Regulated targeting of protein phosphatase 1 to the outer kinetochore by KNL1 opposes Aurora B kinase

Abstract

Regulated interactions between kinetochores and spindle microtubules are essential to maintain genomic stability during chromosome segregation. The Aurora B kinase phosphorylates kinetochore substrates to destabilize kinetochore-microtubule interactions and eliminate incorrect attachments. These substrates must be dephosphorylated to stabilize correct attachments, but how opposing kinase and phosphatase activities are coordinated at the kinetochore is unknown. Here, we demonstrate that a conserved motif in the kinetochore protein KNL1 directly interacts with and targets protein phosphatase 1 (PP1) to the outer kinetochore. PP1 recruitment by KNL1 is required to dephosphorylate Aurora B substrates at kinetochores and stabilize microtubule attachments. PP1 levels at kinetochores are regulated and inversely proportional to local Aurora B activity. Indeed, we demonstrate that phosphorylation of KNL1 by Aurora B disrupts the KNL1-PP1 interaction. In total, our results support a positive feedback mechanism by which Aurora B activity at kinetochores not only targets substrates directly, but also prevents localization of the opposing phosphatase.

Figure 1.

KNL1 recruits PP1γ to kinetochores. (A) Sequence alignment of C. elegans (ce), Caenorhabditis briggsae (cb), Schizosaccharomyces pombe (sp), Saccharomyces cerevisiae (sc), D. melanogaster (dm), Ciona intestinalis (ci), Gallus galls (gg), and human (h) KNL1 homologues showing conservation of N-terminal [S/G]ILK and RVSF motifs. Arrows indicate Aurora B phosphorylation sites (Welburn et al., 2010). Colors indicate conserved amino acids. (B) Binding of KNL1 to PP1γ depends on the RVSF motif. A Coomassie-stained gel shows binding of PP1γ to Ni-NTA agarose resin alone, or resin bound to either His-tagged hKNL1^1–86 wild type or mutants for the conserved PP1-binding motifs. (C) HeLa cells stably expressing GFP^LAP-PP1γ were fixed and stained for KNL1 and Hec1. PP1γ fails to localize to kinetochores after depletion of KNL1 by siRNA. (D) HeLa cells were transfected with KNL1 siRNA and either siRNA-resistant wild-type (wt) KNL1 or the KNL1^RVSF/AAAA mutant, together with GFP-PP1γ. The KNL1^RVSF/AAAA mutant fails to restore kinetochore localization of PP1. (E) For cells treated as in D, the intensity of GFP-PP1γ at kinetochores was calculated relative to control cells for each of the conditions indicated. Each bar represents a mean (± SEM) over multiple cells (n ≥ 5), with ≥40 kinetochores analyzed per cell. Images in C and D are maximum intensity projections of confocal stacks; insets show enlarged views (indicated by the boxed regions) of optical sections showing individual kinetochores. Intensity scaling is consistent between all insets, but insets are scaled differently from the original images to show individual kinetochores more clearly. au, arbitrary units. Bars, 5 µm.

Figure 2.

The KNL1–PP1 interaction is required for cell viability. A tetracycline-repressible ggKNL1 chicken DT40 cell line (Cheeseman et al., 2008) was rescued by constitutive expression of a ggKNL1 mutant with the RVSF motif mutated to alanines. (A) PP1 fails to localize to kinetochores in the KNL1^RVSF/AAAA mutant cells. Images show PP1γ-GFP localization in wild type or KNL1^RVSF/AAAA mutants. (B) KNL1^RVSF/AAAA mutants do not generally affect outer kinetochore assembly. Immunofluorescence images show Hec1 localization in wild type or KNL1^RVSF/AAAA mutants. (C) KNL1^RVSF/AAAA mutant cells are inviable. To determine viability, cells were counted after expression of the wild-type protein was inactivated at t = 0 by tet addition. Cell numbers are averages over multiple experiments (n = 2 for wild type, n = 3 for KNL1^RVSF/AAAA). Bars, 10 µm.

Figure 3.

PP1 opposes Aurora B activity at kinetochores. HeLa cells expressing a kinetochore-targeted Aurora B phosphorylation sensor were treated as indicated and imaged live. (A) Images of YFP emission in control and KNL1-depleted cells are maximum intensity projections of confocal stacks; insets show enlarged views (indicated by boxed regions) of optical sections showing individual kinetochores. Bar, 5 µm. (B) Interkinetochore distances were measured for the indicated conditions. Aligned and misaligned kinetochores were analyzed separately. Centromeres aligned at the metaphase plate are under full tension in KNL1-depleted cells. Each bar represents a mean (± SEM) of >70 kinetochores from ≥12 cells. (C) The YFP/TFP emission ratio was analyzed to measure phosphorylation changes at kinetochores; an increased YFP/TFP emission ratio indicates dephosphorylation. Aligned kinetochores were analyzed in KNL1-depleted cells, either without exogenous KNL1 or expressing siRNA-resistant wild-type KNL1 or KNL1^RVSF/AAAA. Depletion of KNL1 increases phosphorylation on aligned kinetochores, and dephosphorylation is restored by wild-type KNL1 but not by KNL1^RVSF/AAAA. (D) To target exogenous PP1 to kinetochores in KNL1-depleted cells, cells were transfected with CENP-B–PP1γ-mCherry and analyzed as in C. ZM indicates the Aurora B inhibitor ZM447439. Expression of CENP-B–PP1γ-mCherry restores dephosphorylation at aligned kinetochores to levels similar to control cells, but untargeted PP1γ-mCherry does not. Each bar in C and D represents a mean (± SEM) over ≥8 cells, with ≥30 kinetochores analyzed per cell for aligned kinetochores, or a total of at least 60 kinetochores for misaligned kinetochores.

Figure 4.

PP1 at kinetochores stabilizes microtubule attachments. HeLa cells were treated as indicated, fixed, and analyzed for cold-stable microtubules. (A–E) Cells were transfected with KNL1 siRNA and either CENP-B–PP1γ-mCherry to target exogenous PP1 to kinetochores, or CENP-B–mCherry as a control. An untransfected control cell (A) was analyzed the same way and stained for Hec1 to label kinetochores. Images (A–C) are maximum intensity projections of confocal stacks; insets show enlarged views (indicated by the numbered boxed regions) of optical sections showing individual kinetochores. Bars, 5 µm. (D) Cells were classified by the number of aligned kinetochores lacking cold-stable microtubule fibers (n ≥ 45 cells in each group). (E) The intensity of microtubule plus ends at attached kinetochores (n > 60 from multiple cells) was measured in each condition. (F) Cells were transfected with KNL1 siRNA and either siRNA-resistant wild-type KNL1 or the KNL1^RVSF/AAAA mutant (n ≥ 25 cells in each group), and analyzed as in D. Note that the binning in D and F is different because the effect of mutating the RVSF motif is more subtle than the effect of depleting KNL1.

Figure 5.

PP1γ is recruited to kinetochores and dephosphorylates an Aurora B substrate as centromere tension is established. (A–C) HeLa cells transfected with a kinetochore-targeted Aurora B phosphorylation sensor, with or without KNL1 siRNA, were imaged live during recovery from nocodazole (30 ng/ml). Cells were followed for 33 min, which was sufficient time for control cells to reach metaphase. Cells rarely entered anaphase during this time, and any anaphase cells were excluded from the analysis. Representative images (A) show YFP emission. At each time point, the percentage of kinetochores aligned at the metaphase plate was determined (B) and the YFP/TFP emission ratio was calculated (C). Each data point represents seven cells, >15 kinetochores per cell. (D) Images of HeLa cells stably expressing GFP^LAP-PP1γ in early prometaphase, metaphase, or treated with nocodazole. (E) The relative intensity of GFP-PP1γ at kinetochores was calculated during recovery from nocodazole (30 ng/ml). n = 6 cells, multiple kinetochores per cell. Bars, 5 µm.

Figure 6.

Phosphorylation of KNL1 by Aurora B disrupts the interaction with PP1γ. (A) A Coomassie-stained gel shows binding of PP1γ to Ni-NTA agarose resin bound to either His-tagged hKNL1^1–86 wild type, or phosphomimetic or phosphoinhibitory mutants in the SILK and RVSF motifs. (B) Binding was quantified from Coomassie-stained gels in three independent experiments. (C) Hela cells expressing a kinetochore-targeted Aurora B phosphorylation sensor were either treated with nocodazole to release interkinetochore tension, or transfected with Mis12-INCENP-mCherry to redistribute Aurora B to the kinetochore. ZM indicates the Aurora B inhibitor ZM447439. Cells were imaged live, and the YFP/TFP emission ratio was calculated to measure phosphorylation changes at kinetochores. The decrease in YFP/TFP emission ratio indicates increased phosphorylation at aligned kinetochores in cells expressing Mis12-INCENP-mCherry. Each bar represents a mean of ≥10 cells, with ≥30 kinetochores analyzed per cell. (D) HeLa cells stably expressing GFP^LAP-PP1γ were transfected with (bottom) or without (top) Mis12-INCENP-mCherry to increase Aurora B activity at kinetochores. Insets show kinetochores at higher magnification (boxed regions). Bar, 5 µm. (E) Cells were transfected with Mis12-INCENP or KNL1 siRNA, or untransfected cells were analyzed at metaphase or at prometaphase, or treated with nocodazole. The relative intensity of GFP-PP1γ at kinetochores was calculated (n ≥ 9 cells, ≥40 kinetochores per cell) in each case. (F) Model comparing centromeres in low- and high-tension states. In the low-tension state, Aurora B phosphorylates KNL1 and Hec1, which leads to reduced binding of both PP1 and microtubules to kinetochores. In the high-tension state, Aurora B is spatially separated from kinetochore substrates, so KNL1 and Hec1 are dephosphorylated and the binding of PP1 and microtubules is increased.