August 6, 2010;
Aurora kinases and protein phosphatase 1 mediate chromosome congression through regulation of CENP-E.
Opposing roles of Aurora kinases and protein phosphatase 1 (PP1
) during mitosis have long been suggested. Here, we demonstrate that Aurora kinases A and B phosphorylate a conserved residue on the kinetochore
via a motif overlapping this phosphorylation site and binding is disrupted by Aurora phosphorylation. Phosphorylation of CENP-E
by the Auroras is enriched at spindle
poles, disrupting binding of PP1
and reducing CENP-E''s affinity for individual microtubules. This phosphorylation is required for CENP-E
-mediated towing of initially polar chromosomes toward the cell center. Kinetochores on such chromosomes cannot make subsequent stable attachment to spindle
microtubules when dephosphorylation of CENP-E
or rebinding of PP1
is blocked. Thus, an Aurora/PP1
phosphorylation switch modulates CENP-E
motor activity as an essential feature of chromosome congression from poles and localized PP1
delivery by CENP-E
to the outer kinetochore
is necessary for stable microtubule
capture by those chromosomes.
[+] show captions
References [+] :
Figure 1. A Conserved Threonine Close to the Motor Domain of CENP-E Is Phosphorylated by Aurora A and B on the Kinetochores of Unaligned Chromosomes(A) Alignment of CENP-E protein sequences using ClustalW algorithm. α-helical coiled-coil in the CENP-E neck were predicted for human CENP-E using Paircoil (Berger et al., 1995). Heptad repeat positions (a and d) in the coiled-coil are indicated.(B) In vitro kinase assays using Aurora B/INCENP to phosphorylate Xenopus full-length CENP-E or CENP-E1-473 with or without T424A mutation. Coomassie staining of purified proteins and autoradiogram showing incorporation of γ-32P ATP.(C) Coomassie staining and immunoblot for pT422 using human CENP-E1-429 incubated with wild-type (WT) or kinase dead (KD) Aurora B.(D) CENP-E immunoprecipitates from nocodazole-treated DLD-1 cells expressing either WT or T422A MycGFP-CENP-E were blotted with the pT422 antibody. Half the WT immunoprecipitate was treated with λ-phosphatase.(E) Nocodazole-arrested DLD-1 cells expressing MycGFP-CENP-E were treated with Aurora kinase inhibitors and MG132 and blotted for P-TACC3 (Aurora A substrate), P-Histone H3 (Aurora B substrate), and tubulin (loading control). CENP-E immunoprecipitates using a Myc antibody were blotted with the pT422 antibody.(F) PtK2 cells were stained for DNA (Blue), Tubulin (Green), ACA, and pT422 (Red).(G) Nocodazole-arrested HeLa cells were treated with VX-680 and MG132 and stained for CENP-E (Green), pT422 (Red) and DNA (Blue).(H) pT422 fluorescence intensity was normalized to the total CENP-E fluorescence. Plots show the mean of > 15 cells per condition from two independent experiments. ***p < 0.0001 by t test. Error bars represent SEM. See also Figure S1 and Figure S2.
Figure 2. Phosphorylation of T424 of Xenopus CENP-E by Aurora Kinase Reduces CENP-E's Affinity for Microtubules and Reduces Its Run Length In Vitro(A) ATPase rates of Xenopus CENP-E1-473 and phosphorylated CENP-E1-473 measured with increasing concentrations of microtubules. Plots show the mean of three independent experiments and error bars represent SEM. kcat and KmMT values are represented ± SE.(B) Equilibrium binding of WT or T424A CENP-E1-473 (incubated with or without Aurora A) to microtubules in the ADP state. S, supernatant; P, pellet. Percent of CENP-E1-473 bound is shown below (n = 3; Error bars represent SD).(C) Kymographs showing diffusive motion of CENP-E1-473-RFP preincubated with or without Aurora A in 3 mM MgADP at 33°C.(D) Duration of binding (t) was distributed exponentially, and the mean binding time (tmean) was determined by fitting the data into a cumulative distribution function (exp [-t/tmean]). Inset shows 1-cumulative probability of CENP-E1-473-RFP binding time plotted on a log scale. The tmean is 17 ± 0.13 s for CENP-E1-473-RFP, n = 231 and 12 ± 0.07 s for CENP-E1-473-RFP plus Aurora A, n = 240. Values represented ± SE of the curve fit. Probability < 0.0001 that the duration of binding for CENP-E1-473-RFP plus Aurora A was distributed the same as the duration of binding for CENP-E1-473-RFP (by Kolmogorov-Smirnov Test).(E) Kymographs showing processive motion of CENP-E1-473-RFP in the presence of 3 mM MgATP at 33°C. See Movie S1.(F) Run length of CENP-E1-473-RFP was determined by fitting the data into a cumulative distribution function. Inset shows 1-cumulative probability of CENP-E1-473-RFP run length plotted on a log scale. Mean run length is 1.6 ± 0.02 μm for CENP-E1-473-RFP, n = 337 and 1.2 ± 0.02 μm for CENP-E1-473-RFP plus Aurora A, n = 294. See Movie S1. Probability < 0.05 that the run lengths for CENP-E1-473-RFP plus Aurora A were distributed the same as the CENP-E1-473-RFP run lengths (by Kolmogorov-Smirnov Test). Values represent ± SE of the curve fit.(G) Accumulated CENP-E (green) at the end of the microtubules (red). See also Figure S3.
Figure 3. Phosphorylation of CENP-E at T422 Is Required for Chromosome Alignment(A) Diagram of MycLAP-CENP-E transgene with previously identified phosphorylation sites. Splice variations exist within the CENP-E coiled-coil domain and phosphorylation sites are numbered with respect to their position in the CENP-E clone used in this study.(B) Schematic of replacing endogenous CENP-E with a siRNA resistant transgene.(C) Immunoblot showing depletion of CENP-E by siRNA.(D) Immunoblot showing knockdown with control (GAPDH) or CENP-E siRNA in cells expressing various MycLAP-CENP-E transgenes.(E) Box and whisker plots showing the time spent in mitosis for cells expressing MycLAP-CENP-E following transfection of control (GAPDH, green) or CENP-E (red) siRNA. >90 cells per condition are plotted from at least three independent experiments.(F) Immunofluorescence images of cells in which endogenous CENP-E has been replaced with WT or T422A MycGFP-CENP-E. GFP (green); Tubulin (red). See also Figure S4, Figure S5, Figure S6, Movie S2, and Movie S3.
Figure 4. CENP-E Phosphorylation by Aurora Is Required for Congression of Polar Chromosomes(A) Drug treatment scheme to enrich mono-oriented chromosomes near spindle poles.(B) Diagram illustrating attachment status of chromosomes following treatment with drug regime in (A).(C and D) Immunofluorescence images of drug-treated cells in which endogenous CENP-E was replaced with either WT or T422A MycLAP-CENP-E. KNL-1 (green); DNA (red).(E) Percentage of cells from (C) and (D) with polar chromosomes.(F) Number of polar chromosomes in cells from (C) and (D) displaying chromosome misalignment. Bars represent the mean of four independent experiments. ***p < 0.0001 by t test. Error bars represent SEM. See also Figure S7 and Movie S4.
Figure 6. Dephosphorylation of CENP-E T422 by PP1 Is Required for the Stable Biorientation of Chromosomes Congressed from a Spindle Pole by CENP-E(A) Phosphorylated Xenopus CENP-E1-473 (1 μM) was preincubated with rabbit IgG or the pT422 antibody (3 μM) and subsequently mixed with PP1γ (0.2 μM). Phosphorylation of T422 was determined using rhodamine-labeled pT422 antibody and visualized using the rhodamine fluorophore. Anti-Myc immunoblot shows the CENP-E1-473 loading.(B) Reconstructed Z-sections of rhodamine-labeled pT422 antibody-injected live HeLa cells expressing Histone H2B-YFP. Histone H2B-YFP (purple); pT422-Rhod antibody (green).(C) Time-lapse images of antibody-microinjected (green) HeLa cells stably expressing Histone H2B-YFP (purple). Numbered arrows track the movement of four individual chromosomes which congress to the equator, but fail to stably biorient. Stills are taken from Movie S5 and Movie S6.(D) Box and whisker plots showing the time spent in mitosis for microinjected cells. Uninjected cells in the same field of view were also examined. For each antibody, >120 microinjected cells were observed from two independent experiments. (E) Bar graphs showing the percentage of abnormal mitosis for the cells observed in (D).
Akiyoshi, Quantitative proteomic analysis of purified yeast kinetochores identifies a PP1 regulatory subunit. 2009, Pubmed