Table 1.
PCR primers for GST-human NudC truncation constructs.
Fig 1.
NudC co-localizes with Aurora B in mitosis.
(A) Unperturbed mitotic HeLa cells were stained for NudC (green), Aurora B (red) and counterstained with DAPI (blue). Bar, 10 μm. (B) HeLa cells were transfected with Myc-NudC and FLAG-Aurora B (left) or EGFP-NudC and FLAG-Aurora B (right) for 24 h. Cell lysates (1 mg in 250 μl) were immunoprecipitated with anti-Myc antibody and blotted for Aurora B followed by reblotting for NudC (left). A reciprocal immunoprecipitation was performed, in which cell lysates (500 μg in 250 μl, 1 mg in 250 μl or 2 mg in 500 μl) were immunoprecipitated with anti-FLAG antibody followed by blotting for NudC and reblotting for Aurora B (right). Immunoprecipitation with either anti-Myc or anti-FLAG antibody using non-transfected cell lysates was used as a negative control. β-tubulin was used as a loading control. Input, 20 μg total cell lysates. Data are representative of n = 5 independent experiments.
Fig 2.
NudC interaction with Aurora B in mitosis.
(A) HeLa cells were synchronized by a double thymidine block and release protocol as indicated. “P” (prometaphase and metaphase) and “A” (anaphase, telophase and cytokinesis) lysates were prepared from early versus late mitotic cells. Synchronization efficiency was confirmed by a cyclin B1 western blot. α-tubulin was used as a loading control. (B) Lysates from asynchronously cycling (Asy), P or A cells were incubated with GST-NudC fusion protein in GST pulldown assays. GST-NudC bound proteins were immunoblotted for Aurora B. GST binding to lysates from either Asy (this experiment), P or A cells (not shown), served as a negative control. Ponceau S staining showed equal GST-NudC fusion protein used in the pulldown assay. Aurora B binding was quantified as Aurora B signal/input Aurora B normalized against the P sample (mean ± s.e.m.) from 3 independent experiments. *, p < 0.05. (C) Lysates (2 mg in 500 μl) from Asy, P or A cells prepared as in (A) were immunoprecipitated with G1 goat NudC antibody, blotted for Aurora B and reblotted for NudC using 2D9 monoclonal antibody. Asy lysates were also immunoprecipitated with preimmune goat serum (IgG) as a negative control. β-tubulin was used as a loading control. (D) An immunoprecipitation using a different batch of A cell lysate (500 μg in 250 μl) was performed as in (C).
Fig 3.
NudC is phosphorylated by Aurora B in vitro and in vivo.
(A) HeLa cells were transfected with FLAG-Aurora B wild type (WT) or a kinase dead (K106R) mutant Aurora B for 24 h. Aurora B was immunoprecipitated using anti-FLAG antibody and used in IP kinase assays. Substrates used were GST-NudC (lanes 4–6), histone H3 (lanes 1–3) as a positive control, and GST (lane 7) as a negative control. Aurora B WT was also incubated with 2 μM of ZM447439 as a specificity control (lanes 3 and 6). Samples were transferred to a filter, stained by Ponceau S (lower panel) and analyzed by autoradiography (upper panel). *, degradation product. Data are reproducible in 3 independent experiments. (B) HeLa cells were synchronized by an overnight incubation with 100 ng/ml nocodazole (M, mitotic) as indicated. Cells (1 X 106) were labeled with 32P orthophosphate for 4 h in the presence or absence of 2 μM ZM447439 (ZM). Cell lysates (300 μg at 1 mg/ml) were immunoprecipitated for NudC, transferred to a filter, analyzed by autoradiography, and immunoblotted for NudC. 32P-NudC was quantified as 32P-NudC/total immunoprecipitated NudC and normalized against NudC signals in asynchronously cycling (Asy) cells.
Fig 4.
NudC is phosphorylated by Aurora B on T40.
(A) A series of GST-NudC truncations were constructed based on functional domains in human NudC. N1 –N4, NudC truncations that retain the N terminal 49 amino acids (a.a.) but contain various deletions from the C terminus. C1 –C4, NudC truncations that retain most or the entire C terminal nuclear movement domain but contain various deletions from the N terminus. Numbers within brackets refer to amino acid residues in the human NudC protein. CC, coiled-coiled; AR, acidic rich; p23-like CHORD-Sgt domain [51]; NMD, conserved nuclear movement domain. (B) GST-NudC full-length (FL), N- and C- terminal truncation series depicted in (A) were used in Aurora B IP kinase assays. Reactions were transferred to filters, analyzed by autoradiography, blotted for Aurora B, and stained by Ponceau S. Substrates used were GST-NudC (lanes 2–9), histone H3 (lane 1) as a positive control and GST (lane 10) as a negative control. Arrowheads, 32P-labeled GST-NudC proteins in the autoradiogram corresponding to the GST-NudC proteins in the Ponceau stain. *, degradation product. The levels of 32P-GST-NudC signals (autoradiogram)/total GST-NudC (Ponceau) normalized against that of GST-NudC full-length (set as 1) were quantified (mean ± s.e.m.) from 3 independent experiments, except for GST-NudC-N2 which was obtained from one experiment (data not shown). **, p < 0.001; ***, p < 0.04. (C) GST-NudC-N1 was used in Aurora B IP kinase assays. (D) NudC protein sequences from various species share a high degree of sequence homology surrounding amino acid T40. (E) GST-NudC-N1 wild type (WT) and GST-NudC-N1 containing T40A mutation were used in Aurora B IP kinase assays. GST, negative control. Data in C and E are representative of 3 independent experiments.
Fig 5.
Aurora B localization at the kinetochore is not affected in NudC-deficient cells.
(A) HeLa cells were transfected with siLuc or siNudC oligos for 72 h. NudC knockdown was examined by western blotting for NudC. β-tubulin was used as a loading control. (B) Prometaphase cells treated with siRNAs as in (A) were stained for pS326-NudC (red) and Aurora B (green) (enlarged in inset), or with Spc25 (green), and counterstained with DAPI for DNA (blue). In initial experiments, siGLO was co-transfected as an indicator for siRNA oligo uptake. (C) For quantification, cells treated as in (B) were also co-stained with the CREST autoserum to mark the kinetochores. For Aurora B or Spc25 staining, maximum-intensity projections of deconvolved images were measured using AutoDeblur/AutoVisualize software, and their fluorescence intensities (average ± s.d.) relative to that of CREST staining at the kinetochore were quantified, using 10 randomly chosen kinetochores from at least 10 siLuc or siNudC prometaphase cells. n.s., not significant.
Fig 6.
Aurora B distribution in the midbody region in NudC-deficient cells.
(A) HeLa cells were transfected with either siLuc control or siNudC oligos for 72 h. Pairs of interconnected cells undergoing unperturbed cytokinesis were stained for NudC (red), tubulin (green), and counterstained with DAPI (blue). Midbody is enlarged in control cells. Staining with two NudC antibodies, G1 goat or 2D9, gave similar results. Yellow line depicts intercellular distance. (B) Pairs of interconnected siLuc (i–iv) or siNudC (v–viii) cells undergoing cytokinesis were stained for Aurora B (green) and counterstained with DAPI (blue). Midbodies are enlarged in insets. (C) Aurora B staining in the midbody region in siLuc control and siNudC cells in (B) was measured using Nikon NIS-Elements software, and the lengths in μm (mean ± s.e.m.) are presented using GraphPad Prism software. The numbers of midbodies counted in siLuc (n = 84) and siNudC (n = 121) cells were obtained from 3–5 independent experiments. *, p < 0.01. siLuc and siNudC cells were also stained as follows: (D) Aurora B (red) and pTSS-INCENP (green). (E) Borealin (green). (F) pT232 Aurora B (red) and β-tubulin (green). Midbodies are enlarged in insets. (H) PRC1 (red). (I) MKLP1 (red) and β-tubulin (green). All bars, 10 μm. (G) Intercellular distances between interconnected daughter cells were determined by staining with tubulin, as indicated by the yellow line in (A). The lengths in μm (mean ± s.e.m.) were obtained from siLuc (n = 37) and siNudC (n = 49) cells in n = 2 experiments. *, p < 0.01.
Fig 7.
Dynamic phosphorylation of NudC on T40 by Aurora B regulates cytokinesis.
(A) HeLa cells were transfected with siLuc or siNudC oligos for 48 h. Cells were further transfected with EGFP-NudC WT, T40A or T40D for another 24 h. Cell lysates (5 μg) were blotted for NudC followed by β-tubulin as a loading control. n.s., nonspecific band. (B) Cells prepared as in (A) that demonstrated a connection by an intercellular bridge were analyzed. % cells in cytokinesis (mean ± s.e.m.) was determined from 3–5 independent experiments. n, number of cells counted. Statistical significance was calculated using ANOVA. *, p < 0.05; **, p < 0.01. (C) Model of NudC phosphorylation on T40 by Aurora B on cytokinesis and cell abscission. In NudC knockdown cells, Aurora B is widely distributed at the midbody region. This is correlated with over-extension of microtubules (horizontal arrows) in the elongated intercellular bridge and poor cell abscission. Reconstitution with wild-type (WT) NudC or unphosphorylatable T40A NudC rescues abscission and cytokinesis. Reconstitution with the phospho-mimetic T40D NudC is inefficient in completing abscission and cytokinesis. The model suggests that dynamic phospho-regulation on NudC T40 by Aurora B is important in regulating cell abscission and cytokinesis.–, microtubule minus-ends; +, microtubule plus-ends. Horizontal arrows, microtubule sliding and elongation in the intercellular bridge.