Characterization of Organoid Cultures to Study the Effects of Pregnancy Hormones on the Epigenome and Transcriptional Output of Mammary Epithelial Cells

Abstract

The use of mouse derived mammary organoids can provide a unique strategy to study mammary gland development across a normal life cycle, as well as offering insights into how malignancies form and progress. Substantial cellular and epigenomic changes are triggered in response to pregnancy hormones, a reaction that engages molecular and cellular changes that transform the mammary epithelial cells into "milk producing machines". Such epigenomic alterations remain stable in post-involution mammary epithelial cells and control the reactivation of gene transcription in response to re-exposure to pregnancy hormones. Thus, a system that tightly controls exposure to pregnancy hormones, epigenomic alterations, and activation of transcription will allow for a better understanding of such molecular switches. Here, we describe the characterization of ex vivo cultures to mimic the response of mammary organoid cultures to pregnancy hormones and to understand gene regulation and epigenomic reprogramming on consecutive hormone exposure. Our findings suggest that this system yields similar epigenetic modifications to those reported in vivo, thus representing a suitable model to closely track epigenomic rearrangement and define unknown players of pregnancy-induced development.

Keywords: Epigenomics; Mammary organoids; Pregnancy-induced development.

Fig. 1

Pregnancy hormones induce robust upregulation of Casein 2 (Csn2) in post-pregnancy mammary organoid cultures. (a) Genome browser tracks showing distribution of H3K27Ac peaks at the Csn2 locus in luminal ductal MECs harvested from pre- and post-pregnancy mice at distinct pregnancy cycles. (b) mRNA levels (qPCR) of casein genes, Csn2 and Csn3, in pre- and post-pregnancy mammary organoids grown with essential medium, complete medium (supplemented with pregnancy hormones) or complete medium with FGF2 n = 2 biological replicates, with 3 technical replicates each. ***p = 0.0002, **p = 0.002, *p = 0.01 (c) Csn2 mRNA levels (2^deltadeltaCT) in pre-pregnancy mammary organoids grown with complete medium. Normalized to pre- pregnancy organoids treated with complete medium for 3 h. (d) Immunofluorescence (IF) imaging of whole mounted pre- and post-pregnancy mammary organoids before or after complete medium exposure for 12 h and 72 h. KRT8 (blue), KRT5 (red) and CSN2 (green). Scale: 100 μm. For analyses, error bars indicate standard error of the mean (SEM) across samples of same experimental group. p values were defined using Welch’s t-test

Fig. 2

Pregnancy hormones drive changes to the active regulatory landscape of pre-pregnancy mammary organoid cultures. (a) RNAseq quantification of parity gene signature in pre-pregnancy organoids treated with complete medium for 9 days. n = 2 biological replicates. (b) GSEA of gene networks differentially expressed in pre-pregnancy organoids treated with complete medium for 9 days. (c) Venn diagrams comparing H3K27ac peaks of pre-pregnant mammary organoid cultures before complete medium treatment (0 h) and after 3 h, 12 h and 48 h treatment with complete medium. (d) Density plot showing H3K27ac levels at computationally defined DNA binding motifs recognized by the transcription factors ESRa, MYC and STAT5 in pre-pregnancy mammary organoids before complete medium treatment (0 h) and after 3 h, 12 h, 48 h treatment with complete medium

Fig. 3

Altered epigenomic landscape drives post-pregnancy mammary organoids response to re-exposure to pregnancy hormones. (a) RNAseq quantification of parity gene signature in post-pregnancy mammary organoids grown with essential media (no pregnancy hormones) for 9 days. n = 2 biological replicates. (B) Principal component analyses of gene expression datasets from pre- and post-pregnancy organoids treated with and without complete medium for 9 days. (c) Venn diagram comparing total H3K27ac peaks of untreated pre-pregnant mammary organoids and post-pregnancy mammary organoids. (d) Genome browser tracks showing distribution of H3K27ac peaks in untreated pre- and post-pregnancy mammary organoid cultures for the Elf5 and Prlr loci. (e) S-plot showing untreated post-pregnancy organoid exclusive H3K27ac peaks intersected with gene expression from untreated post-pregnancy organoids day 9. (f) Venn diagrams comparing H3K27ac peaks from post-pregnant mammary organoid cultures before complete medium treatment (0 h) and after 3 h, 12 h and 48 h of treatment with complete medium (g) GSEA of gene networks exclusive to post-pregnancy mammary organoids treated with complete medium for 3 h, 12 h and 48 h

Fig. 4

Utilization of organoid cultures to define players in pregnancy-induced development and post-pregnancy epigenome. (a) Representative brightfield images of mammary organoids treated with essential medium or complete medium, supplemented with either DMSO (Dimethyl sulfoxide, control), or EZH2 inhibitor UNC1999 for 48 h. Arrows indicated examples of branching organoids. Scale = 200 μm. (b) Branching quantification of pre- and post-pregnancy mammary organoid cultures treated with essential medium or complete medium, supplemented with either DMSO (control) or EZH2 inhibitor UNC1999. 13 fields of view per well/replicate. n = 25 organoids. ns = not significant; *p = 0.027 differences between post-pregnancy organoids treated with complete media and complete media with UNC1999. Error bars indicate standard error of the mean (SEM) across samples of same experimental group. p values were defined using Students t-test. (c) Size quantification of pre- and post-pregnancy mammary organoid cultures treated with essential medium or complete medium, supplemented with either DMSO (control) or EZH2 inhibitor UNC1999. n = 20 organoids per condition. ns = not significant; *p = 0.018 differences between post-pregnancy organoids treated with complete medium and DMSO and post-pregnancy organoids treated with complete medium and UNC1999. For analyses, error bars indicate standard error of the mean (SEM) across samples of same experimental group. p values were defined using Welch’s t-test. (d) Csn2 mRNA levels (qPCR) in pre- and post-pregnancy mammary organoid cultures treated with either essential medium or complete medium, with DMSO control or UNC1999. *p = 0.0364 differences between post-pregnancy organoids treated with complete medium and DMSO and post pregnancy organoids treated with complete medium and UNC1999. ***p = 0.0009 differences between pre-pregnancy organoids treated with complete medium and UNC1999, and post-pregnancy organoids treated with complete medium and UNC1999. Error bars indicate standard error of the mean (SEM) across samples of same experimental group. p values were defined using Welch’s t-test