Fig 1.
Detailed model of the mitotic checkpoint with bistability.
A) Generic wiring diagrams for positive feedback loops in the mitotic checkpoint network. B) Detailed wiring diagram of the mitotic checkpoint. C) Bifurcation diagram of the detailed model (wiring in panel B). The red square indicates the saddle node bifurcation: the value of unattached kinetochores where the checkpoint ON steady state is created/disappears. In wild type cells, it is located just above 1 unattached kinetochore (nuk=0.9). Equations in S1 Text, and parameters in Table 2. D) Stochastic simulations of the Palframans’s experiments [9]. Reactions in Table 3, parameters in Table 2, initial conditions in Table 1 (checkpoint OFF). Mps1 overexpression is simulated increasing 3X the synthesis of Mps1 for one hour.
Table 1.
Species and initial conditions of used in the model to simulate a checkpoint arrest.
Table 2.
Parameters used in the detailed model for WT simulations. Choice of parameters is given in S1 Text.
Table 3.
Reactions for stochastic simulations.
Fig 2.
Adaptation dynamics with constant unattached kinetochores.
A) Definition of adaptation in the experimental system. UK stands for unattached kinetochores. B) Cumulative distribution of mitotic arrest length. Cells growing in YPD are arrested in G1 and released in nocodazole, indefinitely. Cells carry Mad2-GFP to record checkpoint activation, and Clb2-mCherry to record both mitotic entry (Clb2 rise time) and mitotic exit (Clb2 degradation time). Length of mitotic arrest is defined as the difference between mitotic exit and mitotic entry. C) Definition of adaptation in the model. Adaptation occurs when the system transits from the ON to the OFF state with unattached kinetochores. D) Stochastic simulations of APC/CCdc20 dynamics with fixed number of unattached kinetochores (nuk=10). Reactions in Table 3, parameters in Table 2, initial conditions in Table 1 (checkpoint ON). Filled circles indicate the time APC/CCdc20 crosses the activation threshold (80 mol/cell), after which we assume entry into anaphase. E) Cumulative distribution of simulated mitotic arrests, computed as the difference between the start of the simulation and the time APC/CCdc20 crosses the activation threshold. Representative trajectories are shown in panel (D).
Fig 3.
Adaptation following drug washout.
A) Left panel: Wild-type cells growing in YPRG are arrested in G1 and released in nocodazole for 180 minutes, after which nocodazole is removed from the media. Cells carry Mad2-GFP to record checkpoint activation, and Clb2-mCherry to record mitotic exit (Clb2 degradation time). Adaptation and exit are defined based on Mad2 localization at Clb2 degradation time (see Figs 2A and S2B). Only cells properly degrading Clb2 are included in the percentage (see S2A Fig). Right panel: same percentages for the simulated data, where adaptation is defined as APC/C activation with at least one unattached kinetochore (see Fig 2A). B) Representative stochastic simulations of APC/CCdc20 dynamics (left) from the time kinetochores are allowed to attach (right). Yellow triangles indicate the moment when all kinetochores have attached; filled circles mark the time APC/CCdc20 crosses the activation threshold. Reactions in Table 3, parameters in Table 2, initial conditions in Table 1 (checkpoint ON). C) Left panel, top: scatter plot of the number of unattached kinetochores at APC/C activation in wild-type adapting cells. Black line shows the linear regression, R- and p-values from Pearson’s correlation. Bottom: distribution of APC/CCdc20 activation time in cells adapting (purple) or exiting (green). Data from 100 simulations as in panel (B). Right panel, top: scatter plot of the amount of Mad2 at kinetochores (see Material and Methods) in wild-type adapting cells. The black line shows the linear regression, R- and p-values from Pearson’s correlation. Bottom: histogram of Clb2 degradation time in adapting (purple) or exiting cells (green).
Fig 4.
Dynamics of nocodazole washout in APC-A mutants.
A) Bifurcation diagrams showing the effect of decreasing APC/CCdc20 binding to 88%, 79% and 69% of Wild Type levels. Wild type in black. Equations in S1 Text, and parameters in Table 2. B) Adapting vs exiting cells. Left panel: APC-A cells growing in YPRG were arrested in G1 and released in nocodazole for 180 minutes, after which nocodazole was removed from the media. Cells carry Mad2-GFP to record mitotic checkpoint activation, and Clb2-mCherry to record mitotic exit (Clb2 degradation time). Adaptation and exit are defined based on Mad2 localization at Clb2 degradation time (see S2B Fig). Only cells degrading completely Clb2 are included in the percentage (see S2A Fig). Right panel: same percentages for the simulated data, where adaptation is defined as APC/C reaching the activation threshold with at least one unattached kinetochore. Reactions in Table 3, parameters in Table 2, initial conditions in Table 1 (checkpoint ON). At time 0, kinetochore attachment is allowed. C) Different dynamics of cells exiting and adapting. Upper panel: histograms of APC/C activation time in in cells adapting (purple) or exiting properly (green). Simulations details as in panel (B) right. Lower panel: histograms of Clb2 degradation time in in cells adapting (purple) or exiting (green). Experimental details as in (B) left. D) Comparison of Clb2 degradation time (experiments) and APC/C activation (simulations) in wild-type and APC-A cells. Left, experimental data: cells are the same presented in the barplots in Figs 3A and 4B. Right, results from wash-out simulations. Equations in S1 Text, and parameters in Table 2.
Fig 5.
Dynamics of nocodazole washout upon Mad2 overexpression.
A) Bifurcation diagrams showing the effect of increasing the Mad2 levels to 105%, 115% and 125% of Wild Type levels. Wild type in black. Equations in S1 Text, and parameters in Table 2. B) Barplots show percentages of simulated cells adapting or exiting, where adaptation is defined as APC/C activation with at least one unattached kinetochore. Reactions in Table 3, parameters in Table 2, initial conditions in Table 1 (checkpoint ON). C) Histogram of APC/C activation time in in simulated cells adapting (purple) or exiting (green). Details of simulations as in panel (B). D) Comparison of Clb2 degradation time (experiments) and APC/C activatiom time (simulations) in wild-type and GAL1-MAD2 cells. In the experiments (right), cells growing in YPRG are arrested in G1 and released in nocodazole for 240 minutes, when nocodazole is removed from the media. Cells carry Clb2-mCherry to record Clb2 degradation time. Only cells fully degrading Clb2 are included (see S2A Fig; 95% in wild-type, 87% in GAL1-MAD2). Simulations (left) produced as described in panel B. E) Percentage of cells that missegregate chromosome V. Cells growing in YPRG are arrested in G1 and released in nocodazole for 180 minutes, when nocodazole is removed from the media. Cells carry tetR-GFP/tetO construct to keep track of chrV segregation, and Htb2-mCherry to score nuclear division time. Cells that do not undergo nuclear division are excluded from this barplot (see S5F Fig). Missegregation is defined based on 5 frames (50 minutes) around nuclear division time. A cell is marked as missegregating chrV if after nuclear division either the two GFP dots end in the same cell or only one GFP dot is visible (see S5D and S5E Fig). For statistical comparison we used an exact Fisher test, where the contingency table has wild-type or mutant cells versus correct or wrong segregation of chrV (p-value = 1.3 10-4).