Cdc14 and PP2A Phosphatases Cooperate to Shape Phosphoproteome Dynamics during Mitotic Exit

Abstract

Temporal control over protein phosphorylation and dephosphorylation is crucial for accurate chromosome segregation and for completion of the cell division cycle during exit from mitosis. In budding yeast, the Cdc14 phosphatase is thought to be a major regulator at this time, while in higher eukaryotes PP2A phosphatases take a dominant role. Here, we use time-resolved phosphoproteome analysis in budding yeast to evaluate the respective contributions of Cdc14, PP2A^Cdc55, and PP2A^Rts1. This reveals an overlapping requirement for all three phosphatases during mitotic progression. Our time-resolved phosphoproteome resource reveals how Cdc14 instructs the sequential pattern of phosphorylation changes, in part through preferential recognition of serine-based cyclin-dependent kinase (Cdk) substrates. PP2A^Cdc55 and PP2A^Rts1 in turn exhibit a broad substrate spectrum with some selectivity for phosphothreonines and a role for PP2A^Rts1 in sustaining Aurora kinase activity. These results illustrate synergy and coordination between phosphatases as they orchestrate phosphoproteome dynamics during mitotic progression.

Keywords: Cdc14; PP2A; cell cycle; mitotic exit; phosphatases; phosphoproteomics.

Graphical abstract

Figure 1

Cdc14, PP2A^Cdc55, and PP2A^Rts1 All Contribute to Mitotic Exit Progression (A) Control and cdc14^degron cells were arrested in metaphase by Cdc20 depletion and then released to progress through synchronous mitosis following Cdc20 reinduction. α factor was added to arrest the cells following completion of mitotic exit in G1. Cell-cycle progression was monitored by fluorescence-activated cell sorting (FACS) analysis of DNA content. Protein extracts were prepared at the indicated times and processed for western blotting against the indicated proteins. (B) The fraction of cells with long anaphase (≥2 μm) spindles was scored in aliquots from the experiments in (A), (D), and (E). The mean ± SD of three independent experiments is shown. One hundred cells were scored at each time point in each experiment. (C) Cdc28 was immunoprecipitated at the indicated times and its associated kinase activity against histone H1 was measured in control and mutant cells. A representative autoradiogram and quantification of H1 phosphorylation relative to time point 0 of three independent experiments is presented. The means ± SD are shown. See also Figure S1B for the cell-cycle progression analysis by FACS analysis of DNA content (D) As in (A), but swe1Δ and swe1Δ cdc55Δ cells were used (E) As in (A), but swe1Δ and swe1Δ rts1Δ cells were used. See also Figure S1 for characterization of the cdc14^degron allele, an abundance analysis of the three phosphatases, and characterization of PP2A^Rts3, as well as Figure S2 for cell-cycle analyses following synchronization in G1.

Figure 2

Evidence for Substrate Specificity and Overlap of Cdc14, PP2A^Cdc55, and PP2A^Rts1 (A) Control and cdc14^degron cells were arrested and released as described in Figure 1A. Protein extracts were prepared at the indicated times from strains in which Ask1, Orc6, or Cbk1 were fused to an HA epitope tag. A representative FACS analysis of DNA content is shown. (B) As in (A), but swe1Δ and swe1Δ cdc55Δ cells were used. (C) As in (A), but swe1Δ and swe1Δ rts1Δ cells were used.

Figure 3

Phosphoproteomics Reveals Phosphatase Contributions to Mitotic Exit (A) Schematic of the experiment using TMT10plex to analyze phosphoproteome changes in a time-resolved manner. Two TMT10plex sets were used to generate each experimental dataset. Each isobaric mass tag is represented by a different color. One set was used to label the control and the mutant samples at times 0, 10, 20, 30, and 60 min. The other set was used to label samples at 5, 15, 25, 40, and 90 min. After mixing, phosphopeptide enrichment and liquid chromatography-tandem mass spectrometry (LC-MS/MS) were performed. (B) Table summarizing the three experimental datasets. See Data S1, S2, and S3 for all phosphosite intensities in the cdc14^degron, cdc55Δ, and rts1Δ datasets, respectively. The numbers before and after filtration for singly phosphorylated sites with a localization probability score of > 0.75 are indicated. The number of represented proteins is indicated in parentheses. (C) Overlap of phosphosites between the three experimental datasets. (D) Normalized median intensity profile and distribution of the central 90% of the phosphosites that undergo a 1.5-fold decrease in phosphorylation abundance through mitotic exit in the control. The respective controls are in black, cdc14^degron in red, cdc55Δ in blue, and rts1Δ in green. The number of phosphosites in each category is indicated. (E) As in (D), but phosphosites that show a 1.5-fold abundance increase in the control are shown. (F) As in (D), but phosphosites undergoing a transient 1.5-fold decrease in the control are shown. (G) As in (D), but phosphosites undergoing a transient 1.5-fold increase in the control are shown. See also Figure S3 for further details of the phosphoproteome analysis, including heatmap and cluster analysis, as well as Data S1, S2, and S3 (sheets 2–8) for a full list of the phosphosites.

Figure 4

Cdc14 Enacts Dephosphorylation Order by Targeting the Cdk Signature Motif (A) Normalized mean intensity profiles of phosphosites dephosphorylated during mitotic exit in the control strain. Sites were classified according to dephosphorylation timing (left). The same group of phosphosites in the cdc14^degron strain are plotted (right). The number of sites in each group is indicated. (B) As in (A), but comparing control and cdc55Δ datasets. (C) As in (A), but comparing control and rts1Δ datasets. (D) Normalized median intensity profiles over time of phosphosites dephosphorylated in the control strain that adhere to the three indicated kinase consensus motifs (left). The same phosphosites in the cdc14^degron strain are plotted (right). The number of sites in each group is indicated. (E) As in (D), but comparing the control and cdc55Δ datasets. (F) As in (D), but comparing the control and rts1Δ datasets. See also Figure S4 and Data S4 for a biological replicate of this analysis and for the temporal analysis of protein phosphorylation.

Figure 5

PP2A^Cdc55 Shapes an Anaphase-Specific Phosphorylation Wave (A) Profile plot of six Net1 phosphosites transiently phosphorylated in the control (black) and hyperphosphorylated in the cdc55Δ strain. The same phosphosites are also plotted in rts1Δ and cdc14^degron cells. (B) Profile plot of three Net1 phosphosites transiently phosphorylated in the control (black) as well as in the cdc55Δ strain. (C) 29 sites identified by correlation analysis in control and cdc55Δ cells. (D) IceLogo motif analysis of the 29 phosphosites identified in (C). The phosphorylated residue is at position 0. Larger letter size indicates increasing enrichment; the threshold for enrichment detection was p = 0.01. (E) Median intensity profile and interquartile range of transiently phosphorylated phosphosites that adhere to the two indicated kinase consensus motifs. (F) Median phosphosite abundance over all 10 time points in control, cdc55Δ, and rts1Δ cells, grouped by phosphoacceptor amino acid. 2,948 and 3,491 serine sites and 711 and 880 threonine sites entered the analysis from the cdc55Δ and rts1Δ datasets, respectively. cdc55Δ^∗p = 0.038; ^∗∗p = 0.0002. rts1Δ^∗p = 0.0103; NS, not significant, unpaired t test. See also Figure S5 for design of the correlation analysis, full list of phosphosites identified, and global phosphosite abundance in the cdc14^degron dataset.

Figure 6

PP2A^Cdc55 and PP2A^Rts1 Sustain NDR and Aurora Kinase Motif Phosphorylation (A) Median intensity profile and interquartile range of 264 phosphosites that are stable or phosphorylated in control, but dephosphorylated in rts1Δ cells. (B) IceLogo motif analyses of these sites; the threshold for enrichment detection was p = 0.01. (C) Median intensity profile and interquartile range of 166 phosphosites from (A) that are present in the cdc55Δ dataset. (D) Median intensity profile and interquartile range of 259 phosphosites that are stable or phosphorylated in control, but dephosphorylated in cdc55Δ cells. (E) IceLogo motif analyses of these sites; the threshold for enrichment detection was p = 0.01. (F) Median intensity profile and interquartile range of 174 phosphosites from (D) that are present in the rts1Δ dataset. See also Figure S6 for the identification of PP2A^Cdc55 and PP2A^Rts1 targets.

Figure 7

Phosphatase Interplay Promotes Timely Mitotic Progression (A) Profile plot of Cdc14 phosphosites in control, cdc55Δ, and rts1Δ cells. (B) A physical interaction between Cdc14 and PP2A^Cdc55. Cells were synchronized in metaphase and released. Protein A-tagged Cdc14 was precipitated from cell extracts at the indicated times and coprecipitation Cdc55 was analyzed by western blotting. Cell-cycle progression was monitored by FACS analysis of DNA content. (C) Phosphatase-substrate interactions. Protein A-tagged Cbk1 was precipitated and the coprecipitation of PP2A^Cdc55 was analyzed in extracts from metaphase arrested cells with or without Cdc14 depletion following auxin (IAA) addition. (D) As in (C), but protein A-tagged Orc6 was precipitated. (E) Protein A-tagged Cbk1 was precipitated and the coprecipitation Cdc14 was analyzed in extracts from metaphase-arrested control or cdc55Δ cells. (F) As in (E), but protein A-tagged Orc6 was precipitated. (G) Mitotic progression of control cdc14^degron and cdc55Δ cdc14^degron cells. Cell-cycle progression was monitored by FACS analysis of DNA content and western blotting against the indicated proteins. See also Figure S7 for analysis of Igo2 and Zds2 phosphosites, Cdc14 phosphorylation by electrophoretic mobility shift, and mitotic exit kinetics of rts1Δ cdc14^degron cells.