Localization of apical membrane protein Gp135 around the centrosome and its transport during epithelial polarization

(A) Single MDCK cells were cultured in Matrigel for 12 h. Immunostaining was performed with the following markers: apical membrane marker glycoprotein 135 (Gp135, green), centrosome marker γ-tubulin (magenta), acetyl-tubulin (white), and DAPI for nuclei (blue). Single confocal sections through the middle of a cyst are shown. The order of polarization is arranged from single cell (1-cell), metaphase (Meta), telophase (Telo), cytokinetic pre-abscission (Pre-Abs), post-cytokinesis (Post-CK), to lumen open (LO). Arrows indicate Gp135 clusters around centrosomes, while arrowheads point to centrosome positions. Scale bar: 5 μm.

(B) Illustration showing the use of a fixed-size oval tool to select and measure Gp135 intensity in the chosen region. 1. Cytosol. 2. Cytokinesis (CK) bridge. 3. Centrosome (Cent.).

(C–D) Boxplots of Gp135 intensity surrounding centrosomes at different stages normalized to the mean in metaphase (C) and at different cellular regions at the cytokinetic pre-abscission stage normalized to the mean in the cytosol (D). We analyzed >15 cells for each stage or region in three independent experiments. Statistical analyses were performed via one-way ANOVA and Dunn’s multiple comparisons (****p<0.0001, ***p<0.001, **p<0.01, ns: not significant). The midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values (a.u., arbitrary units).

(E) Illustration showing how the polarity index (LN-C [nucleus–centrosome distance] divided by rN [average nuclear radius]) and θN-C (angle between the N-C and N-N axes) are calculated.

(F–G) Boxplots depicting centrosome positions at different stages, shown by the polarity index (F) and θN-C (G). We analyzed >16 cells per measurement in three independent experiments. Statistical analyses used one-way ANOVA and Dunn’s multiple comparisons (*p<0.05, **p<0.01, ****p<0.0001, ns: not significant). Midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values.

(H) Localization overview summarizing the localization of the centrosome (magenta) and recycling endosome/apical membrane components (green) during cell-division-directed polarization in Matrigel culture. White lines depict intercellular bridge microtubules, and nuclei are shown in blue. Arrows indicate apical membrane component localization around centrosomes and their transport from the centrosome to the AMIS.

Timing of centrosome migration and Gp135 targeting to the AMIS

(A) Time-lapse snapshots of MDCK cells expressing EGFP-Gp135 (green) and PACT-mKO1 (magenta, centrosome marker) in Matrigel. Nuclei were labeled with SiR-DNA (blue) before live imaging. Z-projection images of a dividing cell are shown. Time stamps show hours and minutes, with 0:00 set at the first frame of anaphase onset. Scale bar: 10 μm.

(B–D) Change in polarity index, θN-C, and LO-C over time. Each data point represents the average at a given time (>10 cells in Matrigel culture from three independent experiments). The lines show the means, and the shaded regions indicate SD values.

(E) Fluorescent profiles of EGFP-Gp135 along the line connecting the two nuclei in a cell doublet. White arrows indicate the central value used to indicate the level of Gp135 on the AMIS (a.u., arbitrary units). Time stamps show hours and minutes, with 0:00 set to the first frame after anaphase onset. Scale bar: 10 μm.

(F) Change in the central value in the EGFP-Gp135 fluorescent profile over time. The value of each time point was normalized to the value at 0:00. Each data point represents the average fluorescent intensity at a given time (10 cells in Matrigel culture from three independent experiments). The line shows the mean, and the shaded region indicates SD values (a.u., arbitrary units).

Impact of centriole depletion on apical lumen formation and Gp135 protein levels

(A) Time-lapse snapshots of p53-KO and centrinone (CN)-treated cells expressing EGFP-Gp135 (green) and PACT-mKO1 (magenta, centrosome marker) in Matrigel. Nuclei were labeled with SiR-DNA (blue) before live imaging. Z-projection images of dividing cells are shown. Time stamps show hours and minutes, with 0:00 set at the first frame of anaphase onset. Scale bar: 10 μm.

(B–C) Change in θN-C and LO-C over time. Each data point represents the average at a given time (>3 p53-KO cells in Matrigel culture from three independent experiments). The lines show the means, and the shaded regions indicate SD values.

(D) Change in the central value of the EGFP-Gp135 fluorescent profile over time. The value of each time point was normalized to the value at 0:00. Each data point represents the average fluorescent intensity at a given time (n = 3 [p53-KO; olive] and 3 [p53-KO + CN; gray]). The lines show the means, and the shaded regions indicate SD values (a.u., arbitrary units).

(E) Single MDCK WT and p53-KO cells, with or without CN treatment, after 72 h of Matrigel culture. Single confocal sections through the middle of cysts are shown with immunofluorescent signals of indicated markers: apical membrane Gp135 (green), centrosome marker γ-tubulin (magenta), and DAPI for nuclei (blue). “L” denotes the lumen. WT cells treated with CN remain at the single-cell stage. The insets display enlarged images of the region in the yellow box. Arrowheads indicate the presence of centrosomes. Scale bar: 10 μm. The bottom panels represent a larger view of the 3D culture. Scale bar: 50 μm.

(F) Quantification of the proportion of MDCK cysts with a single central lumen after being cultured for different durations (24, 48, 72 h). The lumen structure was identified by Gp135 staining. n = 92 (24 h), 231 (48 h), 78 (72 h) cysts (MDCK WT); n = 22 (24 h), 133 (48 h), 59 (72 h) cysts (WT + CN); n = 154 (24 h), 170 (48 h), 204 (72 h) cysts (p53-KO); n = 105 (24 h), 128 (48 h), 149 (72 h) cysts (p53-KO + CN) analyzed for each measurement in three independent experiments. Statistical analyses used two-way ANOVA and Tukey multiple comparisons (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001, ns: not significant). Values represent the mean ± SD.

(G) Boxplot of Gp135 intensity on the apical membrane normalized to the mean in WT cells. n = 25 (MDCK WT), 26 (p53-KO), 25 (p53-KO + CN) cysts were analyzed in three independent experiments. Statistical analyses were performed via one-way ANOVA and Dunn’s multiple comparisons (ns: not significant, ****p<0.0001). The midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values (a.u., arbitrary units).

Centrosome migration in CEP164, ODF2, CEP120, and PCNT KO cells during epithelial polarization

(A) Illustration depicting individual centrosomal structures (DA, SDA, and PCM proteins) and the state of microtubules on the centrosome in cells with knockout (KO) of CEP164, ODF2, p53/CEP120, or PCNT genes.

(B, F) Single MDCK cells of different genotypes were cultured in Matrigel for 12 h. Images show post-cytokinesis (CK) cells (B) or cells during pre-abscission (pre-Abs, F) with labeled markers: Gp135 (green), γ-tubulin (magenta), acetyl-tubulin (white), and DAPI for nuclei (blue). Images shown are single confocal sections through the middle of cells and side-view x-z cross-sections. Scale bar: 5 μm.

(C–D, G–H) Boxplots of centrosome positions, represented by θN-C and LO-C, for post-CK cells (C– D, n = 16 [WT], 10 [CEP164-KO], 15 [ODF2-KO], 20 [CEP120-KO], 14 [PCNT-KO]) and pre-Abs cells (G–H, n = 36 [WT], 14 [CEP164-KO], 16 [ODF2-KO], 12 [CEP120-KO], 14 [PCNT-KO]).

Cells were analyzed in three independent experiments. Statistical analyses used one-way ANOVA and Dunn’s multiple comparisons (ns: not significant). Midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values.

(E) Boxplot of Gp135 intensity at the AMIS normalized to the mean in WT cells. n = 25 (WT), 10 (CEP164-KO), 12 (ODF2-KO), 10 (CEP120-KO), and 10 (PCNT-KO) cell doublets were analyzed in three independent experiments. Statistical analyses were performed via one-way ANOVA and Dunn’s multiple comparisons (ns: not significant). The midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values (a.u., arbitrary units).

Par3 as the upstream regulator of centrosome migration and polarized vesicle transport

(A) Single MDCK cells, stably expressing pLKO-GFP-shRNA to knock down specified genes, cultured in Matrigel for 12 h. Images show post-cytokinesis cells with labeled markers: γ-tubulin (magenta), Gp135 (white), pLKO-GFP (green), and DAPI for nuclei (blue). Z-projection images between two centrosomes and the side-view x-z cross-section (bottom) are shown. “Ctrl” indicates the expression of scrambled-sequence shRNA. Arrowheads point to mislocalized centrosomes. Scale bar: 5 μm.

(B–D) Boxplots of centrosome positions, represented by θN-C, LO-C (B–C, n = 22 [sh-Ctrl], 20 [sh-Rab11a], 20 [sh-Sec15a], 20 [sh-MyoVb], 44 [sh-Par3], 20 [sh-Crb3], 20 [sh-Cdc42]), and LIC (D, n = 10 [sh-Ctrl], 10 [sh-Rab11a], 10 [sh-Sec15a], 10 [sh-MyoVb], 16 [sh-Par3], 10 [sh-Crb3], 10 [sh-Cdc42]), in post-cytokinesis cells. Cells were analyzed in three independent experiments. Statistical analyses used one-way ANOVA and Dunn’s multiple comparisons (ns: not significant, ***p<0.001, **p<0.01, *p<0.05). Midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values.

(E) Single WT and Rab11a-KO MDCK cells were cultured in Matrigel for 12 h with labeled markers: Par3 (green), γ-tubulin (magenta), acetyl-tubulin (white), and DAPI for nuclei (blue). Single confocal sections of pre-abscission cells are shown. The bridge region is enlarged from the yellow box. Scale bar: 5 μm.

(F) Single confocal sections of p53-KO cells, with or without centrinone (CN) treatment, cultured in Matrigel for 12 h with labeled markers: Par3 (green), γ-tubulin (magenta), acetyl-tubulin (white), and DAPI for nuclei (blue). The bridge region shown in the yellow box is enlarged. The bottom panels display Z-projected images to demonstrate the complete depletion of centrosomes. Scale bar: 5 μm.

Distribution of trafficking and polarity proteins during cytokinesis-directed de novo epithelial polarization

Centrosome migration in suspended 3D culture during cytokinetic pre-abscission

(A) Time-lapse snapshots of MDCK cells expressing EGFP-Gp135 (green) and PACT-mKO1 (magenta, centrosome marker) in Matrigel. Nuclei were labeled with SiR-DNA (cyan) before live imaging. Z-projection images of an aggregated cell doublet are shown. Arrows indicate Gp135 on the outer surface of the cell aggregate. Arrowheads point to mislocalized centrosomes that did not move over 3 h. Time stamps show hours and minutes, with 0:00 set at the beginning of live cell imaging. Scale bar: 10 μm.

(B, E) MDCK cells, with or without aphidicolin (Aphi) synchronization, were cultured in Matrigel for 6h with labeled markers: Rab11a (B, green), Par3 (E, green), γ-tubulin (magenta), and DAPI for nuclei (blue). Single confocal sections are shown. Two types of cell doublets, formed by cell division (post cytokinesis [CK]) or aggregation, were compared. Arrows indicate Gp135 clusters around centrosomes. Arrowheads point to mislocalized centrosomes. Scale bar: 5 μm.

(C–D) Boxplots of centrosome positions, represented by θN-C (C) and LIC (D), in cell doublets (n = 15 [unsynchronized control], 20 [Aphi-synchronized cells]). Cells were analyzed in three independent experiments. Statistical analyses used an unpaired two-tailed Mann–Whitney U test (*p<0.05, ****p<0.0001). Midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values.

(F) MDCK cells were cultured in low-attachment microwells for 24h with labeled markers: Gp135 (green), γ-tubulin (magenta), Par3 (white), and DAPI for nuclei (blue). A single confocal section through the middle of a cell doublet is shown. Yellow arrows indicate the Par3signal at the edge of cell-cell contacts. The 3D reconstruction reveals Par3 forming a ring around the edge between cell doublets. Scale bar: 10 μm.

(G) Time-lapse snapshots of MDCK cells expressing EGFP-Gp135 (green) and PACT-mKO1 (magenta, centrosome marker) in low-attachment microwells. Nuclei were labeled with SiR-DNA (cyan) before live imaging. Z-projection images of a dividing cell are shown. White and yellow arrows indicate Gp135 on the outer surface of the cell doublet and around the centrosome, respectively. Arrowheads point to the centrosome position. Time stamps show hours and minutes, with 0:00 set at the first frame of anaphase onset. Scale bar: 10 μm.

(H–J) Change in polarity index (H), θN-C (I), and LO-C (J) over time. Each data point represents the average at a given time (three cells in low-attachment microwells from three independent experiments). The lines show the means, and the shaded regions indicate SD values.

Summary of centrosome migration, apical membrane component trafficking, and Par3 recruitment in different experiments and culture conditions

(A) In conventional Matrigel culture, centrosomes move directionally toward AMIS following cell division (magenta arrow). Apical membrane components like Gp135, Crb3, and Cdc42 follow the centrosomes (green arrow). Par3 first emerges at the cytokinesis site, which regulates centrosome positioning and polarized vesicle trafficking during polarization.

(B) Loss of centrosomes diminishes the effectiveness of apical membrane component trafficking (smaller green arrow), yet Par3 still localizes to the cytokinetic bridge. Additionally, centrosome loss mainly impacts the initial stages of epithelial polarization rather than the later stages of lumen formation.

(C) Loss of Par3 results in randomized centrosome positioning, obstructs polarized trafficking, and traps apical membrane components around mislocalized centrosomes.

(D) Two-cell aggregates without cell division cannot effectively guide centrosome migration and polarized vesicle trafficking. Despite Par3 being recruited to the cell-cell interface, apical membrane components remain trapped around mislocalized centrosomes.

(E) MDCK cells suspended in the ECM-free condition. Centrosomes migrate to the center of cell doublets during cytokinesis (magenta arrow), while Apical membrane components transport in the opposite direction (green arrow). After mitosis, Par3 exhibits a pattern distinct from that observed in Matrigel culture.

Distribution of Rab11a, Gp135, and aPKC proteins and centrosome position during de novo epithelial polarization

(A, F) Single MDCK cells were cultured in Matrigel for 12 h with labeled markers: Gp135 (green), γ-tubulin (magenta), and Rab11a (white) in (A), and aPKC (green), γ-tubulin (magenta), and Gp135 (white) in (F). DAPI staining was applied for nuclei (blue). Single confocal sections are shown through the middle of the cells. The centrosome region during telophase (Telo) and the AMIS region in post-cytokinesis (Post-CK) cells, as shown in the yellow box, are enlarged. Yellow arrows indicate Rab11a and aPKC around centrosomes or at the AMIS. White arrows indicate Gp135. Arrowheads point to the centrosome position. LO: lumen open. Scale bar: 5 μm.

(B, D) Illustration showing LO-C (origin to centrosome distance) and LIC (inter-centrosomal distance) for cell doublets. The center point between the two nuclei is taken as the origin (O).

(C, E) Boxplots depicting centrosome positions at different stages, shown by LO-C (C) and LIC (E). We analyzed >15 cells per measurement in three independent experiments. Statistical analyses used one-way ANOVA and Dunn’s multiple comparisons (*p<0.05, **p<0.01, ****p<0.0001). Midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values.

Clustering of Crb3 and Cdc42 around centrosomes during de novo epithelial polarization

(A–B) Single MDCK cells stably expressing EGFP-Crb3 (A, green) or EGFP-Cdc42 (B, green) were cultured in Matrigel for 12 h with labeled markers: γ-tubulin (magenta), acetyl-tubulin (white), and DAPI for nuclei (blue). Single confocal sections through the middle of cells are shown. The centrosome, cytokinesis (CK) bridge, and AMIS region are enlarged from the boxed area in pre-abscission (Pre-Abs) and post-CK cells. White arrows indicate EGFP-Crb3 (A) or EGFP-Cdc42 (B) around centrosomes. Arrowheads point to the centrosome position. Yellow arrows indicate the transport of Crb3 and Cdc42 from the centrosome to the mid-membrane. Scale bar: 5 μm.

Differential centrosome migration of MDCK cells in adherent 2D culture versus 3D culture conditions during cytokinetic pre-abscission

(A) Time-lapse snapshots of MDCK cells expressing EGFP-Gp135 (green) and PACT-mKO1 (magenta, centrosome marker) on a cover glass. Nuclei were labeled with SiR-DNA (blue) before live imaging. Z-projection images of a dividing cell are shown. Time stamps show hours and minutes, with 0:00 set at the first frame of anaphase onset. Scale bar: 10 μm.

(B–C) Change in θN-C and LO-C over time. Each data point represents the average at a given time (6 cells on cover glass from three independent experiments). The lines show the means, and the shaded regions indicate SD values.

Assessment of centrinone (CN) treatment and generation of p53-KO MDCK cells

(A) Bar graph showing progressive centrosome depletion after CN treatment in MDCK cells. The distribution of centrosome numbers over time after the addition of CN is shown.

(B) MDCK cells are present three days after CN addition and an untreated control. Scale bar: 10 μm. Insets show enlargements of the image in the boxed area. Scale bar: 1 μm.

(C) Generation of the p53-KO MDCK cell line. The diagram illustrates two single-guide RNA (sgRNA) sequences and their target sites in p53 exon 4 (blue and green boxes). The red boxes indicate the position of the protospacer-adjacent motif (PAM) sites and the predicted Cas9 cut sites are indicated by red arrowheads. Genotyping results show a genomic mutation in the p53-KO cell line with a deletion (del) within exon 4 of the p53-encoding gene.

(D) Binding of p53 antibody and mutation induced in the p53-encoding gene.

(E) Immunofluorescence images showing p53expression in MDCK WT and p53-KO cells with or without CN treatment. Scale bar: 10 μm.

(F) Immunoblot results for p21 protein in MDCK WT and p53-KO cells with or without CN treatment.

(G) Change in polarity index over time. Each data point represents the average at a given time (four p53-KO cells in Matrigel culture from three independent experiments). The lines show the means, and the shaded regions indicate SD values.

(H) Protocol for CN pre-treatment and the time point of Matrigel culture in the presence of CN.

Impact of centriole depletion on MDCK cell proliferation and lumen formation

(A) Single MDCK WT and p53-KO cells, with or without centrinone (CN) treatment, after 8–12h of Matrigel culture. Single confocal sections through the middle of cysts are shown with immunofluorescent signals of indicated markers: Gp135 (green), γ-tubulin (magenta), acetyl-tubulin (white), and DAPI for nuclei (blue). Acentrosomal cells fail to cluster Gp135-positive vesicles and affect lumen formation in the two-cell stage (Pre-Abs, pre-abscission; LO, lumen open). Scale bar: 5 μm.

(B) Boxplot of Gp135 intensity at the AMIS normalized to the mean in WT cells. n = 22 (MDCK WT), 26 (p53-KO), 28 (p53-KO + CN) post-CK cells were analyzed in three independent experiments. Statistical analyses were performed via one-way ANOVA and Dunn’s multiple comparisons (ns: not significant, ***p<0.001, ****p<0.0001). The midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values (a.u., arbitrary units).

(C) Proliferation curves of MDCK WT and p53-KO cells, with or without CN treatment. The cells with CN treatment underwent a 3-day pre-treatment before examination. Values are presented as the mean ± SD.

(D) Proportion of lumen opening in MDCK cell doublets after 24 h of Matrigel culture based on Gp135 staining. n = 17 (MDCK WT), 30 (p53-KO), and 45 (p53-KO + CN) cell doublets were analyzed in three independent experiments. Statistical analyses used two-way ANOVA and Tukey multiple comparisons (ns: not significant, ****p<0.0001). Values represent the mean ± SD.

Generation of CEP164 and ODF2-KO MDCK cell lines

(A, E) Diagrams illustrating two single-guide RNA (sgRNA) sequences and their targeting sites within CEP164 exon 4 and ODF2 exon 7 (blue and green boxes). The red boxes indicate the position of the protospacer-adjacent motif (PAM) sites and the predicted Cas9 cut sites are indicated by red arrowheads. Genotyping results of the genomic mutations are shown. The predicted genomic mutations (del: deletion, ins: insertion) are illustrated.

(B, F) Diagrams illustrating the binding sites of the antibody and the position of the mutation induced in the CEP164- and ODF2-encoding genes.

(C, G) Immunofluorescence images depicting CEP164and ODF2 expression in MDCK WT control and KO cells with glutamylated tubulin and CEP120, respectively, as the centrosome marker. Scale bar: 5 μm. Insets present enlargements of the images within the boxed areas. Scale bar for insets: 1 μm.

(D, H) Boxplots of the CEP164 and ODF2 signal at the mother centriole. Fluorescent intensity values of KO cells were normalized to the mean of WT cells. n = 70 (MDCK WT), 51 (CEP164-KO) in (D), and n = 50 (MDCK WT), 50 (ODF2-KO) in (H) were analyzed in three independent experiments. Statistical significance was determined by an unpaired two-tailed Mann–Whitney U test (****p<0.0001). The midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values (a.u., arbitrary units).

Generation of p53/CEP120 and PCNT-KO MDCK cell lines

(A, E) Diagrams illustrating two single-guide RNA (sgRNA) sequences and their targeting sites within PCNT exon 4 and CEP120 exon 3 (blue and green boxes). The red boxes indicate the position of the protospacer-adjacent motif (PAM) sites, and the predicted Cas9 cut sites are indicated by red arrowheads. Genotyping results of the genomic mutations are shown. The predicted genomic mutations (del: deletion, ins: insertion) are illustrated.

(B, F) Diagrams illustrating the binding sites of the antibody and the position of the mutation induced in the PCNT- and CEP120-encoding genes.

(C, H) Immunofluorescence images depicting PCNT and CEP120expression in MDCK WT control and KO cells with acetyl-tubulin and γ-tubulin, respectively, as the centrosome marker. Scale bar: 5 μm. Insets present enlargements of the images within the boxed areas. Scale bar for insets: 1 μm.

(D) Boxplots of the PCNT signal at the centrosome. Fluorescent intensity values of KO cells were normalized to the mean of WT cells. n = 20 (MDCK WT), 20 (PCNT-KO) were analyzed in three independent experiments. Statistical significance was determined by an unpaired two-tailed Mann– Whitney U test (****p<0.0001). The midlines and boxes show the mean ± SD, with whiskers indicating minimum and maximum values (a.u., arbitrary units).

(G) Western blot confirming the loss of CEP120 in p53/CEP120-KO cells.

Characterization of MDCK RNAi and localization of Par3 during de novo epithelial polarization

(A–F) MDCK cells stably expressing different pLKO-GFP-shRNA constructs targeting Rab11a (A), Sec15a (B), MyoVb (C), Par3 (D), Crb3 (E), and Cdc42 (F). The knockdown efficiency was verified by blotting with appropriate antibodies (A, D, F) or quantitative RT-PCR (B, C, E), with GAPDH used as a control. For quantitative RT-PCR, statistical significance was determined by an unpaired two-tailed Mann–Whitney U test (*p<0.05, **p<0.01). Values represent the mean ± SD.

(G) Single MDCK cells stably express pLKO-GFP-shRNA to knock down Par3 after 12 h of Matrigel culture. Images show post-cytokinesis (CK) cells with labeled markers: γ-tubulin (magenta), Rab11a (white), pLKO-GRP (green), and DAPI for nuclei (blue). Z-projection images between two centrosomes are shown. “sh-Ctrl” indicates the expression of scrambled-sequence shRNA. The enlarged images show the centrosome position in the yellow-boxed area. Scale bar: 5 μm.

(H) Single MDCK cells after 12h of culture in Matrigel. Immunostaining signals of the indicated markers are shown: centrosome marker γ-tubulin, polarity protein Par3, acetyl-tubulin, and DAPI. Single confocal sections through the middle of cysts are displayed. The enlarged images display the Par3signals as they first emerge near the cytokinesis bridge, as indicated in the yellow boxed area. Arrowheads point to the centrosomes (LO, lumen open). Scale bar: 5 μm.

Localization of cell-cell junction components during de novo epithelial polarization and generation of Rab11a-KO MDCK cell lines

(A–B) Single MDCK cells after 12h of culture in Matrigel. Immunostaining signals of the indicated markers are shown: centrosome marker γ-tubulin, polarity protein Par3, acetyl-tubulin, actin, E-cadherin (E-cad), ZO-1, and DAPI. Single confocal sections through the middle of cysts are displayed. The enlarged images display the Par3and ZO-1 signals as they first emerge near the cytokinesis bridge, as indicated in the yellow boxed area. Arrowheads point to the centrosomes. Scale bar: 5 μm.

(C) Diagram illustrating two single-guide RNA (sgRNA) sequences and their targeting sites within Rab11a exon 1 and exon 2 (blue and green boxes). The red boxes indicate the position of the protospacer-adjacent motif (PAM) sites and the predicted Cas9 cut sites are indicated by red arrowheads. Genotyping results of the genomic mutations, obtained via gel-purified PCR products covering exon 1 or 2 of the Rab11a-encoding gene in the Rab11a-KO cell line, are shown. The predicted genomic insertion (ins) within exons 1 and 2 of the Rab11a-encoding gene is illustrated.

(D) Diagram illustrating the mutation position predicted in the Rab11a-encoding gene.

(E) Western blot confirming the loss of Rab11a in two Rab11a-KO lines.

Centrosome position relative to the cytokinesis site in polarized epithelial sheets

(A) Experimental design: Aphidicolin (Aphi)-treated cells (synchronized at the S phase) were seeded into Matrigel and observed within 8 h to exclude the possibility of cell doublet formation through cell division.

(B) Bright-field images of MDCK cells cultured in low-attachment microwells. The images are merged with the DAPI (cyan) signal to show cell division occurring over time.

(C–D) Polarized MDCK cyst and epithelial sheet with labeled markers: the centrosome markers PACT-mKO1 and γ-tubulin, acetyl-tubulin, F-actin, and DAPI. Insets show enlargements of the image in the boxed area. “L” indicates the lumen in (C). The z-projection of the top view and the X-Z cross-section of the side view (bottom) are shown in (D). Telo, telophase; Pre-Abs, pre-abscission. Arrowheads point to the centrosome position. Scale bar: 10 μm.

(E) Time-lapse snapshots of MDCK cells expressing EGFP-Gp135 and PACT-mKO1 cultured on a Transwell insert to form polarized epithelial sheets. The microtubule probe SiR-tubulin (white) was applied before live cell imaging. The X-Y top view and the X-Z cross-section of the side view (bottom) are shown. Time stamps show hours and minutes, with 0:00 set to the first frame after anaphase onset. Arrowheads point to the centrosomes. The yellow arrow indicates the abscission of the cytokinesis bridge. Scale bar: 10 μm. A schematic depicts the centrosomes moving to the apical membrane before bridge abscission.

(F) Change in distance from the centrosome to the apical membrane, as measured in each frame (data from three independent experiments, n = 5 cells). The graph presents the mean ± SD.

Uncropped Western blots are shown in this manuscript

(A) Western blots showing that p21 is present in MDCK WT and p53-KO cells with or without centrinone (CN) treatment are shown in Figure 3—figure supplement 1F.

(B) Western blots showing the knockout of CEP120 from the p53-KO cell line are shown in Figure 4—figure supplement 2G.

(C–D) Western blots show the depletion of Rab11a and Par3 in MDCK WT cells, as shown in Figure 5—figure supplement 1A and D.

Uncropped Western blots are shown in this manuscript

(A) Western blots showing the depletion of Cdc42 in MDCK WT cells are shown in Figure 5—figure supplement 1F.

(B) Western blots showing the knockout of Rab11a from MDCK WT cells are shown in Figure 5— figure supplement 2E.