Panel a shows widefield images of merged, ROI, VE-cad, EHD4, and BAR comparing GFP-SNX9 and PACSIN2-GFP, with arrowheads indicating labeled structures. Panel b shows similar images but with MICAL-L1 instead of EHD4. Panels c and d present dot plots of Manders split colocalization coefficients for PACSIN2-GFP or GFP-SNX9 with EHD4 and MICAL-L1 at AAJs. Panel e displays widefield images of PACSIN2 and EHD4 depleted HUVECs immunostained for SNX9 and VE-cadherin. Panels f and g show dot plots indicating the percentage of SNX9-positive AAJs of total AAJs per cell in HUVECs transduced with shControl and shPACSIN2 or shEHD4. Panel h contains confocal images of HUVECs immunostained for VE-cadherin and dynamin-2. Panel i quantifies the percentage of dynamin-2-positive AAJs of total AAJs per cell in HUVECs. Panel j presents a widefield time-lapse sequence depicting dynamin-2-GFP recruitment to internalizing VE-cadherin-mScarlet-positive AAJs. Panel k shows maximum intensity projections of confocal z-stacks of HUVECs immunostained for VE-cadherin, dynamin-2, and either PACSIN2 or SNX9. Panel l displays dot plots showing Manders split colocalization coefficients between junctional PACSIN2 or SNX9 with dynamin-2.
SNX9 associates with dynamin-2 during asymmetric junction remodeling. (a and b) Widefield images of GFP-SNX9 and PACSIN2-GFP (green) expressing HUVECs immunostained for VE-cadherin (magenta) and either EHD4 (a) or MICAL-L1 (b) (yellow). The white squares indicate the ROIs that are magnified in the right panels. Green arrowheads indicate AAJs with BAR proteins and yellow arrowheads indicate AAJs with EHD4 (a) or MICAL-L1 (b). (c and d) Dot plots of Mander’s split colocalization coefficient of PACSIN2-GFP or GFP-SNX9 with EHD4 (c) and MICAL-L1 (d) at AAJs. n = 32 AAJs for PACSIN2/EHD4, 32 for SNX9/EHD4, 42 for PACSIN2/MICAL-L1, and 34 for SNX9/MICAL-L1 from three independent experiments. (e) Widefield images of PACSIN2- and EHD4-depleted HUVECs immunostained for SNX9 (magenta) and VE-cadherin (green). (f and g) Dot plots indicating the percentage of SNX9-positive AAJs of total AAJs per cell in HUVECs transduced with shControl and shPACSIN2 (f) or shControl and shEHD4 (g). n = 32 for shControl, 32 for shPACSIN2 in f and n = 34 for shControl, and 31 for shEHD4 in g from three independent experiments. (h) Confocal images of HUVECs immunostained for VE-cadherin (magenta) and dynamin-2 (green). Green arrowheads indicate dynamin-2–positive AAJs. (i) Quantification of the percentage of dynamin-2–positive AAJs of total AAJs per cell in HUVECs. n = 33 cells from three independent experiments. (j) Widefield time-lapse sequence depicting dynamin-2–GFP recruitment to internalizing VE-cadherin–mScarlet-positive AAJs (See Video 10). Yellow arrowheads indicate AAJs that undergo internalization events. (k) Maximum intensity projection of confocal z-stacks of HUVECs immunostained for VE-cadherin (blue), dynamin-2 (magenta), and either PACSIN2 or SNX9 (green). Magenta arrowheads indicate regions of dynamin-2 recruitment, and green arrowheads indicate regions of PACSIN2/SNX9 recruitment at AAJs. (l) Dot plots showing Mander’s split colocalization coefficient between junctional PACSIN2 or SNX9 and dynamin-2. n = 34 AAJs for PACSIN2, and 33 for SNX9 from three independent experiments. Data are presented as mean ± SD of experiment means. Statistical analysis was performed using a two-sided Student’s t test in c, d, f, and g and a Mann–Whitney U test in l. Circles indicate individual data points, and triangles indicate means of individual experiments with colors corresponding to the experimental replicates. P values are indicated on the graphs. Scale bars = 30, 10, 5, and 3 μm.