Involvement of the Chemokine Receptor CXCR4 and Its Ligand Stromal Cell-Derived Factor 1A in Breast Cancer Cell Migration Through Human Brain Microvascular Endothelial Cells

In this study, we have characterized the signaling pathways mediated by CXCR4 in breast cancer cells and its role in breast cancer cell invasion and migration. Stromal cell-derived factor 1A (SDF-1A; CXCL12) stimulation of breast cancer cells resulted in phosphoinositide 3-kinase (PI-3K) activation, AKT phosphorylation, and activation of the FKHRL1 transcription factor. In addition, SDF-1A induced activation of the focal adhesion kinase (FAK) as well as the migration of breast cancer cells. Expression of SDF-1A, the ligand of CXCR4, was about 2-fold higher in microdissected human breast epithelial cancer cells as compared with normal epithelial cells. Immunohistochemical analysis indicated that SDF-1A expression is consistently higher in primary breast tumor cells than in normal breast epithelial cells. Furthermore, SDF-1A induced blood vessel instability, through increased vascular permeability, resulting in the penetration of breast tumor cells through the human brain microvascular endothelial cells (HBMEC). Notably, the migration of breast cancer cells was inhibited by the PI-3K inhibitor, Wortmannin, and the Ca inhibitor BAPTA/AM, indicating that transendothelial breast cancer cell migration induced by SDF-1A is mediated by activation of the PI-3K/AKT pathway and Ca-mediated signaling. Blockade of the CXCR4/SDF1 signaling pathway with anti-CXCR4 antibody also decreased transendothelial breast cancer cell migration as well as vascular permeability. This study focuses on novel interactions between highly relevant signaling pathways in breast cancer cells and brain microvascular endothelial cells and may provide insights into the molecular mechanisms of CXCR4/SDF-1A-mediated breast cancer metastasis to the brain. (Mol Cancer Res 2004;2(6):327–38) Introduction Chemokines, which are a group of low molecular weight proteins, mediate several cellular functions. They play an important role in the regulation of hematopoiesis, leukocyte maturation, angiogenesis, trafficking, and in the homing of T and B lymphocytes as well as the development of lymphoid tissue (1-6). On the basis of the arrangement of their first two of four conserved cysteine residues, the chemokine superfamily has been divided into four subfamilies: a (C-X-C), h (C-C), g (C), and y (C-X-X-X-C; refs. 1, 2, 6). Stromal cell-derived factor 1a (SDF-1a) is a member of the CXC or a-chemokine subfamily and is the only known ligand for the chemokine receptor CXCR4 (7-9). SDF1 is a highly conserved gene localized to chromosome 10q11.1 (10), and its mRNA is predominant in all tissues (10, 11). SDF1 was originally described as pre–B-cell growth-stimulation factor and is implicated in lymphocyte maturation (12). Human SDF1 is chemotactic for T lymphocytes, monocytes, and neutrophils (5); however, its expression is not restricted to the leukocyte lineage. High levels of SDF-1a expression were also found in the human pancreas, spleen, ovary and small intestine, and lower expression levels were associated with the brain, colon, and placenta (10-13). Mice lacking the SDF1 gene die perinatally. These mice have severely impaired lymphopoiesis and abnormally low numbers of B-cell and myeloid bone-marrow precursors as well as a defective ventricular septum of the heart and defects in the central nervous system (14). These abnormalities suggest that SDF1 may play a role in diverse cellular aspects during morphogenesis and development. SDF-1a interacts specifically with the seven-transmembrane G protein-coupled receptor, CXCR4 (15, 16). Expression of CXCR4 was observed in T lymphocytes, monocytes, and neutrophils (11), which mediates the chemotactic response to SDF1 by these cells. Expression of CXCR4 was also found in human neurons, cultured rodent neurons, glial cells (17, 18), microglial cells (19), and endothelial cells (20, 21). CXCR4 expression was observed to be up-regulated in glioblastomas (22, 23). In addition, CXCR4 plays a role in AIDS and is the cofactor necessary for CD4-mediated infection of T cells by HIV (8, 24-27). CXCR4/SDF1 interactions have been implicated in modulation of the immune response by inducing the Received 10/28/03; revised 5/17/04; accepted 5/24/04. Grant support: NIH grant NS39558 (S. Avraham), the Susan G. Komen Fellowship (S. Avraham), the Milheim Foundation (S. Avraham), CA97153 (H. Avraham), and K18 PAR-02-069 (H. Avraham). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Note: This work was done during the term of an established investigatorship from the American Heart Association (H. Avraham). This article is dedicated to Charlene Engelhard for her continuing friendship and support for our research program. Requests for reprints: Hava Karsenty Avraham, Division of Experimental Medicine, Beth Israel Deaconess Medical Center, Harvard Institutes of Medicine, 4 Blackfan Circle, Boston, MA 02115. Phone: (617) 667-0073; Fax: (617) 9756373. E-mail: [email protected] Copyright D 2004 American Association for Cancer Research. Mol Cancer Res 2004;2(6). June 2004 327 Research. on October 12, 2017. © 2004 American Association for Cancer mcr.aacrjournals.org Downloaded from macrophage-mediated apoptosis of CD8 T cells (26). Interestingly, as observed in SDF1-deficient mice, mice lacking CXCR4 also exhibited hematopoietic, cardiac, and cerebellar defects (14, 28). The expression of chemokine receptors, such as CXCR4 and CCR7, is tightly correlated with the metastatic properties of breast cancer cells (29). This study by Muller et al. showed that the level of CXCR4 is higher in malignant breast tumors than in their normal healthy counterparts, suggesting that its expression level correlates with increased metastasis-associated mortality. In vivo, neutralizing the interaction of CXCR4/ SDF1 significantly impaired the metastasis of breast cancer cells and cell migration (29). Furthermore, Kato et al. (30) have shown that the expression of CXCR4 in surgically resected invasive ductal carcinomas (n = 79) is significantly correlated with the degree of lymph node metastasis. Another study has also described that breast cancer cells metastasized to the lungs express very high levels of CXCR4 as compared with the parental cells, suggesting the pivotal role of CXCR4 in breast cancer cell metastasis (31). These results are further substantiated by the fact that CXCR4 is one of the few genes that is up-regulated in bone-metastasized breast cancer cells (32). Consistent with these studies, knockdown of endogenous CXCR4 gene expression in breast cancer cells resulted in significant inhibition of breast cancer cell migration in vitro (33). In spite of its implication in cancer cell invasion and metastasis, little is known about CXCR4/SDF1 signaling events in breast cancer cells. Here, we demonstrate that the interaction of CXCR4 with its ligand, SDF-1, induced activation of the phosphoinositide 3-kinase (PI-3K)/AKT signaling pathway and the phosphorylations of focal adhesion kinase (FAK) and FKHRL1 in breast cancer cells. Our results indicate that activation of the PI-3K/AKT signaling pathway by CXCR4/SDF1 is implicated in breast cancer cell migration through the human brain microvascular endothelial cells (HBMEC). In addition, we provide evidence for the differential expression of SDF-1 in normal and malignant breast tumor tissues. We propose here a molecular link between the CXCR4-mediated signaling pathway and breast cancer cell metastasis. Results Expression of CXCR4 in Breast Cancer Cells To investigate the signaling pathways mediated by the CXCR4/SDF1 receptor ligand in breast cancer, we examined two breast cancer cell lines, MDA-MB-231 and DU4475, which are well characterized in terms of their metastatic potential and properties. MDA-MB-231 cells are invasive, and metastasize to lung from the primary mammary fat pad tumors of nude mice (34). The DU4475 cells were derived from a 70-year-old female patient with advanced breast cancer and are tumorigenic (35). Because the ability of these breast cancer cells to migrate in response to SDF1 suggested the functionality of the CXCR4 cell surface receptors, we, therefore, measured the cell-surface expression of CXCR4 on the MDA-MB-231 and DU4475 cells. DU4475 cells expressed significantly higher levels of CXCR4 (65%), as compared with the MDA-MB-231 cells (15%) (Fig. 1A). To determine whether CXCR4 was functional, we did a chemotactic assay. As shown in Fig. 1B, both the MDA-MB-231 and DU4475 cells migrated across the transwell membrane coated with fibronectin. Activation of the PI-3K/AKT Pathway by SDF-1a Previous studies showed that SDF-1a induced the activation of PI-3K in lymphocytes (36). Thus, we chose to analyze whether SDF-1a stimulates PI-3K in breast cancer cells. PI-3K activity was increased by SDF-1a treatment both in DU4475 and MDA-MB-231 cells (Fig. 2A). In the DU4475 cells, maximum activity was reached around 10 minutes of stimulation, and decreased by 30 minutes. Activation of PI-3K in the MDA-MB-231 cells was delayed as compared with the DU4475 cells. The maximum PI-3K activity in the MDA-MB-231 cells was observed 15 minutes after SDF-1a treatment, and declined by 30 minutes. AKT is a downstream target of PI-3K and is implicated in a variety of cellular functions, such as survival, transcription, and translation. Recent studies show evidence that AKT is also associated with extracellular matrix invasion by regulating matrix metalloproteinase 9 (MMP9; refs. 37, 38). Therefore, we examined whether SDF-1a activates AKT in breast cancer cells. The phosphorylation of AKTwas monitored by immunoblotting with anti-phospho-AKT antibody. In MDAMB-231 cells, the phosphorylation of AKTwas detected within 5 minutes of stimulation with SDF-1a, and decreased progressively by 30 minutes (Fig. 2B, left panel). In contrast, AKT was constitutively phosphorylated in DU4475 cells, although 5 minutes incubation with SDF-1a showed a modest increase in phospho-AKT activity (Fig. 2B, right panel). The level of PTEN is correlated with the constitutive activation of AKT (39). Therefore, we measured the level of PTEN by Western blot analysis in MDA-MB-231 and DU4475 cells, before and after SDF-1a stimulation. The level of PTEN was not significantly affected by the SDF-1a treatment (data not shown), suggesting that the activation of AKT in the MDA-MB-231 and DU4475 cells does not correlate with the level of PTEN. Activation of FKHRL1 by SDF-1a FKHRL1, which is regulated by the PI-3K pathway, is a direct target of AKT phosphorylation and is involved in the regulation of cell cycle progression and cell death (40). In MDA-MB-231 cells, the phosphorylation of FKHRL1 was increased after treatment with SDF-1a. Its phosphorylation reached a maximum level by 10 minutes (Fig. 3A), and then gradually declined to the basal level by 30 minutes. In contrast, DU4475 cells showed a constitutive phosphorylation of FKHRL1, which was unchanged in the presence of SDF-1a. One of the major target genes of FKHRL1 is the Fas ligand, which is a mediator of cell apoptosis. On growth factor withdrawal, FKHRL1 is unphosphorylated and translocates to the nucleus. Within the nucleus, FKHRL1 activates its target genes, such as FasL . To examine whether SDF-1a stimulation regulates FasL expression in breast cancer cells, we transfected into MDA-MB-231 cells the FHRE-LUC reporter gene in which three canonical FHREs were inserted into the 5Vof a base promoter controlling the expression of the luciferase gene (40). Although we found that SDF-1a suppressed FasL promoter Lee et al. Mol Cancer Res 2004;2(6). June 2004 328 Research. on October 12, 2017. © 2004 American Association for Cancer mcr.aacrjournals.org Downloaded from activity at the 4 and 48 hour time points (black bars) as compared with the untreated cells (white bars), its effect was minimal (Fig. 3B, top panel). We next compared apoptotic cell death between untreated and SDF-1a-treated MDA-MB-231 cells. Although the level of apoptosis was reduced at higher concentrations of SDF-1a (200 ng/mL), which is far above physiologic concentrations, its effect was still minimal (Fig. 3B, middle panel). An MTT [3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide] assay was also used to measure changes in cell proliferation on SDF-1a stimulation. SDF-1a did not show any significant effects on cell proliferation (Fig. 3B, bottom panel). To detect apoptosis at the single cell level, we stained MDA-MB-231 cells using Annexin V in the presence or absence of SDF-1a. Consistent with the results in Fig. 3B, we could not detect any significant differences in the percentages of necrotic and apoptotic cells between the SDF-1a treated (10 and 100 ng/mL) and non-treated cells (data not shown). Although CXCR4/SDF-1a signaling has been described to have effects on cell growth and survival in other cell types, we were unable to detect effects on the in vitro growth and survival of the breast cancer cells used in our studies. SDF-1a Increases the Chemotactic Migration of MDAMB-231 Cells through FAK Activation Cell adhesion is a critical step in the migration of tumor cells, and is regulated by a cytoskeletal complex of focal adhesion components. Thus, we examined whether focal adhesion proteins, such as FAK and related focal adhesion kinase (RAFTK), are stimulated by SDF-1a. FAK phosphorylation was measured by immunoprecipitation with anti-FAK antibody and by Western blot analysis with anti-phosphotyrosine (4G10) antibody. We observed a rapid phosphorylation of endogenous FAK in the MDA-MB-231 cells on SDF-1a stimulation (Fig. 4A, left panel). In contrast, DU4475 cells showed a constitutive phosphorylation of FAK, even without SDF-1a stimulation (Fig. 4A, right panel). Next, to examine the role of FAK phosphorylation in breast cancer cell migration, we introduced FAK COOH-terminal domain (FRNK) into MDA-MB-231 cells. FRNK is a dominant-negative inhibitor, which promotes FAK dephosphorylation and inhibits cell motility (41). FRNK is expressed endogenously under the control of an alternative intronic promoter (41). The expression of FRNK was confirmed by GFP expression (Fig. 4B) as well FIGURE 1. Expression of CXCR4 in MDA-MB-231 and DU4475 cell lines. A. CXCR4 expression at the surface of breast cancer cells. Cells were incubated with anti-CXCR4 antibody (12G5; gray line ) or isotype control (black line ) and stained with FITC-conjugated anti-mouse IgG. Similar results were obtained in two additional experiments. B. Transmigration of MDA-MB-231 and DU4475 cells on SDF-1a treatment. The transmigration of cells on SDF-1a treatment was measured in a transwell assay as described in Materials and Methods. For MDA-MB-231 cells, the filters were coated with fibronectin (10 Ag/ mL). To enumerate the migrated cells, cells were stained using a Hema3 kit. Columns , mean of four separate experiments; bars , SD. Role of CXCR4/SDF-1a in Breast Cancer Metastasis Mol Cancer Res 2004;2(6). June 2004 329 Research. on October 12, 2017. © 2004 American Association for Cancer mcr.aacrjournals.org Downloaded from as by Western blot analysis (Fig. 4C). The anti-FAK antibody cross-reacted with endogenous FAK and virus-expressing FRNK (Fig. 4C). As seen in Fig. 4D, overexpression of FRNK reduced the migration of MDA-MB-231 cells on SDF-1a stimulation, demonstrating that FAK phosphorylation is implicated in SDF-1a/CXCR4-mediated breast cancer cell migration. Interestingly, we did not notice any significant phosphorylation of another focal adhesion component, RAFTK, in either the MDA-MB-231 or DU4475 cells (data