Activated platelets induce Weibel-Palade-body secretion and leukocyte rolling in vivo: role of P-selectin

Abstract

The presence of activated platelets and platelet-leukocyte aggregates in the circulation accompanies major surgical procedures and occurs in several chronic diseases. Recent findings that activated platelets contribute to the inflammatory disease atherosclerosis made us address the question whether activated platelets stimulate normal healthy endothelium. Infusion of activated platelets into young mice led to the formation of transient platelet-leukocyte aggregates and resulted in a several-fold systemic increase in leukocyte rolling 2 to 4 hours after infusion. Rolling returned to baseline levels 7 hours after infusion. Infusion of activated P-selectin-/- platelets did not induce leukocyte rolling, indicating that platelet P-selectin was involved in the endothelial activation. The endothelial activation did not require platelet CD40L. Leukocyte rolling was mediated solely by the interaction of endothelial P-selectin and leukocyte P-selectin glycoprotein ligand 1 (PSGL-1). Endothelial P-selectin is stored with von Willebrand factor (VWF) in Weibel-Palade bodies. The release of Weibel-Palade bodies on infusion of activated platelets was indicated by both elevation of plasma VWF levels and by an increase in the in vivo staining of endothelial P-selectin. We conclude that the presence of activated platelets in circulation promotes acute inflammation by stimulating secretion of Weibel-Palade bodies and P-selectin-mediated leukocyte rolling.

Figure 1.

Time course of induction of leukocyte rolling in mesenteric venules after infusion of activated platelets. The number of leukocytes rolling per minute was observed by phase-contrast intravital microscopy. Infusion of activated platelets induced a significant increase in leukocyte rolling compared with uninjected mice after 2 hours and 4 hours. Leukocyte rolling decreased by 7 hours after infusion. ^*P < .01, compared to uninjected mice. Average ± standard error of the mean (SEM) is shown.

Figure 2.

Leukocyte rolling in mesenteric venules 4 hours after platelet infusion. (A) The number of leukocytes rolling per minute was determined by phase-contrast intravital microscopy. Infusion of resting platelets, platelet releasate, or buffer did not induce increased leukocyte rolling as compared with uninjected controls. Infusion of activated platelets induced significantly increased leukocyte rolling compared to infusion of resting platelets (^*P < .001). Average ± SEM is shown. (B) Leukocyte rolling observed by fluorescence intravital microscopy on infusion of activated (left) or resting platelets (right). Rhodamine 6G was infused to label the leukocytes. Arrows point to the vessel wall. Bar = 50 μm.

Figure 3.

Leukocyte rolling in mesenteric venules 4 hours after infusion of CD40L^-/- platelets. The number of leukocytes rolling per minute was determined by phase-contrast intravital microscopy. Infusion of activated CD40L^-/- platelets induced increased leukocyte rolling compared with resting platelets indicating that platelet CD40L is not activating the endothelium (P < .01). Average ± SEM is shown.

Figure 4.

Leukocyte rolling in mesenteric venules 4 hours after platelet infusion depends on platelet and endothelial P-selectin. The number of leukocytes rolling per minute was observed by phase-contrast intravital microscopy. In WT recipients, leukocyte rolling after infusion of P-selectin^-/-–activated platelets was significantly less compared to infusion of WT-activated platelets (P < .001). Infusion of WT-activated platelets in either P-selectin^-/- or PSGL-1^-/- (▪) recipients induced significantly less leukocyte rolling compared to infusion in WT recipients (P < .001). Average ± SEM is shown.

Figure 5.

Weibel-Palade–body secretion induced by infusion of activated platelets. (A) In vivo P-selectin expression in mesenteric venules. Fluorescent microspheres (1 μm) coupled to anti–P-selectin antibody were infused in mice given injections with activated or resting platelets 4 hours prior to intravital microscopy. Arrows point to the venule wall. Many microspheres were observed binding to mesenteric venules of a mouse preinfused with activated platelets (left) compared to venules of a mouse given an infusion of resting platelets (right). No microspheres decorated the arterioles (⋆). Bar = 100 μm. Representative images are shown. (B) Plasma VWF levels after infusion of VWF^-/- platelets. Plasma VWF was determined using ELISA 4 hours after infusion of activated or resting VWF^-/- platelets in WT mice. VWF level of a pool of plasma of WT mice was defined as 1 U/mL. There was a significant increase in circulating VWF after infusion of activated platelets as compared to resting platelets (P < .02). Average ± SEM is shown.