We describe techniques for differentiation induction of two breast epithelial lines, HC11 and EpH4. While both require fetal calf serum, insulin, and prolactin to produce milk proteins, EpH4 cells can fully differentiate into mammospheres in three-dimensional culture. These complementary models are useful for signal transduction studies of differentiation and neoplasia.
Cadherins play an important role in the regulation of cell differentiation as well as neoplasia. Here we describe the origins and methods of the induction of differentiation of two mouse breast epithelial cell lines, HC11 and EpH4, and their use to study complementary stages of mammary gland development and neoplastic transformation.
The HC11 mouse breast epithelial cell line originated from the mammary gland of a pregnant Balb/c mouse. It differentiates when grown to confluence attached to a plastic Petri dish surface in medium containing fetal calf serum and Hydrocortisone, Insulin and Prolactin (HIP medium). Under these conditions, HC11 cells produce the milk proteins β-casein and whey acidic protein (WAP), similar to lactating mammary epithelial cells, and form rudimentary mammary gland-like structures termed "domes".
The EpH4 cell line was derived from spontaneously immortalized mouse mammary gland epithelial cells isolated from a pregnant Balb/c mouse. Unlike HC11, EpH4 cells can fully differentiate into spheroids (also called mammospheres) when cultured under three-dimensional (3D) growth conditions in HIP medium. Cells are trypsinized, suspended in a 20% matrix consisting of a mixture of extracellular matrix proteins produced by Engelbreth-Holm-Swarm (EHS) mouse sarcoma cells, plated on top of a layer of concentrated matrix coating a plastic Petri dish or multiwell plate, and covered with a layer of 10% matrix-containing HIP medium. Under these conditions, EpH4 cells form hollow spheroids that exhibit apical-basal polarity, a hollow lumen, and produce β-casein and WAP.
Using these techniques, our results demonstrated that the intensity of the cadherin/Rac signal is critical for the differentiation of HC11 cells. While Rac1 is necessary for differentiation and low levels of activated RacV12 increase differentiation, high RacV12 levels block differentiation while inducing neoplasia. In contrast, EpH4 cells represent an earlier stage in mammary epithelial differentiation, which is inhibited by even low levels of RacV12.
In normal tissues or tumors, cells have extensive opportunities for adhesion to their neighbors in a three-dimensional organization, and this is mimicked in culture by high density cell growth. Cell-to-cell adhesion is mediated mainly through cadherin receptors, which define cell and tissue architecture. Interestingly, it was recently demonstrated that cadherins also play a powerful role in signal transduction, especially in survival signaling1. Paradoxically, some of these cell-to-cell adhesion signals emanating from cadherins were recently found to be shared by both differentiation and neoplasia2. Here, we describe methods of induction and assessment of differentiation in two representative types of mouse breast epithelial cell lines, HC11 and EpH4.
The HC11 mouse breast epithelial cell line can provide a useful model for the study of epithelial cell differentiation. HC11 cells are a COMMA-1D-derived cell line, originating from the mammary gland of a mid-pregnant Balb/c mouse3. In contrast to other COMMA-1D derivative clones, the HC11 clone has no requirement for exogenously added extracellular matrix or cocultivation with other cell types for the in vitro induction of the endogenous β-casein gene by lactogenic hormones3. This cell line has been used extensively in differentiation studies because it has retained important characteristics of the normal mammary epithelium: HC11 cells can partially reconstitute the ductal epithelium in a cleared mammary fat pad4. Moreover, they can differentiate in a two-dimensional (2D) culture when grown to confluence attached to a plastic Petri dish surface in the presence of a steroid such as Hydrocortisone or Dexamethasone, in addition to Insulin and Prolactin (HIP medium) lacking epidermal growth factor (EGF), an inhibitor of differentiation5,6,7. Under these conditions, HC11 cells produce milk proteins such as β-casein and WAP, which are detectable by Western blotting within 4 days following induction. At the same time, a portion of HC11 cells forms rudimentary mammary gland-like structures termed "domes" in a stochastic manner. Domes are visible 4–5 days following induction and gradually increase in size up to day 10, concomitant with an increase in β-casein production8. Interestingly, HC11 cells possess mutant p539, and therefore represent a preneoplastic state. For this reason, the HC11 model is ideally suited to study signaling networks of differentiation in conjunction with neoplasia in the same cell system.
EpH4 cells, a derivative of IM-2 cells, are a nontumorigenic cell line originally derived from spontaneously immortalized mouse mammary gland epithelial cells isolated from a mid-pregnant Balb/c mouse10. EpH4 cells form continuous epithelial monolayers in 2D culture, but do not differentiate into glandular-like structures10,11. However, following 3D growth in a material consisting of a mixture of extracellular matrix proteins produced by EHS mouse sarcoma cells12 (EHS matrix, matrix, or Matrigel, see Table of Materials), in addition to stimulation with HIP, EpH4 cells can recapitulate the initial stages of mammary gland differentiation. Under these conditions, EpH4 cells form spheroids (also called mammospheres) that exhibit apical-basal polarity and a hollow lumen, and are capable of producing the milk proteins β-casein and WAP, similar to lactating mammary epithelial cells. Contrary to HC11 cells, which are undifferentiated, and some express mesenchymal markers13, EpH4 cells exhibit a purely luminal morphology14. EpH4 cells have also been reported to produce milk proteins in 2D culture through stimulation with dexamethasone, insulin, and prolactin15. However, this approach precludes the study of regulatory effects that mimic the mammary gland microenvironment in 3D culture.
1. Plating HC11 Cells
2. Differentiation Induction, Monitoring, and Quantitation of HC11 Cells
3. Plating and 3D Growth of EpH4 Cells in EHS Matrix
NOTE: The matrix is liquid at temperatures <10 °C and solid at temperatures above. Store at -80 °C. Thaw at 4 °C the night before use. Pre-chill the tissue culture plates and pipette tips at -20 °C prior to handling and keep the matrix, plates, and pipette tips on ice to prevent it from solidifying.
4. Differentiation Induction of EpH4 Cells Grown in 3D (Figure 2)
NOTE: Besides differentiation, EpH4 cells can also undergo tubulogenesis when stimulated with HGF (hepatocyte growth factor) in 3D culture. Tubular outgrowths can be seen after 10 days of HIP and HGF stimulation.
5. Tubulogenesis Induction of EpH4 Cells Grown in 3D (Figure 3A and 3C)
6. Quantitation of Differentiation: Western Blotting for β-casein
It has long been known that the differentiation of epithelial cells and adipocytes requires confluence and engagement of cadherins2. We and others demonstrated that cell-to-cell adhesion and engagement of E- or N-cadherin and cadherin-11, as occurs with the confluence of cultured cells, triggers a dramatic increase in the activity of the small GTPases Rac and cell division control protein 42 (Cdc42), and this process leads to activation of interleukin–6 (IL6) family cytokines and Stat3 (signal transducer and activator of transcription-316,17)1,18,19. Because many components of the cadherin/Rac/IL6/Stat3 pathway may participate in both differentiation and neoplastic transformation, they provide molecular handles for the study of the interrelatedness of these two diametrically opposed processes.
Using the techniques described above, we demonstrated a striking dependence of differentiation upon the strength of the Rac signal in HC11 cells: While endogenous cRac1 is required for differentiation, and at low levels mutationally-activated RacV12 causes a distinct increase in differentiation capacity, high RacV12 levels trigger a dramatic block of differentiation while inducing neoplasia (Figure 1B). In contrast, even low levels of RacV12 expression blocked differentiation in EpH4 cells2. Furthermore, even though RacV12 activates Stat3 in many cell systems, including HC1116, mutationally-activated Stat3C blocked differentiation while inducing neoplastic transformation20.
We further explored the differentiation properties of EpH4 cells in a 3D matrix culture. Whereas β-casein production peaked at 8–10 days, cyclin D-1 expression was maximal at 4–6 days after HIP stimulation (Figure 3B). These findings support an inverse relationship between proliferation and differentiation in the EpH4 model. Using DAPI (nuclear) staining and confocal microscopy to determine the positioning of individual epithelial cells, we observed inner cell death and the formation of hollow lumens in mammospheres (Figure 3A,C). We further showed that addition of HGF (20 ng/mL) at 48 h to the HIP medium resulted in the formation of tubular structures (Figure 3A,C), consistent with earlier electron microscopy studies21. These approaches using the EpH4 model allow the characterization of specific features of mammary epithelial differentiation and their association with distinct signal transduction pathways.
Figure 1: Assessment and quantitation of HC11 cell differentiation. (A) A dome formed in HC11 cells expressing low levels of GFP-RacV12 at 10 days following induction of differentiation. Left: Phase-contrast; Right: Fluorescence of the same field (photos not published before). Scale bar = 100 µm; Magnification = 240x; 20x objective. (B) Effect of RacV12 upon HC11 differentiation: Low (lanes 19–24), intermediate (lanes 13–18), or high (lanes 7–12) levels of a RacV12-GFP fusion construct were expressed in HC11 cells. Following growth in the absence of EGF for 24 h, cells were induced to differentiate with HIP addition, or not, as indicated. Detergent extracts were prepared at the indicated number of days later and probed for β-casein or β-actin as a loading control. Note the dramatic increase in β-casein in low RacV12-GFP expressing cells (lanes 21–24 vs. 3–6), and the dramatic reduction upon expression of high RacV12-GFP (lanes 9–12). NE = Control cultures, grown in the absence of EGF, but not induced to differentiate (Niit et al.2, reproduced with permission). Please click here to view a larger version of this figure.
Figure 2: Schematic of overlay method for growing EpH4 cells in 3D matrix culture. (A) The wells of a 6 well plate were initially coated with 100% EHS matrix that was allowed to solidify at 37 °C forming a gelled bed of basement membrane measuring approximately 2–5 mm in thickness. EpH4 cells were seeded onto this bed as a single cell suspension in HIP medium with 20% matrix. This was overlayed with 10% matrix in HIP medium. The 10% matrix medium was replaced every other day. Cells proliferated and began to form mammospheres after 4–5 days in culture (see Protocols section). (B) Phase contrast images of EpH4 cells in 2D (monolayer) and 3D (mammosphere) culture were captured using a microscope equipped with a digital camera (300x magnification). Representative images at each stage are shown. Please click here to view a larger version of this figure.
Figure 3: Events in 3D acinar morphogenesis of EpH4 in 3D culture. (A) EpH4 cells were grown in matrix-coated 6 well plates and maintained in complete 3D medium as in Figure 2. During the early stages of morphogenesis, cells proliferated, formed clusters, and organized into two groups: (1) an outer layer of polarized cells and (2) an inner cluster of disorganized cells which underwent cell death, corresponding to apoptosis as shown previously12. The latter led to the formation of a hollow lumen, characterized by the darkening of the center of the mammospheres as seen by phase contrast microscopy. Addition of HGF (20 ng/mL) to the medium resulted in the formation of tubular structures. More than 60% of the HGF-induced aggregates showed tubulogenesis, compared to untreated control aggregates, which did not. A representative phase contrast photograph of cells at each stage was captured using a microscope equipped with a digital camera (300x magnification; Scale bar = 100 µm). Results are representative of at least three experiments. Schematic was adapted from Starova et al.22. (B) EpH4 cells grown in 3D were harvested on the indicated days. Protein concentrations were normalized and equal protein amounts (20 μg) were subjected to SDS-PAGE, followed by Western blotting and probing with the indicated antibodies. A homogenized mammary gland from a lactating mouse (MGT) was used as an in vivo comparison. p120RasGAP was used as an independent loading control. Cyclin D1 expression increased transiently coinciding with proliferating cells over the first 3–4 days. In contrast, β-casein expression peaked at 8–10 days, corresponding to the formation of mature mammospheres. Results are representative of two experiments. (C) Lumen formation and tubulogenesis of mammospheres was confirmed using DAPI staining of nuclear DNA (fluorescent blue) in aggregates grown on matrix-coated glass coverslips. Cells were fixed in 3% paraformaldehyde, permeabilized with 25 μg/mL digitonin, and nonspecific binding was blocked with 3% BSA. Cells were then stained for nuclear DNA with DAPI (blue). Coverslips were mounted onto glass slides using a mounting medium. Photomicrographs were taken of the central XY-axis plane using a multiphoton confocal microscope (Scale bar = 100µm). Images in Panels B and C are adapted from Starova et al.22. Please click here to view a larger version of this figure.
HC11 cells are ideally suited for the study of differentiation in conjunction with neoplastic transformation. An added advantage is that HC11 cells are easily infectable with Mo-MLV-based retroviral vectors to express a variety of genes. In our hands, EpH4 cells were more difficult to infect with the same retroviral vectors than HC112.
Cell-to-cell contact and growth arrest are key prerequisites for HC11 cell differentiation. Thus, to achieve uniform differentiation across the cell layer, it is important to achieve a uniform distribution of cells in the Petri dish at the time of seeding. This is especially important if quantitation of differentiation is desired. Trypsinization must be optimized so that a single cell suspension is achieved and cells do not lose viability because of the manipulations.
Cells being trypsinized must be subconfluent (50–70%), because cells grown to high densities tend to strongly adhere to each other, and separating them is difficult. To avoid drying the cells, aspirate with the Petri dish flat on the floor of the laminar flow hood, then tilt to drain quickly. Cover the Petri dish with the lid. If necessary, you can accelerate the action of the trypsin by placing the Petri dish in a 37 °C incubator.
Because the cells are very confluent during the 10 days following induction, it is important to replace the nutrients that are depleted. When the medium is changed, care should also be taken to avoid detachment of the cells, which may not adhere very well to the plastic due to high cell density. Still, in our experience it is not necessary to use Petri dishes coated with fibronectin, collagen, or CellTak to obtain good differentiation results, unlike differentiating preadipocytes such as 3T3L123 or Balb/c3T3 derivatives.
Accurate protein determination for the blotting experiments is critical. Because serum proteins tend to attach to the tissue culture grade plastic, it is important to scrape and wash the cells in a 1.5 mL plastic tube that does not attract proteins and to add the lysis buffer on the cell pellet rather than lysing the cells directly on the Petri dish24.
Regarding protein extraction, it is very important to keep all washing and lysis solutions ice-cold and the Petri dishes or EpH4 mammospheres on ice throughout. This is because in the absence of calcium in the PBS wash the cadherins are not engaged and as a result Stat3-ptyr705 is rapidly dephosphorylated25.
The size of the Petri dishes described for HC11 cells is sufficient for 3–4 Western blots (i.e., the amount of protein recovered per Petri dish is approximately 50 µg per 3 cm dish). We prefer to use 3 cm Petri dishes for differentiation experiments to facilitate protein extraction. If a 6 well tray is used instead, it is difficult to extract proteins from one well in the cold while keeping the rest of the wells sterile. Besides, the CO2 concentration is important for differentiation and it is difficult to maintain it for the rest of the wells as proteins are being extracted from one well on the bench.
In our experience, unlike the differentiation of preadipocytes23, several lots of fetal calf serum are able to support the growth as well as the differentiation of HC11 or EpH4 cells. One lot of newborn calf serum tested did not support differentiation, although it could support normal growth: the cells appeared to differentiate at first but lost their ability to differentiate after three passages in newborn calf serum.
In contrast to HC11 cells, differentiation of EpH4 cells is tightly linked to the surrounding 3D matrix architecture. We describe the steps required for EpH4 differentiation in 3D culture modified from previous studies26,27,28 and subsequent protein extraction for analysis of β-casein production. The EHS matrix is liquid at low temperatures (<10 °C) and solidifies at higher temperatures. It is normally stored at -80 °C, but working aliquots can be stored at -20 °C. Thaw the matrix at 4 °C the night before use and pre-chill tissue culture plates and pipette tips at -20 °C prior to handling. For an accurate assessment of protein concentration it is important to eliminate the matrix completely before extraction with cold PBS washing because the matrix is rich in protein, which may confound the protein determination results.
The EpH4 dependence upon the matrix for differentiation has been demonstrated in multiple experimental models: Reichman et al.10 showed that EpH4 cells induced by lactogenic hormones produced β-casein within areas of laminin deposition and intensified cytokeratin expression. Somasiri et al.11 demonstrated that PI3-kinase is required for adherens junction-dependent spheroid formation and differentiative milk protein gene expression. Furthermore, Brinkmann et al.21 demonstrated that expression of transfected HGF cDNA in EpH4 cells induced branching tubulogenesis of spheroids in a 3D culture model, analogous to the observed HGF-induced tubule formation seen here (Figure 3A,C). These findings illustrate the advantages of the EpH4 cell line model for studying signaling events that regulate mammary gland differentiation.
EpH4 cells can also form mammospheres when plated at concentrations of fetal calf serum lower than the 10% described. Interestingly, they can form mammospheres at lower concentrations, or even in the absence of 20% or 10% matrix, when plated on top of a layer of 100% matrix22,27. In the absence of matrix in the cell suspension, cooling down the cells is avoided. An added advantage is that all cells are found in a single plane, so they are easier to photograph.
Significance: Cadherin engagement in cultured cells, which approximates the physiological state of a cell in vivo, can activate the Rac/IL6/Stat3 pathway. This may be especially important in epithelial cell differentiation. Using the techniques described, our results exposed the intensity of the signal emanating from this pathway as a central determinant in the balance between cell proliferation vs. differentiation, two fundamentally opposed processes2. The in-depth investigation of the E-cadherin/Rac/Stat3 axis may uncover novel amphibolous components of the pathway depending upon the level of expression, that could be exploited as targets for cancer therapy. For such targets, complete inhibition would not be necessary to reverse the neoplastic phenotype. Partial inhibition would even be beneficial, because the residual amounts may actually block transformation and promote differentiation. This may vary with the target as well as the type of tumor in question, as shown from the different behavior of RacV12 vs. Stat3C, in HC11 vs. EpH4 cells2,20.
Future applications: The cadherin/Rac/IL6/Stat3 pathway may play a similar role in the differentiation of other epithelial cells, such as the human breast MCF10A or the canine kidney MDCK29 cell lines, as well as other types of differentiation which depend upon cell confluence, such as of preadipocytes and myoblasts30, which express different types of cadherins. Finally, this technique can be exploited for the examination of the role of Stat531 and other components of differentiative pathways.
The authors have nothing to disclose.
The HC11 cell line was kindly provided by Dr. D. Medina (Houston, TX). The authors are grateful to Dr. Andrew Craig of Queen's University for many reagents and valuable suggestions. EpH4 cells were a gift from Dr. C. Roskelley (UBC, Vancouver). Colleen Schick provided excellent technical assistance for 3D culture studies.
The financial assistance of the Natural Sciences and Engineering Research Council of Canada (NSERC), the Canadian Institutes of Health Research (CIHR), the Canadian Breast Cancer Foundation (CBCF, Ontario Chapter), the Canadian Breast Cancer Research Alliance, the Ontario Centres of Excellence, the Breast Cancer Action Kingston (BCAK) and the Clare Nelson bequest fund through grants to LR is gratefully acknowledged. BE received grant support from CIHR, CBCF, BCAK and Cancer Research Society Inc. PTG is supported by a Canada Research Chair, Canadian Foundation for Innovation, CIHR, NSERC and Canadian Cancer Society. MN was supported by a studentship from the Terry Fox Training Program in Transdisciplinary Cancer Research from NCIC, a Graduate award (QGA), and a Dean's Award from Queen's University. MG was supported by postdoctoral fellowships from the US Army Breast Cancer Program, the Ministry of Research and Innovation of the Province of Ontario and the Advisory Research Committee of Queen's University. VH was supported by a CBCF doctoral fellowship and a postdoctoral fellowship from the Terry Fox Foundation Training Program in Transdisciplinary Cancer Research in partnership with CIHR. Hanad Adan was the recipient of an NSERC summer studentship. BS was supported by a Queen's University graduate award.
30% Acrylamide/0.8% Bis Solution | Bio Rad | 1610154 | Western Blotting |
Anti-Cyclin D1 antibody | rabbit, Santa Cruz | sc-717 | Western Blotting |
Anti-p120 antibody | mouse, Santa Cruz | sc-373751 | Western Blotting |
Anti-β actin antibody | mouse, Cell Signalling technology | 3700 | Western Blotting |
Anti-β casein antibody | goat, Santa Cruz Biotechnology | sc-17971 | Western Blotting |
Aprotinin | Bio Shop | APR600 | Lysis Buffer |
Bicinchoninic Acid Solution | Sigma | B9643-1L-KC | Protein Determination |
Bovine serum albumin | Bio Shop | ALB007.500 | Protein Determination |
Clarity Western ECL Substrate | Bio Rad | 170-5061 | Western Blotting |
Copper(II) sulphate | Sigma | C2284-25ML | Protein Determination |
DAPI | Thermofisher Scientific | D1306 | Staining |
Digitonin | Calbiochem, Cedarlane Laboratories Ltd | 14952-500 | Staining |
EDTA | Bio Shop | EDT001.500 | |
Epidermal Growth Factor | Sigma | E9644 | Medium for HC11 Cells |
Fetal calf serum | PAA | A15-751 | Cell Culture Medium |
Goat- Anti Rabbit-HRP | Santa Cruz | SC-2004 | Western Blotting |
Hepatocyte Growth Factor recombinant | Gibco | PHG0321 | |
Hepes | Sigma | 7365-45-9 | Cell Culture |
Horse Anti-Mouse HRP | Cell Signalling Technology | 7076 | Western Blotting |
Hydrocortisone | Sigma | H0888 | HIP Medium |
insulin | Sigma | I6634 | HIP Medium |
Laminar-flow hood | BioGard Hood | Cell Culture | |
Leupeptin | Bio Shop | LEU001.10 | Lysis Buffer |
Matrix (Engelbreth-Holm-Swarm matrix, Matrigel) | Corning | CACB 354230 | |
Mouse Anti-Goat HRP | Santa Cruz | sc-8360 | Western Blotting |
Mowiol 4-88 Reagent | Calbiochem, Cedarlane Laboratories Ltd | 475904-100GM | Staining |
Multi-photon confocal microscope | Leica TCS SP2 | ||
Na3VO4 | Bio Shop | SOV850 | Lysis Buffer |
Na4P207 | Sigma | 125F-0262 | Lysis Buffer |
NaCl | Bio Shop | SOD001.1 | Western Blotting |
NaF | Fisher Scientific | 7681-49-4 | Lysis Buffer |
Nikon digital Camera | Coolpix 995 | ||
Nitrocellulose | Bio Rad | 1620112 | Western Blotting |
NP-40 | Sigma | 9016-45-9 | |
Paraformaldehyde | Fisher Scientific | 30525-89-4 | Staining |
Phase Contrast Microscope | Olympus IX70 | Cell Culture | |
Phenylmethylsuphonyl fluoride | Sigma | 329-98-6 | Lysis Buffer |
pMX GFP Rac G12V | Addgene | 14567 | |
Prolactin | Sigma | L6520 | HIP Medium |
RPMI-1640 | Sigma | R8758 | Cell Culture Medium |
Tissue Culture Dish 35 | Sarstedt | 83.3900. | Cell Culture |
Tissue Culture Plate-24 well | Sarstedt | 83.1836.300 | Cell Culture |
Transfer Apparatus | CBS Scientific Co | EBU-302 | Western Blotting |
Tris Acetate | Bio Shop | TRA222.500 | Western Blotting |
Trypsin | Sigma | 9002-07-7. | Cell Culture |
Tween-20 | Bio Shop | TWN510.500 | Western Blotting |
Veritical Gel Electrophoresis System | CBS Scientific Co | MGV-202-33 | Western Blotting |