Generation of Mosaic Mammary Organoids by Differential Trypsinization

Summary

The mammary gland is a bilayered structure, comprising outer myoepithelial and inner luminal epithelial cells. Presented is a protocol to prepare organoids using differential trypsinization. This efficient method allows researchers to separately manipulate these two cell types to explore questions concerning their roles in mammary gland form and function.

Abstract

Organoids offer self-organizing, three-dimensional tissue structures that recapitulate physiological processes in the convenience of a dish. The murine mammary gland is composed of two distinct epithelial cell compartments, serving different functions: the outer, contractile myoepithelial compartment and the inner, secretory luminal compartment. Here, we describe a method by which the cells comprising these compartments are isolated and then combined to investigate their individual lineage contributions to mammary gland morphogenesis and differentiation. The method is simple and efficient and does not require sophisticated separation technologies such as fluorescence activated cell sorting. Instead, we harvest and enzymatically digest the tissue, seed the epithelium on adherent tissue culture dishes, and then use differential trypsinization to separate myoepithelial from luminal cells with ~90% purity. The cells are then plated in an extracellular matrix where they organize into bilayered, three-dimensional (3D) organoids that can be differentiated to produce milk after 10 days in culture. To test the effects of genetic mutations, cells can be harvested from wild type or genetically engineered mouse models, or they can be genetically manipulated prior to 3D culture. This technique can be used to generate mosaic organoids that allow investigation of gene function specifically in the luminal or myoepithelial compartment.

Introduction

The mammary gland (MG) is a tree-like, tubular epithelial structure embedded within an adipocyte rich stroma. The bilayered ductal epithelium comprises an outer, basal layer of contractile, myoepithelial cells (MyoECs) and an inner layer of luminal, secretory epithelial cells (LECs), encircling a central lumen¹. During lactation when the outer MyoECs contract to squeeze milk from the inner alveolar LECs, the MG undergoes numerous changes that are under the control of growth factors (e.g., EGF and FGF) and hormones (e.g. progesterone, insulin, and prolactin). These changes cause the differentiation of specialized structures, alveoli, which synthesize and secrete milk during lactation¹. The mammary epithelia can be experimentally manipulated using techniques in which either epithelial tissue fragments, cells, or even a single basal cell are transplanted into host mammary fat pads, precleared of endogenous mammary parenchyma, and allowed to grow out to reconstitute an entire, functional epithelial tree^2,3,4,5. Transplantation is a powerful technique, but it is time-consuming and impossible if a mutation results in early embryonic lethality (prior to E14) that prevents the rescue of transplantable mammary anlage. Furthermore, investigators frequently wish to research the roles of the two different compartments, which are derived from lineage-restricted progenitor cells. While Cre-lox technology allows differential genetic manipulation of MyoECs and LECs, this is also a time-consuming and expensive undertaking. Thus, since the 1950s, investigators have used in vitro mammary organoids as a relatively easy and efficient way to address questions concerning mammary tissue structure and function^6,7.

In early protocols describing the isolation and culture of primary mammary epithelial cells, investigators found that a basement membrane matrix (BME), composed of a plasma clot and chicken embryo extract, was required for MG fragments grown on a dish⁶. In the following decades, extracellular matrices (ECMs, collagen, and jellylike protein matrix secreted by Engelbreth-Holm-Swarm murine sarcoma cells) were developed to facilitate 3D culture and better mimic the in vivo environment^7,8,9,10. Culturing cells in 3D matrices revealed by multiple criteria (morphology, gene expression, and hormone responsiveness) that such a microenvironment better models in vivo physiological processes^9,10,11,12. Research using primary murine cells identified key growth factors and morphogens necessary for the extended maintenance and differentiation of organoids¹³. These studies have set the stage for the protocol presented here, and for the culture of human breast cells as 3D organoids, which is now a modern clinical tool, allowing for drug discovery and drug testing on patient samples¹⁴. Overall, organoid culturing highlights the self-organization capacities of primary cells and their contributions to morphogenesis and differentiation.

Presented here is a protocol to culture murine epithelia that can be differentiated into milk-producing acini. A differential trypsinization technique is used to isolate the MyoECs and LECs that comprise the two distinct MG cell compartments. These separated cell fractions can then be genetically manipulated to overexpress or knockdown gene function. Because lineage-intrinsic, self-organization is an innate property of mammary epithelial cells^15,16,17, recombining these cell fractions allows researchers to generate bilayered, mosaic organoids. We begin by enzymatically digesting the adipose tissue, and then incubating the mammary fragments on a tissue culture dish for 24 h (Figure 1). The tissue fragments settle on polystyrene dishes as bilayered fragments with their in vivo organization: outer myoepithelial layer surrounding inner luminal layers. This cellular organization allows for the isolation of the outer MyoECs by trypsin-EDTA (0.05%) treatment for 3-6 min followed by a second round of trypsin-EDTA (0.05%) treatment that detaches the remaining inner LECs (Figure 2). Thus, these cell types with different trypsin sensitivity are isolated and can subsequently be mixed and plated in ECM (Figure 3). The cells undergo self-organization to form bilayered spheres, comprising an outer layer of MyoECs surrounding inner LECs. Lumen formation occurs as the cells grow in a medium containing a cocktail of growth factors (see recipes for Growth Medium)¹³. After 5 days, organoids can be differentiated into milk-producing acini by switching to Alveologenesis Medium (see recipes and Figure 3F) and incubated for another 5 days. Alternatively, organoids will continue to expand and branch in Growth Medium for at least 10 days. Organoids can be analyzed using immunofluorescence (Figure 3D-F) or released from the ECM using a recovery solution (see Table of Materials) and analyzed via other methods (e.g., immunoblot, RT-qPCR).

Protocol

All methods described here have been approved by the Institutional Animal Care and Use Committee (IACUC) of the University of California, Santa Cruz.

1. Day 1: Mammary gland digestion

1. Prepare to harvest the MGs from mature female mice 10-14 weeks of age.
   1. Perform the harvesting on an open bench under aseptic conditions.
   2. Sterilize all surgical supplies, cork boards, and pins by autoclaving and soaking in 70% alcohol for 20 min prior to surgery.
   3. Anesthetize animals with sodium pentobarbital (2X anesthetic dose of 0.06 mg/g body weight) delivered via intraperitoneal injection with a 0.5 mL insulin syringe.
   4. Monitor the level of anesthesia by pinching the animal's toes to check for a reflex response and commence the protocol only after the animal is fully anesthetized.
   5. Place the animal on its back, pin its appendages to the corkboard, and wipe down its abdomen and chest with ethanol.
2. To harvest the #2, 3, 4, and 5 MGs from one mouse (i.e., 8 MGs, Figure 1A), identify the midline between the two hind legs and make a small incision (1 cm) on the abdominal skin with sharp scissors, then extend the cut up to the neck¹⁸.
3. Follow by making small cuts laterally towards the legs and arms to allow for the release of the skin using a cotton swab. Pull the skin away and stretch it tight before pinning it down on one side (Figure 1B, step 1)¹⁸. Remove the MGs by cutting under them, and remove the lymph nodes from the #4 glands (Figure 1B, steps 2, 3)¹⁸. Repeat the procedure on the other side of the body.
4. Collect the MG tissue in 50 mL of 4 °C Dulbecco's Modified Eagle's Media (DMEM)/Nutrient Mixture F12 (F12) supplemented with 5% fetal bovine serum (FBS) and 1X Antibiotic-Antimycotic (Anti-Anti)¹⁸.
5. Chop the glands in a 35 mm dish or on a ceramic plate using a razor blade or tissue chopper. Rotate the plate every five manual chops or every round on the tissue chopper until the tissue pieces can fit through a 1 mL micropipette tip with ease (~0.1 mm/fragment, Figure 1C).
6. Digest the MGs in Digestion Medium (see Table 1) for 14 h in a 6 well low adhesion dish at 37 °C, 5% CO₂.

2. Day 2: Isolation of mammary epithelial tissue fragments

1. After 14 h of digestion, gently mix by pipetting the digested tissue 10X using a 1 mL micropipette to break down any remaining stroma or adipose tissue, ensuring that neither bubbles nor excess mechanical force are generated.
   NOTE: If there is incomplete digestion after 14 h, this could be due to the accumulation of sheared DNA. In this case, add 1 µL of 1 mg/mL deoxyribonuclease I (DNase I) per 2 mL of Digestion Medium. Incubate for another 30 min at 37 °C, 5% CO₂.
2. Collect tissue in a 15 mL tube and rinse the well used for digestion with 2-3 mL of tissue culture grade 1X Dulbecco's Phosphate Buffered Saline (DPBS) free of Ca²⁺ and Mg²⁺. Centrifuge at 600 x g for 10 min.
3. Evacuate the supernatant containing the lipid layer and medium, and then resuspend the pellet in 5 mL of DPBS and switch to a new 15 mL tube. Centrifuge at 600 x g for 10 min.
4. During centrifugation place a 70 µm nylon cell strainer in a 50 mL tube and prewet the strainer using 10 mL of 37 °C DMEM/F12.
5. Resuspend tissue fragments from protocol step 2.4 using 5 mL of DPBS and pass the suspension through a prewet 70 µm nylon cell strainer to remove stromal cells and single cells (Figure 1D).
6. Collect the tissue fragments on the cell strainer. Rinse 4X with 10 mL of 37 °C DMEM/F12 (Figure 1D).
   CAUTION: Incomplete rinsing will result in cultures contaminated with non-epithelial cells.
7. Release the tissue fragments by holding the strainer tab with gloved fingers, inverting the strainer over a 60 mm tissue culture dish and passing 1 mL aliquots of Maintenance Medium (see Table 1) through the bottom of the strainer 4X (Figure 1E).
8. Check the strainer for tissue fragment remnants, which will be visible by the naked eye, and rinse the strainer 1X more with 1 mL of Maintenance Medium if any fragments are still adhering to the strainer. The rinsed tissue fragments should now be free of stromal cells.
9. Quickly examine the 60 mm dish containing the MG fragments from protocol step 2.9 under an inverted microscope (4X or 10X objective, Figure 1F). A typical preparation of eight MGs yields ~500 fragments. Look for single cells or fat droplets and whether there are contaminating cells.
   NOTE: If there are contaminating cells, repeat the filtration step by collecting the medium and fragments from the 60 mm dish using a 5 mL pipette and filtering the fragment again through a fresh 70 µm strainer, repeating protocol steps 2.6, 2.8-2.10.
10. Incubate 24 h at 37 °C, 5% CO₂, allowing the tissue fragment to adhere and generate bilayered fragments (Figure 2A).
    NOTE:If the fragments have not settled by 24 h, continue to incubate until adhered. If the fragments do not adhere well, the separation of the cell compartments will not work. If researchers are concerned about adhesion, the tissue culture plates can be treated to promote fragment attachment (e.g., poly-L-lysine).

3. Day 3: Differential trypsinization of myoepithelial and luminal epithelial cells

1. To separate MyoECs from LECs, empty the media from the dish, rinse 1x with 1 mL of DPBS and add 1 mL of fresh trypsin-EDTA (0.05%), and carefully monitor the digestion under an inverted microscope (10X or 20X objective, Figure 2B,C,F). The detachment of the outer MyoEC layer will require 3-6 min, depending on trypsin-EDTA (0.05%) strength.
   NOTE: Please see Figure 2 for representative images showing this process. Under brightfield illumination, the MyoECs appear rounded and have a brighter appearance in comparison to the LECs, which remain adhered in the center and appear darker.
2. Collect the MyoEC fraction in a 15 mL tube containing 2 mL 10% FBS/DPBS. Without disturbing the LECs, gently rinse the 60 mm dish with 2 mL of DPBS and then dispose of the DPBS (Figure 2H-I).
   NOTE: The usual recovery for MyoECs is within a range of (3.5 x 10⁶-1.5 x 10⁶) depending on the size of the MGs.
3. To remove the LEC fraction, add 1 mL of trypsin-EDTA (0.05%) to the dish again and incubate 7-15 min, monitoring carefully to prevent overdigestion. Quench the trypsin-EDTA (0.05%) on the dish with 2 mL of 10% FBS/DPBS. Collect the LEC fraction in a new 15 mL tube.
   NOTE: The usual recovery for LECs is within a range (2 x 10⁶-4.2 x 10⁶) depending on the size of the MGs. Routinely, the purity of both fractions as assayed by immunohistochemistry is ~90%¹⁹ (Figure 2E).
4. Centrifuge each fraction for 5 min at 300 x g to remove the trypsin-EDTA (0.05%). Resuspend the pellet in 250 µL of Maintenance Medium and count each cell population using a hemocytometer or automated cell counter. Place the cells on ice while counting.
   NOTE: If genetic manipulation of the cell fractions is desired, primary cells can be grown on low adhesion dishes and infected with lentivirus²⁰.

4. Day 3: Combining and embedding cell fractions in an extracellular matrix

NOTE: Once the MyoEC and LEC fractions have been collected and counted, they can be combined. The typical MyoEC/LEC ratio is 1:3 (Figure 3A)¹⁹. Different studies can be performed. For example, to perform mosaic studies, fractions can be generated from wild type (WT) and mutant (Mut) mice and combined (MyoEC/LEC: WT/WT; WT/Mut; Mut/WT; Mut/Mut)²¹, or fractions can be combined using different ratios of MyoECs/LECs¹⁹.

1. Based on cell counts, calculate the number of wells (8 well chamber, see Table of Materials) that need to be prepared for 12,000 cells/well (e.g. 3,000 MyoECs:9,000 LECs).
2. Establish the base layer for 3D culture by adding 90 µL of 50% ECM (50% ECM/50% DMEM/F12, without phenol red - see the Table of Materials) to each well. Ensure there are no bubbles and the wells are coated evenly. To solidify the base layer, incubate the slides at 37 °C, 5% CO₂ for 30 min.
   NOTE: The ECM needs to stay ice-cold until this step, otherwise it will polymerize prematurely and lead to uneven base coating and polymerization. Using a base ECM enhances organoid growth in a single plane, which aids image capture. This also obtains faster organoid growth and uses less ECM²².
3. During polymerization, prepare the cell mixes. Pellet the MyoEC and LEC fractions at 300 x g for 5 min and resuspend each cell fraction in 10% ECM/90% Growth Medium so each well has 100 µL (see Table 1 for how to make Growth Medium).
   NOTE: For ease of preparation, replicate wells using the same cell mixes. These can be combined and prepared in one tube (e.g., four wells of the same cell mix can be prepared in 400 µL of 10% ECM/90% Growth Medium).
4. Add 100 µL of each cell mix in 10% ECM/90% Growth Medium to each well and allow organoids to settle for 20 min at 37 °C, 5% CO₂.
5. Once the cells have settled, gently add 100 µL of Growth Medium by gently pipetting down the chamber wall of each well. Incubate the slides at 37 °C, 5% CO₂ (see Figure 3A for the total composition of each well).
   NOTE: The Rho Kinase inhibitor, R-spondin, and Nrg1 are factors that have been identified as important for the long-term culture of organoids grown from both primary murine mammary cells and human breast cancer cells^13,14. In addition, the stem cell factors B27 and N2 extend the time that organoids can be cultured.
6. Image cells every 24 h to track their growth and gently renew the Growth Medium every 2-3 days using 100 µL/well (Figure 3B-E).
   NOTE: Use extreme care when renewing the medium. Tilt the chamber slides to collect the medium at one corner of the wells. Remove ≤100 µL of the medium and replenish with care to leave the ECM layer undisturbed.
7. If researchers are interested in investigating lactation/alveologenesis, switch to Alveologenesis Medium (see Table 1) on day 5 and continue to renew the medium every 2-3 days until day 10 (Figure 3F) or beyond by passaging using recovery solution (see the Table of Materials)¹³.
   NOTE: At this point, organoids can be isolated from the ECM by incubating at 4 °C in 400 µL of 4 °C recovery solution (see Table of Materials for protocol and reagent info).

5. Day 5 or 10: Fixing and immunostaining organoids

1. Remove the medium carefully by gently pipetting off the media (a bulb pipette works best). Rinse each well using 200 µL of 1x Dulbecco's Phosphate Buffered Saline (DPBS, see recipes).
2. Fix the organoids using cold (4 °C) 4% (w/v) paraformaldehyde (PFA, see recipes) for 10 min at room temperature.
   CAUTION: PFA is hazardous. Wear personal protective equipment (lab coat, gloves, and safety glasses). This step should be performed inside a fume hood.
   NOTE: The ECM is dissolved by the PFA treatment. Incomplete ECM removal can lead to background staining when the organoids are analyzed by immunofluorescence.
3. Remove the 4% PFA and add 200 µL of 0.2% (w/v) glycine/DPBS (see Table 1) to each well. Incubate the slides at room temperature for 30 min or 4 °C overnight on a rocking surface set to a slow setting.
   NOTE: The organoids can be stored for 1-3 days in DPBS at 4 °C prior to the next step.
4. Permeabilize the organoids using DPBS + 0.25% Triton X-100 (PBST, see Table 1) for 10 min at room temperature.
5. Block the organoids using 5% donkey serum (DS) (or other serum matching the species of the secondary antibody) in DPBS for 1 h on a rocking surface.
   NOTE: This step can be performed overnight at 4 °C on a rocking surface.
6. Prepare primary antibodies in 1% DS/DPBS. Use 125-200 µL for each well. Perform immunostaining by incubating the organoids in primary antibody overnight at 4 °C on a rocking surface.

6. Day 11: Complete immunofluorescence

1. Wash each well 2X with 200 µL PBST for 5 min. Add secondary antibody in 1% DS/DPBS using 125-200 µL for each well. Incubate at room temperature on a rocking surface for 45 min.
2. Wash each well 2X using 200 µL PBST per well.
3. Stain the nuclei using Hoechst DNA dye in DPBS (1:2,000 in DPBS) for 5 min at room temperature.
4. Remove all liquid left on the well by gently suctioning with a vacuum.
5. Carefully remove the chambers and gasket, place one drop (~30 µL) of mounting media (see Table of Materials) on each well and coverslip, taking care to remove bubbles. Allow the slide to dry in a dark space for 1-2 days. Seal with clear nail polish. Image the organoids on a confocal microscope (Figure 3E-F).

Representative Results

The protocol presented here describes a method for investigating specific lineage contributions of mammary epithelial cells by making use of mosaic organoids. To obtain primary murine cells for organoids, the mammary gland epithelium must first be isolated from the surrounding adipocyte rich stroma (Figure 1). This process is described briefly here and is also described in a previously published study¹⁸. To obtain enough cells, it is recommended that #2, 3, 4, and 5 MGs be removed (Figure 1A). An important step key to isolating a pure population of epithelial cells is removal of the lymph nodes from the #4 MGs, which are rich in immune cells that will contaminate the preparation (Figure 1A,B). The MGs were minced to generate fragments ~0.1 mm in size (Figure 1B). The tissue fragments were then enzymatically digested, a process occurring in the presence of collagenase, to release epithelia from stroma, and in the absence of trypsin, to prevent the digestion of proteins such as cadherins that maintain cell-cell contacts. The digested tissue was then centrifuged to remove lipids and filtered through a cell strainer and washed (Figure 1C). Epithelial fragments, adhering to the strainer, were released by inverting the filter and washing the membrane, which transferred the epithelial fragments onto a polystyrene dish (Figure 1D). These fragments appeared as small, branched structures (Figure 1E).

The purified epithelial fragments were incubated for 24 h. They settled down onto the dish and adhered, forming flat, pancake-like structures with an outer layer of MyoECs encircling inner LECs (Figure 2A-B). Figure 2C shows the edge of such a pancake-like structure from a wild type animal. Trypsin treatment differentially detached the MyoECs, which detached first and appeared as bright, rounded cells that encircled the core of remaining cuboidal LECs (Figure 2C,F). The detachment of the MyoECs was carefully monitored using brightfield microscopy and occurred within 3-6 min. Once the MECs were collected, LECs were subsequently detached through a second, longer trypsin treatment of 7-15 min. The time required for cell detachment depends on the trypsin concentration and freshness. The overall purity of the two cell compartments was ~90%, as assayed by counting cells that were KRT14-positive and E-Cadherin (CDH1)-negative in the MyoEC fraction and cells that were KRT14-negative and E-Cadherin-positive in the LEC fraction (Figure 2D-E)¹⁹. We discovered that some of the MyoECs were removed from the top of the pancake-like structure as well as from the outer edges. This was observed by using tissue fragments collected from mice labeled with an inducible, fluorescent basal marker (Cytokeratin 14 (KRT14)-CreERT1; R26RYFP/+) and injected with 75 mg/kg tamoxifen 5 days prior to harvest. In Figure 2F,G the detachment of MyoECs from around the edges of the pancake structure is readily apparent. This occurred within the first 2 min of trypsin treatment (Figure 2F). In addition, YFP-KRT14-positive cells were observed on top of the structure, where they rounded up after trypsin treatment and were removed by the rinse/collection step (Figure 2G). The unlabeled core of LECs (Figure 2H), which contained few or no YFP-KRT14-positive cells, (Figure 2I) subsequently detached in the second round of trypsin treatment.

The MyoEC and LEC fractions were collected, combined, and embedded into 10% ECM plated onto a 50% ECM base. This allowed for better optical resolution of the organoids that grew primarily along the base layer (Figure 3A). After 24 h, the cells assembled into aggregated structures that largely lacked a lumen (Figure 3B). After 48 h, nascent organoids formed as the central lumens hollowed and appeared as a lighter internal space (Figure 3C). After 10 days, the organoids were large, branched structures with well-developed lumens. Mosaic organoids generated from MyoECs harvested from wild type mice and LECs harvested from ACTb-EGFP mice were fixed in situ, immunostained with an antibody against the basal marker alpha-smooth muscle actin (SMA), and stained with the Hoechst DNA stain to show the nuclei. In the figures, the top and section views show different sets of images collected as a Z-stack and reconstructed into a 3D view (Figure 3D). The top view reveals the branched morphology of the organoids (Figure 3E'). The section view shows the bilayered epithelial structure and open lumen of these organoids (Figure 3E''). These organoids can also be differentiated at Day 5 using Alveologenesis Medium and incubated for an additional 5 days (Figure 3F). The organoids grew larger, had more branches, and contained milk. Differentiated organoids were generated as described above and immunostained with an antibody directed against the milk marker, whey acidic protein (WAP, Figure 3F). WAP is a soluble protein secreted into milk. Much of this liquid was lost when the cells were fixed and immunostained in situ. Therefore, in the top and section views, WAP staining is visible intracellularly in secreting cells and extracellularly in milk that was trapped at the cell surface during fixation (Figure 3F), although in section view a small organoid appears to contain liquid milk (Figure 3F'' boxed overlay).

Figure 1: Mammary fragment isolation. (A) Labeled schematic of a mouse's 5 MGs with unlabeled, contralateral paired MGs. (B) Images of mouse MGs with the #4 MG boxed and magnified to show how to identify the lymph node for removal. (C) Image of chopped MGs in a 6 well low adhesion plate with a ruler showing the size of the tissue pieces (~0.1 mm each). (D-E) Schematic illustrating protocol steps 2.7-2.9. (D) MG fragments were filtered through a 70 µm strainer and rinsed 4X. (E) The strainer was then inverted over a 60 mm polystyrene tissue culture dish and fragments were released into the dish. (F) Image showing the filtered tissue fragments collected on a 60 mm dish that are free of stroma. The arrows point to the smallest fragments that are collected on the 70 µm strainer. Scale bar = 100 µm. Please click here to view a larger version of this figure.

Figure 2: Differential trypsinization. (A) Brightfield image showing a tissue fragment adhered on a polystyrene dish, forming a pancake-like structure. (B-C) The first differential trypsinization step detached MyoECs that are clearly visible as bright, rounded cells after 3 min. (D) Immunofluorescent images of MyoECs (bottom) and LECs (top) using the Cytokeratin 14 (KRT14) cell marker for MyoECs, and E-Cadherin (CDH1) cell marker for LECs. (E) The expression of KRT14 and CDH1 was used to quantify the yield and purity of the differentially trypsinized cell fractions. (F-I). Representative phase-contrast and fluorescence (YFP) images of tissue fragments from Cytokeratin 14 (KRT14)-CreERT1; R26RYFP/+ MGs. Mice were injected with 75 mg/kg tamoxifen 5 days prior to harvest. (F) Detaching MyoECs (arrows) during the initial trypsin-EDTA (0.05%) 2 min after incubation. (G) A sprinkling of KRT14-YFP-MyoECs (arrows) on top of a pancake of unlabeled LECs. (H) Brightfield image of LECs after initial trypsinization and MyoEC detachment. (I) After MyoEC detachment, KRT14-YFP-MyoECs are no longer visible as shown by the absence of YFP expression. Scale bars = 30 µm (A, brightfield) 100 µm (F, brightfield), 50 µm (G, fluorescence), 100 µm (H,I). Panels A-E of this figure are modified from Macias et al.¹⁹. Please click here to view a larger version of this figure.

Figure 3: Three-dimensional organoid culture. (A) Schematic representation of single cells embedded in 10% ECM/90% Growth Medium and grown on a 50% ECM/50% DMEM base layer (protocol step 4.5). (B-C) Illustrations and phase-contrast images showing the rapid self-organizing capacities of mammary organoids generated from differentially trypsinized and recombined MyoECs and LECs at 24 h (B) and 48 h (C). Images collected using a digital widefield microscope (D) Schematic representation illustrating the top (left) or section (right) views used in E-F to show immunostained organoids. (E) Schematic representation of a single well of an 8 well chamber slide containing mammary organoids grown for 5-10 days in Growth Medium. (E'-E") Top view (E') and section view (E") of immunostained organoids at day 10 of growth. MyoECs are marked with smooth actin muscle (SMA) in pseudocolor magenta. The LECs are from ACTb-EGFP mice and are shown in pseudocolor green. Nuclei were stained with Hoechst dye. (F) Schematic representation of a single well of an 8 well chamber slide containing mammary organoids grown for 5 days in Growth Medium and 5 days in Alveologenesis Medium. (F'-F"). Top view (F') and section view (F") of immunostained organoids at day 10 of growth. The MyoECs are unmarked. The LECs from ACTb-EGFP mice are shown in pseudocolor green. The milk protein, whey acidic protein (WAP), is shown in pseudocolor yellow in the LECs and coating the inside of the organoids' lumens. Nuclei were stained with Hoechst dye. Images collected using a spinning disk confocal microscope and reconstructed in 3D using Imaris (E', E') or bottom section ~30 slices (F', F"). Scale bars = 100 µm (C), 20 µm (E), 40 µm (F). Please click here to view a larger version of this figure.

10 mL	Digestion Medium
Amount	Reagent	Notes
9.45 mL	DMEM/F12
100 µL	Antibiotic-Antimycotic (100X)
0.04 g	Class 3 Collagenase
0.04 g	Class 2 Dispase
50 µL	Gentamicin	Final Concentration: 500 µg
2.5 mL	Fetal Bovine Serum	Final concentration: 5% (v/v)
Pass through 0.22µm filter
50 mL	Maintenance Medium
Amount	Reagent	Notes
49.47 mL	DMEM/F12
0.5 mL	Antibiotic-Antimycotic (100X)
2.5 mL	Fetal Bovine Serum	Final concentration: 5% (v/v)
25 µL	Insulin	Final concentration: 250 µg
5 µL	EGF	Final concentration: 500 ng
10 mL	Growth Medium
Amount	Reagent	Notes
9.6455 mL	DMEM/F12, no phenol red
100 µL	N-2 Supplement (100x)
200 µL	B27 supplement without vitamin A (50x)
10 µL	Nrg1	Stock: 100µg/mL
42.5 µL	R-spondin	Stock: 10 µg/mL
1 µL	Rho inhibitor Y-27632	Stock: 10 µM
1 µL	EGF	Stock: 0.1 µg/µL
10 mL	Alveologenesis Medium
Amount	Reagent	Notes
9.6355 mL	DMEM/F12, no phenol red
100 µL	N-2 Supplement (100x)
200 µL	B27 supplement without vitamin A (50x)
10 µL	Nrg1	Stock: 100µg/mL
42.5 µL	R-spondin	Stock: 10 µg/mL
1 µL	Rho inhibitor Y-27632	Stock: 10 µM
5 µL	Ovine Pituitary Prolactin	Final concentration: 1 µg/mL
1 µL	Dexamethasone	Final concentration: 5 µg/mL
5 µL	Insulin	Final concentration: 5 µg/mL
1 L	10X DPBS
Amount	Reagent	Notes
80 g	NaCl
2 g	KCl
14.4 g	NaH2PO4
2.4 g	KH2PO4
1 L	di H2O
Fill to 800 mL before adding dry reagents and dissolve. Fill volume to 1 L. Adjust pH to 7.4. Autoclave to sterilize.
1 L	1X DPBS
Amount	Reagent	Notes
100 mL	10X PBS
900 mL	di H2O
1 L	PBST
Amount	Reagent	Notes
100 mL	10X PBS
2.5 mL	Triton X-100
250 mL	4% Paraformaldehyde
Amount	Reagent	Notes
10 g	Paraformaldehyde
200 mL	di H20	water must be at 60 °C
25 mL	10X DPBS
50 µL	10 N Sodium Hydroxide
Pass through a 0.45 µm filter to sterilize and assure pH is 7.4
10 mL	1 % Donkey Serum
Amount	Reagent	Notes
100 µL	Sterile Filtered Donkey Serum
9.9 mL	1X DPBS
10 mL	0.2% Glycine
Amount	Reagent	Notes
0.02 g	Glycine
10 mL	1X DPBS

Table 1: Solution recipes.

Discussion

Here, a method is presented detailing how researchers can generate 3D organoid cultures using primary MG cells. The difference between this and other protocols is that we detail a method to separate the two, distinct MG cell compartments: the outer basal MyoECs and inner LECs. Our method employs a two-step trypsin-EDTA (0.05%) treatment that we call differential trypsinization¹⁹. This procedure allows researchers to isolate MyoECs and LECs without using sophisticated flow cytometry and thus can be used for studying MGs harvested from a wide variety of mammalian species that may not have the well-characterized biomarkers required for FACS. The ability to segregate the two cell subpopulations enables researchers to genetically modify the isolated cells independently or recombine cells from animals harboring genetic mutations or labels, and thus generate mosaic organoids in 3D culture. A limitation of the current protocol is that the stromal compartment is not included in the culturing conditions. However, new methods are being developed to coculture stromal components with organoids generated from either primary cells or cell lines to better recapitulate in vivo ECM^23,24,25,26, and these methods may be adapted to this protocol. In addition, it is important to note that while this protocol achieves a great enrichment of the MyoEC and LEC fractions (~90% purification), the fractions do not represent pure cell lineages.

The success of this protocol relies on a number of key steps. First, it is important to gently but thoroughly digest the MG tissue. Overdigestion of the tissue will lead to cell death and lower recovery of epithelial cells. Incomplete digestion will result in stromal and adipose cell contamination, which will interfere with later analyses (e.g., immunofluorescence, protein analysis, and mRNA measurements). Second, it is important to thoroughly rinse the MG tissue to remove contaminating cells in protocol step 2.8. In protocol step 2.9, the MG tissue fragments are released into a 60 mm dish. Researchers should monitor the released fragments immediately, before they adhere to the dish. If fat droplets or single cells are observed, protocol steps 2.6 and 2.8-2.11 must be repeated. To do this, the medium and tissue fragments are collected from the dish, placed into a new 70 µm strainer, washed 4X with 37 °C DMEM/F12 and then released into a new 60 mm dish. Third, it is essential to watch the first trypsin-EDTA (0.05%) incubation closely because the MyoECs can detach within the first 3 min, but they can also adhere for up to 6 min. There have been instances when the trypsin-EDTA (0.05%) was suboptimal, and incubation proceeded for 10 min with successful purification of MyoECs. However, >10 min of trypsinization resulted in the simultaneous collection of MyoECs and LECs. It is also important that the dish remain undisturbed during the first incubation. Otherwise, contamination of the MyoEC fraction with LECs can occur. The reverse is also true; if MyoECs are not completely detached from the dish, they will contaminate the LEC fraction. If researchers are using reporter mice that label MyoECs or LECs exclusively, it is easier to visualize the separation under a fluorescence microscope (Figure 2F-I). Finally, if researchers plan on fixing organoids for immunofluorescence analyses, the pH (7.4) and temperature (4 °C) of the 4% PFA is important for successful dissociation of the ECM. If the organoids are collected for other analyses (e.g., protein and mRNA measurements), it is important that the recovery solution be at 4 °C. If the ECM is not dissolving, incubation with the recovery solution can be extended by 10 min (i.e., 30 min total incubation). However, longer incubation periods will lead to loss of 3D structure and cell death. The recovery protocol (listed in the Table of Materials) specifies the use of wide-bore tips. This is important for maintaining the 3D structure of the organoids as well as the integrity of the cells.

In addition to these four key steps, there are two factors that influence the success of the protocol. First, organoid growth can be limited by genetic mutations that reduce cell proliferation and therefore reduce organoid growth in ECM. If only a few organoids are obtained, the subsequent fixation step frequently results in their loss. To address this, the number of cells embedded within the ECM should be increased while retaining the ratio of MyoECs:LECs (protocol steps 4.1-4.2). Second, once the cells are transferred into an ECM it is important to watch their growth daily and be vigilant about media renewal (every 2-3 days). This protocol specifies phenol red free reagents for better visualization, but the same success and growth is achieved using phenol red positive reagents. The days when medium renewal occurs prior to fixation (protocol step 4.6) should be performed with extreme care to reduce cell loss. The 10% ECM top layer is delicate; therefore washes or medium renewal should be performed by pipetting fluid down the chamber walls to minimize mechanical disturbances.

Differentiation of the organoids into milk-producing acini requires treatment with differentiation supplements: hydrocortisone or dexamethasone, insulin, and prolactin. In this protocol, dexamethasone is recommended. In addition, while prolactin is commercially available, the prolactin used in this protocol was obtained from the National Hormone and Peptide Program. Again, it is very important to leave the organoids undisturbed when changing the Alveologenesis Medium. Differentiation requires a minimum of 5 days. This can be extended another 3-5 days, but the base layer of ECM degrades after 10-12 days. Differentiated organoids are filled with milk and their lumens appear darker.

This is an efficient technique that can be used to address compartment-specific, lineage contributions to mammary epithelial morphogenesis and differentiation. With this technique, researchers can generate mosaic organoids comprising differentially genetically manipulated MyoECs and LECs²¹, or MyoECs and LECs obtained from mice harboring different genetic mutations. This allows researchers to better understand the contributions of lineage-specific cell compartments to organ morphogenesis and the acquisition of specialized functions such as milk production.

Materials

Name	Company	Catalog Number	Comments
15 ml High-Clarity Polypropylene Conical Tube (BD Falcon)	Fisher Scientific	352096
24 well ultra-low attachment plate (Corning)	Fisher Scientific	CLS3473-24EA
35 mm TC-treated Easy-Grip Style Cell Culture Dish (BD Falcon)	Fisher Scientific	353001
50 ml High-Clarity polypropylene conical tube (BD Falcon)	Fisher Scientific	352098
60 mm TC-treated Easy-Grip Style Cell Culture Dish (BD Falcon)	Fisher Scientific	353004
70 µM nylon cell strainer (Corning)	Fisher Scientific	08-771-2
Antibiotic-Antimycotic (100X)	Thermo Fisher Scientific	15240062	Pen/Strep also works
B27 supplement without vitamin A (50x)	Thermo Fisher Scientific	12587010
B6 ACTb-EGFP mice	The Jackson Laboratory	003291
BD Insulin syringe 0.5 mL	Thermo Fisher Scientific	14-826-79
Class 2 Dispase (Roche)	Millipore Sigma	4942078001
Class 3 Collagenase	Worthington Biochemical	LS004206
Corning Cell Recovery solution	Fisher Scientific	354253	Follow the guidelines for use – Extraction of Three-Dimensional Structures from Corning Matrigel Matrix
Corning Costar Ultra-Low Attachment 6-well	Fisher Scientific	CLS3471
Dexamethasone	Millipore Sigma	D4902-25MG
DMEM/F12, no phenol red	Thermo Fisher Scientific	11039-021
DNase (Deoxyribonuclease I)	Worthington Biochemical	LS002007
Donkey anti-Goat 647	Thermo Fisher Scientific	A21447	Use at 1:500, Lot: 1608641, stock 2 mg/mL, RRID:AB_2535864
Donkey anti-Mouse 647	Jackson ImmunoResearch	715-606-150	Use at 1:1000, Lot: 140554, stock 1.4 mg/mL
Dulbecco's Modified Eagle Medium: Nutrient Mixture F-12 (DMEM/F12)	Thermo Fisher Scientific	11330-057
Dulbecco's phosphate-buffered saline (DPBS)	Thermo Fisher Scientific	14190-250	Without Mg2+/Ca2+
EGF	Fisher Scientific	AF-100-15-100ug
Fetal Bovine Serum	VWR	97068–085	100% US Origin, premium grade, Lot: 059B18
Fluoromount-G (Southern Biotech)	Fisher Scientific	0100-01	Referred to as mounting media in text
Gentamicin	Thermo Fisher Scientific	15710064
Glycine	Fisher Scientific	BP381-5
Goat anti-WAP	Santa Cruz Biotech	SC-14832	Use at 1:250, Lot: J1011, stock 200 µg/mL, RRID:AB_677601
Hoechst 33342	AnaSpec	AS-83218	Use 1:2000, stock is 20mM
Insulin	Millipore Sigma	I6634-100mg
KCl	Fisher Scientific	P217-500
KH2PO4	Fisher Scientific	P285-500
KRT14–CreERtam	The Jackson Laboratory	5107
Matrigel Growth Factor Reduced (GFR); Phenol Red-Free; 10 mL	Fisher Scientific	CB-40230C	Lot: 8204010, stock concentration 8.9 mg/mL
MillexGV Filter Unit 0.22 µm	Millipore Sigma	SLGV033RS
Millicell EZ SLIDE 8-well glass, sterile	Millipore Sigma	PEZGS0816	These chamber slides are great for gasket removal but other brands can work well (e.g. Lab Tek II).
Mouse anti-SMA	Millipore Sigma	A2547	Use at 1:500, Lot: 128M4881V, stock 5.2 mg/mL, RRID:AB_476701
N-2 Supplement (100x)	Thermo Fisher Scientific	17502048
NaCl	Fisher Scientific	S671-3
NaH2PO4	Fisher Scientific	S468-500
Nrg1	R&D	5898-NR-050
Ovine Pituitary Prolactin	National Hormone and Peptide Program		Purchased from Dr. Parlow at Harbor-UCLA Research and Education Institute
Paraformaldahyde	Millipore Sigma	PX0055-3
Pentobarbital	Millipore Sigma	P3761
R26R-EYFP	The Jackson Laboratory	6148
Rho inhibitor Y-27632	Tocris	1254
R-spondin	Peprotech	120-38
Sodium Hydroxide	Fisher Scientific	S318-500
Sterile Filtered Donkey Serum	Equitech-Bio Inc.	SD30-0500
Sterile Filtered Donkey Serum	Equitech-Bio Inc.	SD30-0500
Triton X-100	Millipore Sigma	x100-500ML	Laboratory grade
Trypsin EDTA 0.05%	Thermo Fisher Scientific	25300-062