmid, MPA, or PGR overexpression plasmid combined with MPA (n = 4). *P < 0.05, using one-way ANOVA. (C) Putative binding area of PGR on KLF15 promotor region and ChIP-qPCR analysis of recruitment of PGR to the putative binding area on KLF15 promoter region after MPA treatment (n = 3). *P < 0.05, using t-test. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 4

PGR expression in endometriosis. (A) Representative photomicrographs of PGR protein expression in EM and control group. The left photos were shown in 200× magnification, the right photos were shown in 400× magnification. (B) Immunohistochemical H-score of PGR proteins in proliferative (n = 6) and mid-secretory phase (n = 6) endometrium from women with (n = 12) or without (n = 12) endometriosis. (C) RT-qPCR analysis of PGR expression in different phases of endometrium in patients with (n = 12) or without endometriosis (n = 12). *P < 0.05. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 5

E-cadherin and vimentin expression in endometriosis. (A, D) Representative photomicrographs of E-cadherin and vimentin protein expression in EM and control group. The left photos were shown in 200× magnification, the right photos were shown in 400× magnification. (B, E) Immunohistochemical H-score of E-cadherin and vimentin proteins in proliferative and mid-secretory phase endometrium from women with (n = 12) or without endometriosis (n = 12). (C, F) RT-qPCR analysis of E-cadherin and vimentin expression in different phases of endometrium in patients with (n = 12) or without endometriosis (n = 12). *P < 0.05. **P < 0.01, ***P < 0.001. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 6

KLF15 regulates EMT of Ishikawa cells. (A) Representative images of E-cadherin and vimentin expression in Ishikawa cells transfected with KLF15 siRNA. (B, C) Western blot and RT-qPCR analysis of E-cadherin and vimentin expression after KLF15 knockdown (n = 4). (D) Chamber transwell assays of cellular invasion or migration (n = 6). (E) Wound healing assay analysis Ishikawa cells migration rate after KLF15 siRNA transfection (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, using t-test. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 7

RNA-seq analysis of genes related to KLF15 downregulation. (A) Volcano diagram of upregulated and downregulated genes after KLF15 knockdown. (B) Heatmap of differential gene expression after KLF15 knockdown. Gene Ontology analysis revealed upregulated (C) and downregulated (D) gene function after KLF15 knockdown (n = 3). (E) RT-qPCR analysis of 16 DE gene expression (n = 3), *P < 0.05, *P < 0.01, **P < 0.001, ****P < 0.0001, using t-test. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 8

KLF15 modulates EMT via TWIST2 expression. (A) Putative binding area of KLF15 on TWIST2 promoter region. (B) Chromatin immunoprecipitation (ChIP) qPCR analysis of recruitment of KLF15 to PP2 of TWIST2 was increased (n = 3). (C) Knockdown of TWIST2 significantly inhibits KLF15 overexpression-induced EMT markers expression (n = 3). LV-KLF15: KLF15 overexpression. LV-NC, Negative control overexpression plasmid. *P < 0.05, **P < 0.01. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.

Figure 9

KLF15 is regulated by PGR, and KLF15 promotes TWIST2 induced EMT of endometrial epithelial cells during embryo implantation. A full colour version of this figure is available at https://doi.org/10.1530/JOE-23-0319.