This article is mentioned in:

Introduction

Breast cancer is the most commonly occurring malignancy in women worldwide. Various histological and molecular subtypes of breast cancer have been identified with different biologic implications (1–3). Apocrine differentiation (metaplasia) is frequently observed in the mammary epithelium; however, invasive apocrine carcinomas are rare, constituting less than 5% of all breast carcinomas (1,4). Apocrine lesions of the breast are characterized by over-expression of the androgen receptor (AR) along with loss of estrogen receptor-α (ER-α) and the progesterone receptor (PR) (5–7). Molecular studies of invasive apocrine breast carcinoma have shown a specific molecular apocrine profile based on the AR expression that divides ER-negative breast carcinomas into two different clusters: ER−/AR− (basal) and ER−/AR+ (molecular apocrine cluster) (8).

Breast cancer cell lines are widely used for experimental research as models of various subtypes of breast carcinomas (9–12). The MDA-MB-453 breast cancer cell line, obtained from a malignant pleural effusion of a 48-year-old female, has been suggested as a model for the molecular apocrine breast subtype (13–15). The cell line exhibits a characteristic apocrine carcinoma steroid receptor profile: ER-α-negative, PR-negative, and AR-positive (8,13). Increased proliferation in response to androgens is the key feature of the cell line, which can be blocked by anti-androgens, such as flutamide (13,16,17). Her-2/neu activity has been well documented in this cell line, as well as the existence of a functional cross-talk between AR and Her-2/neu, involving the MAPK/ERK1/2 pathway (13,14). These features also tend to characterize a substantial proportion of invasive apocrine carcinomas of the breast (18–20).

The aim of the present study was to further characterize the MDA-MB-453 cell line and to correlate it with the results obtained from the apocrine breast carcinoma samples of patients.

Materials and methods

Breast cancer cell lines and treatment

Human breast cancer cell line MDA-MB-453 was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and used for analysis and experiments, while MDA-MB-231, MCF-7 and BT-474 cell lines from the ATCC served as controls. MDA-MB-453, MDA-MB-231 and BT-474 were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco-BRL, Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) at 37°C in a humidified CO2 incubator. The MCF-7 cell line was cultured in Improved modified Eagle’s medium (IMEM, Invitrogen) supplemented with 10% FBS. Prior to the experiments, the cells were cultured in phenol-red-free DMEM (Invitrogen) for 24 h, and then in serum-free DMEM overnight. U0126 (MEK1/2 inhibitor, Cell Signaling Technology, Danvers, MA, USA), a highly selective inhibitor of MEK1/2, was used.

Breast tissue samples

Formalin-fixed paraffin-embedded blocks of 8 cases of apocrine carcinoma (seven invasive and one in situ case) that corresponded to the cell line profile (ER-α-, PR−, AR+ and Her-2/neu −/+) were selected from the previously well-characterized cohort (18) and used for comparative analysis.

Protein expression analysis

Protein expression was analyzed by Western blot analysis in cell lysates or by immunocytochemistry (ICC) and immunohistochemistry (IHC) on formalin-fixed and paraffin-embedded cell blocks prepared from the MDA-MB-453 cell line and tissue blocks from breast tissue samples (the list of antibodies is shown in Table I). For ICC and IHC analysis, commercially available detection kits and automated staining procedures were employed (18). Cell lysates, immunoprecipitations and Western blot analysis were carried out using standard procedures as previously described (21,22). Signals were detected by enhanced chemiluminiscence (ECL; Amersham, GE Healthcare Biosciences, Pittsburgh, PA, USA).

Table I List of the antibodies used in the study.

Table I

List of the antibodies used in the study.

Antibody	Manufacturer
ER-α (clone 6F11), PR (clone 16), Her-2/neu (Clone CB11), cyclin D1	Ventana Medical Systems
EGFR, MAPK1/2 (ERK1/2), pERK1/2 (Thr202/Tyr204)	Cell Signaling Technology
p16 (mouse monoclonal IgG1κ)	Cell Marque
Cyclin D1 (clone SP4 rabbit IgG), GCDFP-15 (clone 23A3 mouse IgG2A κ), ER-α (Cat no. RB-9016-P)	Neomarkers
EGFR (PharmDX diagnostic kit), Topoisomerase-IIα (Clone Ki-S1), Cytokeratin 5/6 (D5/16B4), p63 (clone 4A4)	DakoCytomation
Rb (IF8): sc-102, pRb(Ser795), p16 (N-20: sc-467), p53 (DO-1): sc-126, Actin [(I-19): sc-1616], AR (clone AR441: sc-7305), pEGFR (Tyr845), Her-2 (Neu (CB11): sc-52349), p-Neu (Tyr 1248)-R: sc-12352-R	Santa Cruz Biotechnology

A positive p16INK4A expression was defined as the presence of nuclear and/or cytoplasmic staining, and the percentage of positive cells was recorded (23,24). For ER-α, PR, AR, cyclin D1, p53, p63 and topoisomerase-IIα expression, only nuclear labeling was scored (24,25). For Her-2/neu and EGFR proteins, only membranous staining was considered positive. The scoring was carried out according to the manufacturer’s (EGFR, Dako, Carpinteria, CA, USA) and American Society of Clinical Oncology/College of American Pathologists’ guideline recommendations (Her-2/neu protein) (26). GCDFP-15 and CK5/6 proteins were considered positive if any membranous/cytoplasmic staining was observed.

Conventional cytogenetics

G-banding procedures were performed on metaphase cells. Metaphase chromosomes were banded with Wright trypsin and karyotypes were described following the established international guidelines (27).

Fluorescent in situ hybridization (FISH)

FISH was performed to evaluate copy numbers at EGFR, TOP2A and HER-2/neu loci. Chromosome enumeration probes CEP7 and CEP17 were used as indicators of chromosome copy numbers (Abbott Molecular Inc., Des Plaines, IL, USA). A total of 100 nuclei were scored per sample. A ratio of HER-2/CEP17 >2.2 was defined as gene amplification; a ratio of 1.8–2.1 was interpreted as borderline, and a ratio of <1.8 was defined as negative. The same criteria were used for the interpretation of EGFR/CEP7 and TOP2A/CEP17 ratios, respectively. For statistical purposes, equivocal FISH results (ratio of 1.8–2.1) were considered negative (27). Polysomy 7 and 17 were defined as ≥3 CEP signals per cell (18,28).

Flow cytometry

Flow cytometry (FC) was applied to measure the S-phase fraction of MDA-MB-453 cells. Cells at ~50% confluence were harvested and 1 ml cold 70% ethanol was slowly added to the cell pellet while vortexing. Ethanol-fixed cells were treated with 100 μg/ml RNaseA and 50 μg/ml propidium iodide (PI) in PBS at room temperature for 30 min. Flow cytometry of cell cycle distribution was performed using a FACSCalibur flow cytometer (BD-Biosciences, San Jose, CA, USA). The results were analyzed using Multicycle for Windows.

Gene mutation analysis

These assays included the detection of 12 mutations in codons 12 and 13 of the K-RAS gene as well as the V600E mutation in exon 15 of the B-RAF gene. The analysis was performed using a Mutector II assay, using proprietary Shift Termination Assay (STA) technology (TrimGen Corporation, Sparks, MD, USA) following the previously described protocol (29). For quality control of the K-RAS mutation analysis, we used a colon cancer biopsy known to be positive for K-RAS mutation, while a case of malignant melanoma served as a positive control for B-RAF mutation.

Statistical analysis

Quantitative data of the experimental studies were expressed as the mean ± SD. The student’s t-test was used to test the differences between responses in the exposed and control groups. Statistical Package for the Social Sciences version 17.0 (SPSS, Chicago, IL, USA) was used for statistical analysis. P<0.05 was considered to be statistically significant.

Results

Profiling of MDA-MB-453 cell line

A summary of the key findings and comparison with the apocrine carcinoma tissue samples is shown in Table II. The MDA-MB-453 cell line was negative for ER-α, PR and Her-2/neu protein on ICC, whereas AR was found to be positive by Western blot analysis. Notably, the ICC assay demonstrated no nuclear staining but predominantly cytoplasmic distribution of AR, whereas the GCDFP-15 protein was absent. Western blot analysis revealed low levels of Her-2/neu and phosphorylation of the Her-2/neu protein at Tyr1248 in comparison with BT-474 cells, an ER-positive breast cancer cell line with HER-2/neu gene amplification. Western blot analysis revealed that levels of total EGFR and phosphorylation of EGFR at Tyr845 were lower in MDA-MB-453 cells compared with EGFR- over-expressing MDA-MB-231 cells. In contrast, the ICC assay revealed a strong EGFR protein expression (score 3+). FISH analysis revealed neither HER-2/neu (HER2/CEP17 was 3.96/3.68, ratio: 1.08) nor EGFR gene amplification (EGFR/CEP7 was 3.23/2.90, ratio: 1.11). Gains of chromosomes 7 and 17 (CEP7 and 17) observed with FISH were further confirmed using conventional cytogenetic analysis, which revealed a hypertriploid clone characterized by extensive numerical and structural abnormalities including gains of chromosomes 7, 11, 17, 19 and 21 and losses of chromosomes X, 3, 4, 9, 13, 14, 16 and 18.

Table II A summary of the key findings in the cell line and apocrine carcinoma tissue samples.

Table II

A summary of the key findings in the cell line and apocrine carcinoma tissue samples.

Feature	MDA-MB-453	Apocrine carcinoma tissue samples
Estrogen/progesterone receptor	Negative	Negativeb
Androgen receptor	Positivea	Positiveb
GCDFP-15 protein	Negative	Positivec
HER-2/neu status	Low protein expressiona/no gene amplification	High protein expression common/gene amplification b
EGFR status	Low protein expressiona/no gene amplification	High protein expression common/no gene amplification b
KRAS gene status	Mutated (Gly13Asp GGC>GAC)	Not mutated
BRAF gene status	Not mutated	Not mutated
p16INK4A protein	Lost due to gene deletion at 9p.21/no protein detected	Present (variable cytoplasmic/nuclear staining)
Cyclin D1 protein	Highly over-expressed	Low expression
Polysomy 7 and 17 (CEP7 and 17)	Present	Frequently observedb
Basal markers (CK5/6, p63)	Low expression	Low expressiond

a By Western blot analysis;

b Vranic et al (18);

c Tavassoli et al (1);

d Weigelt et al (4).

Of particular functional importance was the deletion of the 9p21 locus that harbors the INK4A gene (p16INK4A) and INK4B gene (p15), as evidenced by the loss of the p16INK4A protein expression in Western blot or ICC analysis (Fig. 1A).

Figure 1 p16INK4A protein was completely absent in MDA-MB-453 cells (A) whereas apocrine carcinoma tissue samples predominantly retained a membranous and cytoplasmic pattern of staining (B).

CK5/6 was completely absent, whereas the p63 protein expression with defined nuclear staining was observed in approximately 10% of the cells.

MDA-MB-453 cells exhibit a high proliferation rate

MDA-MB-453 cells exhibited an excessive proliferation rate under optimal culture conditions as its S-phase fraction measured up to 70% by flow cytometry. This rate was consistent with the cyclin D1 over-expression (positive in ~70% of the cells), inactivation of retinoblastoma protein (Rb), measured by its phosphorylation at Ser795, and the loss of p16INK4A protein expression. The cell line also exhibited a high p53 protein expression measured by both Western blot and ICC analysis. Consequently, topoisomerase-IIα was also strongly expressed in approximately 70% of the cells (without underlying TOP2A gene amplification, TOP2A/CEP17: 2.86/ 2.80, ratio: 1.02). When cultured in a serum-free medium for 24 h, the cell line remained highly proliferative with an S-phase fraction of approximately 40%. The cell line was also capable of sustained proliferation in a serum-free medium for up to 48 h, as demonstrated by the S-phase fraction.

MDA-MB-453 cells harbor a K-RAS mutation and exhibit constitutive activation of the MAPK/ERK pathway

Mutational analysis of the K-RAS gene revealed codon 13 mutation (Gly 13 Asp GGC>GAC), a mutation that constitutively activates K-Ras in the MDA-MB-453 cell line (Fig. 2), whereas there is no gene alteration in the B-RAF gene. Western blot analysis confirmed that ERK1/2 is highly phosphorylated in cells irrespective of the concentration of the fetal calf serum, and culture conditions. In addition, MDA-MB-453 retained high ERK activity even after 48 h in serum-free medium, and was highly sensitive to the specific MEK1/2 inhibitor U0126. Low concentrations of U0126 (0.1 μM) led to a partial inhibition of ERK1/2 phosphorylation after 12 h. The exposure of MDA-MB-453 cells to U0126 at a concentration of 10 μM for various time periods (from 15 min to 24 h) led to a markedly reduced S-phase fraction as measured by flow cytometry (Fig. 3A and B). Thus, 12 h treatment of the MDA-MB-453 cells with U0126 resulted in a significantly reduced proliferation rate in the experimental group in comparison with the control group (22.6 vs. 38.3%, reduction ~42%, p<0.001, Student’s t-test). Western blot analysis with phospho-specific anti-ERK1/2 antibody also showed that U0126 abolished ERK1/2 phosphorylation in MDA-MB-453 cells.

Figure 2 Mutational analysis of the K-RAS gene revealed codon 13 mutation (Gly 13 Asp GGC>GAC), a mutation that constitutively activates K-Ras in the MDA-MB-453 cell line.

Figure 3 The exposure of MDA-MB-453 cells to U0126 at a concentration of 10 μM at different time periods (from 15 min to 24 h) led to a dramatically reduced S-phase fraction in the cells as measured by flow cytometry. Black arrows indicate S-phase fraction after 24 h in control group (A) where S-phase was ~39%, and exposed group (B) where S-phase was ~6%.

Breast tissue sample analysis

Of 5 tested patient samples of the apocrine carcinomas, 1 exhibited an absence of p16INK4A protein expression, 3 samples (two invasive and one in situ carcinoma) had predominantly cytoplasmic p16INK4A expression (observed in 20–80% of the tumor cells), whereas only 1 case retained a nuclear p16INK4A expression intensity, compared with the positive control (Fig. 1B). Cyclin D1 expression was observed in 3 out of 5 cases. However, only 2 of these cases reached more than 10% of the positive cells with predominantly weak to moderate nuclear intensity.

K-RAS and B-RAF mutational analysis performed on 6 invasive apocrine carcinoma samples showed no gene alterations in any of the studied cases. Notably, these cases had a strong EGFR protein over-expression on IHC (scores 2+/3+ on IHC) without underlying EGFR gene amplification (18).

Discussion

The present study describes complex and multiple cytogenetic and molecular alterations in the MDA-MB-453 cell line that is currently considered to be a model for breast apocrine carcinoma. Conventional cytogenetic analysis revealed that MDA-MB-453 is a hypertriploid cell line that corresponds closely to its metastatic origin and high malignant potential (10). We also observed considerable discrepancies between published cytogenetic findings regarding the cell line and our own results, although certain previously reported cytogenetic findings, including gains of chromosomes 7 and 17 and loss of the 9p.21 locus (CDKN2 gene) (30), are also observed in our study. Thus, previous studies (31,32) reported HER-2/neu gene amplification in the MDA-MB-453 cell line, whereas our analysis did not support those findings. FISH and conventional cytogenetic analysis revealed a gain (polysomy) of CEP17 in only a small population of cells (~5% of the counted cells) that carried HER-2/neu gene amplification. Levels of total Her-2/neu expression and phosphorylation of Her-2/neu at Tyr1248, often associated with HER-2/neu gene amplification (33), were lower in comparison with the positive cell control (BT-474 cell line). Belsches-Jablonski et al (34) also revealed a low Her-2/neu protein expression in the MDA-MB-453 cell line, whereas Kao et al (11) found low copy numbers of the HER-2/neu gene in the cell line using a quantitative polymerase chain reaction, and thus classified MDA-MB-453 cells as an HER-2-negative cell line. Consequently, the classification of the cell line varies from luminal (12,31,32) or ‘weakly luminal’ (10) to an apocrine cell line model (8,13–15). We propose that the observed disparities are caused by the prolonged cell culture, which may modify genetic/phenotypic properties of the cell line (9,10), and further stress the importance of a laboratory self validation of the cell line.

Our study reports for the first time a codon 13 mutation in the K-RAS gene followed by constitutive MAPK/ERK activation in the MDA-MB-453 cell line (35,36). The K-RAS gene is an integral part of the MAPK signaling pathway, whose activation strongly correlates with the degree of the K-RAS gene activation (36). The MAPK/ERK signaling pathway is crucial in the regulation of cell proliferation, differentiation, survival and metastasis through ultrasensitive switch-like responses to various stimuli depending on the strength and duration of stimulation (35,37–40). Some of these effects, such as proliferation, may be abolished by the targeted inhibition, as we demonstrated using the specific MEK1/2 inhibitor U0126 in our study.

K-RAS gene mutations (usually restricted to codons 12, 13 and 61) are frequently observed in human tumors (~15%) (41) and established breast cancer cell lines (35). However, Hollestelle et al (35) reported PTEN but not K-RAS gene mutation in this cell line, although these authors found that 13% of the studied breast cancer cell lines harbored K-RAS gene mutations. Furthermore, results obtained from the cell line were not in concordance with the data from the breast cancer samples, which is consistent with findings of previous reports that have confirmed the lack of K-RAS mutations in breast carcinomas (36,42–45). Notably, the lack of K-RAS mutations in apocrine carcinomas may become a potential therapeutic benefit, since the tested cases were EGFR-positive and thus are potentially amenable to targeted anti-EGFR therapy (44). B-RAF gene mutations, mainly involving the V600E locus, have also been described in a wide range of human tumors with the highest frequency in malignant melanoma (41). These mutations are rarely observed in breast carcinoma (41). Neither the cell line nor apocrine breast carcinoma samples harbored the B-RAF gene mutations.

Another significant finding in this study was the deletion of the 9p21–22 region, responsible for the complete loss of the p16INK4A protein as confirmed by ICC and Western blot analysis. This finding is consistent with a recent study that reported an allelic loss at 9p21 loci in MDA-MB-453 cells with a barely detectable p16INK4A protein expression (29). p16INK4A, the product of the CDKN2 gene, one of the key cell cycle regulators, inhibits phosphorylation of the retinoblastoma protein, and thus acts as a negative regulator of the cell cycle (46). In contrast, loss of p16INK4A was found in only 1 of 5 tested apocrine carcinomas of the breast with loss p16INK4A protein expression, whereas the remaining 4 exhibited pre-dominantly cytoplasmic patterns of distribution of the p16INK4A protein. Notably, several investigators previously reported a cytoplasmic distribution of the p16INK4A protein in a subset of breast carcinomas (47,48), which correlated with a high histological grade, loss of ER and PR, p53 protein over-expression and accelerated tumor proliferation, the parameters frequently featured in apocrine carcinoma and the MDA-MB-453 cell line. Similarly, the cell line exhibited a high cyclin D1 expression, which was not observed in the breast tissue samples.

In conclusion, the observed cytogenetic and molecular alterations in the MDA-MB-453 cell line were not consistently present in the apocrine tumor samples. Our cell culture results also differ in certain respects from previously published data. MDA-MB-453 is not an ideal model of apocrine carcinomas as recently defined using molecular methods. Notably, EGFR-positive apocrine carcinomas of the breast do not tend to harbor K-RAS gene mutations observed in the cell line, which indicates that targeted therapy with EGFR inhibitors may be a viable alternative in some patients.

Acknowledgements

This study was supported by the National Institutes of Health Grant, DK070016 (Z.-Y. Wang). Dr Semir Vranic was a research fellow at Creighton University Medical Center, Omaha, NE, USA, and was supported by a UICC American Cancer Society Beginning Investigators Fellowship (ACSBI) (ACS/08/004) funded by the American Cancer Society. We appreciate the technical assistance of Kay Krogman, Deborah Jankovich, and Kristin Bonnstetter (Creighton Medical Laboratories, Creighton University School of Medicine, Omaha, NE, USA) and Dr Faruk Skenderi (Sarajevo University School of Medicine, Bosnia and Herzegovina). We are also indebted to Dr Warren Sanger and Ms. Marilu Nelson (The University of Nebraska Medical Center, Omaha, NE) for their excellent cytogenetics support.

References