Myosin-1C augments endothelial secretion of von Willebrand factor by linking contractile actomyosin machinery to the plasma membrane

Abstract

Blood endothelial cells control the hemostatic and inflammatory response by secreting von Willebrand factor (VWF) and P-selectin from storage organelles called Weibel-Palade bodies (WPBs). Actin-associated motor proteins regulate this secretory pathway at multiple points. Before fusion, myosin Va forms a complex that anchors WPBs to peripheral actin structures, allowing for the maturation of content. After fusion, an actomyosin ring/coat is recruited and compresses the WPB to forcibly expel the largest VWF multimers. Here, we provide, to our knowledge, the first evidence for the involvement of class I myosins during regulated VWF secretion. We show that the unconventional myosin-1C (Myo1c) is recruited after fusion via its pleckstrin homology domain in an actin-independent process. This provides a link between the actin ring and phosphatidylinositol 4,5-bisphosphate (PIP2) at the membrane of the fused organelle and is necessary to ensure maximal VWF secretion. This is an active process requiring Myo1c ATPase activity because inhibition of class I myosins using the inhibitor pentachloropseudilin or expression of an ATPase-deficient Myo1c rigor mutant perturbs the expulsion of VWF and alters the kinetics of the exocytic actin ring. These data offer a novel insight into the control of an essential physiological process and provide a new way in which it can be regulated.

Graphical abstract

Figure 1.

WPB proximal myosin motors. (A) Volcano plot of myosin isoforms in close proximity to WPBs, previously identified by Rab27a-targeted APEX2 proximity proteomics. Blue represents significantly enriched in unstimulated cells; green, significantly enriched in PMA stimulated cells; magenta, significantly enriched in HAI stimulated cells; and gray, not statistically significant, compared with mock transfected HUVECs. Paired t test. (B) Unstimulated HUVECs were fixed and subject to IF analysis to localize Myo9B (green) in relation to VWF (blue) and actin (magenta). Myo9B staining was present in the cytoplasm and at the end of actin stress fibers reminiscent of focal adhesions. In some cases, VWF localized proximal to Myo9B puncta. Scale bar, 10 μm. Inset 1 μm (C) western blotting of tubulin and Myo9b in HUVEC lysate after 2 rounds of electroporation of 300 pMoles luciferase (LUC) and Myo9B targeting siRNA. Representative blot. (D) VWF secretion in response to PMA, HAI, and thrombin was assessed by NIR dot blot (n = 3). (E) Western blotting of tubulin and Myo1c in HUVEC lysate after 2 rounds of electroporation of 500 pM LUC and Myo1c targeting siRNA. (F-G) LUC and Myo1c KD HUVEC were exposed to PMA (100 ng/mL) or (G) VEGF165 (40 ng/mL), and VWF secretion was quantified by NIR dot blot (n = 3). t test. ∗∗∗P < .005; ∗∗P < .01. (H-I) Western blotting and (I) densitometry of Myo1c, VEGFR2, and GAPDH in LUC and Myo1c KD HUVEC (n = 6). (J-K) HUVECs were treated with the pan class I myosin inhibitor PCLP for 16 hours and endogenous levels of GAPDH and VEGRF2 determined by western blotting. One-way analysis of variance (ANOVA), ∗P < .05 (n = 3). GAPDH, glyceraldehyde 3-phosphate dehydrogenase; KD, knock down; NIR, near infrared.

Figure 2.

ECs use Myo1c as part of the WPB exocytic machinery. (A) Super-resolution imaging and immunofluorescent localization of endogenous Myo1c (green), actin (magenta), and VWF (blue) in unstimulated or PMA (100 ng/mL) stimulated HUVEC in the presence and absence of 1 μM of the actin polymerization inhibitor cytochalasin E (CCE). Scale bar, 10 μm. Inset, 1 μm. Myo1c is recruited independently of actin but was dependent on stimulation with PMA. (B) Myo1c-GFP encapsulates WPB after fusion as determined by live cell super-resolution spinning disk imaging of PMA stimulated (100 ng/mL) HUVEC coexpressing a Myo1c-GFP and the WPB fusion marker P.sel.lum.mCherry. Scale bar, 1 μm. Arrows indicates point of collapse/fusion of vesicle (C) Live cell imaging of LifeAct-GFP and P.sel.lum.mCherry expressing HDMEC indicated the utility of actin rings to expel VWF after stimulation. Scale bar, 10 μm. Inset, 1 μm. Z stacks of 0.5 μm were acquired continuously for 10 minutes (Zeiss LSM 800). (D) Confocal imaging and IF analyses of endogenous Myo1c in HDMEC that were left untreated or stimulated with PMA, CCE, or CCE and PMA. Arrows illustrate where Myo1c is recruited to fused/collapsed WPB. Scale bar, 10 μm.

Figure 3.

The PH domain of Myo1c is required for its recruitment during WPB exocytosis. (A) A schematic of the proposed spatiotemporal dynamics of Myo1c recruitment during WPB exocytosis. (B) Live cell imaging of the GFP-PIPK1γ87 and P.sel.lum.mCherry in secretagogue (HAI) stimulated HUVEC illustrates postfusion recruitment. Two exocytic events are seen here. Scale bars are 1 μm. White and magenta arrows indicate independent fusion events. (C) The PIP2 sensor PH-PLCδ1-GFP is also recruited after fusion. Scale bars are 1 μm. (D) A schematic of the Myo1c structural domains and location of truncation or site directed mutations. (E) HUVEC coexpressing LifeAct-Ruby and Myo1c-Tail+3IQ-GFP indicated the importance of the myosin head domain for interacting with actin. Scale bars are 10 μm. (F) Myo1c-Tail+3IQ-GFP is recruited to WPBs after fusion. (G-H) GFP-tagged Myo1c fusion proteins harboring mutations in their PH domain (K892A/R903A) are not recruited to WPBs during exocytosis. Scale bars are 10 μm (F,G,H). Inset scale bars are 1 μm. For live cell confocal imaging experiments, 0.5 μm Z stacks were acquired continuously for 5 to 10 minutes (Zeiss LSM 800).

Figure 4.

HUVEC exposed to PCLP for 16 hours exhibit a VWF trafficking defect. HUVEC were exposed to dimethyl sulfoxide (DMSO) or a range of concentrations of PCLP and incubated overnight (16 hours). (A) Immunoblotting of the resulting lysates displayed changes in pro-VWF and mature VWF in relation to tubulin. Densitometry indicated a decrease in total VWF (B) and mature-VWF levels (C) alongside a dose-dependent increase in the ratio of pro/mature VWF (D). n = 4 Ratio paired t test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .005 (E) 16-hour incubation with PCLP resulted in inhibition of regulated secretion of VWF in response to thrombin and (F) HAI. (G) HUVEC preincubated with DMSO or PCLP for 16 hours were stimulated with HAI for 10 minutes before application of 5 dyne/cm² shear stress. VWF strings were visualized by immunofluorescence and confocal microscopy. (H-I) The number (H) and length (I) of VWF strings secreted under flow in response to HAI in the presence or absence of DMSO or PCLP (n = 3). (J) IF analyses using anti-LAMP1 (green) and anti-VWF (blue) antibodies indicated numerous VWF positive lysosomes in PCLP-treated cells. (K) IF analyses using anti-TGN46 (yellow) and anti-VWF (blue) antibodies indicated a gross defect in TGN morphology (fragmented and swollen) in PCLP-treated cells. Scale bars, 10 μm. Inset scale bars, 1 μm.

Figure 5.

Acute inhibition of class I myosins and Myo1c depletion perturbs the expulsion of VWF. (A) Schematic of Myo1c domains and mechanism of inhibition by PCLP. (B) Pharmacological inhibition of the ATP binding domain with 10 to 40 μM PCLP reduces VWF release. (n = 6) ∗P < .05; ∗∗P < .01, ratio paired t test. (C) Schematic of live cell imaging approach to study WPB fusion dynamics and VWF expulsion. Scale bar, 1 μm. (D) HUVECs were electroporated with the P.sel.lum.mCherry and GFP-VWF constructs and imaged by confocal microscopy. Preincubation with 20 μM PCLP increased the time taken for VWF to be expelled after loss of the fusion marker (P.sel.lum.mCherry). Paired t test, ∗∗P < .01. (n = 3; DMSO: 9 cells, 63 events; PCLP: 9 cells, 38 events; mean ± standard error of the mean [SEM]). (E) A frequency distribution of events. (F) LUC and Myo1c KD HUVEC were used to test whether these effects were specific to Myo1c or a broader effect of class I myosins. Western blotting determined the efficiency of target protein KD. (G) Myo1c siRNA depletion increased the time taken for VWF to be expelled after loss of P.sel.lum.mCherry. ∗P < .05; paired t test.

Figure 6.

Inhibition of Myo1 ATPase activity through pharmacological inhibition or point mutation (G108R) affects the actomyosin machinery associated with exocytosis. (A) Schematic of live cell imaging approach to study actin dynamics during WPB exocytosis. Scale bar, 1 μm. (B) PMA stimulated HUVEC coexpressing LifeAct-GFP and P.sel.lum.mCherry in the presence or absence of PCLP. The percentage of WPB fusion events that recruited an actin ring were unchanged in DMSO and PCLP (20 μM) treated cells. (C) The lifetime (seconds) of LifeAct-GFP signal at fusion sites was quantified in DMSO and PCLP-treated HUVEC. The distribution of frequency of events is presented here (n = 5 DMSO: 15 cells, 81 events; PCLP: 18 cells, 93 events; mean ± SEM). (D) Schematic of site-directed mutagenesis for the generation of a Myo1c rigor mutant. (E) HUVEC coexpressing GFP-tagged Myo1c constructs and P.sel.lum.mCherry were stimulated with PMA and the percentage of exocytic events that recruit GFP-Myo1c WT or G108R was quantified (n = 3; WT: 8 cells and 119 events; G108R: 8 cells and 58 events). (F) HUVEC coexpressing GFP-tagged Myo1c constructs and P.sel.lum.mCherry were stimulated with PMA and the duration of GFP signal in a ring shape forming at the site of WPB fusion was quantified. (n = 3; WT: 9 cells and 79 events; PCLP: 9 cells and 42 events). (G) The distribution of frequency closely resembles actin ring dynamics--panel C. For live cell confocal imaging experiments, 0.5 μm Z stacks were acquired continuously for 5 to 10 minutes (Zeiss LSM 800). (H) HUVEC expressing GFP, GFP-Myo1c (WT) or (G108R) were stimulated with PMA (100 ng/mL) for 10 minutes and labeled for external VWF (red) and total VWF (blue). Scale bar, 1 μm. (I) Quantification of the ratio of externalized VWF to total VWF. ∗P < .05, 1-way ANOVA; n = 3 NTC. Arrows indicate swollen intracellular VWF signal in cells expressing the G108R point mutant. NTC, nontransfected control.