Analysis of Cyclin Specificity with Respect to RXL Docking Sites
(A) Equal amounts of cyclin-Cdk1 complexes were used in a phosphorylation assay with purified T33-Sic1ΔC, T33-Sic1ΔC-1234rxl, and T33H1-Sic1ΔC. The first two mutant proteins contained the original sequence of T33, QA³³TPQKPSQNL, while the last contained the Histone H1 peptide sequence PK³³TPKKAKKL in place of the T33 site of T33-Sic1ΔC.
(B–D) Quantified specificity profiles showing the k_cat/K_M values obtained from the experiments shown in (A), using the indicated cyclin-Cdk1 complexes to phosphorylate T33-Sic1ΔC (B), T33H1-Sic1ΔC (C), or T33-Sic1ΔC-1234rxl (D). In the case of Clb cyclins, activity was also measured with hydrophobic patch-mutated cyclins (hpm, black bars). The dark gray bars in (B) and (C) denote the relative hp-dependent docking effect, and these values are bound to the right-hand scale of the Y axes.
(E) The k_cat/K_M profiles for the control substrate, the H1-based peptide PKTPKKAKKL.
(F) The impact of different RXL motif-bearing docking sites on the phosphorylation specificity of T5, T33, and S76 was studied with Cln2-Cdk1 and Clb5-Cdk1, using substrate constructs with a single RXL motif left unmutated: Sic1ΔC, Sic1ΔC-234rxl, Sic1ΔC-134rxl, Sic1ΔC-124rxl, Sic1ΔC-123rxl, and Sic1ΔC-1234rxl.

Cyclins Differentially Modulate the Phosphorylation Site Consensus Sequence of Cdk1
(A) Differential effect of lysines on the specificity profiles of Cln2-Cdk1 (open bars) or Clb2-Cdk1 (dark bars), using T33-Sic1ΔC variants with lysine introduced at various positions within the sequence QA³³TPQAPSQ.
(B) Differential effect of proline on the specificity profiles of Cln2-Cdk1 (open bars) or Clb2-Cdk1 (dark bars), using T33-Sic1ΔC variants with proline introduced at various positions within the sequence QA³³TPQAASQ.
(C) Examples of Cln2 versus Clb5 cyclin specificity, using different T33-Sic1ΔC and T33-Sic1ΔC-1234rxl variants bearing the motifs with highest differential specificity. The impact of LP motif-dependent specificity of Cln2 was determined with the LP competitor peptide. Red letters indicate the amino acid substitutions into different positions of the T33 site.
(D) Cln2 versus Clb5 phosphorylation consensus site specificity profiles showing the docking-site-independent k_cat/K_M values for eight different physiological phosphorylation sites of Sic1. Sic1ΔC constructs in which only the indicated single phosphorylation site is left unmutated were used in the phosphorylation assays. Cln2-Cdk1 combined with the LP competitor peptide and Clb5hpm-Cdk1 were used to eliminate the impact of the docking sites.

Phosphorylation Screen of Cyclin-Specific Cdk1 Targets
Purified protein substrates were tested in a standard kinase assay using four representative complexes of cyclin-Cdk1. The apparent k_cat/K_M values are provided in Table S2.
(A) Type I substrates specific for Cln2-Cdk1. Lower panels indicate activity in the presence of the competitor peptide based on the LP docking site of Sic1.
(B) Type II substrates specific for Clb5- and Clb3-Cdk1. Lower panels indicate activity with the hpm versions of each Clb.
(C) Type III substrates showing hp-dependent specificity for Clb3 and no specificity for Clb5 and/or Clb2.
(D) Gradual specificity profile of the “mitotic targets” (Type IV substrates) presented together with selected examples of Clb-hpm-Cdk1 profiles. See also Table S2.

Intrinsic Specificity of Cyclin-Cdk1 Complexes toward the Histone H1-Based Model Substrate PKTPKKAKKL Increases Gradually in the Series Cln2-Clb5-Clb3-Clb2, Correlating with Cell Cycle Progression
(A) Western blotting analyses of endogeneous cyclin levels in the course of the cell cycle. The time points were taken after the release of cells from α factor-induced G1 arrest. The strains carried the indicated cyclin with a C-terminal 3HA tag.
(B) Michealis-Menten curves of Histone H1 peptide (PKTPKKAKKL) phosphorylation by four different cyclin-Cdk1 complexes. Arrows indicate the change of Cdk1 kinetic properties during cell cycle progression.
(C) Steady-state kinetic parameters determined for four representative cyclin-Cdk1 complexes using Histone H1-based peptide PKTPKKAKKL as a substrate.
(D) Silver-stained SDS gel showing the purified preparations of cyclin-Cdk1 complexes used in this study.
(E) Analyses of the levels of activating phosphorylation at T169 and inhibitory phosphorylation at Y19 of cyclin-Cdk1 preparations. Equal protein amounts of each cyclin-Cdk1 preparation were loaded and blotted using specific antibody for Cdk1 (upper panel). The second panel shows the active site titration of cyclin-Cdk1 preparations using an excess of purified Swe1. Equal protein amounts of cyclin-Cdk1 preparations were phosphorylated with Swe1 until Cdk1 was totally inactivated. The third panel shows the relative levels of pY19 initially present in the preparations, determined using western blotting with specific antibody. The lower panel shows western blotting to determine the relative levels of activating phosphorylation at T169 (see the Supplemental Experimental Procedures for details).
(F) iTRAQ four-plex analysis of the relative levels of activating phosphorylation at T169 in cyclin-Cdk1 preparations. Cyclin/Cdk1 complexes were separated by 10% PAGE, and the Cdk1 band was excised and digested with trypsin. Eluted peptides were labeled with iTRAQ reagents and analyzed by tandem mass-spectrometry (see the Supplemental Experimental Procedures for details). Intensities of four reporter ions at m/z 114, 115, 116, and 117 are presented. Analyses were performed in triplicate, and error bars indicate standard error of the mean.

A Hydrophobic Stretch in Sic1 Enhances Cln2-Specific Phosphorylation Relative to Clb-Cdk1
(A and B) A 10 amino acid stretch that enhances the Cln2 specificity was located by systematic truncation of Sic1.
(C) A five alanine mutation in the docking site of Sic1 (vllpp mutation) reduced the phosphorylation specificity for Cln2-Cdk1 but not for Clb5-Cdk1. Similar results were obtained using a competitor peptide based on the 10 amino acid stretch (the LP peptide, VLLPPSRPTS) between residues 136 and 145 as indicated in (A).
(D) The LP peptide is equally effective in the inhibition of Cln2 specificity for sites T5, T33, and S76, as demonstrated with Sic1ΔC-based constructs described in Figure 2F.
(E) The rate of Cln2-dependent phosphorylation of Sic1Δ can be reduced by addition of either RXL- or AXL-containing competitor peptides (based on the docking sequence of RVNRILFPT of Sic1: RVNRILFPT and RVNAIAFPT), whereas only the RXL peptide reduces the specificity of Clb5.
(F) A scheme showing the observed docking interactions between Sic1 and Clb5 or Cln2. The dashed arrows indicate the mild cooperative effect of the leucines of the RXL motifs on Cln2 specificity.

Analysis of Changes in Cdk1 Substrate Specificity during the Cell Cycle and of the Physiological Importance In Vivo of Different Substrate Docking Sites
(A and B) Cells constitutively expressing the indicated Sic1ΔC-3HA construct were released from α factor arrest, and at the indicated time points Phos-Tag SDS-PAGE and western blotting were used to assess phosphorylation state (see Figure S1). The constructs contained either the Clb2-specific P-X-T-P-X-K motif (A) or the Cln2-specific P-X-T-P-K-X motif (B) at positions T5, T33, T45, and T76. Other Cdk1 sites were mutated to alanines, and T2 was left unchanged. As controls, we also tested the same constructs with mutations of the LP and the RXL docking sites. Sic1 bands on the western blot were scanned using the GelDoc (GE), and each phosphorylated species was quantified. The relative steady-state phosphorylation at each time point was calculated as a ratio of pS/S_T, where S_T is the total signal of Sic1ΔC-3HA and pS is the total signal of phosphorylated bands multiplied by the number of phosphates each band contains. See also Figure S1. Analyses were performed in duplicates, and error bars indicate standard error of the mean.
(C and D) The importance of Clb5-specific RXL motifs (C) and the Cln2-specific LP motif (D) in the degradation of Sic1 was tested by overexpressing different Sic1 constructs under the GAL1 promoter. To assess the effect of the LP docking site, the strain used in (D) was sensitized by reducing Cln-Cdk1 activity in the cell by deleting Cln2.
(E) Western blotting analysis of Sic1 levels after the release of cells from a G1 arrest. Strains carrying CEN vectors with GAL-SIC1 or GAL-SIC1-1234rxl -vllpp were arrested in G1 with α factor for 2.5 hr. SIC1 expression was induced in the arrested cells by addition of galactose for 45 min. Both galactose and α factor were removed and the cells were released into glucose-containing medium. Sic1-3HA levels were analyzed at different time points using western blotting. In the lower panel the combined quantified Sic1 profiles from two independent experiments are presented. The error bars indicate standard error of the mean.
(F) To demonstrate how altered Cdk1 specificity toward Sic1 affects the timing of DNA replication, we performed flow cytometry of DNA content for cells taken at different time points of the α factor release experiment presented in (E).
(G) To explore further how altered Cdk1 specificity toward Sic1 affects the control of DNA replication, we performed flow cytometry of DNA content in asynchronous cells expressing the Sic1 mutant lacking cyclin docking motifs. Strains carrying CEN vectors with GAL-SIC1 or GAL-SIC1-1234rxl-vllpp were grown to log phase, and the expression of Sic1 was initiated by addition of galactose. Cells expressing Sic1 with mutations in the docking sites caused DNA rereplication (DNA > 2N). Cells expressing wild-type Sic1 did not exhibit any signs of rereplication.

Dynamics of Cdk1 Substrate Specificity during the Cell Cycle
(A) A schematic view of the dynamic changes of Cdk1 specificity during the cell cycle.
(B) Modified Michaelis-Menten formula with the highlighted inhibitory term describing how the Cdk1 substrate pool (S_TOT) can act as a competitive inhibitor relative to any particular substrate (S_X). The lower panel on the left side presents a schematic view of a hypothetical system based on sequential accumulation of three cyclins with gradually changing specificity, and the panel on the right side presents a similar system with a single mitotic cyclin (rising linear lines). The gradual phosphorylation and decrease of the unphosphorylated Cdk1 substrate pool during the progression of the cell cycle is depicted as the declining black line. In the three-cyclin system, the inhibitory terms (1 + S_TOT/K_M,TOT) are kept relatively low and equal throughout the course of the cyclin synthesis, because as the substrate pool gets gradually phosphorylated S_TOT for each cyclin decreases in correlation with K_M,TOT. In the system with a single cyclin, the inhibitory term is high in the early stages of the cell cycle, potentially delaying timely phosphorylation of premitotic targets.