Spatial-temporal regulation of the prostanoid receptor EP2 co-ordinates PGE2-mediated cAMP signaling in decidualizing human endometrium

Abstract

Decidualization denotes the differentiation of endometrial stromal cells into specialized decidual cells, essential for embryo implantation and pregnancy. The process requires coordination of progesterone and cAMP signaling, which converge on downstream transcription factors. PGE2 and relaxin, acting, respectively, through Gαs-coupled GPCRs EP2 and RXFP1, are putative candidates for generating cAMP in differentiating stromal cells. Here, we show that PGE2 is less efficacious than relaxin in elevating intracellular cAMP levels in primary stromal cells but more effective at driving the expression of decidual genes. PGE2-and relaxin-induced cAMP generation involves receptor internalization, but EP2 is endocytosed into very early endosomes (VEEs). Perturbation of VEE machinery through depletion of key trafficking proteins; APPL1 and GIPC, dysregulates PGE2-dependent cAMP profiles and disrupts key decidual signaling pathways, resulting in a disordered differentiation response. We demonstrate that regulation of EP2 via internalization is essential for coordinated activation of the downstream signaling cascades that govern decidualization.

Keywords: Endocrinology; Health sciences; Internal medicine; Medical specialty; Medicine; Reproductive medicine.

Graphical abstract

Figure 1

PGE2 and relaxin induce distinct cAMP profiles (A) Schematic depiction of experimental procedures to establish primary EnSC cultures from endometrial biopsies. (B) Concentration-dependent cAMP induction in EnSCs generated by a 5-min stimulation with either PGE2 or relaxin. Data are mean ± SD, n = 3. (C) Temporal cAMP profiles generated by either 1μM PGE2 or 1μM relaxin, or in combination. Plots from individual patients are represented by dashed lines with bold lines indicating mean values, n = 3. (D) Schematic depiction of PGE2 (left panels) and relaxin (right panels) receptor subtypes and their associated signaling cascades. (E) Expression of PTGER2, PTGER4, and RXFP1 transcripts in EnSC cultures (GEO: GSE246591). Data points from individual patients are color-matched and shown together with mean TPM values ±SD. Different letters indicate statistical difference (p < 0.05) between groups (ANOVA and Tukey’s multiple comparison test), n = 4. (F) UMAPs depicting receptor expression in stromal cells from single-cell RNA sequencing of whole endometrial biopsies (GEO: GSE247962). Data are pooled from 12 individual patients. Positive cells and the level of receptor expression are shown by color key as indicated. (G) The relative change in cAMP to 5-min stimulation with PGE2 in EnSCs following siRNA depletion of PTGER2 and PTGER4. (H) The relative change in cAMP to 5-min stimulation with relaxin in EnSCs following siRNA depletion of RXFP1. For G and H, data points from individual patients are color-matched and shown together with bar graphs denoting mean. Different letters indicate statistical difference (p < 0.05) from NT siRNA, stimulated cells (ANOVA and Dunnett’s multiple comparison test), n = 3.

Figure 2

PGE2, and not relaxin, drives decidualization (A) RT-qPCR analysis of the relative changes in transcripts for decidual genes following 2 and 4-day treatment with PGE2 or relaxin in combination with MPA. (B) Changes in levels of transcripts for PTGER2 (left panel) and RXFP1 (right panel) in EnSCs treated with 8-bromo-cAMP and MPA from 3 h up to 4 days (GEO: GSE246591), n = 4. (C) Extended temporal cAMP profiles in EnSCs stimulated with either PGE2 or relaxin for up to 3 h, n = 3. (D) Schematic representation of the experimental procedures used for the resensitization assay (left panel). The magnitude of secondary cAMP induction in EnSCs following a pre-stimulation, desensitizing challenges with PGE2 or relaxin for variable times (right panel). Data are normalized as a percentage of control responses (no pre-stimulation). Plots from individual patients are normalized to the magnitude of cAMP signal in control (no pre-stimulation) cells, n = 3. For A–D, individual cultures are represented by dashed lines with bold lines indicating mean values. (E) Induction of cAMP by PGE2 (left panel) or relaxin (right panel) following inhibition of receptor internalization with the dynamin-inhibitor, Dyngo-4A. Data points from individual patients are color-matched and shown together with bar graphs denoting mean. Different letters indicate statistical difference (p < 0.05) from untreated, unstimulated cells (ANOVA and Dunnett’s multiple comparison test), n = 3.

Figure 3

Internalized EP2 receptors traffic through very early endosomes (A) Schematic representation of endosomal compartments and their cargo. (B) Representative images from FLAG-βAR and FLAG-EP2 as captured by TIRF microscopy, 5 min post isoproterenol and PGE2 stimulation, respectively. Arrows indicate visible endosomal lumen in FLAG-βAR positive endosomes. Quantitation of endosome diameter (right panel). 182 endosomes (FLAG-EP2) and 264 endosomes (FLAG-βAR) were measured from 4 independent cultures. ∗∗∗∗ denotes a p value of <0.0001 via Student’s t test. (C) FLAG-EP2 (green) and APPL1 (red) positive endosomes with co-localization detected as orange staining, 5 min post PGE2 stimulation (left panels) and quantitation of overlay against the average of all cells (right panel). Arrows indicate FLAG-EP2 and APPL1 positive endosomes. Data are obtained from 30 representative regions chosen at random, n = 3. For B and C, scale bars = 20μm, inset = 10μm.

Figure 4

PGE2, but not relaxin, mediated cAMP signaling are regulated by APPL1 and GIPC (A) RT-qPCR analysis of transcripts for APPL1 (left panel) and GIPC (right panel) following their depletion in EnSCs by siRNA. Data from individual patients are color-matched and shown with bar graphs denoting mean values. Differing letters indicate significance from NT siRNA controls (p < 0.05) (ANOVA and Dunnett’s multiple comparison test, n = 3). (B) Induction of cAMP after 5-min stimulation with PGE2 (left panel) and relaxin (right panel) following depletion of APPL1 and GIPC by siRNA. Data from individual patients are color-matched and shown with bar graphs denoting mean values. Differing letters indicate significance from NT siRNA, unstimulated controls (p < 0.05) (ANOVA and Dunnett’s multiple comparison test, n = 6). (C) Schematic representation of experimental procedures used to assess resensitization of EP2 receptors depleted of APPL1 and GIPC (left panel). The cAMP signal from a 5-min PGE2 challenge (2^nd response) following an identical desensitization challenge (1^st response) or control, and ligand washout (right panel). Data from individual patients are color-matched and shown with bar graphs denoting mean values. Differing letters indicate significance from NT siRNA, unstimulated controls (p < 0.05) (ANOVA and Dunnett’s multiple comparison test, n = 3).

Figure 5

Loss of APPL1 and GIPC inhibits decidualization (A) Schematic depiction of treatment protocol for RNA sequencing in EnSCs. (B) Relative changes in absorbance from an XTT assay to assess cell viability in untreated EnSCs depleted of APPL1 and GIPC for 4 days. Individual patients are shown as dashed lines, with mean values depicted by the bold lines, n = 3. (C) RT-qPCR analysis of APPL1 (left panel) and GIPC (right panel) 4 days following siRNA depletion. Data from individual patients are color-matched with bar-graphs denoted mean values. Differing letters indicate significant between groups (p < 0.05) (ANOVA and Tukey’s multiple comparison test, n = 3). (D) Heatmap of differentially expressed genes (Bonferroni correction, p < 0.05) identified from bulk RNA-sequencing between untreated (Day 0) NT siRNA cells and those treated with PGE2/MPA for 4 days. A total of 645 differentially expressed genes were identified (435 upregulated and 210 downregulated) with each gene scaled (z-score) across treatments to show changes with APPL1 and GIPC depletion. Red, blue and white colors represent high, medium and low expression, respectively, as per key. Data are n = 3. (E) Venn diagram depicting the number of differentially expressed genes (Bonferroni correction, p < 0.05) identified when comparing APPL1 and GIPC depleted EnSCs to their treatment-matched (untreated; day 0 or PGE2/MPA treated; day 4) NT siRNA cells. Circles are relative in size to the number of genes. (F) Selected Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment of differentially expressed genes (full lists of KEGG enrichment terms are available in Tables S2-5). The size of circles is relative to the number of genes in each enrichment term, and the color represents p value calculated as a result of enrichment degree. (G) RT-qPCR analysis showing relative changes in decidual genes IGFBP1 (left panel) and PRL (right panel) following 4-day treatment with 8-bromo-cAMP and MPA (C + M) in EnSCs depleted of APPL1 or GIPC. Data from individual patients are color-matched with bar-graphs denoted mean values. Differing letters indicate significant compared to untreated (p < 0.05) (ANOVA and Dunnett’s multiple comparison test), n = 3. (H) Summary schematic detailing the rapid recycling and resensitization of EP2 receptors through the VEEs and its contribution to decidualization of EnSCs.