Fig 2: Schematic representation of AKTIP structural and functional association with the ESCRT machinery.(A) AKTIP localizes at the midbody in proximity with the annular structure formed by MKLP1 and with that formed by the ESCRT I subunit TSG101 and VPS28. AKTIP super-structure is flanked by the rings formed by the ESCRT III subunits CHMP4B, CHMP2A and IST1. (B) AKTIP is recruited to the midbody in a MKLP1 dependent, CEP55 independent way. AKTIP and TSG101 cooperate in the recruitment of CHMP4B to the midbody through independent routes, likely involving the ESCRT I subunit VPS28, that interacts with both TSG101 and AKTIP. AKTIP and TSG101 are needed for the recruitment of IST1 to the midbody, and act in a common pathway leading to abscission.

Fig 3: KIF5B promotes ALIX, CHMP4B and TSG101 recruitment to the midbody and accurate abscission timing. a Western blot showing efficacy of siRNA (12.5 nM) induced KIF5B knockdown (KD) after 2 days of transfection in comparison to constant expression levels of ALIX and actin in both control (ctrl.) and KIF5B siRNA-treated cells. b Cumulative frequency plot showing the time interval between ICB formation and abscission upon control or KIF5B siRNA treatment as indicated (n ≥ 70 cells per treatment from four independent experiments; control: 73.6 ± 2.0 min; KIF5B KD: 141 ± 6.8 min [mean time 50% of cells completed abscission ± SEM]; P < 0.001). c FRAP analysis of ALIX dynamics at the midbody. Normalized ALIX intensities of control and KIF5B-depleted cells stably expressing ALIX-mCherry are plotted (number of FRAP experiments: ctrl. = 21, KIF5B KD = 16). d Scatter plot showing the time interval between bridge formation and first appearance of ALIX-mCherry at the midbody (MB) in control cells or upon KIF5B siRNA treatment as indicated (n ≥ 40 cells per treatment from four independent experiments; control: 53.3 ± 2.3 min; KIF5B KD: 79.4 ± 4.5 min [± SEM]; P < 0.001). e Scatter plot showing the time interval between bridge formation and first appearance of CHMP4B-GFP at the midbody (MB) (n ≥ 40 cells per treatment from four independent experiments; control: 52.5 ± 2.4 min; KIF5B KD: 82.3 ± 3.4 min [± SEM]; P < 0.001). f Scatter plot showing the time interval between bridge formation and first appearance of GFP-TSG101 at the midbody (n ≥ 40 cells per treatment from four independent experiments; control: 42.7 ± 1.8 min; KIF5B KD: 73.5 ± 3.7 min [± SEM]; P < 0.001). g Scatter dot plots showing the total distance of ALIX-positive vesicles in cytokinetic cells (ctrl. vs. KIF5B KD) during a time period of 7 min (n ≥ 30 cells per treatment from four independent experiments; total number of analyzed vesicles ctrl. ≥ 2800 and KIF5B KD ≥ 4700; ctrl.: 5.96 ± 0.10 µm; KIF5B KD: 5.63 ± 0.08 µm [± SEM]; P = 0.0103). h Scatter dot plots showing the mean speed of ALIX-positive vesicles in cytokinetic cells (ctrl. vs. KIF5B KD; n ≥ 30 cells per treatment from four independent experiments; total number of analyzed vesicles ctrl. ≥ 4300 and KIF5B KD ≥ 5400; ctrl.: 0.54 ± 0.008 µm/sec; KIF5B KD: 0.47 ± 0.006 µm/sec [± SEM]; P < 0.0001). i Scatter dot plots showing the maximum speed of ALIX-positive vesicles in cytokinetic cells (ctrl. vs. KIF5B KD; n ≥ 30 cells per treatment from four independent experiments; total number of analyzed vesicles ctrl. ≥ 1400 and KIF5B KD ≥ 4000; ctrl.: 1.8 ± 0.04 µm/sec; KIF5B KD: 1.4 ± 0.02 µm/sec [± SEM]; P < 0.0001). j KIF5B depletion affects midbody morphology. Cells were fixed and stained for RacGAP1 (green), CEP55 (magenta) and tubulin (grey). Overview images of cytokinetic cells as well as projections of the midbody region (RacGAP1 and CEP55 staining) for fixed control cells (left) or following KIF5B KD (right) are shown. Depletion of KIF5B leads to enlarged and less compact midbody rings. Scale bars = 5 µm. k Percentage of irregularly shaped midbodies upon depletion of ALIX, KIF5B or KLC1 in cytokinetic cells. Fixed cells were stained for tubulin, CEP55 and RacGAP1 and the midbody (MB) morphology was analyzed by SIM from four independent experiments (% of cells showing abnormal MB morphology, ctrl.: 13.13 ± 2.86%, n ≥ 150 cells; ALIX KD: 37.48 ± 3.99%, n ≥ 150 cells; KIF5B KD: 37.99 ± 5.85%, n ≥ 150 cells; KLC1 KD: 29.3 ± 3.48, n ≥ 60 cells). Depletion of ALIX, KIF5B or KLC1 significantly increases the percentage of irregularly shaped midbodies (ctrl. vs. ALIX KD, P = 0.0037; ctrl. vs. KIF5B KD P = 0.0087; ctrl. vs. KLC1 KD, P = 0.015)

Fig 4: Cytokinetic defects upon ALIX knockdown. a Western blot showing knockdown (KD) efficiency of siRNA-induced ALIX depletion after 3 days of transfection. b Cumulative frequency plot showing the time interval between ICB formation and abscission upon control and ALIX siRNA treatment as indicated (n ≥ 60 cells per treatment from three independent experiments; control: 63.8 ± 1.5 min; ALIX KD: 169.2 ± 8.4 min [mean time 50% of cells completed abscission ± SEM]; P < 0.001). c Scatter plot showing the time interval between bridge formation and first appearance of CHMP4B-GFP at the midbody (MB) in live cell imaging analysis upon control (ctrl.) and ALIX siRNA (ALIX KD) treatment as indicated (n ≥ 60 cells per treatment from three independent experiments; control: 59.9 ± 1.4 min; ALIX KD: 137.8 ± 10.6 min [± SEM]; P < 0.001). d Quantification of the length of the ICB from live cell imaging data (n ≥ 100 cells from three independent experiments; control: 6.33 ± 0.1 µm; ALIX KD: 9.5 ± 0.29 µm [± SEM]; P < 0.001). e Altered morphology of the midbody upon ALIX depletion. Cells were fixed and stained for RacGAP1 (green), CEP55 (magenta) and tubulin (grey). Overview images as well as projections of the midbody region (RacGAP1 and CEP55 staining) are shown. Control cells show a compact midbody. In contrast to control cells, ALIX-depleted cells display enlarged midbodies with filamentous extensions. Scale bars = 5 µm