Establishment, growth and in vivo differentiation of a ... - Springer Link

Breast Cancer Res Treat (2007) 103:247–257 DOI 10.1007/s10549-006-9358-x

PRECLINICAL STUDY

Establishment, growth and in vivo differentiation of a new clonal human cell line, EM-G3, derived from breast cancer progenitors Marketa Brozova Æ Zdenek Kleibl Æ Irena Netikova Æ Jan Sevcik Æ Eva Scholzova Æ Jana Brezinova Æ Alena Chaloupkova Æ Pavel Vesely Æ Pavel Dundr Æ Marie Zadinova Æ Luboslava Krasna Æ Eva Matouskova

Received: 13 April 2006 / Accepted: 31 July 2006 / Published online: 25 October 2006 Springer Science+Business Media B.V. 2006

Abstract A new clonal cell line, EM-G3, was derived from a primary lesion of human infiltrating ductal breast carcinoma. The line consisted of cuboidal cells with occasional appearance of more differentiated branched cells apparently involved in cell-to-cell communication. The EM-G3 cells, population doubling time 34 h, are dependent on the epidermal growth factor. Multicolor fluorescence in situ hybridization (mFISH) analysis demonstrated a stable diploid genome with several genetic changes. Immunocytochemical analysis of EM-G3 M. Brozova Æ A. Chaloupkova Æ P. Vesely Æ L. Krasna Æ E. Matouskova (&) Department of Cell Biology, Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, Flemingovo nam. 2, 166 37 Prague 6, Czech Republic e-mail: [email protected] Z. Kleibl Æ J. Sevcik Æ E. Scholzova First Faculty of Medicine, Institute of Biochemistry and Experimental Oncology, Charles University in Prague, Prague, Czech Republic I. Netikova General Faculty Hospital in Prague, Laboratory of Cytostatics, Prague, Czech Republic J. Brezinova Institute of Hematology and Blood Transfusion, Prague, Czech Republic P. Dundr First Faculty of Medicine, Department of Pathology, Charles University in Prague, Prague Czech Republic M. Zadinova First Faculty of Medicine, Institute of Biophysics and Informatics, Charles University in Prague, Prague, Czech Republic

in vitro revealed positivity for keratins (K) K5, K14, K18, nuclear protein p63, epithelial membrane antigen (EMA) and other proteins indicative of a pattern of mammary epithelium bipotent progenitors. Detection of integrins a-6, b-1, and protein CD44 by cDNA array also pointed to the character of basal/stem cells. In contrast, dominant cells in the human original tumor showed the luminal character (K18+, K19+, K5–, K14–, and p63–). However, cells with the immunocytochemical profile similar to that of cultured EM-G3 cells were found in minor clusters in the patient’s tumor sections. The EM-G3 cells formed limited tumors in nu/nu mice. The cells in mouse tumors were organized in primitive ductal-like structures consisting of 1–3 large central luminal-like cells (EMA+) surrounded by peripheral myoepithelial-like cells (p63+/EMA–). The large central cells gradually disintegrated, forming a pseudolumen. Apparently, EM-G3 cells are able to partially differentiate in vivo as well as in vitro. Our results indicate that EM-G3 cells were derived from a premalignant population of common progenitors of luminal and myoepithelial cells that were immortalized in an early stage of tumorigenesis. Keywords Breast cancer Æ Cell line Æ Progenitor cells Æ Differentiation Æ Immunophenotyping

Introduction A substantial part of knowledge about breast cancer (BC) biology is based on studies of permanent BC cell lines that provide an unlimited source of homogenous self-replicating material. About 100 permanent cell lines have been obtained since the first line BT20 was

123

248

established [1, 2]. Most of them are of metastatic origin and, therefore, represent cell populations of the late stage of tumor progression [2]. More than two-thirds of all studies reporting BC cell lines were performed on only three lines derived from pleural effusions, i.e. late stage metastases: MCF7, T-47D, and MDA-MB-231. Permanent cell lines often show chromosome instability and significant genomic rearrangements [3]. The BC cell lines from primary solid tumors are rare [4, 5]. The mammary gland ductal and lobular system consists of two major epithelial cell types: (1) Luminal cells form a secretory layer producing milk in the acini. (2) Myoepithelial cells, shielding the luminal cell compartment, provide contractile apparatus for milk transport during breast-feeding. Epithelial cell types are distinguished by the immunocytochemical markers they express. The most reliable markers of mammary epithelial cells are keratins (K; proteins forming intermediate filaments in epithelial cells) and some other proteins, such as p63, epithelial membrane antigen (EMA), epithelial specific antigen (ESA), or smooth muscle actin (SMA) [6]. Luminal cells are characterized by expression of K7, K8, K18, K19, and EMA, whereas the majority of myoepithelial cells express K5, K14, p63, and SMA. Cells of invasive carcinomas usually display the differentiated luminal epithelial phenotype K18+/K19+. Luminal and myoepithelial cell lineages are believed to share common progenitors occurring at various stages of their differentiation in the mammary epithelium [7–12]. These progenitors are referred to as basal epithelial cells, which are located between basal/suprabasal and luminal cell layers in the proximity of the basement membrane [13, 14]. Progenitor cells are in high proportion also present in terminal duct lobular units (TDLU), the most mitotically active parts of the mammary gland [14, 15]. Stem cells or progenitor cells in TDLU are supposed to be targets of malignant transformation as most neoplastic breast lesion arise in this particular site [16, 17]. During the past years, successful isolation, cultivation and characterization of mammary stem/progenitor cells made it possible to trace differentiation of breast cells in vitro [8, 12, 15] and to identify tumor-initiating cells [18–20]. Petersen et al. isolated and subsequently immortalized K19+ suprabasal putative precursor cells from the luminal epithelial compartment [17]. These cells were able to form TDLU-like structures within the reconstituted basement membrane. We previously succeeded in selective expansion of mammary luminal epithelial cells using the optimized 3T3 feeder layer technique [21] and in development of a number of primary cell lines with a limited lifespan

123

Breast Cancer Res Treat (2007) 103:247–257

from individual breast carcinomas [22]. Here, we report the establishment and characterization of a K19–/K18+/p63+ clonal cell line, EM-G3, which differ from dominant cells of the original breast tumor. The cells of this line are able to form primitive ductal-like structures after subcutaneous application into nu/nu mice demonstrating a differentiation potential. The reported results indicate that EM-G3 cells originated from a common progenitor of luminal and myoepithelial cells immortalized in an early stage of BC development. Materials and methods Patient characteristics, tumor sample The patient’s informed consent approved by the Ethical committee was obtained prior to surgery. The sample was obtained in the year 2001 from a 70-year-old woman who underwent lumpectomy at the General Faculty Hospital in Prague. Histological verification of the tumor revealed infiltrating ductal breast carcinoma grade II, pT1cN0M0 with microcalcifications, fibrogenesis, and intraductal propagation. Estrogen receptor (ER) and progesterone receptor (PR) positivity was assessed at 85 and 90%, respectively, membrane positivity for c-erbB-2 was present in about 80% of cancer cells. In the year 1996 the patient underwent radical mastectomy for BC in the contralateral breast. The patient is still alive 10, respectively, 5 years after BC diagnoses (2006). Specimen dissociation and cell cultivation Primary epithelial cells, designated EM, were isolated and cultured using the 3T3 feeder layer technique as described previously [22]. Briefly, the tumor tissue (0.3 cm3) was cut to pieces and digested in 0.05% collagenase A (Roche, Penzberg, Germany) at 37C for 5 h with shaking. The obtained cells were seeded for 3 h into the first tissue culture flask to remove rapidly attaching fibroblasts. Then, the medium with unattached cells was transferred to another 25 cm2 tissue culture flask containing 1-day-old feeder layer of lethally irradiated NIH 3T3 cells (100 Gy; 2.5 · 104 cells/cm2). The medium used for epithelial cell cultivation was H-MEM supplemented with nonessential amino acids, 0.5 lg/ml hydrocortisone, 5 lg/ml insulin, 10–10 M cholera toxin, 5 ng/ml epidermal growth factor (EGF), 10 bovine and 2% fetal bovine serum, penicillin, and streptomycin. In the tenth passage 3T3 feeder cells were discontinued. Eighteen months after the primary culture the 100th passage was reached.


The EM cells were cultured up to passage 130, and then cloned in a 96-well-multidish by seeding two cells per well. After 24 h, the wells containing only one cell were selected. Satisfactorily growing single cell colonies were transferred to 35 mm tissue culture dishes and clones were further passaged. One of them, EM-G3, was chosen for more extensive studies. Tumorigenicity in nu/nu female mice Animal experiments were approved by the faculty Committee for animal care. EM-G3 cells at passages 141 and 195 were subcutaneously injected together with Matrigel (BD Biosciences, San Jose, USA) into eight immunodeficient CD1 nu/nu female mice. Nine million of cells in 0.1 ml medium and 0.05 ml Matrigel per mouse were applied. Mice were observed for 3 months. The cells from tumors were seeded back to the culture 1 and 3 months post injection. Immunocytochemistry Cultured EM-G3 cells and paraffin sections of the original and xenograft tumors were immunocytochemically analyzed as described previously using monoclonal antibodies against K5, K14, EGF receptor, SMA (Novocastra, New Castle, UK), K18, K19, EMA, ESA (DAKO, Hamburg, Germany), ER (Sigma, Deisenhofen, Germany), PR (Immunotech, Heidelberg, Germany), and p63 (BD Biosciences) [22]. Fluorescence in situ hybridization for HER2/neu, cyclin D1, c-myc, Rb, p53 PathVysion DNA probe kits for HER2/neu, cyclin D1, c-myc, RB1, and p53 genes (Abbott-Vysis, Downers grove, IL, USA) were used for FISH analysis of these genes in EM-G3 cells according to the manufacturer’s instructions. PathVysion Her-2 DNA Probe Kit (Abbott-Vysis) was used for FISH analysis in paraffin sections of the original patient’s tumor sample as described previously [23]. Multicolor fluorescence in situ hybridization The EM-G3 cells in the exponential growth phase at passages 142 and 212 were treated by 0.05% colcemid for 5 h at 37C. The cells were harvested by trypsinEDTA, centrifuged, hypotonized by 0.075 M KCl and fixed by methanol : acetic acid (3:1). For mFISH and multicolor banding (mBAND) 24 XCyte MetaSystems 24 color whole chromosome painting probe kit and

249

XCyte 12 MetaSystems mBAND kit for human chromosome 12 (Meta Systems, Watertown, MA, USA) were used as described previously [24]. cDNA array GEArray Q series Human Cancer PathwayFinder array kit was obtained from SuperArray Inc. (see www.superarray.com for details). Total RNA (3 lg) from EM-G3 cells was used as a template to generate 32 P-dCTP (Amersham Biosciences, Piscataway, NJ, USA) labeled cDNA probes, according to the manufacturer’s instructions. The cDNA probes were denatured and hybridized overnight at 60C with the SuperArray membrane. The membrane was washed and analyzed by scanning the Kodak BioMax X-Omat X-ray film (Amersham Biosciences). Mutation analysis of the p53, PIK3CA, BRAF and K-Ras in EM-G3 cells Genomic DNA from 2 · 106 EM-G3 cells in passage 182 was isolated using Wizard DNA isolation Kit (Promega, Mannheim, Germany). The entire p53 (OMIM 191170) coding sequence (exons 2–11) including intron–exon junctions was PCR amplified in nine fragments (primers and PCR conditions available upon request). The exons with mutation hot spots of the phosphatidylinositol-3kinase catalytic subunit a (PIK3CA—OMIM 171834; exons 7, 9, 20), BRAF (OMIM 164775; exons 11, 14) and K-Ras (OMIM 190070; exon 2) genes were amplified as described previously [25, 26]. The PCR products of the analyzed genes were sequenced using BigDye Terminator Version 3.1 on ABI310 sequencer (Applied Biosystems, Foster City, CA, USA).

Results Establishment of EM-G3 cells The first passage of epithelial cells was performed 4 weeks after primary cell seeding. From the third passage on, the epithelial cells started to grow quickly in the presence of feeders. Starting from passage 10, the feeders were omitted and the parental EM cells (see materials and methods) reached a stable population doubling time of about 34 h. The clone EM-G3 was selected from parental EM cells at the 130th passage. Currently, over 215 passages of the cell line EMG3 (including the EM precloning passages) have been performed.

123

250

From the primary culture, the EM cells and the EM-G3 clone were dependent on the presence of EGF and other growth factors (insulin, hydrocortisone, cholera toxin) in the medium. Without EGF, the cell growth was inhibited and in several days the cells started to enlarge, flatten, disintegrate, and eventually died. Morphological examinations The EM-G3 line demonstrated a typical epithelial morphology of cuboidal or polygonal cells growing in monolayer. Occasionally, individual cells in the culture (~1/150 cells) changed to a more differentiated branched cell with marked protrusions displaying a number of cell-to-cell contacts (Fig. 1a and b). Tumorigenicity in nu/nu female mice Subcutaneous injections of EM-G3 cells into nu/nu mice resulted in tumor development in seven out of eight mice in the course of a 2-week period. In 4 weeks tumors reached a size of about 30–50 mm3, stopped growing and remained stable for the next 2 months. Histological examinations revealed spherical tumors covered by a fibrous capsule (Fig. 2). Islands of EM-G3 were organized in primitive ductal-like structures characterized by 1–3 large central epithelial cells expressing luminal markers, surrounded by smaller epithelial cells expressing myoepithelial markers (see immunocytochemical analysis and Fig. 3l, t, u, v, and x). The central cells gradually disintegrated, resulting in formation of a pseudolumen in the cell island (Fig. 3v and x). One, respectively, 3 months after EM-G3 injection, the cells from the mouse tumor nodule were seeded back to the culture, using collagenase dissociation and the 3T3 cells as feeders. Interestingly, collagenase treatment resulted in isolation of spherical cellular clusters (Fig. 3y) instead of single cell suspension. Cells seeded after 1-month growth in mice migrated out off spherical clusters and proliferated intensively. The morphology of EM-G3 was preserved

Fig. 1 Morphology of EMG3 cells in vitro. (a) Six-dayold colony of cultured cells. (b) Confluent culture. Note the large cell types with long protrusions (arrows), seemingly involved in cell-tocell communication witnessing for cell differentiation (phase contrast; scale bars = 20 lm)

123


(Fig. 3y). However, the cells isolated 3 months after the injection survived in the culture in form of spherical clusters without any apparent outgrowth of cells. Immunocytochemical analysis In vitro cultured EM-G3 cells, paraffin sections of the original tumor specimen, and EM-G3 nu/nu mice tumors were analyzed using a panel of antibodies (Table 1). EM-G3 in vitro The mammary epithelial character of EM-G3 cells in vitro was proved by positivity for a luminal marker K18 and for basal/myoepithelial markers K5 (Fig. 3c), K14, and p63 (Fig. 3k). Cultured cells were negative for the most typical marker of fully differentiated luminal cells K19 (Fig. 3g) and for a marker of mature myoepithelial cells, SMA. Strong membrane positivity for EGFR was determined, whereas c-erb-B2 was negative. Weak to medium membrane and/or cytoplasmic (but not nuclear) positivity was detected for ER (Fig. 3o), PR, EMA and ESA (Table 1). Original tumor Immunohistochemical analysis revealed two different cellular populations in the patient’s tumor sample: (1) The major tumor cell population (Table 1) displayed immunomarkers typical for fully differentiated luminal cells. They were K5– (Fig. 3a) and K14–, while K18+ and K19+ (Fig. 3e); most of the cells were EMA+, c-erb-B2+, SMA– and p63– (Fig. 3i). Nuclei of these tumor cells in paraffin sections were ER+ (Fig. 3m) and PR+. (2) The minor cellular regions were immunohistochemically highly similar to EM-G3 cells. These cell islands were K5+ (Fig. 3b), K14+, K19– (Fig. 3f), p63+ (Fig. 3j), and ER– (Fig. 3n). A pattern similar to EM-G3 was also displayed by cells forming TDLU randomly interspersed within the tumor specimen (Fig. 3 q, r, and s).


251

Fig. 2 Tumor formation in nu/nu mice by EM-G3. (a) External view on xenografted tumor in a nu/nu mouse 4 weeks following EM-G3 injection. (b) Appearance of encapsulated tumor nodule growing subcutaneously without any signs of invasiveness

EM-G3 in nu/nu mouse tumors Immunohistochemical analysis of xenografted EM-G3 cells in nu/nu tumors revealed a profile identical to the EM-G3 cells in vitro: K5+ (Fig. 3d and v), K14+ (Fig. 3u), p63+ (Fig. 3l), K19– (Fig. 3h), ER– (Fig. 3p), PR–, EGFR+ (Fig. 3x), and SMA–. Expression of K18 was very weak (Table 1). Interestingly, the xenografted cells, though of clonal origin, were mostly organized in islands composed of two cell types: large central cells expressing luminal markers: EMA (Fig. 3t) and ESA; and peripheral cells expressing myoepithelial markers: p63 (Fig. 3l), K14 (Fig. 3u), and others. Only K5 remained expressed in all cells, central and peripheral (Fig. 3d), during the entire process of differentiation. The large central cells gradually disintegrated, leaving a hole in the colony center (Fig. 3v and x, partially also 3u). cDNA array Analysis of mRNA with cDNA arrays (Fig. 4) showed high expression levels of apoptosis activator Apaf1, membrane proteins CD44, and TGFR1b, integrin family members a2, a3, a4, a6, aV, and b1, p53 regulator MDM2, p21Ras GTP-ase activating molecule RasA1, and extracellular matrix-associated protein thrombospondin I. Expression of EGFR confirmed the immunocytochemical results, and expression of integrins a6/b1 and CD44 gave evidence of the stem/progenitor character of the EM-G3 cells [8]. Array data were submitted to the GEO database (http:// www.ncbi.nlm.nih.gov/geo/) under accession number GSM81460. Karyology The mFISH allows simultaneous visualization of chromosomes, complex chromosomal rearrangements, and all numerical aberrations in a single hybridization

experiment. Prior to mFISH, karyology was optimized (different conditions of colcemid treatment, Giemsa staining). Fifteen metaphase spreads at passage 142 were examined, in each of them complex karyotype with derivative chromosomes 3, 6, 12, and 13 and monosomy of chromosomes 15 and 20 were detected. Different chromosomes were randomly lost. Karyotype was evaluated as 44,XX,der(3)t(3;9)(q26;?), der(6)t(6;15)(p21;q12), der(12)t(12;12)(p13.2;q?)dup(12) (q21.2q21.3)dup(12)(q21.2q21.3),der(13)t(13;20)(p11;?), –15,–20 (Fig. 3z). Rearrangement of chromosome 12 was revised using mBAND for chromosome 12 and duplication of the long arms on derivative chromosome 12 was confirmed. The same results were found in the passage 212 when control mFISH was performed to prove the stability of the EM-G3 karyotype. In 15 examined metaphases only one extra aberration was found—derivative chromosome 9, originated as a result of deletion of both short and long arms of chromosome 9—der(9)del(9)(p13)del(9)(q21). All other chromosomal rearrangements, described in the passage 142 were present. Fluorescence in situ hybridization for HER2/neu, cyclin D1, c-myc, Rb, and p53 genes Amplifications of oncogenes (HER2/neu, cyclin D1, cmyc) and deletions of tumor suppressor genes (Rb, p53) are frequent events in human BC. Using FISH, 20 metaphases and 200 interphase nuclei were analyzed for each gene in cultured EM-G3 cells. We did not find any deletions or amplifications of HER2/neu, cyclin D1, c-myc, p53, and Rb genes. Two specimens of paraffin sections from the original patient’s tumor sample were analyzed for HER2/neu amplification. In both sections amplification of HER2/neu (compared to centromeric probe of chromosome 17) was 4.1 and 3.9, respectively (eight signals on average). However, the cells with two HER2/neu signals were found at the edges of the specimens.

123

252

123


Breast Cancer Res Treat (2007) 103:247–257 b Fig. 3 Immunocytochemical analysis of the original tumor

sample, EM-G3 cells in vitro and differentiating EM-G3 in vivo. Outgrowth of EM-G3 from the nu/nu tumor. Karyotype of EMG3 cells. a–d immunostaining for K5. The majority of cells in the patient’s tumor sample were K5– (a), ‘minor islands’ of irregular cells (b, arrow) and TDLUs inside the tumor sample (q) were K5+, EM-G3 cells were K5+ in vitro (c) as well in nu/nu mice (d). e–h immunostaining for K19. The majority of cells in the patient’s tumor were K19+ (e), the minority were K19– (f, arrow), cultured EM-G3 cells (g), and EM-G3 cells in nu/nu mouse (h) were K19–. i–l immunostaining for p63. The majority of cells in the patient’s tumor showed nuclear p63– (i), infrequent p63+ irregular cells were located at the edges of invasive tumor islands (j, arrows), all cultured cells in vitro were p63+ (k), EM-G3 cells in nu/nu mice were organized in primitive ductal-like structures with p63+ in the peripheral cells and p63– in the central luminallike cells (l). m–p immunostaining for ER. The majority of cells in the tumor showed ER+ nuclear staining typical for differentiated glandular-luminal cells (m), the minority of cells in the tumor sample were ER– (n, arrows), cultured EM-G3 cells were ER– except for cytoplasmic or membrane expression (o), islands of EM-G3 cells in nu/nu mice were ER– (p). q–s, immunostaining of TDLUs in the patient’s tumor sample. Staining with antibodies to K5 (q) and SMA (r) in serial sections showed basal/progenitor epithelial cells (K5+/SMA–), (s) p63 immunostaining showed positive nuclei in TDLU cells (also typical for basal/progenitor cells), arrowheads show individual p63– cancer cells scattered around TDLUs. t–x, immunostaining of differentiating EM-G3 cells in nu/nu tumors. Staining for EMA antibody (t) showed large central EMA+ luminal-like cells surrounded by smaller peripheral myoepithelial-like cells (EMA–) in developing primitive ductal structures. Staining for K14 (u) revealed strong expression in peripheral myoepithelial-like cells, lower expression of this antigen in two luminal-like cells (arrows) and an early pseudolumen on the right site of this colony (arrowhead). Staining for K5 (v) showed example of a colony in which the central luminal-like cells disintegrated, forming a pseudolumen. Staining for EGFR (x) revealed strong membrane positivity expressed in intercellular contacts. (y) Outgrowth of EM-G3 cells from a ‘mammosphere’ (arrow) obtained by enzymatic dissociation of the nu/nu tumor. (z) mFISH karyotype of cultured cells EM-G3 showed stable diploid character 44, XX, with monosomy of chromosomes 15 and 20, and translocations on chromosomes 3(9), 6(15), 12(12), and 13(20). Altered chromosomes are marked by arrows. Scale bars = 20 lm

Mutation analysis of the p53, PIK3CA, BRAF, and K-Ras in EM-G3 cells The p53 gene belongs to the most frequently inactivated tumor suppressor genes in the human cancers. The results of mutation analysis of the whole coding sequence (exons 2–11) with adjacent intron–exon junctions revealed only the presence of homozygous nucleotide change at the position 14,168 within intron 7 (T-to-G). This genetic alteration is referred as polymorphism in IARC p53 mutation database (http:// www-p53.iarc.fr/Polymorphism.html). The PIK3K-AKT and RAS-RAF-MEK-ERKMAPK pathways are often affected in human cancer. Recently, it has been published that >90% of mutations in the PIK3CA gene occur in exons 7, 9, 20

253

coding for crucial C2, helical and kinase domains [27]. The frequency of mutations in these exons was determined by Li et al. as 35% in BC patients [28]. Mutations of k-Ras oncogene and its downstream signaling substrate BRAF are frequent in many human cancers. Ras gene mutations are clustered in codons 12, 13, and 61, whereas the most frequent mutation E600V in BRAF affects exon 15 [29]. Mutation analysis of hot spots for PIK3CA, BRAF, and k-Ras oncogenes in EM-G3 cells did not reveal any mutations in analyzed sequences.

Discussion In this report we present the establishment and characterization of a spontaneously immortalized clonal cell line, EM-G3, derived from a primary lesion of ductal breast carcinoma. The key question was to determine the origin of EMG3 cells. The EM-G3 cultured in vitro displayed the immunoprofile of basal cells (K5+/K14+/p63+/SMA–/ K19–). Using immunohistochemical analysis we have tried to identify EM-G3-related cells in the patient’s tumor sample. The majority of cancer cells displayed the characteristics typical for fully differentiated luminal epithelial cells (K5–/p63–/K18+/K19+). However, we have found two types of regions (both randomly interspersed within the tumor specimen) that contained cells with immunoprofile similar to EM-G3: the ‘minor islands’ of irregular cells (Fig. 3b, f, j, and n) and the TDLUs (Fig. 3q, r, and s). Besides hallmarks of basal cells, the EM-G3 expressed some luminal markers such as K7, K8 (data not shown), EMA, ESA, and K18. The mixed expression of basal/myoepithelial and luminal markers, and namely co-expression of K5/K14/K18, with negativity for K19/SMA, indicates that the EM-G3 represent the immature cells originated either from the population of bipotent progenitors (K5+/K14+; according to Birnbaum et al.) or from a later stage, i.e. luminalcommitted progenitors (K5+/K18+; according to Gusterson et al.) present in TDLUs [11, 14]. Moreover, TDLUs are believed to contain adult stem cells necessary for mammary gland proliferation during pregnancy [14], and their cells are frequent targets of malignant transformation [16, 17]. Therefore, we suppose that spontaneously immortalized cells of TDLUs are the most probable source of EM-G3 cells. To support the ‘stemness’ pattern of the EM-G3 cells we have performed the cDNA microarray, focused on gene expression profiling of BC cells. Using this approach we detected the expression of other basal

123

254


Table 1 Immunocytochemistry of EM-G3 cells in vitro and in the xenograft mouse model in comparison to cancer and non-cancer mammary epithelial cells presented in the histopathological paraffin sections of the patient’s tumor

Antigen

Cell line EM-G3 In vitro

K5 K14 K18 K19 SMA p63 ER PR EGFR EMA ESA

+ + + – – +a –b –b + + +

Patient breast cancer sample In vivo (nu/nu mice)

+ + +/– – – +a – – + +c +c

a

Positive nuclei in the majority of cells

b

Nuclear negativity, weak membrane/cytoplasmic positivity

c

Positive central cells in ductal-like structures in nu/nu mice

Major cancer cell population

– – + + – – +a +a N.D + +

Non-cancer mammary epithelial cells Luminal

Myoepithelial

– – + + – – + + N.D + +

+ + – – + +a – – N.D – –

N.D not done

markers typical for different types of stem cells (e.g. CD44, integrins a6 and b1) [20, 30]. Only a few human BC cell lines are able to form tumors in nude mice. We have shown that EM-G3 formed limited tumor nodules. However, the EM-G3 cells were not simply proliferating in nu/nu mice but, in addition, were able to differentiate. Despite of the clonal origin, the EM-G3 cells formed structures reminding of simple ducts, consisting of central cells expressing luminal antigens (EMA/ESA), and peripheral cells expressing myoepithelial antigens (p63/K14), similarly as in the normal mammary tissue. To our knowledge, EM-G3 is the first spontaneously immortalized human clonal cell line derived Fig. 4 Expression analysis of EM-G3 cells in vitro using cDNA array: a representative DNA array (a) of 96 cancerrelated genes showed expression of genes depicted in panel (b) (strong expression = black boxes, moderate expression = gray boxes)

123

from primary BC with such a clear differentiation potential. The observations of clone differentiation into two-cell-type mammary structures also support the hypothesis of development of myoepithelial cells from luminal cells, or at least from a common precursor [7, 8]. However, although the 3D microenviroment has substantial influence on cell differentiation, expression of K18 was nearly lost in vivo and K19 or SMA (typical for mature cells) was not gained. It has been described that breast progenitor cells express low amount of K18, often detectable only in frozen sections [31, 32]. In accordance with the evidence that K5 is a marker of mammary stem/progenitor cells [33] is our observa-


tion that this keratin remained positive in all cells (luminal-like and myoepithelial-like) present in primitive ductal structures until central cell disintegration (see Fig. 3d and v). Similar differentiation in vivo resulting in lumen formation was described for the BC44 mouse mammary epithelial cell line [34]. The EM-G3 cells isolated from nu/nu mouse tumors have been successfully seeded back to the culture 4 weeks after injection, when the tumor reached its definitive size. The cells migrated and proliferated from spherical clusters similar to ‘mammospheres’ formed in vitro from cancer-initiating cells [20]. However, 3 months after tumor growth in nude mice the viability of xenografted EM-G3 was substantially decreased. This could be explained either by bad vascularization of the encapsulated tumor (no vessels were found in the nodule) or by a higher degree of cell differentiation at that time. The EM-G3 cells did not form colonies in soft agar (data not shown) indicating their low-malignant potential. The further indication that EM-G3 cells are not true BC cells comes from their dependence on EGF in the medium. Typically, EGF is survival factor for mammary epithelial cells and stimulates the proliferation of rapidly dividing mammary epithelial cells present within breast tumors [35]. There are only a few examples of bona fide BC cells that depend on exogenous EGF for growth in vitro [4, 34, 36]. That said, EGF dependent cells typically isolated from breast tumors have not been found to be immortal in vitro, so this phenotype does distinguish the EM-G3 cells from other similar cells isolated from BC. Recent investigations showed that early changes (involving genomic DNA re-arrangements or mutations in critical genes) occur in stem or progenitor cells years before apparent primary cancer lesion development. During a subsequent period of time altered progenitors accumulate genomic insults, which in some of them reach the level allowing independent cellular progression [37]. Nascent cancer cells bear the characteristics of their progenitors in a certain phase of their development toward mature cell population. Using LOH analysis it has been demonstrated that major tumor cell populations harbor common genetic alterations with an adjacent ‘histologically normal’ tissue [38, 39]. The hypothesis that cancer originates from mutations in the stem cell population has been well established for haematopoietic cells [40, 41]. However, the evidence that solid cancers arise from mutations in stem cells was less definitive [40]. Recently, CD44+/CD24– human BC stem/progenitor cells able to generate tumors in the mammary fat pad of immunocompromised NOD/SCID mice have been

255

isolated [18–20], indicating that also mammary tumors are initiated in mutated stem cells. Our results on EMG3 character and origin support the hypothesis of a cancer precursor, i.e. an immortalized progenitor cell developing different unstable clones with various degrees of differentiation and mitotic activity. The BC cell lines show frequent and substantial chromosomal aberrations (http://cgap.nci.nih.gov/). Karyotype of the cell line EM-G3 in passage 142 was complex with monosomy of chromosomes 15 and 20 and derivative chromosomes 3, 6, 12, and 13. However, in comparison with established BC cell lines, EM-G3 cells revealed only ‘subtle’ genetic changes without amplifications of HER2/neu, cyclin D1, c-myc oncogenes or deletions of tumor suppressor genes p53 and Rb. These relatively small chromosomal changes were sufficient for spontaneous immortalization of EM-G3 cell progenitors and, together with lack of mutations in analyzed oncogenes and p53, support the speculation that the EM-G3 line represents a cell population in an early stage of neoplastic progression. The only karyotype difference between passages 142 and 212 was derivation of chromosome 9. This result has proved genomic stability of the EM-G3 cell line compared to established BC cell lines [42]. Contrary to EM-G3 in vitro, in the major cancer cell population of the original patient’s tumor amplification of the HER2/neu gene was identified. This observation further indicated that cultured EM-G3 cells are not true BC cells, but a premalignant immortal cell population present within the tumor specimen. In conclusion, the EM-G3 cells represent a unique spontaneously immortalized clonal cell line evidently derived from premalignant BC progenitors. Such an origin together with the high growth rate, ability to form tumors in vivo and preserved differentiation potential are promising attributes for further exploitation of the EM-G3 cells for research on BC progression. Acknowledgments We thank Mrs. Eva Taislova for excellent technical assistance. This work was supported by grants NR8145– 3 and NR8345–4 from the Grant Agency of the Ministry of Health of the Czech Republic, Grant No. 93/2005/C from the Grant Agency of Charles University, Project AVOZ50520514 from the Academy of Sciences of the Czech Republic, and Research Project MSM0021620808 from the Ministry of Education, Youth and Sports of the Czech Republic

References 1. Lasfargues EY, Ozzello L (1958) Cultivation of human breast carcinomas. J Natl Cancer Inst 21:1131–1147 2. Lacroix M, Leclercq G (2004) Relevance of breast cancer cell lines as models for breast tumours: an update. Breast Cancer Res Treat 83:249–289

123

256 3. Jones C, Payne J, Wells D, Delhanty JD, Lakhani SR, Kortenkamp A (2000) Comparative genomic hybridization reveals extensive variation among different MCF-7 cell stocks. Cancer Genet Cytogenet 117:153–158 4. Wang CS, Goulet F, Lavoie J, Drouin R, Auger F, Champetier S, Germain L, Tetu B (2000) Establishment and characterization of a new cell line derived from a human primary breast carcinoma. Cancer Genet Cytogenet 120:58–72 5. O’Hare MJ (1991) Breast cancer. In: Masters JRW (eds) Human cancer in primary culture, 1st edn. Kluwer Academic Publishers, Dordrecht 6. Taylor-Papadimitriou J, Stampfer M, Bartek J, Lewis A, Boshell M, Lane EB, Leigh IM (1989) Keratin expression in human mammary epithelial cells cultured from normal and malignant tissue: relation to in vivo phenotypes and influence of medium. J Cell Sci 94:403–413 7. Pechoux C, Gudjonsson T, Ronnov-Jessen L, Bissell MJ, Petersen OW (1999) Human mammary luminal epithelial cells contain progenitors to myoepithelial cells. Dev Biol 206:88–99 8. Clayton H, Titley I, Vivanco M (2004) Growth and differentiation of progenitor/stem cells derived from the human mammary gland. Exp Cell Res 297:444–460 9. Deugnier MA, Teuliere J, Faraldo MM, Thiery JP, Glukhova MA (2002) The importance of being a myoepithelial cell. Breast Cancer Res 4:224–230 10. Bocker W, Moll R, Poremba C, Holland R, Van Diest PJ, Dervan P, Burger H, Wai D, Ina DR, Brandt B, Herbst H, Schmidt A, Lerch MM, Buchwallow IB (2002) Common adult stem cells in the human breast give rise to glandular and myoepithelial cell lineages: a new cell biological concept. Lab Invest 82:737–746 11. Birnbaum D, Bertucci F, Ginestier C, Tagett R, Jacquemier J, Charafe-Jauffret E: Basal and luminal breast cancers: basic or luminous? (review). Int J Oncol 25:249–258 12. Stingl J, Raouf A, Emerman JT, Eaves CJ (2005) Epithelial progenitors in the normal human mammary gland. J Mammary Gland Biol Neoplasia 10:49–59 13. DiRenzo J, Signoretti S, Nakamura N, Rivera-Gonzalez R, Sellers W, Loda M, Brown M (2002) Growth factor requirements and basal phenotype of an immortalized mammary epithelial cell line. Cancer Res 62:89–98 14. Gusterson B, Ross D, Heath V, Stein T (2005) Basal cytokeratins and their relationship to the cellular origin and functional classification of breast cancer. Breast Cancer Res 7:143–148 15. Gudjonsson T, Villadsen R, Nielsen HL, Ronnov-Jessen L, Bissell MJ, Petersen OW (2002) Isolation, immortalization, and characterization of a human breast epithelial cell line with stem cell properties. Genes Dev 16:693–706 16. Wellings SR, Jensen HM, Marcum RG (1975) An atlas of subgross pathology of the human breast with special reference to possible precancerous lesions. J Natl Cancer Inst 55:231–273 17. Petersen OW, Gudjonsson T, Villadsen R, Bissell MJ, Ronnov-Jessen L (2003) Epithelial progenitor cell lines as models of normal breast morphogenesis and neoplasia. Cell Prolif 36(Suppl 1):33–44 18. Dontu G, Al Hajj M, Abdallah WM, Clarke MF, Wicha MS (2003) Stem cells in normal breast development and breast cancer. Cell Prolif 36(Suppl 1):59–72 19. Al Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF (2003) Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA 100:3983–3988

123

Breast Cancer Res Treat (2007) 103:247–257 20. Ponti D, Costa A, Zaffaroni N, Pratesi G, Petrangolini G, Coradini D, Pilotti S, Pierotti MA, Daidone MG (2005) Isolation and in vitro propagation of tumorigenic breast cancer cells with stem/progenitor cell properties. Cancer Res 65:5506–5511 21. Matouskova E, Dudorkinova D, Krasna L, Vesely P (2000) Temporal in vitro expansion of the luminal lineage of human mammary epithelial cells achieved with the 3T3 feeder layer technique. Breast Cancer Res Treat 60:241–249 22. Krasna L, Dudorkinova D, Vedralova J, Vesely P, Pokorna E, Kudlackova I, Chaloupkova A, Petruzelka L, Danes J, Matouskova E (2002) Large expansion of morphologically heterogeneous mammary epithelial cells, including the luminal phenotype, from human breast tumours. Breast Cancer Res Treat 71:219–235 23. Hermanova M, Lukas Z, Kroupova I, Kleibl Z, Novotny J, Nenutil R, Pazourkova M, Brazdil J, Kren L, Dite P (2003) Relationship between K-ras mutation and the expression of p21WAF1/CIP1 and p53 in chronic pancreatitis and pancreatic adenocarcinoma. Neoplasma 50:319–325 24. Brezinova J, Zemanova Z, Ransdorfova S, Sindelarova L, Siskova M, Neuwirtova R, Cermak J, Michalova K Prognostic significance of del(20q) in patients with hematological malignancies. Cancer Genet Cytogenet 160:188–192 25. Li SY, Rong M, Grieu F, Iacopetta B (2006) PIK3CA mutations in breast cancer are associated with poor outcome. Breast Cancer Res Treat 96:91–95 26. Mayr D, Hirschmann A, Lohrs U, Diebold J (2006) KRAS and BRAF mutations in ovarian tumors: a comprehensive study of invasive carcinomas, borderline tumors and extraovarian implants. Gynecol Oncol (in press) 27. Campbell IG, Russell SE, Choong DY, Montgomery KG, Ciavarella ML, Hooi CS, Cristiano BE, Pearson RB, Phillips WA (2004) Mutation of the PIK3CA gene in ovarian and breast cancer. Cancer Res 64:7678–7681 28. Li SY, Rong M, Grieu F, Iacopetta B (2006) PIK3CA mutations in breast cancer are associated with poor outcome. Breast Cancer Res Treat 96:91–95 29. Wooster R, Futreal AP, Stratton MR (2005) Sequencing Analysis of BRAF Mutations in Human Cancers. Methods Enzymol 407:218–224 30. Tumbar T, Fuchs E (2006) Epithelial skin stemm cells. In: Lanza R, Gearhart J, Hogan B (eds) Essentials of stem cell biology, 1st edn. Elsevier, Amsterdam, Academic Press, New York 31. Clarke CL, Sandle J, Parry SC, Reis-Filho JS, O’Hare MJ, Lakhani SR (2004) Cytokeratin 5/6 in normal human breast: lack of evidence for a stem cell phenotype. J Pathol 204:147–152 32. Laakso M, Loman N, Borg A, Isola J (2005) Cytokeratin 5/ 14-positive breast cancer: true basal phenotype confined to BRCA1 tumors. Mod Pathol 18:1321–1328 33. Boecker W, Buerger H (2003) Evidence of progenitor cells of glandular and myoepithelial cell lineages in the human adult female breast epithelium: a new progenitor (adult stem) cell concept. Cell Prolif 36(Suppl 1):73–84 34. Deugnier MA, Faraldo MM, Janji B, Rousselle P, Thiery JP, Glukhova MA (2002) The EGF controls the in vivo developmental potential of a mammary epithelial cell line possessing progenitor properties. J Cell Biol 159:453–463 35. Rosfjord EC, Dickson RB (1999) Growth factors, apoptosis, and survival of mammary epithelial cells. J Mammary Gland Biol Neoplasia 4:229–237 36. Paine TM, Soule HD, Pauley RJ, Dawson PJ (1992) Characterization of epithelial phenotypes in mortal and immortal human breast cells. Int J Cancer 50:463–473

Breast Cancer Res Treat (2007) 103:247–257 37. Smalley M, Ashworth A (2003) Stem cells and breast cancer: a field in transit. Nat Rev Cancer 3:832–844 38. Deng G, Lu Y, Zlotnikov G, Thor AD, Smith HS (1996) Loss of heterozygosity in normal tissue adjacent to breast carcinomas. Science 274:2057–2059 39. Lakhani SR, Chaggar R, Davies S, Jones C, Collins N, Odel C, Stratton MR, O’Hare MJ (1999) Genetic alterations in ‘normal’ luminal and myoepithelial cells of the breast. J Pathol 189:496–503

257 40. Marx J (2003) Cancer research. Mutant stem cells may seed cancer. Science 301:1308–1310 41. Reya T, Morrison SJ, Clarke MF, Weissman IL (2001) Stem cells, cancer, and cancer stem cells. Nature 414:105–111 42. Wenger SL, Senft JR, Sargent LM, Bamezai R, Bairwa N, Grant SG (2004) Comparison of established cell lines at different passages by karyotype and comparative genomic hybridization. Biosci Rep 24:631–639

123