Two Interlinked Bistable Switches Govern Mitotic Control in Mammalian Cells

Abstract

Distinct protein phosphorylation levels in interphase and M phase require tight regulation of Cdk1 activity [1, 2]. A bistable switch, based on positive feedback in the Cdk1 activation loop, has been proposed to generate different thresholds for transitions between these cell-cycle states [3-5]. Recently, the activity of the major Cdk1-counteracting phosphatase, PP2A:B55, has also been found to be bistable due to Greatwall kinase-dependent regulation [6]. However, the interplay of the regulation of Cdk1 and PP2A:B55 in vivo remains unexplored. Here, we combine quantitative cell biology assays with mathematical modeling to explore the interplay of mitotic kinase activation and phosphatase inactivation in human cells. By measuring mitotic entry and exit thresholds using ATP-analog-sensitive Cdk1 mutants, we find evidence that the mitotic switch displays hysteresis and bistability, responding differentially to Cdk1 inhibition in the mitotic and interphase states. Cdk1 activation by Wee1/Cdc25 feedback loops and PP2A:B55 inactivation by Greatwall independently contributes to this hysteretic switch system. However, elimination of both Cdk1 and PP2A:B55 inactivation fully abrogates bistability, suggesting that hysteresis is an emergent property of mutual inhibition between the Cdk1 and PP2A:B55 feedback loops. Our model of the two interlinked feedback systems predicts an intermediate but hidden steady state between interphase and M phase. This could be verified experimentally by Cdk1 inhibition during mitotic entry, supporting the predictive value of our model. Furthermore, we demonstrate that dual inhibition of Wee1 and Gwl kinases causes loss of cell-cycle memory and synthetic lethality, which could be further exploited therapeutically.

Keywords: Cdk1-Cyclin B; Greatwall-kinase; HeLa Cdk1as cell line; PP2A-B55; bistability; cell cycle; hysteresis; mitosis; mitotic control.

Graphical abstract

Figure 1

Bistable Switches of Mitotic Control (A–C) Schematic signal-response (SR) diagram for Cdk1 auto-activation (A), PP2A:B55 feedback regulation (B), and mitotic substrate phosphorylation by interlinked kinase–phosphatase switches (C). (D) Model of the hysteresis assay based on Cdk1 inhibition. The Cdk1 inhibitor shifts the SR curve to the right in a concentration-dependent manner: more inhibitor is required to induce mitotic exit (right curve) than to block mitotic entry (middle curve). (E) Summary: Cdk1 inhibitor blocks mitotic entry and promotes mitotic exit.

Figure 2

Hysteresis of Mammalian Mitotic Control (A) Experimental protocol to measure threshold 1NM-PP1 concentrations at constant CycB levels in cdk1as cells (see Figure S1 for characterization of cdk1as cells and STAR Methods for details of how they were generated). (B) Still images from time-lapse videos (see also Video S1) of HeLa cdk1as cells released from 20-hr 1NM-PP1 arrest (entry, top two panels) or treated with 2 μM 1NM-PP1 90 min after mitotic arrest in MG132 (exit, bottom two panels). Time in minutes from release into MG132 (top panels) or from 1NM-PP1 add-back (bottom panels) is shown. Scale bars indicate 10 μm. (C) Galleries of nuclei from HeLa cdk1as cells, imaged 4 hr after entry and exit treatments (1NM-PP1 concentrations in μM as indicated above the panels). For each panel, one hundred nuclei were randomly chosen by Olympus SCANR software. Interphase nuclei appear rounded and enlarged; mitotic nuclei are rod-shaped and condensed. The red lines indicate the borders upon which 1NM-PP1 is becoming active for either entry or exit. (D) Quantifications of entry and exit experiments in HeLa cdk1as cells (see Figure S2 for similar results with U2OS cdk1as cells). The values are means of three biological repeats (N = 100 per repeat), and error bars indicate SD. (E) Temporal dynamics of mitotic entry at increasing 1NM-PP1 concentrations from live-cell imaging analysis. (F) Levels of mitotic cyclins in G2- and M-arrested cells analyzed by immuno-blotting with indicated antibodies. Each panel shows four steps of a serial 1:1 dilution of extracts.

Figure 3

Contribution of Cdk1 and PP2A:B55 Feedbacks to Hysteresis (A) Wiring diagram of the mitotic switch that forms the basis of the modeling for this paper. (B) Percentage of cells entering into (red) and staying in (blue) mitosis at different Cdk1 inhibitor concentrations. Control (left column), Wee1-inhibited (second column), Gwl-depleted (third column; see Figure S3A for confirmation of depletion and Figure S3B for a higher resolution of 1NM-PP1 concentrations for this condition), and both Wee1-inhibited and Gwl-depleted (right column) cells are shown, with means of 3 or 4 biological repeats (N = 100 per repeat) and errors bars indicating SD (see Figure S3C for comparison of IC₅₀s for the different experimental conditions and Figure S3D for time course data). (C) Signal-response curves for mitotic substrate phosphorylation in single cells as a function of inhibitor concentration. Steady-state substrate phosphorylation is high (red), low (blue), or intermediate (orange). (D) Simulated population response of mitotic cells assuming log-normal CycB distribution among individual cells (see Figure S3F for our estimation of cyclin B concentration in the population, on which this model was based).

Figure 4

Testing Predictions of the Model (A) Explanation of the synchronization protocol to obtain cells arrested in prophase (see Figure S4A for a simulation of this experiment based on the model). (B) Images of cyclin B1-GFP-tagged HeLa cdk1as cells released from 1NM-PP1 arrest and retreated with 1 μM 1NM-PP1 25 min after release (see Figure S4B for characterization of the GFP knockin cells). Time after re-addition of 1NM-PP1 is indicated in minutes; scale bars correspond to 20 μm. Left panels show cells moving back to G2 (chromosome decondenses and cyclin B is exported and cells flatten); middle panels show cells moving to mitotic arrest (chromosome condensation, nuclear envelope breakdown, and cell rounding); right panels show cells that remain in prophase (chromosomes partially condensed, cyclin B remains nuclear, and cells remain rounded up). See also Video S2. (C) Measurements of cytoplasmic versus nuclear cyclin B1 ratio in single cells (top panels). Quantification of cell-cycle states in cells treated as described in (B) 4 hr after 1NM-PP1 addition (bottom). N > 20 in three independent experiments, error bars indicate SD. (D) Representative images of cells fixed 4 hr after 1NM-PP1 re-addition and stained by immuno-fluorescence with indicated antibodies. Scale bar, 10 μM. (E) Quantification of Cdk1 substrate phosphorylation, Cdk1 Y15 phosphorylation, and nuclear surface of cells in the G2/M and prophase steady states (data are from three independent experiments; N > 40 per experimental repeat; error bars indicate SD; p values were assessed using an independent two-sample t test; significant [^∗∗∗p < 0.001; ^∗p < 0.05] p values are indicated). (F) Images of HeLa cells progressing through mitosis after combined Gwl depletion and Wee1 inhibition. Time intervals are indicated in hr:min; scale bars indicate 10 μm; see also Videos S3 and S4. (G) Quantification of mitotic phenotypes from live-cell imaging in HeLa cells. (H) Cell-cycle profiles of MDA MB 231 cells following inducible Gwl knockout (see Figure S4C for confirmation of depletion) and Wee1 inhibition based on Edu (replicating population) and PI (DNA content) staining and FACS analysis. (I) Average values for Edu-positive and negative S phase populations of MDA MB 231 cells after Gwl depletion and or Wee1 inhibition. Values are from three independent experiments; N = 10,000 per experiment; error bars indicate SD (see Figure S4D for quantification of all cell-cycle phases). (J) Proliferation assays following increasing time of Wee1 inhibition in MDA MB 231 cells following Ctr and Gwl gRNA/Cas9 expression. Cells were counted based on Hoechst staining and nuclear segmentation 6 days after MK1775 pulse treatment. Values are from three independent experiments; N > 500 per experiment; error bars indicate SD. p values were assessed using an independent two-sample t test. Significant (^∗∗∗p < 0.001) or non-significant (p > 0.05) p values are indicated. (K) Colony-formation assays in MDA MB 231 cells following Ctr and Gwl gRNA/Cas9 expression and 24-hr incubation with MK1775 at the indicated concentrations. 5,000 cells were plated into each 6-well plate and stained and imaged 2 weeks after MK1775 pulse treatment. See also Figure S4.