Abstract
Background/Aim: During the bone metastatic process, tumor cells and bone cells drive a vicious cycle stimulating growth and activity of each other. We here address how low molecular weight protein tyrosine phosphatase (LMW-PTP) could be involved in this process. Materials and Methods: We targeted LMW-PTP by siRNA and evaluated the effect of various soluble factors released to the culture medium by the MDA-MB-435 breast cancer cell line, in RAW 264.7 osteoclastogenesis. Results: We showed that these soluble factors did not change RAW 264.7 osteoclastogenic potential. The knockdown of the LMW-PTP slow isoform decreased osteoclastogenesis of RAW 264.7, showing less active Src. The knockdown of LMW-PTP and its slow isoform decreased the release of IL-8 but not IL-6 in MDA-MB-435. Conclusion: The LMW-PTP slow isoform can be an important protein in bone metastatic disease, with a fundamental role in the interplay between tumor cells and osteoclasts, through the regulation of Src activity and IL-8 secretion.
Breast cancer ranks among the most prevalent malignancies in women. Breast carcinoma frequently metastasizes to bone and approximately 70% of patients develop bone metastases (1). The process of metastasis, including the spread and growth of tumor cells in distant organs, is paramount to the definition of malignancy; once a tumor metastasizes to bone, it is usually incurable (2).
The development of bone metastases is associated with numerous debilitating skeletal-related events (SREs), pathological fractures, hypercalcemia, radiotherapy and surgery. SREs are associated with a considerable decrease of patients' quality of life and increased morbidity and mortality (3).
Breast cancer cells in bone do not directly resorb bone but can regulate the activity of osteoblasts and consequently osteoclasts causing disruption of bone remodeling (4). Bone derived growth factors will stimulate metastatic cancer cells, further promoting tumor growth in bone, driving the vicious cycle of bone metastases (5).
Patients with bone metastases are treated with anti-tumor drugs targeting tumor cells and with bone modifying agents, such as bisphosphonates, that inhibit osteoclastic bone resorption. Among patients undergoing bisphosphonates therapy 25% do not respond to treatment (6, 7); however, the mechanism underlying this therapeutic differential response is currently unknown. Therefore, it is important to address this question and understand which molecules are involved in this process.
Low-molecular weight protein tyrosine phosphatase (LMW-PTP), a polymorphic enzyme widely expressed in different tissues, has been considered as an important signaling molecule in osteoblast biology and bone metabolism (8, 9). Regarding tumorigenesis, LMW-PTP has recently been associated with different types of cancers (10), such as breast, lung (10), prostate (11) and colorectal (12). The two main isoforms of LMW-PTP, fast and slow, seem to have different roles in the development of breast tumor (13), with the fast genotypes having been positively associated with cancer (14, 15). LMW-PTP has also been shown to be involved in the regulation of Src activity (8, 9), an important protein for osteoclastic activity.
Based on the above, we hypothesized that LMW-PTP may be involved in bone metastatic disease and its isoforms may have a differential role in the communication between tumor cells and osteoclasts. Since the soluble factors produced by human or mouse breast cancer cells, such as interleukin-6 (IL-6) (16) and interleukin-8 (IL-8) (17), can directly stimulate osteoclast differentiation from late human or mouse osteoclast precursors (18), we tested how a specific knockdown (KD) of LMW-PTP and its slow isoform (KD LMW-PTP slow) in the MDA-MB-435 breast carcinoma cell line (13) can influence the differentiation of RAW 264.7 murine monocytic cells in osteoclasts. Our results show that the specific KD LMW-PTP slow decreases the potential of osteoclastic differentiation of RAW 264.7 cells probably due to the decrease of Src activation and lower levels of IL-8 produced by this KD.
Materials and Methods
Cell culture. MDA-MB-435, a breast cancer epithelial cell line, was obtained from ATCC (American Type Culture Collection, Manassas, VA, USA), cultured in Dulbecco's modified Eagle's Medium (DMEM) with 10% fetal bovine serum (FBS) and 1% pennicilin/streptomycin (ThermoFisher Scientific, Waltham, MA, USA). The conditioned medium (CM) was harvested when cells reached 80% confluence. The choice of this cell line was based on previous work from our group (13).
RAW 264.7 cells, murine monocyte cells for osteoclast differentiation, were kindly provided by Dr. Michael Rogers (Aberdeen, Scotland). These cells were maintained in DMEM, 10% heat inactivated FBS and 1% penicillin/streptomycin. For the differentiation in osteoclasts, cells were plated in 96-well plates, 950 cells/well in alpha-minimum essential medium Eagle (α-MEM; ThermoFisher Scientific) with 10% heat inactivated FBS, 1% penicillin/streptomycin and 100 ng/ml receptor activator of NFkB (RANKL). On day three, cells were supplemented with RANKL 100 ng/ml and either fresh medium or 60% CM and cultured for another 2 days. Supernatants were collected on day 5 for tartarate resistant acid phosphatase (TRAP) quantification and cells were stained for TRAP.
LMW-PTP knockdowns. Five different siRNA sequences were designed to specifically knockdown the total and the slow isoform (KD LMW-PTP slow). One scramble (non-targeted; KD NT) siRNA was used as control. The lentiviruses containing these sequences were a kind gift from Prof. Luis Moita (iMM Lisboa, Faculdade de Medicina, Portugal) who designed the sequences and constructed the lentiviral vectors. Infection and clone selection were performed as described by Alho et al. (19).
TRAP staining. Cells were stained for TRAP using the commercial Acid Phosphatase, Leukocyte (TRAP) Kit (Sigma-Aldrich, St. Louis, MO, USA) without the final step with acid hematoxilin. Osteoclasts were identified as multinucleated TRAP-positive cells, with more than three nuclei.
TRAP quantification. TRAP quantification was performed in the supernatant of cells using the commercial kit MouseTRAP™Assay (TRACP ELISA's – ids, UK), a solid phase immunofixed enzyme activity assay for the determination of osteoclast derived TRAP form 5b (TRACP 5b) in mouse serum. Results are expressed in U/l of TRAP.
Src phosphorylation. Src activation status was determined by Western blot analysis. MDA-MB-435 cells were seeded in 6-well plates and allowed to grow. At approximately 80% confluence, cells were washed once with PBS, lysed in 200 μl 2x SDS-loading buffer with protease and phosphatase inhibitors cocktails (Sigma-Aldrich) and heated to 95°C for 10 min. Samples were loaded onto a 10% polyacrylamide gel and electrophoresis was performed using a Mini-PROTEAN Tetra cell (BioRad, Hercules, CA, USA). Proteins were transferred onto a Protran BA85 nitrocellulose membrane (Whatman, GE Healthcare; Rochester, Kent, UK) using a Mini-PROTEAN Tetra Cell transfer system (BioRad). Membranes were blocked in PBST, 5% nonfat dry milk for 1 h, incubated overnight with the primary antibody and for 1h with the secondary antibody. Antibody detection was performed using SuperSignal West Pico Chemiluminescent HRP Substrate (Pierce, ThermoFisher Scientific) according to the manufacturer's instructions and signal was visualized on radiographic film. Antibodies against total Src (#2108), phospho Y527Src (#2107), non-phospho Y416Src (#2101), as well as anti-rabbit, anti-mouse peroxidase-conjugated antibodies were purchased from Cell Signaling Technologies (Boston, MA, USA). Total Src was used as loading control.
IL-6 and IL-8 quantification. IL-6 and IL-8 were measured in cell culture supernatants. Briefly, supernatants from MDA-MB-435, KD NT, KD LMW-PTP and KD LMW-PTP slow at approximately 80% confluence were harvested, centrifuged at 200 × g for 5 min and assayed immediately using a commercial ELISA kit for IL-6 and IL-8 quantification (Human IL-6 Quantikine HS ELISA kit and Human IL-8/CXCL8 Quantikine ELISA kit; R&D systems, Minneapolis, MN, USA). Results are expressed in pg/ml.
Results
Conditioned medium from MDA-MB-435 breast cancer cell line does not change osteoclastogenesis in RAW 264.7 cell line. Osteoclastogenesis was assessed using RAW 264.7 cells. In the total absence of RANKL (0 ng/ml), RAW cells did not differentiate into osteoclasts (negative control, Figure 1A and Figure 1F). In contrast, supplementation with 100 ng/ml RANKL (positive control, Figure 1B and Figure 1F) induced osteoclastogenesis. MDA-MB-435 CM did not change osteoclastogenesis compared to the positive control (Figure 1C and F). Since there were no differences between MDA-MB-435 (parental cell line) and KD NT, all results from KDs have been compared with the parental cell line.
KD LMW-PTP slow decreases the ability of breast cancer-derived factors to induce osteoclastogenesis. Comparison between the differentiation of RAW 264.7 cells in the presence of CM from different MDA-MB-435 knockdowns showed that only the KD LMW-PTP slow decreased osteoclastogenesis (Figure 1E and F).
KD LMW-PTP slow decreases activated Src in MDA-MB-435. Src is only fully activated when Tyr 527 is dephosphorylated and Tyr 416 phosphorylated. The increase in Tyr 527 phosphorylation and Tyr 416 dephosphorylation on the KD LMW-PTP slow compared with MDA-MB-435 (Figure 2) represented a decrease in Src activation, whereas total Src (loading control) was not changed.
Osteoclast differentiation depending on conditioned medium. RAW 264.7 cells were stimulated with RANKL (100 ng/ml), stained for TRAP and TRAP was quantified in cells' supernatants as described in the “Materials and Methods” section. A: Negative control: differentiation medium without RANKL; B-F: representative images of osteoclasts derived from RAW 264.7 cells in different differentiation media; B. positive control: differentiation medium with 100 ng/ml RANKL; C-E. osteoclasts' differentiation in conditioned medium derived from tumor cells: C. MDA-MB-435; D: MDA-MB-435 KD LMW-PTP; E: MDA-MB-435 KD LMW-PTP slow. F: TRAP quantification in cells' supernatants. Data are mean±standard deviation from 3 independent experiments. *p<0.05 compared with differentiation with conditioned medium from MDA-MB-435.
LMW-PTP knockdown decreases IL-8 secretion but not IL-6 by MDA-MB-435. Secretion of IL-6 and IL-8 by tumor cells in bone is crucial for the maintenance of the vicious cycle of bone metastasis through osteoclasts' stimulation. Our results showed that LMW-PTP knockdown in MDA-MB-435 decreased the secretion of IL-8 (Figure 3) but not IL-6 (Figure 4).
Discussion
There are three main players in the bone vicious cycle: primary tumor cells, metastatic cancer cells and bone osteoclasts. In this work, our aim was to address how LMW-PTP isoforms influence this process.
Our results show that, in the absence of RANKL, factors released by MDA-MB-435 do not stimulate differentiation of RAW 264.7 into osteoclasts. However, the KD LMW-PTP slow in MDA-MB-435 decreased the osteoclastogenic potential of RAW 264.7 even in the presence of RANKL. LMW-PTP has been associated with the control of Src activity and Src also controls LMW-PTP activity (8, 9, 20, 21). Therefore, in order to further explore how this phosphatase can be involved in the regulation of osteoclastogenesis, we evaluated Src activity and the release of IL-6 and IL-8 by MDA-MB-435, since these cytokines are secreted by breast cancer cells in bone and are known to modulate osteoclastic activity (16).
Phosphorylation of Src. Total Src was used as loading control.
Src activity is controlled by two tyrosine residues, Y527 and Y416. Src is activated when Y527 is non-phosphorylated and Y416 is phosphorylated (21). Our results show that in KD LMW-PTP slow, Src is less active than in the MDA-MB-435 parental cell line. Src has an important role in physiological and pathological processes, such as cell proliferation and tumorigenesis (22). Recently, Ferreira et al. (20) showed that knockdown of LMW-PTP reverts chemoresistence in a chemoresistant cell line, Lucena-1, due to a decrease in Src activation. This is in accordance with our results, since Src is less active in the KD LMW-PTP slow. The relation between LMW-PTP and Src was also demonstrated by Fernandes et al. (9) suggesting the formation of a supra-molecular interaction involving focal adhesion kinase (FAK)/Src/LMW-PTP that could have an important role in osteoblast adhesion.
A mechanism by which Src inactivation is important for tumor cell invasion may be the interaction between this protein and metalloproteinases (MMPs) (21). It is known that, in the vicious cycle of bone metastases, MMPs are important proteins in the cross-talk between tumor and bone cells (5). We may speculate that, by regulating Src, LMW-PTP can have the ability to interact with MMPs, thus playing a role in bone metastases development.
KD LMW-PTP and KD LMW-PTP slow decreased IL-8 secretion in the MDA-MB-435 cell line. *p<0.05 compared toMDA-MB-435.
KD LMW-PTP and KD LMW-PTP slow did not change IL-6 secretion in MDA-MB-435.
Regarding IL-8 results, Bendre et al. described IL-8 as a potent direct activator of osteoclastic differentiation independently of RANKL, involving the IL-8 receptor (CXCR1) on the surface of osteoclasts and their precursors (17, 23). Our results show that LMW-PTP KDs secrete lower levels of IL-8 than the MDA-MB-435 parental cell line. Therefore, LMW-PTP could interfere with the production of IL-8 and, thus, decrease osteoclastogenesis in RAW 264.7 cells exposed to conditioned medium from MDA-MB-435 with the LMW-PTP slow isoform knocked-down. The fact that only the KD LMW-PTP slow decreased osteoclastogenesis suggests that: (i) the slow isoform has the ability to change the secretion of IL-8 but (ii) other factors involved in osteoclasts' differentiation are being changed by the fast isoform, thereby explaining that, although the secretion of IL-8 is decreased in the two knock-downs, only the conditioned medium from KD LMW-PTP slow in MDA-MB-435 decreased the differentiation of osteoclasts. Furthermore, our results suggest that the role on LMW-PTP in the vicious cycle of bone metastasis is independent of osteoblasts, since these cells are not present in our experimental system.
IL-6 is less expressed in MDA-MB-435 than IL-8 (24) and some authors have reported that MDA-MB-435 does not produce IL-6 (25). Our results show that, although the concentration of IL-6 in the supernatant of MDA-MB-435 cells is much lower than IL-8, this cytokine is secreted by this breast cancer cell line, though independently of LMW-PTP expression. Therefore, we conclude that alterations in osteoclastogenesis caused by the KD LMW-PTP slow in MDA-MB-435 are independent of IL-6 secretion.
In a recent study, we proposed that, in breast cancer cell lines, the LMW-PTP slow isoform may have an oncogenic role and the fast isoform an opposite, anti-oncogenic role (13). However, the role of the two isoforms in tumor progression has not been addressed. Based on our previous findings associating LMW-PTP with tumor cells migration (19), both LMW-PTP isoforms seem to be involved in tumor progression. Taken together, these results may indicate that the slow isoform not only is involved in tumorigenesis (13) and tumor cells migration (19) but also increases the affinity of MDA-MB-435 towards bone, which suggests that the slow isoform may be responsible for the increased interaction between breast cancer cells and osteoclasts. Given that there are no studies regarding LMW-PTP isoforms and tumor progression in bone, further studies are required to address this question in vivo.
Conclusion
Taken together, our results show that the expression of the LMW-PTP slow isoform in tumor cells may be a prognostic marker for tumors in bone; increased expression of the LMW-PTP slow isoform may be a marker for a high propensity for tumors to metastasize to bone. Further research on the interaction between LMW-PTP, Src, IL-8 and MMPs could provide new approaches for developing novel drugs for bone metastatic disease.
Acknowledgements
The Authors wish to thank Dr. Luis Moita from Instituto de Medicina Molecular, Faculdade de Medicina de Lisboa, Portugal, for his kind gift of lentiviruses vectors with siRNA sequences targeting LMW-PTP and Dr. Michael Rogers, University of Aberdden, Scotland from his kind gift of RAW 264.7 cell line. The assistance of Sandra Casimiro for the Src Western Blot, Maria Bettencourt and Ricardo Pires for technical support is greatly appreciated. Irina Alho is supported by Fundação para a Ciência e Tecnologia, FCT grant: SFRH/BD/44716/2008.
- Received February 22, 2016.
- Revision received March 31, 2016.
- Accepted April 1, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
