Imatinib as a key inhibitor of the platelet-derived growth factor receptor mediated expression of cell surface heparan sulfate proteoglycans and functional properties of breast cancer cells

Authors

  • Christina J. Malavaki,

    1. Laboratory of Biochemistry, Department of Chemistry, University of Patras, Greece
    Current affiliation:
    1. Department of Pharmacy, Laboratory of Molecular Pharmacology, University of Patras, Greece
    Search for more papers by this author
  • Andreas E. Roussidis,

    1. Laboratory of Biochemistry, Department of Chemistry, University of Patras, Greece
    Search for more papers by this author
  • Chrisostomi Gialeli,

    1. Laboratory of Biochemistry, Department of Chemistry, University of Patras, Greece
    2. Foundation for Research and Technology/Institute of Chemical Engineering Sciences, Patras, Greece
    Search for more papers by this author
  • Dimitris Kletsas,

    1. Laboratory of Cell Proliferation and Ageing, Institute of Biosciences and Applications, National Centre for Scientific Research‘Demokritos’, Athens, Greece
    Search for more papers by this author
  • Theodore Tsegenidis,

    1. Laboratory of Biochemistry, Department of Chemistry, University of Patras, Greece
    Search for more papers by this author
  • Achileas D. Theocharis,

    1. Laboratory of Biochemistry, Department of Chemistry, University of Patras, Greece
    Search for more papers by this author
  • George N. Tzanakakis,

    1. Laboratory of Histology, Medical School, University of Crete, Heraklion, Greece
    Search for more papers by this author
  • Nikos K. Karamanos

    Corresponding author
    1. Foundation for Research and Technology/Institute of Chemical Engineering Sciences, Patras, Greece
    • Laboratory of Biochemistry, Department of Chemistry, University of Patras, Greece
    Search for more papers by this author

Correspondence

N. Karamanos, Laboratory of Biochemistry, Department of Chemistry, University of Patras, 26110 Patras, Greece

Fax: +30 2610 997153

Tel: +30 2610 997915

E-mail: n.k.karamanos@upatras.gr

Abstract

Cell surface heparan sulfate proteoglycans (HSPGs), syndecans and glypicans, play crucial roles in the functional properties of cancer cells, such as proliferation, adhesion, migration and invasion. Platelet-derived growth factor (PDGF)/PDGF receptor (PDGF-R) mediated signaling, on the other hand, is highly associated with cancer progression. Specifically, PDGF-Rα and PDGF-Rβ expressions documented in breast cancer tissue specimens as well as breast cancer cell lines are correlated with tumor aggressiveness and metastasis. Imatinib (Glivec®) is a tyrosine kinase inhibitor specific for PDGF-Rs, c-ΚΙΤ and BCR-ABL. In this study we evaluated the effects of imatinib on the properties of breast cancer cells as well as on the expression of HSPGs in the presence and absence of PDGF-BB. These studies have been conducted in a panel of three breast cancer cell lines of low and high metastatic potential. Our results indicate that imatinib exerts a significant inhibitory effect on breast cancer cell proliferation, invasion and migration as well as on the cell surface expression of HSPGs even after exposure of PDGF. These effects depend on the aggressiveness of breast cancer cells and the type of HSPG. It is suggested that imatinib may be of potential therapeutic usefulness in breast cancer regimes.

Abbreviations
ECM

extracellular matrix

GAPDH

glyceraldehyde-3-phosphate dehydrogenase

HSPG

heparan sulfate proteoglycan

MAPK

mitogen-activated protein kinase

PDGF

platelet-derived growth factor

PDGF-R

PDGF receptor

PI3K

phosphatidylinositol 3-kinase

Introduction

Breast cancer is the most common type of cancer in women worldwide and the incidence has constantly increased in recent years. In response to this, the scientific community has focused interest on discovering new molecular targets for this disease. During previous decades a variety of therapies were proposed for the treatment of breast cancer, but the majority of chemotherapeutic agents have not managed to inhibit cell proliferation or invasion effectively. The recent new developments in drug discovery and molecular-targeted therapies are aimed at this problem, and they have produced improvements in patients' response and survival with fewer side effects, leading to an improvement in quality of life [1].

Growth factors mediate their signal starting the activation of cell surface transmembrane receptors to intracellular effectors that control critical functions in cancer cells [2-4]. Platelet-derived growth factor (PDGF) is a mitogenic growth factor able to mediate differentiating, proliferating and migrating roles in developing and developed cells [5, 6]. The members of the PDGF family (PDGF-AA, -BB, -AB, -CC and –DD) exert their effect by binding to two structurally related receptor tyrosine kinases, denoted PDGF-Rα and PDGF-Rβ [7]. PDGF binding induces dimerization, both homo- and hetero-dimerization of the receptors, leading to their autophosphorylation on tyrosine residues and binding of downstream signaling molecules. After receptor activation, several intracellular pathways are stimulated, leading to cell proliferation and several other crucial processes [8, 9]. In general, aberrant tyrosine kinase activity, and/or tyrosine kinase overactivity, is associated with human cancer and other hyperproliferative diseases. Specifically, PDGF-R signaling has important functions during embryogenesis, and its overexpression is associated with several pathological conditions such as fibrotic and vasculoproliferative diseases and cancer [10-13].

In some solid tumors, the PDGF-R-mediated growth stimulation of cancer cells is well known [14]. Expression of PDGF-Rα and PDGF-Rβ has been detected in tissue specimens from different breast cancer types and in breast cancer cell lines [15]. Brown et al. [16] studied the expression of PDGF-Rα, PDGF-Rβ and PDGF-AB in three hormone-negative breast cancer cell lines (SKRB-3, MDA-175, MDA-MB-231) and detected both receptors in all three cell lines, PDGF-AB in high amounts in MDA-175 and a faint reactivity in the other two cell lines [16]. Studies in breast tumors and surrounding stroma cells indicate that the expression of PDGF-Rs is implicated in breast cancer. PDGF-Rα has been associated with tumor aggressiveness and PDGF-Rβ with tumor metastasis. Even when the expression has not been established in tumor cells, the surrounding stroma seems to express PDGF-Rs indicating not only an autocrine but also a paracrine activation and regulation of PDGF-Rs [14].

The extracellular matrix (ECM) is a three-dimensional scaffold of glycoproteins, proteoglycans and glycosaminoglycans that provides not only structural support to the host tissue but also signals cells maintaining tissue homeostasis and functions [17]. The key representative ECM macromolecules in cancer progression are the ECM-interacting cell surface proteoglycans. Syndecans and glypicans, type I transmembrane and glycosyl-phosphatidylinositol anchored to the cell membrane respectively, belong to the heparan sulfate proteoglycans (HSPGs) family. They are mediators of cell–cell and cell–ECM interactions and interact with various factors regulating cell signaling activity and subsequently various cellular properties [18-20]. Breast cancer is characterized by significant quantitative changes of extracellular network components. Consequently, changes in unique ECM properties of tumor cells and their microenvironment may lead to changes in cancer cell behavior during cancer progression [18, 21]. It has been reported that PDGF-mediated signaling coordinates ECM expression and vice versa [22, 23].

Several strategies for inhibition of aberrant tyrosine kinase activity, such as antisense oligonucleotides, antigenic stimulation and small molecular inhibitors, have been developed [24]. However, imatinib (Glivec®, imatinib mesylate), a 2-phenylaminopyrimidine derivate, is an ATP competitive inhibitor of several tyrosine kinases, such as BCR-ABL, c-KIT and PDGF-Rs, which regulate major cellular events in a number of solid tumors. This drug is already used for the treatment of chronic myelogenous leukemia [25] and gastrointestinal stromal tumors [26] and has shown encouraging in vitro results in colon [27] and small cell lung cancer cell lines. However, the clinical trials in colon and small cell lung cancer were not encouraging [28, 29].

With respect to the emerging data on the role of PDGF-R in cancer progression and the importance of cell surface proteoglycans, we found it important to evaluate the effects of imatinib, a selective inhibitor of PDGF-R, on PDGF/PDGF-R-mediated cellular functional properties and the expression of HSPGs in a panel of low and highly invasive breast cancer cells.

Results

Expression of PDGF-Rs and ligands in breast cancer cells

The expression of PDGF-Rs was evaluated at mRNA level by RT-PCR and at protein level by western blotting. As shown in Fig. 1A,B both PDGF-R isoforms (PDGF-Rα and PDGF-Rβ) are expressed in all epithelial breast cancer cell lines (MCF-7, MDA-MB-231 and ZR-75-1) and normal mammary cells (MCF-12A). It has been previously reported [30] that these three breast cancer cell lines express the ligand PDGF-BB; here we report that PDGF-AA is also expressed in all cell lines tested (Fig. 1C).

Figure 1.

(A) RT-PCR and (B) western blotting analysis showing the gene and protein expression, respectively, of PDGF-Rα and PDGF-Rβ, and (C) the gene expression of PDGF-AA by RT-PCR in normal mammary MCF-12A cells and breast cancer cells (MDA-MB-231, MCF-7 and ZR-75-1). (D) The localization of PDGF-Rα, -Rβ is also assessed by immunocytochemical detection in all breast cell lines.

Further analysis regarding the expression of PDGF-Rs was carried out by immunocytochemical staining using specific antibodies against PDGF-Rα and PDGF-Rβ. As shown in Fig. 1D, both PDGF-Rα and PDGF-Rβ were detectable in all cell lines in variable amounts. The profiles obtained are in accordance with the expression studies at transcriptional and protein levels.

PDGF-BB enhances breast cancer cell proliferation

The presence of the PDGF-Rs and particularly of PDGF-Rβ in its membrane form as well as the receptors ligands (PDGF-AA and PDGF-BB) highlights the importance of the PDGF/PDGF-R signaling axis in the growth and functional properties of breast cancer cells. We therefore evaluated the effects of PDGF-BB – this ligand can activate both PDGF-Rα and PDGF-Rβ receptor types – on cell proliferation.

The addition of exogenous PDGF-BB (0.5–10 ng·mL−1) in serum-free culture medium induces cell proliferation of all breast cell lines in a mode dependent on the cell line and the ligand concentration used (Fig. 2A). The stimulation was higher in MDA-MB-231, MCF-7 and ZR-75-1, whereas normal MCF-12A cells were less affected. Evaluation of the dose-dependent PDGF-BB action revealed that a concentration of 2 ng·mL−1 significantly enhances ( 0.001) the proliferation of breast cancer cells compared with 0.5 ng·mL−1, whereas a concentration of 10 ng·mL−1 slightly enhances cell proliferation compared with 2 ng·mL−1. Therefore a concentration of 2 ng·mL−1 was used for evaluating PDGF-mediated effects.

Figure 2.

(A) Effects of increasing concentrations of PDGF-BB (0.5, 2 and 10 ng·mL−1) on normal mammary MCF-12A cell and breast cancer cell (MDA-MB-231, MCF-7 and ZR-75-1) proliferation. The results are expressed as a percentage of the control. Values were estimated as the mean ± SD of three separate experiments in triplicate. (B) Effect of PDGF-BB (2 ng·mL−1) on the activation of PDGF-Rβ and (C) the signaling effectors Akt and Erk1/2 in breast cancer cells. Statistically significant differences among the control cells and PDGF-treated cells are shown by † 0.01.

PDGF-BB enhances PDDF-Rβ phosphorylation and signaling

The effect of PDGF-BB on the activation of PDGF-Rα and PDGF-Rβ was examined using specific antibodies recognizing the phosphorylation of Y720 for PDGF-Rα and Y1021 for PDGF-Rβ. It is noted that PDGF-BB significantly enhances the phosphorylation of PDGF-Rβ only in ZR-75-1 and MDA-MB-231 breast cancer cells (Fig. 2B). No phosphorylation enhancement of PDGF-Rβ was found in normal mammary cells MCF-12A, whereas in MCF-7 cells the effect was not significant (data not shown). Moreover, phosphorylated PDGF-Rα was not detected either in control or in PDGF-BB-treated breast cancer cells. These data indicate that the effects of PDGF-BB on breast cancer cells of a high metastatic potential (MDA-MB-231 and ZR-75-1) are almost exclusively mediated via the activation of PDGF-Rβ.

The Ras-MEK-Erk and phosphatidylinositol 3-kinase (PI3K) Akt signaling pathways are synergistically regulated and have been shown to crosstalk in breast cancer cell lines. PI3K/Akt activity has critical roles in cell survival and proliferation, whereas high Erk activity induces growth arrest and differentiation [30]. As reported, the breast cell lines used in the present study exhibit elevated Erk activity constitutively [31]. The PDGF-BB/PDGF-Rβ signaling axis was further evaluated at the level of PI3K/Akt and p44/p42 (Erk1/2) intracellular proteins, which are related to cell survival and growth. As shown in Fig. 2C, all breast cancer cell lines exhibit constitutively phosphorylated Akt and Erk1/2; MDA-MB-231 and ZR-75-1 exhibit a higher constitutive activation of Akt compared with MCF-7. It should be noted that, although PDGF-BB does not further enhance the phosphorylation of Akt in all cell lines, treatment with PDGF-BB resulted in a significant upregulation of Erk phosphorylation in the highly metastatic MDA-MB-231 cells and in ZR-75-1 cells. No significant effect was observed for the MCF-7 cells (Fig. 2C). These data suggest that the PDGF-BB/PDGF-Rβ-mediated signaling resulting in the enhancement of Erk1/2 phosphorylation may well be attributed to the survival and enhancement of cell proliferation of breast cancer cells.

Imatinib exerts a G2/M cytostatic activity and inhibits PDGF-mediated cell proliferation

We have previously reported [32] that imatinib significantly inhibits cell growth of breast cancer cells. This was further confirmed by Weigel et al. [33] for a panel of breast cancer cell lines, where a remarkable apoptosis was noted for long-term treated cells (> 72 h). It should be noted that following a 24 h cell culture in the presence of imatinib apoptosis lower than 1% was documented. Moreover, flow cytometric cell cycle analysis in all three breast cancer cell lines showed that imatinib inhibits cell proliferation exhibiting a significant cytostatic increase of the G2/M phase with a significant decrease at the S phase (Table 1).

Table 1. Flow cytometric cell cycle analysis of breast cancer cell lines with treatment of imatinib for 24 h. An asterisk indicates a statistically significant difference between the imatinib-treated cells and control cells:  0.01
 G0/G1SG2/M
ZR-75-1
Control51.94%36.63%11.43%
Imatinib54.52%18.41%27.07%
% differenceMinor−50%*+136%*
MCF-7
Control68.06%19.96%11.99%
Imatinib67.52%9.01%23.47%
% differenceMinor−54%*+96%*
MDA-MB-231
Control48.48%34.46%17.06%
Imatinib46.73%30.11%23.16%
% differenceMinor−13%*+35%*

It has been reported [16] that imatinib produces growth inhibition correlated with the PDGF family. Taking into consideration the effectiveness of the PDGF-BB/PDGF-Rβ system, pointed out above, we further evaluated the effects of imatinib on cell proliferation and PDGF-Rβ phosphorylation and intracellular signaling of the PDGF-BB-treated breast cancer cells. Notably, imatinib, apart from its direct inhibitory effect on cell proliferation, significantly inhibits the PDGF-BB-mediated stimulation of breast cancer cell proliferation in a dose-dependent manner. This inhibitory effect was profound even at the lower imatinib concentration used (3 μm) in the present study (Fig. 3). These results may well be correlated with the data reported previously by Weigel et al. [33] indicating that the activation of PDGF-Rβ in breast cancer cell lines is significantly suppressed by imatinib. However, taking into consideration that the phosphorylation/activation of PDGF-Rβ by PDGF-BB depends on the breast cancer cell line, it is plausible to suggest that alternative pathways may also be responsible for the PDGF-BB-mediated effects on cell proliferation.

Figure 3.

Effects of PDGF-BB (2 ng·mL−1), imatinib (3 μm and 10 μm) and their combination on normal mammary MCF-12A cell and breast cancer cell (MDA-MB-231, MCF-7 and ZR-75-1) proliferation. The results are expressed as mean ± SD of three separate experiments in triplicate. Statistically significant differences among the imatinib-treated cells and control cells as well as (PDGF-BB + imatinib)- and PDGF-treated cells are shown by † 0.01 and * 0.01, respectively.

Imatinib modulates the PDGF-BB-mediated expression of cell surface HSPGs

Cell surface HSPGs play critical roles in cancer cell properties [19]. Following pilot experiments in our laboratories we examined the involvement of cell surface HSPGs in breast cancer malignancy under various treatment conditions and particularly the syndecans-2 and -4 as well as glypican-1. It was found that the mRNAs encoding for syndecan-2, syndecan-4 and glypican-1 are all expressed in the normal and breast cancer cell lines tested (Fig. 4A). As shown in Fig. 4A, syndecan-2 and syndecan-4 are overexpressed in the highly aggressive MDA-MB-231 breast cancer cell line compared with the low metastatic MCF-7 breast cancer cells and normal mammary cells. On the other hand, glypican-1 mRNA is significantly higher in both breast cancer cell lines (MDA-MB-231 and MCF-7) compared with normal cells (Fig. 4A). The expression of these HSPGs at protein level in breast cancer cell lines, previously reported by Dedes et al. [34], was in agreement with the profiles obtained by RT-PCR. In order to evaluate the effects of PDGF-BB in the expression of syndecans-2 and -4 and glypican-1, breast cancer cells were treated with PDGF-BB, imatinib and their combination. Notably PDGF-BB affected the expression of these proteoglycans depending on the cell line and the type of HSPG (Fig. 4B–D). In particular, gene expression of syndecan-2 is not affected by PDGF-BB in any cell line (Fig. 4B). On the other hand, the expression of syndecan-4 is suppressed in MDA-MB-231 cells (Fig. 4C), whereas that of glypican-1 is stimulated only in the highly metastatic MDA-MB-231 cells (Fig. 4D). The inhibitory effect on syndecan-4 by PDGF-BB may be related to an altered adhesive profile of treated cells, whereas the stimulatory effect of PDGF-BB on glypican-1 is likely to be correlated with the aggressive and/or migration ability of MDA-MB-231 cells.

Figure 4.

(A) Gene expression profiles of HSPGs syndecan-2, -4 and glypican-1 by normal mammary MCF-12A cells and breast cancer cells (MDA-MB-231 and MCF-7). The results are expressed as mean ± SD of separate experiments in triplicate. Statistically significant differences among the breast cancer cells and normal MCF-12A cells are shown by * 0.001. Effects of PDGF-BB (2 ng·mL−1), imatinib (3 μm and 10 μm) and their combination on gene expression of (B) syndecan-2, (C) syndecan-4 and (D) glypican-1 at mRNA levels by MDA-MB-231 and MCF-7 breast cancer cells. RT-PCR is shown in the upper panel and the results of the semiquantitative analysis are presented below in the histogram. The results are expressed as mean ± SD of separate experiments in triplicate. Statistically significant differences among the imatinib-treated cells and control cells as well as (PDGF-BB + imatinib)- and PDGF-treated cells are shown by † 0.01 and * 0.01, respectively.

On the other hand, the activity of imatinib on the expression of HSPGs in the PDGF-BB- treated breast cancer cells is mainly inhibitory. In particular, treatment with imatinib reduced the expression of syndecan-2 (Fig. 4B), syndecan-4 (Fig. 4C) and glypican-1 (Fig. 4D) both in MCF-7 and MDA-MB-231. Although the gene expression studies presented here should be further evaluated at the protein level, these data clearly indicate that imatinib is a potent inhibitor of the PDGF-induced expression of HSPGs.

Effects of PDGF-BB and imatinib on cellular invasion and migration

Taking into account that PDGF-BB and imatinib affect the expression of cell surface proteoglycans known to contribute to the ability of cancer cells to grow, migrate and invade [18], we further evaluated the effects of imatinib in cell growth/confluence, invasion and migration. We first examined the effects of imatinib on cell culture confluence following long-term treatment of breast cancer cells with imatinib for 16 days, with culture medium and imatinib replacement every 24 h. As shown in Fig. 5A, cell culture confluence was gradually and effectively inhibited by imatinib, indicating that the inhibitory effects of imatinib on the expression of cell surface HSPGs, implicated in cell–cell and cell–matrix interactions, may account for such growth effects.

Figure 5.

(A) Effects of imatinib (10 μm) on ZR-75-1 breast cancer cell growth/confluence. Representative images are given following treatment for 11 and 16 days. (B) Effects of PDGF-BB (2 ng·mL−1) and imatinib (10 μm) plus PDGF-BB on the invasive ability of MDA-MB-231 and MCF-7 breast cancer cells. Cells were preincubated for 90 min with or without imatinib and then PDGF-BB was added for 24 h. Cells were harvested, resuspended in serum-free medium and loaded on the upper chamber, which contained a layer of ECM matrix. Medium supplemented with 10% FBS was used as a chemoattractant. Invasion time was 24 h. (C) Effects of PDGF-BB (2 ng·mL−1) and imatinib (10 μm) plus PDGF-BB on the motility of MDA-MB-231 and MCF-7 breast cancer cells. Migration profiles were obtained following the wound healing assay, presented as the percentage of wound closure. The results are expressed as mean ± SD of separate experiments in triplicate. Statistically significant differences among the imatinib-treated cells and control cells as well as (PDGF-BB + imatinib)- and PDGF-treated cells are shown by † 0.01 and * 0.01, respectively.

We have previously reported [32, 35] that imatinib dose-dependently inhibits the breast cancer cellular invasion on matrigel. Here, we further examined whether PDGF-BB and imatinib affect the invasion of human breast cancer cells. For this purpose the highly invasive MDA-MB-231 and low invasive MCF-7 cells were cultured in the absence or presence of imatinib and PDGF-BB and their ability to invade an ECM membrane was determined by a cell invasion assay, as described in 'Materials and methods'. As shown in Fig. 5B, PDGF-BB increases the invasion of the highly metastatic MDA-MB-231 breast cancer cell line (~30%) within 24 h. Notably, imatinib decreases the invasive potential of MDA-MB-231 cells (~25%), but its suppressive effect is more profound for the PDGF-BB-enhanced invasion (~45%). Although PDGF-BB has no significant effect in the invasiveness of the low metastatic MCF-7 cells, imatinib resulted in a decrease (~25%) of the invasion of these cancer cells (Fig. 5B). The inhibitory effects of imatinib on cell invasion were in accordance with those obtained for their ability to migrate. In terms of cell migration, imatinib exhibited an inhibitory profile for both concentrations tested. The similarly reported effect of imatinib, 3 and 10 μm, low and high concentration, may indicate that the action of imatinib has reached a ‘plateau’. The observed inhibition was further maintained even under the action of PDGF-BB, which induced cell migration (Fig. 5D).

Discussion

Aberrant regulation of growth factors and their receptors has been shown to be involved in many diseases such as cancer, cardiovascular diseases and developmental defects. Breast carcinomas are known to express PDGF; however, the availability of PDGF-Rs on the target cell is naturally crucial to the PDGF-dependent cellular responses and is subject to regulation. For this reason, we evaluated the expression of the mature forms of PDGF-Rα and PDGF-Rβ in breast cancer cell lines. All cell lines tested expressed PDGF-AA and PDGF-BB ligands of PDGF-Rs. Thus, treatment of cells with PDGF-BB affected PDGF-R activation levels. The cellular effects of the PDGF-R signaling pathway are mainly mediated by the Erk/MAPK and PI3K/Akt signaling cascade [13, 23, 36-38]. In all breast cancer cell lines tested, both Erk1/2 and Akt were constitutively phosphorylated, while PDGF-BB treatment resulted in further enhancement of Erk1/2 activation, especially in the highly metastatic MDA-MB-231 cells. Moreover, activation of PDGF-Rs from their natural ligand results in enhanced cell proliferation, cellular invasion and migration indicating that PDGF-BB-mediated signaling plays an important role in breast cancer progression.

We have previously shown that imatinib is a powerful inhibitor of cell proliferation and invasion of breast cancer epithelial cells [33, 37]. In vitro preclinical investigation on cancer cell lines encourages the thorough study of tyrosine kinase pathways that signal cancer cells, especially when the overall effect of imatinib is inhibitory on cell growth progression, in clinically relevant experiments [33]. Our studies further indicated a significant cytostatic increase of the G2/M phase with a significant decrease at the S phase in the presence of imatinib, while the inhibitory action on breast cancer cell growth was significantly apparent even in PDGF-BB-treated cells. The inhibitory action of imatinib is in accordance with downregulated PDGF-Rβ phosphorylation activation as well as Akt signal transduction, as reported by other research groups [33]. However, in other studies in multiple myeloma, imatinib is reported to induce the phosphorylation levels of Erk1/2 dose-dependently, while at the same time exhibiting inhibition of cell proliferation [39].

At the functional level our studies revealed that the invasiveness, assayed by the ECM-like trans-well system, is affected by the action of PDGF-BB depending on the cell line. It is well known that MDA-MB-231 cells exhibit higher invasion index than MCF-7 [34]. In this study, we have shown that PDGF-BB mediates the activation of PDGF-Rβ mainly in MDA-MB-231 cells and only slightly in MCF-7 cells. The PDGF-enhanced invasive potential of MDA-MB-231 cells is significantly suppressed, following PDGF-R blocking of activation by imatinib. The significant suppressive effect of imatinib on PDGF-induced invasion is in agreement with the receptor activation in these cells. The effects of both PDGF-BB and imatinib on migration were suppressive to a relatively low degree. In previous studies by other research groups, the importance of PDGF-Rβ activity for migration and invasion of medulloblastoma cells via transactivating EGFR, using imatinib as a selective inhibitor of the receptor's activity, was highlighted [40]. It is therefore plausible to suggest that other pathways may also be implicated in the invasion and migration profiles of these breast cancer cells.

At the level of matrix macromolecules implicated in breast cancer migration, adhesion and progression we demonstrated the effect of imatinib on the gene expression level of certain HSPGs, key cell surface effectors. Specifically, analysis of cell surface HSPG expression by RT-PCR analysis showed significant differences between the normal and cancer cells as well as between the highly invasive and the low invasive cell lines. The differential expression of HSPGs between the normal and cancer cell lines may indicate a correlation between the expression of specific HSPGs and the invasiveness and migration of breast cancer cells [41-44]. Comparing the two types of cancer cell lines with different invasiveness it can be seen that the expression of syndecan-4 and the overexpression of syndecan-2 are associated with high invasiveness, i.e. MDA-MB-231. The expression of glypican-1 gene in the MDA-MB-231 cells may be indicative of higher metastatic potential. Furthermore, treatment of the breast cancer cells with imatinib resulted in inhibition of the PDGF-BB-mediated expression of HSPGs, associated with the observed results of the long-term effect of imatinib on breast cancer, as well as with its inhibitory effect on the invasive and migratory potential of breast cancer cells.

It seems that the overall effect of imatinib depends on the total expression of tyrosine kinases and thus the potency of imatinib depends on which pathway is overactivated in a particular cancer type. The overall data, illustrated in Fig. 6, indicate the regulatory role of PDGF-R in breast cancer and at the same time the possibility that specific inhibitors against these receptors, such as imatinib, might be used for molecular treatment of breast cancer. Learning more about the function and biology of growth factors and their receptors is crucial to the understanding of the mechanisms behind these diseases and for the development of new therapies.

Figure 6.

Overall effect of imatinib on breast cancer cells. PDGF-BB enhances the activation of PDGF-Rβ and its prime signaling effector Erk1/2, whereas Akt activation is not affected. On the other hand, imatinib blocks PDGF-Rβ activation as well as breast cancer cell proliferation mediated by PDGF-BB. Moreover, imatinib impedes PDGF-induced invasive and migratory potential of breast cancer cells. At the level of matrix effectors implicated in breast cancer progression, the PDGF-mediated expression pattern of HSPGs, syndecan-2, -4 and glypican-1 is suppressed by the imatinib action. These data indicate the regulatory role of PDGF-R in breast cancer and at the same time present imatinib as a promising inhibitor of breast cancer progression.

Materials and methods

Chemicals and cell lines

Cell culture chemicals, i.e. Eagle's minimal essential medium (EMEM), Dulbecco's modified Eagle's medium (DMEM)/Ham's F12 medium (1 : 1), RPMI 1640, fetal bovine serum (FBS), l-glutamine, sodium pyruvate, sodium bicarbonate, nonessential amino acids, penicillin, streptomycin, amphoterecin B, gentamycin and HEPES were purchased from Biochrom KG (Berlin, Germany). Insulin, glucose, epidermal growth factor, cholera toxin and hydrocortisone were all obtained from Sigma Chemicals (Steinhelm, Germany). All other chemicals used were of the best available grade. The breast cancer cell lines MCF-7 (HTB 22: human breast adenocarcinoma, ERα+ of low invasive potential), MDA-MB-231 (HTB26: human breast adenocarcinoma, ERα– of high invasive potential) and ZR-75-1 (CRL-1500: human ductal carcinoma, ERα+ of high invasive potential) as well as the normal mammary breast cell line MCF-12A (CRL 10782: human mammary epithelial cells, ERα+) were obtained from the American Type Culture Collection (ATCC).

Cell cultures

MDA-MB-231 and MCF-7 breast cancer cells were cultured as monolayers at 37 °C in a humidified atmosphere of 5% (v/v) CO2 and 95% air. As has been described previously [45] EMEM was supplemented with 10% FBS, 2 mm l-glutamine, 1.0 mm sodium pyruvate, 1.5 g·L−1 sodium bicarbonate, 0.1 mm nonessential amino acids, 0.01 mg·mL−1 insulin and a cocktail of antimicrobial agents (100 IU·mL−1 penicillin, 100 μg·mL−1 streptomycin, 10 μg·mL−1 gentamicin sulfate and 2.5 μg·mL−1 amphoterecin B). ZR-75-1 cells were grown in RPMI 1640 medium containing 10% (v/v) FBS, 2 mm l-glutamine, 1.0 mm sodium pyruvate, 1.5 g·L−1 sodium bicarbonate, 10 mm HEPES, 4.5 g·L−1 glucose and the above mentioned cocktail of antimicrobial agents. Control flasks were treated with 0.1% (v/v) dimethylsulfoxide. Normal mammary MCF-12A cells were grown in DMEM/Ham's F12 (1 : 1) medium containing 5% horse serum, 20 ng·mL−1 hydrocortisone and the same cocktail of antimicrobial agents as mentioned before [46]. According to pilot experiments with respect to growth rate and doubling time, the medium was changed every 3 days. The cells were harvested after treatment with 0.25% (w/v) trypsin in NaCl/Pi containing 0.1% (w/v) Na2EDTA. All experiments were conducted in serum-free conditions. The cells prior to treatment with imatinib and/or PDGF-BB were cultured in serum-starved conditions for 24 h. Fresh serum-free medium was added to the cultures for another 12 h period. Imatinib was then added for 90 min to the serum-free cultures. PDGF-BB treatment was for 10 min in the signal transduction studies and for 24 h in the expression studies.

Cell proliferation

To evaluate the effect of PDGF-BB and imatinib on cell proliferation, cells were seeded into 24-well plates at a density of (20–30) × 103 cells per well. As has been previously described [47], 24 h after plating new medium was added in the presence or absence of PDGF-BB and imatinib at final concentrations of 0.5, 2 and 10 nm for PDGF-BB and 3 and 10 μm for imatinib. Then, 16–18 h before the experiment ended, solution containing [methyl-3H]-thymidine (100 Ci·mmol−1) in 0.1 μCi·mL−1 final concentration was added. When the experiment ended the culture medium was aspirated and the cells were washed with NaCl/Pi and fixed with ice-cold trichloroacetic acid (5% w/v) for 10 min. The culture plates were then washed extensively under running tap water and air-dried. DNA was solubilized by the addition of 0.2 mL of 1% (w/v) SDS in 0.3 m NaOH with continuous shaking. After 15 min the lysates were transferred into vials containing 2 mL of ready safe scintillation cocktail (Lumac-LSC, Groningen, The Netherlands) and subjected to scintillation counting.

Flow cytometry

Cell cycle distribution was estimated by flow cytometry. Breast cancer cells were cultured in the absence and presence of imatinib. According to the protocol previously reported [32], after trypsinization cells were washed with ice-cold NaCl/Pi, fixed in 50% (v/v) ethanol and stained for DNA content with a solution containing propidium iodide (50 mg·mL−1) and RNase (10 mg·mL−1). The fluorescence of stained cells was measured using a FACS Scan flow cytometer (Becton Dickinson, Menlo Park, CA, USA) equipped with modfit software.

RNA isolation and RT-PCR

Total cellular RNA was isolated after cell lysis with guanidium isothiocyanate using the SV total RNA isolation system (Promega GmbH, Mannheim, Germany). The amount of isolated RNA was quantified by measuring its absorbance at 260 nm. All total RNA preparations were free of DNA contamination as assessed by RT-PCR analysis and the absorbance ratios (A260/A280). Reverse transcription of RNA was performed using the Qiagen® OneStep RT-PCR kit (Qiagen GmbH, Hilden, Germany), on a Perkin-Elmer 2400 GeneAMP PCR. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was amplified as internal control. The amplification products were separated by electrophoresis on a 2% agarose gel containing Gel Star® stain (BioWhittaker, Rockland, ME, USA). Bands were visualized with a UV lamp and gels were photographed with a charge-coupled device camera. The sequences of primers as well as their prime characteristics for the genes of interest are provided in Table 2. For semiquantitative analysis the PCR products were expressed as the relative density of each molecule band compared with the band of GAPDH. Image analysis was performed using the program unidocmv version 99.03 for Windows (UVI Tech, Cambridge, UK).

Table 2. PCR primers used to amplify the genes under investigation
GeneUpstream/DownstreamBase pairs of PCR productAnnealing temperature of primers Tan (°C)Cycles
Syndecan-2 GGGAGCTGATGAGGATGTAG/CACTGGATGGTTTGCGTTCT 3946035
Syndecan-4 CTCCTAGAAGGCCGATACTTCT/GGACCTCCGTTCTCTCAAAGAT 3456035
Glypican-1 ATCACCGACAAGTTCTGGGGTA/CATCTTCTCACTGCACAGTGTC 3176035
GAPDH ACATCATCCCTGCCTCTACTGG/AGTGGGTGTCGCTGTTGAAGTC 261  

Immunocytochemistry

Slides were thawed for 15 min at room temperature before staining. All staining steps were conducted at room temperature and included 5-min washes of NaCl/Tris with Triton X-100 between the steps listed below. Cells were fixed in 3% paraformaldehyde for 15 min and then blocked with NaCl/Tris with Triton X-100, including 3% FBS. All incubations were 1 h long. Primary antibodies for PDFGF-Rα and PDGF-Rβ (1 : 50) were incubated to label all cell lines mentioned above. Secondary biotinylated multi-link polyclonal antibody (DakoCytomation, Glostrup, Denmark) was then applied at a 1 : 200 dilution. Slides were incubated with an avidin-biotin complex which was conjugated to horseradish peroxidase (DakoCytomation). Nuclear staining was performed with DAB for 5 min. Coverslips were applied after treatment with Merck Entellan® neu Mounting Media.

Western blot

Cells were lysed with ice-cold lysis buffer RIPA (50 mm Tris/HCl pH 7.3, 150 mm NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS). Cell extracts were resolved by 10% SDS/PAGE and transferred to polyvinylidene difluoride membranes (Macherey-Nagel, Düren, Germany). Membranes were blocked and incubated for 16–20 h at 4 °C with primary antibody recognizing the certain ECM molecule each time. The immune complexes were detected after incubation with the appropriate peroxidase-conjugated secondary antibody with the SuperSignalWest Pico Chemiluminescent substrate (Pierce). Protein expression of each molecule was expressed as relative intensity, normalized to α-tubulin.

Invasion and migration studies

To determine whether imatinib and PDGF-BB affect the invasiveness potential of the cancer cell lines MDA-MB-231 and MCF-7, a commercially available cell invasion assay (Chemicon International Inc.) and migration assay (Ibidi) were used according to the manufacturer's instructions. Cells were incubated in the presence or absence of 10 μm imatinib and 2 pg·mL−1 PDGF-BB at 37 °C under 5% CO2. After 48 h, the cells from the inner chamber were removed and the lower surface of the polycarbonate membrane was stained with the solution provided. Following this the membrane was lysed in 10% (v/v) acetic acid and the stain incorporated in the cells was measured by a colorimetric reading at 560 nm.

Antibodies, growth factors, inhibitors

Cell activation of PDGF-Rα and PDGF-Rβ was performed using PDGF-BB growth factor. The antibodies used for western blotting, immunocytochemistry and immunoprecipitation are as follows: anti-pAkt 1/2/3 (sc-16646, Santa Cruz), anti-pErk1/2 (sc-16982, Santa Cruz), anti-PDGF- Rα (sc-431 Santa Cruz), anti-PDGF-Rβ (sc-432, Santa Cruz), anti-phosphotyrosine (PY20, Neomarker), anti-tubulin-α (T9026, Sigma) and peroxidase-conjugated anti-rabbit (AP132P, Chemicon) and anti-mouse (A-4416, Sigma).

Statistical analysis

All values are given as mean ± standard deviation of three separate experiments in triplicate. Differences between cell lines were evaluated using the Student t test (graphpad instat version 3.0 software) and considered statistically significant at a level of  0.01.

Acknowledgements

We thank Alexander S. Onassis Public Benefit Foundation for providing financial support to Ch.M. We thank Novartis Pharma AG (Basel, Switzerland) for generously providing imatinib mesylate.

Ancillary