SEARCH

SEARCH BY CITATION

Keywords:

  • p130Cas;
  • c-Kit;
  • Mammary luminal progenitor;
  • Differentiation;
  • Basal-like breast cancer

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

It has recently been proposed that defective differentiation of mammary luminal progenitors predisposes to basal-like breast cancer. However, the molecular and cellular mechanisms involved are still unclear. Here, we describe that the adaptor protein p130Cas is a crucial regulator of mouse mammary epithelial cell (MMEC) differentiation. Using a transgenic mouse model, we show that forced p130Cas overexpression in the luminal progenitor cell compartment results in the expansion of luminal cells, which aberrantly display basal cell features and reduced differentiation in response to lactogenic stimuli. Interestingly, MMECs overexpressing p130Cas exhibit hyperactivation of the tyrosine kinase receptor c-Kit. In addition, we demonstrate that the constitutive c-Kit activation alone mimics p130Cas overexpression, whereas c-Kit downregulation is sufficient to re-establish proper differentiation of p130Cas overexpressing cells. Overall, our data indicate that high levels of p130Cas, via abnormal c-Kit activation, promote mammary luminal cell plasticity, thus providing the conditions for the development of basal-like breast cancer. Consistently, p130Cas is overexpressed in human triple-negative breast cancer, further suggesting that p130Cas upregulation may be a priming event for the onset of basal-like breast cancer. STEM Cells2013;31:1422–1433


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

The mammary gland epithelium is composed of two cell layers, the outer basal cells and the inner luminal cells. The basal layer mainly consists of myoepithelial cells, which are responsible for milk ejection during lactation [1]. The luminal compartment includes a mixture of hormone receptor-positive and negative cells and milk-producing cells throughout lactation [2, 3]. Recent studies have delineated a mammary epithelial cell hierarchy comprising a small population of basal stem cells, a larger luminal progenitor population, and a series of differentiated cells [4–6]. Although several approaches have enabled the isolation and characterization of distinct mammary cell subpopulations [7–10], little is known about the molecular mechanisms underlying lineage commitment and differentiation. The identification of cell fate determinants in the normal mammary epithelium is emerging as crucial to understand the basis of breast cancer diversity [11].

Intriguingly, the basal-like breast cancer (BLBC) subtype, characterized by the expression of basal cell markers and a triple-negative (TN) profile (estrogen receptor ER-, progesterone receptor PR-, and HER2-) [12], is supposed to originate from luminal progenitors. Moreover, the gene expression profile of mammary luminal progenitors was found to be more similar to the BLBC profile than to that of any other tumor subtype [13, 14]. Furthermore, the deficiency of the tumor suppressor Brca1, which predisposes to BLBC in humans, is associated with defective differentiation and a basal-like phenotype of luminal progenitors [13, 15, 16]. Importantly, the targeting of BRCA1 deletion into different mammary cell subpopulations revealed that only luminal ER– progenitors initiate tumors reminiscent of human BLBCs [14].

It has recently been reported that the c-Kit tyrosine kinase receptor is a marker of mammary luminal progenitors [13, 17] and is required for their survival, but its possible role in the control of differentiation has been poorly investigated [16, 17].

p130Cas is an adaptor molecule that, given its multimodular structure and extensive tyrosine phosphorylation, can act as scaffold for numerous partners, thus regulating many cellular processes, including survival, proliferation, and migration [18, 19]. Overexpression of p130Cas has been reported in many human cancers [20]. In particular, p130Cas overexpression correlates with intrinsic resistance to tamoxifen in ER+ breast cancer [18, 20]. Furthermore, we previously reported that high levels of p130Cas synergize with the HER2 oncogene in driving mammary cell transformation and invasion [21–23]. In addition, tyrosine phosphorylation profiling of different breast cancer cell lines identified p130Cas as a component of signaling networks characteristic of BLBC [24].

Here, using a MMTV-transgenic mouse model [21], we demonstrate that p130Cas overexpression in mammary luminal progenitor cells leads to the acquirement of basal cell features and impairment of differentiation in response to lactogenic stimuli. The abnormal c-Kit activation, which results from p130Cas overexpression, is responsible for the alterations in mammary progenitor cell behavior. Finally, we show that p130Cas is overexpressed in human TN breast cancers, thus raising the possibility that p130Cas upregulation in mammary luminal progenitors might be a priming event in the development of BLBC.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Antibodies and Reagents

Antibodies against p130Cas used for immunoblotting were purchased from Becton Dickinson (BD, Franklin Lakes, NJ, USA, http://www.bd.com). Anti-p130Cas mouse monoclonal antibodies (1H9 mAb) for immunoprecipitation were produced in our laboratory as previously described [21]. Anti-p130Cas (clone CAS-14) from Labvision Thermo Scientific (Fremont, CA, USA, http://www.labvision.com) were used for immunohistochemistry. Antibodies anti-CD49f-PE (clone GoH3), anti-CD24-FITC (clone M1/69), anti-Sca1-APC, and anti-Sca1-PE (clone D7) were from eBioscience (San Diego, CA, USA, http://www.ebioscience.com) and anti-CD61-APC (clone HMbeta3-1) from Invitrogen (Carlsbad, CA). Antibodies anti-pTyr, anti-phospho-c-Kit (Y721), c-Kit, anti-actin, anti-p63, and anti-β-casein were from Santa Cruz (Palo Alto, CA, USA, http://www.scbt.it) and anti-phospho-c-Kit (Y568/Y570) from Genetex (San Antonio, TX, USA, http://www.genetex.com). Antibodies against GAPDH were from Millipore (St. Charles, MO, USA, http://www.millipore.com). Antibodies anti-keratin 14 and anti-keratin 5 were from Covance (Princeton, NJ, USA, http://www.covance.com), and anti-keratin 18 from Progen (Heidelberg, Germany). Antibodies anti-α-smooth muscle actin (SMA) were from Sigma (St. Louis, MI, USA, http//www.sigma.com) and anti-phospho-Stat5 (Y694) from Cell Signaling (Danvers, MA, USA, http://www.cellsignal.com). Growth factor reduced Matrigel and Collagen Type I were purchased from BD (Franklin Lakes, NY). Fetal bovine serum (FBS), fetal calf serum, and basic fibroblast growth factor were from Invitrogen (Grand Island, NY, USA, http//www.invitrogen.com). Soluble c-Kit ligand (SCF) was purchased from Peprotech (Brussels, Belgium, http://www.peprotech.com), epidermal growth factor (EGF), heparin, insulin, hydrocortisone, linoleic acid, dexamethasone, and prolactin from Sigma.

Mice, Mammary Gland Dissection, and Preparation of Single Mammary Cells

The use of animals was in compliance with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health and was approved by the Animal Care and Use Committee of the University of Turin. The fourth mammary glands from 12-week-old virgin or 12.5-day pregnant Mouse Mammary Tumor Virus (MMTV)-p130Cas transgenic mice [21] and wt Friend leukemia Virus B (FVB) littermates were collected after removal of intramammary lymph nodes. Two independent MMTV-p130Cas animal lines (L25 and L37) were used. Glands from virgin mice were selectively collected from animals in the estrous phase of the estrous cycle, as determined by inspection of vaginal smears. Glands were finely chopped and then digested in Dulbecco's modified Eagle's medium/F-12 medium (Gibco, Grand Island, NY, http://www.invitrogen.com) containing 1 mg/ml collagenase A (Roche Applied Science, Indianapolis, IN, USA, http://www.roche-applied-science.com). After 2 hours at 37°C, a mixture of epithelial-enriched fragments (organoids) and nonepithelial single cells was obtained. The epithelial-enriched fraction resulting from a series of differential centrifugations was resuspended in 0.64% NH4Cl for 5 minutes, then washed with Phosphate Buffered Saline before digestion with 0.25% trypsin-EDTA (Gibco). Single cells for staining with anti-CD61 antibodies were prepared using a nonenzymatic cell dissociation solution (Sigma). Incubation with 1 mg/ml DNase I (Sigma) for 3 minutes was performed before filtration through a 40 μm mesh (BD).

Cell Labeling, Flow Cytometry, and Sorting

Mammary single-cell suspensions were incubated with fluorochrome-conjugated antibodies in PBS/0.2% bovine serum albumin, for 30 minutes at 4°C. Analysis was carried out on a FACSCalibur (BD) flow cytometer. Dead cells, doublets, and higher order clumps were excluded. Sorting was carried out on MoFlo High Speed (DakoCytomation, Glostrup, Denmark, http://www.dakocytomation.com) cell sorter. For sorting experiments, mammary cells were obtained from a pool of at least four mice (two fourth glands/mice). Leibovitz's L-15 medium (Invitrogen) was used during preparation and processing of mammary cells for sorting and subsequent three-dimensional (3D) culture. Prior to sorting, cells were stained with 7-Amino-Actinomycin D (Invitrogen) for dead cell exclusion.

Mouse Mammary Epithelial Cell Culture

Culture of primary mouse mammary epithelial cells (MMECs) and KIM-2 cells [25] onto plastic is described in supporting information Data.

3D Colony-Forming Cell Assays

For 3D colony-formation cell (CFC) assays, mouse mammary single cells were plated onto growth factor reduced Matrigel in eight-well chamber slides (BD) (5,000 cells per well). Each well was filled with complete EpicultB medium supplemented with 2.5% Matrigel. At day 12 of culture, phase-contrast images were taken using a Zeiss Observer Z.1 microscope (×5/0.12 objective) and the AxioVision Rel 4.8 software.

In Vitro Lactogenic Differentiation Assay

Primary MMECs, harvested from 12.5-day pregnant mice, were seeded onto growth-factor reduced Matrigel in eight-well chamber slides (2 × 105 cells per well) and cultured in F-12 (Gibco) supplemented with 10% FBS, 1 μg/ml hydrocortisone, 5 μg/ml insulin, and 5 ng/ml EGF (growth medium) for 4 days, shifted to growth medium devoid of EGF and FBS for 1 day, and then stimulated with prolactin (3 μg/ml) for 2 days.

Lentiviral Vectors and Transduction of Primary Mouse Mammary Cells

D818V substitution is the murine equivalent of the human gain-of-function mutation D816V, found in several malignancies [26]. Mouse D818V c-Kit mutant cDNA was generated using the Quickchange Lightening site-directed mutagenesis kit (Agilent Technologies, Palo Alto, CA, USA, http://www.agilent.com) according to the manufacturer's instructions. Primers used were: fwd, CGGGCTAGCCAGAGTCATCAGGAATGATTC; rev, GAATCATTCCTGATGACTCTGGCTAGCCCG). The I.M.A.G.E. mouse c-Kit cDNA (IRAVp968F02141D clone, sequence verified, Source BioScience LifeSciences, Nottingham, U.K., http:/www.lifesciences. sourcebioscience.com) was used as template. After confirming mutagenesis by sequence analysis, D818V c-Kit mutant ORF was inserted into a pCCL lentiviral vector using the restriction sites AgeI/SalI. Generation of pCCL–p130Cas lentiviral vector was previously described [22]. A pLKO.1 lentiviral vector carrying a shRNA directed to mouse c-Kit (c-Kit shRNA) was selected in the pLKO.1 target gene shRNA set (cat. n. RMM4534, clone ID TRCN0000023669), purchased from Open Biosystem (Huntsville, AL, USA, http://www.thermoscientificbio.com). pLKO.1 scramble (Scr shRNA) vector (Addgene, Cambridge, MA, USA, http://www.addgene.com) was used as negative control. Lentiviral particles were generated as previously described [22] and concentrated by ultracentrifugation (50,000g, 2 hours). Freshly isolated single mammary cells were infected in suspension as previously described [27]. After 6 hours, transduced cells (now in clumps) were washed and plated onto Matrigel. For 3D CFC assays, clumps were redissociated into single cells by trypsinization and filtered through a 40 μm mesh prior to plating. Puromycin (Sigma) (0.8 μg/ml) was added 32 hours after infection with PLKO.1 vectors.

RNA Extraction, Retrotranscription, and Quantitative Real-Time PCR

Total RNA was extracted using Trizol reagent (Invitrogen), following the manufacturer's recommendations and then treated with DNase I (Promega, Madison, WI, USA, http://www.promega.com). Retrotranscription was carried out using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Warrington, Cheshire, U.K., http://www.invitrogen.com). Up to 20 ng of cDNA was used per quantitative real-time PCR (qRT-PCR) reaction and each analysis reaction was run in triplicate. qRT-PCR reactions were performed using Platinum qPCR SuperMix-UDG w/ROX (Invitrogen) and the Universal Probe Library (Roche Applied Science) probes on an ABI Prism 7300HT detection system (Applied Biosystems). Results were analyzed with the 2−deltadeltaCt method using the 18S rRNA predeveloped TaqMan Assay (Applied Biosystems) as internal control. RPL7 gene was also used as endogenous control. qRT-PCR reactions for RPL7 gene expression analysis were performed using Platinum SYBR Green qPCR SuperMix-UDG w/ROX. Primers are listed in supporting information Data.

Protein Extraction, Immunoprecipitation, and Western Blot Analysis

To prepare protein extracts from sorted cells, proteins were recovered from Trizol after ethanol precipitation of DNA. After protein precipitation with acetone, the protein pellet was extensively washed with 0.3 M guanidine hydrochloride in 95% ethanol/2.5% glycerol (1:1) and solubilized in 20 mM Tris-HCl, pH 7.4 and 2% SDS for 20 minutes at 60°C. To evaluate keratin protein expression, organoids were lysed using the same buffer as above. Primary mammary cells cultured onto Matrigel were released from gels using BD cell recovery solution and then lysed in Laemmli buffer. Whole mammary gland protein extracts were obtained by grinding the tissue in liquid nitrogen and solubilizing the resulting powder in 50 mM Tris pH 7.5, 1% Triton X-100, 150 mM NaCl, 1 mM EDTA, 10 mM NaF, 1 mM sodium vanadate, and complete protease inhibitor cocktail (Roche Applied Science). Mammary cells cultured onto tissue-culture plastic were lysed using the same buffer as above. Immunoprecipitation and Western blot analysis were performed as described in [22]. Protein band intensities were determined using the Image J software.

Immunofluorescence Analysis

Gels with colonies were embedded in OCT and sectioned into 5 μm slices. Sections were processed as described in supporting information Data. Images were taken using a Zeiss Observer Z.1 microscope (×20/0.50 objective) with the ApoTome module.

Immunohistochemistry Procedures

Human investigations were performed with informed consent and were preceded by local institutional review board approval.

Tissue arrays were prepared using the Tissue Arrayer instrument (ATA-100, Chemicon International, Temecula, CA, USA, http://millipore.com) from 51 invasive TN breast carcinomas. Three to four cores were obtained as previously described [28]. Normal breast tissues were obtained from reduction mammoplasties (N = 5). Samples were routinely fixed in 10% formaldehyde buffer (pH 7.4) for 24 hours, paraffin-embedded, and processed for immunohistochemical analysis. Slides were incubated with anti-p130Cas (Labvision Thermo Scientific) (1 μg/ml) for 1 hour at room temperature, after antigen retrieval (citrate buffer, at 98°C for 40 minutes). Staining was detected with EnVision System-HRP Labeled Polymer anti-mouse (DakoCytomation) and developed with the LiquidDAB Substrate Pack (BioGenex, San Ramon, CA, USA, http://www.biogenex.com). Nuclei were counterstained with Mayer hemallum. Images were taken using a Leica DM 2000 microscope.

Statistical Analysis

Unless otherwise indicated, data are presented as mean ± SEM and two-tailed Student's t test was used for comparisons, with p-values *, p < .05; **, p < .01; ***, p < .001 considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Expression of Endogenous and MMTV-Driven p130Cas in the Mouse Mammary Epithelium

To characterize the expression pattern of p130Cas within the mammary epithelium, we assessed p130Cas expression in distinct mouse mammary cell populations. The mammary epithelium comprises basal (CD24+/Low CD49fHigh) and luminal (CD24+/High CD49fLow) cells [5], which can be further divided into two functionally distinct subpopulations, the ER+ (CD24+/High Sca-1+) and ER– (CD24+/High Sca-1) luminal cells [2]. Mammary cells from wild-type (Wt) virgin mice (Fig. 1A) were sorted into basal (CD24+/Low CD49fHigh) and luminal (CD24+/High CD49fLow) cells by flow cytometry. Expression analysis of the luminal marker keratin 18 (K18) and the basal marker α-SMA, by qRT-PCR (supporting information Fig. S1A) and Western blot (Fig. 1C), confirmed the purity of the sorted cell populations. Interestingly, basal cells showed higher expression of p130Cas than luminal cells, at both transcript and protein level (Fig. 1B, 1C). To further analyze p130Cas expression, we sorted the ER+ (CD24+/High Sca-1+) and ER– (CD24+/High Sca-1) luminal cell subpopulations (Fig. 1D). As expected, ERalpha expression was found to be higher in the CD24+/High Sca-1+ than in the CD24+/High Sca-1 cell fraction (supporting information Fig. S1B). Expression of p130Cas was instead comparable between these two luminal subpopulations (Fig. 1E). Together, these data show that, although detected in all cell subsets, p130Cas expression is highly enriched in the basal cell compartment.

thumbnail image

Figure 1. Expression of endogenous and Mouse Mammary Tumor Virus (MMTV)-driven p130Cas in the mammary epithelium. (A): Representative flow cytometry dot plots of primary mouse mammary cells isolated from virgin 12-week-old mice and stained with antibodies against CD24 and CD49f. Sort gates for luminal (CD24+/HighCD49fLow) and basal (CD24+/LowCD49fHigh) cells are indicated as I (red) and II (orange), respectively. (B): Quantitative real-time PCR (qRT-PCR) analysis of p130Cas mRNA expression in mammary luminal and basal cell populations from 12-week-old virgin Wt mice after cell sorting by flow cytometry as in (A). Quantification of results from two independent preparations for each cell population is shown. (C): Western blot analysis of p130Cas, keratin 18, and α-SMA protein expression in mammary luminal and basal cell populations, sorted as in (A). GAPDH was used as loading control. Ig, antibodies used for cell staining, prior to sorting by flow cytometry, are indicated (arrows). Quantification of results from two independent preparations for each cell population is shown on the right. (D): Representative flow cytometry dot plots of primary mouse mammary cells isolated from virgin 12-week-old mice and stained with antibodies against CD24 and Sca-1. Sort gates for luminal ER+ (CD24+/High Sca-1+) and ER- (CD24+/High Sca-1) cells are indicated as I (purple) and II (blue), respectively. Gate for basal cell exclusion is also indicated (III, green). (E): qRT-PCR analysis of p130Cas mRNA expression in mammary luminal ER+ and ER- cell populations from 12-week-old virgin Wt mice after cell sorting by flow cytometry as in (D). Quantification of results from two independent preparations for each cell population is shown. (F): qRT-PCR expression analysis of myc-p130Cas mRNA in mammary luminal and basal cell populations sorted as in (A) from 12-week-old virgin Tg mice by flow cytometry. TgL25 and TgL37, two independent Tg lines. (G): qRT-PCR expression analysis of myc-p130Cas mRNA in unsorted and as in (D) sorted luminal ER- (CD24+/High Sca-1) and ER+ (CD24+/High Sca-1+) cell subpopulations from 12-week-old virgin TgL25 and TgL37 mice. Quantification of data from at least two independent cell separations for each mouse line is shown. (H): Western blot analysis of p130Cas and ERalpha expression levels on protein extracts from mammary luminal ER- (CD24+/High Sca-1) and ER+ (CD24+/High Sca-1+) cell populations, sorted as in (D). GAPDH was used as loading control. Abbreviations: α-SMA, smooth muscle actin; ER, estrogen receptor.

Download figure to PowerPoint

To understand whether regulation of p130Cas levels might be crucial to maintain the cellular homeostasis within the mammary epithelium, we analyzed MMTV-p130Cas mice [21], which overexpress p130Cas under the control of the mouse mammary tumor virus promoter. The MMTV promoter has been extensively used to address transgenes to the mammary gland but its expression pattern is still not well-defined. Therefore, we isolated mammary luminal (CD24+/High CD49fLow) and basal (CD24+/Low CD49fHigh) cells from virgin MMTV-p130Cas (Tg) mice and assessed p130Cas transgene expression by qRT-PCR. Noteworthily, qRT-PCR primers were designed to discriminate between the endogenous p130Cas gene and the p130Cas transgene, which encodes a myc-tagged p130Cas (myc-p130Cas) (supporting information Fig. S2A). In two independent Tg lines (TgL25 and TgL37), mammary luminal cells displayed strong expression of myc-p130Cas mRNA compared to basal cells, in which the transgene was nearly undetectable (Fig. 1F). Moreover, isolation of ER+ (CD24+/High Sca-1+) and ER– (CD24+/HighSca-1) luminal cells from transgenic mammary glands revealed that p130Cas transgene mRNA is predominantly expressed in the ER– (CD24+/High Sca-1) fraction (Fig. 1G). Accordingly, Tg ER– (CD24+/HighSca-1) but not ER+ (CD24+/High Sca-1+) luminal cells showed a striking increase in p130Cas protein expression compared to their Wt counterparts (Fig. 1H).

These results demonstrate that endogenous p130Cas expression is enriched in basal cells whereas MMTV promoter drives p130Cas overexpression into luminal cells, primarily into the ER– subset. Notably, since we found that two independent MMTV-p130Cas lines exhibit the same transgene expression pattern, a preferential luminal expression should be taken into account when studying MMTV-driven transgenes.

p130Cas Overexpression Confers Basal Features to Luminal ER– Cells

Mammary luminal ER– cells are mainly progenitor cells with high proliferative potential, primarily committed to the luminal lineage, but also able to generate multiple mammary cell types in vitro [2, 29, 30]. To investigate the consequences of the increase in p130Cas levels in this cell compartment, we compared the cellular composition of the mammary epithelium between Tg and Wt virgin mice by fluorescence-activated cell sorter (FACS) analysis. Tg mammary glands showed a significant increase in the number of luminal (CD24+/High CD49fLow) cells without any significant difference in the number of basal (CD24+/Low CD49fHigh) cells compared to Wt (Fig. 2A, upper panel). In addition, the amount of ER– (CD24+/High Sca-1) and ER+ (CD24+/High Sca-1+) luminal cells was estimated and, as shown in Figure 2A (lower panel), an expansion of the ER– (CD24+/High Sca-1) population was observed in Tg glands. Moreover, we assessed the levels of keratin 18 (K18) and keratin 5 (K5), luminal and basal epithelial cell markers, respectively, in freshly isolated mammary organoids from Wt and Tg mice. Consistent with the FACS, Western blot analyses revealed an increased expression of K18 in Tg organoids, while no significant difference was found in K5 expression (Fig. 2B).

thumbnail image

Figure 2. Mouse Mammary Tumor Virus (MMTV)-driven p130Cas overexpression alters the cell composition of the mammary epithelium and confers basal cell traits to luminal ER- cells. (A): Representative flow cytometry dot plots of primary mouse mammary cells isolated from virgin estrus-matched Wt and MMTV-p130Cas (Tg) mice and stained with anti-CD24 and anti-CD49f (upper panels) or anti-CD24 and anti-Sca-1 (lower panels) antibodies. Luminal (CD24+/HighCD49fLow) and basal (CD24+/LowCD49fHigh) cell populations are depicted as green and blue dots, respectively, luminal ER- (CD24+/High Sca-1) and ER+ (CD24+/High Sca-1+) cell populations as blue and purple dots. Graphs on the right show the absolute number of luminal and basal cells (upper panel; **, p < .01) and luminal ER- (CD24+/High Sca-1) and ER+ (CD24+/High Sca-1+) cells (lower panel; *, p < .05) per gland obtained from Wt and Tg mice (10 mice per group). (B): Total protein extracts from freshly isolated mammary organoids from virgin Wt and Tg mice were probed for keratin 5 and keratin 18 expression. GAPDH is provided as loading control. Quantification of protein levels is shown on the right (seven mice per group). (C): Expression (left panel) and quantification (right panel) of p63 levels on total protein extracts, prepared as above. GAPDH is provided as loading control (seven mice per group). (D): Box-and-whisker plots representing relative p63 mRNA expression measured by quantitative real-time PCR (qRT-PCR) in freshly isolated mammary organoids from virgin Wt and Tg mice (12 mice per group, left panel) (**, p < .01). (E): qRT-PCR expression analysis of p63 (left panel) and deltaNp63 (right panel) mRNA in ER+ (CD24+/High Sca-1+) and luminal ER- (CD24+/High Sca-1) mammary cell subpopulations, sorted from Wt and Tg mice by flow cytometry as in Figure 1D. Results are from three independent cell isolation experiments for each mouse group (**, p < .01). (F): Western blot analysis of p63 protein expression in luminal ER+ (CD24+/High Sca-1+) and ER- (CD24+/High Sca-1) mammary cell subpopulations, sorted from Wt and Tg virgin mice as in Figure 1D. (G): Representative phase-contrast images of colonies arising from three-dimensional (3D) culture of single mammary cells derived from Wt and Tg virgin mice. Scale bar = 100 μm. Box-and-whisker plots represent the cloning efficiency (CFC per 1,000 cells) of Wt and Tg mammary cells in 3D culture (*, p < .05). Stacked bar graphs show the percentage of acinar and solid colonies. At least six independent experiments were performed per animal group (***, p < .001). (H): Representative images of immunofluorescence staining with antibodies to keratin 5 (green) and keratin 18 (red) of sections of colonies from 3D cultures of Wt and Tg mammary single cells (upper and lower panel, respectively). Nuclei were stained with DAPI (blue). Scale bar = 50 μm. (I): Total protein extracts from colonies recovered from 3D cultures of Wt and Tg mammary cells were probed for keratin, keratin 18, and p130Cas expression levels. GAPDH was used as loading control. Graphs represent data from three independent cultures per animal group (*, p < .05; **, p < .01). Abbreviations: CFC, colony-forming cells; ER, estrogen receptor.

Download figure to PowerPoint

Unexpectedly, we observed that the expression of p63, an additional basal cell marker, was higher in Tg than Wt mammary organoids, both at protein and mRNA level (Fig. 2C, 2D). To determine which cell population might contribute to the overall increase in p63 expression, ER+ (CD24+/High Sca-1+) and ER– (CD24+/High Sca-1) luminal cells were sorted from both Wt and Tg glands and tested for p63 mRNA expression (Fig. 2E). Whereas both Wt and Tg ER+ cells displayed very low levels of p63 mRNA, Tg ER– (CD24+/High Sca-1) cells showed a marked upregulation of p63 compared to their Wt counterparts. In particular, high levels of the p63 isoform lacking the amino-terminal domain (deltaNp63), which is reported as uniquely expressed in the basal cell compartment [31–33], were detected in the Tg luminal ER– fraction (Fig. 2E). Notably, Western blot analysis confirmed the aberrant expression of p63 in Tg luminal ER– cells (Fig. 2F), thus explaining the increase in p63 protein levels observed in Tg mammary organoids despite the expansion of the luminal cell population.

The frequency and differentiation ability of Tg mammary progenitor cells was then evaluated by performing colony-forming cell (CFC) assays. Single mammary cells were prepared from Tg and Wt mice and assessed for their clonogenicity in 3D cultures. Cells from Tg glands generated a higher number of colonies than controls (Fig. 2G). Interestingly, while the majority of colonies arising from Wt cells showed an acinar morphology, Tg colonies were mostly solid and irregularly shaped (Fig. 2G). As shown in Figure 2H, acinar colonies from Wt cultures primarily consisted of K18+ cells, but also comprised few sparse K5+ cells. Conversely, solid Tg colonies were composed mainly of K5+ cells and of K18+ cells (Fig. 2H) to a lesser extent. Accordingly, Western blot analyses showed a marked increase of K5 levels along with a decrease of K18 levels in 3D Tg cultures compared to Wt (Fig. 2I). Specifically, CFC assays performed on sorted cell populations from Wt and Tg glands revealed that Tg ER– (CD24+/High Sca-1) luminal cells generated mostly solid colonies, enriched in K5+ cells, which resembled colonies derived from basal (CD24+/Low Sca-1) rather than ER– (CD24+/High Sca-1) luminal Wt cells (supporting information Fig. S3A, S3B). Overall, these data indicate that p130Cas overexpression promotes the expansion of ER– mammary progenitor cells together with a shift from a luminal to a basal pattern in their differentiation potential.

p130Cas Overexpression Impairs Lactogenic Differentiation

During pregnancy luminal progenitors are massively committed to differentiate into alveolar milk-producing cells [3, 6]. Before evaluating the effects of p130Cas overexpression on alveolar differentiation, we analyzed the expression of p130Cas in the mammary epithelium of Wt mice at distinct developmental stages. Endogenous p130Cas mRNA and protein levels were relatively high in virgin mice, declined during late gestation and lactation, and tended to rise in early involution (Fig. 3A, 3B). Similarly, we observed a reduction of p130Cas transcript and protein expression in KIM-2 mammary cells [25] upon treatment with prolactin and dexamethasone (supporting information Fig. S4A, S4B), further confirming that p130Cas downregulation occurs during lactogenic differentiation.

thumbnail image

Figure 3. p130Cas overexpression opposes lactogenic differentiation in vivo and in vitro. (A): Quantitative real-time PCR analysis of p130Cas mRNA expression in freshly isolated organoids from Wt mouse mammary glands at various stages of development. Virgin, number of mice (N) = 9; 16dP (16 days of pregnancy), N = 3; 7dL (7 days of lactation), N = 3; 7dL + 2dI (7 days of lactation + 2 days of involution), and N = 3 (*, p < .05; **, p < .01). (B): Wt mammary organoid preparations as above were probed for p130Cas protein expression by immunoblotting. Beta-casein expression is also shown. Actin is provided as loading control. Quantification of p130Cas protein expression is on the right. (C): Western blot analysis of phospho-Stat5 and β-casein protein levels in Wt and Mouse Mammary Tumor Virus (MMTV)-p130Cas (Tg) mammary glands collected at 12.5 days of pregnancy. Actin was used as loading control. Quantification is shown on the right (six animals per group) (*, p < .05; **, p < .01). (D): Representative flow cytometry profile showing the expression of CD61 in the CD24+CD49fLow cell population of mammary cells isolated from 12.5-day pregnant Wt and Tg mice. Colorless histograms represent staining with isotype control antibody. Bar graph indicates the percentage of CD61+ luminal progenitor cells in the CD24+CD49fLow mammary cell population of 12.5-day pregnant Wt and Tg mice (six animals per group) (***, p < .001). (E): Western blot analysis of phospho-Stat5 and β-casein protein levels in primary mammary epithelial cells from Wt and Tg mice, cultured onto Matrigel, and treated or not with prolactin (Prl) for the indicated days. Actin was used as loading control. Quantification is shown on the right. Bars represent the means ± SEM of three independent experiments (*, p < .05; **, p < .01).

Download figure to PowerPoint

Next, we examined mammary glands of Tg and Wt 12.5-day-pregnant mice for the expression of alveolar differentiation markers. As shown in Figure 3C, the levels of phospho-Stat5 and β-casein were lower in Tg glands relative to controls. Concomitantly, the proportion of luminal progenitor cells (CD61+CD49fLowCD24+/High), which include alveolar precursors [3, 6], was found to be significantly higher in Tg glands than controls (Fig. 3D), thus indicating an accumulation of immature luminal cells in Tg mammary glands. In addition, when primary mammary epithelial cells from Tg mice were cultured onto Matrigel in the presence of prolactin, they showed a marked reduction of both β-casein and phospho-Stat5 compared to control cells (Fig. 3E). These data demonstrate that sustained overexpression of p130Cas impairs mammary lactogenic differentiation.

p130Cas Overexpression Results in Hyperactivation of the Tyrosine Kinase C-Kit

It has recently been shown that the majority of luminal ER– cells express high levels of the c-Kit tyrosine kinase receptor [17]. As p130Cas overexpression in our transgenic mice occurs primarily in luminal ER– cells and previous studies suggested a functional link between p130Cas and c-Kit [34], we investigated the interaction between these two molecules in primary MMECs. First, we assessed the phosphorylation of p130Cas in response to c-Kit activation upon treatment with the SCF. Notably, both c-Kit and p130Cas underwent a biphasic phosphorylation in response to SCF (Fig. 4A). In addition, c-Kit was found to coimmunoprecipitate with p130Cas (Fig. 4B).

thumbnail image

Figure 4. p130Cas overexpression results in hyperphosphorylation of c-Kit in primary mouse mammary epithelial cells. (A): Wt mouse mammary epithelial cells (MMECs) were starved for 12 hours and left unstimulated or stimulated for the indicated times with 100 ng/ml SCF. Total protein extracts were prepared and immunoprecipitated with antibodies (Abs) directed to phospho-tyrosine (phospho-Tyr). Immunoprecipitates were probed with Abs to p130Cas (right upper panel). Total cell extracts were probed with Abs to p130Cas (input, right lower panel), phospho-c-Kit (Y721), and c-Kit (left panels). (B): Total protein extracts from Wt MMECs were immunoprecipitated with Abs to p130Cas or control IgG (Ctr). Immunoprecipitates and total protein extracts (input) were blotted with Abs to c-Kit and p130Cas. (C): MMECs from Wt and MMTV-p130Cas (Tg) mice were starved for 12 hours and left unstimulated or stimulated for the indicated times with 100 ng/ml SCF. Total cell extracts were probed with Abs to phospho-c-Kit (Y721) and c-Kit (upper and lower panel, respectively). Histograms show levels of c-Kit phosphorylation, normalized to total c-Kit protein, for each time point in Wt and Tg MMECs. Bars represent the means ± SEM of three independent experiments (*, p < .05; **, p < .01). Abbreviation: SCF, soluble c-Kit ligand.

Download figure to PowerPoint

Next, c-Kit phosphorylation (Y721) was also evaluated in SCF-treated and untreated MMECs isolated from Tg mice. Interestingly, c-Kit tyrosine phosphorylation was higher in Tg than Wt cells, even in the absence of SCF, indicating that high levels of p130Cas enhance c-Kit phosphorylation independently of its ligand (Fig. 4C). Together, these data highlight that p130Cas and c-Kit may be part of the same signaling network in MMECs, with p130Cas playing a crucial role in the regulation of c-Kit activity.

Constitutive Activation of C-Kit Promotes Basal Cell Commitment and Interferes with Lactogenic Differentiation

Since p130Cas overexpression leads to hyperactivation of c-Kit, we evaluated whether constitutive c-Kit activation could be sufficient by itself to affect mammary cell differentiation. To achieve c-Kit constitutive activation, MMECs were transduced with lentiviruses expressing c-Kit with an activating point mutation (D818V c-Kit). In parallel, MMECs were transduced with lentiviral vectors either empty (Ctr) or carrying p130Cas cDNA (p130Cas). As shown in Figure 5A, cells expressing the D818V c-Kit mutant (D818V c-Kit) as well as p130Cas overexpressing cells (p130Cas) displayed higher levels of c-Kit phosphorylation (Y568/570) than Ctr cells. To determine whether abnormal c-Kit activation alters mammary progenitor cell behavior, D818V c-Kit, Ctr, and p130Cas transduced cells were assayed for their clonogenicity in 3D culture. Unlike Ctr cells, D818V c-Kit and p130Cas cells gave rise mainly to solid rather than acinar colonies (Fig. 5B). Immunostaining for K5 and K18 revealed that solid colonies from D818V c-Kit and p130Cas cells were enriched for K5+ cells compared to acinar colonies of Ctr cultures (Fig. 5B). Consistently, significantly higher levels of K5 and lower levels of K18 were found in protein extracts from D818V c-Kit and p130Cas cells relative to Ctr (Fig. 5C). In sum, like p130Cas overexpression, constitutive c-Kit activation biases mammary progenitor differentiation toward the basal cell lineage. Moreover, similarly to p130Cas cells, D818V c-Kit cells were refractory to lactogenic differentiation, since they expressed lower amounts of β-casein than Ctr cells upon treatment with prolactin (Fig. 5D). In conclusion, these data indicate that constitutive c-Kit activation promotes basal cell commitment and interferes with lactogenic differentiation, thus entirely mirroring the effects of p130Cas overexpression.

thumbnail image

Figure 5. Constitutive c-Kit activation mimics p130Cas overexpression in primary mouse mammary epithelial cells. (A): Mouse mammary epithelial cells (MMECs) were transduced by empty lentiviral vectors (Ctr, left and right panels) or carrying D818V c-Kit cDNA (D818V c-Kit, left panels) or p130Cas cDNA (p130Cas, right panels), seeded onto Matrigel for 3 days and then shifted for 1 day in growth medium without epidermal growth factor and fetal bovine serum. Total protein extracts from Ctr, D818V c-Kit, and p130Cas cells were probed for phospho-c-Kit (Y568/Y570), c-Kit, and p130Cas expression. Actin was used as loading control. (B): Representative phase-contrast images of colonies arising from three-dimensional (3D) culture of Ctr, D818V c-Kit, and p130Cas transduced MMECs (upper panels). Scale bar = 100 μm. Stacked bar graphs on the left show the percentage of acinar and solid colonies. Graphs represent data from three independent experiments. Representative images of immunofluorescence staining with antibodies to keratin 5 (green) and keratin 18 (red) of sections of colonies arising from 3D cultures of Ctr, D818V c-Kit, and p130Cas transduced MMECs (lower panels). Nuclei were stained with DAPI (blue). Scale bar = 50 μm (**, p < .01). (C): Total protein extracts from colonies, recovered from 3D cultures of Ctr, D818V c-Kit, p130Cas transduced MMECs, were probed for keratin 5 and keratin 18 expression levels. GAPDH was used as loading control. Quantification of data from three independent experiments is shown on the right. (*, p < .05). (D): Ctr, D818V c-Kit, and p130Cas transduced mammary cells, cultured onto Matrigel, were treated with prolactin (Prl) for 2 days. Beta-casein protein expression was determined by Western blot analysis. Actin was used as loading control. Quantification of results from three independent experiments is shown on the right (*, p < .05; **, p < .01).

Download figure to PowerPoint

C-Kit Downregulation Re-establishes Proper Differentiation of p130Cas Overexpressing Mammary Epithelial Cells

To directly address whether c-Kit mediates the effects of p130Cas overexpression, we investigated whether c-Kit downregulation was sufficient to re-establish proper differentiation in p130Cas overexpressing cells. MMECs isolated from Tg mice were transduced to express a short hairpin RNA against c-Kit (shRNA c-Kit) or a scramble shRNA (shRNA Scr). In parallel, Wt cells were infected with shRNA Scr lentiviral vectors. It has recently been reported that c-Kit silencing results in massive apoptosis of MMECs in vitro [17]. As shown in Figure 6A, c-Kit expression was reduced by approximately 40% in Tg shRNA c-Kit compared to Tg shRNA Scr cells. Notably, the extent of this downregulation was sufficient to significantly reduce c-Kit phosphorylation (Y568/570) (Fig. 6A), without compromising cell viability. Importantly, in 3D CFC assays, Tg shRNA c-Kit cells, unlike Tg shRNA Scr cells, gave raise to acinar and solid colonies in a proportion resembling that of Wt shRNA Scr colonies (Fig. 6B). Moreover, colonies from Tg shRNA c-Kit cells were more similar to Wt shRNA Scr than to Tg shRNA Scr colonies for their content in both K18 and K5 (Fig. 6C). In addition, Tg shRNA c-Kit cells behaved like Wt shRNA Scr cells in response to prolactin, as they expressed higher levels of β-casein compared to Tg shRNA Scr cells (Fig. 6D). These results demonstrate that c-Kit hyperactivation is required for p130Cas overexpression to promote a basal differentiation pattern and to oppose lactogenic differentiation in primary mammary epithelial cells.

thumbnail image

Figure 6. c-Kit downregulation re-establishes proper differentiation of Mouse Mammary Tumor Virus (MMTV)-p130Cas primary mammary epithelial cells. (A): Mouse mammary epithelial cells (MMECs) transduced with lentiviral vectors carrying a scramble shRNA (Scr shRNA) and in Tg mammary cells transduced with vectors carrying a shRNA against c-Kit (c-Kit shRNA) were cultured as described in Figure 5A. Phospho-c-Kit (Y568/Y570) and c-Kit protein expression in Wt and Tg are shown. Actin was used as loading control. Histograms show c-Kit phosphorylation (middle panel) and total c-Kit (right panel) levels, normalized to actin. Bars represent the means ± SEM of three independent experiments (*, p < .05; **, p < .01). (B): Representative phase-contrast images of colonies arising from three-dimensional (3D) culture of Wt Scr shRNA, Tg Scr shRNA, and Tg c-Kit shRNA transduced MMECs. Stacked bar graphs show the percentage of acinar and solid colonies. Graphs represent data from three independent experiments (*, p < .05; **, p < .01). Scale bar = 100 μm. (C): Total protein extracts from colonies, recovered from 3D cultures of Wt Scr shRNA, Tg Scr shRNA, and Tg c-Kit shRNA transduced mammary cells, were probed for keratin 5 and keratin 18 expression levels. GAPDH was used as loading control. Graphs represent data from three independent experiments (*, p < .05). (D): Wt Scr shRNA, Tg Scr shRNA, and Tg c-Kit shRNA transduced MMECs, cultured onto Matrigel, were treated with prolactin (Prl) for 2 days. Beta-casein protein expression is shown. Actin was used as loading control. Quantification of data from three independent experiments is shown (*, p < .05; **, p < .01).

Download figure to PowerPoint

p130Cas Is Upregulated in Human TN Breast Cancer

Increasing evidence supports the hypothesis that defective differentiation of mammary luminal progenitors predisposes to BLBC [13, 15]. Since our analyses of MMTV-p130Cas mice showed that p130Cas overexpression impairs the differentiation of mammary luminal progenitors, we tested whether p130Cas upregulation might occur in human TN breast cancers, which are commonly associated with a basal-like phenotype [12]. Immunohistochemical analysis was performed on 51 human TN breast cancers and 5 normal breast tissue samples from reduction mammoplasties (Fig. 7A–7I). Forty-four tumor samples were positive for p130Cas expression. As shown in the table in Figure 7K, p130Cas staining intensity was scored as low (1+) in 5 samples, intermediate (2+) in 21 samples, and high (3+) in 18 samples. Moreover, in approximately 91% of positive samples, more than 50% of cells was positive for p130Cas staining. The pattern of p130Cas staining in normal breast tissues was not homogeneous and the intensity ranged from 0 to 1+ (absent to low). Specifically, p130Cas was nearly undetectable in the ductal mammary epithelium, while the lobular epithelium displayed a low intensity of p130Cas staining with few cells more strongly stained. These results indicate that p130Cas is overexpressed in approximately 75% of the TN breast cancers analyzed, thus further indicating that p130Cas upregulation might be a priming event in the development of BLBC.

thumbnail image

Figure 7. p130Cas expression in triple-negative (TN) breast cancer. (A--C): Representative images of normal breast tissue obtained from reduction mammoplasties (N = 5). Staining intensity ranges from absent in ductal epithelium to low (+1) in lobular epithelium. A few cells in the lobules are more intensively stained (+2). (D--I): Representative images of TN breast cancer with low (1+, D and G), intermediate (2+, E and H), and high (3+, F and I) intensity of p130Cas staining. Magnification, ×20, panels A and B; ×10, panels C, D, E, and F; ×40 panels G, H, and I. Scale bar = 65 μm. (K): Classification of the TN breast cancer samples based on p130Cas staining intensity.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Here, we demonstrate that p130Cas overexpression boosts mammary luminal progenitor cell plasticity, leading to a shift from luminal to basal commitment and impairing lactogenic differentiation. We also show that these alterations depend on the hyperactivation of the tyrosine kinase c-Kit induced by high levels of p130Cas.

Defective luminal differentiation has previously been reported as a consequence of altered expression of several cell-surface receptors [35–37] or transcription factors [38–40]. In this study, we show that upregulation of the adaptor protein p130Cas in luminal progenitors is sufficient to promote signaling in support of a basal differentiation program.

In particular, the switch in the differentiation pattern observed in the mammary progenitors of MMTV-p130Cas mice might rely on the ability of p130Cas to upregulate the transcription factor p63, specifically the deltaNp63 isoform, which is reported as peculiar to the basal compartment in the mammary epithelium [31–33]. Indeed, in primary mammary epithelial cells, maintenance of basal cell characteristics depends on deltaNp63 expression, while ectopic deltaNp63 expression in luminal cells is enough to confer a basal phenotype [31]. Moreover, other studies show that deltaNp63 directly induces the expression of basal-restricted keratins [41].

Additionally, an intrinsic characteristic of basal cells is the expression of several genes related to epithelial-mesenchymal transition (EMT) [8, 42], whereas the maintenance of the luminal cell identity is supposed to result from homotypic cell-cell contacts [29]. Interestingly, recent reports indicate that elevated expression of p130Cas may contribute to the EMT program, by promoting the expression of typical mesenchymal genes and the disassembly of cell-cell junctions [43, 44], further supporting a possible role for p130Cas in the transition from luminal to basal cell phenotype.

Luminal progenitors are characterized by a distinctive molecular profile compared to other cell subsets in the mammary epithelium [8]. In particular, the tyrosine kinase c-Kit has been identified as a marker of these progenitors [13, 17] but the functional significance of c-Kit expression in the mammary gland has been only partially understood. In this regard, a recent study revealed that c-Kit is required for the survival and growth of mammary progenitor cells in vitro [17].

In primary MMECs, p130Cas associates with c-Kit and is phosphorylated in response to SCF (the c-Kit ligand). Remarkably, mammary epithelial cells overexpressing p130Cas exhibit high levels of c-Kit phosphorylation, even in the absence of its soluble ligand. As for other tyrosine kinase receptors, dimerization subsequent to ligand binding must occur in order to induce autophosphorylation of c-Kit [45]. Given the ability of p130Cas to form oligomers as well as to complex with c-Kit, it is tempting to speculate that ligand-independent c-Kit activation, in the context of p130Cas overexpression, may be due to constitutive receptor dimerization resulting from intracellular clustering of p130Cas molecules. Alternatively, the involvement of other proteins might explain an indirect induction of c-Kit activity following p130Cas overexpression. In this regard, members of the Src family kinase as well as the phosphatase Shp2 represent some of the best candidates, given their ability to bind p130Cas [19] and to regulate c-Kit phosphorylation status [46].

Our data demonstrate that abnormal activation of c-Kit in primary mammary epithelial cells promotes differentiation toward the basal cell lineage and interferes with lactogenic differentiation. The consequences of c-Kit overexpression, but not specifically of c-Kit hyperactivation, on mammary cell differentiation have been, in part, investigated previously. In vitro studies in the HC11 cell line indicate that c-Kit overexpression does not affect differentiation [16], but do not provide any evidence that high levels of c-Kit indeed result in a higher activity of this receptor. Conversely, in line with our observations, in vivo transplantation experiments with c-Kit overexpressing primary mammary cells suggest that c-Kit overexpression may cause a block of differentiation [17].

Finally, we show that upregulation of p130Cas occurs in human TN breast cancers, which are reported to be commonly associated with a basal-like phenotype [12]. Accumulating evidence indicates that BLBCs originate from mammary luminal progenitor cells [13, 14]. Although several genetic aberrations, such as loss of PTEN, p53 mutations, or Brca1 inactivation, have been described as more frequently associated with this breast cancer subtype [47, 48], the molecular pathogenesis of the majority of BLBCs remains to be elucidated.

Noteworthily, BRCA1 germline mutations, which predispose to BLBC, are associated with distinctive cell alterations in the mammary epithelium even before the onset of cancer. Specifically, it has been described that disease-free breast tissue from BRCA1mut/+ patients is characterized by luminal progenitors displaying basal cell features and a block in maturation [13, 15]. Similar differentiation defects were observed in Brca1-deficient mice [16]. Hence, the existence of an aberrant luminal cell population seems to be a prerequisite for the development of BLBCs. Interestingly, we have observed that p130Cas overexpression causes defects in the differentiation of mammary luminal progenitors, closely resembling those associated with Brca1 deficiency. However, while lack of Brca1 may accelerate tumorigenesis by promoting accumulation of genetic alterations [48], p130Cas overexpression by itself is not sufficient to drive mammary cell transformation [21]. Furthermore, p130Cas and Brca1 are both linked to the luminal progenitor marker c-Kit, with p130Cas overexpression resulting in abnormal activation of c-Kit and Brca1 deficiency inducing c-Kit expression in mammary epithelial cells. Although c-Kit expression has been found very rarely in nonbasal-like breast tumors, only a fraction of BLBCs exhibit c-Kit expression [13, 49, 50]. Thus, we can speculate that deregulation of this receptor might contribute to early perturbations in the luminal progenitor population rather than to maintenance of BLBCs. In conclusion, our data indicate that p130Cas upregulation may serve to prime the mammary epithelium to the development of BLBCs, adding a new piece of information on the possible mechanisms underlying the onset of this aggressive breast cancer subtype.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

This study reveals for the first time that the expression of p130Cas adaptor is characterized by a precise spatial and temporal regulation in the mammary gland epithelium. Moreover, the regulation of p130Cas levels is crucial for maintaining cellular homeostasis within the mammary epithelium. Specifically, p130Cas overexpression in mammary luminal progenitors, through abnormal activation of the tyrosine kinase c-Kit, biases commitment toward the basal cell fate and opposes differentiation in response to lactogenic stimuli. Nevertheless, p130Cas upregulation occurs in human TN breast cancers. In conclusion, based on the emerging hypothesis on the cell of origin of BLBC, we propose that high levels of p130Cas may render mammary luminal progenitor cells prone to originating BLBC.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

We would like to thank C. Brisken (EPFL, Lausanne, Switzerland) and J. Stingl (CR-UK CRI, Cambridge, U.K.) for devices to isolate and analyze primary mouse mammary epithelial cells, C.J. Watson (University of Cambridge, U.K.) for kindly providing KIM2 cells, E. Santopietro and L. Conti (MBC, Turin) for technical assistance, V. Poli (MBC, Turin), E. Hirsch (MBC, Turin), and P. P. Pandolfi (Harvard Medical School, Boston, MA) for critical reading of the manuscript. This work was supported by AIRC (Grants IG2008 and IG2011), MIUR (FIRB Giovani 2008), PiSTEM, and Progetto Sanità Finalizzata. G. Tornillo and B. Bisaro are supported by FIRC fellowships.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSION
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
sc-12-1049_sm_SupplFigure1.tif736KSupplemental Figure 1. Expression of keratin 18, alpha-SMA and ERalpha in sorted cells. (A) Histograms show keratin 18 (left panel) and alpha-SMA (right panel) mRNA relative expression in sorted luminal and basal cells, as assessed by qRT-PCR analysis. (B) Histograms show ERalpha mRNA relative expression in unsorted and sorted luminal CD24+/High Sca-1- and CD24+/High Sca-1+ cells, as assessed by qRT-PCR analysis.
sc-12-1049_sm_SupplFigure2.tif200KSupplemental Figure 2. Primers for detection of MMTV-p130Cas transgene transcript. (A) RT-PCR showing p130Cas transgene (myc-p130Cas) expression. Total cDNA prepared from Wt and Tg mammary cells was used as template. Beta-actin expression is also included.
sc-12-1049_sm_SupplFigure3.tif229KSupplemental Figure 3. p130Cas overexpression alters the differentiation potential of ER- luminal progenitor cells. (A) Phase contrast images of colonies arising from 3D culture of luminal ER- (CD24+/High Sca-1-) cells purified from Wt (left panel) or Tg (middle panel) mammary glands. Images of colonies derived from basal (CD24+/Low Sca-1-) cells isolated from Wt glands are shown in the right panels. Scale bar, 100 micron. Images are representative of three independent experiments. (B) Representative images of immunofluorescence staining with antibodies to keratin 5 (green, K5) and keratin 18 (red, K18) of colonies arising from sorted Wt CD24+/High Sca-1- (top panels), Tg CD24+/High Sca-1- (middle panels) and Wt (CD24+/Low Sca-1-) basal cells (bottom panels). Nuclei were stained with DAPI (blue). Scale bar, 50 micron.
sc-12-1049_sm_SupplFigure4.tif2202KSupplemental Figure 4. p130Cas is downregulated during differentiation of KIM-2 cells. (A) KIM-2 cells were treated with prolactin and dexamethasone for the number of days indicated and probed for p130Cas and beta-casein mRNA expression (left and right panel, respectively) by qRT-PCR analysis. (B) KIM- 2 cells, treated as above, were analyzed for p130Cas protein expression by immunoblotting. Actin was used as loading control. Bars represent the means ± SEM of three independent experiments.
sc-12-1049_sm_SupplData.pdf26KSupplementary Data

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.