Fig 2: ANKS1A deficiency results in a marked loss of the microtubule network in aged mice.a–d 3DSIM images of the MT networks stained with alpha-tubulin antibody for both young (P45) and aged (20-month old) mice. a, b ANKS1A+/+ set, n = 6 cells from 2 mice; ANKS1A−/−, n = 5 cells from 2 mice. c, d ANKS1A+/+ set, n = 9 cells from 3 mice; ANKS1A−/−, n = 9 cells from 3 mice. Scale bar for a, c, 5 μm. e A representative TEM image illustrating the intermediate filament and microtubule. Scale bar, 500 nm. f–j 3D reconstruction of the microtubule and intermediate lattice. The image of the microtubules (green) and intermediate filament (orange) was reconstructed and superimposed on a single plane. For g, h (P45 mice), ANKS1A+/+ set, n = 10 cells from 2 mice; ANKS1A−/−, n = 10 cells from 2 mice. For j, k (20-month-old mice), ANKS1A+/+ set, n = 9 cells from 2 mice; ANKS1A−/−, n = 9 cells from 2 mice. Scale bar for f, i, 500 nm.

Fig 3: ANKS1A loss results in defective rotational polarity.a Schematic depicting the flow of CSF (225°) in the AD region. Double staining for FOP (green, SDA marker) and γ-Tubulin (red, BF marker) to visualize rotational polarity. b Images of the AD region obtained by 3DSIM, and BBs magnified for rotational polarity analysis (top panels). Each black arrow points from the FOP staining toward the γ-Tubulin-positive dot, indicating the rotational polarity axis of each individual BF (bottom panels). Scale bar, 1 μm. c Histogram showing the distribution of the rotational vector angles. Error bars on the graph represent CSD. d En face electron micrographs from the apical surface of ANKS1A WT and KO samples (top panels). Schematics showing the BBs (circular) with BFs (triangular) appear in the bottom panels. Scale bar, 2 μm. e The TEM data shown in d were used to draw histograms. Error bars on the graphs represent CSD.

Fig 4: ANKS1A loss results in uncoordinated beating of motile cilia.a Schematic of LWs showing three regions around the adhesion area (marked by a circle) or the movement of fluorescent beads (green) on a live LW. AD anterior–dorsal, AV anterior–ventral, PM posterior–medial. b, c Whole-mount staining of LWs using GT335 (glutamylated tubulin) or AcTub (acetylated tubulin) antibodies at P20. The black arrow (~225°) indicates the direction of CSF flow in the AD region. The abnormal phenotype of KO ependymal cilia was observed from at least five different pairs of WT and KO littermates. Scale bar for b, c, 20 μm. d SEM analysis of LWs from littermates at P30. Scale bar, 20 μm. This result was reproducibly observed from at least three different pairs of WT and KO littermates. e High-speed video imaging analysis of the same fluorescent bead at different time points. f Thirty-five consecutive frames of Supplementary Movie 1 were merged into a single picture. Scale bar for e, f, 200 μm. g Data presented in d were used to calculate an average bead speed based on at least three independent experiments. Data represent mean ± S.D. Each point on the graph represents the speed of an individual bead.

Fig 5: Inducible ANKS1A ablation results in a rotational polarity defect, affecting the SDAs.a Control and iKO mice were generated via five separate daily TM injections from P30 to P35. b The mice were sacrificed at P42 for live LW preparation, and the experiments were performed essentially as described in Fig. 3. c Thirty-five consecutive frames of Supplementary Movie 3 were merged into a single picture. Scale bar for b, c, 200 μm. d Data presented in b were used to calculate an average bead speed based on three independent experiments. Data represent mean ± SD. e Double staining for both ODF2 (green) and CNTRL (red) permitted the visualization of rotational polarity. f, g Experiments were performed as described in Fig. 4b, c. h, i Experiments were performed as described in Fig. 2e, f. Scale bar for f, h, 1 μm. Control set, n = 99 cells; ANKS1A iKO, n = 71 cells.