Involvement of epidermal growth factor receptor overexpression in the promotion of breast cancer brain metastasis

Authors

  • Fang Nie MD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Jian Yang PhD,

    1. Department of Systems Medicine and Bioengineering, The Methodist Hospital Research Institute, Weill Cornell Medical College, Cornell University, Houston, Texas
    Search for more papers by this author
  • Song Wen MD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Yan-Li An MD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Jie Ding MS,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Sheng-Hong Ju PhD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Zhen Zhao MD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Hua-Jun Chen MD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Xin-Gui Peng MD,

    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    Search for more papers by this author
  • Stephen T. C. Wong PhD,

    1. Department of Systems Medicine and Bioengineering, The Methodist Hospital Research Institute, Weill Cornell Medical College, Cornell University, Houston, Texas
    Search for more papers by this author
  • Hong Zhao PhD,

    1. Department of Systems Medicine and Bioengineering, The Methodist Hospital Research Institute, Weill Cornell Medical College, Cornell University, Houston, Texas
    Search for more papers by this author
  • Gao-Jun Teng MD, PhD

    Corresponding author
    1. Jiangsu Key Laboratory of Molecular and Functional Imaging, Department of Radiology, Zhong-Da Hospital, Medical School, Southeast University, Nanjing, China
    • Department of Radiology, Zhong-Da Hospital, Southeast University, 87 Ding-Jia-Qiao Road, Nanjing 210009, China

    Search for more papers by this author
    • Fax: (011) 86-25-83311083


Abstract

BACKGROUND:

Brain-metastatic breast cancer (BMBC) is increasing and poses a severe clinical problem because of the lack of effective treatments and because the underlying molecular mechanisms are largely unknown. Recent work has demonstrated that deregulation of epidermal growth factor receptor (EGFR) may correlate with BMBC progression. However, the exact contribution that EGFR makes to BMBC remains unclear.

METHODS:

The role of EGFR in BMBC was explored by serial analyses in a brain-trophic clone of human MDA-MB-231 breast carcinoma cells (231-BR cells). EGFR expression was inhibited by stable short-hairpin RNA transfection or by the kinase inhibitor erlotinib, and it was activated by heparin-binding epidermal growth factor-like growth factor (HB-EGF). Cell growth and invasion activities also were analyzed in vitro and in vivo.

RESULTS:

EGFR inhibition or activation strongly affected 231-BR cell migration/invasion activities as assessed by an adhesion assay, a wound-healing assay, a Boyden chamber invasion assay, and cytoskeleton staining. Also, EGFR inhibition significantly decreased brain metastases of 231-BR cells in vivo. Surprisingly, changes to EGFR expression affected cell proliferation activities less significantly as determined by a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, an anchorage-independent growth assay, and cell cycle analysis. Immunoblot analysis suggested that EGFR drives cells' invasiveness capability mainly through phosphoinositide 3-kinase/protein kinase B and phospholipase C γ downstream pathways. In addition, EGFR was involved less in proliferation because of the insensitivity of the downstream mitogen-activated protein kinase pathway.

CONCLUSIONS:

The current results indicated that EGFR plays more important roles in cell migration and invasion to the brain than in cell proliferation progression on 231-BR cells, providing new evidence of the potential value of EGFR inhibition in treating BMBC. Cancer 2012. © 2012 American Cancer Society.

INTRODUCTION

Brain metastasis is the most common type of malignancy identified in the brain, and breast cancer is the second most common cause of the development of brain metastasis.1, 2 The most feared complication of breast cancer, brain metastasis develops in 10% to 20% of patients with metastatic breast cancer and mostly occurs late during disease progression.3 The incidence appears to be rising as we achieve improved treatments for extracranial disease, but the highly selective blood-brain barrier creates a “sanctuary” site for breast tumors.3, 4

Unfortunately, there are no effective therapies for brain-metastatic breast cancer (BMBC), and the mechanisms underlying its establishment and progression are largely unknown. Identification of the mechanisms that regulate BMBC is a prerequisite for the development of new, efficient treatments.5 To date, only a few studies have documented genes that may contribute to BMBC through the investigation of brain-colonizing variants of human breast cancer carcinoma cell lines as well as retrospective patient tissues. These genes are likely to be dispensable for primary tumor initiation and growth and may or may not be part of gene expression profiles exhibited by the primary tumor.6 Bos et al7 identified 17 genes that may mediate the spread of metastatic breast cancer cells to the brain as a brain-metastases signature (BrMS), including heparin-binding epidermal growth factor-like growth factor (HB-EGF), which is a ligand of epidermal growth factor receptor (EGFR). Also, retrospective studies have demonstrated higher levels of EGFR expression in patients with BMBC than that in patients with primary breast cancer.8-11 Those studies indicate that the EGFR pathway may play an important role in BMBC. However, the exact contribution that EGFR makes to BMBC progression remains unclear.

EGFR is a transmembrane protein that belongs to the erythroblastic leukemia viral oncogene homolog (ErbB) tyrosine kinase family, which consists of EGFR (ErbB-1), human epidermal growth factor receptor 2 (HER2) (ErbB-2), HER3 (ErbB-3), and HER4 (ErbB-4). Upon ligand-dependent homodimerization or heterodimerization with all members of the ErbB family, EGFR can be activated functionally to lead to kinase activation and initiation of signaling12 and is believed to play a major role in survival, growth, and proliferation of mammalian cells.13 EGFR deregulation has been observed in breast cancer and is associated with a poor clinical outcome.14-16 Despite evidence for the role of EGFR in breast cancer and the development of EGFR-blocking agents, including small-molecule inhibitors and monoclonal antibodies, clinical trials of EGFR inhibitors have exhibited limited efficacy in decreasing tumor growth in terms of what would be expected in the treatment of EGFR-overexpressing breast cancer.17, 18 These results raise the possibility that EGFR may contribute to breast cancer by affecting characteristics other than proliferation, such as metastasis.

These observations motivated us to explore the mechanistic role of EGFR in BMBC progression. In the current study, a brain-seeking derivative of the human MDA-MB-231 breast carcinoma cell line (MDA-MB-231-BR cells) was used to elucidate the function of EGFR in BMBC by characterizing its role in cell proliferation, migration, and invasion using both in vitro and in vivo models. Subsequent analyses revealed that inhibition of EGFR by short-hairpin RNA (shRNA) knockdown or erlotinib (a small-molecule EGFR inhibitor) in MDA-MB-231-BR cells resulted in significant inhibition of cell migration in vitro and prevented metastatic colonization of the brain in vivo. However, inhibition or activation of EGFR expression mildly affected cell proliferation, providing new evidence of the potential value of EGFR inhibition for treating BMBC.

MATERIALS AND METHODS

Cell Culture and Knockdown of Epidermal Growth Factor Receptor With Short-Hairpin RNA Plasmid

The human MDA-MB-231-BR “brain-seeking” breast cancer cell line (231-BR cells) was described previously.19 The 231-BR cell line transfected with enhanced green fluorescent protein (EGFP) was kindly provided by Dr Patricia S. Steeg (National Cancer Institute, National Institutes of Health, Bethesda, Md).9 The MDA-MB-231 parental cell line (231-P) was purchased from American Type Culture Collection (ATCC) (Manassas, Va). Cells were maintained in high-glucose Dulbecco's Modified Eagle Medium (DMEM) (GIBCO, Grand Island, NY) supplemented with 10% fetal bovine serum (FBS) (GIBCO).

Clones of 231-BR cells with stable knockdown of EGFR expression were generated using commercial plasmids for introduction of shRNAs (sc-29301-sh and sc-44340-sh; Santa Cruz Biotechnology, Santa Cruz, Calif). Control shRNA was used to generate control cell lines. The protocol recommended by the manufacturer was followed. shRNA plasmids were added to 231-BR cells for 48 hours, and the cells subsequently were cultured in medium containing 1 μg/mL puromycin (Merck and Company, Inc., Whitehouse Station, NJ). The cells were maintained and expanded in this selective medium to select individual clones with stable knockdown. EGFR levels relative to control clones and 231-P cells were determined by flow cytometry and immunoblot analysis.

Flow Cytometry for the Detection of Epidermal Growth Factor Receptor Expression

Cells were fixed in 4% paraformaldehyde, blocked using 1% (weight/volume) bovine serum albumin (BSA) (Sigma Chemical Company, St. Louis, Mo) in phosphate-buffered saline (PBS), incubated with 1:200 rabbit anti-EGFR primary antibody (Santa Cruz Biotechnology) for 45 minutes at 4°C, and stained with 1:200 Alexa Fluor 532 anti-rabbit immunoglobulin G secondary antibody (Invitrogen, Carlsbad, Calif) for 30 minutes at room temperature. Fluorescence intensity was analyzed with a FACS Calibur Flow Cytometer (Becton Dickinson, San Jose, Calif).

Immunoblot Analysis of Epidermal Growth Factor Receptor and Downstream Signaling Proteins

After serum starvation, cells were treated with 100 ng/mL HB-EGF (Peprotech Inc., Rocky Hill, NJ) for 30 minutes or were treated with various concentrations of erlotinib (Selleck Chemicals LLC, Houston, Tex) for 24 hours before lysis. Cell lysates and immunoblots were performed as described in Gril et al.20 All primary antibodies were obtained from Santa Cruz Biotechnology, and secondary horseradish peroxidase-conjugated antibodies were obtained from GE Healthcare (Port Washington, NY). Densitometric analysis was performed using ImageJ software (National Institutes of Health, Bethesda, Md).

Cell Adhesion Assay

Ninety-six-well plates were precoated with fibronectin (Millipore Corporation, Billerica, Mass). Nonspecific binding sites were blocked by incubation with 1% BSA in DMEM for 1 hour at 37°C. Then, 100,000 cells in 100 μL of DMEM were added to each well and allowed to adhere for 30 minutes at 37°C. The wells were washed and fixed, and adherent cells were observed under a fluorescent microscope and counted.

Wound-Healing Migration Assay

Cells were plated in 6-well plates and grown to confluence. After serum starvation overnight, cells were scraped using a 1-mL pipette to induce a “wound.” The scraped off cells were washed out, and the remaining cells were cultured in fresh medium in the absence or presence of 100 ng/mL HB-EGF. Images of labeled fields in the scraped wound were taken at the time of wound induction (0 hours) and 18 hours after cell wounding. The area of the wound was measured using ImageJ software for statistical analysis.

Boyden Chamber Invasion Assay

Serum-starved cells were treated with 100 ng/mL HB-EGF or 60 μM and 70 μM erlotinib for 24 hours. Cell invasion was examined with the use of 24-well BD Biocoat Matrigel Cell Invasion Chamber (Becton Dickinson. East Rutherford, NJ), as previously described.21 Instead of staining the invaded cells with a Diff-Quik kit, the cells were stained with 4′,6-diamidino-2-phenylindole (DAPI), observed under a fluorescent microscope, and counted.

Cell Proliferation Assay

Cells were plated at a density of 3 × 103 cells per well in 96-well plates. After serum starvation overnight, the cells were cultured in the absence or presence of 100 ng/mL HB-EGF. The number of viable cells was followed for 6 days after cell seeding by adding 0.5 mg/mL 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma Chemical Company) to each well. After 4-hour incubation at 37°C, dimethyl sulfoxide was added, absorbance was measured at 490 nm, and growth curves were drawn. The results are presented as the fold growth compared with day 0.

Anchorage-Independent Cell Proliferation

Culture medium (0.5 mL) containing 0.4% agarose, cells (10,000/well) and 10% FBS were overlaid onto the bottom layer of 24-well plates containing 0.6% agarose in the same culture medium (0.5 mL). The plates were cultured in the absence or presence of 100 ng/mL HB-EGF or 60 μM and 70 μM erlotinib for 2 weeks. At the end of the culture, colonies that measured >100 μm in greatest dimension were counted.

Cell Cycle Analysis

The cell cycle was analyzed by flow cytometry according to standard procedures. In brief, the cells were harvested, washed, and gently resuspended in 250 μL of hypotonic fluorochrome solution (PBS, 50 μg propidium iodide, 0.1% sodium citrate, and 0.1% Triton X-100) with RNase A (100 U/mL) to stain cell nuclei. The DNA content was analyzed on the FACS Calibur Flow Cytometer (Becton Dickinson).

Cytoskeleton Staining and Confocal Microscopy

The cells were fixed with 4% paraformaldehyde, blocked in 1% BSA in PBS, permeabilized in 0.5% Triton X-100 in PBS, and stained with 100 nM rhodamine-conjugated phalloidin (Cytoskeleton Inc., Denver, Colo) for 45 minutes at room temperature in the dark. Nuclei were counterstained with DAPI. Fluorescence images were obtained with conventional fluorescein isothiocyanate, tetramethyl rhodamine isothiocyanate, and DAPI excitation settings on the laser confocal microscope (Olympus FluoView FV1000; Olympus Corporation, Tokyo, Japan).

In Vivo Animal Experiments

Animal procedures were conducted in accordance with approvals and guidelines of the Institutional Animal Care and Use Committee (IACUC) of the Medical School of Southeast University. Under isoflurane/O2 anesthesia, 30 female BALB/c nude mice (4 weeks old; Laboratory Animal Center, Academy of Military Medical Science, Beijing, China) were injected into the left cardiac ventricle with 231BR-control shRNA or cells 231BR-EGFR shRNA-1 cells (2 × 105 cells in 0.1 mL serum-free medium; n = 15 mice per cell line).

The mice were killed under CO2 asphyxiation after 4 weeks, and the brains were excised for imaging. The whole brain was cut into 4 equal, 2-mm-thisk sections to detect the fluorescence intensity of deep-seated tissues. EGFP fluorescence was detected in whole brain and sectioned brain with the use of a Maestro EX In Vivo Spectral Imaging System (Cambridge Research and Instrumentation, Woborn, Mass). Data processing software was used to analyze fluorescence intensity, and the photon count rate (counts/second/mm2) was calculated to compare dynamic changes in signal intensity.

After fluorescence imaging, brain sections (10 μm thick) were cut serially, 1 section from every 300 μm was stained with hematoxylin and eosin (H&E), and immunohistochemistry for EGFR and Ki-67 was performed according to standard procedures. Thirty H&E-stained serial sections every 300 μm through the whole brain were analyzed for the presence of metastatic lesions under a microscope. Every large metastasis (>50 μm2) and micrometastasis (≤50 μm2) was counted in each section for quantification analysis. EGFR-positive and Ki-67-positive cells were counted on 10 randomly chosen visual fields from 1 section per mouse (n = 5 mice per cell model) at ×400 magnification. All analyses were carried out by 2 investigators who were blinded to experimental group assignment.

Statistical Analysis

Statistical analysis was performed using SPSS software (SPSS for Windows, version 19.0, 2010; SPSS Inc., Chicago, Ill). To determine differences across multiple groups, we used 1-way analyses of variance. Independent 2-sample t tests were conducted to detect difference in means the between 2 groups. The data are summarized as means ± standard deviations. P values < .05 were considered statistically significant.

RESULTS

Expression of Epidermal Growth Factor Receptor in 231-BR Cells and Epidermal Growth Factor Receptor Knockdown Cells

The MDA-MB-231 cell line was selected through 6 rounds of in vivo passage for a brain-metastatic subline (231-BR),19 which had 100% brain metastasis efficiency.20, 21 EGFR expression was higher in 231-BR cells (3.2-fold; quantified by immunoblot analysis; P < .001) than in 231-P cells (Fig. 1a,b), although estrogen receptor, progesterone receptor, and HER2 all were negative.

Figure 1.

The analysis of epidermal growth factor receptor (EGFR) expression by immunoblotting and flow cytometry is illustrated. (a) Immunoblotting of EGFR expression in the human MDA-MB-231 parental cell line (231-P), brain-seeking breast carcinoma cells (231-BR), short-hairpin RNA transfected 231-BR cells (231BR-control shRNA cells, 231BR-EGFR shRNA-1 cells, and 231BR-EGFR shRNA-2 cells). Tubulin expression was used as a loading control. (b) Flow cytometry analysis of EGFR expression is illustrated. (c) Immunoblotting of phosphorylated-EGFR (p-EGFR) is illustrated at various concentrations in erlotinib-treated 231-BR cells.

To understand how EGFR contributes to BMBC progression, shRNA knockdown or kinase inhibitor was used to inhibit EGFR expression in 231-BR cells, and the cells were characterized for their tumorigenicity. One single-cell, stable clone with the strongest degree of EGFR knockdown derived from each EGFR shRNA was selected for subsequent studies (231BR-EGFR shRNA-1 cells and 231BR-EGFR shRNA-2 cells). EGFR-knockdown 231-BR cells reduced EGFR protein levels by 88.5% (P < .001) in the 231BR-EGFR shRNA-1 clone and by 79.9% (P < .001) in the 231BR-EGFR shRNA-2 clone relative to the control clone (231BR-control shRNA), as determined by immunoblot analysis (Fig. 1a). Flow cytometry analysis confirmed much lower cell surface expression levels of EGFR for 231BR-EGFR shRNA-1 cells and 231BR-EGFR shRNA-2 cells (Fig. 1b). Immunoblot analysis also revealed that erlotinib decreased autophosphorylation levels of EGFR in 231-BR cells in a dose-dependent manner, and 24 hours of 60 μM and 70 μM erlotinib treatments inhibited phosphorylated EGFR (p-EGFR) by 72.4% (P < .001) and 89.2% (P < .001), respectively (Fig. 1c), and these 2 concentrations of erlotinib (inducing similar EGFR inhibition as 2 shRNA knockdown clones) were used in subsequent studies.

Epidermal Growth Factor Receptor Plays Important Roles in Cell Migration and Invasion on 231-BR Cells In Vitro

To identify the in vitro biologic correlation of EGFR on brain-metastatic progression, we used an adhesion assay, a wound-healing assay, and a Boyden chamber invasion assay on 231-BR cells. To analyze 3-dimensional cell invasion properties, the number of 231BR-EGFR shRNA-1 cells and 231BR-EGFR shRNA-2 cells that invaded had a very pronounced decrease (62.33% [P < .01] and 51.67% [P < .01], respectively) in contrast to 231BR-control shRNA cells. In addition, similar anti-invasion effects were observed in the erlotinib treatment groups (44.23% and 55.64% for 60 μM and 70 μM erlotinib, respectively). After 100 ng/mL HB-EGF stimulation, 231BR-control shRNA + EGF produced significantly more invading cells (43.34% more) than control cells (P < .01) (Fig. 2a,b).

Figure 2.

Epidermal growth factor receptor (EGFR) reveals important roles in cell migration and invasion in brain-seeking human MDA-MB-231 breast carcinoma (231-BR) cells in vitro. (a,b) A Boyden chamber invasion assay is illustrated in (a) representative images of 231-BR cell invasion (original magnification, ×200) and (b) a graph indicating the percentage of invasion compared with the control cell group. shRNA indicates short-hairpin RNA; EGF, epidermal growth factor. (c,d) A wound-healing assay is illustrated in (c) phase-contrast images of wound closure in 231-BR cells (Top) at the time of wound induction (0 hours) and (Bottom) 18 hours after cell wounding (original magnification, ×100; solid lines indicate wound edges) and (d) a graph that indicates the percentage of wound closure compared with the control cell group. (e,f) A cell adhesion assay is illustrated in (e) representative images of the adhesion of 231-BR cells (original magnification, ×100) and (f) a graph indicating the percentage of adhesion compared with control cells. Single asterisk indicates P < .05 versus control; double asterisks, P < .01 versus control; triple asterisks, P < .001 versus control.

In the wound-healing cell migration analysis, EGFR knockdown markedly reduced the 2-dimensional migration of 231BR-EGFR shRNA-1 and 231BR-EGFR shRNA-2 cells at 18 hours postscratch by 53.91% (P < .001) and 45.22% (P < .001), respectively, compared with control cells; whereas HB-EGF stimulation increased cell migration by 22.61% (P < .05) (Fig. 2c,d). Also, in an analysis of cell adhesion properties, the 231BR-EGFR shRNA-1 and 231BR-EGFR shRNA-2 cell groups produced less adherent cells than controls by 66.20% (P < .001) and 59.89% (P < .001), respectively, and 231BR-control shRNA + EGF cells increased significantly by 45.13% (P < .001) compared with control cells (Fig. 2e,f). These data indicate that EGFR knockdown resulted in pronounced inhibition of cell migration and invasion, and HB-EGF stimulation increased such migration and invasion dramatically, indicating the important role of EGFR in cell migration and invasion properties in 231-BR cells.

Epidermal Growth Factor Receptor Plays Less Important Roles in Cell Proliferation on 231-BR Cells in Vitro

Growth curve analysis revealed that 231-BR cell growth was slowed by EGFR knockdown, which produced the greatest reduction of 13.77% (P < .05) in 231BR-EGFR shRNA-1 cells and 9.07% (P < .05) in 231BR-EGFR shRNA-2 cells at day 4 compared with control cells. HB-EGF stimulation promoted cell growth with the greatest promotion rate of 12.16% at day 4 compared with control cells (P < .05) (Fig. 3a).

Figure 3.

Epidermal growth factor receptor (EGFR) reveals less important roles in cell proliferation on brain-seeking human MDA-MB-231 breast carcinoma (231-BR) cells in vitro. (a) A 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) proliferation assay is illustrated. (b) An anchorage-independent growth assay is illustrated. Colonies were counted after 14 days in culture. (c,d) The cell cycle was analyzed to determine cell cycle distribution and apoptosis of 231-BR cells. Single asterisk indicates P < .05 versus control; double asterisks, P < .01 versus control. shRNA indicates short-hairpin RNA; EGF, epidermal growth factor.

Anchorage-independent growth was examined to determine 3-dimensional proliferation by colony formation in soft agar. 231-BR cell colony formation was inhibited by 15.09% (P < .05) in 231BR-EGFR shRNA-1 cells and was stimulated by 16.79% (P < .05) in 231BR-control shRNA + EGF cells compared with control cells. However, colony formation did not differ significantly between 231BR-EGFR shRNA-2 cells or erlotinib-treated cells and control cells (P > .05) (Fig. 3b).

Cell cycle analysis revealed that shRNA-mediated suppression of EGFR induced mild G1-phase arrest (G1, 52.11% ± 0.99% of 231BR-EGFR shRNA-1 cells and 50.39% ± 2.17% of 231BR-EGFR shRNA-2 cells) compared with control cells (G1, 46.19% ± 1.07% of control cells); whereas, after HB-EGF stimulation, cell cycle progression was slightly promoted (G1, 41.65% ± 2.36% of cells). In addition, <1% of cells were in sub-G1 phase for all 4 cell populations, demonstrating that there was no significant apoptosis (Fig. 3c,d). Taken together, despite of the results indicating the marked involvement of EGFR in cell migration and invasion in vitro, results from the MTT assay, soft agar growth, and cell cycle analysis demonstrated that 231-BR cell proliferation was affected only mildly by EGFR knockdown and HB-EGF stimulation.

Epidermal Growth Factor Receptor Is Involved in Cell Migration and Invasion Through Cytoskeletal Reorganization of 231-BR Cells

Cancer cell motility and invasion involves a complex and integrated series of events that are controlled primarily by regulation and reorganization of the actin cytoskeleton.22, 23 Regulation of actin polymerization is responsible for the formation of protrusive structures, such as lamellipodia (large, broad membrane protrusion), filopodia (slender, spiky membrane protrusion), and ruffle formation (a membrane protrusion that does not form focal adhesions), that are essential for tumor cell movement and invasion. In the current investigation, we stained F-actin filaments to study cytoskeletal remodeling and generation of membrane protrusions. Figure 4 illustrates that, after HB-EGF stimulation, 231BR-control shRNA + EGF cells displayed more membrane ruffles, lamellipodia, and filopodia extensions than control cells. In contrast, EGFR knockdown induced less formation of protrusive structures, indicating that EGFR may drive invasion and migration through a mechanism mediated by cytoskeletal reorganization.

Figure 4.

Epidermal growth factor receptor (EGFR) is involved in cell migration through cytoskeletal reorganization of brain-seeking human MDA-MB-231 breast carcinoma (231-BR) cells. F-actin filaments were stained by rhodamine-conjugated phalloidin, and fluorescence images were obtained by confocal microscopy. Red arrow, membrane ruffles; white arrows, lamellipodial; yellow arrow, filopodia (scale bar = 10 μm). shRNA indicates short-hairpin RNA; EGF, epidermal growth factor; DAPI, 4′,6-diamidino-2-phenylindole; EGFP, enhanced green fluorescent protein.

Western Blot Analysis of Proteins Involved in Epidermal Growth Factor Receptor Downstream Signaling Pathways

To further assess the downstream effects of the signaling pathways involved in EGFR knockdown and stimulated 231-BR cell function, the expression and activation of proteins involved in signaling downstream of EGFR were examined by Western blot analysis. Much lower expression of total EGFR and markedly decreased levels of EGFR phosphorylation (at tyrosine residue 845 [Tyr845], Tyr992, and Tyr1068), phosphorylated AKT, phosphorylated 40-kDa proline-rich Akt substrate (PRAS40), and phosphorylated phospholipase C γ1 (PLCγ1) were observed in 231BR-EGFR shRNA-1 cells; and significantly increased levels of phosphorylated EGFR, phosphorylated AKT, phosphorylated PRAS40, and phosphorylated PLCγ1 were observed in 231BR-control shRNA + EGF cells compared with control cells. However, phosphorylated extracellular signal regulated kinase (ERK) and phosphorylated p38 were not affected by EGFR knockdown or stimulation in 231-BR cells (Fig. 5).

Figure 5.

These are images from immunoblot analyses of the proteins involved in epidermal growth factor receptor (EGFR) downstream signaling pathways. Brain-seeking MDA-MB-231 breast carcinoma (231-BR) cells were lysed for immunoblot analysis. Proteins were analyzed for phosphorylated EGFR (p-EGFR) (tyrosine 845 [Tyr845], Tyr992, and Tyr1068), EGFR, phosphorylated protein kinase B (p-AKT) (serine 473 [Ser473]), AKT, phosphorylated 40-kDa proline-rich Akt substrate (p-PRAS40) (threonine 246 [Thr246]), PRAS40, phosphorylated phospholipase C γ1 (p-PLCγ1) (Tyr771), PLCγ1, phosphorylated extracellular signal regulated kinases 1 and 2 (p-ERK1/2), ERK1/2, phosphorylated p38 kinase (p-p38), p38, and tubulin. The prefix “p-” refers to phosphorylation of the residues in parentheses. shRNA indicates short-hairpin RNA; EGF, epidermal growth factor.

Characterization of the In Vivo Brain Metastases of Breast Cancer Brain-Seeking Cells

A mouse xenograft model was used to quantify the effect of EGFR knockdown on in vivo brain metastases. The 231BR-EGFR shRNA-1 clone, which had greater EGFR knockdown efficiency, was selected for the in vivo studies.

Ex vivo EGFP images of whole brains and sectioned brains from mice that were injected with 1 of the 2 cell lines revealed discrete EGFP foci throughout the brain (Fig. 6a). Mice that were injected with 231BR-EGFR shRNA-1 cells had significantly fewer metastatic foci with a 2.8-fold reduction of whole-brain EGFP intensity and a 2.2-fold reduction of sectioned-brain EGFP intensity compared with mice that were injected with 231BR-control shRNA cells (P < .05) (Fig. 6a, Table 1).

Figure 6.

The brain-seeking human MDA-MB-231 breast carcinoma (231-BR) cell in vivo model system for quantification of brain metastases is illustrated. (a) Ex vivo enhanced green fluorescent protein (EGFP) brain images were captured using the Maestro EX In vivo Spectral Imaging System (Cambridge Research & Instrumentation, Woborn, Mass). These representative images show (Left) whole-brain and (Right) sectioned-brain images from (Top) 231BR-control short-hairpin RNA (shRNA) and (Bottom) 231BR-epidermal growth factor receptor (EGFR) shRNA-1 groups. (b) These are representative hematoxylin and eosin (H&E)-stained sections from a mouse brain that contain lesions from (Left) large metastases) and (Right) micrometastases (original magnification, ×200). (c,d) In vivo expression of (Top) EGFR and (Bottom) Ki-67 is illustrated in 231-BR brain metastases foci. (c) Representative images of brain-metastatic lesions from each experimental group are indicated by brown staining; nuclei are counterstained purple with hematoxylin (original magnification, ×400). (d) Graphs indicate the percentage of EGFR-positive and Ki-67-positive cells compared with the control model. Triple asterisks indicate P < .001 versus control.

Table 1. In Vivo Analysis of Control and Epidermal Growth Factor Receptor Knockdown MDA-MB-231-BR Clones
  Mean±SD
Cell LineNo. of MiceMicrometastases (≤50 μm2)Large Metastases (>50 μm2)Whole-Brain EGFP Intensity (Counts/s/mm2)Sectioned-Brain EGFP Intensity (Counts/s/mm2)
  • Abbreviations: EGFP, enhanced green fluorescent protein; EGFR, epidermal growth factor receptor; SD, standard deviation; shRNA, short-hairpin RNA.

  • a

    Mice were injected with 2 × 105 231BR-control shRNA cells or with 231BR-EGFR shRNA-1 cells through the left cardiac ventricle. Four weeks later, the mice were killed for ex vivo brain imaging and immunohistology. Fluorescence intensity and the number of metastases were determined as described in the text (see Materials and Methods).

  • b

    P < .01 versus control.

  • cP < .05 versus control.

231BR-control shRNA15102.4±31.875.13±2.622.28±0.442.95±0.54
213BR-EGFR shRNA-11558.34±27.91b2.60±1.83b0.81±0.49a1.32±0.53a

The number of large metastases (>50 μm2) and micrometastases (≤50 μm2) in H&E-stained brain sections was counted as previously described.20 The results indicated that most of the metastases in each section could be classified as micrometastases, comprising few tumor cells (Fig. 6b). Mice that were injected with 231BR-EGFR shRNA-1 cells had a 1.8-fold decrease in the number of micrometastases and a 2-fold decrease in the number of large metastases compared with the control model (P < .01) (Table 1).

Immunohistochemical analysis revealed that the expression of total EGFR was decreased drastically in the EGFR knockdown group compared with the control group (P < .001) (Fig. 6c,d). Ki-67-positive cells were measured as the percentage of nuclei in tumor regions, and the percentages were 42.09% ± 11.14% in the 231BR-EGFR shRNA model, indicating no statistical difference compared with the control group (46.12% ± 9.85%; P > .05) (Fig. 6c,d) and no difference in proliferation activity between the 2 groups. Collectively, the inhibition of brain-metastatic colonization of the EGFR knockdown 231-BR cell line was paralleled by reduced EGFR staining in the remaining lesions in vivo, confirming the EGFR contribution to in vivo brain metastases in 231-BR cells.

DISCUSSION

EGFR overexpression has been reported as an important molecular characteristic of BMBC. It also has been reported that EGFR inhibition by neutralizing antibodies or lapatinib (a dual inhibitor of EGFR and HER2 tyrosine kinases) significantly reduced BMBC in animals.7, 20 Palmieri et al9 demonstrated that the transfection of HER2 into 231-BR cells activated EGFR signal transduction, with EGFR as a leading heterodimerization partner, and increased brain-metastatic colonization, providing in vivo evidence that EGFR signaling activation promotes BMBC. In the current study, 2 independent EGFR-knockdown, brain-trophic 231-BR cell lines were developed by shRNA transfection and were used to investigate the functional roles and underlying mechanisms of EGFR in BMBC progression, providing direct evidence that EGFR can alter the brain-metastatic phenotype of breast cancer cells.

We observed that EGFR played very important roles in 231-BR cell migration/invasion in vitro and in brain metastases in vivo. Surprisingly, we noted that EGFR had less effect on cell proliferation, although EGFR is potently mitogenic for many normal and tumor cells. This finding supports previous work indicating that cells either migrate or proliferate, but not both at the same time, a phenomenon termed the migration-proliferation dichotomy.24-26 The determining factors of the cellular choice between migration and proliferation are unknown. In the current study, a set of signals (PLCγ1, AKT, PRAS40) was activated, coupled with F-actin remodeling within pseudopods at the leading edge in EGFR-stimulated 231-BR cells, and the signals were significantly inhibited coupled with less formation of protrusive structures in EGFR knockdown 231-BR cells. However, another set of signals (ERK1/2, p38) was not significantly affected by EGFR knockdown or stimulation.

The major downstream effectors of the EGFR signaling pathway include the phosphoinositide 3-kinase (PI3K)/AKT, Ras/mitogen-activated protein kinase (RAS/MAPK), and PLCγ/PKC pathways. It has been demonstrated that activation of the Akt and PLCγ plays an important role in cell migration through cytoskeletal reorganization,27, 28 which is a fundamental process that is required during in vivo metastases. The MAPK family, including ERK1/2 and p38, reportedly is involved mostly in cellular proliferation and differentiation.29 EGFR inhibition would be expected to decrease the activation of ERK1/2 and p38, but their activation was affected less by EGFR knockdown in our study, possibly because of an activating mutation of k-Ras (an upstream effector of MAPK) in the MDA-MB-231 cell line.30 In view of this, it is possible that EGFR may contribute to the invasive and metastatic properties of breast cancer cells to the brain mainly through the AKT and PLCγ1 downstream pathways. These findings provide evidence of the potential value of combining an EGFR inhibitor and an MAPK pathway inhibitor to prohibit BMBC. Our group previously demonstrated that combined treatment with a mammalian target of rapamycin (PI3K/AKT downstream) inhibitor and a mitogen-activated protein kinase kinase (MAPK upstream) inhibitor produced more antibrain metastasis effects than single-drug treatment.21

Animal model systems are lacking for the study of BMBC progression. The triple-negative MDA-MB-231 human breast carcinoma cell line has served as the mainstay of brain metastasis studies.1 Studies have indicated that triple-negative breast cancer (TNBC) (estrogen receptor negative, progesterone receptor negative, and HER2 unamplified) is significantly more likely to develop brain metastasis.31-33 Higher EGFR expression has been reported in TNBC.34, 35 Therefore, an MDA-MB-231 brain-seeking derivative (231-BR cells) was used to investigate the function of EGFR in BMBC in the current study. The 231-BR model mimics clinical cases, because the model develops multiple brain metastases in mouse brain. Additional studies in other model systems may reveal other new molecular mechanisms that contribute to BMBC, and such a study is currently underway in our laboratory.

Along with EGFR, HER2,9 HER3,36 and vascular endothelial growth factor37 also reportedly are up-regulated in BMBC. More recent studies have focused on the significant association between HER2 and brain metastasis in patients with HER2-positive breast cancer.1, 9, 38, 39 Our current study indicates that EGFR may be a potential molecular target of BMBC, especially for the subgroup of patients with the triple-negative phenotype that overexpresses EGFR. Currently, there are no accurate diagnostic methods for determining the level of EGFR expression in a tumor; therefore, the clinical benefits from anti-EGFR therapies for breast cancer are limited. More work is required for the precise determination of EGFR status to improve the clinical outcome of EGFR-targeting agents.

Previously, EGFR and its ligands were linked to breast cancer infiltration of the lungs, but not of the bones or liver.11, 40 However, cell intravasation into the brain, which is protected by a unique and complex blood-brain barrier, appears to be distinct from that into other organs.

In summary, the current results demonstrate that EGFR makes more significant contributions to brain-seeking breast cancer cell invasiveness in vitro and to brain metastases in vivo and that EGFR makes a lesser contribution to cell proliferation in vitro and in vivo. These findings suggest that EGFR may be a potential diagnostic and therapeutic target for BMBC.

Acknowledgements

We thank Dr. Patricia S. Steeg (National Cancer Institute, Bethesda, MD) for the 231-BR cell line.

FUNDING SOURCES

Dr. Teng has received research funding from the National Natural Science Foundation of China (NSFC grants 30910103905 and 30870703).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary