Medroxyprogesterone acetate inhibits wound closure of human endometrial epithelial cells and stromal fibroblasts in vitro

Abstract

Mucosal integrity in the endometrium is essential for immune protection. Since breaches or injury to the epithelial barrier exposes underlying tissue and is hypothesized to increase infection risk, we determined whether endogenous progesterone or three exogenous progestins (medroxyprogesterone acetate (MPA), norethindrone (NET), and levonorgestrel (LNG)) used by women as contraceptives interfere with wound closure of endometrial epithelial cells and fibroblasts in vitro. Progesterone and LNG had no inhibitory effect on wound closure by either epithelial cells or fibroblasts. MPA significantly impaired wound closure in both cell types and delayed the reestablishment of transepithelial resistance by epithelial cells. In contrast to MPA, NET selectively decreased wound closure by stromal fibroblasts but not epithelial cells. Following epithelial injury, MPA but not LNG or NET, blocked the injury-induced upregulation of HBD2, a broad-spectrum antimicrobial implicated in wound healing, but had no effect on the secretion of RANTES, CCL20 and SDF-1α. This study demonstrates that, unlike progesterone and LNG, MPA and NET may interfere with wound closure following injury in the endometrium, potentially conferring a higher risk of pathogen transmission. Our findings highlight the importance of evaluating progestins for their impact on wound repair at mucosal surfaces.

Figure 1

MPA inhibits wound closure by endometrial epithelial cells. (A) Wound closure of endometrial epithelial cells grown in transwell inserts at 24 h post-scratch following pre-treatment for 48 h with 100 nM progesterone. Each symbol represents a single individual (n = 10). (B) Representative images of endometrial epithelial cells grown in 96-well plates scratched using the IncuCyte WoundMaker in the absence (control) or presence of MPA (100 nM) at 0, 12, 24, and 48 h at  × 10 magnification using the IncuCyte Zoom. Time of scratch is 0 h. The solid white line represents the epithelial wound front. (C) Relative wound density of endometrial epithelial cells grown in 96-well plates was measured at the indicated time points for 24 h post-scratch in the presence (100 nM) or absence of MPA. Data shown is from another representative patient than (B). Each symbol represents the mean + /− SEM from a single timepoint. Each timepoint is the average of triplicate or quadruplicate wells. (D) Wound closure of endometrial epithelial cells grown in transwell inserts at 24 h post-scratch following pre-treatment with 100 nM MPA, NET, and LNG for 48 h prior to scratch. MPA, NET, LNG, and Progesterone were maintained in cell culture media at 100 nM following scratch. Each symbol represents a single individual (n = 22).

Figure 2

MPA decreases barrier integrity and increases paracellular permeability of endometrial epithelial cells treated with MPA. (A) Representative transepithelial resistance (TER) profile of endometrial epithelial cells grown in transwell inserts pre- and post-scratch during exposure to 100 nM MPA, NET, LNG, or Progesterone. Scratch was performed at 0 h. Each symbol represents the mean + /− SEM from a single time point. Data shown is from a representative patient. (B) TER of endometrial epithelial cells grown in transwell inserts 2 h and (C) 24 h following scratch treated with 100 nM MPA, NET, and LNG. Each symbol represents a single individual (n = 11). (D) FITC-Dextran 4KDa flux assay 24 h post-scratch performed over 6 h. Cells were treated with MPA, NET, or LNG at 100 nM for 48 h prior to scratch, and maintained in cell culture media at 100 nM post-scratch. Bars represent mean + /− SEM. Each symbol represents a single individual (n = 6). Values are normalized to the control unscratched condition which is set to 100. *p < 0.05; **p < 0.01.

Figure 3

MPA and NET inhibit wound closure by endometrial stromal fibroblasts. (A) Representative images of endometrial stromal fibroblasts grown in 96-well plates scratched using an IncuCyte WoundMaker in the absence (control) or presence of MPA (100 nM) at 0, 12, 24, and 48 h at  × 10 magnification using the IncuCyte Zoom. Time of scratch is 0 h. The solid white line represents the fibroblast wound front. (B) Relative wound density of endometrial stromal fibroblasts grown in 96-well plates 24 h post-scratch in the presence (100 nM) or absence of MPA. Data shown is from a representative patient. Each symbol represents the mean + /− SEM from a single timepoint. (C) Wound closure of endometrial stromal fibroblasts grown in 24-well plates at 24 h post-scratch following treatment with 100 nM MPA, NET, LNG, or Progesterone for 48 h prior to scratch. MPA, NET, LNG, and Progesterone were maintained in cell culture media at 100 nM following scratch. Each symbol represents a single individual (n = 5). (D) Relative wound density of endometrial stromal fibroblasts grown in 96-well plates 24 h post-scratch following treatment with 100 nM MPA, NET, LNG, or Progesterone. Each symbol represents a single individual (MPA n = 10; NET n = 7; LNG n = 7; Progesterone n = 7). *p < 0.05.

Figure 4

MPA and NET exhibit a dose-dependent suppression of wound closure in endometrial epithelial cells and stromal fibroblasts. (A) Wound closure by endometrial epithelial cells grown in transwell inserts and (B) wound closure by endometrial stromal fibroblasts grown in 24-well plates 24 h post-scratch following treatment with 0.1, 1, 10, and 100 nM MPA, NET, and LNG for 48 h prior to scratch. MPA, NET, and LNG were maintained in the cell culture media following scratch. Bars represent mean + /− SEM. Each symbol represents a single individual (epithelial cell n = 6; stromal fibroblast n = 5). *p < 0.05; **p < 0.01.

Figure 5

Wound closure is not accompanied by cell proliferation over 48 h. Cell proliferation following injury of (A) endometrial epithelial cells grown in transwell inserts and (B) endometrial stromal fibroblasts grown in 24-well plates. Cells were treated with MPA, NET, or LNG at 100 nM for 48 h prior to scratch, and maintained in cell culture media at 100 nM following scratch. Cell proliferation was determined using a CellTiter cell proliferation assay performed 48 h post-scratch. Bars represent mean + /− SEM. Each symbol represents a single individual (epithelial cell n = 6; stromal fibroblast n = 9).

Figure 6

MPA inhibits the upregulation of HBD2 by polarized endometrial epithelial cells following injury. Secretion of (A) HBD2, (B) CCL20, (C) RANTES, and (D) SDF-1α by endometrial epithelial cells grown in transwell inserts 48 h post-scratch. Cells were treated for 48 h with 100 nM MPA, NET, and LNG prior to scratch and maintained in cell culture media following scratch. Bars represent mean + /− SEM. Secretion values were normalized to the control wells which were set to 100. Each symbol represents a single individual (Control n = 8; MPA n = 8; NET n = 4; LNG; n = 4). Mean secretion in controls samples: 850 pg/ml (HBD2), 575 pg/ml (CCL20), 37 pg/ml (RANTES), and 59 pg/ml (SDF-1α). Control untreated unscratched control, Control + SC scratched control, MPA unscratched MPA treated, MPA + SC scratched MPA treated, NET unscratched NET treated, NET + SC scratched NET treated, LNG unscratched LNG treated, LNG + SC scratched LNG treated. *p < 0.05.