The endometrial response to chorionic gonadotropin is blunted in a baboon model of endometriosis

Abstract

Endometriosis-associated infertility has a multifactorial etiology. We tested the hypothesis that the endometrial response to the early embryonic signal, human chorionic gonadotropin (hCG), alters over time in a nonhuman primate model of endometriosis. Animals with experimental or spontaneous endometriosis were treated with hCG (30 IU/d), from d 6 after ovulation for 5 d, via an oviductal cannula. Microarray analysis of endometrial transcripts from baboons treated with hCG at 3 and 6 months of disease (n=6) identified 22 and 165 genes, respectively, whose levels differed more than 2-fold compared with disease-free (DF) animals treated with hCG (P<0.01). Quantitative RT-PCR confirmed abnormal responses of known hCG-regulated genes. APOA1, SFRP4, and PAPPA, which are normally down-regulated by hCG were up-regulated by hCG in animals with endometriosis. In contrast, the ability of hCG to induce SERPINA3 was lost. Immunohistochemistry demonstrated dysregulation of C3 and superoxide dismutase 2 proteins. We demonstrate that this abnormal response to hCG persists for up to 15 months after disease induction and that the nature of the abnormal response changes as the disease progresses. Immunohistochemistry showed that this aberrant gene expression was not a consequence of altered LH/choriogonadotropin receptor distribution in the endometrium of animals with endometriosis. We have shown that endometriosis induces complex changes in the response of eutopic endometrium to hCG, which may prevent the acquisition of the full endometrial molecular repertoire necessary for decidualization and tolerance of the fetal allograft. This may in part explain endometriosis-associated implantation failure.

Figures

Figure 1

Experimental paradigm in the nonhuman primate of simulated early pregnancy in an induced model of endometriosis. Baboons (P. anubis) were inoculated with menstrual endometrium during two consecutive menses. In the subsequent cycle, a cannula was secured into the oviduct during the window of implantation. At this time eutopic endometrium was harvested by endometrectomy serving as unstimulated control endometrial tissue. Three and six months (mo) after induction of disease, hCG was infused into the oviductal lumen, mimicking the secretory behavior of the preimplantation embryo. Eutopic endometrium was subsequently harvested after 4–5 d of hCG infusion. Microarray analysis was performed to determine whether eutopic endometrium responded aberrantly to the early embryonic signal as a proposed mechanism for endometriosis-associated infertility.

Figure 2

Morphological characterization and ASMA immunostaining of eutopic endometrium after hCG infusion. hCG infusion in DF baboons (A and C) resulted in the generation of a luminal epithelial plaque (A) and the induction of ASMA in the endometrial stroma (C). In animals with 6 months endometriosis (B and D) hCG failed to elicit both an epithelial plaque (B) or ASMA production (D) in endometrial stromal cells. Bar, 100 μm.

Figure 3

Dendogram of unsupervised hierarchical clustering of eutopic endometrial gene expression in a model of simulated early pregnancy in baboons with endometriosis. Baboons with experimental and spontaneous endometriosis were treated with hCG during the window of implantation (6mo+hCG). After Illumina Beadchip interrogation of endometrial transcripts from 6mo+hCG animals (n = 6) and DF animals similarly treated with hCG (DF+hCG; n = 6), nonhierarchical clustering was performed using centered correlation and average linkage analysis. Spont1/2, Animals with spontaneous disease; induced 1/2/3/4, animals with experimentally induced disease. Animals in italics were considered as outliers and excluded from further analysis.

Figure 4

The eutopic endometrial molecular response to hCG is blunted in baboons with endometriosis. Total endometrial RNA was extracted from baboons with 3 months (n = 4) and 6 months (n = 5) endometriosis and DF animals (n = 5) after hCG infusion during the window of implantation. Additional endometrial RNAs were extracted from unstimulated, DF animals (DF-hCG; n = 6). qRT-PCR analysis revealed that mRNA levels of APOA1 (A), SFRP4 (B), and PAPPA (C) were significantly increased in both 3mo+hCG (3 mo) and 6mo+hCG (6 mo) animals compared with DF controls (DF+hCG). Conversely, mRNA levels of SERPINA3 (D) were significantly decreased after hCG infusion in animals with endometriosis compared with DF+hCG control. Columns represent the median normalized values against ribosomal 18S. Error bars, Interquartile range. a, Significant difference to DF+hCG, P < 0.05; Mann-Whitney.

Figure 5

C3 and SOD2 proteins are down-regulated by hCG in the eutopic endometrium of baboons with endometriosis. The distribution of C3 (left panel) and SOD2 (right panel) was determined in DF-hCG unstimulated animals, DF animals after hCG infusion (DF+hCG), and throughout the progression of endometriosis from 3 to 12 months of disease. Minimal levels of C3 (A) and SOD2 (G) proteins were detectable in endometria of DF-hCG animals. After hCG infusion in DF+hCG animals, both C3 (B) and SOD2 (H) endometrial protein levels were markedly increased. The ability of hCG to induce C3 and SOD2 was dramatically reduced in animals with endometriosis at 3mo+hCG (C and I), 6mo+hCG (D and J), 12mo+hCG (E and K), and 15mo+hCG (F and L). Inserts in F and L represent preimmune controls of C3 and SOD2, respectively. Bars, 50 μm.

Figure 6

Molecular dysregulation of the endometrial response to hCG is apparent throughout progression of disease in baboons with endometriosis. Total endometrial RNA was extracted from baboons with 3 months (mo; n = 4), 6 months (n = 5), 12 months (n = 3), and 15 months (n = 2) endometriosis and DF animals (DF+hCG; n = 5) after hCG infusion during the window of implantation. Additional RNAs were extracted from unstimulated DF animals (DF-hCG; n = 6) and unstimulated animals with 1 month endometriosis (n = 4). qRT-PCR analysis revealed a loss of CG-induced down-regulation of endometrial expression of APOA1 (A) and SFRP4 (B). In contrast, SERPINA3 mRNA was up-regulated by hCG in DF animals (C). Columns represent medians normalized values against ribosomal 18S. Error bars, Interquartile range. a, Significant difference to DF-CG, Mann Whitney U test, P < 0.05; b, a significant difference to DF+hCG; Kruskal Wallis ANOVA, P < 0.05).

Figure 7

LHCGR immunostaining of eutopic endometrium from DF and endometriotic animals. Immunohistochemistry was performed using polyclonal antibodies against the extracellular domain of the LHCGR. LHCGR protein was detectable in glandular epithelial cells in the functionalis (A) and basalis (B) endometrium of unstimulated DF baboons (DF-hCG) during the window of implantation. After hCG infusion in DF animals (DF+hCG), glandular LHCGR protein levels reduced (C), whereas increased levels were observed surrounding the spiral arteries, predominantly in the basalis endometrium (D). Normal baseline levels of LHCGR were present in unstimulated functionalis (E) and basalis (F) eutopic endometrium of animals with endometriosis at 1 month of disease. However, the hCG-induced down-regulation of glandular LHCGR in animals with endometriosis was not evident in 6mo+hCG animals (G), whereas the up-regulation adjacent to the spiral arteries remained evident (H). Bar, 50 μm. SA, Spiral artery.