Activation of Endoplasmic Reticulum Stress in Granulosa Cells from Patients with Polycystic Ovary Syndrome Contributes to Ovarian Fibrosis

Abstract

Recent studies report the involvement of intra-ovarian factors, such as inflammation and oxidative stress, in the pathophysiology of polycystic ovary syndrome (PCOS), the most common endocrine disorder of reproductive age women. Endoplasmic reticulum (ER) stress is a local factor that affects various cellular events during a broad spectrum of physiological and pathological conditions. It may also be an important determinant of pro-fibrotic remodeling during tissue fibrosis. In the present study, we showed that ER stress was activated in granulosa cells of PCOS patients as well as in a well-established PCOS mouse model. Pharmacological inducers of ER stress, tunicamycin and thapsigargin, were found to increase the expression of pro-fibrotic growth factors, including transforming growth factor (TGF)-β1, in human granulosa cells, and their expression also increased in granulosa cells of PCOS patients. By contrast, treatment of PCOS mice with an ER stress inhibitor, tauroursodeoxycholic acid or BGP-15, decreased interstitial fibrosis and collagen deposition in ovaries, accompanied by a reduction in TGF-β1 expression in granulosa cells. These findings suggest that ER stress in granulosa cells of women with PCOS contributes to the induction of pro-fibrotic growth factors during ovarian fibrosis, and that ER stress may serve as a therapeutic target in PCOS.

Figure 1

The expression of UPR genes and pro-fibrotic growth factors in granulosa-lutein cells from control (n = 10) and PCOS (n = 11) patients. The expression levels of (A–E) UPR genes, XBP1(S), HSPA5, ATF4, ATF6, and CHOP mRNA and (F,G) pro-fibrotic growth factors, TGF-β1 and CTGF mRNA in granulosa-lutein cells were measured by real-time PCR and normalized to that of GAPDH. Increased UPR gene expression was indicative of ER stress activation. The values represent means ± SEM. *p < 0.05.

Figure 2

Phospho-IRE1, phospho-PERK, CHOP, and TGF-β1 protein expression levels and the area of fibrotic tissue in the ovaries of control and PCOS patients. Immunohistochemical analysis was performed on the ovaries of 3 PCOS and 3 control patients. (A–D) Cross-sections of ovaries were stained with an anti-phospho-IRE1, anti-phospho-PERK, anti-CHOP, or anti-TGF-β1 antibody, counterstained with hematoxylin. (E) Controls for background level stained with isotype IgG and hematoxylin. (F) Cross-sections of ovaries were stained with Masson’s trichrome stain. Fibrotic tissue was stained blue. (A–D,F) (a–d) show the representative sections. Lower panels (c,d) show highly magnified views corresponding to (a,b). (e) show the quantitative analysis of (A–D) immunohistochemical staining and (F) Masson’s trichrome staining. (E) A right panel (b) shows a highly magnified view corresponding to a left panel (a). The scale bars in (A–D) (a,b) and (E) (a) indicate 50 μm, while those in (A–D) (c,d) and (E) (b) indicate 20 μm. The scale bars in (F) (a,b) and (F) (c,d) indicate 200 μm and 50 μm, respectively. *p < 0.05. GC, granulosa cell layer; NC, negative control.

Figure 3

XBP1(S) and HSPA5 mRNA expression levels, and CHOP, phospho-IRE1, phospho-PERK, and TGF-β1 protein expression levels in ovaries of control and PCOS mice. Three-week-old female mice were divided into two groups. The control group (n = 5) was s.c. injected daily with sesame oil for 20 days. The PCOS group (n = 5) was s.c. injected daily with DHEA (6 mg/100 g of body weight) for 20 days. The ovaries were collected on day 21. (A,B) Cross-sections of ovaries from control and PCOS mice were hybridized with a DIG-labeled antisense XBP1(S) or HSPA5 probe. (C–F) Cross-sections of ovaries from control and PCOS mice were stained with an anti-CHOP antibody counterstained with hematoxylin, or an anti-phospho-IRE1, anti-phospho-PERK, or anti-TGF-β1 antibody. (G) Controls for background level (a,b) hybridized with sense probe, (c) stained with isotype IgG, and (d) stained with isotype IgG and hematoxylin. (A–F) (a–d) show the representative sections. Lower panels (c,d) show highly magnified views corresponding to (a,b). (e) show the quantitative analysis of (A,B) in situ hybridization and (C–F) immunohistochemical staining. The scale bars in (A–F) (a,b) and (G) (a–d) indicate 100 μm. The scale bars in (A,B) (c,d) and (C–F) (c,d) indicate 20 μm and 50 μm, respectively. *p < 0.05. NC, negative control; ISH, in situ hybridization; IHC, immunohistochemistry.

Figure 4

Effects of ER stress on TGF-β1 and CTGF mRNA expression levels and TGF-β1 secretion by cultured human granulosa-lutein cells. (A–E) Granulosa-lutein cells were treated with tunicamycin at 2.5 μg/mL for 0, 3, 9, or 24 h. (F–K) Granulosa-lutein cells were pre-incubated with TUDCA at 1 mg/mL for 24 h, followed by treatment with tunicamycin at 2.5 μg/mL for 24 h. (L,M) Granulosa-lutein cells were transfected with siRNA (50 nM) or negative control siRNA (50 nM) for 24 h and then, treated with tunicamycin (2.5 μg/mL) for 24 h. (A,F,M) TGF-β1, (B,G) CTGF, (C,I,L) XBP1(S), (D,J) HSPA5, and (E,K) CHOP mRNA expression levels in granulosa-lutein cells were measured by real-time PCR and normalized to that of GAPDH. The values represent means ± SEM of triplicate or quadruplicate experiments, relative to the mean control value. (H) The bioactive TGF-β1 level in cultured cell supernatants was measured by ELISA. The values represent means ± SEM of sextuplicate experiments. The results are representative of at least three independent experiments using three different samples. The letters denote significant differences between groups. *p < 0.05. Tm, tunicamycin; TUDCA, tauroursodeoxycholic acid.

Figure 5

Effects of ER stress inhibitors on ovarian fibrosis in PCOS mice. Three-week-old female mice were divided into four groups. The control group (n = 5) was s.c. injected daily with sesame oil, followed by the oral administration of saline for 20 days. The PCOS group (n = 5) was s.c. injected daily with DHEA (6 mg/100 g of body weight), followed by the oral administration of saline for 20 days. The PCOS + TUDCA group (n = 5) was s.c. injected daily with DHEA, followed by the oral administration of TUDCA (50 mg/100 g of body weight) for 20 days. The PCOS + BGP-15 group (n = 5) was s.c. injected daily with DHEA, followed by the oral administration of BGP-15 (3 mg/100 g of body weight) for 20 days. The ovaries were collected on day 21. (A) Cross-sections of ovaries were stained with Masson’s trichrome stain. Fibrotic tissue was stained blue. (B–F) Cross-sections of ovaries were stained with collagen type I, collagen type IV, TGF-β1, phospho-IRE1, and phospho-PERK. (G) Cross-sections of ovaries were hybridized with a DIG-labeled antisense XBP1(S) probe. (A–G) (a–h) show the representative sections. Lower panels (e–h) show highly magnified views corresponding to (a–d). (i) show the quantitative analysis of (A) Masson’s trichrome staining, (B–F) immunohistochemical staining, and (G) in situ hybridization. The scale bars in (A–C) (a–d) and (A–C) (e–h) indicate 50 μm and 20 μm, respectively. The scale bars in (D–G) (a–d) and (D–G) (e–h) indicate 100 μm and 50 μm, respectively. (H–J) XBP1(S), TGF-β1, and CTGF mRNA expression levels in mouse ovaries were measured by real-time PCR and normalized to that of GAPDH. The letters denote significant differences between groups. TUDCA, tauroursodeoxycholic acid.