Mammary organoid (MaO) models are only available for a few traditional model organisms, limiting our ability to investigate mammary gland development and cancer across mammals. This study established equine mammary organoids (EqMaOs) from cryopreserved mammary tissue, in which mammary tissue fragments were isolated and embedded into a 3D matrix to produce EqMaOs. We evaluated viability, proliferation and budding capacity of EqMaOs at different time points during culture, showing that although the number of proliferative cells decreased over time, viability was maintained and budding increased. We further characterized EqMaOs based on expression of stem cell, myoepithelial and luminal markers, and found that EqMaOs expressed these markers throughout culture and that a bilayered structure as seen in vivo was recapitulated. We used the milk-stimulating hormone prolactin to induce milk production, which was verified by the upregulation of milk proteins, most notably β-casein. Additionally, we showed that our method is also applicable to additional non-traditional mammalian species, particularly domesticated animals such as cats, pigs and rabbits. Collectively, MaO models across species will be a useful tool for comparative developmental and cancer studies.

The mammary gland arises from the ectoderm to form a rudimentary ductal tree that undergoes developmental expansion first during puberty and later during pregnancy and lactation, where it forms a network of ducts and alveoli capable of milk secretion (Oakes et al., 2006). Mammary development and lactogenesis have been largely studied using rodent models, for which there are many tools available, including transgenic mice and several well-established culture models (Darcy et al., 2000; Smits et al., 2007; Inman et al., 2015; Nguyen-Ngoc et al., 2015). One of these models is mammary organoids (MaOs), which are self-organizing tissues grown in a 3D matrix (Fatehullah et al., 2016; Sumbal et al., 2020; Srivastava et al., 2020).

Thus far, MaOs have been established for mouse (Nguyen-Ngoc et al., 2015; Jardé et al., 2016; Jamieson et al., 2017; Rubio et al., 2020), rat (Darcy et al., 2000) and human (Miller et al., 2017). MaO models for other mammalian species, however, are lacking, with the exception of a few non-traditional model species, including cow (Ellis, 1998; Martignani et al., 2018), pig (Zhang et al., 2018; Hurley, 2019), dog (Cocola et al., 2017) and marmoset (Wu et al., 2016). The availability of MaOs from non-traditional model species is important from the perspective of providing a physiologically relevant model for comparative mammary development studies. Although it is well accepted that mammary development varies between species, most notably in terms of lactogenesis strategies, there are no current methods to perform multiple-species comparative studies to elucidate the exact mechanisms underlying these variations (Rauner et al., 2018; Hughes, 2021). Moreover, the mammary gland field will greatly benefit from a comparative approach to mammary cancer using non-traditional model species, based on the observation that some mammals, such as horses, have a remarkably low incidence of mammary cancer (Boyce and Goodwin, 2017; Rauner et al., 2018; Ledet et al., 2020). A potential mechanism for horses to resist mammary cancer has been identified and studied using 2D mammosphere-derived epithelial cells, where these cells respond to DNA damage by undergoing apoptosis (Ledet et al., 2018), similar to what has been reported for peripheral blood lymphocytes and fibroblasts from elephants, a cancer-resistant long-lived mammal (Abegglen et al., 2015). Having an organoid model for horses, and other non-traditional model species, will provide a more physiologically relevant model to further explore this mechanism for mammary cancer resistance, and explore additional mechanisms, which eventually can result in the identification of novel therapeutics to treat mammary cancer.

However, establishing MaOs from non-traditional model species can present some challenges, e.g. due to lack of known markers to unequivocally identify progenitor/stem cells in the mammary gland. Because of these challenges, the most straightforward way to establish MaOs for various mammals would be to use a method that does not rely on sorting cells based on marker expression. Several candidate methods can be used to accomplish this, including establishing MaOs from fragments of mammary tissue, as shown for mouse and cow (Nguyen-Ngoc et al., 2015; Martignani et al., 2018), or from dissociated mammospheres, as shown for dog, which become enriched for mammary stem/progenitor cells after being cultured from dissociated mammary tissue under ultra-low attachment conditions (Cocola et al., 2017). Finally, MaOs may also be established from luminal or myoepithelial lineages using cell-sorting methods or a differential trypsin treatment of mammary tissue fragments (Jamieson et al., 2017; Rubio et al., 2020). Furthermore, establishing MaOs from non-traditional species can be difficult due to inconsistent availability of tissue; therefore, being able to establish MaOs from cryopreserved tissue is crucial.

Here, we have established and characterized MaOs from cryopreserved equine mammary tissues, by isolating mammary tissue fragments (MTFs) that were then embedded into a 3D matrix. To this end, we modified a method previously used to establish mouse MaOs (Nguyen-Ngoc et al., 2015) to allow the establishment of MaOs from cryopreserved mammary tissue, which has a more fibrous stroma compared with the fat pad in mice. Our salient findings were that equine MaOs are viable and proliferative, and they bud and show expression of mammary markers at both the mRNA and protein level. We also demonstrated that they have higher expression of milk proteins upon prolactin supplementation. Finally, we showed that our protocol can be used to establish MaOs from various other domesticated, non-traditional model species. Together, this study represents the first report on the establishment and characterization of MaOs from the horse, and shows that this method is applicable to other domesticated mammals, such as cats, pigs and rabbits.

Equine mammary organoids (EqMaOs) can be successfully generated from mammary tissue fragments (MTFs)

We assessed whether we could successfully establish EqMaOs from mammary gland tissue fragments (MTFs), using a modified protocol for establishing murine MaOs (Nguyen-Ngoc et al., 2015). A key difference between the two protocols was that equine mammary gland tissue samples used to isolate MTFs were pre-digested using collagenase type II and cryopreserved for long-term storage (Fig. 1), whereas the mouse study isolated MTFs from fresh mammary gland tissues. Briefly, cryopreserved equine mammary gland tissue was thawed and, after mincing, digested again with collagenase type II to generate MTFs. For a detailed description of this protocol, see Materials and Methods and Fig. 1 (Fig. 1B-E). Moreover, our modified protocol included additional steps to allow for the successful establishments of MaOs from MTFs that are isolated from mammals with more fibrous mammary gland compared with the adipose-rich murine mammary gland. First, tissues were strained using a 500 µM strainer to separate small MTFs from larger pieces of undigested stroma immediately after the second collagenase digestion. Second, any trituration was avoided as this resulted in the release of fibrous strings that contaminated the cultures and led to a poor MaO yield by breaking down MTFs into single cells. We also included a DNase treatment step to detach the isolated MTFs from any contamination, followed by several centrifugation steps to dispose of these contaminants in the supernatant, keeping the fragments in a pellet. A successful MTF isolation was characterized by the observation of several large MTFs, in addition to some single cells, nerve tissue fragments and unknown fibrous/cellular clusters (Fig. S1A), whereas failed isolations yielded only a few small MTFs with many single cells and fibrous strings (Fig. S1B).

Fig. 1.

Isolation of mammary tissue fragments from cryopreserved mammary tissue. Overview of tissue collection and cryopreservation (A), followed by mammary tissue fragment isolation (B). Arrows show fibrous contamination that can be seen with the naked eye. Scale bars: 100 μm. For a detailed description, see the Materials and Methods.

Fig. 1.

Isolation of mammary tissue fragments from cryopreserved mammary tissue. Overview of tissue collection and cryopreservation (A), followed by mammary tissue fragment isolation (B). Arrows show fibrous contamination that can be seen with the naked eye. Scale bars: 100 μm. For a detailed description, see the Materials and Methods.

Next, equine MTFs from a successful isolation were embedded in growth factor reduced (GFR) Matrigel and supplemented with organoid media. The formation of equine MaOs from MTFs was visible within 1 day of culturing, after which MaOs became progressively rounder (Fig. S2A). This was corroborated using time lapse imaging to visualize an equine MTF forming into an EqMaO (Movie 1). Additionally, we observed that MaOs were still viable and showed continued growth at 14 days of culture (Fig. S2B).

Increasing epidermal growth factor (EGF) concentration affects the percentage of budding EqMaOs

Based on previous studies reporting that budding morphology in murine MaOs, generated using the MTF isolation method, increases when the concentration of growth factors is increased (Simian et al., 2001; Fata et al., 2007), we decided to evaluate whether this is the same for EqMaOs. For this assessment, we defined a bud as a rounded protrusion with smooth outer edges, and EqMaOs were defined as budding when they had at least one bud. Briefly, EqMaOs were cultured for 3 days with increasing concentrations of EGF (2.5, 5, 10 and 20 nM) then fixed, and images were captured to analyze the percentage of budding MaOs. Two controls were included, with one control being untreated (plain organoid medium with 0 nM EGF) and the other being a vehicle control, consisting of PBS with 10% 10 mM acetic acid. The latter control served to evaluate whether the vehicle that carries EGF has any effect on EqMaO budding, and no statistically significant difference was found in the percentage of budding EqMaOs between the untreated (39.83±6.38%) and vehicle (36.84±8.1%1) controls (P=0.3999). Overall, there was an increase in percentage of budding EqMaOs when EGF concentration was increased, with a statistically significant difference between the untreated (0 nM) control and organoids grown in either 10 or 20 nM EGF (Fig. S3).

EqMaOs are viable and proliferative, and continue budding throughout a 14-day culture period

To evaluate in more depth whether EqMaOs remain viable during long-term culture, we grew EqMaOs for 3, 6 and 14 days, dissociated them into a single cell suspension, and determined the percentage of live, damaged and dead cells using flow cytometry. EqMaOs contained 75-80% viable cells at all three culture times tested, without any significant difference across these time points (P=0.2546) (Fig. 2Ai). A subset of EqMaOs were also labeled for live and dead cells in situ, and showed the presence of some sparse dead cells, corroborating the findings of the flow cytometric analysis (Fig. 2Aii). Next, we assessed the percentage of proliferating cells in EqMaOs using EdU, and calculated percentage of EdU-positive cells relative to DAPI-stained cells using ImageJ. EqMaOs cultured for 3 and 6 days had 22.77±0.34% and 18.89±2.52% proliferating cells, respectively, without any significant difference between these two time points (P=0.3818) (Fig. 2B). Proliferating cells were still observed in EqMaOs cultured for 14 days, albeit at a reduced percentage of 7.74±5.13%, which was significantly different from the 3-day-old (P=0.0034) and 6-day-old (P=0.0145) EqMaOs (Fig. 2B). Finally, we also evaluated the percentage of budding EqMaOs (at least one bud) over time and found that the percentage was significantly higher in 6- and 14-day-old EqMaOs when compared with 3-day-old EqMaOs (P=0.0027 and P=0.0150, respectively), with an additional significant increase in 14-day-old compared with 6-day-old EqMaOs (P=0.0246) (Fig. 2C).

Fig. 2.

Viability, proliferation and budding capacity of equine mammary organoids (EqMaOs) over time. (Ai) Flow cytometric analysis of EqMaO-dissociated cells using a Live:Dead/Cytotoxicity Assay Kit to assess the percentage of live, damaged or dead cells. (Aii) Representative images of whole-mount EqMaOs stained with the Live:Dead/Cytotoxicity Assay Kit to show live (green), damaged (co-stained) or dead (red) cells (ii). (B) Percentage of EdU-positive cells (i) and representative images of EdU-labeled and DAPI-stained MaOs (ii). (C) Percentage of budding EqMaOs (at least one bud) (i) and representative images of budding EqMaOs (arrows indicate EqMaOs with at least one bud) (ii). A one-way paired ANOVA was used for B and C; ns, not significant; *P≤0.05, **P≤0.01. Data are presented as the average of three biological replicates±s.d. Horse A (circle), horse B (square) and horse C (triangle). Scale bars: 100 µm.

Fig. 2.

Viability, proliferation and budding capacity of equine mammary organoids (EqMaOs) over time. (Ai) Flow cytometric analysis of EqMaO-dissociated cells using a Live:Dead/Cytotoxicity Assay Kit to assess the percentage of live, damaged or dead cells. (Aii) Representative images of whole-mount EqMaOs stained with the Live:Dead/Cytotoxicity Assay Kit to show live (green), damaged (co-stained) or dead (red) cells (ii). (B) Percentage of EdU-positive cells (i) and representative images of EdU-labeled and DAPI-stained MaOs (ii). (C) Percentage of budding EqMaOs (at least one bud) (i) and representative images of budding EqMaOs (arrows indicate EqMaOs with at least one bud) (ii). A one-way paired ANOVA was used for B and C; ns, not significant; *P≤0.05, **P≤0.01. Data are presented as the average of three biological replicates±s.d. Horse A (circle), horse B (square) and horse C (triangle). Scale bars: 100 µm.

EqMaOs express stem, luminal, and myoepithelial cell markers

We next evaluated the expression of genes associated with a stem-like specific gene expression, based on a murine mammary gland study (Williams et al., 2009), and evaluated expression of Ly1 antibody reactive clone (Lyar), CD44 and hyaluronan-mediated motility receptor (HMMR/CD168), in equine MaOs over time. Across all three tested time points, EqMaOs expressed Lyar and CD44, and this expression was not significantly different (P>0.05) (Fig. 3A). HMMR/CD168 was also expressed at all three tested time points, although its expression was significantly reduced at day 14 (P=0.0190) when compared with days 3 and 6 of culture (Fig. 3A). We also assessed gene expression of the luminal epithelial markers estrogen receptor alpha (ERα), progesterone receptor (PR) and cytokeratin 18 (CK18) (Kumar et al., 1987; Su et al., 1996), and found all three were expressed in EqMaOs over time, with ERα and PR showing reduced expression at day 14 (P=0.0211 for both genes), and CK18 being similarly expressed across all three time points (P>0.05) (Fig. 3B). Next, we also evaluated expression of the myoepithelial marker CK14, which is predominantly expressed in basal epithelial cells of the mammary gland (Su et al., 1996), and found expression at all three time points without any significantly different expression over time (P>0.05) (Fig. 3C). Finally, as the expression of PR and ERα was decreased by 14 days of culture (Fig. 3B), which could correlate with functional maturation of the epithelium, we evaluated the expression of ELF5 and GATA3, two genes that are responsible for proliferation and differentiation of mammary epithelium during pregnancy and lactation (Zhou et al., 2005; Asselin-Labat et al., 2007). However, both genes were expressed in EqMaOs without any significantly different expression over time (P>0.05) (Fig. 3D).

Fig. 3.

Gene expression of stem, luminal and myoepithelial cell markers in EqMaOs. Relative expression (2−ΔΔCT) of: the stem cell markers Ly1 antibody reactive clone (Lyar), CD44 and hyaluronan mediated motility receptor (HMMR/CD168) (A); the luminal cell markers estrogen receptor alpha (ERα), progesterone receptor (PR) and cytokeratin 18 (CK18) (B); the myoepithelial cell marker cytokeratin 14 (CK14) (C); and genes important for mammary development and lactation, E74 like ETS transcription factor 5 (ELF5) and GATA-binding protein 3 (GATA3), using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a reference gene. A one-way paired ANOVA was used; ns, not significant; *P≤0.05. Horse A (circle), horse B (square) and horse C (triangle). Data are presented as the average of three biological replicates±s.d., with exception of ERα, PR, ELF5 and GATA3 (two biological replicates±s.d.).

Fig. 3.

Gene expression of stem, luminal and myoepithelial cell markers in EqMaOs. Relative expression (2−ΔΔCT) of: the stem cell markers Ly1 antibody reactive clone (Lyar), CD44 and hyaluronan mediated motility receptor (HMMR/CD168) (A); the luminal cell markers estrogen receptor alpha (ERα), progesterone receptor (PR) and cytokeratin 18 (CK18) (B); the myoepithelial cell marker cytokeratin 14 (CK14) (C); and genes important for mammary development and lactation, E74 like ETS transcription factor 5 (ELF5) and GATA-binding protein 3 (GATA3), using glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a reference gene. A one-way paired ANOVA was used; ns, not significant; *P≤0.05. Horse A (circle), horse B (square) and horse C (triangle). Data are presented as the average of three biological replicates±s.d., with exception of ERα, PR, ELF5 and GATA3 (two biological replicates±s.d.).

To further explore expression of the myoepithelial marker CK14 and the luminal epithelial marker CK18 at the protein level, we performed immunohistochemistry (IHC) on serial sections of formalin-fixed paraffin-embedded (FFPE) EqMaOs after 3, 6 and 14 days of culture. Sections of FFPE equine mammary tissue were included as positive control and, as expected, showed positive CK14 staining in myoepithelial cells within bilayered mammary epithelia and CK18 staining in the inner luminal cells (Fig. 4A). Likewise, a similar positive staining distribution was found in the EqMaOs, especially starting at 6 days of culture, where CK14 staining was restricted to the basal layer and, albeit somewhat more dispersed, CK18 was expressed in the inner cell layer of EqMaOs (Fig. 4A). To corroborate these findings, we also performed immunofluorescence (IF) on whole-mount fixed EqMaOs, using an antibody against another myoepithelial marker, smooth muscle actin (SMA), and labeled F-actin using phalloidin. This approach is commonly used to evaluate proper bilayered orientation in mouse MaOs (Ewald et al., 2008; Nguyen-Ngoc et al., 2015). We observed that SMA was restricted to the outer basal layer in 3- and 6-day-old EqMaOs, whereas SMA appeared to be more restricted to the central spherical body of the 14-day old MaOs (Fig. 4B).

Fig. 4.

Protein expression of myoepithelial and luminal cell markers in EqMaOs. (A) Representative immunohistochemistry (IHC) images of the myoepithelial cell marker cytokeratin 14 (CK14) and the luminal cell marker cytokeratin 18 (CK18) in 3-, 6- and 14-day-old EqMaOs. Equine mammary gland tissue was included as a control. Scale bars: 50 µm. (B) Representative double immunofluorescence images of the myoepithelial cell marker smooth muscle actin (SMA, green) and F-actin (red) in whole-mount fixed 3-, 6- and 14-day-old EqMaOs. Scale bars: 100 µm.

Fig. 4.

Protein expression of myoepithelial and luminal cell markers in EqMaOs. (A) Representative immunohistochemistry (IHC) images of the myoepithelial cell marker cytokeratin 14 (CK14) and the luminal cell marker cytokeratin 18 (CK18) in 3-, 6- and 14-day-old EqMaOs. Equine mammary gland tissue was included as a control. Scale bars: 50 µm. (B) Representative double immunofluorescence images of the myoepithelial cell marker smooth muscle actin (SMA, green) and F-actin (red) in whole-mount fixed 3-, 6- and 14-day-old EqMaOs. Scale bars: 100 µm.

EqMaOs show increased expression of milk proteins after prolactin stimulation

To evaluate the functionality of EqMaOs, we cultured EqMaOs in organoid media for 7 days, followed by another 7 days of culture in organoid media supplemented with 1 µg/ml prolactin, an approach modeled after prolactin experiments with murine MaOs (Sumbal et al., 2020). EqMaOs cultured in organoid medium for 14 days were included as controls. We first performed RT-qPCR to assess whether the expression of β-lactoglobulin, casein α subunit 1 (CSNS1), casein β (CSN2) and casein κ (CSN3), genes encoding milk proteins present in equine milk (Sharifi-Rad et al., 2013; Wodas et al., 2020), was upregulated after prolactin treatment. Expression of β-lactoglobulin, CSNS1 and CSN3 was increased in prolactin-stimulated EqMaOs, although this did not reach statistical significance (P=0.1347, P=0.1229 and P=0.1182, respectively), likely due to the variation in the amount of upregulation observed between the three biological replicates (Fig. 5A). In contrast, CSN2 was not detectable in untreated EqMaOs, but was expressed in prolactin-stimulated EqMaOs (Fig. 5A).

Fig. 5.

β-Lactoglobulin and casein expression in EqMaOs after prolactin stimulation. (A) Expression of β-lactoglobulin, casein α subunit 1 (CSNS1), casein β (CSN2) and CSN3 (casein κ) in EqMaOs stimulated with prolactin compared with an untreated control. A one-tailed paired Student's t-test was used and data are presented as the average of three biological replicates±s.d. ns, not significant. Horse A (circle), horse B (square) and horse C (triangle). (B) Representative immunohistochemistry (IHC) images of β-lactoglobulin and β-casein expression in untreated and prolactin-stimulated EqMaOs. Equine mammary gland tissue was included as a control. (C) Representative IHC of Oil Red O staining of untreated and prolactin-stimulated EqMaOs. Equine mammary gland tissue was included as control. Scale bars: 50 µm.

Fig. 5.

β-Lactoglobulin and casein expression in EqMaOs after prolactin stimulation. (A) Expression of β-lactoglobulin, casein α subunit 1 (CSNS1), casein β (CSN2) and CSN3 (casein κ) in EqMaOs stimulated with prolactin compared with an untreated control. A one-tailed paired Student's t-test was used and data are presented as the average of three biological replicates±s.d. ns, not significant. Horse A (circle), horse B (square) and horse C (triangle). (B) Representative immunohistochemistry (IHC) images of β-lactoglobulin and β-casein expression in untreated and prolactin-stimulated EqMaOs. Equine mammary gland tissue was included as a control. (C) Representative IHC of Oil Red O staining of untreated and prolactin-stimulated EqMaOs. Equine mammary gland tissue was included as control. Scale bars: 50 µm.

We then performed immunohistochemistry on FFPE untreated and prolactin-stimulated EqMaOs to evaluate β-lactoglobulin and β-casein expression at the protein level and included sections of FFPE equine mammary gland tissue as controls. Similar to the RT-qPCR results, we observed β-lactoglobulin expression in both untreated and prolactin-stimulated EqMaOs, although labeling was slightly more apparent in the prolactin-stimulated EqMaOs, whereas β-casein was not detectable in untreated EqMaOs but was present in EqMaOs stimulated with prolactin (Fig. 5B). In the equine mammary gland tissue control, β-lactoglobulin was clearly expressed and β-casein was slightly detectable (Fig. 5B). Moreover, we performed Oil Red O staining on these FFPE sections to detect the presence of triglycerides and lipids, and noticed some sparse positive red staining in all samples, with no clear difference between untreated and prolactin-stimulated EqMaOs (Fig. 5C). The latter could be due to the use of FFPE sections, as Oil Red O is usually used in frozen sections and detects only some lipoproteins in paraffin-embedded sections as the dehydration step in this procedure removes most of the fat (Tracy and Walia, 2002).

MaOs can be established successfully from other non-traditional model species using the MTF protocol

To evaluate whether our protocol could also be used to establish MaOs from other non-traditional mammalian species, we included mammary gland tissues from cat (Felidae, FeMaO), pig (Suidae, SuMaO), deer (Cervidae, CeMaO), rabbit (Leporidae, LeMaO) and prairie vole (Cricetidae, CrMaO). MTFs were isolated from cryopreserved mammary gland tissues from these five mammals (Table S1) and all isolations contained MTFs, although the prairie vole and deer MTFs were rather sparse, despite multiple MTF isolation attempts (Fig. S4A). Similar to the horse, we observed different types of contamination in the MTF isolations, including nerve tissue and fibrous fragments across all species (Fig. S4B). Muscle fragments were also noted but only in the smaller mammals (e.g. cat, rabbit and prairie vole) (Fig. S4B). When MTFs from these five mammals were cultured in growth factor reduced Matrigel and organoid media, MaOs were successfully established from all species and could be cultured for at least 6 days, except CrMaOs, where most MaOs died off by day 3 of culture (Fig. 6A). CeMaOs survived up until 6 days in culture but remained small and did not show much growth (Fig. 6A). FeMaos, SuMaOs and LeMaOs were then further characterized as follows. First, the percentage of budding MaOs (1≥buds) was analyzed in 3-day-old MaOs and averaged around 30% for all three species (Fig. 6B). Subsequent double IF labeling for SMA and F-actin showed positive SMA expression on the outside of the MaOs, which is indicative of a bilayered organoid structure, except in LeMaOs where SMA was scattered throughout (Fig. 6C). We also stained MaOs for activated caspase 3, a marker for apoptosis, and found some positive labeling in MaOs from all three species, which was primarily restricted to the outside of the MaOs (Fig. 6D). Culturing these MaOs for up to 14 days showed increased branching, which was particularly impressive in LeMaOs (Fig. 6E), and when MaOs sank through the Matrigel to the bottom of the culture plates, they started to form 2D adherent monolayers containing cells with both mesenchymal-like and cobblestone-like morphologies. (Fig. S5).

Fig. 6.

Generation of MaOs from other non-traditional model species. (A) Representative bright-field images of MaOs from cat (Felidae, FeMaO), pig (Suidae, SuMaO), deer (Cervidae, CeMaO), rabbit (Leporidae, LeMaO) and prairie vole (Cricetidae, CrMaO). (B) Percentage of budding MaOs (at least one bud) after 3 days of culture. Data are presented as the average percentage of budding MaOs for three biological replicates±s.d. (C) Representative double immunofluorescence (IF) images of the myoepithelial cell marker smooth muscle actin (SMA, green) and F-actin (red) in whole-mount fixed 3-day-old FeMaOs, SuMaOs and LeMaOs. (D) Representative IF images of active caspase 3 (green) and DAPI-stained (blue) 3-day-old FeMaOs, SuMaOs and LeMaOs. (E) Representative bright-field images of FeMaOs, SuMaOs and LeMaOs after 14 days of culture. Scale bars: 50 µm (black); 100 µm (white).

Fig. 6.

Generation of MaOs from other non-traditional model species. (A) Representative bright-field images of MaOs from cat (Felidae, FeMaO), pig (Suidae, SuMaO), deer (Cervidae, CeMaO), rabbit (Leporidae, LeMaO) and prairie vole (Cricetidae, CrMaO). (B) Percentage of budding MaOs (at least one bud) after 3 days of culture. Data are presented as the average percentage of budding MaOs for three biological replicates±s.d. (C) Representative double immunofluorescence (IF) images of the myoepithelial cell marker smooth muscle actin (SMA, green) and F-actin (red) in whole-mount fixed 3-day-old FeMaOs, SuMaOs and LeMaOs. (D) Representative IF images of active caspase 3 (green) and DAPI-stained (blue) 3-day-old FeMaOs, SuMaOs and LeMaOs. (E) Representative bright-field images of FeMaOs, SuMaOs and LeMaOs after 14 days of culture. Scale bars: 50 µm (black); 100 µm (white).

There is an increasing recognition in the field of mammary gland biology, both in health and disease, that research needs to be expanded into understudied, non-traditional mammal species, as this allows comparative studies to be carried out that should result in newly gained knowledge based on the variation in mammary gland development and cancer incidence across mammals (Rauner et al., 2018; Harman et al., 2021; Hughes, 2021). Here, we report a novel method to establish equine (Eq), feline (Fe), porcine (Su) and leporine (Le) mammary organoids (MaOs), and propose these MaOs as useful models for comparative studies investigating species-level differences in mammary development, natural cancer variation, mammary gland anatomy and physiology, as well as mammary gland evolution.

Specifically, we describe a protocol using mammary tissue fragments (MTFs), synonymous to previously described murine ‘epithelial pieces’ (Fata et al., 2007) or ‘primary mammary organoids’ (Nguyen-Ngoc et al., 2015) and bovine ‘mammary organoids’ (Martignani et al., 2018), which can be embedded in Matrigel and used to grow MaOs for various non-traditional model species, with an initial focus on horses. To avoid confusion in the description of mammary tissue fragments (MTFs) as ‘organoids’, we propose that the word ‘organoid’ should be restricted to 3D-matrix-embedded mini-organs and not be used for MTFs. In addition to using MTFs to generate MaOs, other methods have also been described. For example, embedding single cells from dissociated canine mammospheres in a 3D matrix has been reported, but these MaOs appeared to lack complex structures or form stellate colonies (Cocola et al., 2017), an observation we made ourselves using equine mammospheres (data not shown). The use of sorted CD49fhigh and CD49flow cells to establish mammospheres from marmosets that were then grown into MaOs seemed more successful as no stellate colonies were observed, although these marmoset MaOs did not appear to have extensive branching structures (Wu et al., 2016). The use of stem cells to generate MaOs is another approach (Rosenbluth et al., 2020), but this is only feasible for those species with well-established cell sorting protocols for adult mammary gland stem cells and/or reprogramming protocols for induced pluripotent stem cells (IPSCs). To date, sorting protocols for adult mammary gland stem cells are primarily established for mouse (Stingl et al., 2006), human (Gudjonsson et al., 2002; Qu et al., 2017) and cow (Rauner and Barash, 2012; Cravero et al., 2015). Although the IPSC technology has been described for various mammals, both domesticated and wild (Stanton et al., 2019), it is not routinely used for generation of MaOs, at least not to our knowledge.

Owing to our interest in the naturally low incidence of mammary cancer in horses, despite the equine mammary gland being similar to the human breast in relation to development, anatomy and function (Rauner et al., 2018), we decided to establish EqMaOs using an optimized MTF protocol followed by an in-depth analysis. Epidermal growth factor (EGF) was included in the organoid medium, and we found that increasing concentrations of EGF increased the percentage of 3-day-old budding EqMaOs significantly, a finding similar to studies with murine MaOs (Simian et al., 2001; Fata et al., 2007), and the percentage of budding EqMaOs increased over time when cultured with the same concentration of EGF. The percentage of proliferating cells was similar after 3 and 6 days of culture, but showed a significant decrease after 14 days of culture, perhaps suggesting that EqMaO maintenance beyond 2 weeks of culture could be limited. Although we did not require extended culture periods for this particular study, future studies could be aimed at extending maintenance of EqMaOs by supplementing the organoid culture medium with neuregulin 1 and R-spondin 1, which have been shown to extend the culture period of murine MaOs up to 2.5 months (Jardé et al., 2016). Despite the decrease in proliferative cells in 14-day-old EqMaOs, we did not observe a decrease in expression of genes associated with mammary stem cells, with the exception of hyaluronan-mediated motility receptor (HMMR/CD168), suggesting these cells to still be present in EqMaOs after 14 days of culturing. Last, and similar to what is described for murine MaOs grown in Matrigel (Nguyen-Ngoc et al., 2015), we observed that the myoepithelial marker smooth muscle actin (SMA) appeared to be more restricted to the spherical center, and not the buds or branches, of the EqMaO.

To demonstrate the functionality of EqMaOs, we stimulated them with prolactin and evaluated the expression of several milk proteins. Gene expression analysis showed that β-lactoglobulin, casein α subunit 1 (CSNS1) and casein κ (CSN3) were upregulated in prolactin-treated EqMaOs compared with an untreated control, although this did not reach statistical significance, and that casein β (CNS2) mRNA was not detected in untreated control EqMaOs but was expressed at varying levels in prolactin-treated EqMaOs. Interestingly, this variable upregulation of these different milk protein-encoding genes was highly tissue source dependent, suggesting that EqMaOs established from different horses show a preference for upregulating specific milk proteins. For example, the highest CSN2 expression was observed in prolactin-treated EqMaOs from horse A, but prolactin-treated EqMaOs from this horse showed the lowest upregulation of the genes β-lactoglobulin, CSN1 and CSN3. Variations in bovine milk composition between individuals is well established and influenced by many variables such as genetics, environmental factors, herd, breed and age (McLean et al., 1984; Wedholm et al., 2006; Gustavsson et al., 2014); thus, we propose that a similar variation could also exist between horses and that these differences could be recapitulated in mammary organoids. At the protein level, we found that equine β-lactoglobulin was readily detectable in mammary gland tissue from a virgin mare, suggesting that this protein may be more constitutively expressed in the equine mammary gland, even in the absence of pregnancy and lactation. The protein β-casein did show differential expression in prolactin-stimulated EqMaOs versus untreated controls, demonstrating that prolactin was able to induce an upregulation of this key milk protein. These results are in line with a study on milk induction in murine MaOs, where β-casein was detected in prolactin-stimulated MaOs but not in an untreated control (Sumbal et al., 2020).

We had variable success in growing MaOs from additional, non-traditional model species. MaOs from domesticated mammals, such as cats, pigs and rabbits, were readily established and remained healthy over a 14-day culture period, whereas MaOs established from wild mammals such as deer and prairie voles did not develop well or quickly died off in culture, respectively. The latter could be due to technical difficulties related to dissecting mammary gland parenchyma and/or isolating sufficient MTFs from these species. Further analysis of FeMaOs, SuMaOs and LeMaOs showed that MaOs from cat, pig, and rabbit had a similar percentage of budding MaOs after 3 days of culture, although there appeared to be large variability between biological replicates in SuMaOs. The expression of smooth muscle actin (SMA), a myoepithelial marker, was expressed in the basal cells of FeMaOs and SuMaOs, demonstrating an intact bilayered morphology. However, SMA labeling was not restricted to basal cells in LeMaOs and appeared to be present throughout the mammary organoids. One potential explanation could be that LeMaOs need more time in culture to form a bilayered tissue morphology when compared with MaOs from other mammals that become bilayered within 3 days of culture. Taken together, this study describes the establishment of MaOs from non-traditional model species, which we propose will provide a new resource for species-specific and/or comparative mammary developmental, evolutionary and cancer focused studies.

The isolation of mammary tissue fragments is based on the protocol described by Nguyen-Ngoc and colleagues (2015) for fresh mouse mammary gland tissue but has been optimized for successful use in cryopreserved tissues from mammals with fibrous mammary stroma.

Ethics statement

As all tissues were collected after euthanasia/culling, no Institutional Animal Care and Use Committee approval is needed.

Mammary tissue collection, pre-digestion and cryopreservation

Mammary tissues were collected from female animals euthanized/culled for reasons unrelated to this study by the following: research labs at Cornell University, including horse (Equus caballus), pig (Sus scrofa), rabbit (Oryctolagus cuniculus) and prairie vole (Microtus ochrogaster); local hunters for deer (Odocoileus virginianus); or Marshall BioResources for cats (Felis catus; North Rose, NY, USA). Details of breed, age and history are provided in Table S1. Mammary parenchyma was collected using sterile technique from an area near the nipple, to maximize the proportion of glandular tissue to fat, and placed in phosphate-buffered solution (PBS) (Corning) for a maximum of 1-2 h on ice until processing. Tissue was washed with Hanks’ Balanced Salt solution (HBSS) (Thermo Fisher) with 5% antibiotic-antimycotic (Thermo Fisher) before mincing using sterile scissors into 3-5 mm3 pieces and digested in enzyme solution A, composed of 1.2 mg/ml collagenase type II (Worthington Biochemical), 100 U/ml hyaluronidase (Sigma Aldrich), 5% fetal bovine serum (FBS) (Atlanta Biological), 5 µg/ml human recombinant insulin (Thermo Fisher) and 1 µg/ml hydrocortisone (Sigma Aldrich) in 1:1 DMEM/Ham's F12 (Corning) for 3 h with rocking at 37°C. After incubation, tissues were softened by triturating with a wide-bore pipette and washed three times by centrifugation at 1250 g at room temperature in HBSS with 2% FBS. Around 1 g of tissue was then resuspended in freezing medium A, comprising FBS with 10% sterile DMSO (Sigma Aldrich), aliquoted into cryovials (Corning) and transferred to liquid nitrogen for long-term storage (Fig. 1A).

Protocol for mammary tissue fragment isolation from cryopreserved mammary tissue

Tubes and pipettes used to hold or transfer mammary tissue fragments (MTFs) must be coated with 1% BSA solution in PBS immediately before use to prevent MTFs sticking to surfaces. This can be accomplished by adding 10 ml of the coating solution to tubes, inverting the tubes briefly and removing the solution. Pipettes can be coated by aspirating the solution into the pipette and dispensing it. This 1% BSA solution can be used repeatedly during the procedure.

  • 1.

    Thaw cryopreserved mammary tissue in a 37°C water bath, swirling the tube until tissue is thawed.

  • 2.

    Immediately transfer tissue to a 15 ml tube containing 10 ml sterile PBS by carefully pouring to wash off freezing medium A. Fragments should be uniform in size and sink to the bottom of the tube (Fig. 1Bi).

  • 3.

    Centrifuge tissue at 1250 g for 2 min at room temperature and remove supernatant using a pipette.

  • 4.

    Resuspend 1 g mammary tissue in 2 ml PBS, pour into a sterile petri dish and cut into 1 mm3 pieces using sterile scissors (Fig. 1Bii).

  • 5.

    Transfer tissue to a 1% BSA-coated 50 ml tube and add 10 ml of enzyme solution B, which is composed of 2 mg/ml collagenase type II, 5% FBS, 5 µg/ml insulin (Gibco), 50 µg/ml gentamycin (Thermo Fisher) and 2 mg/ml trypsin-EDTA (Corning) in 1:1 DMEM/Ham's F12. Tissue should readily sink when swirling the tube (Fig. 1Biii).

  • 6.

    Incubate, rocking for 30-45 min at 37°C. Check the digestion every 10 min to ensure MTFs are being released from larger pieces of tissue, but are not being digested into single cells. Tissue is optimally digested when it is fully suspended and MTFs are floating (Fig. 1Biv).

  • 7.

    Transfer MTFs through a 500 µm cell strainer (pluriSelect Life Science) into a BSA-coated 50 ml tube using a BSA-coated 10 ml pipette to separate MTFs from larger pieces of undigested tissue. Transfer the tissue fragments into a BSA-coated 15 ml tube using a clean BSA-coated pipette.

  • 8.

    Centrifuge MTFs at 1250 g for 5 min at room temperature and discard supernatant. The pellet will comprise different layers of MTFs, as well as contaminating fibrous tissue and single cells (Fig. 1Ci).

  • 9.

    Rinse MTFs by adding 10 ml 1:1 DMEM/Ham's F12 with 1% penicillin/streptomycin and carefully resuspend the pellet using a BSA-coated 10 ml pipette. Centrifuge at 1250 g for 5 min at room temperature and discard supernatant by decanting.

  • 10.

    Resuspend pellet in 2 ml DMEM/Ham's F12 with 1% penicillin/streptomycin and add 50 µl 5 mg/ml DNase (Sigma Aldrich). Gently and slowly hand rock the tube for 5 min at room temperature. This will cause MTFs to detach from single cells and fibrous contamination. Check a small sample under a microscope to see whether MTFs are detached from contamination; additional DNase in 50 µl aliquots may be added if needed up to 150 µl total.

  • 11.

    Add 10 ml DMEM/Ham's F12 with 1% penicillin/streptomycin, centrifuge the tube at 1250 g for 5 min at room temperature and remove supernatant by decanting. The pellet will now appear smaller and more uniform (Fig. 1Cii).

  • 12.

    Add 10 ml DMEM/Ham's F12 with 1% penicillin/streptomycin to the tube and gently resuspend the pellet with a 1% BSA-coated pipette, making sure to resuspend single cells that may be adhered to the tube.

  • 13.

    Centrifuge at 1250 g for 3-4 s at room temperature and remove supernatant by decanting. The pellet will appear smaller with some darker cells (Fig. 1Di).

  • 14.

    Repeat steps 12-13 three more times until pellet becomes smaller and is enriched with MTFs (Fig. 1Dii). While performing this step, contamination (e.g. fibrous material) that is visible to the naked eye can be removed using an uncoated pipette tip (Fig. 1Diii).

  • 15.

    Quantify MTFs by resuspending pellet in 1 ml DMEM/Ham's F12 with 1% penicillin/streptomycin. At this point, small fragments should be visible to the naked eye without any fibrous contamination (Fig. 1E).

  • 16.

    Pipette three 10 µl droplets of the sample onto a glass slide, making sure to thoroughly mix the sample beforehand. Using a cell counter and microscope, count the number of MTFs present per droplet.

  • 17.

    Take the average of the three MTF counts and use following equation to calculate the total number of MTFs in the sample:

    One gram of tissue should produce ∼2000-5000 MTFs, depending on several factors, such as mammary tissue origin and previous handling, including digestion and cryopreservation.

  • 18.

    Either bring into culture (as described in the next section) or cryopreserve 2000 MTFs/tube in freezing solution B, consisting of organoid medium supplemented with 20% DMEM, 4.5 g/l glucose and glutagro without sodium pyruvate (Cellgro), 10% FBS and 10% sterile DMSO.

Plating MTFs and mammary organoid (MaO) culturing

Depending on the desired application, MTFs are plated in different ways, as outlined in following sub-sections. For all MTF plating, growth factor reduced (GFR) Matrigel (Corning) was thawed on ice in the refrigerator for 24 h before use. Prior to plating, all pipette tips used for Matrigel transfer were chilled inside 5 ml conical tubes on ice and culture plates were pre-warmed in a 37°C incubator for at least 5 min. After counting MTFs, 1:1 DMEM/Ham's F12 was added up to 10 ml, centrifuged once at 1250 g for 3-4 s at room temperature, and the supernatant was removed by decanting. The tube containing the MTF pellet was placed on ice for 5 min to avoid the Matrigel solidifying prematurely. MTFs were resuspended in Matrigel (10 MTFs/µl) and plated. For all experiments, mammary organoids (MaOs) were cultured at 5% CO2 and 37°C. With the exception of MaOs analyzed for viability, MaOs were fixed with 4% paraformaldehyde (PFA) (Sigma Aldrich) in PBS for 10 min, washed three times with PBS for 10 min and stored in PBS at 4°C.

Protocol for culture in 96-well plates

This method is best used for high-throughput experiments with expensive treatments. Organoids plated in triplicate wells will serve as technical replicates.

  • 1.

    Add Matrigel to the MTF pellet to reach a concentration of 10 MTFs/µl, and carefully pipette up and down to resuspend the pellet evenly.

  • 2.

    Plate one 5 µl dome of Matrigel into each well of a pre-warmed 96-well plate (Corning).

  • 3.

    Place plate in a 37°C incubator for 10-15 min for the Matrigel to solidify.

  • 4.

    Add 125 µl of organoid media, composed of 1:1 DMEM/Ham's F12 with 1% penicillin-streptomycin (Thermo Fisher), 1% insulin-transferrin-selenium (Thermo Fisher) and 5 nM epidermal growth factor (EGF) (Sigma Aldrich) (unless were indicated otherwise).

  • 5.

    Change the media every 3 days.

  • 6.

    For details of passing MaOs, see supplementary Materials and Methods and Fig. S6.

Protocol for culture in 24-well plates fitted with coverslips

This method is ideal for culturing MaOs for immunohistochemistry, collecting RNA or generating MaOs for flow cytometry.

  • 1.

    If planning to culture MaOs for less than 1 week, transfer sterile glass coverslips into the desired wells of a 24-well plate using sterile forceps and place the plate in a 37°C incubator for at least 5 min prior to plating. If planning to culture MaOs for more than 1 week, a Matrigel underlay should first be made to prevent them from attaching to coverslips and forming monolayers. To this end, place sterile glass coverslips (12 mm diameter) inside a sterile Petri dish and put the Petri dish on ice in a refrigerator for at least 10 min for the coverslips to chill. Plate 50 µl of Matrigel onto the coverslip and use the tip to spread the Matrigel, stopping 1-2 mm from the edges, and place in a 37°C incubator to solidify for at least 20-30 min. Transfer Matrigel-covered glass coverslips into the desired wells of a 24-well plate using sterile forceps and place plate in a 37°C incubator for at least 5 min prior to MaO plating.

  • 2.

    Add the desired amount of Matrigel (diluted 1:3 in 1:1 DMEM/Ham's F12 if using Matrigel-covered coverslips) to the MTF pellet to reach a 10 MTFs/µl concentration, and carefully pipet up and down to resuspend the pellet evenly.

  • 3.

    Plate 50 µl MTF domes onto the coverslip and place the 24-well plate in a 37°C incubator to solidify for at least 20-30 min.

  • 4.

    Add 1 ml organoid media, with or without 1 µg/ml sheep recombinant prolactin (Sigma Aldrich), and change the medium every 3 days. Specifically, organoids were first cultured for 7 days in organoid media, after which media were removed and replaced with organoid media supplemented with prolactin for another 7 days of culture.

Protocol for culture in MatTek dishes

This method is ideal for microscopy imaging of immunofluorescent MaOs. In order to reduce any autofluorescence and/or to prevent potential estrogenic effects, Phenol Red-free Matrigel should be used.

  • 1.

    Place a MatTek dish onto a sterile Petri dish and put on ice in the refrigerator for at least 10 min for the MatTek dish to chill.

  • 2.

    Plate 100 µl of undiluted Matrigel onto the glass area of the MatTek dish. Move the dish from side to side to spread the liquid Matrigel over the surface of the glass or manually spread with a pipette tip, and place the MatTek dish in a 37°C incubator to solidify for at least 20-30 min.

  • 3.

    Add the desired amount of Matrigel, diluted 1:3 in 1:1 DMEM/Ham's F12, to the MTF pellet to reach a 10 MTFs/µl concentration, and carefully pipet up and down to resuspend the pellet evenly.

  • 4.

    Plate 100 µl MTF domes onto the MatTek dish and place in a 37°C incubator to solidify for at least 20-30 min.

  • 5.

    Add 3 ml of organoid media, containing the experimental treatment where applicable.

  • 6.

    Check the MatTek dish daily for media evaporation and change the media every 3 days or as needed.

Viability, proliferation and budding assessments

To assess viability, 3-, 6- and 14-day-old MaOs were dissociated into single cell suspensions using trypsin-EDTA and labeled using a Live:Dead/Cytotoxicity Assay Kit (Abcam), according to the manufacturer's instructions. Fluorescence was measured with a LSR Fortessa X-20 flow cytometer (BD Biosciences). We collected 10,000 events and analyzed data with FlowJo version 10.4.1 software. For representative fluorescence pictures, MaOs were incubated with the Live:Dead/Cytotoxicity Assay Kit for 30 min and imaged on an Olympus Fluoview FV3000 confocal microscope.

To assess proliferation, 3-, 6- and 14-day-old MaOs were cultured in organoid media supplemented with 10 µM EdU for 4 h at 5% CO2 and 37°C, and fixed as described above. EdU staining was performed using the EdU assay/EdU staining Proliferation Kit (Abcam), according to the manufacturer's instructions. Nuclei were labeled with 4′,6-diamidino-2-phenylindole (DAPI, Thermo Fisher) (1:20,000 in PBS) for 5 min at room temperature. For data analysis, z-stack images (10 nm) from 20 MaOs within one randomly selected field were taken on an Olympus Fluoview FV3000 confocal microscope (Olympus) and the number of EdU-positive cells were compared with the number of DAPI-stained cells using automated cell counting of single-color images in ImageJ.

To assess budding, bright-field images of MaOs were obtained using a Zeiss Axio Observer inverted microscope (Zeiss). One randomly selected field within three individual wells of the same condition were used to determine the percentage of budding organoids (at least one bud).

Time-lapse imaging

Organoids were plated in MatTek dishes, as described above, and bright-field images were taken every 15 min at 5× magnification on a Zeiss Axio Observer inverted microscope fitted with a built-in incubator maintained at 37°C with 5% CO2.

Immunohistochemistry and immunofluorescence

Equine MaOs were plated in 24-well plates, cultured for 3, 6 and 14 days, and prepared for histology (see supplementary Materials and Methods and Fig. S7). EqMaOs were sent to the Anatomic Pathology section of the Cornell University College of Veterinary Medicine's Animal Health Diagnostic Center for paraffin wax embedding and sectioning (https://www.vet.cornell.edu/animal-health-diagnostic-center/laboratories/anatomic-pathology). Paraffin wax-embedded MaO sections were deparaffinized and rehydrated, and permeabilized with Tris-buffered saline (TBS) with 0.5% Tween-20 (Sigma Aldrich) for 10 min at room temperature. After three washes with wash buffer [Tris-buffered saline (TBS) with 0.1% Tween-20] for 5 min, antigen retrieval was performed using sodium citrate buffer (pH 6) by microwaving on low power for 20 min, followed by another three washes with wash buffer for 5 min. Sections were blocked with anti-goat blocking solution, consisting of TBS with 1% BSA and 10% normal goat serum (Thermo Fisher) for 30 min at 37°C. Anti-cytokeratin (CK) 14 anti-CK 18 antibodies (Abcam ab7800 and ab668, respectively) that stain myoepithelial and luminal markers, respectively, were diluted 1:100 in diluent solution (TBS with 1% BSA); anti-β-lactoglobulin (Abcam, ab112893) antibody was diluted 1:1000 and anti-β-casein (Abcam, ab76025) antibody was diluted 1:100. An equal concentration of mouse IgG (Abcam, ab18443) was used as isotype control. Primary antibody incubation was carried out at 4°C overnight, followed by three washes with wash buffer for 5 min. A 0.3% hydrogen peroxide solution in TBS was applied for 15 min and slides were washed once with wash buffer for 5 min. The secondary antibodies goat anti-mouse IgG-HRP (Jackson ImmunoResearch), were diluted 1:500 and incubated for 1 h at room temperature, followed by three washes with wash buffer for 5 min. Biotin-labeled anti-mouse IgG antibodies (Jackson), diluted 1:500 and incubated for 1 h at room temperature, were used to amplify signal and was followed by incubation with streptavidin-HRP (Sigma Aldrich), diluted to 2 µg/ml, for 1 h at room temperature. A 3-amino-9-ethylcarbazole (AEC) substrate kit was used for colorimetric visualization. For Oil Red O staining, slides were deparaffinized and rehydrated, followed by a 5 min wash in distilled (DI) water, followed by a 5 min wash in 60% isopropyl alcohol in DI water. Next, slides were stained using a stock of 0.3 g Oil Red O (O0625, Sigma Aldrich) dissolved in 100 ml 100% isopropyl alcohol, which was diluted 1:3 in DI water for 15 min at room temperature for staining and washed twice for 5 min with DI water. Slides were counterstained with Gills 2 Hematoxylin, according to manufacturer's instructions (Thermo Fisher).

Immunofluorescence was used on fixed whole gels containing 3-, 6- and 14-day old MaOs to detect the presence of F-actin and smooth muscle actin (SMA). To this end, gels were permeabilized with 0.5% Triton X-100 for 30 min at room temperature, blocked with anti-goat blocking solution for 2 h at room temperature, and incubated with a directly conjugated anti-mouse SMA antibody (Abcam, ab202368) (1:500 in TBS with 1% BSA) overnight at 4°C. Gels were washed three times with PBS for 10 min, incubated with phalloidin (Abcam, ab176753) (1:1000 in TBS with 1% BSA) for 30 min at room temperature and washed another three times with PBS for 10 min. Gels were stored in PBS at 4°C until imaging on an Olympus Fluoview FV3000 confocal microscope.

RNA isolation and quantitative reverse transcription PCR (RT-qPCR)

Total RNA was isolated from a pool of EqMaOs using the Total RNA Purification Kit (Norgen Biotek), according to the manufacturer's instructions. RNA was treated with DNAse (Promega) and reverse transcribed using the M-MLV reverse transcriptase (Promega), according to the manufacturer's instructions. RT-qPCR was performed using SYBR Green Master Mix (Thermo Fisher) on a QuantStudio 3 Real-Time PCR System. The ΔΔCt=ΔCt (sample)−ΔCt (reference) method was used to determine relative fold gene expression with glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a housekeeping gene. Primers including Ly1 antibody reactive clone (Lyar), CD44, hyaluronan mediated motility receptor (HMMR/CD168), estrogen receptor alpha (ERα), progesterone receptor (PR), cytokeratin 18 (CK18), cytokeratin 14 (CK14), E74 like ETS transcription factor 5 (ELF5), GATA-binding protein 3 (GATA3), β-lactoglobulin, casein α subunit 1 (CSNS1), casein β (CSN2), casein κ (CSN3) and GAPDH were designed to target horse genes (taxid:9796) using NCBI Primer-BLAST. Primer sequences are shown in Table S2. Analyses were performed using three technical replicates and negative controls consisting of nuclease-free water controls and samples without reverse transcriptase were included.

Statistical analysis

Student's one-tailed paired t-tests and paired one-way ANOVA statistical analyses were performed in GraphPad Prism 9.1.2 (n=3, unless explicitly stated otherwise). Statistical analysis of the percentage budding organoids was carried out in RStudio (v 1.1.456). For budding analysis, data were grouped into ‘budding’ or ‘not budding’ categories and a binomial generalized linear mixed model (GLMM) was used, followed by pairwise comparisons using Tukey's method. An alpha level of 0.05 was used for all tests.

We are grateful to Sarah Johnston for the graphic design of Fig. 1A. We thank Stephen Parry for his assistance with statistical analysis of budding mammary organoids. We thank James Miller for help with immunohistochemistry and Dr Chinatsu Mukai for her assistance in the time-lapse imaging of mammary organoids. Last, we thank the Cornell Histology Core for processing our samples. Illustrations in Fig. 1, Fig. S6 and Fig. S7 were created with Biorender.com.

Author contributions

Conceptualization: A.P.B., G.R.V.d.W.; Methodology: A.P.B., R.M.H., J.R.W.; Validation: G.R.V.d.W.; Formal analysis: A.P.B., J.R.W.; Investigation: A.P.B., G.R.V.d.W.; Data curation: A.P.B.; Writing - original draft: A.P.B.; Writing - review & editing: R.M.H., G.R.V.d.W.; Supervision: R.M.H., G.R.V.d.W.; Project administration: G.R.V.d.W.; Funding acquisition: G.R.V.d.W.

Funding

This work was funded by a Cornell University Feline Health Center (FHC) grant and by funding from the Albert C Bostwick Foundation to G.R.V.d.W. A.P.B. was supported by a Liz Hanson Graduate Fellowship. Open access funding provided by the Cornell University Feline Health Center and the Albert C Bostwick Foundation. Deposited in PMC for immediate release.

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.200412.

Abegglen
,
L. M.
,
Caulin
,
A. F.
,
Chan
,
A.
,
Lee
,
K.
,
Robinson
,
R.
,
Campbell
,
M. S.
,
Kiso
,
W. K.
,
Schmitt
,
D. L.
,
Waddell
,
P. J.
,
Bhaskara
,
S.
et al. 
(
2015
).
Potential mechanisms for cancer resistance in elephants and comparative cellular response to DNA damage in humans
.
JAMA
314
,
1850
-
1860
.
Asselin-Labat
,
M.-L.
,
Sutherland
,
K. D.
,
Barker
,
H.
,
Thomas
,
R.
,
Shackleton
,
M.
,
Forrest
,
N. C.
,
Hartley
,
L.
,
Robb
,
L.
,
Grosveld
,
F. G.
,
van der Wees
,
J.
et al. 
(
2007
).
Gata-3 is an essential regulator of mammary-gland morphogenesis and luminal-cell differentiation
.
Nat. Cell Biol.
9
,
201
-
209
.
Boyce
,
S. D.
and
Goodwin
,
S. L.
(
2017
).
Mammary gland neoplasia in a Canadian mare: Challenges of diagnosis and treatment in a rural setting
.
Can. Vet. J.
58
,
628
-
630
.
Cocola
,
C.
,
Molgora
,
S.
,
Piscitelli
,
E.
,
Veronesi
,
M. C.
,
Greco
,
M.
,
Bragato
,
C.
,
Moro
,
M.
,
Crosti
,
M.
,
Gray
,
B.
,
Milanesi
,
L.
et al. 
(
2017
).
FGF2 and EGF are required for self-renewal and organoid formation of canine normal and tumor breast stem cells
.
J. Cell. Biochem.
118
,
570
-
584
.
Cravero
,
D.
,
Martignani
,
E.
,
Miretti
,
S.
,
Accornero
,
P.
and
Baratta
,
M.
(
2015
).
Bovine CD49 positive-cell subpopulation remarkably increases in mammary epithelial cells that retain a stem-like phenotype
.
Vet. Sci. Res. J.
102
,
1
-
6
.
Darcy
,
K. M.
,
Zangani
,
D.
,
Lee
,
P.-P. H.
and
Ip
,
M. M.
(
2000
).
Isolation and culture of normal rat mammary epithelial cells
. In
Methods in Mammary Gland Biology and Breast Cancer Research
(ed.
M. M.
Ip
and
B. B.
Asch
), pp.
163
-
175
.
Boston, MA
:
Springer
.
Ellis
,
S. E.
(
1998
).
Mechanisms controlling ductal morphogenesis in the ruminant mammary gland
.
PhD thesis
,
Virginia Polytechnic Institute and State University
,
Blacksburg
,
VA
.
Ewald
,
A. J.
,
Brenot
,
A.
,
Duong
,
M.
,
Chan
,
B. S.
and
Werb
,
Z.
(
2008
).
Collective epithelial migration and cell rearrangements drive mammary branching morphogenesis
.
Dev. Cell
14
,
570
-
581
.
Fata
,
J. E.
,
Mori
,
H.
,
Ewald
,
A. J.
,
Zhang
,
H.
,
Yao
,
E.
,
Werb
,
Z.
and
Bissell
,
M. J.
(
2007
).
The MAPK(ERK-1,2) pathway integrates distinct and antagonistic signals from TGFα and FGF7 in morphogenesis of mouse mammary epithelium
.
Dev. Biol.
306
,
193
-
207
.
Fatehullah
,
A.
,
Tan
,
S. H.
and
Barker
,
N.
(
2016
).
Organoids as an in vitro model of human development and disease
.
Nat. Cell Biol.
18
,
246
-
254
.
Gudjonsson
,
T.
,
Villadsen
,
R.
,
Nielsen
,
H. L.
,
Rønnov-Jessen
,
L.
,
Bissell
,
M. J.
and
Petersen
,
O. W.
(
2002
).
Isolation, immortalization, and characterization of a human breast epithelial cell line with stem cell properties
.
Genes Dev.
16
,
693
-
706
.
Gustavsson
,
F.
,
Buitenhuis
,
A. J.
,
Johansson
,
M.
,
Bertelsen
,
H. P.
,
Glantz
,
M.
,
Poulsen
,
N. A.
,
Lindmark Månsson
,
H.
,
Stålhammar
,
H.
,
Larsen
,
L. B.
,
Bendixen
,
C.
et al. 
(
2014
).
Effects of breed and casein genetic variants on protein profile in milk from Swedish Red, Danish Holstein, and Danish Jersey cows
.
J. Dairy Sci.
97
,
3866
-
3877
.
Harman
,
R. M.
,
Das
,
S. P.
,
Bartlett
,
A. P.
,
Rauner
,
G.
,
Donahue
,
L. R.
and
Van de Walle
,
G. R.
(
2021
).
Beyond tradition and convention: benefits of non-traditional model organisms in cancer research
.
Cancer Metastasis Rev.
40
,
47
-
69
.
Hughes
,
K.
(
2021
).
Development and pathology of the equine mammary gland
.
J. Mammary Gland Biol. Neoplasia
26
,
121
-
134
.
Hurley
,
W. L.
(
2019
).
Review: Mammary gland development in swine: embryo to early lactation
.
Animal
13
,
s11
-
s19
.
Inman
,
J. L.
,
Robertson
,
C.
,
Mott
,
J. D.
and
Bissell
,
M. J.
(
2015
).
Mammary gland development: cell fate specification, stem cells and the microenvironment
.
Development
142
,
1028
-
1042
.
Jamieson
,
P. R.
,
Dekkers
,
J. F.
,
Rios
,
A. C.
,
Fu
,
N. Y.
,
Lindeman
,
G. J.
and
Visvader
,
J. E.
(
2017
).
Derivation of a robust mouse mammary organoid system for studying tissue dynamics
.
Development
144
,
1065
-
1071
.
Jardé
,
T.
,
Lloyd-Lewis
,
B.
,
Thomas
,
M.
,
Kendrick
,
H.
,
Melchor
,
L.
,
Bougaret
,
L.
,
Watson
,
P. D.
,
Ewan
,
K.
,
Smalley
,
M. J.
and
Dale
,
T. C.
(
2016
).
Wnt and Neuregulin1/ErbB signalling extends 3D culture of hormone responsive mammary organoids
.
Nat. Commun.
7
,
13207
.
Kumar
,
V.
,
Green
,
S.
,
Stack
,
G.
,
Berry
,
M.
,
Jin
,
J.-R.
and
Chambon
,
P.
(
1987
).
Functional domains of the human estrogen receptor
.
Cell
51
,
941
-
951
.
Ledet
,
M. M.
,
Oswald
,
M.
,
Anderson
,
R.
and
Van de Walle
,
G. R.
(
2018
).
Differential signaling pathway activation in 7,12-dimethylbenz[a] anthracene (DMBA)-treated mammary stem/progenitor cells from species with varying mammary cancer incidence
.
Oncotarget
9
,
32761
-
32774
.
Ledet
,
M. M.
,
Harman
,
R. M.
,
Fan
,
J. C.
,
Schmitt-Matzen
,
E.
,
Diaz-Rubio
,
M. E.
,
Zhang
,
S.
and
Van de Walle
,
G. R.
(
2020
).
Secreted sphingomyelins modulate low mammary cancer incidence observed in certain mammals
.
Sci. Rep.
10
,
20580
.
Martignani
,
E.
,
Accornero
,
P.
,
Miretti
,
S.
and
Baratta
,
M.
(
2018
).
Bovine mammary organoids: a model to study epithelial mammary cells
.
Methods Mol. Biol.
1817
,
137
-
144
.
McLean
,
D. M.
,
Graham
,
E. R. B.
,
Ponzoni
,
R. W.
and
McKenzie
,
H. A.
(
1984
).
Effects of milk protein genetic variants on milk yield and composition
.
J. Dairy Res.
51
,
531
-
546
.
Miller
,
D. H.
,
Sokol
,
E. S.
and
Gupta
,
P. B.
(
2017
).
3D Primary Culture Model to Study Human Mammary Development
.
Methods Mol. Biol.
1612
,
139
-
147
.
Nguyen-Ngoc
,
K.-V.
,
Shamir
,
E. R.
,
Huebner
,
R. J.
,
Beck
,
J. N.
,
Cheung
,
K. J.
and
Ewald
,
A. J.
(
2015
).
3D culture assays of murine mammary branching morphogenesis and epithelial invasion
.
Methods Mol. Biol.
1189
,
135
-
162
.
Oakes
,
S. R.
,
Hilton
,
H. N.
and
Ormandy
,
C. J.
(
2006
).
The alveolar switch: coordinating the proliferative cues and cell fate decisions that drive the formation of lobuloalveoli from ductal epithelium
.
Breast Cancer Res.
8
,
207
.
Qu
,
Y.
,
Han
,
B.
,
Gao
,
B.
,
Bose
,
S.
,
Gong
,
Y.
,
Wawrowsky
,
K.
,
Giuliano
,
A. E.
,
Sareen
,
D.
and
Cui
,
X.
(
2017
).
Differentiation of human induced pluripotent stem cells to mammary-like organoids
.
Stem Cell Rep.
8
,
205
-
215
.
Rauner
,
G.
and
Barash
,
I.
(
2012
).
Cell hierarchy and lineage commitment in the bovine mammary gland
.
PLoS ONE
7
,
e30113
.
Rauner
,
G.
,
Ledet
,
M. M.
and
Van de Walle
,
G. R.
(
2018
).
Conserved and variable: understanding mammary stem cells across species
.
Cytometry A
93
,
125
-
136
.
Rosenbluth
,
J. M.
,
Schackmann
,
R. C. J.
,
Gray
,
G. K.
,
Selfors
,
L. M.
,
Li
,
C. M.-C.
,
Boedicker
,
M.
,
Kuiken
,
H. J.
,
Richardson
,
A.
,
Brock
,
J.
,
Garber
,
J.
et al. 
(
2020
).
Organoid cultures from normal and cancer-prone human breast tissues preserve complex epithelial lineages
.
Nat. Commun.
11
,
1711
.
Rubio
,
S.
,
Cazares
,
O.
,
Macias
,
H.
and
Hinck
,
L.
(
2020
).
Generation of mosaic mammary organoids by differential trypsinization
.
J. Vis. Exp.
157
,
e60742
.
Sharifi-Rad
,
J.
,
Alfatemi
,
M. H.
and
Sharifi-Rad
,
M.
(
2013
).
Horse milk; the composition, equine milk proteins, milk allergy and homology between mammal species with horse
.
Brit. Biomed.
1
,
001
-
004
.
Simian
,
M.
,
Hirai
,
Y.
,
Navre
,
M.
,
Werb
,
Z.
,
Lochter
,
A.
and
Bissell
,
M. J.
(
2001
).
The interplay of matrix metalloproteinases, morphogens and growth factors is necessary for branching of mammary epithelial cells
.
Development
128
,
3117
-
3131
.
Smits
,
B. M. G.
,
Cotroneo
,
M. S.
,
Haag
,
J. D.
and
Gould
,
M. N.
(
2007
).
Genetically engineered rat models for breast cancer
.
Breast Dis.
28
,
53
-
61
.
Srivastava
,
V.
,
Huycke
,
T. R.
,
Phong
,
K. T.
and
Gartner
,
Z. J.
(
2020
).
Organoid models for mammary gland dynamics and breast cancer
.
Curr. Opin. Cell Biol.
66
,
51
-
58
.
Stanton
,
M. M.
,
Tzatzalos
,
E.
,
Donne
,
M.
,
Kolundzic
,
N.
,
Helgason
,
I.
and
Ilic
,
D.
(
2019
).
Prospects for the use of induced pluripotent stem cells in animal conservation and environmental protection
.
Stem Cells Transl. Med.
8
,
7
-
13
.
Stingl
,
J.
,
Eirew
,
P.
,
Ricketson
,
I.
,
Shackleton
,
M.
,
Vaillant
,
F.
,
Choi
,
D.
,
Li
,
H. I.
and
Eaves
,
C. J.
(
2006
).
Purification and unique properties of mammary epithelial stem cells
.
Nature
439
,
993
-
997
.
Su
,
L.
,
Morgan
,
P. R.
and
Lane
,
E. B.
(
1996
).
Expression of cytokeratin messenger RNA versus protein in the normal mammary gland and in breast cancer
.
Hum. Pathol.
27
,
800
-
806
.
Sumbal
,
J.
,
Chiche
,
A.
,
Charifou
,
E.
,
Koledova
,
Z.
and
Li
,
H.
(
2020
).
Primary mammary organoid model of lactation and involution
.
Front. Cell Dev. Biol.
8
,
68
.
Tracy
,
R. E.
and
Walia
,
P.
(
2002
).
A method to fix lipids for staining fat embolism in paraffin sections
.
Histopathology
41
,
75
-
79
.
Wedholm
,
A.
,
Hallén
,
E.
,
Larsen
,
L. B.
,
Lindmark Månsson
,
H.
,
Karlsson
,
A. H.
and
Allmere
,
T.
(
2006
).
Comparison of milk protein composition in a Swedish and a Danish dairy herd using reversed phase HPLC
.
Acta Agric. Scand. A Anim.
56
,
8
-
15
.
Williams
,
C.
,
Helguero
,
L.
,
Edvardsson
,
K.
,
Haldosén
,
L.-A.
and
Gustafsson
,
J.-A.
(
2009
).
Gene expression in murine mammary epithelial stem cell-like cells shows similarities to human breast cancer gene expression
.
Breast Cancer Res.
11
,
R26
.
Wodas
,
L.
,
Mackowski
,
M.
,
Borowska
,
A.
,
Puppel
,
K.
,
Kuczynska
,
B.
and
Cieslak
,
J.
(
2020
).
Genes encoding equine β-lactoglobulin (LGB1 and LGB2): Polymorphism, expression, and impact on milk composition
.
PLoS One
15
,
e0232066
.
Wu
,
A.
,
Dong
,
Q.
,
Gao
,
H.
,
Shi
,
Y.
,
Chen
,
Y.
,
Zhang
,
F.
,
Bandyopadhyay
,
A.
,
Wang
,
D.
,
Gorena
,
K. M.
,
Huang
,
C.
et al. 
(
2016
).
Characterization of mammary epithelial stem/progenitor cells and their changes with aging in common marmosets
.
Sci. Rep.
6
,
32190
.
Zhang
,
S.
,
Chen
,
F.
,
Zhang
,
Y.
,
Lv
,
Y.
,
Heng
,
J.
,
Min
,
T.
,
Li
,
L.
and
Guan
,
W.
(
2018
).
Recent progress of porcine milk components and mammary gland function
.
J. Anim. Sci. Biotechnol.
9
,
77
.
Zhou
,
J.
,
Chehab
,
R.
,
Tkalcevic
,
J.
,
Naylor
,
M. J.
,
Harris
,
J.
,
Wilson
,
T. J.
,
Tsao
,
S.
,
Tellis
,
I.
,
Zavarsek
,
S.
,
Xu
,
D.
et al. 
(
2005
).
Elf5 is essential for early embryogenesis and mammary gland development during pregnancy and lactation
.
EMBO J.
24
,
635
-
644
.

Competing interests

The authors declare no competing or financial interests.

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.

Supplementary information