An intimate interplay with platelets is an initial key issue for tumor cells in terms of hematogenous metastasis. Tumor cells activate platelets by different pathways and receive, upon forming a platelet cloak, protection from immune surveillance and support in metastatic niche creation. Therapeutic intervention with this early interaction is promising to antagonize the whole metastatic cascade. Here we aimed to investigate the capability of low molecular weight heparin (LMWH), unfractionated heparin (UFH), and a non-anticoagulant heparin derivative or FXa inhibitor fondaparinux to interfere with platelet activation by tumor cells. Coagulation-dependent and independent pathways of platelet activation by three tumor cell lines, and interference therewith were analyzed by fluorigenic thrombin formation assay, platelet aggregometry, ATP and VEGF release and endothelial tube formation assay. LMWH and UFH were found to repress various routes of platelet activation, reflected by attenuated endothelial tube formation. This confirms the duality of anti-coagulative and anti-adhesive properties of heparin. While non-anticoagulative heparin (RO-heparin) depressed platelets’ ATP and VEGF release by contact inhibition sufficiently, fondaparinux just attenuated tissue factor mediated thrombin generation. Concluding, these data suggest that LMWH as a guideline-based drug for anticoagulative strategies in oncology is promising to provide additional benefit for interference with metastatic activities.

Tumor cell–platelet interactions are regarded as an initial crucial step in hematogenous metastasis. Platelets protect tumor cells from immune surveillance in the blood, mediate vascular arrest, facilitate tumor extravasation, growth, and finally angiogenesis in the metastatic foci. Tumor cells aggregate platelets in the bloodstream by activation of the plasmatic coagulation cascade and by direct contact formation. Antimetastatic activities of unfractionated or low molecular weight heparin (UFH/LMWH) can undoubtedly be related to attenuated platelet activation, but molecular mechanisms and contribution of contact formation vs. coagulation remain to be elucidated. Using a set of non-anticoagulant heparin derivatives varying in size or degree of sulfation as compared with UFH, we provide insight into the relevance of contact formation for platelet activation. Light transmission aggregometry and ATP release assays confirmed that only those heparin derivatives with P-selectin blocking capacities were able to attenuate breast cancer cell-induced platelet activation, while pentasaccharide fondaparinux was without effects. Furthermore, a role of P-selectin in platelet activation and signaling could be confirmed by proteome profiler arrays detecting platelet kinases. In this study, we demonstrate that heparin blocks tumor cell-induced coagulation. Moreover, we identify platelet P-selectin, which obviously acts as molecular switch and controls aggregation and secretion of procoagulant platelets.

Metastasis is responsible for the majority of cancer associated fatalities. Tumor cells leaving the primary tumor and entering the blood flow immediately interact with platelets. Activated platelets contribute in different ways to cancer cell survival and proliferation, e.g. in formation of the early metastatic niche by release of different growth factors and chemokines. Here we show that a direct interaction between platelets and MV3 melanoma or MCF7 breast cancer cells induces platelet activation and a VEGF release in citrated plasma that cannot be further elevated by the coagulation cascade and generated thrombin. In contrast, the release of platelet-derived chemokines CXCL5 and CXCL7 depends on both, a thrombin-mediated platelet activation and a direct interaction between tumor cells and platelets. Preincubation of platelets with therapeutic concentrations of unfractionated heparin reduces the tumor cell initiated VEGF release from platelets. In contrast, tumor cell induced CXCL5 and CXCL7 release from platelets was not impacted by heparin pretreatment in citrated plasma. In defibrinated, recalcified plasma, on the contrary, heparin is able to reduce CXCL5 and CXCL7 release from platelets by thrombin inhibition. Our data indicate that different chemokines and growth factors in diverse platelet granules are released in tightly regulated processes by various trigger mechanisms. We show for the first time that heparin is able to reduce the mediator release induced by different tumor cells both in a contact and coagulation dependent manner.

No abstract

Heparanase (HPSE) is an endo-β-D-glucuronidase that cleaves heparan sulphate (HS) chains of proteoglycans (HSPGs). Besides a remodelling of the extracellular matrix, HPSE increases the bioavailability of pro-angiogenic mediators, such as HS-associated vascular endothelial growth factor (VEGF), thereby contributing to metastatic niche formation. Notably, HPSE also induces release of VEGF from tumour cells independent of its enzymatic activity, but the underlying molecular mechanisms remain unresolved. We found that exogenous addition of latent HPSE stimulates VEGF release from human MV3 melanoma cells. The same effect was noted upon direct stimulation of thrombin receptor (protease-activated receptor 1 [PAR-1]) by Thrombin Receptor Activator Peptide 6 (TRAP-6). The matricellular ligand cysteine-rich 61 protein (Cyr61) was identified as pathway component since Cyr61 knockdown in MV3 cells abolished the VEGF release by TRAP-6 and HPSE. Since both TRAP-6 and HPSE mediated an up-regulation of phosphorylated focal adhesion kinase, which could be blocked by antagonizing PAR-1, we postulated a crosstalk between latent HPSE and PAR-1 in promoting pro-angiogenic pathways. To test this hypothesis at a molecular level, we applied dynamic mass redistribution (DMR) technique measuring intracellular mass relocation as consequence of direct receptor activation. Indeed, latent HPSE evoked a concentration-dependent DMR signal in MV3 cells as TRAP-6 did. Both could be modulated by targeting G-protein receptor signalling in general or by the PAR-1 inhibitor RWJ 56110. Using cells devoid of cell surface HS synthesis, we could confirm HPSE effects on PAR-1, independent of HSPG involvement. These data indicate, for the first time, a crosstalk between latent HPSE, thrombin receptor activation and G-protein signalling in general.