Unattached kinetochores catalyze production of an anaphase inhibitor that requires a Mad2 template to prime Cdc20 for BubR1 binding

Abstract

Premature anaphase onset is prevented by the mitotic checkpoint through production of a "wait anaphase" inhibitor(s) that blocks recognition of cyclin B and securin by Cdc20-activated APC/C, an E3 ubiquitin ligase that targets them for destruction. Using physiologically relevant levels of Mad2, Bub3, BubR1, and Cdc20, we demonstrate that unattached kinetochores on purified chromosomes catalytically generate a diffusible Cdc20 inhibitor or inhibit Cdc20 already bound to APC/C. Furthermore, the chromosome-produced inhibitor requires both recruitment of Mad2 by Mad1 that is stably bound at unattached kinetochores and dimerization-competent Mad2. We show that purified chromosomes promote BubR1 binding to APC/C-Cdc20 by acting directly on Mad2, but not BubR1. Our results support a model in which immobilized Mad1/Mad2 at kinetochores provides a template for initial assembly of Mad2 bound to Cdc20 that is then converted to a final mitotic checkpoint inhibitor with Cdc20 bound to BubR1.

Fig. 1. Unattached kinetochores on purified chromosomes recruit Mad2

(A) Schematic of chromosome purification from mitotic HeLa cells stably expressing YFP-H2B histone. Cells were collected after 16 hours in colcemid, lysed, cell debris removed by pelleting and the chromosome containing supernatant was fractionated on sequential sucrose gradients. (B) Morphology of purified chromosomes detected by fluorescence of YFP-H2B on coverslips without fixation. (C) Protein constituents of purified chromosomes assessed by immunoblotting after pelleting. (D) Tubulin levels remaining in purified chromosomes, along with a dilution series of the initial cellular input. (E) Indirect immunofluorescence for detection of Mad1, Bub1, CENP-E and Mad2 on isolated chromosomes. (Blue) Chromosomes stained with DAPI; (Green) Anticentromere (ACA) antibodies; (right panel) merged image. (F) Purified recombinant Mad2 before and after covalently labeling with rhodamine, assessed by Coomassie staining. (G) Purified chromosomes were incubated with rhodamine-labeled Mad2, fixed, stained for (blue) DAPI and (green) ACA, and imaged by deconvolution microscopy. (H) Chromosomes were incubated for 10 min with anti-Mad1 antibody, then with rhodamine-labeled Mad2, and finally fixed, stained, imaged as in (G), and scored for Mad2 localization.

Fig. 2. Kinetochores amplify production of an APC/C^Cdc20 inhibitor

(A) Purified recombinant human Cdh1, Cdc20, Mad2, Bub3 and BubR1, assessed by Coomassie staining. (B) Purified recombinant human E1, UbcH10, and N-terminal of cyclin B, assessed by Coomassie staining. (C) Interphase Xenopus APC/C after immunoprecipitation with immobilized antibodies to Cdc27 and visualized by silver stain. (D) Schematic of APC/C ubiquitination activity assays. (E) Equal molar amounts of Mad2, Bub3, and BubR1 were incubated with Cdcd20 either alone or in various combinations, in the absence of unattached kinetochores. APC/C activity was assessed either as the degree of cyclin B ubiquitination (top panel) or as the depletion of the unubiquitinated pool (bottom panel). (F-H) Kinetochores on purified chromosome amplify inhibition of mitotic APC/C^Cdc20. (F) Schematic of Xenopus extract preparation for isolation of mitotic APC/C by immunoprecipitation. (G) Immunoprecipitated mitotic APC/C: hyperphosphorylation retards mobility relative to interphase APC/C after immunoblotting for Cdc27. Hyperphosphorylation is lost upon phosphatase treatment. (H) Comparison of mitotic and interphase APC/C activity (quantified by diminished abundance of lower mobility ubiquitin-conjugated cyclin B species) assessed after addition of increasing amounts of BubR1, Bub3, and Mad2 to Cdc20, either in the presence or absence of chromosomes. (I) BubR1, Bub3, Mad2 and chromosomes were incubated with either (red bar) Cdc20 or (blue bar) Cdh1 activators, prior to APC/C activity determination. Incubations with Cdc20 rendered the APC/C almost fully inactive, while Cdh1 incubations had no effect on the activity of APC/C. (J-L) Chromosomes (blue squares) at a final concentration equaling ten unattached kinetochores per cell volume or just buffer (red triangles) were added to increasing concentrations of (J) Mad2, (K) Bub3/BubR1, or (L) both, incubated for 1 hr prior to addition of APC/C, and then assayed for APC/C ubiquitination of myc-cyclin B_1-102. APC/C activity was quantified by the intensity of remaining unubiquitinated cyclin B. Mad2 inhibition of APC/C in the presence of chromosomes increased, while BubR1 inhibition remained unchanged. Chromosomes further amplified APC/C inhibition when added to the combination of Mad2, BubR1, and Bub3 at physiological concentrations.

Fig. 3. Mad1-dependent Mad2 conformational change and dimerization are required for kinetochore-mediated amplification of a Cdc20 inhibitor

(A) Schematic of Mad2 template model (De Antoni et al., 2005). (B) Table of Mad2 mutant properties (De Antoni et al., 2005; Fang et al., 1998a; Luo et al., 2000; Luo et al., 2004; Sironi et al., 2002). (C) Purified recombinant human Mad2^ΔC, Mad2^RQ, and Mad2^RQ-ΔC visualized by Coomassie staining. (D) Increasing quantities of (red triangles) Mad2^wt, (blue squares) Mad2^ΔC, (green circles) Mad2^RQ, or (yellow diamonds) Mad2^RQ-ΔC were incubated for 1 hour with Cdc20 prior to APC/C addition and assayed for APC/C ubiquitination of myc-cyclin B_1-102. (E) Mad2RQ inhibition of Cdc20 activation of APC/C assessed after incubating increasing quantities with Cdc20 either in the (blue squares) presence or (red triangles) absence of chromosomes before assaying cyclin B ubiquitination. (F) Testing chromosome amplification of a Cdc20 inhibitor after increasing concentrations of (green circles) Mad2^RQ or (blue squares) Mad2^wt, along with BubR1, Bub3, and Cdc20, either in the presence or (red triangles) absence of chromosomes. (G) Purified chromosomes were incubated with Mad1 antibody for 1 hour prior to addition of a five-fold excess of Mad2 to Cdc20, and finally addition to APC/C ubiquitination assays and (H) inhibition of AP/CC^Cdc20 was measured.

Fig. 4. Inhibition of APC/C activation is not achieved solely by sequestration of Cdc20

(A) Checkpoint components including Cdc20 and chromosomes were incubated, followed by APC/C addition and assay for its activity. (B) Immunoprecipitated APC/C was first incubated with Cdc20 to form an active complex (“Pre-activated APC/C”). APC/C was affinity recovered and subsequently incubated with chromosomes and increasing amounts of BubR1, Bub3, and Mad2, and APC/C activity was assayed. (C) Quantitation of (red triangles) co-incubated and (blue squares) pre-activated Cdc20-stimulated APC/C activity.

Fig. 5. Catalytic production by unattached kinetochores of a diffusible Cdc20 inhibitor through sequential involvement of Mad2 and BubR1

(A) Chromosomes were incubated with BubR1, Bub3, Mad2, and Cdc20 (1:1:1:5); the chromosomes were subsequently removed by centrifugation, and (blue) the supernatant fraction was assayed for activation of APC/C for cyclin B ubiquitination. (Red) Parallel assay was done without chromosome removal or (white) without initial chromosome addition. (B) Chromosomes were initially incubated with Mad2 and Cdc20 (1:5); the chromosomes were subsequently removed, and BubR1/Bub3 added to the supernatant fraction after chromosome removal (green). The resulting ubiquitination activity was compared to (blue) inhibition produced by incubating chromosomes with Mad2, BubR1, and Bub3 prior to chromosome removal or (white) no chromosomes added. (C) Cdc20 was incubated with a 5-fold excess Mad2 and chromosomes. At the indicated time points, (blue squares) a fraction of the incubation was removed and assayed for APC/C activity. Chromosome-mediated catalysis was compared to (red triangles) chromosome-independent inhibitor production over time.

Fig. 6. Unattached kinetochores facilitate BubR1, Bub3, and Mad2 association with APC/C^Cdc20

(A)APC/C was incubated with pre-incubated combinations of Cdc20, Mad2, BubR1, Bub3 and chromosomes, recovered, peptide-eluted from Affiprep beads, and analyzed for bound components by immunoblotting. (B-E) APC/C was incubated with Cdc20 and increasing amounts of BubR1, Bub3, and Mad2^wt (or Mad2^RQ) either in the (blue squares) presence or (red triangles) absence of chromosomes, and treated as in (A). The amounts of eluted (B) BubR1, (C) Bub3, and (D) Mad2^wt (or Mad2^RQ) were measured against a dilution series of purified protein and quantified relative to the amount of Cdc20 bound to APC/C. Values were plotted as fold change over the initial (1x) non-chromosome incubated eluted amounts. (E) The relative stoichiometry of BubR1 molecules to Mad2 molecules associated with the APC/C complex. BubR1 relative stoichiometry to Mad2 increased above 1:1 when chromosomes were present, but remained approximately 1:1 or below when Mad2^wt was replaced with Mad2^RQ regardless of the presence of chromosomes. (F) APC/C incubated with combinations of Cdc20, Mad2, BubR1, Bub3 and chromosomes was recovered, eluted from Affiprep beads, fractionated over a sucrose gradient, and analyzed by immunoblotting for bound components.

Fig. 7. Kinetochores catalyze the production of a BubR1-Cdc20 inhibitor without stably associated Mad2

(A) Individual components or (B-E) combinations of Cdc20, Mad2, BubR1, Bub3, and/or chromosomes were incubated for 1hr and subsequently fractionated over a Superose-6 filtration column. (E) Mixtures containing chromosomes were removed by pelleting the chromosomes prior to loading onto the column. Fractions eluted from the column were analyzed for BubR1, Bub3, Cdc20, and Mad2 content by immunoblotting for those components. (F) Model for generation of a “wait anaphase” mitotic checkpoint inhibitor by sequential production of Mad2-Cdc20 and BubR1-Cdc20 inhibitors. Cytosolic Mad2 in an initially open conformation is recruited to unattached kinetochores via an immobilized Mad1:Mad2 heterodimer. This second molecule of Mad2 binds in an activated conformation that is poised for capture of Cdc20 either while kinetochore bound or after release. This transient Mad2-Cdc20 complex promotes handoff of Cdc20 to BubR1, thereby inhibiting ability of that Cdc20 to activate ubiquitination by APC/C of cyclin B, both by sequestering Cdc20 from APC/C and by inhibiting Cdc20 while APC/C bound.