Rab10, but not Rab8, is essential for tubular endosome formation. (A,B) Rab8-DKO and Rab10-KO HeLaM cells generated by the CRISPR/Cas9 system. Cell lysates of parental cells (WT), Rab8-DKO cells, Rab8-DKO cells stably expressing Myc-tagged Rab8A (Rab8-DKO+Myc–Rab8A) cells (A), Rab10-KO cells and Rab10-KO cells stably expressing Rab10 (Rab10-KO+Rab10) cells (B) were analyzed by immunoblotting (IB) with the antibodies indicated. (C,D) Tubular endosomes in Rab8-DKO cells (C) and Rab10-KO cells (D). Rab8-DKO cells and Rab8-DKO+Myc–Rab8A cells stably expressing EGFP–Rab10 were immunostained with anti-Rab8 antibody (C). Rab10-KO cells and Rab10-KO+Rab10 cells were immunostained with antibodies against Rab8 and MICAL-L1 (D). The insets are magnified views of the boxed areas. Scale bars: 20 µm. (E) Quantification of the percentage of cells containing at least one EGFP–Rab10 tubule >20 µm in length. (F) Quantification of the percentage of cells containing at least one Rab8 tubule >20 µm in length. (G) Quantification of the percentage of cells containing at least one MICAL-L1-tubule >20 µm in length. The bars represent the mean±s.e.m. of data from three independent experiments (n=3; more than 20 cells were analyzed in each experiment). **P<0.01; N.S., not significant (unpaired two-tailed Student's t-test). (H) Histogram analysis of the length of MICAL-L1-positive tubules in Rab10-KO cells and Rab10-KO+Rab10 cells. Total tubule length (i.e. the sum of the length of all tubules) in individual cells was classified into five categories: <5 µm, 5–20 µm, 20–50 µm, 50–100 µm or >100 µm, and the number of cells containing tubules having the total tubule length in each category was counted as described in the Materials and Methods. The numbers of cells in each category as a percentage of the total cell population are expressed as stacked bar graphs. The bars represent the mean±s.e.m. of data from three independent experiments (n=3; more than 30 cells were analyzed in each experiment). ***P<0.001 (Pearson's χ² test).

Microtubules are necessary for tubular endosome extension. (A) Dynamics of EGFP–Rab10-positive tubules. HeLaM cells stably expressing EGFP–Rab10 were analyzed by live-cell imaging on a spinning-disk confocal microscope. Image stacks of the regions of interest were captured, and maximum intensity projections were obtained. The time-lapse images of the boxed area captured at 4.8 s intervals are shown on the right. Arrowheads point to the fission sites of tubular endosomes. Scale bars: 20 µm (main image) and 2 µm (inset). (B) Effect of DMSO (control), cytochalasin D and nocodazole on EGFP–Rab10-positive tubules. HeLaM cells stably expressing EGFP–Rab10 were treated for 1 h with 0.1% DMSO, 10 µM cytochalasin D or 10 µg/ml nocodazole. The insets are magnified views of the boxed areas. Scale bar: 20 µm. (C) Percentage of cells containing EGFP–Rab10-positive tubules as shown in B. The bars represent the mean±s.e.m. of data from three independent experiments (n=3; more than 20 cells were analyzed in each experiment). *P<0.05; ***P<0.001 (Dunnett's test). (D) Colocalization between EGFP–Rab10 and mStr–EMTB (a microtubule marker). Typical images of HeLaM cells stably co-expressing EGFP–Rab10 and mStr–EMTB were captured with a conventional confocal microscope (upper panels) or OSR (Olympus Super Resolution; lower panels). The super-resolution images correspond to the boxed area in the upper right panel. The insets in the lower panels are magnified views of the boxed areas. Arrowheads point to a microtubule and EGFP–Rab10 double-positive tubule. Scale bar: 20 µm.

KIF13A/B are novel Rab10-interacting proteins. (A) Schematic representation of mouse KIF13A/B. Amino acid (aa) numbers are shown on both sides of each KIF13 protein. The Rab10-binding homology domain (RHD) of KIF13A/B determined in this study is represented by orange lines. (B) Sequence alignment and predicted secondary structure of the RBD or RHD of mouse KIF13A/B, EHBP1 and MICAL1. The amino acid residues that are conserved and that are similar in more than three sequences are shown against a black background and against a shaded background, respectively. The predicted secondary structure is indicated above the corresponding sequences. (C) A 3D-homology model of the KIF13A RHD and the crystal structure of the Rab10–MICAL1 (RBD) complex (PDB 5LPN). The predicted 3D-structure of the KIF13A RHD (orange) is superimposed on the MICAL1 RBD (blue) in complex with Rab10 (silver). The α-helix 1 and α-helix 2 in the KIF13A RHD of the boxed area correspond to the secondary structure shown in B. (D) Schematic representation of KIF13A/B-tail mutants without a motor domain. (E) Interaction between Rab10 and KIF13A/B-tail. mStr–Rab10 and EGFP–KIF13A/B-tail were co-expressed in COS-7 cells, and their associations were analyzed by co-immunoprecipitation assays with glutathione–Sepharose beads coupled with GST–GFP nanobody, followed by immunoblotting with the antibodies indicated.

The RHD of KIF13 is required for its localization at EGFP–Rab10-positive tubules. (A) Localization of Myc–KIF13A (WT and ΔRHD). HeLaM cells stably co-expressing EGFP–Rab10 (green) and Myc–KIF13A (WT or ΔRHD; magenta) were examined. The insets are magnified views of the boxed areas. Scale bars: 20 µm. (B) Line plot profiles of the yellow arrows in the insets in A. (C) Schematic representation of a KIF13A(ΔRHD) mutant that lacks an RHD, shown in orange. (D) PCCs for the relationship between EGFP–Rab10 and Myc–KIF13A (WT or ΔRHD) as shown in A. The bars represent the mean±s.e.m. of data from three independent experiments (n=3; more than five images were analyzed in each experiment). *P<0.05 (unpaired two-tailed Student's t-test). (E) Localization of Myc–KIF13A in Rab10-KO cells. Rab10-KO cells stably expressing Myc–KIF13A (green) were immunostained with the anti-MICAL-L1 antibody. The insets are magnified views of the boxed areas. Scale bars: 20 µm. (F) Percentage of cells containing at least one Myc–KIF13A-positive tubule >20 µm in length as shown in E. The bars represent the mean±s.e.m. of data from three independent experiments (n=3; more than 20 cells were analyzed in each experiment). *P<0.05 (unpaired two-tailed Student's t-test). (G) Unaltered expression of KIF13A/B in Rab10-KO cells. Cell lysates of parental (WT) cells, Rab10-KO cells and Rab10-KO cells stably expressing Rab10 (Rab10-KO+Rab10) were analyzed by immunoblotting (IB) with the antibodies indicated.

Both the motor domain and the RHD of KIF13A are necessary for tubular endosome formation. (A) KIF13A-KO, KIF13B-KO and KIF13-DKO HeLaM cells generated by the CRISPR/Cas9 system. Cell lysates of parental (WT), KIF13A-KO, KIF13B-KO and KIF13-DKO cells were analyzed by immunoblotting with the antibodies indicated. (B) Tubular endosomes in KIF13A-KO, KIF13B-KO and KIF13-DKO cells. Each cell line was immunostained with anti-MICAL-L1 antibody. The insets are magnified views of the boxed areas. Scale bar: 20 µm. (C) Percentage of cells containing at least one MICAL-L1-positive tubule >20 µm in length as shown in B. **P<0.01; ***P<0.001 (Dunnett's test). (D) Rescue of KIF13A-KO cells by exogenous expression of KIF13A mutants. KIF13A-KO cells stably co-expressing EGFP–Rab10 (green) and Myc–KIF13A (WT, ΔRHD or tail) (magenta) were tested. The insets are magnified views of the boxed areas. Scale bars: 20 µm. (E) Similar level of Myc–KIF13A expression (WT, ΔRHD or tail) in KIF13A-KO cells. Cell lysates of parental (WT) cells, KIF13A-KO cells and KIF13A-KO cells stably expressing Myc–KIF13A (WT, ΔRHD or tail) were analyzed by immunoblotting with the antibodies indicated. (F) Percentage of cells containing at least one EGFP–Rab10-positive tubule >20 µm in length as shown in D. The bars represent the mean±s.e.m. of data from three independent experiments (n=3; more than 20 cells were analyzed in each experiment). ***P<0.001 (one-way analysis of variance followed by the Tukey–Kramer test).