Colocalization patterns of EXOC6A and myosin-Va at preciliary vesicles, ciliary vesicles, and ciliary sheath during ciliogenesis.

(A) RPE1-based mCherry-Myo-Va-GTD cells were treated with Dox for 24 h, followed by serum starvation for 2h. Cells were then stained with antibodies against EXOC6A (green) and polyglutamylated tubulin (Glu-tub) (gray) and analyzed using immunofluorescence confocal microscopy. (B and C) RPE1-based mCherry-Myo-Va-GTD-inducible cells were treated as described in A and immunostained with antibodies against EXOC6A (green), CEP120 (centriole marker), and CEP164 (distal appendage marker, gray). Images were captured using 3D-SIM or ultra-expansion microscopy (U-ExM) with an LSM880 confocal system. (D) CFLEM analysis of the localization of GFP-EXOC6A during ciliogenesis. RPE1-based GFP-EXOC6A-inducible cells were treated with Dox for 24 h and subjected to serum starvation for 2 or 24 h to observe the CVs or cilia membrane, respectively. Images were taken via SIM and TEM and then merged based on their relative localization. GFP-EXOC6A signals were located at PCVs, CVs (i-ii), and the ciliary membrane (iii-iv). (E) RPE1-based GFP-EXOC6A and mCherry-Myo-Va-GTD-inducible cells were treated with Dox for 24 h and analyzed via 3D-SIM using the indicated antibodies. Right panels show fluorescence profile plots. EXOC6A colocalized with myosin-Va at the ciliary sheath, while ARL13B (i) and INPP5E (iii) are the ciliary shaft markers. Scale bars are 1 μm.

Spatial–temporal localization of EXOC6A and Myo-Va during ciliogenesis.

(A) RPE1-based mCherry-Myo-Va-GTD-inducible cells were treated with Dox for 24 h, serum starved, released at different time points, and analyzed via 3D-SIM using the indicated antibodies. The white arrow indicates a tubule extending from the ciliary membrane. (B, C) Schematic model showing the localization patterns of EXOC6A (green) and Myo-Va (red) during ciliogenesis, and the percentages of cells with different types of staining patterns (%) are shown in C. Scale bars are 1 μm.

FRAP analysis of Myo-Va- and EXOC6A-labeled vesicles during ciliogenesis.

(A and B) RPE1-based GFP-EXOC6A and mCherry-Myo-Va-GTD-inducible cells were treated with Dox for 24 h and serum starved for 30 min. Regions of CVs (A) or the ciliary membrane (B) of the cells were photobleached using the LSM880 confocal microscope with the 405 nm laser (related to Movies 1 and 2). Signal intensity of EXOC6A is shown in the right panel. (C and D) Dynamic localization of GFP-EXOC6A at CVs (C) or the ciliary membrane (D) during ciliogenesis. RPE1-based GFP-EXOC6A-inducible cells were treated with Dox for 24 h and serum starved for 30 min. Images were taken using an Elyra 7 high-speed live-cell imaging system. Series of images in (C) shows the progression of fusion of GFP-EXOC6A-associated vesicles into CVs (white arrow). Series of images in (D) shows the process by which GFP-EXOC6A-associated vesicles fuse into the ciliary pocket (white arrow) and simultaneously exit or reintegrate into the ciliary pocket (red arrow). Some images also show the release of tubule-like structures in the ciliary membrane (yellow arrow), or the excretion of vesicles in the ciliary membrane (orange arrow). These data suggested that the components of CVs, the ciliary pocket, and the ciliary membrane are highly dynamic (related to Movies 4 and 5). Scale bars are 2 μm.

Myo-Va and EHD1 are required for EXOC6A-labeled CV formation, and EXOC6A is associated with Myo-Va.

(A) Myo-Va is required for EXOC6A-labeled CV formation. RPE1-based Myo-Va KO cells and WT cells immunostained with antibodies against EXOC6A (green) and Glu-tub (red). Quantification of cells with EXOC6A-positive CVs (middle panel) and immunoblotting results (right panel) are shown. (B) EXOC6A is not required for Myo-Va-labeled CV formation. RPE1-based mCherry-Myo-Va-GTD-inducible cells against EXOC6A KO background and WT cells were immunostained with antibodies against Glu-tub (red). For easy differentiation, the mCherry-Myo-Va-GTD signal is converted to green. Quantification of cells with Myo-Va-associated CVs (middle panel) and immunoblotting results (right panel) are shown. (C) EHD1 is required for EXOC6A-labeled CV formation. RPE1 cells treated with siControl or siEHD1 for 48 h were immunostained with antibodies against EXOC6A (green) and Glu-tub (red). Quantification of cells with EXOC6A-positive CVs (middle panel) and immunoblotting results (right panel) are shown. (D) EXOC6A is not required for EHD1-labeled CV formation. RPE1-based EXOC6A KO and WT cells were immunostained with antibodies against EHD1 (green) and Glu-tub (red). Quantification of cells with EHD1-positive CVS is shown (right panel). (E) Co-IP experiments analyzed the association between EXOC6A and its potential binding proteins. Cell lysates were immunoprecipitated (IP) with EXOC6A or Myo-Va antibodies, followed by immunoblotting with the indicated antibodies. Error bars in A–D represent mean ± s.d. from at least 3 independent experiments with 100 cells per experiment. P-value was determined with two-tailed Student’s t-test. P < 0.05 was considered statistically significant. NS, not significant. Scale bars are 1 μm.

Transportation of EXOC6A-associated vesicles to the M-centriole occurs via a dynein-, actin-, and MT-dependent pathway

(A) Schematic showing the protocols of siRNA (left) or drug (right) treatment. (B) RPE1 cells were treated with siControl or siGolgin160 and analyzed via confocal fluorescence microscopy or immunoblotting using the indicated antibodies. (C) RPE1 cells were treated with ciliobrevin D (i) or nocodazole (ii) and analyzed via confocal fluorescence microscopy using the indicated antibodies. (D-F) Cells with EXOC6A-associated CVs (D), a positioned Golgi (E), or a cilia ratio (F) were analyzed via 3D-SIM and quantified. Cells with a positioned Golgi were defined as those showing all GM130-labeled Golgi signals concentrated within a 7 μm radius surrounding the Glu-tub-labeled centrioles. (G) RPE1-based mCherry-myosin-Va-GTD-inducible cells were treated with siARP2 as described in A (left) and analyzed via confocal microscopy or immunoblotting using the indicated antibodies. (H, I) Quantifications of the cilia ratio (H) and EXOC6A-associated CVs (I) is shown. (J) RPE1 cells were treated with CK666 or high-dose (10 μM) CytoD as described in A (right panel) and analyzed via confocal microscopy using the indicated antibodies. Quantification of the cilia ratio (H) and EXOC6A-associated CVs (I) is shown. * in (B, right panel) indicates non-specific bands. Error bars represent mean ± s.d. from at least 3 independent experiments with 100 randomly selected cells. P-value was determined with two-tailed Student’s t-test. P < 0.05 was considered statistically significant. NS, not significant. Scale bars are 1 μm.

Loss of EXOC6A inhibits ciliogenesis, and some cells exhibit an abnormal ciliary morphology when passing through the CV block.

(A-C) RPE1-based mCherry-Myo-Va-GTD-inducible WT and EXOC6A KO cells were treated with Dox for 24 h, serum starved (SS) at different time points, and analyzed via confocal fluorescence microscopy with the indicated antibodies. mCherry-Myo-Va-GTD signal is artificially converted to green for discrimination. (B) RPE1 WT and EXOC6A KO cells were serum starved at different time points and analyzed via confocal fluorescence microscopy with the indicated antibodies. For comparison, the intensities of ARBL13B and Glu-tub of EXOC6A KO cells with non-enhanced and enhanced signals are shown in B (lower panel). (C) Cilia ratio measured through ARL13B labeling (non-enhanced signal). (D and E) Quantitation of non-enhanced ARL13B (D) and non-enhanced Glu-tub intensity on cilia (E) of WT or EXOC6A KO cells is shown. (F) WT and EXOC6A KO cells were serum starved for 72 h. Morphologies of normal or abnormal cilia, including DA, CVs, and ciliary membranes, were examined via EM (F) and quantified (G; N is the number of cells examined). Error bars in C, D, and E represent mean ± s.d. from at least 3 independent experiments from 100 randomly selected cells. P-value was determined with two-tailed Student’s t-test. P < 0.05 was considered statistically significant. NS, not significant. Scale bars are 1 μm in A and B. Scale bars are 500 nm in F.

EXOC6A is required for the recruitment of MKS module proteins to the peripheral TZ.

(A) WT and EXOC6A KO cells were fixed for 2 h (A) or 24 h (B) after serum starvation and analyzed via confocal fluorescence microscopy using antibodies against TZ proteins. Glu-tub staining labeled the centriole. Quantification of the fluorescence signals of various MKS proteins and NPHP8 on ciliated WT or EXOC6A KO cells is shown in the right panel. Error bars represent the mean ± s.d. from at least 3 independent experiments with 100 randomly selected cells per experiment. P-value was determined with two-tailed Student’s t-test. P < 0.05 was considered statistically significant. NS, not significant. Scale bars, 1 μm.

Ciliogenesis model and comparison between WT and EXOC6A KO cells.

EXOC6A is located at the PCV/CV/ciliary sheath, and Myo-Va mediated the transportation of EXOC6A vesicles to the mother centriole via a dynein-, microtubule-, and actin-dependent pathway. EXOC6A-associated vesicles are continuously recruited and fused or excreted from the CVs or ciliary membrane. Depletion of EXOC6A impairs ciliogenesis, and most cells are arrested at the CV stage. There is a lack of MKS module proteins of the TZ in cilia with defects in EXOC6A KO cells. PCV, preciliary vesicle; CV, ciliary vesicle; DA, distal appendage; sDA, subdistal appendage.

Correlation of spatial localization of EXOC6A and Myo-Va during the early stages of ciliogenesis.

(A) RPE1-based mCherry-Myo-Va-GTD-inducible cells were treated with DOX for 24 h and serum starved for 0, 15, and 30 min. Fluorescence intensities of EXOC6A (green) or Myo-Va-GTD (red) signals within a 2 μm radius surrounding the Glu-tub-labeled centrioles (white) were quantified and are shown in B (EXOC6A) and C (Myo-Va-GTD). Correlation R (Pearson correlation coefficient) between EXOC6A and Myo-Va is shown in D. Correlation R was analyzed using Zeiss Zen blue software. Error bars in B–D represent mean ± s.d. from at least 3 independent experiments with 100 randomly selected cells. P-value was determined with one-way ANOVA. P < 0.05 was considered statistically significant. NS, not significant. Scale bars are 1 μm.

Live-cell imaging of GFP-EXOC6A co-localized with Myo-Va-GTD at the ciliary membrane

(A) RPE1-based inducible cells expressing GFP-EXOC6A and mCherry-MyoVa-GTD were treated with Dox for 24 h, serum starved for 30 min, and SiR-tubulin was added to label centriole and axoneme. (B) Single channel image (green box in A) captured at the starting time point (0 sec). (C) Single channel image (red box in A) captured at frame 21 (197.577 sec). Scale bars are 1 μm.

EXOC6A deletion does not interfere with the removal of CP110 from the mother centriole.

WT and EXOC6A KO cells were fixed 24 h after serum starvation and analyzed via fluorescence confocal microscopy using the indicated antibodies. (B) Percentages of cells with one or two CP110 dots are shown. Error bars represent the mean ± s.d. from at least 3 independent experiments with 100 cells per experiment. P-value was determined with two-tailed Student’s t-test. NS, not significant. Scale bars are 1 μm.

Low doses of Cytochalasin D (CytoD) promote cilia elongation, whereas higher concentrations (greater than 4 μM) inhibit ciliogenesis.

(A) RPE1 cells were treated with different concentrations of Cytochalasin D (CytoD), including control (0.1% DMSO), 200 nM, 1 μM, 2 μM, 4 μM, 7 μM, and 10 μM. Quantification of ciliary length labeled with ARL13B antibody (B) and cilia ratio (C) are shown. P-values were determined with a two-tailed Student’s t-test. P < 0.05 was considered statistically significant. Error bars represent the mean ± s.d. of 100 randomly selected cells from at least 3 independent experiments. Scale bars are 2 μm.

Rescue of ciliogenesis defects in EXOC6A knockout cells by GFP-EXOC6A re-expression.

(A) Representative immunofluorescence images of RPE1 cells (WT, WT+GFP-EXOC6A, EXOC6A KO, and EXOC6A KO+GFP-EXOC6A) stained for ARL13B (red) and Glu-tubulin (white). (B) Quantitative analysis of ARL13B-labeled cilia phenotypes. Data are presented as the mean ± s.d. of 100 randomly selected cells from at least three independent experiments. Scale bars are 1 μm.