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

Primary mammary organoids were derived from either pre-pregnant or post-pregnant female Balb/c mice as previously described [2]. In short, pipettes and tubes were pre-coated with 5% BSA solution (in 1X PBS, Gibco #A10008–01). Female mice (~15 weeks old) were euthanized via CO₂ asphyxiation, and mammary glands from 3 animals (thoracic and inguinal mammary glands pairs) were removed and collected into 10 cm culture dishes. The pooled glands were then minced under sterile conditions using scalpels, cutting around 50 times in a crisscross pattern to loosen the tissue of the glands. The minced glands were then transferred to a 50 mL falcon tube that contained 20 mL of collagenase solution which consisted of AdDf+++ (Advanced DMEM F12 (Dulbecco’s Modified Eagle Medium/F-1,5 mM GlutaMax 5 mM HEPES, 1x Penicillin/Streptomycin), FBS (1%, Corning #35–010-CV), Insulin (5μg/mL, Sigma #I9278) and Collagenase A (2 mg/mL, type IV from Clostridium histolyticum, Sigma #C5138). The tubes with the glands and collagenase solution were shaken at 200 RPM for 30–40 min until solution became cloudy, and large chunks of tissue dissipated. After digestion, 1 mL of FBS was added, and then solution was passed up and down a 5 mL pipette 10 times to ensure complete disassociation of the tissue. The solution was spun down in a centrifuge at 300 x g for 5 min at room temperature and the fat-containing supernatant was removed. The pellet was resuspended in 10 mL of AdDf+++ and then passed through a 100 μm strainer (Falcon #352360) to ensure no large tissue pieces would proceed. An additional room temperature 5-min centrifugation step at 300 x g was used to wash any remaining enzyme from the pelleted epithelium. The pellet was then resuspended in 10 mL of AdDf+++, pulse centrifugated to 500 x g, then had supernatant removed. This was repeated a total of 3 times.

The organoid pellet was resuspended using a pre-chilled pipette tip in desired amount of Matrigel (100%, Corning #354230). In a 37 °C pre-warmed 24 well plate, three small domes were made using a total of 50 μl of Matrigel in a triangle-shaped pattern. The plates were then inverted and placed into a CO₂ incubator (5% CO₂, 37 °C) for 20 min to allow Matrigel to solidify. Each of the wells was filled with 0.5 mL of Essential organoid medium AdDf+++, supplemented with 1x ITS (Insulin/Transferrin/Sodium Selenite, Gibco #41400–045) and FGF-2 (Final concentration: 5 nm, PeproTech #450–33) for 6 days. Medium was changed every two days. The cultured mammary organoids were then grown in medium lacking FGF2 for 24 h and then incubated with complete medium (AdDf+++, supplemented with ITS (Final Concentration:1x, Insulin/Transferrin/Sodium Selenite, Gibco, #41400–045), 17-β-Estradiol (Final concentration: 40 ng/mL, Sigma #E2758), Progesterone (Final concentration: 120 ng/mL, Sigma #P8783), Prolactin (Final concentration: 120 ng/mL,Sigma #L4021). Organoids were isolated from Matrigel using Cell Recovery solution (0.5 mL, Corning #354253).

Organoid visualization and image collection were performed on a Nikon Eclipse TI microscope utilizing NIS-Elements BR software (Nikon). For branching and size quantification, at least 13 fields of view and 20 to 25 organoids were analyzed. The area (size) of mammary organoids in each image was measured via ImageJ. A branching classification was given to organoids displaying three or more elongated buds.

Medium was removed from each of the wells and then washed with 0.5 mL 1x PBS. RNA was extracted by adding Trizol (0.5 mL, Thermo Fisher Scientific, #15596018) to each well with organoids. Reverse transcription was carried out using SuperScript III ™ kit (Thermo Fisher Scientific). RTqPCR was performed using a Quantstudio 6 with SYBR Green Master mix (Applied Biosystems #4368577). Each reaction was run in at least duplicate. Relative gene expression was calculated via the deltadeltact method in which the values for the measured genes were normalized to the house keeping gene, mouse β-actin mRNA. Primers sequences targeting mRNA of β-actin, Csn2 and Csn3 genes were designed as previously described [1].

Each well was washed with 0.5 mL of 1X PBS, followed by the addition of Cell recovery solution (0.5 mL, Corning #354253). Culture plates were them incubated at 4 °C for 30 min or until Matrigel domes were no longer visible, followed by the addition of 1 mL of AdDf+++ and gentle pipetting all organoids were released from Matrigel. Organoids were then transferred to a 1% BSA (in 1X PBS, Gibco #A10008–01) pre-coated 15 mL conical tube, and spun at 500 x g for 5 min at room temperature (RT). Organoid pellets were gently resuspended in 1 mL of 4% PFA (in 1X PBS, Electron Microscopy Sciences, #15711) for 1 h at RT, following centrifugation at 500 x g for 5 min at RT. Fixed organoids were permeabilized with 1 mL of Permeabilization solution (0.5% Triton X-100, Sigma #93443 in 1X PBS) and incubated for 30 min at RT, followed by centrifugation at 500 x g for 5 min at RT. Permeabilized organoids were washed once with 1 mL of washing solution (0.1% Tween-20 solution, MP, #9005-64-5 in 1X PBS), followed by centrifugation at 500 x g for 5 min at RT. Fixed and permeabilized organoid pelleted were gently resuspended with 0.5 mL of blocking buffer (1X PBS; 300 mM Glycine, Fischer-Scientific, #G45–212; 10 mg/mL BSA, Sigma #A2153; 5% Goat serum, Abcam #ab7481) and incubated for 1 h at RT, followed by centrifugation at 500 x g for 5 min at RT. Blocked organoids were then resuspended in blocking buffer containing indicated concentrations of antibodies, and incubated overnight at 4 °C with constant agitation. Staining solution was removed, and organoids were washed with 1 mL of washing solution, centrifugation at 500 x g for 5 min at RT for a total of 3 times. Stained organoids were then stained with Propidium Iodide (PI, for nuclear staining, Invitrogen, 1:1000 dilution), at room temperature for 15 min. Staining solution was removed, and organoids were washed with 1 mL of washing solution and then underwent centrifugation at 500 x g for 5 min at RT for a total of 3 times. At the end of the last centrifugation, the supernatant was removed, and one drop of Prolong™ Glass Antifade Mountant (Invitrogen #P36982) was added to stained organoids, followed by mounting on a glass slide and allowed to cure overnight in the dark. All imaging was acquired on a Zeiss 780 Confocal Microscope.

Mammary organoids (n = 2–3 wells per technical replicate) were collected, dissociated and resuspended in Trizol (Thermo Fisher Scientific, #10296010) for RNA extraction. Double stranded cDNA synthesis and Illumina libraries were prepared utilizing the Ovation RNA-seq system (V2) (Nugen Technologies, #7102–32). RNA-seq libraries were prepared utilizing the Ovation ultralow DR multiplex system (Nugen Technologies, #0331–32). Each library (n = 2 per experimental condition) was barcoded with Illumina True-seq adaptors to allow sample multiplexing, followed by sequencing on an Illumina NextSeq500, 76 bp single-end run. Bioinformatics analyses were performed with command-line interfaced tools such as FastQC [66] for quality control and Trimmomatic [67] for sequence trimming. We used STAR [68] for mapping reads and deepTools [69] for principal component analysis. Further, we utilized FeatureCounts [70] for assigning reads to genomic features and DESeq [71] to assess changes in expression levels. Gene Set Enrichment Analysis (GSEA) was used for global analyses of differentially expressed genes [72]. For the parity signature analysis (total of 47 genes [18]), differentially expressed genes above 0.5 log2foldchange were counted as upregulated, while genes below −0.5 log2foldchange were counted as downregulated.

Mammary organoids (n = 2–3 wells per technical replicate) were collected, dissociated, and permeabilized with digitonin, following overnight incubation with antibody against H3K27ac histone marks (Abcam, #ab4729) at 4°C with constant agitation. Antibody-chromatin complexes were fragmented with pA-MNAse and purified utilizing Phenol-Chloroform. Cut&Run libraries (n = 2 per experimental condition) were amplified and barcoded using Clontech DNA Smart ChIP-Seq kit (Clontech, #634866) in accordance with the manufacturer’s instructions, then pooled for sequencing on an Illumina NextSeq500, 76 bp paired-end run. Reads were mapped to the indexed mm9 genome using bowtie2 short-read aligner tool [73] using default settings. Sparse Enrichment Analysis for Cut&Run (SEACR) peak-calling program [74] was used to identify enriched genomic regions with an empirical threshold of n = 0.01, returning the top n fraction of peaks based on total signal within peaks. The stringent argument was implemented, which used the summit of each curve. Further downstream analyses were performed using various command-line interfaced programs including deepTools for investigating H3K27ac peak intensity at DNA binding motifs and bedtools [75] for defining regions that are shared between conditions as well as peaks exclusive to each condition. Open source software such as Enrichr [76, 77] for comparing peaks against publicly available data and GREAT [78] for gene ontology analyses served to support condition-exclusive analyses. Additionally, UCSC’s Genome Browser [79] was used to investigate region specific H3K27ac peak intensity. H3K27ac Cut&Run peaks were utilized as input for an unbiased transcription factor analyses using Analysis of Motif Enrichment (AME) [80] and Find Individual Motif Occurrences (FIMO) [81] was used to computationally define DNA binding motif regions. DESeq2 [82] was utilized to generate genomic regions with differential H3K27ac levels sample groups (FDR < 0.05).