SEARCH

SEARCH BY CITATION

Keywords:

  • focal adhesion kinase;
  • integrin;
  • interleukin-6 (IL-6);
  • irradiation resistance;
  • syndecan

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Syndecan-1 is a cell surface heparan sulfate proteoglycan with various biological functions relevant to tumor progression and inflammation, including cell–cell adhesion, cell–matrix interaction, and cytokine signaling driving cell proliferation and motility. Syndecan-1 is a prognostic factor in breast cancer, and has a predicitive value for neodadjuvant chemotherapy. It is still poorly understood how syndecan-1 integrates matrix-dependent and cytokine-dependent signaling processes in the tumor microenvironment. Here, we evaluated the potential role of syndecan-1 in modulating matrix-dependent breast cancer cell migration in the presence of interleukin-6, and its potential involvement in resistance to irradiation in vitro. MDA-MB-231 breast cancer cells were transiently transfected with syndecan-1 small interfering RNA or control reagents, and this was followed by stimulation with interleukin-6 or irradiation. Cellular responses were monitored by adhesion, migration and colony formation assays, as well as analysis of cell signaling. Syndecan-1 depletion increased cell adhesion to fibronectin. Increased migration on fibronectin was significantly suppressed by interleukin-6, and GRGDSP peptides inhibited both adhesion and migration. Interleukin-6-induced activation of focal adhesion kinase and reduction of constitutive nuclear factor kappaB signaling were decreased in syndecan-1-deficient cells. Focal adhesion kinase hyperactivation in syndecan-1-depleted cells was associated with dramatically reduced radiation sensitivity. We conclude that loss of syndecan-1 leads to enhanced activation of β1-integrins and focal adhesion kinase, thus increasing breast cancer cell adhesion, migration, and resistance to irradiation. Syndecan-1 deficiency also attenuates the modulatory effect of the inflammatory microenvironment constituent interleukin-6 on cancer cell migration.


Abbreviations
ECM

extracellular matrix

ER

estrogen receptor

FAK

focal adhesion kinase

IL

interleukin

MMP

matrix metalloproteinase

NF-κB

nuclear factor kappaB

Sdc-1

syndecan-1

siRNA

small interfering RNA

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Syndecan-1 (Sdc-1) is a cell surface heparan sulfate proteoglycan that is highly expressed in epithelial cells and plasma cells [1]. The core protein of human Sdc-1 comprises 310 amino acids, and is characterized by the presence of highly conserved transmembrane and cytoplasmic domains, which mediate oligomerization and cytoskeletal interactions [1-3]. The extracellular domain of Sdc-1 contains an integrin-binding site and harbors conserved attachment sites for heparan sulfate and chondroitin sulfate carbohydrate chains [1, 4]. These glycosaminoglycans are linear polymers composed of repetitive disaccharide units of uronic acids and variably sulfated N-acetylglucosamine (heparan sulfate) or N-acetylgalactosamine (chondroitin sulfate), and are capable of forming specific ligand-binding motifs [1, 5-7]. Glycosaminoglycan-mediated interactions with a broad range of ligands determine the majority of the physiological functions of Sdc-1: It acts as a coreceptor for cytokines and chemokines, and as an adhesion receptor via interactions with a wide range of matrix molecules [1-3].

All of these cellular processes are of relevance for tumor progression, as dysregulation of these Sdc-1-dependent functions can result in changes in cell proliferation and survival, thus promoting primary tumor growth, and in an altered proteolytic environment and altered cell motility, chemotaxis, and invasiveness, thus promoting metastatic cell behavior. Furthermore, Sdc-1 participates in the generation of a proinflammatory and proangiogenic microenvironment, supporting tumor growth and metastatic spread [5-8]. Sdc-1 is a substrate for matrix metalloproteinases (MMPs), which are important pharmacological targets that promote the process of metastasis via degradation of the extracellular matrix (ECM) and by promoting tumor angiogenesis and cancer cell motility [8-10]. Clinically, a prognostic value has been assigned to Sdc-1 in a variety of malignant diseases [5-7], including breast cancer [11-13], and it has been proposed as a predictive marker for the outcome of neoadjuvant chemotherapy [14, 15]. The finding of a differential role of MMP-generated soluble Sdc-1 ectodomains and the membrane-bound form of Sdc-1 in breast cancer cell proliferation and invasiveness [8], and the association of the prometastatic microRNA miR-10b with reduced Sdc-1 expression [16, 17], suggest that the function of Sdc-1 in breast cancer progression is highly context-dependent, and that Sdc-1 may play different roles during early and late stages of the pathogenetic process. In each step of cancer progression, the tumor microenvironment has been implicated as a major regulator of carcinogenesis [18-20]. Owing to its pleiotropic molecular functions, Sdc-1 is potentially a pivotal element in augmenting both cytokine-mediated and ECM-dependent signaling processes within the tumor microenvironment. However, it is presently only poorly understood how Sdc-1 integrates these two routes of signaling, and what the consequences for malignant cell behavior are.

Among the cytokines in the tumor microenvironment, interleukin (IL)-6 mediates antiapoptotic, proinvasive and immune-stimulatory effects [20]. Like Sdc-1, IL-6 is a prognostic factor, as its elevation in serum correlates with poor prognosis in breast cancer [21, 22]. Notably, IL-6 is also a cytokine with multiple heparin-binding sites, and its interaction with heparan sulfate is thought to concentrate IL-6 in the vicinity of the producing cell type, thus favoring paracrine rather than endocrine stimulation, and rendering IL-6 resistant to proteolytic degradation [2, 23, 24]. The observations that IL-6 modulates Sdc-1 expression in a variety of malignant and benign cell types [25, 26] and that IL-6 expression is dysregulated in Sdc-1-deficient mice [27, 28] and Sdc-1-depleted endometriotic cells [29] suggest the presence of regulatory feedback loops between Sdc-1 and IL-6, with potential consequences regarding IL-6-regulated cancer cell motility. However, a possible functional interplay of Sdc-1 and IL-6-dependent signaling in breast cancer has not been experimentally addressed to date.

Apart from cytokines, the ECM is capable of conveying signals to tumor cells [5-7, 30, 31]. Sdc-1 apparently regulates downstream signaling pathways that are traditionally associated with the integrin family of matrix receptors [3, 4, 7, 16, 31-33], thus mediating cell migration by forming a dynamic linkage between the ECM and the cytoskeleton, and by modulating Rho family members that control the activation of focal adhesion kinase (FAK). Notably, in MDA-MB-231 breast cancer cells, Sdc-1 physically interacts with FAK [16]. Most recently, integrin-mediated activation of FAK has been linked to the resistance of a variety of malignant cells and tumors to radiotherapy [34-37], underscoring the clinical relevance of this signaling pathway. We have recently demonstrated that small interfering RNA (siRNA)-mediated depletion of Sdc-1 in human MDA-MB-231 breast cancer cells is associated with increased activation of FAK signaling, and complex dysregulation of cytokine expression and function [16], suggesting an impact of Sdc-1 on the breast cancer microenvironment.

In the present study, we used an in vitro siRNA approach in highly aggressive, triple-negative human MDA-MB-231 breast cancer cells to elucidate the role of Sdc-1 in matrix-dependent and cytokine-dependent signaling, as exemplified by IL-6 and fibronectin-triggered integrin activation. Our study revealed that loss of Sdc-1 leads to increased integrin-dependent adhesion and cell motility, which is associated with increased FAK activation. Sdc-1 depletion results in decreased constitutive IL-6 signaling, a process modulating Sdc-1-dependent cell motility, but not adhesion to fibronectin. Finally, we uncover a novel role for Sdc-1 in radiation resistance, suggesting that FAK hyperactivation in Sdc-1-depleted cells may exert a radioprotective effect on MDA-MB-231 cells. Overall, our data identify Sdc-1 as an integrator of cytokine-dependent and matrix-dependent signals from the tumor microenvironment, marking this proteoglycan as a promising therapeutic target in breast cancer.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Modulation of MDA-MB-231 breast cancer cell migration by Sdc-1 is a matrix-dependent process regulated by β-integrins and IL-6

Recent results from our group have demonstrated increased invasiveness of Sdc-1-depleted MDA-MB-231 breast cancer cells on matrigel matrices [16], along with cytoskeletal alterations and activation of signaling pathways traditionally associated with integrins. As Sdc-1 can act as a matrix receptor [1], we hypothesized that it may modulate cell migration in a matrix-dependent manner. To test this hypothesis, we investigated the migration of control and Sdc-1 siRNA-transfected cells through transwell polycarbonate filters that had either been left uncoated or had been coated with fibronectin (Fig. 1A). Whereas Sdc-1 depletion significantly reduced cell migration through uncoated filters, by ~ 27% as compared with controls, Sdc-1 siRNA knockdown substantially and significantly increased cell migration through fibronectin-coated filters, by ~ 77% (Fig. 1A), suggesting matrix-dependent effects of Sdc-1 on cell migration. To elucidate the contributing molecular mechanisms, we next investigated two possible pathways related to Sdc-1 function – its role as a cytokine coreceptor, and its role as a modulator of integrin-related signaling. Regarding cytokines, we focused on an investigation of IL-6 function, as: (a) high serum levels of IL-6 are associated with a poor prognosis in breast cancer [38]; (b) IL-6 promotes breast cancer cell motility [39]; (c) Sdc-1 expression has been shown to be regulated by IL-6 [25, 26]; and (d) IL-6 expression is dysregulated in Sdc-1-deficient mice [27, 28], suggesting possible links between IL-6-mediated cancer cell motility and Sdc-1. Incubation of control cells with IL-6 resulted in a moderate (11.6%), but significant, increase in cell migration through polycarbonate filters (Fig. 1B). Notably, IL-6 treatment was able to restore reduced migration of Sdc-1-depleted MDA-MB-231 cells to control cell levels (Fig. 1B). When cell migration through fibronectin-coated filters was investigated, no migration-promoting effect on control cells was seen; however, IL-6 treatment resulted in significant and substantial inhibition of increased cell migration of Sdc-1-depleted cells through fibronectin-coated filters (Fig. 2), indicating a functional interaction of Sdc-1 and the IL-6 pathway in cell migration. The increased migration of Sdc-1-depleted MDA-MB-231 cells through fibronectin-coated filters suggested potential integrin involvement. To address integrin involvement, we pretreated control and Sdc-1-depleted cells with the GRGDSP peptide – a specific inhibitory peptide for α5β1-integrin [40, 41] – prior to performing the fibronectin migration assay. α5β1-Integrin (VLA5) is a specific receptor for fibronectin that binds through the Arg-Gly-Asp (RGD) binding site [42]. In fact, most of the cell adhesion activity of fibronectin is mediated by the central cell-binding domain recognized by VLA5 [43]. In our experimental setting, GRGDSP peptide treatment was able to reduce the increased migration of Sdc-1-depleted cells to control levels (Fig. 2), indicating that Sdc-1 silencing promotes increased MDA-MB-231 cell migration via a β1-integrin-dependent mechanism. In contrast to IL-6, GRGDSP peptide inhibited control cell migration through fibronectin-coated filters. Combined treatment of cells with IL-6 and GRGDSP peptide did not result in an additive effect in control or Sdc-1-depleted cells (Fig. 2). We next addressed the question of whether Sdc-1-modulated migration on fibronectin was linked to altered cell adhesion. As compared with controls, Sdc-1 siRNA knockdown induced a significant increase (> 62%) in adhesion of MDA-MB-231 cells to fibronectin (Fig. 3). In contrast to the modulatory effect on cell migration, IL-6 treatment did not reduce adhesion of Sdc-1-depleted cells to control levels. However, pretreatment with GRGDSP peptide resulted in a substantial and significant decrease in cell adhesion in both Sdc-1-depleted and control cells (Fig. 3).

image

Figure 1. Sdc-1 siRNA knockdown promotes MDA-MB-231 breast cancer cell migration in a fibronectin-dependent and IL-6-dependent manner. (A) In vitro cell migration of control and Sdc-1 siRNA-depleted MDA-MB-231 cells was assayed on uncoated and fibronectin-coated polycarbonate filters as described in 'Experimental procedures'. Sdc-1 depletion inhibited migration through polycarbonate filter membranes, whereas it promoted migration through fibronectin-coated filters. Left panel: representative picture of stained cells on filter membranes. Right panel: quantitative analysis. n ≥ 3,*P < 0.05, **P < 0.005. (B) Stimulation with 20 ng·mL−1 IL-6 moderately enhanced control cell migration through polycarbonate filters. Sdc-1 siRNA knockdown retarded migration of MDA-MB-231 cells, whereas this was rescued by IL-6 stimulation. Left panel: representative picture. Right panel: quantitative analysis. n ≥ 3, **P < 0.01, ***P < 0.001. FN, fibronectin; Sdc-1si, Sdc-1 siRNA.

Download figure to PowerPoint

image

Figure 2. Increased migration of Sdc-1-depleted breast cancer cells is an integrin-dependent process that can be modulated by IL-6. In vitro cell migration of control and Sdc-1 siRNA-depleted MDA-MB-231 cells was assayed on fibronectin-coated polycarbonate filters as described in 'Experimental procedures'. Cells were left untreated or were subjected to preincubation with or without 100 μg·mL−1 GRGDSP peptide (RGD) as an inhibitor of α5β1-integrin–fibronectin binding for 30 min, and/or subjected to incubation with or without 20 ng·mL−1 IL-6 for 16 h. Both the GRGDSP peptide and IL-6 treatment abolished the migration-enhancing effect of Sdc-1-depletion on fibronectin in a nonadditive manner. Control cells were affected by GRGDSP peptide treatment only. Top panel: representative images of stained cells on filter membranes. Bottom panel: quantitative analysis. n ≥ 3, *P < 0.05, **P < 0.005, ***P < 0.0005. Sdc-1si, Sdc-1 siRNA.

Download figure to PowerPoint

image

Figure 3. Increased adhesion of Sdc-1-depleted MDA-MB-231 cells to fibronectin depends on integrins, but not on IL-6. Adhesion of control and Sdc-1 siRNA-treated MDA-MB-231 cells to fibronectin-coated 96-well plates was determined as described in 'Experimental procedures'. Cells were left untreated or subjected to GRGDSP peptide (RGD) treatment (30 min of pretreatment) with or without 20 ng·mL−1 IL-6-treatment for 1 h. Sdc-1 depletion substantially increased adhesion to fibronectin as compared with controls. IL-6 did not influence cell adhesion, whereas interference with integrin–fibronectin interactions resulted in a strong inhibitory effect on both control and Sdc-1-depleted cells. Top panel: representative images of stained cells. Bottom panel: quantitative analysis. n ≥ 3, *P < 0.05, **P < 0.01. Sdc-1si, Sdc-1 siRNA.

Download figure to PowerPoint

IL-6 differentially affects dysregulated FAK activation and nuclear factor kappaB (NF-κB) signaling in Sdc-1-depleted breast cancer cells

Our data suggested that β1-integrins may already promote increased migration of Sdc-1-depleted MDA-MB-231 cells on fibronectin at the step of cell adhesion, whereas IL-6 may promote this process at a later stage or via indirect mechanisms. To determine further mechanistic details, we performed a signal transduction analysis by western blotting for activated forms of FAK, a downstream effector of integrin signaling [32, 42, 44], and of the transcription factor NF-κB as a downstream readout of IL-6 function [45]. As previously demonstrated [16], Sdc-1 depletion resulted in increased activation of FAK in MDA-MB-231 cells (Fig. 4A). IL-6 stimulation for 10 min resulted in a significant and transient increase of FAK activation in control cells. In contrast, IL-6 treatment lead to transient downregulation of increased FAK activation in Sdc-1-depleted cells, generating a mirror image of the control cell response (Fig. 4A). These data demonstrate a modulatory role of IL-6 in FAK activation, and dysregulation of this process in Sdc-1-depleted cells. To study the role of NF-κB, control and Sdc-1-silenced MDA-MB 231 cells were stimulated with IL-6 for 10 and 30 min. In accordance with previous studies, NF-κB was found to be constitutively activated in estrogen receptor (ER)-negative MDA-MB-231 cells [46] (Fig. 4B). In control cells, stimulation with IL-6 resulted in decreased activation of NF-κB. In contrast, Sdc-1-depleted cells showed a lower degree of basal NF-κB phosphorylation. Similarly to control cells, Sdc-1-depleted cells responded to IL-6 stimulation with a decrease in NF-κB activation; however, overall levels of the phosphorylated transcription factor were lower than in controls (Fig. 4B).

image

Figure 4. IL-6 modulates Sdc-1-dependent changes in FAK and NF-κB signaling. MDA-MB-231 cells were subjected to control or Sdc-1 siRNA treatment, and incubated with 25 ng·mL−1 IL-6 for 0, 10 or 30 min; this was followed by western blot analysis of activated FAK (pFAK, Y925) or activated NF-κB (pNF-κB), respectively. Data were normalized to tubulin expression. (A) In control cells, IL-6 treatment transiently increased constitutive FAK activation after 10 min, and this was followed by a decrease after 30 min of incubation. In contrast, Sdc-1 siRNA depletion led to increased basal FAK activation [10]. IL-6 treatment led to transient normalization of FAK phosphorylation to untreated control levels after 10 min of stimulation, and increased levels were restored after 30 min. Top panel: representative western blot. Bottom panel: quantitative analysis. n ≥ 3, *P < 0.05, **P < 0.005. (B) Constitutive NF-κB activation in MDA-MB-231 cells was inhibited by IL-6 treatment both in control and in Sdc-1-depleted cells. Sdc-1 siRNA knockdown resulted in a significantly lower degree of NF-κB activation than in controls. n ≥ 3, *P < 0.05, **P < 0.01, ***P < 0.001. Sdc-1si, Sdc-1 siRNA.

Download figure to PowerPoint

Sdc-1 siRNA knockdown substantially enhances resistance of MDA-MB-231 cells to irradiation

Our results indicated a role for Sdc-1 in FAK activation [16], which can be modulated by IL-6 (Fig. 4A). Integrin-associated pathways such as FAK activation have most recently been identified as as key promoters of cancer cell resistance to irradiation [47]. Therefore, we studied the potential impact of Sdc-1 depletion on the radioresistance of MDA-MB-231 cells in a colony formation assay. In untreated cells, Sdc-1 depletion did not induce significant changes in colony formation (Fig. 5A). Irradiation of control cells with 5 Gy drastically reduced cell survival, by > 70%. In contrast, Sdc-1 siRNA knockdown exerted a substantial protective effect, resulting in a significant and substantial increase in the survival fraction (Fig. 5A). Western blot analysis revealed that exposure of control cells resulted in activation of FAK signaling, whereas Sdc-1 siRNA-treated cells already showed preactivation of FAK under nonirradiated conditions (Fig. 5B).

image

Figure 5. Sdc-1 depletion increases resistance to irradiation. Control and Sdc-1 siRNA-depleted cells were subjected to siRNA treatment for 24 h, and then to irradiation with 5 Gy. Cells were subjected to a colony formation assay (A) or western blot analysis for FAK activation (B) as described in 'Experimental procedures'. (A) The colony formation assay indicated increased resistance of Sdc-1-depleted cells to irradiation in vitro. n ≥ 3, *P < 0.05, ***P < 0.001. (B) Western blot analysis was performed 24 h after sample irradiation. Irradiation induced activation of FAK. Sdc-1-depleted cells showed increased FAK activation prior to irradiation. Sdc-1si, Sdc-1 siRNA.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

In this study, we used an siRNA approach in the well-established human breast cancer cell line MDA-MB-231 to investigate matrix-dependent functions of Sdc-1, and functions depending on the inflammatory microenvironment constituent IL-6. The decrease in cell migration of Sdc-1-deficient cells through polycarbonate is in accordance with the well-established role of Sdc-1 as an adhesion molecule (reviewed in [1]). Regarding matrix-dependent functions, our data suggest that absence of Sdc-1 leads to increased activation of β1-integrins (Fig. 6), resulting in increased cell adhesion to, and migration through, their substrate fibronectin. Apparently, the downregulation of Sdc-1 enhances the binding capacity or activity of β-integrins, similarly to previous findings in Sdc-1-deficient neutrophils, which showed increased adhesion to ICAM-1 that could be blocked by antibodies directed against the β-integrin CD18 and by heparin [27, 48], suggesting a potential involvement of the glycosaminoglycan chains in this process. Previous studies have shown that Sdc-1 is indeed sufficient for adhesion and is capable of deactivating β1-integrin-mediated adhesion in MDA-MB-231 cells, while, in turn, integrins competitively suppress Sdc-1-mediated adhesion upon activation by fibronectin binding [33]. Although these data would be in accordance with a model whereby Sdc-1 may cause a steric hindrance to integrin–substrate interactions, we also demonstrated hyperactivation of the integrin-related FAK signaling pathway in Sdc-1-depleted MDA-MB-231 cells (Fig. 6). These data conform with our previous demonstration of FAK binding to Sdc-1 in MDA-MB-231 cells [16], which was associated with increased Rho-GTPase-dependent cell motility and enhanced invasiveness through basement membrane-like matrices rich in laminin and collagen IV. Similar findings made in different tumor entities further suggest that Sdc-1 acts in concert with integrins, mediating cell migration by forming a dynamic linkage between the ECM and cytoskeleton and by modulating Rho family members that control activation of FAK at sites of focal adhesion [31, 49]. FAK is a key transducer of integrin-mediated signals, and can influence the cytoskeleton, cell adhesion sites and membrane protrusions to regulate cell movement [50]. Recent findings also point at a role of the transmembrane domain of Sdc-1 in regulating focal adhesion disassembly in lung epithelial cells via a process involving the small GTPase Rap1 [51, 52]. This novel role of Sdc-1 in FAK activation provides a mechanistic link to the striking finding that Sdc-1-depleted MDA-MB-231 cells were substantially more resistant to radiotherapy. Our data suggest that Sdc-1 siRNA knockdown exerts a radioprotective effect on MDA-MB-231 cells via preactivation of the FAK pathway (Fig. 6). Overexpression of FAK protects head and neck squamous cell carcinoma cells from radiation-induced cell death [34], whereas siRNA-mediated silencing or pharmacological inhibition of FAK increases the radiosensitivity of different tumor cell lines, including breast cancer and colorectal cancer cells [35-37]. Moreover, β1-integrin inhibition enhances the sensitivity to ionizing radiation and delays the growth of human head and neck squamous cell carcinoma cell lines in 3D cell culture and in xenografted mice. Mechanistically, FAK dephosphorylation in β1-integrin-inhibited cells resulted in dissociation of a FAK–cortactin protein complex, inducing radiosensitization [53]. Our data reveal a previously unknown role of Sdc-1 in resistance to irradiation, which may be linked to the prognostic value of Sdc-1 expression in breast cancer [11-14]. Future therapeutic radiosensitizing approaches should thus not only consider pharmacological targeting of integrins, but also take into account the contribution of Sdc-1 and its modulation of integrin-dependent processes.

image

Figure 6. Sdc-1-dependent modulation of breast cancer cell adhesion, migration and resistance to irradiation as an integrin-dependent and IL-6-dependent process. Sdc-1-dependent modulation of adhesion to fibronectin depends on integrin function, whereas Sdc-1-dependent modulation of cell migration is differentially modulated by integrins and IL-6. Increased activation of FAK in the absence of Sdc-1 may be an underlying cause of increased resistance to irradiation, and is differentially affected by IL-6. Absence of Sdc-1 leads to a decrease in constitutive NF-κB activation in MDA-MB-231 cells, and this is further decreased by IL-6. See text for further details.

Download figure to PowerPoint

Our study also demonstrates an influence of Sdc-1 on the inflammatory microenvironment, which in turn modulates signaling responses, resulting in altered cell motility and invasiveness (Fig. 6). Increased activation of β-integrins and FAK in the absence of Sdc-1 may be linked to the restoration of decreased migration of Sdc-1-depleted cells by IL-6, as Shain et al. [54] reported that the activity of the IL-6 coreceptor gp130 is enhanced by β1-integrin ligation and FAK. As the basal activity of gp130 is frequently low or absent in MDA-MB-231 cells [55, 56], such a potential mechanism could have compensated for the loss of Sdc-1 in the presence of IL-6. Furthermore, the downstream effector of IL-6, NF-κB [45], is activated by the engagement of α5β1-integrin [57]. IL-6 is considered to be an inducer of this critical regulator of cancer cell invasiveness [58-60], and activated NF-κB synergizes with IL-6 gene transcription in a positive feedback loop [45]. Sdc-1 depletion apparently leads to deregulation of this process in MDA-MB-231 cells, consistent with a role for syndecan glycosaminoglycan chain-mediated binding of cytokines and their receptors in a coreceptor function for proinflammatory signaling [2, 7]. Similarly to our findings, IL-6 has a negative impact on the migration of human malignant plasma cell lines [61], and stimulation of HeLa cells and lymphoma cells with IL-6 was shown to suppress their motility on fibronectin [62]. Sdc-1-depleted cells show reduced IL-6-mediated inhibition of constitutive NF-κB activation (Fig. 6). This function may be relevant for the NF-κB-dependent cell proliferation, formation of bone metastases and metastatic osteolysis previously demonstrated for MDA-MB-231 cells in vivo [58].

In this study, we employed an ER-negative breast cancer cell line as an in vitro model system. Although there is evidence for regulation of syndecan family members by estrogen in human breast cancer cells, this steroidal mode of regulation appears to be syndecan-specific: for example, expression of syndecan-2 and of the syndecan-shedding MMP9 [9, 10] in ER-positive MCF-7 cells is regulated in an estrogen-dependent manner, whereas expression of syndecan-4 is not [63]. Whereas epithelial Sdc-1 expression in breast cancer is associated with an ER-negative subtype, stromal expression is associated with ER-positive status, indicating possible divergent regulatory modes and roles of epithelial and stromal Sdc-1 [12, 64, 65]. Regarding adhesion to fibronectin, both ER-positive and ER-negative breast cancer cell adhesion to this matrix substrate was reported to be blocked by antibodies against β1-integrin [66]. However, as ER-negative MDA-MB-231 breast cancer cells overexpress the fibronectin receptor α5β1-integrin [67], whereas ER-positive breast cancer cell lines such as MCF-7 and T47D express considerably lower levels of α5β1-integrin [68], ER-negative cell lines may be less prone to the integrin-modulating function of Sdc-1 depletion.

Although this study has revealed novel functions of Sdc-1 in breast cancer cell adhesion, motility, and irradiation resistance, our in vitro siRNA approach has some limitations, as it did not distinguish between membrane-bound and shed Sdc-1. MMP-mediated shedding of syndecans is particularly active in invasive breast cancer cells [10], as functionally demonstrated by an invasion-promoting effect of soluble Sdc-1 overexpression in MCF-7 cells [8]. As the soluble Sdc-1 ectodomain can act antagonistically to the membrane-bound form, owing to its ability to bind the same ligands [1], it is conceivable that shed Sdc-1 may act as a paracrine factor interfering with integrin–matrix interactions, or with cytokines from the inflammatory microenvironment via its glycosaminoglycan chains. The complexity of Sdc-1 function in breast cancer is further increased by the presence of Sdc-1 not only in tumor cells, but also in the tumor stroma [64, 65], and on endothelial cells and leukocytes, meaning that it can influence both the inflammatory microenvironment and tumor angiogenesis [3, 4, 18, 69]. Although future therapeutic approaches exploiting Sdc-1-dependent functions may have the advantage of simultaneously targeting multiple steps of tumor progression [5, 7, 70], they will also have to take this complexity into account when assessing possible side effects.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

Antibodies and reagents

The antibodies pNFκB p65 (ser596, 39H1), pFAK (Tyr925) and FAK were from Cell Signaling (Beverly, MA, USA). IL-6 was from R&D Systems (Wiesbaden, Germany), and tubulin antibodies were from Sigma (Deisenhofen, Germany). The GRGDSP peptide was from Calbiochem (EMB Biosciences, La Jolla, CA, USA). Media, fetal bovine serum and tissue culture supplies were from Gibco BRL (Karlsruhe, Germany). Unless stated otherwise, all chemicals were from Sigma.

Cell culture

The human breast cancer cell line MDA-MB-231 was from ATCC/LGC Promochem (Wesel, Germany). Cells were maintained in DMEM containing 10% fetal bovine serum, 1% glutamine and 1% penicillin/streptomycin in a humidified atmosphere of 7% CO2 at 37 °C.

siRNA-mediated knockdown of Sdc-1 expression

siRNA knockdown was performed with siRNA #12634 (Ambion, Cambridge, UK), targeting the coding region of Sdc-1, and a negative control siRNA (negative control #1; Ambion). MDA-MB-231 cells were transfected with 40 nm siRNA, by the use of Dharmafect reagent (Dharmacon, Lafayette, CO, USA), according to the manufacturer's instructions. siRNA knockdown was confirmed by quantitative real-time PCR and flow cytometry, as previously described [16].

Western blot

To detect activation of FAK and NF-κB, cells were incubated with 25 ng·mL−1 IL-6 for 0 min, 10 min, and 30 min, respectively, 72 h after siRNA transfection. Total cell lysates were prepared with modified RIPA buffer [8] with proteinase inhibitors. Thirty to 50 μg of protein per lane was separated on 10% gels, and electrotransferred to Hybond nitrocellulose membranes (Amersham, Pharmacia Biotech, Piscataway, NJ, USA). Antibody incubations were performed essentially as previously described [8]. Following 1 h of blocking, detection of phosphorylated forms was performed with rabbit polyclonal primary antibodies against phosphorylated or total forms of NF-κB or FAK, respectively (1 : 1000), and horseradish peroxidase-conjugated anti-(rabbit IgG) (Calbiochem; 1 : 10000). Subsequently, membranes were subjected to an enhanced chemiluminescence reaction and signal quantification with NIH imagej software, normalizing the densitometric values of pNFκB/pFAK to tubulin (as loading control). For tubulin detection, membranes were stripped with 0.2 m glycine buffer (pH 2.5), washed, reincubated with primary antibody, and subjected to the procedure described above.

Cell migration assay

Transwell inserts (Corning Costar, Cambridge, MA, USA) with a 8-μm-pore polycarbonate membrane were coated with 10 μg·mL−1 fibronectin. Forty-eight hours after transfection, 2.5 × 104 cells in 100 μL of serum-free DMEM were placed in the upper compartment, and allowed to migrate towards the lower compartment containing or not containing 20 ng·mL−1 IL-6. After 16 h, nonmigrating cells at the upper filter surface were removed with a cotton swab, and migrated cells on the bottom of the filter were fixed and stained with Diff-Quik dye (Medion, Duedingen, Switzerland). Excised and mounted filter membranes were photographed with a Zeiss Axiovert microscope equipped with axiovision software (Zeiss, Jena, Germany) at × 100 magnification and quantified. To study integrin-specific effects, cells were preincubated for 30 min with or without 100 μg·mL−1 GRGDSP peptide, a specific inhibitor of the α5β1-integrin–fibronectin binding site [40, 41]. Experiments with uncoated filters were performed analogously with a migration time of 48 h.

Adhesion assay

Ninety-six-well plates were coated with 50 μg·mL−1 fibronectin or 10 μg·mL−1 BSA as a negative control overnight at 4 °C, and washed twice with washing solution (0.1% BSA in DMEM). The plate was then incubated for 45 min with 0.5% BSA in DMEM (blocking solution) to block nonspecific binding at room temperature, and washed once. Seventy-two hours after transfection, control and Sdc-1 siRNA-transfected cells were released from the plates with 2 mm EDTA in NaCl/Pi, washed twice, and resuspended in blocking solution. Cells (2.5 × 104 per well) were added to the coated wells, and allowed to attach for 1 h at 37 °C in a cell culture incubator. Nonadherent cells were removed by three gentle washes with NaCl/Pi buffer, and subsequently fixed with 3.7% NaCl/Pi-buffered formaldehyde for 30 min. Attached cells were stained with 1% methylene blue in 0.01% borate buffer (pH 8.5) for 30 min following four washes with borate buffer, the cells were lysed in ethanol/0.1 m HCl (1 : 1), and the released stain was quantified in a Softmax Microplate reader (Molecular Devices, Sunnyvale, CA, USA) at 620 nm.

Colony formation assay

To measure colony-forming ability, 2 × 105 knockdown cells were irradiated, and subsequently trypsinized and counted. Cells (1 × 103) were resuspended in 1 mL of culture medium, plated into 3.5-cm Petri dishes with a 2.5-mm grid (Nunc, Langenselbold, Germany), and incubated for ~ 6 days in a CO2 incubator at 37 °C. Cell colonies with > 50 cells were counted with a microscope (Olympus, Hamburg, Germany). The survival fraction was calculated as follows: plating efficiency treated/plating efficiency control.

Radiation exposure

Irradiation was performed at room temperature with 6-MV photons of a linear accelerator (Varian Medical Systems, Palo Alto, CA, USA). The dose rate was 4.8 Gy·min−1, and a dose of 5 Gy was applied.

Statistical analysis

Data were expressed as mean ± standard error of the mean. Statistical analysis was performed with sigma stat 3.1 software (Systat Software, Point Richmond, CA, USA). A two-tailed unpaired t-test was used when groups passed the normality test, and a Mann–Whitney test was used when the standard deviations of two groups differed significantly. A P-value of < 0.05 was considered to be statistically significant. All calculations were performed on the means of triplicate/sextuplet measurements of at least three independent experiments.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References

The authors would like to thank B. Pers, A. van Dülmen and K. Brüggemann for expert technical assistance. This study was financially supported by the German Academic Exchange Service (DAAD) A/06/90277 (to S. A. Ibrahim), DAAD Al Tawasul Project ID 56808461 (to S. A. Ibrahim and M. Götte), DAAD PROBRAL Project ID 54387857 (to M. Götte and M. S. G. Pavao), DFG-CNPq 444 BRA-113/63/0-1 (to M. Götte and M. S. G. Pavao), EU FP7 INFLAMA (to M. Götte and M. S. G. Pavao), and Maria Möller Stiftung (to M. Götte).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References