Breast tumor-associated osteoblast-derived CXCL5 increases cancer progression by ERK/MSK1/Elk-1/Snail signaling pathway

Abstract

The skeleton is the most common metastatic site for breast cancer, with bone metastasis causing pain as well as risk of pathological fractures. Interaction between tumors and the bone microenvironment creates a vicious cycle that accelerates both bone destruction and cancer progression. This study is the first to analyze the soluble factors secreted by breast tumor-associated osteoblasts (TAOBs), which are responsible for promoting cancer progression. The addition of CXCL5 (chemokine (C-X-C motif) ligand 5), present in large amounts in TAOB-condition medium (TAOB-CM), mimicked the inductive effect of TAOB-CM on breast cancer epithelial–mesenchymal transition, migration and invasion. In contrast, inhibition of CXCL5 in OBs decreased TAOB-mediated cancer progression. Inducement of MCF-7 and MDA-MB-231 cancer progression by TAOB-derived CXCL5 is associated with increased Raf/MEK/ERK activation, and mitogen- and stress-activated protein kinase 1 (MSK1) and Elk-1 phosphorylation, as well as Snail upregulation. Activation of Elk-1 facilitates recruitment of phosphorylated MSK1, which in turn enhances histone H3 acetylation and phosphorylation (serine 10) of Snail promoter, resulting in Snail enhancement and E-cadherin downregulation. Moreover, mice treated with anti-CXCL5 antibodies showed decreased metastasis of 4T1 breast cancer cells. Our study suggests that inhibition of CXCL5-mediated ERK/Snail signaling is an attractive therapeutic target for treating metastases in breast cancer patients.

Introduction

The skeleton is the most common metastatic site for breast cancer, with bone metastasis developing in 65–75% of patients with metastatic breast cancer. Bone metastasis causes pain as well as increases the risk of pathological fractures.1, 2 The interaction between tumors and the bone microenvironment is a vicious cycle that increases both bone destruction and tumor progression: the tumor secretes soluble factors to stimulate osteoclastogenesis, and the bone stroma produces growth factors to enhance tumor progression.1, 2 Metastasis, the major cause of death for cancer patients, is a complex and multistep process in which secondary cancers are formed in other nonadjacent organs.3, 4 The development of metastasis includes several steps that consists of cellular transformation and abnormal growth, new blood or lymphatic vessels formation, dissemination of tumor cells into the circulation, attachment to the target organs and growth in target sites.3, 5 The migratory and invasive ability of cancer cells is required in many of these steps, and is therefore associated with metastasis.

Epithelial–mesenchymal transition (EMT) is a common phenotypic transformation in cancer cells that causes loss of cell–cell adhesion and increases cell motility,6, 7, 8 thereby increasing their metastatic potential. Snail, one of the important transcription factors involved in the regulation of EMT, acts by repressing the expression of the E-cadherin gene.9 A recent study has demonstrated that the Ras/Raf/ERK cascade modulates Snail expression, leading to EMT in various cell types.10 Phosphorylation of transcription factor Elk-1 by extracellular signal-regulated kinases 1 and 2 (ERK1/2) translocates it to the nucleus and then regulates serum response element-dependent transcription by chromatin remodeling.11 Mitogen- and stress-activated protein kinases (MSKs), another downstream target of ERK signaling pathway, play regulatory roles in chromatin remodeling.12 Therefore, we hypothesized that Raf/MEK/ERK may play a pivotal role in mediating the progression of breast cancer by Snail upregulation via MSK1/Elk-1-mediated epigenetic regulation.

Chemokines, a multifunctional family of small cytokine-like proteins, selectively control the recruitment and migration of lymphocytes to infection or injury sites by chemoattraction.13, 14 Recent evidence increasingly shows that chemokines also play a critical role in neoplastic transformation and cancer progression, in addition to their role in development and inflammatory responses.15, 16 Although regulation of bone metastasis has been the focus of intensive investigation, relatively little is known about how osteoblasts influence the tissue specificity of metastasis and progression of breast cancer, or precisely which molecules are involved in the cross-talk of osteoblasts and cancer cells in the tumor microenvironment. Understanding the mechanisms implicated in the interaction between cancer and osteoblasts is essential for the development of biomarkers of metastasis, as well as for the development of more effective therapies. The aim of this study was to determine whether inflammatory CXCL5 (chemokine (C-X-C motif) ligand 5) secreted by tumor-associated osteoblasts induces cancer migration and invasion through the interaction between tumors and the bone microenvironment, and whether ERK/MSK/Elk-1 is involved in the regulation of an oncogenic transcription factor Snail by epigenetic regulation. Our investigations have enhanced our understanding of the cross-talk between osteoblasts and breast cancer via CXCL5 in the metastasis of breast cancer cells, providing a molecular basis explaining the vicious cycle between osteoblasts and breast cancer cells.

Results

TAOBs increased breast cancer migration and invasion

Soluble factors released from the bone matrix may be involved in the progression of cancer.17, 18 To investigate this hypothesis, we collected breast tumor-associated osteoblast culture medium (TAOB-CM), and then assessed the effects of TAOB-CM on breast cancer cell proliferation, migration and invasion (Supplementary Figure S1A). As shown in Figures 1a and b, both MCF-7-OB-CM (20%) and MDA-MB-231-OB-CM (20%) increased migration and invasion of MCF-7 and MDA-MB-231 cells but did not affect cell proliferation in either cell line (Supplementary Figure S1B). In addition, MCF-7-OB-CM and MDA-MB-231-OB-CM also caused MCF-7 and MDA-MB-231 to undergo EMT, including the downregulation of epithelial marker (E-cadherin), and upregulation of fibroblast markers (N-cadherin and vimentin) (Figure 1c).

Figure 1
figure1

TAOB increases breast cancer migration, invasion and EMT. TAOB-CM enhanced migration (a), invasion (b) and EMT (c) in MCF-7 and MDA-MB-231 cells. (d) Breast cancer increased the expression of CXCL5 in OBs. CXCL5 increased migration (e), invasion (f) and EMT (g) in MCF-7 and MDA-MB-231 cells. The invasiveness and migration ability of MCF-7 and MDA-MB-231 cells were quantified by QCM 24-well Cell Migration and Invasion assay. A total of 20% of MCF-7-OB-CM (for MCF-7), MDA-MB-231-OB-CM (for MDA-MB-231), OB-CM (for control) or CXCL5 (0.01–10 nM) acted as chemoattractants for cancer migration and invasion. MCF-7 and MDA-MB-231 cells were treated with MCF-7-OB-CM, MDA-MB-231-OB-CM or CXCL5 (10 nM) for 24 h and the expression of various proteins assessed by immunoblot. For (d), primary osteoblasts were treated with MEM (20%), MCF-7-CM (20%) and MDA-MB-231-CM (20%) for 24 h. CXCL5 levels were assessed by CXCL5 ELISA kit. Each value is the mean±s.d. of three independent experiments. The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). All experiments were performed independently at least 3 times. A full colour version of this figure is available at the Oncogene journal online.

CXCL5 plays an important role in TAOB-mediated cancer progression

We found that TAOBs express large amounts of CXCL5 protein (Figure 1d). Next, we determined the effect of CXCL5 on the proliferation, migration, invasion and EMT of MCF-7 and MDA-MB-231 cells. As shown in Supplementary Figure S1C, after 48 h of treatment of the two cancer cell lines, CXCL5 did not increase cell proliferation in either MCF-7 or MDA-MB-231 cells, but did enhance the migration and invasiveness of MCF-7 and MDA-MB-231 cells (Figures 1e and f). In addition, CXCL5 also caused MCF-7 and MDA-MB-231 cells to undergo EMT, including the downregulation of epithelial marker (E-cadherin), and upregulation of fibroblast markers (vimentin and N-cadherin) (Figure 1g).

To further elucidate the nature of the soluble factors present in the sera of breast cancer patients, we examined the levels of CXCL5 in healthy donors, stage I and stage III patients. The average CXCL5 serum level in the sera of 15 stage III patients was 124.3 pg/ml, whereas the levels of CXCL5 in the sera of healthy donors and stage I patients were lower than those found in stage III patients (Figure 2a). MCF-7 cells were used to assess cell migration and invasion in the presence of 15% of the serum from breast cancer patients. As shown in Figures 2b and c, migration and invasion of MCF-7 cells was dramatically enhanced in the presence of patients’ sera. Sera presenting high levels of CXCL5 exhibited greater enhancement of cell migration (Figure 2d). Our results convincingly prove that patients with invasive breast cancer have soluble factors released from the bone into the sera, which effectively increases cancer cell migration and invasion.

Figure 2
figure2

Sera from breast cancer patients increase breast cancer progression. (a) The levels of CXCL5 in breast cancer patient sera. Breast cancer patient sera enhanced the migratory (b) and invasive (c) ability of breast cancer cells. (d) Cancer patient sera’s enhancement of migration and invasion correlates with CXCL5 levels, which were assessed by ELISA. Horizontal bars represent means. The invasiveness and migration ability of MCF-7 cells were quantified by QCM 24-well Cell Migration and Invasion assay. Healthy donor sera (15%) or breast cancer patient sera (15%) were used as a chemoattractant for cancer migration and invasion. Each value is the mean±s.d. of three independent experiments. The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). The relationship was analyzed using linear regression. All experiments were performed independently at least 3 times.

To better understand the role of CXCL5, we blocked the effect of CXCL5 on MCF-7 and MDA-MB-231 migration and invasion by using CXCL5-specific receptor (CXCR2) neutralization antibody. As shown in Figure 3a, both MCF-7 and MDA-MB-231 express detectable levels of CXCR2. The migration and invasion of MCF-7-OB-CM and MDA-MB-231-OB-CM in MCF-7 and MDA-MB-231 cells was neutralized by treatment with anti-CXCR2 antibody, but not by isotype control (Figures 3b and c). In addition, anti-CXCR2 antibody decreased the stimulatory effects of the chemotaxin CXCL5 in breast cancer cell migration and invasion (Figures 3d and e). Similarly, depletion of CXCL5 by immunoprecipitation from MCF-7-OB-CM and MDA-MB-231-OB-CM also reduced the migration and invasion of MCF-7 and MDA-MB-231 cells (Figures 3f and g). Furthermore, MCF-7 cells were neutralized by treatment with anti-CXCR2 antibody, preventing the induction of cell migration and invasion via the sera of breast cancer patients (Figures 3h and i). These data suggest that the migration and invasion of MCF-7 and MDA-MB-231 induced by CXCL5 is dependent on CXCR2 receptors.

Figure 3
figure3

Neutralizing CXCL5-specific receptor by its specific CXCR2 antibody decreased TAOB-CM-mediated cancer progression. (a) The expression of CXCR2. CXCR2 antibody decreased the effect of TAOB-CM on cell migration (b) and invasion (c). CXCR2 antibody decreased the induction of CXCL5 on cell migration (d) and invasion (e). Depletion of CXCL5 from TAOB-CM decreased the induction of TAOB-CM on cell migration (f) and invasion (g). CXCR2 antibody decreased the effect of breast cancer patients’ sera on cell migration (h) and invasion (i). Depletion of CXCL5 was performed by CXCL5 antibody, followed by immunoprecipitation techniques. Breast cancer cells seeded in upper inserts were treated with or without CXCR2 antibody (5 μg/ml) or immunoglobulin G (IgG; 5 μg/ml), and TAOB-CM (20%), breast cancer patient sera (15%), CXCL5 (10 nM) or CXCL-5 depletion. TAOBs were added into lower wells to act as a chemoattractant agent for 24 h. The invasiveness and migration ability of MCF-7 and MDA-MB-231 cells were quantified by QCM 24-well Cell Migration and Invasion assay. Each value is the mean±s.d. of three independent experiments. The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). All experiments were performed independently at least 3 times.

To evaluate the contribution of CXCL5 to the induction of MCF-7-CM- and MDA-MB-231-CM-associated osteoblasts in cancer progression, we interfered with the effects of CXCL5 by short hairpin RNA (shRNA) transfection. Compared with nontarget shRNA-transfected osteoblasts, CXCL5 shRNA effectively decreased CXCL5 expression in osteoblasts by 80%, as determined by real-time PCR and enzyme-linked immunosorbent assay (ELISA; Supplementary Figure S2A). Knockdown of CXCL5 by shRNA from osteoblasts was sufficient to reverse the effect of MCF-7-CM and MDA-MB-231-CM on the induction of breast cancer migration and invasion (Supplementary Figures S2B and C).

TAOB-CM and CXCL5 activate ERK/MSK1/Elk and increase the expression of Snail

Previous studies have shown that CXCL5 stimulates cancer progression through ERK1/2.19 Consequently, we assessed the effect of TAOB-CM and CXCL5 on mitogen-activated protein kinase cascade expression. The results showed that exposure of MCF-7 and MDA-MB-231 cells to TAOB-CM or CXCL5 resulted in the phosphorylation of Raf, MEK1/2 and ERK1/2 and MSK1 (Figure 4a). On the other hand, the expression of Raf, MEK1/2 and ERK1/2 (unphosphorylated form) was unaltered by TAOB-CM or CXCL5 treatment in either MCF-7 or MDA-MB-231 cells. In addition, the downstream factors of ERK1/2, MSK1 and Elk-1 were activated via phosphorylation and subsequently translocated from the cytosol into the nuclei after TAOB-CM and CXCL5 treatment (Figures 4a and b).

Figure 4
figure4

TAOB-CM and CXCL5 promote cancer progression by mitogen-activated protein kinase (MAPK) cascade activation. The effect of CXCL5 on the activation of MAPKs (a) and the nuclear translocation of MSK1 and Elk-1 (b). The effect of ERK2 shRNA on MSK1 and Elk-1 phosphorylation (c), cell migration (d) and cell invasion (e) in MCF-7 and MDA-MB-231 cells. Cells were transfected with control shRNA or ERK2 shRNA by Lipofectamine 2000 agents (Invitrogen), and then treated with TAOB-CM (20%) or CXCL5 (10 nM) for the indicated times (1 h for MSK1 and Elk-1 and 24 h for E-cadherin, cell migration and invasion assay). The invasiveness and migration ability of MCF-7 and MDA-MB-231 cells were quantified by QCM 24-well Cell Migration and Invasion assay. The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). All experiments were performed independently at least 3 times.

We next investigated the role of ERK1/2 on TAOB-CM- and CXCL5-mediated activation of MSK1 and Elk-1, as well as on cell migration and invasion. Transfection of MCF-7 and MDA-MB-231 cells with ERK2 shRNA reduced the phosphorylation of ERK2 induced by TAOB-CM and CXCL5 (Figure 4c). Selective genetic inhibition of ERK abrogated TAOB-CM- and CXCL5-mediated MSK1 and Elk-1 phosphorylation in cancer cells (Figure 4c). In addition, specific knockdown of ERK2 expression by ERK2 shRNA decreased cell migration and invasion in TAOB-CM- and CXCL5-treated cells (Figures 4d and e). These data suggest that ERK1/2 is the upstream event of MSK1 and Elk-1, which are involved in TAOB- and CXCL5-mediated cancer progression.

Snail is involved in TAOB-CM- and CXCL5-mediated EMT and cell progression

Snail has been reported to decrease E-cadherin expression by means of transcription repression, inducing EMT and cell migration.6, 9 Consequently, we assessed whether TAOB-CM and CXCL5 caused EMT, migration and invasion through Snail. Immunoblot data show that TAOB-CM and CXCL5 increase Snail levels while decreasing the expression of E-cadherin in both MCF-7 and MDA-MB-231 cells (Figures 5a, 1c and g). We next investigated the role of Snail on TAOB-CM- and CXCL5-mediated cell migration and invasion by shRNA transfection. As shown in Figure 5b, in comparison with control shRNA, Snail shRNA reduced Snail expression by 70%. Inhibition of Snail by shRNA transfection blocks the TAOB-CM- and CXCL5-mediated downregulation of E-cadherin in cancer cells (Figure 5c). On the other hand, ERK1/2 phosphorylation was unaltered by TAOB-CM and CXCL5 treatment in either Snail-knockdown MCF-7 or MDA-MB-231 cells. In contrast, selective genetic inhibition of ERK2 abrogated TAOB-CM- and CXCL5-mediated Snail upregulation and E-cadherin downregulation in cancer cells (Figure 5d). These data suggest that Snail is an upstream factor of E-cadherin but not of ERK1/2. Furthermore, Snail shRNA also decreased TAOB-CM- and CXCL5-mediated cancer cell migration and invasion (Figures 5e and f).

Figure 5
figure5

The role of Snail on TAOB-CM- and CXCL5-mediated EMT. (a) TAOB-CM and CXCL5 increased the expression of Snail. (b) Snail shRNA decreased Snail mRNA in both cell lines. (c) The effect of Snail shRNA on Snail and E-cadherin expression. (d) ERK2 shRNA decreased TAOB-CM- and CXCL5-mediated Snail upregulation. Snail shRNA decreased TAOB- and CXCL5-mediated cell migration (e) and invasion (f). Cells were transfected with control shRNA or ERK2 shRNA or Snail shRNA, and then treated with TAOB-CM (20%) or CXCL5 (10 nM) for the indicated times (24 h for E-cadherin, cell migration and invasion assay). Snail mRNA levels were assessed by reverse transcriptase–PCR (RT–PCR) after 72 h of transfection. The invasiveness and migration ability of MCF-7 and MDA-MB-231 cells were quantified by QCM 24-well Cell Migration and Invasion assay. The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). All experiments were performed independently at least 3 times.

MSK1 cooperates with Elk-1 to regulate the expression of Snail by histone acetylation and phosphorylation

Activated MSK1 has been reported to associate with Elk-1, and then to increase the expression of the cancer invasion-related gene by enhancing the phosphorylation and acetylation of histone H3.20 We used the TFsearch computational tool to search for Elk-1-binding sites in Snail promoter. There are three Elk-1-binding sites present at −2000 to +1000 from the transcription start site in Snail promoter, suggesting that Elk-1 may regulate the expression of Snail. To examine whether MSK1 recruits Elk-1, we first determined the interaction of MSK1 and Elk-1 by immunoprecipitation. As shown in Figure 6a, TAOB-CM and CXCL5 increased the association of MSK1 with Elk-1. Chromatin immunoprecipitation assay also verified that Elk-1 bound to the promoter of Snail at the −438 to 388 region after TAOB-CM and CXCL5 treatment (Figure 6b). Acetylation and phosphorylation of histone H3 on the Elk-1-binding region of Snail promoter was thereby increased (Figure 6b). In addition, inhibition of Elk-1 by specific shRNA transfection decreases the upregulation of TAOB-CM and CXCL5 on Snail expression (Figure 6c). Acetylation and phosphorylation of histone H3 on the Elk-1-binding region of Snail promoter was also abrogated by ElK-1 shRNA (Figure 6d). Inhibition of MSK1 also reduces the upregulation of Snail expression by TAOB-CM and CXCL5 in MCF-7 and MDA-MB-231 cells (Figure 6e). Acetylation and phosphorylation of histone H3 on the Elk-1-binding region of Snail promoter was also abrogated by MSK1 shRNA (Figure 6f). These data suggest that MSK1/Elk-1 cooperation is involved in the regulation of Snail transcription.

Figure 6
figure6

MSK1/Elk-1 increases the expression of Snail by epigenetic regulation. (a) The interaction of MSK1 and Elk-1. (b) TAOB-CM and CXCL5 increased the binding of Elk-1 and modification of histone H3 on the Elk-1-binding region of Snail promoter. Inhibition of Elk-1 prevents TAOB-CM- and CXCL5-mediated Snail upregulation (c) and histone modification of Snail promoter (d). Knockdown of MSK1 prevents TAOB-CM- and CXCL5-mediated Snail upregulation (e), and histone modification of Snail promoter (f). Cells were treated with CXCL5 (10 nM) for the indicated times. The interaction of MSK1 and Elk-1 was assessed by immunoprecipitation. Histone acetylation and phosphorylation and MSK1/Elk-1 binding on Snail were determined by chromatin immunoprecipitation. Cells were transfected with control shRNA or ERK2 shRNA or Snail shRNA, and then treated with TAOB-CM (20%) or CXCL5 (10 nM) for the indicated times (1 h for ERK phosphorylation, 6 h for Snail and 3 h for histone modification). The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). All experiments were performed independently at least 3 times.

CXCL5 increases breast cancer metastasis in vivo

We next used animal experiments to determine the role of CXCL5 in breast cancer metastasis in mice. Mouse breast cancer cell line 4T1 was injected into the mammary fat pads in mice, and then the cells were allowed to spontaneously metastasize from a primary tumor in vivo. Before 4T1 transplantation, we assessed whether CXCL5 increases the ability of 4T1 in migration, invasion and EMT. As shown in Figures 7a and b, CXCL5 increased migration and invasion of 4T1 cells. CXCL5 also caused 4T1 cells to undergo EMT, including the downregulation of epithelial marker (E-cadherin), and upregulation of fibroblast markers (vimentin and N-cadherin) (Figure 7c). These data coincide with human breast cancer data. ELISA data show that the sera obtained from 4T1-bearing mice revealed high levels of CXCL5 compared with the sera of control mice (Figure 7d).

Figure 7
figure7

CXCL5 increases the metastasis of breast cancer in vivo. CXCL5 increases migration (a), invasion (b) and EMT (c) in 4T1 cells. (d) Levels of CXCL5 in the sera of mice. (e) CXCL5 increased the number of tumor nodules in the organs. CXCL5 increased metastasis, as revealed by photographs (f) and hematoxylin and eosin (H&E) staining (g). Neutralization of CXCL5 decreased the occurrence of 4T1 metastasis, as revealed by the number of tumor nodules (h), photographs (i) and H&E staining (j). Invasiveness and migration ability of 4T1 cells was quantified by QCM 24-well Cell Migration and Invasion assay. CXCL5 (10 nM) was used as chemoattractant for cancer migration and invasion. 4T1 cells were treated with CXCL5 (10 nM) for 24 h and the expression of various proteins assessed by immunoblot. 4T1 cells were injected into the mammary fat pads of mice. Mice were treated every 3 days with intraperitoneal (i.p.) injections of phosphate-buffered saline (PBS; n=8) or CXCL5 (0.5 mg/kg; n=8). For CXCL5 inhibition, 4T1 was injected into the fat pads of mice, and then mice were treated every 7 days with i.p. injections of immunoglobulin G (IgG; n=7) or CXCL5 antibodies (50 μg/mice; n=8). After 24 days, organs were harvested, cut and H&E stained, and then analyzed by microscopy. The number of tumor nodules in various organs was recorded for analysis of cancer incidence. The asterisk indicates a significant difference with the control, as analyzed by analysis of variance (ANOVA) with Student’s t-test post hoc (*P<0.05). T, tumor. Arrows indicate tumor nodules.

Next, we assessed whether CXCL5 increases the metastasis of 4T1 cells. Treatment of mice by CXCL5 (0.5 mg/kg) increased the metastasis of 4T1 cells in 100% (8 of 8) of the mice, in comparison with 75% (6 of 8) of the control mice. Tumor nodules in various organs (livers, lungs and intestines) of the CXCL5-treated mice were both more numerous and larger than those found in the control group of mice (Figures 7e–g). Mice treated with CXCL5 also contained larger metastatic nodules in lungs, livers and intestines than the lungs of the control mice.

Finally, we assessed whether inhibiting CXCL5 decreases the metastasis of cancer. Treatment of mice by anti-CXCL5 monoclonal antibody inhibited the metastasis of 4T1 cells in 62.5% (5 of 8) of the mice, in comparison with 100% (8 of 8) of the control mice. The tumor nodules in various organs (livers, lungs and intestines) of the CXCL5 antibody-treated mice were far fewer than those found in the control group mice (Figures 7h–j).

Discussion

Bone metastasis has permitted the identification of factors, produced by the tumor cells themselves or by the microenvironment in response to the tumor, that mediate the process of bone destruction.21 This study is the first to investigate the interaction between tumors and the bone microenvironment that results in tumor progression. TAOB-CM, or invasive breast cancer patient sera containing high levels of CXCL5, also significantly increase the migration and invasion of breast cancer cells. Pretreatment of breast cancer cells with CXCL5-specific CXCR2 neutralization antibody decreases TAOB-CM or invasive breast cancer patient sera-mediated cancer cell migration and invasion, suggesting that CXCL5 may be a major factor enhancing the metastatic ability of breast cancer cells.

EMT has been characterized as a dynamic and reversible biological behavior. When cancer cells escape from the primary site by the circulation and metastasize at secondary sites by switching epithelial phenotype into mesenchymal phenotype, these cancer cells may return to an epithelial phenotype at secondary sites.22, 23, 24 EMT, described by a conversion in the cell type-specific protein pattern and promotion of cell movement, is vital for the generation of new tissue types during embryonic development, and plays a critical role in inflammatory response and wound healing of adult tissue.25 It has also been implicated in important steps of cancer development, such as invasion and metastasis.26 Ectopic expression of Snail has been reported to be a sufficient activator of EMT, as well as being a predictor for an aggressive malignancy.27, 28 In breast cancer, increased Snail expression has been associated with lymph node involvement, invasiveness and metastatic potential,28, 29, 30 and also with decreased rates of recurrence-free survival.30 Knockdown of Snail by RNA interference in MDCK-Snail cells induces a complete MET, associated with the increase of E-cadherin, decrease of mesenchymal markers and inhibition of invasiveness.27, 31 Furthermore, inhibition of endogenous Snail leads to a notable decrease of in vivo tumor growth, accompanied by enhanced tumor differentiation and reduced invasiveness.27 In our study, we found that treating MCF-7 and MDA-MB-231 cells with TAOB-CM or CXCL5 dramatically increases Snail expression, accompanied by the decrease of the epithelial marker E-cadherin and increase of mesenchymal markers vimentin and N-cadherin, suggesting that the interaction between the tumor and the bone microenvironment induces EMT of cancer. Selective inhibition of Snail by shRNA decreased the effects of TAOB-CM- and CXCL5-mediated E-cadherin downregulation, and cancer cell migration and invasion, suggesting that the upregulation of Snail plays a critical role in TAOB-CM- and CXCL5/ENA78-mediated cell migration and EMT in breast cancer.

Compared with normal tissue or benign lesions, ERK activity is increased in invasive breast carcinomas.32, 33, 34 As invasive cancer cells are migratory, it is reasonable to postulate that the enzymatic ERK activity could be involved in cancer cell migration.35 Chen et al.32 suggest that the ERK pathway regulates breast cancer cell migration by maintaining Snail expression. We found that treatment of MCF-7 and MDA-MB-231 cells treated with TAOB-CM or CXCL5 resulted in increased phosphorylation of mitogen-activated protein kinase cascade (Raf, MEK1/2 and ERK1/2). Furthermore, inhibition of ERK with ERK2 shRNA completely abrogated TAOB-CM- and CXCL5-mediated cell migration, invasion and Snail upregulation, whereas Snail shRNA did not affect the phosphorylation of ERK1/2. These data suggest that Raf/MEK1/2/ERK1/2 signaling is the major upstream event of Snail.

ERKs are well known to activate several transcription factors and co-activators by phosphorylation, and thus to regulate specific gene expressions by controlling epigenetics and transcription dynamics.17, 36 Transcription factor Elk-1 is phosphorylated (activated) by ERKs and recruited to the serum response elements of target gene promoter. Activated Elk-1 exhibits transcriptional activation through the formation of a complex with co-activators. The recruitment of ElK-1 and MSK1 has been reported to regulate the expression of c-fos and egr-1 stimulated by mitogens, and this transcriptional stimulation correlates with histone H3 acetylation and phosphorylation.37, 38 Our data are the first to demonstrate that CXCL5 increases the activation of MSK1 and Elk-1, which in turn form a transcriptional complex that enhances the acetylation and phosphorylation of histone H3 of Snail promoter, resulting in the upregulation of Snail. TAOB-CM- and CXCL5-mediated MSK1 and Elk-1 are dependent on ERK1/2. Knockdown of MSK1 by specific shRNA prevents the binding of Elk-1 and modification of histone H3 of Snail promoter. Indeed, genetic blockade of Elk-1 attenuated the acetylation and phosphorylation of histone H3 of Snail promoter, suggesting cooperation between MSK1 and Elk-1 in the regulation of Snail.

Taken together, our findings provide the first evidence that TAOB-derived CXCL5 reinforces breast cancer cell migration and invasion. Our results also clearly establish that ERK/MSK1/Elk-1 regulates Snail expression by epigenetic regulation, and enhances cancer progression ability (Figure 8). Neutralization of CXCL5 by CXCL5 antibody significantly decreases the incidence of cancer development in mice. This novel finding could elucidate the interaction between tumor cells and the osteoblast microenvironment, which results in a vicious cycle accelerating the malignancies of late-stage breast cancer patients. These findings on the interaction of cancer cells and osteoblasts provide important insights into the biology of tumors, and establish new rationales for using anti-CXCL5 strategies to target metastasis in breast cancer.

Figure 8
figure8

Proposed interactions between cancer and TAOBs. Breast cancer increases osteoblasts’ secretion of high levels of CXCL5, which in turn enhances the ability of breast cancer cell migration, invasion and EMT. TAOB-derived CXCL5 triggers Raf/MEK/ERK and recruitment of the Elk-1 and MSK1, which in turn enhance transcription of Snail, a repressor of E-cadherin.

Materials and methods

Cell cultures and condition medium

Human primary osteoblasts were obtained from Lonza (Walkersville, MD, USA) and cultured in Osteoblast Growth Media. MCF-7 human breast adenocarcinoma cells (ATCC HTB-22) were cultured in Minimum Essential Medium (MEM; Life Technologies, Inc., Grand Island, NY, USA) with nonessential amino acids, 0.1 mM sodium pyruvate, 10 μg/ml insulin and 10% fetal bovine serum MDA-MB-231 human breast adenocarcinoma cells (ATCC HTB-26) and the 4T1 mouse breast cancer cell line (CRL-2539) were cultured in MEM and RPMI-1640 (Life Technologies, Inc) supplemented with 10% fetal bovine serum. To obtain the MCF-7 and MDA-MB-231-conditioned medium (CM), 2 × 106 cells were seeded in a 100-mm dish and cultivated for 24 h. MCF-7 tumor-associated osteoblasts (MCF-7-OBs) and MDA-MB-231-OBs were generated by culturing human osteoblasts presenting in MCF-7-CM and MDA-MB-231-CM. After washing, the medium was replaced and the supernatants harvested after 24 h of incubation. The results were designated as MCF-7-OB-conditioned medium (MCF-7-OB-CM) and MDA-MB-231-OB-CM (Supplementary Figure S1A). CXCL5 was neutralized using anti-CXCR2 antibody (5 μg/ml, clone 48 311, R&D Systems, Minneapolis, MN, USA). CXCL5 depletion from MCF-7-OB-CM and MDA-MB-231-OB-CM was performed using anti-CXCL5 antibodies (2 μg/ml, Ab9802; Abcam, Cambridge, MA, USA) and Sepharose A/G beads following regular immunoprecipitation techniques. Cytokine depletion was confirmed by CXCL5 ELISA assay kit.

Cell proliferation

Cells were plated in 96-well culture plates. After 24 h of incubation, the cells were treated with vehicle control-CM, various condition media or various concentrations of CXCL5 (R&D Systems Europe, Abingdon, UK) for 72 h. At the end of the assay period, cell proliferation was measured by WST-1 assay. Cell proliferation was determined by Premixed WST-1 Cell Proliferation Reagent (Clontech Laboratories Inc., Mountain View, CA, USA) in accord with the manufacturer’s instructions.

Cell migration and invasion assay

Cell migration and invasion assay was conducted using QCM 24-well Cell Migration and Invasion Assay kits (Millipore Corp., Bedford, MA, USA). Briefly, cells were seeded into the migration chamber, and control medium, various condition media or various concentrations of CXCL5 were added to the bottom wells for 24 h as chemoattractant. At the end of the treatment, the cells were poststained with CyQuant GR dye in cell lysis buffer for 15 min at room temperature. Fluorescence of the migrated and invaded cells was then read using a fluorescence plate reader at excitation/emission wavelengths of 485/520 nm.

Measurement of secreted factors

Human primary osteoblasts were treated with MCF-7-CM or MDA-MB-231-CM for 24 h. After washing, the supernatants were collected. CXCL5 levels were then assessed by CXCL5 ELISA kit.

Serum samples from breast cancer patients

Preoperative blood samples were obtained from 15 stage I and 15 stage III breast cancer patients and 15 healthy donors admitted to the Department of Surgery, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan. Serum was separated by centrifugation and frozen at –80 °C. Approval for these studies was obtained from the institutional review board of Kaohsiung Medical University Hospital, and informed consent was obtained from all patients in accordance with the Declaration of Helsinki Principles.

Immunoblot/immunoprecipitation

Cells were treated with MCF-7-OB-CM (20%), MDA-MB-231-OB-CM (20%) or 10 nM CXCL5 for the indicated times and then lysed by M-PER solution on ice for 15 min. Cell lysate was centrifuged at 14 000 g for 15 min, and the supernatant fraction collected for immunoblot. Equivalent amounts of protein were resolved by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (10–12%) and transferred to polyvinylidene difluoride membranes. After blocking for 1 h in 5% nonfat dry milk in Tris-buffered saline, the membranes were incubated with the desired primary antibody for 1–16 h. The membranes were then treated with appropriate peroxidase-conjugated secondary antibody, and the immunoreactive proteins detected using an enhanced chemiluminescence kit (Amersham, Buckinghamshire, UK) according to the manufacturer’s instructions. The primary antibodies used in this study to target unphosphorylated and phospho-ERK proteins, vimentin, N-cadherin and E-cadherin, were obtained from Cell Signaling Technology (Danvers, MA, USA). Snail antibody was obtained from Abcam, Ltd (Cambridge, UK). β-Actin and glyceraldehyde-3-phosphate dehydrogenase were obtained from Upstate Biotechnology (Lake Placid, NY, USA).

Real-time reverse transcriptase–PCR

RNA isolation was performed using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Complementary DNA was prepared using an oligo (dT) primer and reverse transcriptase (Takara, Shiga, Japan) following standard protocols. Real-time PCR was performed using SYBR Green on the ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA). Each PCR reaction mixture contained 200 nM of each primer, 10 μl of 2 × SYBR Green PCR Master Mix (Applied Biosystems), 5 μl complementary DNA and RNase-free water, with a total volume of 20 μl. The PCR reaction was carried out with a denaturation step at 95 °C for 10 min, and then for 32 cycles at 95 °C for 15 s and 60 °C for 1 min All PCRs were performed in triplicate and normalized to internal control glyceraldehyde-3-phosphate dehydrogenase mRNA. Relative expression was presented using the 2CT method.

Gene knockdown by shRNA

Knockdown of CXCL5, Snail, ERK2, MSK1 and Elk-1 was performed using the lentiviral expression system provided by the National RNAi Core Facility (Taipei, Taiwan). The lentiviruses were produced by co-transfecting HEK293T with pLKO-AS2, pLKO-AS2-CXCL5, pLKO-AS2-Snail, pLKO-AS2-ERK2, pLKO-AS2-MSK1 or pLKO-AS2-Elk-1 shRNA and two packaging plasmids (pCMVVDR8.91 and pMD.G). The efficacy of shRNA plasmids was assessed by real-time PCR and ELISA.

Chromatin immunoprecipitation

Chromatin immunoprecipitation was performed by Agarose ChIP Kit (Pierce, Rockford, IL, USA). Immunoprecipitated DNA was analyzed by quantitative PCR (Light Cycler, as described above) using SYBR Green dye. The primers used were 5′-IndexTermGCTTCCTCCCCAGTGATGT-3′ and 5′-IndexTermTCAGAAGCGCTCAGACCAC-3′ for Snail, yielding a 171-bp fragment.

The in vivo orthotropic tumor metastasis

4T1 cells (1 × 105) were injected into the mammary fat pads of 8-week-old female BALB/c mice. CXCL5 (0.5 mg/kg) was administered by intraperitoneal injection three times every 3 days after transplantation. Animals were killed on day 24 after 4T1 transplantation, and the number of tumor nodules in all organs of these mice was recorded for analysis of metastasis incidence. For CXCL5 inhibition, isotype control immunoglobulin G and CXCL5 neutralization antibody (50 μg/mice, clone 61 905, R&D Systems) were administered every 7 days via intraperitoneal injection three times after administration of 4T1. Animals were killed on day 24 after 4T1 transplantation and the numbers of tumor nodules in the various organs were recorded for analysis of 4T1 metastasis. All animal experiments were approved by the local ethics committee.

Statistical analysis

Data were expressed as means±s.e. Statistical comparisons of the results were made using analysis of variance. Significant differences (P<0.05) between the means of test groups were analyzed by Student’s t-test.

References

  1. 1

    Onishi T, Hayashi N, Theriault RL, Hortobagyi GN, Ueno NT . Future directions of bone-targeted therapy for metastatic breast cancer. Nat Rev Clin Oncol 2010; 7: 641–651.

    CAS  Article  Google Scholar 

  2. 2

    Coleman RE, Lipton A, Roodman GD, Guise TA, Boyce BF, Brufsky AM et al. Metastasis and bone loss: advancing treatment and prevention. Cancer Treat Rev 2010; 36: 615–620.

    Article  Google Scholar 

  3. 3

    Chen Y, Shi HY, Stock SR, Stern PH, Zhang M . Regulation of breast cancer-induced bone lesions by β-catenin protein signaling. J Biol Chem 2011; 49: 42 575–42584.

    Article  Google Scholar 

  4. 4

    Van't Veer LJ, Weigelt B . Road map to metastasis. Nat Med 2003; 9: 999–1000.

    CAS  Article  Google Scholar 

  5. 5

    Hanahan D, Weinberg RA . The hallmarks of cancer. Cell 2000; 100: 57–70.

    CAS  Article  Google Scholar 

  6. 6

    Yuen HF, Chan YK, Grills C, McCrudden CM, Gunasekharan V, Shi Z et al. Polyomavirus enhancer activator 3 protein promotes breast cancer metastatic progression through Snail-induced epithelial-mesenchymal transition. J Pathol 2011; 224: 78–89.

    CAS  Article  Google Scholar 

  7. 7

    Thiery JP, Acloque H, Huang RY, Nieto MA . Epithelial-mesenchymal transitions in development and disease. Cell 2009; 139: 871–890.

    CAS  Article  Google Scholar 

  8. 8

    Yang J, Weinberg RA . Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell 2008; 14: 818–829.

    CAS  Article  Google Scholar 

  9. 9

    Taube JH, Herschkowitz JI, Komurov K, Zhou AY, Gupta S, Yang J et al. Core epithelial-to-mesenchymal transition interactome gene-expression signature is associated with claudin-low and metaplastic breast cancer subtypes. Proc Natl Acad Sci USA 2010; 107: 15 449–15454.

    Article  Google Scholar 

  10. 10

    Jordà M, Vinyals A, Marazuela A, Cubillo E, Olmeda D, Valero E et al. Id-1 is induced in MDCK epithelial cells by activated Erk/MAPK pathway in response to expression of the Snail and E47 transcription factors. Exp Cell Res 2007; 313: 2389–2403.

    Article  Google Scholar 

  11. 11

    Dunn KL, Espino PS, Drobic B, He S, Davie JR . The Ras-MAPK signal transduction pathway, cancer and chromatin remodeling. Biochem. Cell Biol 2005; 83: 1–14.

    CAS  Google Scholar 

  12. 12

    Espino PS, Li L, He S, Yu J, Davie JR . Chromatin modification of the trefoil factor 1 gene in human breast cancer cells by the Ras/mitogen-activated protein kinase pathway. Cancer Res 2006; 66: 4610–4616.

    CAS  Article  Google Scholar 

  13. 13

    Pacheco-Rodriguez G, Kumaki F, Steagall WK, Zhang Y, Ikeda Y, Lin JP et al. Chemokine-enhanced chemotaxis of lymphangioleiomyomatosis cells with mutations in the tumor suppressor TSC2 gene. J Immunol 2009; 182: 1270–1277.

    CAS  Article  Google Scholar 

  14. 14

    Bièche I, Chavey C, Andrieu C, Busson M, Vacher S, Le Corre L et al. CXC chemokines located in the 4q21 region are up-regulated in breast cancer. Endocr Relat Cancer 2007; 14: 1039–1052.

    Article  Google Scholar 

  15. 15

    Kuo PL, Hung JY, Huang SK, Chou SH, Cheng DE, Jong YJ et al. Lung cancer-derived galectin-1 mediates dendritic cell anergy through inhibitor of DNA binding 3/IL-10 signaling pathway. J Immunol 2011; 186: 1521–1530.

    CAS  Article  Google Scholar 

  16. 16

    Georges S, Ruiz Velasco C, Trichet V, Fortun Y, Heymann D, Padrines M . Proteases and bone remodelling. Cytokine Growth Factor Rev 2009; 20: 29–41.

    CAS  Article  Google Scholar 

  17. 17

    Hsu YL, Huang MS, Yang CJ, Hung JY, Wu LY, Kuo PL . Lung tumor-associated osteoblast-derived bone morphogenetic protein-2 increased epithelial-to-mesenchymal transition of cancer by Runx2/Snail signaling pathway. J Biol Chem 2011; 286: 37 335–37346.

    Article  Google Scholar 

  18. 18

    Chirgwin JM, Guise TA . Molecular mechanisms of tumor-bone interactions in osteolytic metastases. Crit Rev Eukaryot Gene Expr 2000; 10: 159–178.

    CAS  Article  Google Scholar 

  19. 19

    Kuo PL, Chen YH, Chen TC, Shen KH, Hsu YL . CXCL5/ENA78 increased cell migration and epithelial-to-mesenchymal transition of hormone-independent prostate cancer by early growth response-1/snail signaling pathway. J Cell Physiol 2011; 226: 1224–1231.

    CAS  Article  Google Scholar 

  20. 20

    Besnard A, Bouveyron N, Kappes V, Pascoli V, Pagès C, Heck N et al. Alterations of molecular and behavioral responses to cocaine by selective inhibition of Elk-1 phosphorylation. J Neurosci 2011; 31: 14296–14307.

    CAS  Article  Google Scholar 

  21. 21

    Roodman GD . Mechanisms of bone metastasis. N Engl J Med 2004; 350: 1655–1664.

    CAS  Article  Google Scholar 

  22. 22

    Thiery JP, Sleeman JP . Complex networks orchestrate epithelial-mesenchymal transitions. Nat Rev Mol Cell Biol 2006; 7: 131–142.

    CAS  Article  Google Scholar 

  23. 23

    Christiansen JJ, Rajasekaran AK . Reassessing epithelial to mesenchymal transition as a prerequisite for carcinoma invasion and metastasis. Cancer Res 2006; 66: 8319–8326.

    CAS  Article  Google Scholar 

  24. 24

    Thiery JP . Epithelial-mesenchymal transitions in development and pathologies. Curr Opin Cell Biol 2003; 15: 740–746.

    CAS  Article  Google Scholar 

  25. 25

    Nieto MA . The ins and outs of the epithelial to mesenchymal transition in health and disease. Annu Rev Cell Dev Biol 2011; 27: 347–376.

    CAS  Article  Google Scholar 

  26. 26

    Peinado H, Olmeda D, Cano A . Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat Rev Cancer 2007; 7: 415–428.

    CAS  Article  Google Scholar 

  27. 27

    Olmeda D, Jordá M, Peinado H, Fabra A, Cano A . Snail silencing effectively suppresses tumour growth and invasiveness. Oncogene 2007; 26: 1862–1874.

    CAS  Article  Google Scholar 

  28. 28

    Blanco MJ, Moreno-Bueno G, Sarrio D, Locascio A, Cano A, Palacios J et al. Correlation of Snail expression with histological grade and lymph node status in breast carcinomas. Oncogene 2002; 21: 3241–3246.

    CAS  Article  Google Scholar 

  29. 29

    Cheng CW, Wu PE, Yu JC, Huang CS, Yue CT, Wu CW et al. Mechanisms of inactivation of E-cadherin in breast carcinoma: modification of the two-hit hypothesis of tumor suppressor gene. Oncogene 2011; 20: 3814–3823.

    Article  Google Scholar 

  30. 30

    Moody SE, Perez D, Pan TC, Sarkisian CJ, Portocarrero CP, Sterner CJ et al. The transcriptional repressor Snail promotes mammary tumor recurrence. Cancer Cell 2005; 8: 197–209.

    CAS  Article  Google Scholar 

  31. 31

    Cano A, Pérez-Moreno MA, Rodrigo I, Locascio A, Blanco MJ, del Barrio MG et al. The transcription factor snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression. Nat Cell Biol 2000; 2: 76–83.

    CAS  Article  Google Scholar 

  32. 32

    Chen H, Zhu G, Li Y, Padia RN, Dong Z, Pan ZK et al. Extracellular signal-regulated kinase signaling pathway regulates breast cancer cell migration by maintaining slug expression. Cancer Res 2009; 69: 9228–9235.

    CAS  Article  Google Scholar 

  33. 33

    Mueller H, Flury N, Eppenberger-Castori S, Kueng W, David F, Eppenberger U . Potential prognostic value of mitogen-activated protein kinase activity for disease-free survival of primary breast cancer patients. Int J Cancer 2000; 89: 384–388.

    CAS  Article  Google Scholar 

  34. 34

    Sivaraman VS, Wang H, Nuovo GJ, Malbon CC . Hyperexpression of mitogen-activated protein kinase in human breast cancer. J Clin Invest 1997; 99: 1478–1483.

    CAS  Article  Google Scholar 

  35. 35

    McCubrey JA, Steelman LS, Chappell WH, Abrams SL, Wong EW, Chang F et al. Roles of the Raf/MEK/ERK pathway in cell growth, malignant transformation and drug resistance. Biochim Biophys Acta 2007; 1773: 1263–1284.

    CAS  Article  Google Scholar 

  36. 36

    Zhang HM, Li L, Papadopoulou N, Hodgson G, Evans E, Galbraith M et al. Mitogen-induced recruitment of ERK and MSK to SRE promoter complexes by ternary complex factor Elk-1. Nucleic Acids Res 2008; 36: 2594–2607.

    CAS  Article  Google Scholar 

  37. 37

    Cassinat B, Zassadowski F, Ferry C, Llopis L, Bruck N, Lainey E et al. New role for granulocyte colony-stimulating factor-induced extracellular signal-regulated kinase 1/2 in histone modification and retinoic acid receptor α recruitment to gene promoters: relevance to acute promyelocytic leukemia cell differentiation. Mol Cell Biol 2011; 31: 1409–1418.

    CAS  Article  Google Scholar 

  38. 38

    Vicent GP, Ballaré C, Nacht AS, Clausell J, Subtil-Rodríguez A, Quiles I et al. Induction of progesterone target genes requires activation of Erk and Msk kinases and phosphorylation of histone H3. Mol Cell 2006; 24: 367–381.

    CAS  Article  Google Scholar 

Download references

Acknowledgements

This study is supported by grants from the National Science Council of Taiwan (NSC 99-2320-B-037-017-MY3), the Excellence for Cancer Research Center Grant, the Department of Health, Executive Yuan, Taipei, Taiwan (DOH101-TD-C-111-002), the Kaohsiung Medical University Hospital (KMUH 99–9I04) and the Kaohsiung Medical University Research Foundation (KMUER008).

Author information

Affiliations

Authors

Corresponding author

Correspondence to Y-L Hsu.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Hsu, Y., Hou, M., Kuo, P. et al. Breast tumor-associated osteoblast-derived CXCL5 increases cancer progression by ERK/MSK1/Elk-1/Snail signaling pathway. Oncogene 32, 4436–4447 (2013). https://doi.org/10.1038/onc.2012.444

Download citation

Keywords

  • breast cancer
  • osteoblasts
  • CXCL5
  • cancer progression

Further reading

Search