Interleukin-1β inhibits estrogen receptor-α, progesterone receptors A and B and biomarkers of human endometrial stromal cell differentiation: implications for endometriosis

Abstract

Human blastocyst nidation in the uterus and successful pregnancy require coordinated endometrial expression of estrogen receptor (ER)-α, progesterone receptors (PR)-A and -B and the gap junction protein, connexin (Cx)43. Our prior work established that inflammation associated with conditions of reduced fecundity, particularly endometriosis, can perturb eutopic decidual function. In the current studies, we have modeled endometrial decidualization in primary human endometrial stromal cell cultures derived from normal controls (NESC) and from the eutopic endometria of women with endometriosis (EESC) to test the hypothesis that a proinflammatory cytokine, interleukin (IL)-1β, can disrupt stromal cell differentiation. The cells were grown under a standard protocol with hormones (10 nM 17β-estradiol, 100 nM progesterone and 0.5 mM dibutyryl cAMP) for up to 7 days in the absence or presence of IL-1β. Time-course experiments showed that IL-1β compromised decidual function in both NESC and EESC, which was accompanied by rapid phosphorylation of ER-α, PR and Cx43 and their cellular depletion. Inhibition of the extracellular signal-regulated kinase (ERK)1/2 pathway by a selective pharmacological blocker (PD98059) or siRNA interference, or the addition of hormones themselves, blocked the phosphorylation of ERK mediators; increased the production of steroid receptors, Cx43, prolactin, insulin-like growth factor binding protein-1 (IGFBP)-1 and vascular endothelial growth factor (VEGF) and accelerated the differentiation. The results indicate that inhibition of IL-1β can enhance decidualization in NESC and EESC in vitro. Strategies to interfere with this pathway might be implemented as an in vivo approach to enhance fertility in women with endometriosis and, potentially, other inflammatory pathologies.

Keywords: connexin 43; cytokines; decidualization; endometriosis; estrogen receptor-α; inflammation; interleukin-1β; progesterone receptors; signal transduction; uterus.

Figure 1

Immunohistochemistry of PR and ERK1/2 in endometrium. Immunoperoxidase histochemistry (brown precipitate) shows nuclear localization of PR in endometrial glands and stroma (left panel). ERK1/2 is identified in the cytoplasm of endometrial glands with both cytoplasmic and nuclear staining in the stroma (middle panel). A negative control, where the primary antibody was substituted with nonimmune serum, is included (right panel). All sections were counterstained with Mayer’s hematoxylin. Magnification ×200. ERK1/2, extracellular signal-regulated kinase (ERK)1/2; PR, progesterone receptors.

Figure 2

Western blots of endometrial proteins in NESC in response to IL-1β and hormones. (A) NESC treated with E₂, P₄ and cAMP for 3 (H 3d) or 7 (H 7d) days show upregulation of ER-α (66 kDa, panel 1), PR-A and -B (116 and 90 kDa, panel 2) and Cx43 (43 kDa, panel 3). Exposure to IL-1β for 3 or 7 days suppressed ER-α (panel 1), PR-A and -B (panel 2) and Cx43 (panel 3) expression, relative to control (Cont) cells. β-Actin levels (42 kDa) are shown. (B) A short time-course revealed the effects of IL-1β on ER-α (panel 1), PR-A/B (panel 2) and pro- (31 kDa) and processed IL-1β (17 kDa, panel 3). β-Actin levels are shown (panel 4). Full length images of the western blots are available on request. Cx43, connexin 43; ER-α, estrogen receptor-α; NESC, normal human endometrial stromal cells.

Figure 3

Model describing IL-1β signaling via ERK1/2, endometrial biomarkers and clinical consequences. The ERK1/2 MAP kinase enzyme cascade following IL-1β ligation is depicted, showing intermediates and signaling mechanisms. Arrows (←) indicate stimulatory effects, whereas ‘block’ symbols ( ⊥) represent inhibitory effects. H refers to the combination of E₂, P₄ and cAMP and postulated, related reproductive pathologies are noted. See text for other abbreviations. MEK1/2, mitogen-activated protein kinase kinase.

Figure 4

Effects of decidualizing hormones and PD98059 (PD) on NESC proteins: Cx43, ERK1/2, p90RSK, ER-α and PR-A/B (western blots). (A) The ERK1/2 inhibitor (PD), hormones (H) or the combination of the two were incubated for 3 (PD + H 3d) or 7 days (PD + H 7d) and their effects on Cx43 (panel 1), phospho-ERK1/2 (panel 2), phospho-p90RSK (panel 3) and β-actin levels (panel 4) were measured. (B) Cx43 (panel 1), ER-α (panel 2), PR-A/B (panel 3) and β-actin levels (panel 4) were also evaluated, relative to control ESC. Molecular weights (kDa) of the specific protein bands are indicated at the right. Full length images of the western blots are available on request.

Figure 5

Decidual hormones and PD alter NESC morphology. NESC cultured for 7 days in the presence of E₂, P₄ and cAMP (H 7d), PD (PD 7d) or the combination (PD + H 7d) were examined for decidual morphology. Cell shape indices were calculated and are presented in the text as mean ± SEM. Magnification ×400.

Figure 6

Hormones and PD increase secretion of decidual biomarkers in NESC (ELISA data). (A) Triplicate NESC cultures were incubated for 7 days with E₂, P₄ and cAMP (H 7d), PD (PD 7d) or the combination (PD + H 7d) and cell supernatants were assayed for (A) PRL, (B) IGFBP-1 and (C) VEGF. Means ± SEM of triplicate wells from a single representative subject are shown. Asterisk depicts differences among comparisons by ANOVA with Scheffé’s post hoc tests (^*P < 0.05). Similar findings were noted in three independent NESC preparations. PRL, prolactin; IGFBP-1, insulin-like growth factor binding protein-1; VEGF, vascular endothelial growth factor.

Figure 7

IL-1β and PD98059 differentially affect nuclear translocation and accumulation of phospho-ERK1/2 in NESC. Immunofluorescence cytochemistry in NESC monolayers was performed with anti-phospho-ERK1/2 antibodies (green signal, upper right panel) 20 min following IL-1β exposure. Control (upper left panel), PD (lower left panel) and PD + IL-1β (lower right panel). Magnification ×400. Pixel quantification and statistical analyses are provided in Table I.

Figure 8

Effects of IL-1β and hormones (H) on NESC phosphoprotein expression. (A) Kinetics of IL-1β effects on phospho-ERK1/2 (P-ERK1/2, panel 1), phospho-p90RSK (P-p90RSK, panel 2), phospho-Cx43 isoforms (at Ser367 and Ser368 residues; panels 3 and 4), phospho-ER-α (Ser118 and Ser167; panels 5 and 6), phospho-PR-A and -B (Ser294, panel 7) and β-actin (panel 8). (B) Brief exposure (20 or 40 min) to IL-1β reduced total Cx43 (panel 1) but increased phospho-MEK1/2, phospho-ERK1/2, phospho-p90 RSK and phosho-p70/85 S6Kinase (panels 2–5). Phosphorylation of ER-α and PR-A and -B is also shown (panels 6–8). β-Actin effects of decidual hormone treatment for 48 h (H 48 h) are presented in the far right lane. Full length images of the western blots are available on request.

Figure 9

Suppression of MAP kinase signal cascade by PD and effects of IL-1β in NESC and EESC. (A) NESC cultured for 20 min to 72 h with PD were evaluated for phospho-MEK1/2 (panel 1), phospho-ERK1/2 (panel 2), phospho-p90RSK (panel 3), phospho-Cx43 (Ser367 and Ser368 isoforms, panels 4 and 5, respectively), non-phosphorylated Cx43 (Ser368, panel 6) and β-actin (panel 7). Similar findings were noted in two independent ESC preparations. (B) The effects of IL-1β (+) on Cx43 (panel 1), ER-α (panel 2), PR-A and -B (panel 3) and β-actin (panel 4) are shown in four independent preparations each of NESC (1–4) and EESC (1–4). (C) In two additional preparations of EESC (5–6) and NESC (5–6), the effects of IL-1β and PD + IL-1β are compared to the control (Cont) for Cx43 (panel 1), ER-α (panel 2), PR (panel 3) and β-actin (panel 4). Full length images of the western blots are available on request. EESC, eutopic endometrial stromal cells.

Figure 10

Effects of ERK1/2 siRNA and hormones (H) on NESC morphology and proteins. (A) siRNA used to inhibit ERK1/2 in the absence and presence (+H) of hormones showed prominent morphological changes (panels 2 and 3, respectively) relative to controls treated with scrambled siRNA (panel 1). (B) ERK1/2 siRNA effects on phospho-ERK1/2 (panel 1), phospho-Cx43 isoforms (panels 2 and 3), ER-α (panel 4), PR-A and -B (panel 5) and β-actin (panel 6) are shown.