Myosin IIa is required for cAMP-mediated VWF secretion from endothelial cells both in vitro and in vivo. (A) Immunostaining of myosin IIa (cyan) and f-actin (magenta) in HUVECs (scale bar, 5 μm). (B) The western blot of the knockdown efficiency of the shRNAs against myosin IIa (shMyoIIa-1 and shMyoIIa-2). The p85 protein was used as a loading control. (C) VWF secretion from HUVECs expressing scrambled (SCR) or myosin IIa shRNAs (shMyoIIa-1 and shMyoIIa-2) with stimulation of forskolin or epinephrine at the indicated concentrations (n = 12; *P < .05, **P < .01). (D) Whole-mount staining to check the knockout (KO) efficiency of myosin IIa in the iEMKO mouse. Mouse ears were fixed and stained with anti-myosin IIa (green) and anti-CD31 (red) to label vascular endothelial cells (scale bars, 60 μm). (E) Normalized plasma levels of VWF in WT (n = 12) and iEMKO (n = 12) mice before (NS > 0.05) and after epinephrine (EPI) stimulation (**P < .01). Results were expressed as percentage of the value of pooled plasma from 5 WT mice. (F) Bleeding times in WT (n = 14) and iEMKO (n = 12) mice after epinephrine stimulation (***P < .001). (G) The graph showed the coverage of the area by thrombus in (H) at 20 min. (***P < .001; mean ± standard error of the mean). (H) FeCl₃-induced thrombus formation in mesenteric vessels of WT (n = 10) and iEMKO (n = 8) mice at different points. Thrombus indicated by rhodamine-labeled platelets and fluorescein isothiocyanate (FITC)-conjugated anti-VWF antibody. All error bars represent standard deviation.