Cdc20 is degraded in checkpoint-arrested cells
A) HeLa cells were pre-synchronised in S phase and subsequently arrested in mitosis with nocodazole for 6 hrs. Cells were left untreated (control) or treated with either MG132 or cycloheximide (CHX) for 2 hrs and the level of BubR1, Cdh1, Cdc20 and Mad2 was determined by quantitative immunoblotting. Protein levels were normalised against actin and the bar diagrams indicate the levels of the indicated proteins with the untreated sample set to 1. B) The levels of BubR1, Cdc20, Cyclin B1 and Mad2 at 0, 20, 40, 60 and 120 min after addition of cycloheximide to either nocodazole- or taxol-arrested cells (synchronised as in A) were analyzed by ECL western blot (whole blot shown in Fig S7). C) The levels of Cdc20 and Mad2 at 0, 20, 40, 60 and 120 min after addition of cycloheximide or rapamycin to nocodazole arrested cells (whole blot shown in Fig S7). D) The level of p70S6K phoshorylated on Thr389 at 0, 30, 60 and 120 min following addition of rapamycin to asynchronous HeLa cells. E and F) Degradation of YFP-Cdc20 in single cells. G2 phase HeLa cells were microinjected with a plasmid encoding YFP-Cdc20 and the fluorescence level assayed through mitosis. Note that the YFP-Cdc20 was encoded by cap-dependent mRNA, whose translation was repressed in mitosis, therefore, we could assay degradation without cycloheximide. Either nocodazole or MG132 was added at the times indicated. Nocodazole was added in prometaphase in E) and in G2 phase in F). Cells are representative of 10 cells in 3 independent experiments.

Cdc20 requires both APC/C and Mad2 binding motifs to be degraded in the SAC
HeLa cells were microinjected with plasmids encoding YFP- wild-type or mutant Cdc20 as indicated and the fluorescence level assayed through mitosis in the presence of nocodazole. Data are representative of 3 independent experiments for each mutant. A) IR mutant (R499E) 9 cells B) C-box mutant (D⁷⁷RYIPHR⁸³ to A⁷⁷AAIAHA⁸³) 13 cells C) Mad2-binding site R132A, 12 cells.

A Cdc20 mutant without lysines overrides the spindle checkpoint
A) Wild-type or the K-less mutant of Cdc20 were expressed and purified from baculovirus-infected SF9 cells, and increasing amounts incubated with APC/C purified from nocodazole-arrested cells. Reactions were analyzed by immunoblotting for Cdc20.. B) G2 phase HeLa cells treated with siRNA to deplete endogenous Cdc20 were injected with plasmids encoding siRNA resistant YFP-tagged wild-type Cdc20 or the K-less mutant and treated with nocodazole. Cells were monitored by time-lapse DIC and fluorescence microscopy at 3 min intervals and the fluorescence levels measured and quantified. The slow decrease in YFP-Cdc20 K-less is due to photo-bleaching. C and D) DIC and fluorescence images of the cells analysed in (B). The mitotic cell (arrowed) expressing the YFP-Cdc20 K-less protein (D) exits the nocodazole block after ∼120 min whereas the cell expressing wild-type YFP-Cdc20 remains arrested for more than 5.5 hrs and degraded the YFP-Cdc20 protein (C). Cells are representative of 17 (wild-type) and 16 (K-less mutant) cells in 2 independent experiments.

Spindle Checkpoint Model
Cdc20 is continually synthesized in mitosis and acts as an activator of the APC/C. However, in the presence of improperly attached kinetochores, Mad2 binds Cdc20 and this allows the loading of Cdc20 onto BubR1. Once BubR1 is bound to Cdc20, Mad2 is not required to sustain the binding and leaves the complex before or after the complex is presented to the APC/C. The APC/C then promotes ubiquitination and degradation of Cdc20 to sustain the checkpoint.

Cdc20 degradation is dependent on APC/C activity and a functional SAC.
A) Cells were treated with siRNA oligos against either GADPH (as in Fig. 1) or APC3 and after 72 hrs the stability of Cdc20 was determined in nocodazole-arrested cells by adding cycloheximide for 30 min and analysis by quantitative immunoblotting (whole blots shown in Fig S7). B) The level of Cdc20 was normalized to Hsp70 and the mean and standard deviation of Cdc20 levels in 3 experiments are shown with the untreated sample set to 1. C) The level of YFP-Cdc20 fluorescence in a mitotic cell treated with APC3 siRNA oligos (which blocked cells in mitosis for several hours in >95% of cells). The cell shown is representative of 6 cells in 2 experiments. D) The level of YFP-Cdc20 fluorescence in mitotic cells treated with siRNA targeting Mad2 or GAPDH in the presence of nocodazole. The data shown are representative of 10 and 6 cells in 2 experiments for cells treated with siRNA against, respectively, GAPDH and Mad2,

Analysis of checkpoint complexes by gel-filtration chromatography
Taxol-arrested cells (A) or taxol-arrested cells treated with MG132 for 2 hrs (B), or taxol-arrested cells treated with MG132 and ZM447439 for 2 hrs (C) were fractionated on a Superose 6 column. Fractions were probed for BubR1 (red), APC4 (green), Cdc20 (black) and Mad2 (grey), and analyzed by quantitative immunoblotting. Values were normalized to the peak fraction value to obtain the profiles shown. (The novel peak of Cdc20 indicated by an asterisk that appears between the APC/C and BubR1 peaks may be Cdc20 bound to the CCT chaperone, since it is not always seen, for example, see Supplemental Fig 4 and 5.) Results are representative of 4 or more independent experiments. (Whole blot for taxol-arrested cells shown in Fig S7. Note that this is probed sequentially for Cdc20, then BubR1 and then Mad2.) D & E) Cdc20 was immunoprecipitated from each of the two peak fractions from nocodazole-arrested cells (fraction 22 and fraction 28 see Fig S4) and analyzed for the presence of BubR1, APC3, Bub3 and Mad2, and for APC/C subunits in the 2 MDa complex (E). F-H) Cells were treated with siRNA oligos against GAPDH, Mad2 or BubR1 for 72 hrs. During this time cells were released from a double thymidine block and treated 9 hrs later with nocodazole and MG132 for 2hrs. Cdc20 was immunoprecipitated from mitotic cells harvested by shake-off. The amount of BubR1 and Mad2 associating with Cdc20 was determined by quantitative immunoblotting and normalized to the amount of Cdc20. The levels of BubR1 and Mad2 in GAPDH treated cells were set to 1. Data are the mean+/-SD of 3 independent experiments in F using BubR1 RNAi oligo #1, and the average of 2 independent experiments in G using RNAi oligo #2. The whole blot is shown in Fig S7. In this case the blot was cut into 3 and the different panels probed for BubR1 or Cdc20 or Mad2.

Dissociation of checkpoint proteins from Cdc20 does not require ubiquitination of Cdc20.
A) Cells were treated with siRNA against GAPDH or siRNAs against APC3 and APC11 and immunoblotted for APC3, APC11, and Hsp70 as a loading control. B) Cells were treated as in A), arrested in mitosis with taxol and the half-life of Cdc20 assayed by adding cycloheximide (CHX). C) Cells treated as in A) were incubated with Taxol for 6 hrs, the samples divided in two and MG132 or MG132 plus ZM447439 added. After 2 hrs Cdc20 was immunopurified, and the amount of BubR1 and Mad2 bound to Cdc20 determined by quantitative immunobloting. (LC= antibody light chain). The whole blot is shown in Fig S7. Note the blot was cut into 3 and panels probed for BubR1 or Cdc20 or Mad2. D) Quantification of the experiment shown in C: the level of bound BubR1 and Mad2 is normalized to the amount of Cdc20, and the level of BubR1 and Mad2 set to 1 in the samples without ZM447439. The mean +/- SD of 3 independent experiments is shown.
E) HeLa cell lines stably expressing FLAG-tagged wt or K-less Cdc20 were released from S phase, MG132 added 9 hrs later for 2 hrs and mitotic cells harvested by shake-off. Samples were split in two and 100 nM Taxol plus MG132 added to both. After 1 hr FLAG-Cdc20 was immunopurified from one and 4 μM ZM447439 added to the other 1.5 hrs before immunoprecipitating FLAG-Cdc20. The amount of BubR1, APC3 and Mad2 bound to Cdc20 was determined by quantitative immunobloting. F) Quantification of the experiment shown in E) with the level of BubR1 and Mad2 normalized to the amount of Cdc20 and the amount of BubR1 and Mad2 bound to Cdc20 set to 1 in experiments without ZM447439. Note that the K-less mutant of Cdc20 co-immunopurifies from taxol-treated cells with more BubR1 and Mad2 than does wt Cdc20. The mean +/- SD of 3 independent experiments is shown. The amount of APC3 associated with K-less and wt Cdc20 dropped by 75% when the checkpoint was inactivated.