Platelet Derived Microparticles (PMP) Interfere with Rolling, Adhesion and Transmigration of Hematopoietic Progenitor Cells (HPC).

We wished to investigate if small particles derived from platelet membranes (PMP) which are present in blood may influence the migration behavior of HPC. PMP were isolated from pooled platelet concentrates derived from human whole blood donations via the buffy coat or PRP methods. After activation of platelets with thrombin or calcium ionophore, PMP were quantitatively enriched by sequential centrifugation at 1.250 and 50.000g. Flow cytometric analysis showed separated forward and side scatter profiles of the larger platelets and smaller PMP, with both platelets and PMP staining highly positive for CD41 and CD61. Purity of PMP was 89–92%. Binding of PMP to HPC was performed at 37°C in cell culture medium at ratios between 3 and 100 PMP/HPC. Binding was assessed flow cytometrically using PMP which were fluorescence-marked with PKH26, or by appearance of CD41/CD61 complex on the surface of HPC. We found that PMP dose-dependently bound human KG-1 or primary CD34⁺ HPC, with half-saturation occurring within seconds to a few minutes, and 10–20 PMPs binding per HPC.

Compared to controls, the binding of PMP-preloaded HPC to immobilized fibronectin showed a mean 10.4-fold reduction with KG-1 cells and a mean 4.2-fold reduction with CD34⁺ primary HPC. The interaction of PMP-pretreated HPC with endothelial cells (EC) was assessed under laminar shear flow in a parallel plate flow chamber at defined wall shear stresses using TNF-alpha prestimulated human umbilical vein endothelial cells (HUVEC). We found a 1.6-fold (KG-1) and 2.2 fold (CD34⁺ HPC) reduction in numbers of rolling and subsequently binding cells at shear stresses of 0.1 dyn/cm², and a 1.5-fold (KG-1) or 1.7-fold (CD34⁺ HPC) reduction at 2 dyn/cm² compared to untreated controls. Both binding and rolling of HPC were dependent on P-selectin in this model, as shown by blocking of the interaction of HPC with EC with MoAB AK-4, and the abrogation of this block with PMP-HPC. Finally, SDF-1alpha-induced transmigration of HPC in transwell chemotaxis chambers was reduced > 3-fold after preincubation of HPC with PMP when more than 10–20 PMP were bound per HPC. Taken together, highly purified PMP are able to bind HPC and to modulate their interaction with endothelium. This process is mediated by (i) suppression of P-selectin mediated rolling and adhesion of HPC to endothelium, (ii) suppression of integrin-mediated adhesion of HPC and (iii) suppression of SDF-1alpha mediated transmigration of HPC. PMPs may therefore be part of a physiological mechanism which can modulate the adhesion behaviour of circulating HPC and protect them from recruitment into sites of inflammation or activated coagulation under conditions of postcapillary or venous shear stress.