Increased mitogen-activated protein kinase kinase/extracellularly regulated kinase activity in human endometrial stromal fibroblasts of women with endometriosis reduces 3',5'-cyclic adenosine 5'-monophosphate inhibition of cyclin D1

Abstract

Endometriosis is characterized by endometrial tissue growth outside the uterus, due primarily to survival, proliferation, and neoangiogenesis of eutopic endometrial cells and fragments refluxed into the peritoneal cavity during menses. Although various signaling molecules, including cAMP, regulate endometrial proliferation, survival, and embryonic receptivity in endometrium of women without endometriosis, the exact molecular signaling pathways in endometrium of women with disease remain unclear. Given the persistence of a proliferative profile and differential expression of genes associated with the MAPK signaling cascade in early secretory endometrium of women with endometriosis, we hypothesized that ERK1/2 activity influences cAMP regulation of the cell cycle. Here, we demonstrate that 8-Br-cAMP inhibits bromodeoxyuridine incorporation and cyclin D1 (CCND1) expression in cultured human endometrial stromal fibroblasts (hESF) from women without but not with endometriosis. Incubation with serum-containing or serum-free medium resulted in higher phospho-ERK1/2 levels in hESF of women with vs. without disease, independent of 8-Br-cAMP treatment. The MAPK kinase-1/2 inhibitor, U0126, fully restored cAMP down-regulation of CCND1, but not cAMP up-regulation of IGFBP1, in hESF of women with vs. without endometriosis. Immunohistochemistry demonstrated the highest phospho-ERK1/2 in the late-secretory epithelial and stromal cells in women without disease, in contrast to intense immunostaining in early-secretory epithelial and stromal cells in those with disease. These findings suggest that increased activation of ERK1/2 in endometrial cells from women with endometriosis may be responsible for persistent proliferative changes in secretory-phase endometrium.

Figure 1

cAMP regulation of CCND1 expression in endometrial stromal fibroblasts from women without vs. with endometriosis. A, Endometrial stromal fibroblasts from patients without (no endo, gray bars, n = 4) endometriosis were treated with vehicle (Ctrl) or 0.5 mm 8-bromo-cAMP (cAMP) for 4 d in 2% serum-containing medium, and CCND1 protein expression was analyzed by Western blot. B, Endometrial stromal fibroblasts from patients without (n = 6) or with (endo, white bars, n = 6) endometriosis were treated as above, and CCND1 mRNA expression was analyzed by quantitative RT-PCR. All values were normalized by RPL19 for total actin and mRNA expression for protein levels. C, Endometrial stromal fibroblasts from patients without (n = 5) or with (n = 4) endometriosis were treated as above, and cell proliferation was analyzed by BrdU incorporation. Bar graphs (least square means ± sem) represent fold changes relative to control (Ctrl, no endo, without cAMP). Means with asterisks and same letters indicate significant differences at P < 0.05 by Student’s t test and two-way ANOVA followed by Bonferroni post hoc test, respectively.

Figure 2

Inhibition of PKA by H89 in endometrial stromal fibroblasts from women without vs. with endometriosis. A and B, Endometrial stromal fibroblasts from patients without (no endo, gray bars, n = 6) or with (endo, white bars, n = 6) endometriosis were treated with DMSO vehicle (no inh), 1 μm H89, or 5 μm H89 for 30 min in 2% serum-containing medium, followed by addition of vehicle (Ctrl) or 0.5 mm cAMP for 4 d. Treatments and medium were refreshed every 2 d. CCND1 (A) and IGFBP1 (B) mRNA expression were analyzed by quantitative RT-PCR. All values were normalized by RPL19. Bar graphs (least square means ± sem) represent fold changes relative to control (no inh/no endo). Means with same letters indicate significant differences within treatments at P < 0.05 using two-way ANOVA by ranks, followed by Bonferroni post hoc test. Means with asterisks indicate significant differences in endo vs. corresponding no endo treatment at P < 0.05 using Mann-Whitney U test.

Figure 3

ERK phosphorylation in endometrial stromal fibroblasts incubated with 2% serum-containing or serum-free medium. A, Endometrial stromal fibroblasts from patients without (no endo, gray bars, n = 6) or with (endo, white bars, n = 6) endometriosis were incubated in 2% serum-containing medium for 0 min, 10 min, 1 h, or 2 d without refreshing the medium (A), treated with vehicle (Ctrl) or 0.5 mm 8-bromo-cAMP (cAMP) in 2% serum-containing medium for 2 d without refreshing the medium (B and C), or incubated in serum-free containing medium for 24 h (D). Cell lysates were analyzed for Tyr₂₀₄-phosphorylated ERK (pERK) and total ERK by Western blot. Values for pERK (C and E) and total ERK (F) were normalized by total ERK and total actin protein, respectively. Representative blots (A, B, and D) and graphical representations (C, E, and F) are shown as least square means ± sem and presented as fold change relative to control (Ctrl, no endo). Means with asterisks and same letters indicate significant differences at P < 0.05 using Student’s t test and two-way ANOVA followed by Bonferroni post hoc test, respectively. G, Endometrial stromal fibroblasts from patients without (no endo, n = 6) or with (endo, n = 6) endometriosis were incubated in serum-free medium for 24 h. ERRFI1 mRNA expression was analyzed by quantitative RT-PCR. Values were normalized by RPL19 gene. Bar graphs (least square means ± sem) represent fold changes relative to control (no endo). Means with asterisks indicate significant differences at P < 0.05 by Student’s t test.

Figure 4

Contribution of MEK/ERK to cAMP regulation of CCND1 and IGFBP1 expression in cultured endometrial stromal fibroblasts. A–C, Endometrial stromal fibroblasts from patients without (no endo, gray bars, n = 9) or with (endo, white bars, n = 7) endometriosis were treated with DMSO (vehicle, no inh) or 10 μm U0126 for 30 min in 2% serum-containing medium, followed by addition of vehicle (Ctrl) or 0.5 mm 8-bromo-cAMP for 4 d. Treatments and medium were refreshed every 2 d. CCND1 (A) and IGFBP1 (B) mRNA expression was analyzed by quantitative RT-PCR, and values were normalized to the RPL19 gene. C, IGFBP1 protein secretion was analyzed by ELISA, and values were normalized to total RNA. All bar graphs (least square means ± sem) represent fold changes relative to control (Ctrl, no endo and no treatment). Means with same letters indicate significant differences within treatments at P < 0.05 using two-way ANOVA by ranks, followed by Bonferroni post hoc test. Means with asterisks indicate significant differences in endo vs. corresponding no endo treatment at P < 0.05 using Mann-Whitney U test.

Figure 5

Immunohistochemistry for ERK phosphorylation in eutopic endometrium of women with or without endometriosis. A, Representative photomicrographs of immunostaining for Tyr₂₀₄-phosphorylated ERK (pERK) (brown) in paraffin sections of human endometrium from women without (no endo, white bars) vs. with (endo, gray bars) endometriosis during different phases of the menstrual cycle. Panels are shown at ×200 magnification. B and C, Graphical representation of pERK staining in stromal cells (B) and epithelial cells (C) is shown as least square means ± sem and presented as percent pERK-positive cells. Means with same letters indicate significant differences across the menstrual cycle at P < 0.05 using Kruskal-Wallis ANOVA, followed by Dunn’s multiple comparison test. Means with asterisks indicate significant differences in endo vs. corresponding no endo treatment at P < 0.05 using Mann-Whitney U test. For PE phase, no endo n = 3 and endo n = 5; for ES, no endo n = 5 and endo n = 6; for MS, no endo n = 3 and endo n = 4); for LS, no endo n = 4. EP, Epithelial cells; ST, stromal cells.