Downregulation of CEP44 in breast cancer correlates with poor prognosis. (A) Immunoblots of CPAP and CEP44 in MCF10A, MDA-MB-436 and MDA-MB-231 cells. Tubulin was used as a loading control. (B) Immunostaining of Centrin-3 (red) and CEP44 (green) in MCF10A, MDA-MB-436 and MDA-MB-231 cells. (C) Quantification of the fluorescence intensity of CEP44 in the cells from (B). (D) CEP44 transcript levels in normal human mammary tissues and breast carcinoma samples based on data sets from TCGA. (E) CEP44 transcript levels in paired tumor and tumor-adjacent tissue samples based on data sets from TCGA. (F-G) Representative images (F) of IHC staining and the relative IHC scores (G) of CEP44 in breast cancer tissues and adjacent normal tissues. (H-I) Overall survival (OS) (H) and recurrence free survival (RFS) (I) were compared between breast cancer patients with high and low CEP44 expression using Kaplan-Meier Plotter. For C, error bars represent the means ± S.E.M for three independent experiments. n.s., not significant, ***p < 0.001, as determined using one-way ANOVA with Dunnett’s multiple comparisons test. For D, E and G, error bars represent the means ± S.E.M for three independent experiments. ***p < 0.001, as determined using Student’s t-test. Bars: 2 μm (B), 100 μm (F, upper), 50 μm (F, lower).

Depletion of CEP44 results in centriole overduplication. (A) Immunostaining of CEP97 (red) and α-tubulin (green) in MCF10A, MDA-MB-436, MDA-MB-231, SKBR3 and HCC1806 cells. DNA was stained with DAPI (blue). (B) Quantification of the percentage of cells with >4 centrioles from (A). (C-F) Immunostaining of Centrin-3 (green) (C) or CEP97 (green) (E) with γ-tubulin (red) in HeLa cells transfected with control- or CEP44-siRNA. DNA was stained with DAPI (blue). Quantification of the percentage of cells with the indicated number of Centrin-3 dots (D) and CEP97 dots (F) from (C and E). (G) Immunostaining of CEP97 (red) in CEP44-GFP-overexpressed MDA-MB-436 cells. (H) Quantification of the percentage of cells with more than 4 centrioles from (G). For B, D, F and H, error bars represent the means ± S.E.M for three independent experiments. n.s., not significant, *p < 0.05, **p < 0.01, ***p < 0.001, as determined using one-way ANOVA with Dunnett’s multiple comparisons test (D-F) or Student’s t-test (H). Bars: 10 μm (A), 5 μm (C, E), 2 μm (G).

Depletion of CEP44 results in premature centriole disengagement. (A) Immunostaining of Centrin-3 (red) and CEP192 (green) in wild-type (WT) and CEP44 knockout (KO) HeLa cells. DNA was stained with DAPI (blue). (B) Quantification of the percentage of cells with centriole disengagement from (A). (C) Immunostaining of CEP192 (green) and Centrin-3 (red) in WT and CEP44 KO HeLa cells. DNA was stained with DAPI (blue). (D) Immunostaining of CENP-F (green) and Centrin-3 (red) in WT and CEP44 KO HeLa cells treated with EdU (S phase marker; purple) for 30 min before fixation. DNA was stained with DAPI (blue). (E-F) Quantification of the percentage of cells with the indicated phenotype (E) and number of centrioles (F) from (D). (G-H) Time-lapse images of WT (G) and CEP44 KO (H) HeLa cells stably expressing H2B-RFP and GFP-Centrin-3. Red arrowheads indicate centrioles after the first duplication. Green arrowheads indicate the overduplicated centrioles after the second overduplication. For B, E and F, error bars represent the means ± S.E.M obtained for three independent experiments. **p < 0.01, as determined using Student’s t-test. Bars: 5 μm (A, C upper, D, G and H), 2 μm (C, lower).

CEP44 recruits CEP57/CEP57L1 to the centriole. (A) Lysates of HEK293T cells co-overexpressing V5-CEP57L1 and Flag or CEP44-Flag were subjected to immunoprecipitation (IP) and immunoblotting with anti-Flag and anti-V5 antibodies. (B) Lysates of HEK293T cells co-overexpressing V5-CEP57 and Flag or CEP44-Flag were subjected to immunoprecipitation (IP) and immunoblotting with anti-Flag and anti-V5 antibodies. (C) Immunoblots of CEP44, CEP57L1 and CEP57 in control and CEP44-depleted HeLa cells. Tubulin was used as a loading control. (D) Immunostaining of CEP57L1 (green) or CEP57 (green) and γ-tubulin (red) in control and CEP44-depleted HeLa cells. DNA was stained with DAPI (blue). (E-F) Quantification of the fluorescence intensity of CEP57L1 (E) and CEP57 (F) in HeLa cells from (D). (G-H) Quantification of the fluorescence intensity of CEP44 in CEP57-depleted (G) and CEP57L1-depleted (H) HeLa cells. For E, F, G and H, error bars represent the means ± S.E.M for three independent experiments. n.s., not significant, ***p < 0.001, as determined using one-way ANOVA with Dunnett’s multiple comparisons test. Bars: 5 μm (D).

Depletion of CEP44 results in multipolar spindle formation. (A) Immunoblots of the indicated proteins in wild-type (WT), CEP44-knockout (KO) and CEP44-rescued HeLa cells. Tubulin was used as a loading control. (B) Immunostaining of α-tubulin (green) and γ-tubulin (red) in wild-type (WT) or CEP44 knockout (KO) HeLa cells. DNA was stained with DAPI (blue). (C-D) Quantification of the percentage of cells with a multipolar spindle (C) or a disorganized spindle (D) from (B). (E) Immunostaining of CEP97 (green) and α-tubulin (red) in WT or CEP44 KO HeLa cells. DNA was stained with DAPI (blue). (F) Quantification of the percentage of cells with a multipolar spindle with centriole overduplication from (E). (G) Time-lapse images of mitotic CEP44 KO HeLa cells stably expressing H2B-RFP and GFP-Centrin-3. (H) Quantification of the percentage of pattern 1 and pattern 2 cells in WT and CEP44 KO cells. Pattern 1: premature centriole disengagement in interphase without centriole overduplication; Pattern 2: premature centriole disengagement in interphase with centriole overduplication. For C, D, F and J, error bars represent the means ± S.E.M obtained for three independent experiments. n.s., not significant, **p < 0.01, ***p < 0.001, as determined using one-way ANOVA with Dunnett’s multiple comparisons test. Bars: 5 μm (B, E and G).

Aurora A phosphorylates CEP44 at S324 in the G2/M phase. (A) Immunostaining of α-tubulin (red) and CEP44 (green) in HeLa cells. DNA was stained with DAPI (blue). (B) Immunoblots of CEP44 and Cyclin B1 at the indicated time points after nocodazole release in HeLa cells. GAPDH was used as a loading control. (C) Immunoblots of lysates from HEK293T cells co-overexpressing the indicated proteins with or without λ-ppase treatment. Tubulin was used as a loading control. The red asterisks indicate the shifted band of CEP44. (D) Immunoblots of lysates from HEK293T cells with or without nocodazole and MLN8237 treatment. Tubulin was used as a loading control. The red asterisk indicates the shifted band of CEP44. (E) Immunoblots of lysates from HEK293T cells co-overexpressing the indicated proteins. Tubulin was used as a loading control. (F) The consensus sequence (R-X-pS/T-L/V) of the Aurora A substrate including NuMA, CPAP, TPX2, TACC3 and CEP44. (G) Immunoblots of lysates from HeLa cells with or without nocodazole treatment. (H) Immunoblots of the indicated proteins in control and Aurora A-depleted HeLa cells after nocodazole treatment. (I) Immunostaining of Aurora A (red) and CEP44 (green) in control and Aurora A-depleted mitotic HeLa cells. DNA was stained with DAPI (blue). (J) Quantification of the fluorescence intensity of CEP44 at spindle microtubules in mitotic cells from (I). For J, error bars represent the mean ± S.E.M. for three independent experiments. ***p < 0.001, as determined using one-way ANOVA with Dunnett’s multiple comparisons test. Bars: 5 μm (A and I).

Phosphorylation of CEP44 by Aurora A maintains spindle integrity. (A) Immunostaining of γ-tubulin (red) and α-tubulin (green) in wild-type (WT) and CEP44 knockout (KO) HeLa cells. DNA was stained with DAPI (blue). (B) Quantification of fluorescence microtubule intensity in mitotic cells from (A). (C) Immunostaining of α-tubulin (red) and Flag (green) in mitotic HeLa cells expressing the indicated proteins. DNA was stained with DAPI (blue). (D-E) Quantification of the microtubule intensity at spindles (D) and intensity of CEP44 at spindle microtubules (E) in mitotic cells from (C). (F) Immunostaining of Centrin-3 (red) and γ-tubulin (green) in control and CEP44 knockdown HeLa cells. DNA was stained with DAPI (blue). (G) Quantification of the normalized intensity of γ-tubulin at spindle MTs in control and CEP44 knockdown HeLa cells. (H) Time-lapse images of mitotic CEP44 KO HeLa cells stably expressing H2B-RFP and GFP-Centrin-3. Chr, chromosome. (I) Quantification of the percentage of cells with the indicated phenotypes from (H). For B, D, E, G and I, error bars represent the means ± S.E.M for three independent experiments. n.s., not significant, ***p < 0.001, as determined using one-way ANOVA with Dunnett’s multiple comparisons test. Bars: 5 μm (A, C, F and H).

Schematic of the role of CEP44 in the cell cycle. In wild-type HeLa cells, CEP44 is localized at the proximal halves of the centriole lumen. For centriole duplication during the S phase, CEP44 exists in the lumen of the procentrioles to maintain disengagement between the old centrioles and the procentrioles. During mitosis, CEP44 translocates to the spindle due to Aurora A phosphorylation at serine 324 to ensure spindle integrity. After depletion of CEP44, premature centriole disengagement occurs following centriole duplication in the S phase, which licenses centriole overduplication as early as the S phase. The supernumerary centrioles result in multipolar spindle formation during mitosis. In addition, a lack of CEP44 localization at the spindle impairs spindle integrity, leading to an increase in the percentage of disorganized spindles. Notably, after Aurora A depletion or inhibition, unphosphorylated CEP44 fails to translocate to spindle microtubules and accumulates at the spindle pole, leading to impaired spindle integrity.