BubR1 is a central component of the spindle assembly checkpoint that inhibits progression into anaphase in response to improper kinetochore–microtubule interactions. In addition, BubR1 also helps stabilize kinetochore–microtubule interactions by counteracting the Aurora B kinase but the mechanism behind this is not clear. Here we show that BubR1 directly binds to the B56 family of protein phosphatase 2A (PP2A) regulatory subunits through a conserved motif that is phosphorylated by cyclin-dependent kinase 1 (Cdk1) and polo-like kinase 1 (Plk1). Two highly conserved hydrophobic residues surrounding the serine 670 Cdk1 phosphorylation site are required for B56 binding. Mutation of these residues prevents the establishment of a proper metaphase plate and delays cells in mitosis. Furthermore, we show that phosphorylation of serines 670 and 676 stimulates the binding of B56 to BubR1 and that BubR1 targets a pool of B56 to kinetochores. Our data suggest that BubR1 counteracts Aurora B kinase activity at improperly attached kinetochores by recruiting B56–PP2A phosphatase complexes.
Accurate segregation of sister chromatids during mitosis depends on their bi-orientation on the mitotic spindle such that the sister chromatids segregate to opposite poles at anaphase. Bi-orientation requires the establishment of proper connections between dynamic microtubules and kinetochores on sister chromatids (Santaguida and Musacchio, 2009; Welburn and Cheeseman, 2008). Kinetochores are able to stably bind to the dynamic plus ends of microtubules and in response to improper binding they activate the spindle assembly checkpoint (SAC) a conserved signaling pathway that delays the onset of anaphase (Musacchio and Salmon, 2007). In addition to their role in the SAC a subset of checkpoint proteins namely the kinases Aurora B, Mps1, Bub1 and BubR1 are required for proper kinetochore–microtubule interactions (Ditchfield et al., 2003; Hewitt et al., 2010; Klebig et al., 2009; Lampson and Kapoor, 2005; Maciejowski et al., 2010; Meraldi and Sorger, 2005; Santaguida et al., 2010; Welburn et al., 2010). Aurora B plays an important role in this process in that it phosphorylates a number of outer kinetochore proteins to destabilize the interaction with microtubules thus ‘resetting’ the kinetochore allowing new proper connections to form (Cheeseman et al., 2006; DeLuca et al., 2006; Welburn et al., 2010). How phosphorylations on outer kinetochore proteins are removed is not clear but both the PP1 and PP2A phosphatases have been implicated. Recently it was shown that PP2A in complex with regulatory subunits of the B56 family (also known as PPP2R5 or B′ subunits) localizes to kinetochores/centromeres of unattached chromosomes and dephosphorylates kinetochore proteins to establish proper kinetochore–microtubule interactions (Foley et al., 2011). However, it is not clear how B56–PP2A gets recruited to the kinetochore.
BubR1 binds to unattached kinetochores and counteracts Aurora B kinetochore phosphorylation by an unknown mechanism (Ditchfield et al., 2003; Lampson and Kapoor, 2005). The central region of BubR1 does not contain any obvious structural motifs but is required for the role of the protein in stabilizing kinetochore–microtubule interactions (Suijkerbuijk et al., 2010). This region of BubR1 is heavily phosphorylated by Cdk1 and Plk1 and the respective phosphorylation of S670 and S676 by these kinases is important for the role of BubR1 in stabilizing kinetochore–microtubule interactions (Elowe et al., 2010; Elowe et al., 2007; Huang et al., 2008).
Here we show that BubR1 binds directly to B56 regulatory subunits of the PP2A phosphatase through a short highly conserved region centered on the S670 phosphorylation site. BubR1 recruits a pool of B56 to the outer kinetochore and preventing their interaction interferes with chromosome segregation. We propose that in response to improperly attached kinetochores Cdk1 and Plk1 stimulates the recruitment of B56–PP2A complexes to the outer kinetochore through BubR1 and, in turn, this counteracts Aurora B kinase activity.
Results and Discussion
BubR1 binds to PP2A regulatory subunits during mitosis
To understand how BubR1 functions during mitosis we used a yeast two-hybrid approach to screen a human placenta cDNA library using full length BubR1 as bait. The most prominent interactors identified were the five members of the B56 family of protein phosphatase PP2A regulatory subunits (Fig. 1A). To determine whether BubR1 interacts with the B56–PP2A complex in vivo we immunopurified endogenous BubR1 from cells arrested in mitosis with nocodazole or from asynchronous growing cells. Although we have not been able to detect any members of the B56 family in BubR1 purifications we consistently enriched the catalytic subunit of PP2A in the mitotic sample (Fig. 1B). Given the evidence presented later we believe that the interaction between BubR1 and PP2AC is mediated by the B56 subunits.
To further validate the interaction we generated a HeLa cell line stably expressing Venus-tagged B56α. The Venus B56α fusion appeared functional in that it co-purified two proteins at almost stoichiometric levels corresponding in size to the PP2A catalytic and PP2A scaffold proteins and localized to kinetochores/centromeres in mitosis (supplementary material Fig. S1A,B). We purified Venus B56α and analyzed its ability to co-purify BubR1 from mitotic or interphase cells. In mitosis we observed a robust binding to BubR1 that was reduced in the interphase sample while binding to the catalytic subunit was constant (Fig. 1C). We also tested binding to Venus tagged B56β and B56γ and compared this to the unrelated B55α subunit. Again we observed efficient co-purification of BubR1 from the mitotic sample and this was more efficient with the B56 subunits than the B55 subunit (supplementary material Fig. S1C).
Combined these results show that BubR1 interacts with the B56–PP2A complex and that this interaction is stronger during mitosis.
B56 subunits bind to a conserved region in BubR1 that is phosphorylated by Plk1 and Cdk1
To determine how B56 subunits interact with BubR1 we made 5 constructs of BubR1 covering different functional domains and analyzed the ability of these to interact with B56α in the yeast two-hybrid system. We detected binding to all fragments containing the region 555–700 (supplementary material Fig. S2A). This region of BubR1 has been reported to be involved in the alignment of chromosomes and formation of stable kinetochore–microtubule interactions and indeed we could confirm this by comparing mitotic progression in RNAi rescue experiments with BubR1 1–483 and BubR1 1–715 (supplementary material Fig. S2B).
To determine if this binding was direct we purified a recombinant fragment of BubR1 encompassing amino acids 516–715 (BubR1 555–700 did not express) and tested whether this could bind recombinant B56α. The two proteins were run individually on a Superdex 200 size exclusion column and fractions analyzed by SDS-PAGE (Fig. 2B). B56α migrated at the expected size while BubR1 516–715 migrated around 90 kDa indicating that this BubR1 fragment is either a multimer or elongated. When the two proteins were mixed prior to loading on the column they co-eluted from the column as a complex of ∼150 kDa showing that B56α can bind directly to BubR1 516–715.
To identify the residues in BubR1 binding to B56α we made a 20mer peptide array covering the residues 555–700 of BubR1 with a three amino acid shift between peptides. Two arrays were treated identically except one was incubated with recombinant B56α protein and then both arrays were incubated with a B56α specific antibody. Three peptides bound strongly to B56α and they covered the region 660–685 (Fig. 2C). To further identify the binding site we generated a second peptide array using a template peptide covering amino acids 663–682 and then performed an alanine scan through the sequence. Incubation with B56α revealed that B56α binding was abolished when L669 or I672 where mutated to alanine while K668 and E674 also contributed to the interaction (Fig. 2D). The identified residues required for B56α to bind BubR1 are strictly conserved and surround the conserved S670 Cdk1 phosphorylation site (Fig. 2E) (Elowe et al., 2010; Huang et al., 2008). To determine if the same residues were required in vivo for binding to B56α we transfected the Venus B56α cell line with mCherry tagged BubR1 WT, BubR1 Δ660–685 and BubR1 L669A/I672A and purified Venus B56α from mitotic cells. While the endogenous BubR1 was equally enriched in the different samples only mCherry BubR1 WT bound to Venus B56α confirming that indeed the residues identified by the peptide array experiments are critical for binding in vivo (Fig. 2F).
To further characterize the interaction we used isothermal titration calorimetry (ITC) and measured the affinity of purified B56α for a BubR1 peptide encompassing amino acids 660–682. This peptide bound to B56α with a Kd of 4.6 µM whereas when L669 and I672 were mutated to alanine the binding was completely abolished (Fig. 2G). To determine if phosphorylations on S670 and S676 affected the binding to B56α we measured the affinity of the same peptide with phosphorylations on either S670 or S676 or both sites phosphorylated. When S670 was phosphorylated the peptide bound to B56α with a Kd of 0.5 µM while phosphorylation of S676 lowered the Kd to 0.8 µM and the double phosphorylated peptide bound with a Kd of 0.12 µM. This reveals that phosphorylation of S670 or S676 stimulates the interaction between B56α and the BubR1 peptide in vitro, which is in line with our observation that BubR1 and B56 subunits interacted stronger during mitosis. Indeed when we inhibited Plk1 in mitotic cells there was a slight reduction in the interaction between BubR1 and B56 (Fig. 2H).
Our results show that B56 subunits bind directly to a conserved region in BubR1 centered on the Cdk1 S670 phosphorylation site and that the conserved L669 and I672 are required for B56 binding. The binding of B56 subunits to BubR1 is stimulated by phosphorylation of S670 and S676 revealing that Cdk1 and Plk1 can stimulate the binding. Since phosphorylation of BubR1 on S670 and S676 appears to happen exclusively at unattached and tension-less kinetochores respectively (Elowe et al., 2010; Huang et al., 2008) our results suggest that the binding between BubR1 and B56 is stimulated in response to improper kinetochore microtubule attachments and that the two proteins interact at the kinetochore.
Binding of BubR1 to B56 subunits is required for proper chromosome segregation
To determine if BubR1 regulates the recently described localization of B56 subunits to the kinetochore (Foley et al., 2011) we analyzed B56α localization in taxol or nocodazole arrested cells depleted of BubR1 and compared it to control treated cells. In taxol treated cells we observed an approximate 50% reduction in total B56α levels at the kinetochore/centromere when BubR1 was depleted while in nocodazole arrested cells the total levels appeared similar (Fig. 3A,B). However when we inspected the line profile of B56α staining through the kinetochore/centromere of sister chromatids in nocodazole arrested cells it was clear that BubR1 depletion resulted in a sharper profile (Fig. 3C). This shows that BubR1 can recruit a subset of B56 to the outer kinetochore in both taxol and nocodazole arrested cells.
To determine if the interaction between BubR1 and B56–PP2A complexes is critical for chromosome segregation we analyzed the ability of fluorescent RNAi resistant BubR1 WT, BubR1 Δ660–685 and BubR1 L669A/I672A to rescue the chromosome alignment defect observed when endogenous BubR1 was depleted. In the absence of BubR1 cells rapidly progressed through mitosis and no clear metaphase plate was visible (Fig. 3D–F) whereas the introduction of BubR1 WT restored mitotic timing and chromosome alignment. Cells complemented with BubR1 Δ660–685 and BubR1 L669A/I672A arrested in mitosis for a prolonged time with several unaligned chromosomes indicating that BubR1 binding to B56–PP2A complexes is critical for chromosome alignment (Fig. 3D–F).
Using a similar approach we also analyzed the phenotype by immunofluorescence under the different rescue conditions (Fig. 4A–C). We treated cells for 2 hours with MG132 to maintain them in mitosis and then either fixed them directly or kept them on ice for 15 minutes before fixation to depolymerize non-kinetochore microtubules. Similar to what we observed by time-lapse microscopy, cells expressing BubR1 WT formed a metaphase plate while BubR1 Δ660–685 and BubR1 L669A/I672A expressing cells had numerous unaligned chromosomes that lacked cold stable microtubules (Fig. 4A–C).
Here we have revealed a novel interaction between BubR1 and the B56–PP2A complex that is critical for chromosome segregation and our work provides insight into how Cdk1 and Plk1 help establish stable kinetochore–microtubule interactions in prometaphase by stabilizing the interaction between BubR1 and B56 subunits to counteract Aurora B. Interestingly the individual phosphorylations on S670 and S676 of BubR1 both stabilize the binding to B56 and combined they further increase the affinity indicating that the levels of B56–PP2A complexes on kinetochores could be fine tuned by combinations of these phosphorylations depending on the microtubule binding state of the kinetochore. While this work was being prepared for submission a study from the Kops lab came to similar conclusions and they identify T680 as a novel Plk1 site that also stimulates the interaction between BubR1 and B56 (Suijkerbuijk et al., 2012). Understanding how these phosphorylations increases the binding between BubR1 and B56 and how Plk1 and Cdk1 monitors microtubule-kinetochore interactions will be an important topic for future research.
Materials and Methods
Yeast two-hybrid screening was performed by Hybrigenics Services, S.A.S.
The full length coding sequence for BubR1 (GenBank accession number gi: 168229167) was PCR-amplified and cloned into pB29 as an N-terminal fusion to LexA (N-BubR1-LexA-C). The construct was checked by sequencing the entire insert and used as a bait to screen a random-primed Human Placenta cDNA library constructed into pP6. 68.1 million clones (6.8-fold the complexity of the library) were screened using a mating approach with Y187 (matα) and L40ΔGal4 (mata) yeast strains. 172 His+ colonies were selected on a medium lacking tryptophan, leucine and histidine, and supplemented with 10 mM 3-aminotriazole to handle bait autoactivation. The prey fragments of the positive clones were amplified by PCR and sequenced at their 5′ and 3′ junctions. A confidence score was attributed to each interaction.
The following antibodies were used at the indicated dilutions for western blot. BubR1 (WB, A300-995A, 1∶1000, Bethyl Laboratories), BubR1 (IF, A300-386A, 1∶1000, Bethyl), BubR1 (monoclonal produced in-house used for IP), CDC20 (sc-13162, 1∶1000, Santa Cruz), PP2A catalytic subunit (05-421, 1∶2000, Millipore), GFP (ab290, 1∶4000, abcam), B56α (610, 615, 1∶3000, BD Biosciences), CREST (Antibodies Inc., 1∶400)
Full length B56α, B56β, B56γ and B55α were amplified from Invitrogen Ultimate clones by PCR and cloned into BamHI and NotI sites of pcDNA5/FRT/TO 3*FLAG-Venus. BubR1 1–483 and 1–715 was obtained by amplifying the corresponding region of BubR1 from pcDNA5/FRT/TO 3*FLAG Venus BubR1 (siRES) (Bolanos-Garcia et al., 2011) and inserting into the BamHI and NotI sites of pcDNA5/FRT/TO 3*FLAG Venus. BubR1 Δ660–685 was obtained by two step PCR using pcDNA5/FRT/TO 3*FLAG Venus BubR1 (siRES) as template and inserted into BamHI and NotI sites of pcDNA5/FRT/TO 3*FLAG Venus. pcDNA5/FRT/TO 3*FLAG Venus BubR1 L669A/I672A was generated by whole plasmid PCR. BubR1, BubR1Δ660–685 and BubR1 L669A/I672A were subcloned into pcDNA3.1 mCherry vector using BamHI and NotI sites. All constructs were verified by sequencing.
The E. coli expression plasmid 2A5A-c001 for expression of B56α was constructed by ligating a PCR amplified insert into vector pCPR0011 through ligation independent cloning. The PCR primers were 2A5A-2FW: tacttccaatccatgTCGTCGTCGTCGCCGCCG and 2A5A-484RV: tatccacctttactgttaACTTGTATTGCTGAGAATACTGTGCATGTTGTAAG. pCPR0011 is a derivative of pNIC28-Bsa4 (GenBank Acc EF198106) where a StrepII tag (WSHPQFEK) has been inserted between the His-tag and the TEV-cleavage site.
The expression plasmid for BubR1 516–715 was constructed by ligating a PCR amplified insert into vector pCPR0005 through ligation independent cloning. (pCPR0005 is a derivative of pNIC28-Bsa4 (GenBank Acc EF198106) where a FLAG-HA tag (DYKDDDDK-YPYDVPDYA) has been inserted between the His-tag and the TEV-cleavage site. The PCR template was entry clone IOH21121 (Invitrogen) containing the BubR1 coding sequence and the PCR primers were BUB1B-1-516FW: tacttccaatccatgCAGGAACAACCTCATTCTAAAGGTCCC and BUB1B-1-715Rev: tatccacctttactgttaTTCTGAAGTCTCATTAGTAAGTTCTAGTTTCTCAG.
Protein expression and purification
Proteins were expressed in the E. coli strain BL21 Rosetta2 (DE3) R3 T1 at 18° for 20 hours using 0.5 mM IPTG. The pellets were resuspended in buffer L (50 mM NaP, 300 mM NaCl, 10% glycerol, 0.5 mM TCEP, pH 7.5) containing protease inhibitors and lyzed with a high-pressure homogeniser at 1000 Bar. The lysate was centrifuged at 18,500 g for 30 minutes and the supernatant filtered through a 0.22 µm PES filter and loaded onto a 1 ml Ni column (GE healthcare) in buffer L with 10 mM immidazole, washed and eluted. The eluate was loaded on a Superdex 200 PG 16/60 equilibrated with SEC buffer (50 mM NaP, 150 mM NaCl, 0.5 mM TCEP, 10% glycerol, pH 7.50) and fractions analyzed by SDS-PAGE. All proteins verified by mass spectrometry.
Size exclusion chromatography
A Superdex 200 10/300 (GE healthcare) column was equilibrated with SEC buffer and 250 µg of B56α or BubR1 516–715 run on the column with a flow-rate of 0.4 ml/minute and 0.5 ml fractions collected. 250 µg B56α and 250 µg BubR1 516–715 were mixed in SEC buffer and incubated on ice for 1 hour before loading on the column.
Peptide array experiments
Peptide arrays were synthesized using a ResPepSL Bioanalytical Instrument (Intavis, Köln, Germany) on derivatized cellulose membranes using standard Fmoc [N-(9-fluorenyl)methoxycarbonyl] chemistry according to the manufacturer's instructions. For B56α binding experiments, purified recombinant B56α was added at the indicated concentrations in PBST (0.05% tween 20) containing 3% BSA and incubated together with the membrane overnight at 4°C. Membranes were washed three times in PBST and bound protein was visualized with anti-B56α antibodies and HRP or fluorescently conjugated antibodies.
The following purified peptides were purchased from Biosyntan GmbH (Berlin, Germany): WT: YSQTLSIKKLSPIIEDSREATHS, L669A/I672A: YSQTLSIKKASPAIEDSREATHS, S670(Phos): YSQTLSIKKL-pS-PIIEDSREATHS, S676(Phos): YSQTLSIKKLSPIIED-pS-REATHS and S670/S676(Phos): YSQTLSIKKL-pS-PIIED-pS-REATHS. The peptides and B56α were extensively dialyzed into 50 mM sodium phosphate, 150 mM NaCl, 0.5 mM TCEP [tris(2-carboxyethyl)phosphine], 5% glycerol, pH 7.5. Protein and peptide concentrations were determined by UV spectroscopy and molar extinction coefficients at 280 nm of 1490 and 47,245 M−1cm−1 for the peptides and B56α, respectively. Isothermal Titration Calorimetry (ITC) experiments were performed at 25°C using an ITC200 instrument (Microcal). 1.5 µl volumes of peptide, at ∼450 µM, were titrated into the ITC sample cell containing ∼35 µM of B56α, until saturation was achieved. Blank titrations of peptide into buffer were performed to determine peptide dilution heat effects and in all cases these were negligible. For the L669A/I672A peptide, no binding was observed at 25°C and the experiment was repeated at 10°C to take into account potential differences in binding ΔCp (change in heat capacity) which could lead to no heat effects being observed at 25°C. The heat of the reaction was obtained by integrating each peak after the injection of peptide and fit to a model describing a single binding site using software provided by the ITC200 manufacturer.
Cells were lysed in lysis buffer (50 mM Tris-HCl pH 7.5, 50 mM NaCl, 1 mM EDTA, 1 mM DDT and 0.1% NP40). Complexes were immunoprecipitated in lysis buffer with antibodies coupled to Protein G-Sepharose 4B (Invitrogen) or GFP-Trap (ChromoTek) beads as indicated and incubated at 4°C for 2 hours or 30 minutes, respectively. Precipitated protein complexes were washed in washing buffer (50 mM Tris-HCl pH 7.5, 1 mg/ml BSA, 20% glycerol and 1 mM DTT) and eluted in 2×SDS sample buffer. The purification of Venus tagged BubR1 constructs was performed as described previously (Bolanos-Garcia et al., 2011).
Tissue culture work
HeLa cells were maintained in DMEM with 10% serum. The following drug concentrations were used: Thymidine 2.5 µM, MG132 10 µM, Nocodazole 200 ng/µl, Taxol 200 ng/µl, Reversine 0.5 µM, ZM447439 2 µM, BI2536 0.5 µM.
Live cell analysis
Live cell analysis and immunofluorescence was performed as described previously (Bolanos-Garcia et al., 2011; Zhang et al., 2012) using a Deltavision Elite microscope (GE Healthcare). The BubR1 RNAi rescue experiments were done as described in Bolanos-Garcia et al. (Bolanos-Garcia et al., 2011) except that cells were transiently transfected with Venus-BubR1 constructs and CFP-Histone H3 together with 100 nM RNAi oligo targeting BubR1 using lipofectamine 2000 (Invitrogen). After 24 hours the cells were filmed using a 40×, 1.35 NA, WD 0.10 objective. All data analysis was performed using the softworx software (GE Healthcare).
For immunofluorescence microscopy cells were synchronized with 2.5 mM thymidine, 10 µM MG132 and 200 ng/µl nocodazole. Cells were pre-fixed with 4% PFA in PHEM buffer (50 mM PIPES, 25 mM HEPES, 10 mM EGTA, 8.5 mM MgSO4, pH 7.0) for 20 seconds, permeabilized in 0.5% Triton X-100 in PHEM buffer for 5 minutes and fixed in 4% PFA in PHEM buffer for 20 minutes. Coverslips were quenched with 25 mM Glycin in PBS for 20 minutes, blocked with 3% BSA in PBS-T (0.1% Tween in PBS) for 30 minutes, incubated with primary antibodies for 1 hour, with Alexa Fluor goat secondary antibodies for 45 minutes and mounted on slides using ProLong Gold Antifade mounting media (Invitrogen). Images were acquired taking z stacks of 200 nm using a 100×/1.4NA objective on a Deltavision Elite Microscope (Applied Precision). Images were analyzed after deconvolution using SoftWoRx (Applied Precision Instrument). Figures were generated by maximum intensity projection of entire cells using Softworx and ImageJ.
We thank Sabine Elowe for plasmids and Stephen Taylor for the HeLa/FRT/TRex cell line. We thank Geert Kops for communicating results before publication and Mikkel Staberg for help with peptide arrays and Mia Funk Nielsen for help with protein production. The authors have no conflict of interest.
T.K. performed the experiments shown in Figs 1 and 2, except that shown in Fig. 2B, which was performed by J.N. Experiments shown in Fig. 2G were performed by W.S., G.Z. performed the experiments shown in Fig. 4, M.S.Y.L those shown in experiments in Fig. 3D–F and T.L. performed the experiments shown in Fig 3A–C. T.K.N and S.P.B produced and purified recombinant proteins. J.N. designed the experiments together with the other authors and wrote the paper.
This work was supported by grants from the Novo Nordisk Foundation, the Lundbeck Foundation, and The Danish Council for Independent Research-FNU (all to J.N.); and a grant from The Danish Council for Independent Research-FSS to G.Z.