Bone marrow-derived mesenchymal stem cells promote invasiveness and transendothelial migration of osteosarcoma cells via a mesenchymal to amoeboid transition

Abstract

There is growing evidence to suggest that bone marrow-derived mesenchymal stem cells (BM-MSCs) are key players in tumour stroma. Here, we investigated the cross-talk between BM-MSCs and osteosarcoma (OS) cells. We revealed a strong tropism of BM-MSCs towards these tumour cells and identified monocyte chemoattractant protein (MCP)-1, growth-regulated oncogene (GRO)-α and transforming growth factor (TGF)-β1 as pivotal factors for BM-MSC chemotaxis. Once in contact with OS cells, BM-MSCs trans-differentiate into cancer-associated fibroblasts, further increasing MCP-1, GRO-α, interleukin (IL)-6 and IL-8 levels in the tumour microenvironment. These cytokines promote mesenchymal to amoeboid transition (MAT), driven by activation of the small GTPase RhoA, in OS cells, as illustrated by the in vitro assay and live imaging. The outcome is a significant increase of aggressiveness in OS cells in terms of motility, invasiveness and transendothelial migration. In keeping with their enhanced transendothelial migration abilities, OS cells stimulated by BM-MSCs also sustain migration, invasion and formation of the in vitro capillary network of endothelial cells. Thus, BM-MSC recruitment to the OS site and the consequent cytokine-induced MAT are crucial events in OS malignancy.

Keywords: bone marrow-derived mesenchymal stem cells; cytokines; osteosarcoma; transendothelial migration; tumour plasticity.

Figure 1

BM‐MSCs migrate toward CM from OS cells and affect the invasive behaviour of tumour cells. (A) BM‐MSCs were allowed to migrate overnight toward CM from OS cells (SaOS‐2, MG‐63 and HOS). Starvation medium (St Med) was used control. Results are expressed as mean ± SEM of five biological replicates; *** P < 0.001 vs. St Med. (B) BM‐MSCs starved for 24 h in the presence or absence of neutralizing antibodies against CXCR4 (20 μg·mL⁻¹) were allowed to migrate toward CM from OS cells. Results are expressed as mean ± SEM of three biological replicates. (C) ELISA of cytokines and growth factors in CM derived from MG‐63 starved for 48 h (mean ± SD, n = 3 technical replicates). (D) BM‐MSC migration toward CM from MG‐63 with neutralizing antibodies against MCP‐1 (5 μg·mL⁻¹, α‐MCP‐1), blocking of GRO‐α receptor (200 nm, SB 225002) and TGF‐β1 receptor (100 μg·mL⁻¹, TβR blk). Results are presented as mean ± SEM of three biological replicates; *** P < 0.001 vs. CM MG‐63. (E) OS cells were mantained for 48 h in St Med or CM obtained from BM‐MSCs starved for 48 h. Cells were then allowed to invade or (F) transmigrate toward complete medium (FBS 10%). Results are presented as mean ± SEM of three biological replicates; * P < 0.05; ** P < 0.01; *** P < 0.005 vs. St Med.

Figure 2

CM derived from OS cells stimulates the BM‐MSCs trans‐differentiation into CAF‐like cells. (A) FACS analysis of NG‐2 and CD31 expression in BM‐MSCs treated for 48 h with CM OS cells. (B) BM‐MSCs were stimulated for 48 h with CM from OS cells and α‐SMA and Col I‐α1 expression was assessed by immunoblot analysis. Results are representative of four biological replicates. (C) Collagen contraction assay of BM‐MSCs treated for 24 h with St Med or HOS CM. Data are expressed as percentages of the relative area of collagen disc following contraction in comparison with an empty well (mean ± SD, n = 3 biological replicates performed in duplicate). * P < 0.05 vs. St Med. (D) Migration assay of HOS cells stimulated for 48 h with CM derived from BM‐MSCs previously activated or not activated by tumour cells (CM BM‐MSCs ^{St/ HOS}). Cells were allowed to migrate toward complete medium (FBS 10%). Untreated cells (St Med) were used as control. Results are the mean ± SEM of three biological replicates. *** P < 0.001 vs. St Med. (E) Invasion and transmigration (F) assays of HOS cells treated as in (D) (mean ± SD, n = 3 independent experiments). ** P < 0.005 vs. St Med; *** P < 0.001 vs. St Med.

Figure 3

Cross‐talk between BM‐MSCs and OS cells promotes the acquisition of an amoeboid‐like motility in cancer cells. (A) Gelatin zymography of CM obtained from SaOS‐2, MG‐63 and HOS cells stimulated or not stimulated for 48 h with CM derived from BM‐MSCs activated by each OS cell line (CM BM‐MSCs ^OS). The white line indicates the junction of two different gels. Image is representative of three independent experiments. (B) Representative images of pull‐down assay of Rac1 and RhoA GTPases (left panel) and related quantification (right panel). The assay was performed on OS cells grown for 48 h in CM from tumour‐activated BM‐MSCs (CM BM‐MSCs OS) or in starvation medium. Rac1‐GTP and RhoA‐GTP expression was normalized with respect to total Rac1 and RhoA in OS lysates. ^### P < 0.001, Rac1‐GTP CM BM‐MSCs vs. Rac1‐GTP St Med. * P < 0.05, RhoA‐GTP CM BM‐MSCs vs. RhoA‐GTP St Med. Results are presented as mean ± SD of three biological replicates. (C) Confocal analysis of F‐actin (FITC phalloidin) and P‐MLC staining in MG‐63 cells treated or not treated with CM BM‐MSCs ^{MG −63} for 48 h. Scale bar: 10 μm. The images are representative of three biological replicates with similar results. (D) Live imaging of MG‐63 cell migration in three‐dimensional collagen lattice. CFSE‐loaded MG‐63 cells were incorporated into the collagen matrix and monitored by confocal fluorescence‐reflection video microscopy. Tumour cells are visualized in green and the back‐scatter signal of the collagen I is shown in white. On the left, arrows indicate the point at which MG‐63, treated with ST medium, shows an elongated morphology. In MG‐63 cells treated for 48 h with CM from tumour‐activated BM‐MSCs (right), arrowheads indicate the rounded shape of the cells squeezing across collagen I fibres. Scale bar: 20 μm.

Figure 4

BM‐MSCs secrete a specific pattern of cytokines and growth factors. (A) CM from BM‐MSCs grown in St Med or HOS CM was collected and analysed with Human Cytokine Antibody Array according to manufacturer's protocol. C5 (+) positive controls; (−) negative controls; (1) GRO; (2) IL‐6; (3) IL‐8; (4) MCP‐1; (5) OPG; (6) TIMP‐1; (7) TIMP‐2. (B) Bar graph reporting the spot density quantified on each membrane with imagej. Expression values > 10 000 arbitrary units were set to identify a threshold for cytokines to be considered.

Figure 5

The pro‐tumourigenic activities of BM‐MSCs depend on GRO‐α, IL‐6, IL‐8 and MCP‐1 secretion. (A) Invasion assay of HOS cells treated with CM BM‐MSCs ^HOS supplemented or not supplemented with neutralizing antibodies against IL‐6 (5 μg·mL⁻¹), IL‐8 (10 μg·mL⁻¹) and MCP‐1 (10 μg·mL⁻¹) and GRO‐α receptor inhibitor (SB 225002, 200 nm). Results are presented as mean ± SEM of three biological replicates. ** P < 0.005 CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + α‐IL‐8 and CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + α‐MCP‐1; *** P < 0.001, CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + α‐IL‐6; ^§§§ P < 0.001, CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + SB 225002. (B) Transendothelial migration assay of HOS cells treated as in (A) (mean ± SEM, n = 3 biological replicates). * P < 0.05, CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + α‐IL‐6; ** P < 0.005, CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + α‐IL‐8; *** P < 0.001, CM BM‐MSCs ^HOS vs. CM BM‐MSCs ^HOS + α‐MCP‐1.

Figure 6

Cross‐talk between BM‐MSCs and OS cells stimulates the recruitment and activation of endothelial cells. (A,B) Quantitative RT‐PCR of VEGF and IL‐8 mRNA expression in OS cells maintained in St Med or CM BM‐MSCs ^OS for 48 h. Results are presented as mean ± SD of three biological replicates. * P < 0.05 vs. St Med; *** P < 0.001 vs. St Med. (C) Semi‐quantitative detection of pro‐angiogenic factors in CM derived from HOS cells stimulated with CM from tumour‐activated BM‐MSCs in comparison with CM derived from BM‐MSCs maintained in St Med for 48 h. CM were collected and analysed with Human Cytokine Antibody Array according to manufacturer's protocol reported in Material and Methods. (D) Migration assay of HUVEC cells towards CM derived from HOS ^St or HOS ^{BM ‐ MSC s}. Results are mean ± SEM from three independent experiments. ** P < 0.05 vs. St Med; *** P < 0.001 vs. St Med. (E) Invasion assay of HUVEC cells treated as in (D). Results are expressed as mean ± SEM of three biological replicates.** P < 0.05 CM HOS ^{BM ‐ MSC s} vs. CM HOS ^St; *** P < 0.001 CM HOS^{BM ‐ MSC s} vs. CM HOS ^St. (F) Capillary morphogenesis assay of HUVECs incubated in CM from HOS ^St or HOS ^{BM ‐ MSC}. Cord formation was examined after 6 h at 37 °C with optical microscope. The total number of junctions is presented as mean ± SEM of three randomly chosen fields for each experimental condition of three biological replicates performed in technical duplicate. ** P < 0.05 vs. St Med.