Cross-regulation between Aurora B and Citron kinase controls midbody architecture in cytokinesis

Abstract

Cytokinesis culminates in the final separation, or abscission, of the two daughter cells at the end of cell division. Abscission relies on an organelle, the midbody, which forms at the intercellular bridge and is composed of various proteins arranged in a precise stereotypic pattern. The molecular mechanisms controlling midbody organization and function, however, are obscure. Here we show that proper midbody architecture requires cross-regulation between two cell division kinases, Citron kinase (CIT-K) and Aurora B, the kinase component of the chromosomal passenger complex (CPC). CIT-K interacts directly with three CPC components and is required for proper midbody architecture and the orderly arrangement of midbody proteins, including the CPC. In addition, we show that CIT-K promotes Aurora B activity through phosphorylation of the INCENP CPC subunit at the TSS motif. In turn, Aurora B controls CIT-K localization and association with its central spindle partners through phosphorylation of CIT-K's coiled coil domain. Our results identify, for the first time, a cross-regulatory mechanism between two kinases during cytokinesis, which is crucial for establishing the stereotyped organization of midbody proteins.

Keywords: Aurora B; Citron kinase; cell division; midbody.

Figure 1.

CIT-K is required for midbody architecture. EM images of midbodies in HeLa Kyoto cells treated with an siRNA directed against either a random sequence (control) or CIT-K for 48 h. The bottom panels are magnification of the insets as indicated by dotted lines. The arrowhead marks a membrane bleb and the arrows indicate gaps between the midbody matrix and the cortex. MM, midbody matrix. Scale bars, 1 µm.

Figure 2.

CIT-K associates with the CPC in vivo. (a) Proportional Venn diagram showing the overlap between the proteins pulled down by the Flag tag alone (284) and the proteins that co-purified with Flag::CIT-K (449). Note that 350 proteins specifically associated with Flag::CIT-K. (b) Partial list of the CIT-K specific interactors identified by affinity purification. A full list can be found in the electronic supplementary material, table S1, worksheet 1. The bait, CIT-K, is highlighted in red and previously known CIT-K binding partners are in black. The three subunits of the CPC are in blue and the kinesin responsible for the translocation of the CPC to the spindle midzone, KIF20A, is in green. (c) HeLa Kyoto cells stably expressing either Flag-tagged CIT-K or Flag alone were synchronized in telophase by thymidine/nocodazole block and release and then protein extracts were used in a pull-down assay using anti-Flag antibodies. The extracts and pull downs were analysed by western blot to detect KIF14, KIF23, Aurora B, KIF20A, PRC1 and Flag. Note that a slower migrating KIF23 band was detected in the pull-down compared with whole cell extracts. This is probably a phosphorylated form that we routinely observe in protein extracts from both telophase cells and purified midbodies. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (d) U2OS cells stably expressing Aurora B::Venus were washed in PBS to remove tetracycline (tet-off inducible system) and kept in tetracycline free media for 24 h. Cells were then fixed and stained to detect CIT-K (red), Aurora B::Venus (green) and tubulin (blue). Insets show a 3× magnification of the cleavage furrow region. The profiles of the green and red fluorescent signals, measured at the centre of the cleavage furrow (dotted line), are shown on the right. Scale bar, 10 µm. (e) Midbodies were purified form HeLa Kyoto cells and fixed and stained to detect CIT-K (red) and tubulin (green). Scale bar, 5 µm. (f) Midbodies were purified from HeLa Kyoto cells and fixed and stained to detect CIT-K (green) and Aurora B (red). The horizontal profiles of the green and red fluorescent signals, measured at the centre of the midbody (dotted lines), are shown on the right. Scale bars, 5 µm.

Figure 3.

CIT-K directly binds to CPC components in vitro. (a) Schematic diagrams illustrating the protein domains of CIT-K and of the three CPC components analysed—Aurora B, Borealin and INCENP. The positions of the different CIT-K and INCENP fragments used for the in vitro pull-down assays are also indicated. Note that CIT-K is not drawn to the same scale as the CPC components. CC1 and CC2 indicate the fragments encompassing the first and second coiled coil regions; C1, cysteine-rich motif; PH, Pleckstrin homology domain; CNH, Citron-Nik1 homology domain. (b) The GST::CIT-K proteins indicated at the top and GST alone were incubated with the in vitro translated and radiolabelled Aurora B, Borealin and INCENP polypeptides indicated at the right, and then pulled down using glutathione beads. The Ponceau S staining of the protein loading is shown at the bottom and the numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (c) Summary of the results from the in vitro pull-down assays using GST-CIT-K fragments and in vitro transcribed and translated (IVTT) radiolabelled CPC polypeptides. The increasing salt concentrations used in the washes are colour-coded and indicated at the top right. The autoradiogram shown in (b) corresponds to the 1 M NaCl experiment highlighted in blue. (d) The GST-tagged CPC proteins indicated at the top and GST alone were incubated with the in vitro translated and radiolabelled CIT-K polypeptides indicated at the right, and then pulled down using glutathione beads. The protein loading is shown at the bottom and the numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (e) Ribbon diagram of the predicted structure of the CIT-K C-terminal CNH domain (residues 1634–1948), which adopts a seven-bladed β-propeller fold. Ribbon is coloured from blue (N-terminus) to red (C-terminus) and the numbers indicate the seven blades.

Figure 4.

CIT-K is required for correct CPC localization and the orderly arrangement of midbody proteins. (a) HeLa Kyoto cells were treated with siRNAs directed against either a random sequence (control) or CIT-K and after 48 h were fixed and stained to detect DNA (blue), tubulin (green) and the CPC component INCENP (red). The shape and thickness of microtubule bundles at the intercellular bridge were used as criteria to stage telophase cells. Insets show a 3× magnification of the midbody. Scale bars, 10 µm. (b) Quantification of INCENP localization defects from the experiment shown in (a). More than 100 mid–late-telophase cells were counted in each experiment, n = 3. Scale bars indicate standard errors. (c) HeLa Kyoto cells stably expressing Flag-tagged wild-type CIT-K, a kinase-dead version (KD) of CIT-K, or Flag alone were treated with siRNAs directed against either a random sequence (control) or the 3′-UTR of CIT-K and after 48 h were fixed and stained to detect tubulin (blue), Flag (green) and the CPC component INCENP (red). The shape and thickness of microtubule bundles at the intercellular bridge were used as criteria to stage telophase cells. Insets show a 3× magnification of the midbody. The horizontal profiles of the red fluorescent signals, measured at the centre of the midbody (dotted lines), are shown at the right of the respective images. Scale bars, 10 µm. (d) Quantification of INCENP localization defects from the experiment shown in (c). Only Flag-positive cells were counted, and more than 100 mid–late-telophase cells were counted in each experiment, n = 3. Bars indicate standard errors. (e) HeLa Kyoto cells were treated with an siRNA directed against either a random sequence (control) or CIT-K and after 48 h were fixed and stained to detect DNA (blue), Aurora B (green) and KIF23 (red). The shape and thickness of microtubule bundles at the intercellular bridge were used as criteria to stage telophase cells. Insets show a 3× magnification of the midbody. Scale bars, 10 µm.

Figure 5.

CIT-K phosphorylates INCENP. (a) Schematic diagram of INCENP structure illustrating the phosphorylated sites identified by MS. The GST-tagged fragments used for the in vitro phosphorylation assays shown in (b), (c) and (d) are depicted at the bottom. (b) GST-tagged INCENP polypeptides, GST alone, and the positive control MBP (myelin basic protein) were incubated with GST-tagged CIT-K kinase domain or KD-CIT-K kinase domain in the presence of [γ-³²P] ATP. The reactions were then separated by SDS-PAGE, transferred onto nitrocellulose membranes, and exposed at −80°C. The Ponceau S staining of the protein loading is shown at the bottom. The asterisks mark the molecular positioning of the respective proteins. The dagger (†) indicates CIT-K auto-phosphorylation. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (c) GST alone, GST-tagged INCENP 783–918 and GST-tagged INCENP mutants (T844A, TSS/AAA and T844A + TSS/AAA), were incubated with GST-tagged CIT-K kinase domain or KD-CIT-K kinase domain in the presence of [γ-³²P] ATP. The reactions were then separated by SDS-PAGE, transferred onto nitrocellulose membranes, and exposed at −80°C. The protein loading is shown at the bottom. An asterisk marks CIT-K auto-phosphorylation. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (d) GST alone and GST-tagged INCENP 783–918 were incubated in the presence or absence of GST-tagged CIT-K kinase domain or KD-CIT-K kinase domain, using non-radioactive ATP. The reactions were then separated by SDS-PAGE and analysed by western blot to detect phosphorylated INCENP. The protein loading is shown at the bottom. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (e) HeLa Kyoto cells were treated with siRNAs directed against either a random sequence (control) or CIT-K for 48 h. During RNAi incubation, cells were synchronized using 2 mM thymidine for 19 h, released for 5 h, treated with 10 µM RO3306 for 13 h, released for 2 h, fixed and stained to detect phosphorylated INCENP (green), tubulin (red) and DNA (blue). Insets show 2× magnification of the midbody. The box plot showing the quantification of pTSS fluorescence levels at the midbody is shown on the right. The intensity of pTSS INCENP fluorescence at the midbody was calculated using the formula shown, where the mean fluorescence intensity was measured at the midbody (I_M) and the mean background fluorescence intensity was measured within an identical area inside the cytoplasm (I_C). The numbers of cells counted are detailed below each plot. Scale bars, 10 µm. **p < 0.01 (Student's t-test). (f) HeLa Kyoto cells were treated as in (e), and stained to detect phosphorylated Aurora B (green), tubulin (red) and DNA (blue). Insets show 2× magnification of the midbody. The box plot showing the quantification of pT232 fluorescence levels at the midbody, calculated as described in (e), is shown on the right. The numbers of cells counted are detailed below each plot. Scale bars, 10 µm. ***p < 0.001 (Student's t-test).

Figure 6.

Aurora B activity is necessary for proper CIT-K localization. (a) Asynchronous HeLa Kyoto cells were treated for 20 min with the Aurora B inhibitor ZM447439 at a final concentration of 5 µM and then fixed and stained to detect CIT-K (red), tubulin (green) and DNA (blue). Scale bars, 10 µm. (b) HeLa Kyoto cells were treated with siRNAs directed against either a random sequence (control) or KIF20A for 48 h and then proteins were extracted, separated by SDS-PAGE, transferred to a membrane, and incubated to detect KIF20A and tubulin (loading control). (c) HeLa Kyoto cells were treated with siRNAs directed against either a random sequence (control) or KIF20A and after 48 h were fixed and stained to detect CIT-K (green), INCENP (red) and tubulin (blue). The shape and thickness of microtubule bundles at the intercellular bridge were used as criteria to stage telophase cells. Insets show a 3× magnification of the midbody. Scale bars, 10 µm. (d) Quantification of CIT-K localization defects from the experiment shown in (c). More than 100 late-telophase cells were counted in each experiment, n = 3. Bars indicate standard errors. (e) EM images of midbodies from HeLa Kyoto cells treated with an siRNA directed against either a random sequence (control) or KIF20A for 48 h. The arrowheads mark the midbody matrix (MM) in KIF20A depleted cells. Scale bars, 1 µm. (f) HeLa Kyoto cells were treated with an siRNA directed against either a random sequence (control) or KIF20A and after 48 h were fixed and stained to detect DNA (blue), tubulin (green) and KIF23 (red). The arrowheads mark the dark zone. Insets show a 3× magnification of the midbody. Scale bars, 10 µm. (g) HeLa Kyoto cells carrying a doxycycline-inducible GFP-tagged PRC1-Baronase transgene were treated with siRNAs directed against either a random sequence (control) or KIF20A and after 24 h incubated in 2 mM thymidine for a further 20 h. Cells were washed and incubated with or without 1 µg/ml doxycycline for 10 h and then fixed and stained to detect CIT-K (red), PRC1-Baronase (green) and tubulin (blue). Scale bars, 10 µm. (h) Quantification of CIT-K localization defects from the experiments shown in (f). More than 100 late-telophase cells were counted in each experiment, n = 3. Bars indicate standard errors.

Figure 7.

Aurora B phosphorylates CIT-K. (a) Schematic diagram of CIT-K structure illustrating the phosphorylated sites identified by MS. The GST- tagged fragments used for the in vitro phosphorylation assays shown in (b), (c) and (d) are depicted at the bottom. (b) GST-tagged CIT-K polypeptides, GST alone and the positive control MBP were incubated with (+) or without (−) recombinant Aurora B in the presence of [γ-³²P] ATP. The reactions were then separated by SDS-PAGE, transferred onto nitrocellulose membranes and exposed at −80°C. The Ponceau S staining of the protein loading is shown at the bottom. Aurora B auto-phosphorylation is marked by an asterisk. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (c) GST-tagged wild-type CIT-K-CC1 (WT) and S699A mutant polypeptides, GST alone and the positive control MBP (myelin basic protein) were incubated with (+) or without (−) recombinant Aurora B in the presence of [γ-³²P] ATP. The reactions were then separated by SDS-PAGE, transferred onto nitrocellulose membranes and exposed at −80°C. The protein loading is shown at the bottom. Aurora B auto-phosphorylation is marked by an asterisk. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (d) The GST-tagged wild-type CIT-K-C1+PH peptide (WT), along with the S1385A-S1386A-T1387A (TripleA) and S1474A mutant polypeptides, GST alone and the positive control MBP (myelin basic protein) were incubated with (+) or without (−) recombinant Aurora B in the presence of [γ-³²P] ATP. The reactions were then separated by SDS-PAGE, transferred onto nitrocellulose membranes and exposed at −80°C. The protein loading is shown at the bottom. Aurora B auto-phosphorylation is marked by an asterisk. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (e) HeLa Kyoto cells stably expressing Flag::CIT-K or Flag::CIT-K-S699A were treated with an siRNA directed against the CIT-K 3′-UTR for 48 h, blocked in metaphase by thymidine/nocodazole block, released for 90 min and then treated with 10 µM RO3306 for further 15 min. Proteins were extracted and used in a pull-down assay with anti-Flag antibodies. The extracts and pull downs were analysed by western blot to detect KIF14, KIF23, Aurora B and Flag::CIT-K. The numbers on the left indicate the sizes in kilodaltons of the molecular mass marker. (f) HeLa Kyoto cells stably expressing Flag::CIT-K or Flag::CIT-K-S699A were treated with siRNAs directed against either a random sequence (control) or 3′-UTR CIT-K for 48 h. During RNAi incubation, cells were synchronized using 2 mM thymidine for 19 h, released for 5 h, treated with 10 µM RO3306 for 13 h, released for 2 h, fixed and stained to detect Flag (red), tubulin (green) and DNA (blue). All images are maximum intensity projections of the three most central z sections; z step = 0.25 µm. Scale bars, 10 µm. (g) Quantification of CIT-K midzone localization from the experiments showed in (f). No less than 50 early–mid-telophase cells were counted in each experiment, n = 4. Bars indicate standard errors.

Figure 8.

Cartoon illustrating the cross-regulation between Aurora B and CIT-K in midbody assembly. The red arrows indicate the phosphorylation of the CIT-K CC1 and C1 domains by Aurora B and the phosphorylation of the INCENP TSS motif by CIT-K. In the presence of CIT-K (a), midbody components including the CPC, centralspindlin, KIF20A and KIF14 are properly aligned along the midbody. By contrast, after depletion of CIT-K (b) KIF14 is not recruited to the midbody [12,14], the CPC and KIF20A are dispersed and centralspindlin is mis-positioned on one side of the midbody.