Effect of aftiphilin depletion on VWF secretion is specific. (A and B) Rescue of aftiphilin knockdown by mutant aftiphilin cDNA. HEK293 cells were transfected with aftiphilin siRNA and incubated for 3 d. They were then transfected again with the siRNA and a full-length VWF plasmid, with or without an siRNA-resistant aftiphilin construct. VWF secretion assays were carried out 48 h later. (A) Western blotting showing the expression of siRNA-resistant aftiphilin plasmid restored the expression of aftiphilin to 61 ± 6% of mock level (n = 6). Actin was used as a loading control. (B) The aftiphilin depletion phenotype was partially rescued by mild expression of the siRNA-resistant aftiphilin plasmid. Error bars represent SEs; n = 6. (The asterisks, squares and circles show pairs of samples compared using t tests: *p = 0.015, significantly different; □, p = 0.059; ○, p = 0.2; ●, p = 0.2, statistically not different). (C) Effect of aftiphilin depletion is specific to VWF. ssHRP, a constitutive secretion marker, was co-transfected into HEK293 cells in the second round of siRNA transfection and left for 1 d. The ratio of the amount of ssHRP released in 6 h to that remained in the cells was plotted in the graph. The triangles and diamonds show pairs of samples compared using t tests. There was no significant difference between mock cells and those transfected with aftiphilin (▴, p > 0.05) or epsinR siRNA(◊, p > 0.05). Error bars represent SEs; n = 6.

WPBs are made in aftiphilin-depleted cells. (A and B) HEK293 cells were nucleofected with aftiphilin siRNA (B) or mock-treated (A), and after 3 d, were further nucleofected with full-length VWF and siRNA. Fixation and immunolabeling with anti-VWF were carried out 2 d later. Insets, magnified view of the boxed areas. Bar, 10 μm. (C–F) HUVECs were nucleofected with aftiphilin siRNA (E and F) or mock-treated (C and D). After 3 d, the cells were nucleofected again with siRNA and GFP-VWF and incubated for 1 d, followed by fixation and immunolabeling with anti-aftiphilin and anti-GFP. Secondary antibodies used were Alexa Fluor 594-anti-rabbit and Alexa Fluor 488-anti-mouse. Insets, magnified view of the boxed areas. Bar, 10 μm. (G–J) γ-Synergin–depleted cells are capable of making WPBs. HUVECs were nucleofected with γ-synergin siRNA (I and J) or mock-treated (G and H). After 3 d, the cells were microinjected with siRNA and GFP-VWF and incubated for 1 d, followed by methanol/acetone fixation and immunolabeling with anti-γ-synergin (G and I) and anti-GFP (H and J). Insets, magnified view of the boxed areas. Bar, 10 μm. (K–L) Subcellular fractionation shows the efficiency of WPB production in aftiphilin-depleted cells. (K) Mock-treated (blue diamonds) and aftiphilin-depleted (red circles) HEK293 cells, co-transfected with full-length VWF, were homogenized and the post-nuclear supernatants were fractionated using a density gradient of 20–60% sucrose. The dense peak corresponds to the WPB fraction. (L) VWF multimer analysis of fractions from the light and dense peaks on a 1.4% SDS-agarose gel. Highly multimerized VWF is found in the WPB-enriched fractions from both mock and aftiphilin-depleted cells. (M–N) AP-1-depleted cells cannot make WPBs. HUVECs were nucleofected with AP-1 siRNA or mock-treated, followed by a 2-d incubation and a second round of nucleofection. Mock-treated and AP-1-depleted cells were plated onto the same dish. After 2 d of incubation, cells were fixed and analyzed by immunofluorescence microscopy using anti-γ-adaptin (M) and anti-VWF (N). Bar, 10 μm.

Recruitment of P-selectin to WPBs in aftiphilin-depleted cells. HUVECs were transfected with aftiphilin siRNA (E–H) or mock-treated (A–D) as described above, fixed and immunolabeled with rabbit anti-aftiphilin (A and E), mouse anti-P-selectin (B and F), and sheep anti-VWF (C and G). D and H are merged images of A, B and C, and E, F and G, respectively. It was then followed by Alexa Fluor 594 donkey anti-rabbit, Alexa Fluor 488 donkey anti-mouse, and Alexa Fluor 680 donkey anti-sheep. Projected Z-stacks are shown here. Bar, 10 μm.

Recruitment of mature WPB markers Rab27 and CD63 in aftiphilin-depleted cells. (A–H) Rab27 recruitment. HUVECs were nucleofected with aftiphilin siRNA (E–H) or mock-treated (A–D), and after 3 d, were microinjected with GFP-Rab27a and siRNA. The cells were fixed 2 d later and labeled with anti-aftiphilin (B and F; red in A and E), anti-GFP (C and G; green in A and E) and anti-VWF (D and H; blue in A and E). The secondary antibodies used were Texas Red-anti-rabbit, FITC-anti-mouse, and Cy5-anti-sheep. (C, D, G, and H) Magnified view of the boxed areas. Bar, 10 μm. (I–P) CD63 recruitment. HUVECs were nucleofected with aftiphilin siRNA (M–P) or mock-treated (I–L), and incubated for 3 d before another round of nucleofection. After 2 d, the cells were fixed and immunolabeled with anti-aftiphilin (J and N; red in I and M), anti-VWF (K and O; green in I and M) and anti-CD63 (L and P; blue in I and M). The secondary antibodies were Texas red-anti-rabbit, FITC-anti-sheep, and Cy5-anti-mouse. (K, L, O, and P) Magnified view of the boxed areas. Bar, 10 μm.

Multimerization analyses of secreted VWF from mock and aftiphilin-depleted cells. (A) WPBs are released in an unregulated manner from aftiphilin-depleted cells, as revealed by VWF multimer analysis of constitutively released VWF. Constitutively secreted VWF was harvested from HEK293 cells transfected with VWF and siRNA duplexes against aftiphilin or AP-1. Samples were analyzed on 1.4% SDS-agarose gels. The figure is a densitometry trace analysis of multimers, with the data normalized to the total signal. The multimeric state of VWF increases from right to left on the multimer profile graph, with the dimer at the far right peak and increments of dimer units leftwards. Octamers or smaller VWF multimers are defined here as LMW multimers; VWF of higher multimeric states are defined as HMW multimers, which can only be found in WPBs. HMW multimers are released from aftiphilin-depleted cells constitutively. (B) Comparison of the multimeric states of VWF released in the presence and absence of PMA. Secretion assay samples were collected from aftiphilin-depleted and mock-treated HEK293 cells. The VWF multimers were then analyzed as above. Compared with the control (dark blue line), VWF released constitutively from aftiphilin-depleted cells has more HMW multimers (red line), which is a hallmark of secretion from WPBs (light blue line). The two traces of VWF secreted from aftiphilin-depleted cells are very similar (red and orange lines), reflecting the significant decrease in PMA responsiveness as seen in Figure 1; the slightly higher amount of HMW multimers in the presence of PMA echoes our observation in Figure 1C that there is 20% residual responsiveness to PMA in aftiphilin knockdown.

Aftiphilin and γ-synergin affect the ratio of VWF released in the constitutive and regulated pathways. (A) HEK293 cell lysates from mock and knockdown cells were analyzed by Western blotting. Actin was used as the loading control. The bottom panel shows the protein levels of the p200/aftiphilin/γ-synergin complex when one subunit is targeted by siRNA in the HEK293 system. (B) Effect of siRNAs against AP-1 effectors on VWF secretion assays. HEK293 cells were transfected with siRNA and incubated for 3 d, before the co-transfection of full-length VWF and siRNA. VWF secretion assays were carried out when the cells reached 90% confluence. After 30 min of constitutive secretion (left, blue bars), the medium was collected and fresh medium containing the secretagogue PMA was added. The medium was collected 30 min later. To determine the PMA-responsive pool of VWF (right, red bars), the amount released constitutively was subtracted from the PMA releasate. A representative experiment is shown here. (C) Aftiphilin and γ-synergin depletion led to significant increase in constitutive secretion and reduction in regulated secretion. Here, the data were normalized by dividing by the mock data from the same experiments, and pooled from multiple secretion assays (n = 7 for aftiphilin, n = 6 for γ-synergin, and n = 5 for epsinR). The error bars represent SEs (t test results compared with mock: *p < 0.01; ○ = 0.03). (D) Immunofluorescence showing effective knockdown of aftiphilin in siRNA-treated HEK293 cells. Bar, 10 μm. (E–L) Partial colocalization of aftiphilin and VWF in the perinuclear region. HUVECs were fixed with 6% paraformaldehyde, permeabilized with saponin, and immunolabeled with sheep anti-VWF (F and H; red in E) and rabbit anti-aftiphilin (G and I; green in E). The secondary antibodies used were Texas Red-anti-sheep and FITC-anti-rabbit. Bar, 10 μm. (F–I) Magnified view of the boxed areas in E. (J–L) Partial colocalization of γ-synergin and VWF in the perinuclear region. HUVECs were fixed and permeabilized using methanol and acetone, followed by immunolabeling with sheep anti-VWF (K; red in J) and rabbit anti-γ-synergin (L; green in J). The secondary antibodies were as above. Bar, 10 μm. (K and L) Magnified view of the boxed area in J.