Fig 2: Selected endosomal Rab RNAi screen identified Rab22A as a regulator of RE dynamics AIFM images of KIF13A-YFP-transfected control and Rab-knockdown HeLa cells. TN: average tubule number (mean ± SEM, n = 3).B, CGraphs represent the measurement of KIF13A-positive TN (B) and TL (C) in HeLa cells of Fig 1A (mean ± SEM). n = 3. nc: total number of cells. nt: total number of tubules. *P = 0.05, **P = 0.01, ***P = 0.001 and ns = not significant (unpaired Student's t-test).DIFM images of KIF13A-YFP and mCherry-Rab7A/11A/22A-transfected HeLa cells.ELive cell imaging of GFP/mCherry-Rab22A with respect to mCherry-Rab11A or GFP-Rab7A in HeLa cells. Magnified view of insets (at 0, 20, 40 s) are shown separately.FPull-down of different His-KIF13A domains using HeLa cell lysate and then probed with indicated Rab proteins. The bead-bound His-KIF13A in each pull-down was shown on the Coomassie-stained gel. *, non-specific bands. Note, part of this experiment was shown in Fig 5F.Data information: In (A, D, E), arrowheads and arrows point to the KIF13A-/Rab22A-positive tubular REs and E/SEs, respectively. Scale bars: 10 µm.

Fig 3: Effect of 2xFYVE and p40PX expression on the development of the inner pre-AC in MCMV infected cells. Balb/3T3 cells were transfected with EGFP-2xFYVE and YFP-p40PX and MCMV infected (?m138-MCMV, MOI 10) at 42 h p.t. After 6 h of infection (48 h p.t.), the cells were fixed, permeabilized, and stained with mAb against IE1 and either rabbit polyclonal (A) anti-Rab5a, (B) anti-Rab10, (C) anti-Rab11a, or (D) anti-Evectin-2, followed by non-cross reactive AF680-anti-mouse IgG2a and AF594-anti-rabbit IgG. Arrows point to the juxtanuclear accumulation in un-transfected cells. Cell borders are indicated by fine dotted lines. Bars, 10 µm. The 3D colocalization of Rab5a, Rab10, Rab11a, and Evectin-2 with (E) EGFP-2xFYVE or (F) YFP-p40PX in uninfected (orange bars) and MCMV-infected (blue bars) cells, based on M1/M2 coefficients of pixel overlap measured across the Costes-algorithm thresholded z-stacks of confocal images. Data represent the mean ± SEM per cell (n = 8–12). Significance: * = p = 0.05, ** = p = 0.01, *** = p = 0.001.

Fig 4: IN1 treatment prevents recycling cargo sorting into the Rab11a-dependent pathway of the recycling circuit in MCMV infected cells. Cells infected with ?m138-MCMV were treated with 3 µm IN1 or appropriate dilution of the solvent (control) at 4 hpi and exposed 45 min to Tf-AF594 (µg/mL) followed by staining against either Rab5a or Rab11a and IE1 proteins. Representative images of intracellular distribution of internalized Tf-AF594 and Rab5a (A) or Rab11a (B) relative to the nuclear distribution of the IE1 protein. Inserts present zoomed area box. Bars, 10 µm. (C) 3D reconstruction of the Rab11a- and Tf-AF594-labeled TfRs in IN1-treated MCMV infected cells. The right panel presents the 3D view of the entire z-stack of 0.3 µm sections (32 slices) from the boxed area in Figure 4B using the Image J Volume Viewer plugin. The image shown on the right presents the 3D view of the entire stack, and the images on the left give the view across the vertical section through the stack. (D) Quantification of intracellular Tf-AF594 in control (Ø) and IN1-treated cells by ImageJ analysis. Dots represent the Total Corrected Cell Fluorescence (TCCF) of individual cells in the representative experiment (N = 3) and horizontal bars the median value. (E) Quantification of Rab5a and Rab11a intensity in control and IN1-treated cells by ImageJ analysis. Dots represent the Total Corrected Cell Fluorescence (TCCF) of individual cells in a representative experiment (N = 3) and horizontal bars the median value (F). The 3D colocalization of Tf with organelle markers and organelle markers with Tf in control and IN1-treated cells based on M1/M2 coefficients of pixel overlap measured across the Costes-algorithm thresholded z-stacks of confocal images. Data represent the mean ± SEM per cell (n = 8–12). Significance: * = p = 0.01, *** = p = 0.001.

Fig 5: Interaction of a subunit isoforms with small GTP-binding proteins. (a) Interaction of a subunit isoforms and Rab7. FLAG-tagged a isoforms and various V5-fused forms of Rab7 were co-expressed in HEK293T cells. The cells were lysed, and lysates were immunoprecipitated with an anti-FLAG antibody. The precipitates were analysed using antibodies against FLAG (upper panel), the A subunit of the V1 sector (upper middle panel) and V5 (lower middle panel). As a control, cells were co-transfected with an empty vector and a recombinant plasmid harbouring V5-fused Rab7 (Control). W, D and C indicate wild-type, dominant-negative GDP-bound (T22N) and constitutively active GTP-bound (Q67L) Rab7, respectively. About 5% of the cell lysate used for immunoprecipitation was also subjected to Western blotting with an anti-V5 antibody (lower panel). Dominant-negative Rab7 co-precipitated with FLAG-a3 (middle panel, arrow). Numbers below blots represent relative signal intensities of V5-Rab7 co-precipitated with a isoform normalised to that co-precipitated with a3. Relative intensities less than 0.1 are not shown. (b) Interaction of a3 and dominant-negative Rab proteins. FLAG-a3 and the V5-fused dominant-negative form of Rab7, Rab11B or Rab27A were co-expressed in HEK293T cells. Immunoprecipitation was carried out as described in (a). Rab7 and Rab27A were detected in the precipitate (middle panel, closed and open arrows, respectively). Numbers below blots represent relative signal intensities of V5-fused Rab protein co-precipitated with a3 normalised to that of V5-Rab7. Relative intensities less than 0.1 are not shown. (c) Interaction of a3 and the Rac, Ran or H-Ras GTP-binding protein. FLAG-a3 and V5-fused Rac, Ran or H-Ras were co-expressed in HEK293T cells. Immunoprecipitation was carried out as described in (a). Numbers below blots represent relative signal intensities of V5-fused GTP-binding protein co-precipitated with a3 normalised to that of V5-Rab7. Relative intensities less than 0.1 are not shown. Unprocessed scans of immunoblots in (a–c) are shown in Supplementary Fig. S6. (d) Model illustrating the role of a3 in secretory lysosome trafficking. During osteoclast differentiation, the a3 isoform, the level of which is increased, recruits GDP-bound Rab7, and GDP is subsequently exchanged for GTP via an unknown GEF. Activated Rab7 anchors secretory lysosomes to microtubules, in collaboration with other machineries that are omitted here. Interestingly, a3 is also required for the localisation of GTP-bound Rab7 to lysosomes. The dotted line between a3 and GTP-bound Rab7 indicates a functional interaction between these two proteins.