Monocyte chemotactic protein 1 promotes lung cancer-induced bone resorptive lesions in vivo

Abstract

Lung cancer is the leading cause of cancer-related deaths. The morbidity and mortality of lung cancer have markedly increased in the past decade with at least 75% of patients with lung cancer having evidence of metastases at the time of diagnosis. It frequently metastasizes to bone resulting in osteolytic lesions with unknown mechanisms. The aim of this study was to identify factors that mediate lung cancer-induced osteoclast activity in vivo. Using a human cytokine antibody array, we first determined cytokine levels in a conditioned medium collected from non-small cell lung cancer A549 and H1299 cells and the non-neoplastic human bronchial epithelial BEAS2B cells. Both A549 and H1229 cells produced significantly higher amount of several cytokines including monocyte chemotactic protein 1 (MCP-1) and interleukin 8 (IL-8) compared with BEAS2B cells. These findings were confirmed by ELISA. From clinical serum specimens, we also observed that MCP-1 and IL-8 levels were increased in lung cancer patients with bone metastases compared with the patients with localized tumor. Next, we investigated the effects of MCP-1 on osteoclast formation in vitro using murine bone marrow-derived monocytes. A549 conditioned medium induced osteoclast formation that was inhibited by neutralizing antibodies against MCP-1. Finally, A549 cells were stably transfected with MCP-1 short hairpin RNA. The MCP-1 knockdown A549 cells were implanted into the tibia of severe combined immunodeficient mice for 4 weeks. The MCP-1 knockdown significantly diminished A549 cell growth. We conclude that MCP-1 promotes lung cancer-induced osteoclast activity and thus bone resorptive lesions in vivo.

Figure 1

Lung cancer cells produced high amounts of MCP-1 and IL-8. The cytokines' production in CM collected from lung cancer cells was measured using a human cytokine antibody array. (A) Images of the array that were assayed with CM obtained from control media, BEAS2B, H1299, and A549 cells. Details of the procedure are described in Materials and Methods. The signals of MCP-1, RANTES, ENA-78, GRO, GROα, IL-8, VEGF, CXCL16, MMP-9, and TIMP-4 were labeled with rectangles. (B) Radiographs of the arrays were scanned to determine the density of each cytokine. Data are presented as a mean value of fold increase relative to control media. The value for each scan was adjusted based on the intensity of positive control spots on each membrane.

Figure 2

Confirmation of cytokine array data by ELISA. ELISA assays were performed with the CM used for the cytokine antibody array. Data are presented as mean ± SD from triplicates. *P < .001 compared with the control media. ^†P < .001 compared with BEAS2B.

Figure 3

Box plot of serum MCP-1 and IL-8 levels in lung cancer patients and healthy donors. Serum levels of MCP-1 and IL-8 in lung cancer patients with localized cancer (n = 18) and bone metastases (n = 7) and in healthy donors (n = 20) were measured by ELISA. The minimum detectable level of MCP-1 is typically less than 5.0 pg/ml. Statistical significance was determined using Student's t test. *P < .001 compared with healthy donors. ^†P < .001 compared with the patients with localized cancer. ^‡P < .01 compared with the patients with localized cancer. Horizontal lines in the box represent the first, second (the median), and third quartiles; whiskers (vertical lines) represent extension of data up and down.

Figure 4

Neutralizing antibodies for MCP-1 inhibited lung cancer A549 CM-induced osteoclast formation. Nonadherent MBMC was cultured at 1 x 10⁵ per well in 96-well plates for 7 to 10 days. The cells were incubated with recombinant mouse M-CSF (10 ng/ml) and/or RANKL (50 ng/ml) and 10% CM collected from the BEAS2B, H1299, or A549 cells in the presence of various doses of neutralizing antibodies against MCP-1 (1 µg/ml), IL-8 (200 ng/ml), and RANKL (1 µg/ml). (A) Representative micrographs of MBMC cell cultures stained for TRAP. (B) The number of osteoclast-like multinucleated cells per well was quantified. Samples were evaluated in quadruplicate. Results are reported as mean (±SD). Data were analyzed using one-way analysis of variance. *P < .001 compared with BEAS2B CM-treated group. **P < .01 compared with A549 CM-treated group. ^†P < .001 compared with A549 CM-treated group.

Figure 5

Knockdown MCP-1 in A549 cells diminished the tumor growth in bone. (A) ELISA was performed in the CM from the selected A549 cells that were stably transfected with shRNA targeting of either MCP-1 or a scrambled control. *P < .001 compared with wild type (WT) or control shRNA cells. (B) MCP-1 knockdown in A549 cells minimally decreased the tumor cell proliferation as determined by MTS assay. *P < .01, significant difference from the control cells by t test. (C) Single-cell suspension of the selected A549 cells was injected into the right tibia of SCID mice. The tumor cells were allowed to grow for 4 weeks, at which time the mice were killed. Evidence of tumor-induced bone osteolysis was evaluated by x-ray, H&E staining, and TRAP staining. On the x-ray film, arrows point to the tumor-induced osteolysis. On TRAP staining, T represents tumor cells, B represents bone, OC represents osteoclast, and arrows point to the increased osteoclast activities. Osteoclast number per millimeter bone surface is determined by bone histomorphometry. Tumor volume versus non-bone soft tissue volume was quantified by bone histomorphometry. Results are reported as mean ± SD. *P < .001 compared with the control cells.