Modulatory effects of heparin and short-length oligosaccharides of heparin on the metastasis and growth of LMD MDA-MB 231 breast cancer cells in vivo

Abstract

Expression of the chemokine receptor CXCR4 allows breast cancer cells to migrate towards specific metastatic target sites which constitutively express CXCL12. In this study, we determined whether this interaction could be disrupted using short-chain length heparin oligosaccharides. Radioligand competition binding assays were performed using a range of heparin oligosaccharides to compete with polymeric heparin or heparan sulphate binding to I(125) CXCL12. Heparin dodecasaccharides were found to be the minimal chain length required to efficiently bind CXCL12 (71% inhibition; P<0.001). These oligosaccharides also significantly inhibited CXCL12-induced migration of CXCR4-expressing LMD MDA-MB 231 breast cancer cells. In addition, heparin dodecasaccharides were found to have less anticoagulant activity than either a smaller quantity of polymeric heparin or a similar amount of the low molecular weight heparin pharmaceutical product, Tinzaparin. When given subcutaneously in a SCID mouse model of human breast cancer, heparin dodecasaccharides had no effect on the number of lung metastases, but did however inhibit (P<0.05) tumour growth (lesion area) compared to control groups. In contrast, polymeric heparin significantly inhibited both the number (P<0.001) and area of metastases, suggesting a differing mechanism for the action of polymeric and heparin-derived oligosaccharides in the inhibition of tumour growth and metastases.

Figure 1

The shortest length oligosaccharide of heparin capable of significantly competing ¹²⁵I-CXCL12 from solid-phase heparin or heparan sulphate (50 μg ml⁻¹) was determined using radioligand competition binding assays. (A) Initially a range of soluble oligosaccharides between 26 and 8 monosaccharide units in length and heparin (7 μg ml⁻¹), were assessed with solid-phase heparin (50 μg ml⁻¹). The oligosaccharides dp14 (filled), dp12 (open), dp10 (hashed) and dp8 (dots) were then (B) used at a range of concentrations to compete CXCL12 from heparin. Finally (C), the dp14, dp12 and dp10 oligosaccharides were competed against a heparan sulphate solid-phase for CXCL12 binding at a range of concentrations (^*P<0.001). Data are representative of three individual experiments; bars show mean values±s.e.m.

Figure 2

The effects of dp12 on the migration of breast cancer cells. The number of LMD MDA-MB 231 cells migrating through 8 μM pores towards 12.5 nM CXCL12 was measured in the presence of varying concentrations of the dp12 inhibitor (^*P<0.001). The results are expressed as a chemotactic index normalised to the number of migrating cells in the presence of CXCL12 with no dp12. Data are representative of three individual experiments; bars show mean values±s.e.m.

Figure 3

The anticoagulation properties of dp12. (A) Anti-Xa levels in plasma isolated from whole blood receiving doses of heparin, tinzaparin and an equivalent dose (weight) of dp12; therapeutic levels are between 0.5–1.0 anti-Xa IU ml⁻¹. (B) Anti-Xa levels in whole blood containing varying doses of dp12.

Figure 4

Representative haemotoxylin and eosin staining of tumours (red arrows) in animals treated with (A) PBS, (B) heparin and (C) dp12 for 28 days after intravenous administration of LMD MDA-MB 231 cells.

Figure 5

Murine breast cancer model. Following administration of LMD MDA-MB 231 cells, three groups of seven mice were treated for 28 days with subcutaneous doses either of heparin (3.3 mg kg⁻¹ per day) given twice daily, 0.1 ml PBS given once daily, or dp12 (4.0 mg kg⁻¹ per day) also given once daily. A total of 20 sections from each mouse lung were H&E stained and assessed blindly using Leica Q win software for both the number (A) of metastases and the area (B) of each metastasis identified (^*P<0.001); bars show mean values±s.e.m.

Figure 6

CXCL12 and CXCR4 expression in mouse lung and tumours. CXCL12 expression on the surface of the vascular endothelium was assessed by immunohistochemical staining using a CXCL12 mouse monoclonal antibody (A) in mouse lung treated with PBS, heparin and dp12 for 28 days following administration of LMD MDA-MB 231 cells (red arrows); negative section is isotype control. There was strong CXCL12 staining on the vascular endothelium in the PBS-treated group, which was unaltered in the dp12 group, whereas CXCL12 was absent from the vascular endothelium of the heparin-treated mice. The levels of human CXCR4 expression (B), relative to PBS group, in tumours present in the mouse lung treated with heparin and dp12 for 28 days following administration of LMD MDA-MB 231 cells were assessed using real-time PCR normalised to human GAPDH. Data are representative of three individual experiments; bars show mean values±s.e.m.