Development of Polarity-Reversed Endometrial Epithelial Organoids

Abstract

The uterine epithelium is composed of a single layer of hormone responsive polarized epithelial cells that line the lumen and form tubular glands. Endometrial epithelial organoids (EEO) can be generated from uterine epithelia and recapitulate cell composition and hormone responses in vitro. As such, the development of EEO represents a major advance for facilitating mechanistic studies in vitro. However, a major limitation for the use of EEO cultured in basement membrane extract and other hydrogels is the inner location of apical membrane, thereby hindering direct access to the apical surface of the epithelium to study interactions with the embryo or infectious agents such as viruses and bacteria. Here, a straightforward strategy was developed that successfully reverses the polarity of EEO. The result is an apical-out organoid that preserves a distinct apical-basolateral orientation and remains responsive to ovarian steroid hormones. Our investigations highlight the utility of polarity-reversed EEO to study interactions with E. coli and blastocysts. This method of generating apical-out EEO lays the foundation for developing new in vitro functional assays, particularly regarding epithelial interactions with embryos during pregnancy or other luminal constituents in a pathological or diseased state.

Figure 1

Generation and characterization of apical-out organoids. (A) Schematic of approach to generate apical-out endometrial epithelial organoids (AO-EEO or AO); (B) Characterization of AO. (i–ii) cartoons of epithelial cells in apical-in organoids (AI) exhibiting apical surface facing the lumen or in AO exhibiting apical surface exposed to the outer environment; (iii–iv) bright field images of organoids grown in Cultrex BME or in suspension culture for eight days showing gross morphology; scale bars, 100 µm; (v–vi) Immunofluorescence staining of apical surface marker zonula occludens 1 (ZO1) on paraffin sections of AI or AO; scale bars, 20 µm and 60 µm for v and vi respectively. Sections were counterstained with Hoechst to detect DNA; Areas outlined in red boxes in v and vi magnified to demonstrate apical and basal surfaces, separated by vertical line; (vii–viii) scanning electron microscopy (SEM) images of AI or AO showing surface microarchitecture; scale bars, 50 µm. (C) Time-dependent screening of organoids flipping from day 0 to day 8 detected by ZO1 immunofluorescence (i–iv and vi–ix), categorized as apical-in, apical-out, or mixed polarity and quantified in (v, x); scale bars, 50 µm (i, vi) and 150 µm (ii–iv and vii–ix). Schematic in A created with Biorender.com.

Figure 2

Apical-out organoids show similar response to uterine hormones as apical-in organoids. (A) Schematic showing the experimental design to treat AI or AO organoids with steroid hormones. (B) Representative images of immunofluorescence staining of PGR (red) and ZO1 (green) on paraffin sections of AI or AO treated with vehicle or E2 for 2 days and/or E2 + MPA for 6 days. Organoids were counterstained with Hoechst (blue) (i–vi and vi–ix); scale bars, 20 µm (i–vi) and 40 µm (vi–ix). Quantification of PGR-positive cells in AI and AO (v and x); (C) Representative images of immunofluorescence staining of ENPP3 (red) and ZO1 (green) on paraffin sections of AI or AO organoids treated with vehicle or E2 for 2 days and then with E2 + MPA for 6 days. Organoids were counterstained with Hoechst (blue). All scale bars in C are 40 µm (i–ii and iv–v). Quantification of ENPP3 (red) signal intensity (iii, vi). Bar graphs in B and C represent mean ± s.d. Solid dots represent the average organoid per biological replicate. **P < 0.01, ***P < 0.001, ****P < 0.0001 (Student’s t test). (D) Representative TEM images of AI or AO organoids treated with vehicle, E2 or E2 + MPA. Areas outlined in boxes are magnified in panels below to demonstrate the distribution of cilia (red filled triangles), microvilli (black filled triangles), tight junctions (white arrow), and secretory vacuoles (red asterisk) on the apical surface. Letter ‘L’ indicates the lumen of organoids. Scale bars in D are indicated within each image. Schematic in A created with Biorender.com.

Figure 3

Apical-out organoids show susceptibility to E. coli infection. (A) Schematic showing the experimental design to infect AI or AO with E. coli^GFP. (B) Bright field images from live epifluorescence imaging of AI or AO organoids infected with GFP-labeled E. coli^GFP for 0 hpi, 48 hpi, and 96 hpi; All scale bars are 100 µm. (C) Quantification of GFP fluorescence intensity. (D) Representative images of immunofluorescence staining of ZO1 (red) and GFP (green) using antibodies against GFP from AI or AO infected with GFP-labeled E. coli for 96 h. Organoids were counterstained with Hoechst to detect DNA; All scale bars, 40 µm. (E) Quantification of GFP fluorescence intensity. Plots in C and E represent mean ± s.d. Solid dots represent the average GFP intensity per biological replicate. **P < 0.01, ***P <0.001, ****P < 0.0001 (Student’s t-test). Schematic in A created with Biorender.com.

Figure 4

Apical-out organoids attach with blastocysts. (A) Schematic showing experimental design to coculture of mouse blastocysts and organoids. (B) Bright field, GFP epifluorescence and merged images of AO treated with vehicle or E2 for 2 days and E2 + MPA for 6 days, and then cocultured with mouse GFP labeled blastocysts for 24 h. scale bars, 100 µm. (C) Quantification of mouse blastocysts attached with vehicle or hormone treated AO organoids from three independent trials.