Human Chorionic Gonadotropin modulates CXCL10 Expression through Histone Methylation in human decidua

Abstract

The process of implantation, trophoblast invasion and placentation demand continuous adaptation and modifications between the trophoblast (embryonic) and the decidua (maternal). Within the decidua, the maternal immune system undergoes continued changes, as the pregnancy progress, in terms of the cell population, phenotype and production of immune factors, cytokines and chemokines. Human chorionic gonadotropin (hCG) is one of the earliest hormones produced by the blastocyst and has potent immune modulatory effects, especially in relation to T cells. We hypothesized that trophoblast-derived hCG modulates the immune population present at the maternal fetal interface by modifying the cytokine profile produced by the stromal/decidual cells. Using in vitro models from decidual samples we demonstrate that hCG inhibits CXCL10 expression by inducing H3K27me3 histone methylation, which binds to Region 4 of the CXCL10 promoter, thereby suppressing its expression. hCG-induced histone methylation is mediated through EZH2, a functional member of the PRC2 complex. Regulation of CXCL10 expression has a major impact on the capacity of endometrial stromal cells to recruit CD8 cells. We demonstrate the existence of a cross talk between the placenta (hCG) and the decidua (CXCL10) in the control of immune cell recruitment. Alterations in this immune regulatory function, such as during infection, will have detrimental effects on the success of the pregnancy.

Figure 1

In vitro model of decidualization. A human endometrial stromal cell line (HESCs) was decidualized with 10 nM estradiol, 1 uM medroxyprogesterone acetate, and 500 uM 8-bromo cyclic AMP (cAMP) for 7 days. (A) Morphological changes associated with the decidualization process. The left panel shows the untreated HESCs, while the right panel shows the decidualized cells (DCS) after treatment with estradiol, medroxyprogesterone acetate, and 8-bromo cyclic AMP. (B) Increased expression of Prolactin and IGFBP1 in decidualized cells (DSC) compared to HESC. Data presented as mean ± SD and are shown for 6 independent experiments with each done in triplicate. *p < 0.001.

Figure 2

Recruitment of immune cells by stromal cell- derived factors. Supernatants from HESC and decidualized HESC (DSC) cells were collected after 5 days of culture (equal confluence) and used as conditioned media (CM) for testing their effect on the migration of peripheral blood mononuclear cells (PBMC) using a two- chamber migration assay. PBMCs were added to the upper chamber for 24 h. The migrated cells (lower chamber) were collected, phenotyped, and counted by flow cytometry. (A) Representative figure of the different cell types collected in the lower chamber recruited by HESC and DSC’s conditioned media. Both CD4 and CD8 T cells migrated towards HESC CM and DSC CM. (B) Quantification of the total number of T cells recruited towards conditioned media obtained from HESC and DSC. Data shown are for 6 independent experiments with each group done in triplicate. *p < 0.001. (C) Flow Cytometry Analysis of CD4 and CD8 T cells. Percentage of CD4 and CD8 T cells recruited by DSC CM compared to HESC CM. DSC CM recruits significantly lower number of CD4 and CD8 T cells compared to HESC CM. Data presented as mean/SEM and are for 6 independent experiments with each group done in triplicate. *p < 0.05. **p < 0.01.

Figure 3

Regulation of CXCL10 expression in stromal versus decidual cells. (A) Effect of cAMP and decidualization on CXCL10 mRNA expression. HESC were treated with 500 uM 8-bromo cAMP alone or decidualized with 10 nM estradiol, 1 uM medroxyprogesterone acetate, and 500 uM 8-bromo cyclic AMP for 7 days. Note the significant inhibition of CXCL10 mRNA expression by cAMP in HESC, similar to the decrease observed in the decidualized cells (DSC) *p < 0.05. Data presented as mean ± SD and are from 3 independent experiments with each group done in triplicate. (B) Quantification of CXCL10 protein expression in HESC treated with cAMP or decidualized with estrogen, progesterone and cAMP. cAMP treatment and decidualized cells (DSC) are associated with decreased CXCL10 protein expression. *p < 0.001. **p < 0.05. Data presented as mean ± SD and are from 3 independent experiments with each group done in triplicate. (C) Effect of estrogen, progesterone and cAMP on CXCL10 mRNA expression. HESC were treated with 10 nM estradiol, 1 uM medroxyprogesterone acetate (MPA), and 500 uM 8-bromo cAMP individually or under the combination of estradiol, MPA and cAMP for 7 days. CXCL10 mRNA was determined by qPCR. Note that only cAMP has a major effect on CXCL10 mRNA expression. Data presented as mean ± SD and are from 3 independent experiments with each group done in triplicate.

Figure 4

Regulation of CXCL10 expression by hCG in human endometrial stromal cells (HESC). (A) HESC were treated with hCG 100 IU/ml for 24 h and CXCL10 mRNA expression was determined by qPCR. Note the significant decrease of CXCL10 mRNA expression after hCG treatment. Data presented as mean ± SD and are for 6 independent experiments with each group done in triplicate *p < 0.005. (B) Time response to hCG treatment (100 IU/ml). HESC were treated with hCG (100 IU/ml) for 2, 12 and 24 h and CXCL10 mRNA expression was determined by qPCR. CXCL10 mRNA expression decreased within 2 hours of hCG treatment and remained low up to 24 hours of treatment. (C) Dose response to hCG treatment. HESC were treated with increasing concentrations of hCG and CXCL10 mRNA expression was determined by qPCR. CXCL10 mRNA expression in HESC decreased in a dose dependent manner.

Figure 5

Regulation of H3K27me3 expression by hCG and cAMP in human endometrial stromal cells. (A) hCG and cAMP induce H3K27me3 in HESC. HESC were treated with hCG (100 IU/mL) or cAMP (500 uM) and H3K27me3 expression was determined by Western blot analysis. Representative western blot analysis showing hCG and cAMP enhanced H3K27me3 expression in HESCs. (B) Quantification of the western blot analysis showed in figure A. Data is presented as Relative Intense Units (RIU) as an average of 3 independent experiments. *p < 0.001. (C) Dose response effect of hCG on H3K27me3 expression. HESC were treated with increasing concentrations of hCG for 24 h followed by the determination of H3K27me3 expression by Western blot. hCG induces a dose dependent increase on H3K27me3 expression. Data is representative of 3 independent experiments done in triplicate. (D) Quantification of the western blot analysis showed in figure C. Data is presented as Relative Intense Units (RIU) as an average of 3 independent experiments.

Figure 6

Regulation of EZH2 expression by hCG in human endometrial stromal cells. (A) Treatment of HESC with the EZH2 inhibitor GSK126 (2μM) decreases H3K27me3 expression. HESC were treated with GSK126 (2μM) for 24 h followed by the evaluation of H3K27me3 expression by Western blot analysis. Note the decrease of H3K27me3 expression in the presence of GSK126. Representative Western blot analysis from 6 independent experiments. (B) Treatment of HESC with the EZH2 inhibitor GSK126 (2μM) increases CXCL10 mRNA expression. HESC were treated with GSK126 (2μM) for 24 h followed by the evaluation of CXCL10 mRNA expression by qPCR. Treatment with GSK126 is associated with significant increase in CXCL10 mRNA expression. *p < 0.002. Data shown as mean + SD and are for 6 independent experiments. (C) Effect of hCG on EZH2 mRNA expression. HESC were treated with hCG (100 IU/mL) for 24 h and CXCL10 mRNA expression was determined by qPCR. hCG treatment enhances EZH2 mRNA expression in HESC. *p < 0.05. Data shown as mean + SD and are for 3 independent experiments.

Figure 7

Identification of a novel site of epigenetic regulation in the CXCL10 promoter in term human decidual tissue. (A) Diagram of the CXCL10 promoter region. Eight specific primers were designed to cover the full length of the CXCL10 promoter region (IP-10 = CXCL10). (B) Chromatin immunoprecipitation was performed on human decidual samples collected from normal term non-labored placentas delivered by cesarean delivery. Decidual samples were crosslinked, sonicated and immunoprecipitated using an anti-H3K27me3 antibody. The DNA enriched with the H3K27me3 was purified and amplified by qPCR for the different regions of the CXCL10 promoter. Immunoprecipitation of H3K27me3 and PCR amplification of the DNA sequence bound by the histone demonstrated a strong band at Region 4 (505–601 base pairs upstream of the CXCL10 gene) in the promoter region of CXCL10 gene. No bands were detected at regions 1–3 or 5–8, indicating that the histone mark was not enriched in those regions. (C) PCR amplification of the DNA from Region 4 of the CXCL10 promoter that was bound by the H3K27me3 after immunoprecipitation using an anti-H3K27me3 antibody. Representative figure of 6 independent decidual samples. (D) Quantification of the enrichment of the DNA sequence in the different regions of the CXCL10 promoter. Note that only Region 4 reveled binding to the H3K27me3 after ChIP qPCR. n = 6 human decidual tissue samples. (E) Relative CXCL10 mRNA expression levels from decidual tissue samples in relation to stromal cells. *p < 0.05. n = 6 human tissue samples.

Figure 8

hCG regulates CXCL10 expression and secretion by increasing H3K27me3 histone enrichment in Region 4 of the CXCL10 promoter. (A) HESC were treated with either vehicle or hCG (100 IU/ml). The cells were crosslinked, sonicated and immunoprecipitated using an anti-H3K27me3 antibody. The DNA enriched with the H3K27me3 was purified and amplified for Region 4 of the CXCL10 promoter. The results were quantified by ChIP qPCR. Note the significant enrichment of Region 4 bound to H3K27me3 in cells treated with hCG. *p < 0.05. Data shown are for 3 independent experiments. (B) Enrichment of the H3K27me3 mark in the CXCL10 promoter from HESC undergoing decidualization. HESC were decidualized with estrogen, progesterone and cAMP for 5 days. Chromatin immunoprecipitation was performed in HESC and decidualized stromal cells using an anti-H3K27me3 antibody. The enrichment was quantified using ChIP qPCR. Note the significant enrichment of Region 4 bound to H3K27me3 in decidualized cells compared to HESC. **p < 0.002. Data shown for 3 independent experiments.

Figure 9

Effect of LPS on CXCL10 expression and secretion. (A) Human term decidua tissues were treated with LPS (100 ng/ml) for 24 followed by chromatin immunoprecipitation using an anti-H3K27me3 antibody. The DNA enriched with the H3K27me3 was purified and amplified by qPCR for Region 4 of the CXCL10 promoter. Note the decreased binding of the H3K27me3 to region 4 of the CXCL10 promoter observed in cells treated with LPS compared to the control (PBS) treated group. Representative figure of 6 independent experiments. (B) Quantification of the enrichment of the DNA sequence of region 4 of the CXCL10 promoter bound to H3K27me3. The enrichment was quantified using ChIP qPCR and reveled decreased enrichment in decidual samples exposed to LPS. p < 0.05. (C) Effect of LPS on CXCL10 mRNA expression. Human term decidua tissue samples were treated with LPS (100 ng/ml) for 24. CXCL10 mRNA expression was determined by qPCR. LPS treatment induces a significant increase on CXCL10 mRNA expression in human term decidua tissues n = 6, *p < 0.05. (D) Human term decidua tissues were treated with LPS (100 ng/ml) for 24 followed by evaluation of CXCL10 protein expression by Luminex. LPS induces CXCL10 protein secretion in decidual samples. n = 6, *p < 0.001.

Figure 10

Role of JMJD3 in the regulation of CXCL10 expression and secretion. (A) Expression of JMJD3 in human decidual tissue. Human term decidual tissues were treated with LPS (100 ng/ml) and JMJD3 expression was determined by qPCR. LPS treatment enhances JMJD3 mRNA expression. n = 6, *p < 0.05. (B) Inhibition of JMJD3 function with the inhibitor GSKJ4 prevents LPS-induced CXCL10 expression. Human term decidual tissues were treated with LPS (100 ng/ml) in the presence or absence of GSKJ4, an inhibitor of JMJD3 for 24 h. CXCL10 mRNA expression was determined by qPCR. The presence of GSKJ4 inhibits LPS induced CXCL10 expression. *p < 0.05. (C) Quantification of the total number of T cells recruited towards conditioned media obtained from HESC and DSC in the presence or absence of LPS treatment (100 ng/ml). Transwell migration assays were performed using peripheral blood mononuclear cells and conditioned media from HESC and DSC treated with LPS or vehicle (control). LPS treatment is associated with increased number of T cells collected in the lower chamber (recruitment chamber). n = 6 independent experiments. Each group was done in triplicates. *p < 0.001. (D) Flow Cytometry Analysis of CD4⁺ and CD8⁺ T cells. Percentage of CD4⁺ and CD8⁺ T cells recruited by DSC CM in relation to HESC CM was determined by flow cytometry. CD4⁺ and CD8⁺ T cells were identified in the Transwell migration assays. Representative flow cytometry from four independent experiments.

Figure 11

hCG and CXCL10 expression ratio throughout early gestation. Circulating levels of hCG and CXCL10 were measured using ELLA assay in serum samples collected from normal pregnant women during the first 4 weeks of pregnancy. (A) hCG is detected in the serum as early as 9 days after embryo transfer and shows a steady increase in the following days. (B) Ratios of CXCL10 and hCG across gestational age were determined in blood samples collected from IVF pregnancies. The line represents the mean estimation by LOWESS and the colored areas the 95% confidence intervals of the mean estimation. CXCL10:hCG ratios decreased in normal pregnancies in relation to gestational age.

Figure 12

Regulation of CXCL10 by hCG and its relevance during pregnancy. (A) hCG, through cAMP enhances the expression of EZH2 increasing H3K27me3 histone methylation which binds to region 4 of CXCL10 promoter region and consequently inhibiting CXCL10 expression. (B) Trophoblast derived hCG inhibits CXL10 expression by decidual cells preventing T cell recruitment.