Cyclin D-Cdk4,6 Drives Cell-Cycle Progression via the Retinoblastoma Protein's C-Terminal Helix

Abstract

The cyclin-dependent kinases Cdk4 and Cdk6 form complexes with D-type cyclins to drive cell proliferation. A well-known target of cyclin D-Cdk4,6 is the retinoblastoma protein Rb, which inhibits cell-cycle progression until its inactivation by phosphorylation. However, the role of Rb phosphorylation by cyclin D-Cdk4,6 in cell-cycle progression is unclear because Rb can be phosphorylated by other cyclin-Cdks, and cyclin D-Cdk4,6 has other targets involved in cell division. Here, we show that cyclin D-Cdk4,6 docks one side of an alpha-helix in the Rb C terminus, which is not recognized by cyclins E, A, and B. This helix-based docking mechanism is shared by the p107 and p130 Rb-family members across metazoans. Mutation of the Rb C-terminal helix prevents its phosphorylation, promotes G1 arrest, and enhances Rb's tumor suppressive function. Our work conclusively demonstrates that the cyclin D-Rb interaction drives cell division and expands the diversity of known cyclin-based protein docking mechanisms.

Keywords: Cdk; E2F; G1/S; Rb; cell-cycle regulation; cyclin; docking; kinase; phosphorylation.

Figure 1.. Cyclin D-Cdk4,6 Complexes Target Rb for Phosphorylation by Docking a C-Terminal Helix

(A) Schematic of the cyclin-Cdk docking sites and the 14 accessible Cdk phosphorylation sites on Rb. (B) Reported interactions and mutations of cyclin-Cdk docking sites on Rb. (C) In vitro kinase assays using the denoted cyclin-Cdk complexes and histone H1 or variants of Rb. Rb^RxLmut lacks all C-terminal RxL sequences that dock cyclin hydrophobic patches. Rb^{LxCxE cleft mut} lacks the LxCxE docking cleft. Rb^ΔC-term lacks the C-terminal amino acids 893–928. The Coomassie-stained gels showing equal amounts of substrate used in each reaction are placed below each autoradiograph. (D and E) Quantification of H1 kinase assays (D) and Rb kinase assays (E) using the denoted cyclin-Cdk complexes. Data are mean ± SEM; n = 2. (F) The ratio of kinase activity of the denoted cyclin-Cdk complexes toward Rb and Rb^ΔC-term. Data are mean ± SEM; n = 2. (G) Helical wheel projection of the predicted Rb C-terminal helix. The black and gray arrows indicate the beginning and end of the helix, respectively. Circles represent hydrophilic residues, diamonds represent hydrophobic residues, triangles represent potentially negatively charged residues, and pentagons represent potentially positively charged residues. Colored amino acids correspond to alanine or proline substitutions in (H). (H) In vitro kinase assays of Rb protein variants by cyclin D1-Cdk4. ΔCdk denotes Rb lacking the 14 Cdk phosphorylation sites. ΔHelix denotes Rb lacking the C-terminal helix amino acids 895–915. Helix mut. denotes an Rb variant where the predicted docking interface residues F897, L901, and R908 are substituted with alanines. Data are mean ± SEM; n = 2. (I) Immunoblot analysis of a GST-pull-down binding assay. The GST-tagged Rb bait proteins were incubated with 3X FLAG-tagged cyclin D1 prey proteins. * Denotes a degradation product below full-length GST-Rb. (J) Quantification cyclin D1-Cdk4 pulled down in (F). Data are mean ± SEM; n = 3. (K and L) Quantification of in vitro kinase assays of the denoted Rb variant with (K) cyclin D1-Cdk4 or (L) cyclin E1-Cdk2. Data are mean ± SEM; n = 3. (M) Comparison of the docking mechanisms recognized by cyclin D-Cdk4,6 and cyclin E-Cdk2.

Figure 2.. The Rb C-Terminal Helix Is Sufficient to Recruit Cyclin D1-Cdk4,6 Complexes for Phosphorylation

(A) Schematic of the engineered GST-Cdk phosphorylation site fusion protein containing a GST tag and the Rb amino acids 775–787 containing a single Cdk site fused to either no docking site, the Rb C-terminal helix docking site (+Helix), the Rb C-terminal helix docking site with the three interface residues substituted with alanines (+Helix mut.), or the Rb C-terminal helix docking site in reverse (+Helix rev.). (B and C) In vitro kinase assays of the indicated engineered GST-Cdk phosphorylation site fusion protein by (B) cyclin D1-Cdk6 or(C) cyclin E1-Cdk2. Data are mean ± SEM; n = 3. (D and E) Schematic of docking interaction between cyclin D-Cdk4,6 complexes and (D) the engineered GST-Cdk phosphorylation site fused to the Rb C-terminal helix docking site in comparison to (E) mutant versions of the engineered GST-Cdk phosphorylation site fusion protein.

Figure 3.. D-Type Cyclins Recognize the Rb C-Terminal Helix

(A) In vitro kinase assays with all six cyclin D-Cdk4,6 complexes and the denoted Rb variants. The Coomassie-stained gels showing equal amounts of substrate used in each reaction are placed below each autoradiograph. Fold changes ± SEM of wild type to Helix mutant phosphorylation are: 30 ±10for cyclin D1-Cdk4; 19 ± 3 for cyclin D2-Cdk4; 90 ± 20 for cyclin D3-Cdk4; 17 ± 6 for cyclin D1-Cdk6; 19 ± 8 for cyclin D2-Cdk6; 30 ± 20 for cyclin D3-Cdk6. Representative experiments shown (out of two independent experiments). (B) In vitro kinase assays of the denoted Rb variant with cyclin D1-Cdk6 or cyclin D1^{HP mut.}-Cdk6. The Coomassie-stained gels showing equal amounts of substrate used in each reaction are placed below each autoradiograph. Representative experiments shown (out of two independent experiments). (C) In vitro kinase assays using the denoted cyclin D1-Cdk6 complexes with GST-Rb775–787(S780)+Helix docking or GST-Rb775–787(S780)+RxL docking. Cyclin D1^HPmut.- denotes mutation of the hydrophobic patch on cyclin D1. Cyclin D1^{LxCxE mut} denotes mutation of the LxCxE motif on cyclin D1. The Coomassie-stained gels showing equal amounts of substrate used in each reaction are placed below each autoradiograph. Representative experiments shown (out of three independent experiments). (D and E) Quantification of the phosphorylation of the engineered GST-Cdk phosphorylation site fusion protein containing helixdocking from Rb (+Helix) or RxL docking from Cdc6 (+RxL) by wild type cyclin D1-Cdk6 versus (D) cyclin D1^{HP mut} -Cdk6 or (E) cyclin D1^{LxCxE mut} -Cdk6. Data are mean ± SEM; n = 3. (F) In vitro kinase assays with cyclin D1 fused to Cdk2. GST-Rb775–787(S780) without docking or GST-Rb775–787(S780)+Helix docking were used as substrates. The Coomassie-stained gel showing equal amounts of substrate used in each reaction is placed below the autoradiograph. Representative experiments shown (out of two independent experiments). (G) Schematic of the docking interactions between cyclin D1, D2, and D3 complexes and their substrates. Cyclins D1–D3 are capable of recognizing substrate proteins through helix-based docking, independent of the Cdk present in the complex.

Figure 4.. A C-Terminal Docking Helix Is Present in the Metazoan Rb Protein Family

(A–D) Helical wheel projection of the predicted C-terminal helix for human (A) p107 and (C) p130, and in vitro kinase assays of human (B) p107 and (D) p130 variants by cyclin D1-Cdk4. DC-term denotes truncating p107 after amino acid position 1014 and p130 after amino acid position 1122. Helix mut. denotes alanine substitution of the predicted docking interface residues indicated in blue and green in (A) and (C). Data are mean ± SEM; n = 2. (E) C-terminal helix sequence motif generated using 682 sequences of metazoan Rb family members. In the logo, blue denotes hydrophilic residues, green denotes neutral amino acids, and black denotes hydrophobic amino acids. Below the logo, C-terminal helix residues for Rb, p107, and p130 are aligned. The C-terminal helix residues that were tested for importance to cyclin D-docking are colored according to Figure 1G and (A) and (C). (F) A C-terminal Rb helix was not found outside metazoan proteome sequences.

Figure 5.. The Cyclin D-Cdk4,6-Rb Interaction Promotes the G1/S Transition, Rb Dissociation from Chromatin, and E2F1 Activation

(A) Map of PiggyBac integration constructs containing doxycycline-inducible human RB1 gene fused to fluorescent Clover and 3X FLAG affinity tag sequences. (B) Composite phase contrast and Clover fluorescence images showing expression of Clover-3X FLAG-Rb in HMECs with and without 500 ng/mL doxycycline. (C-E) Cell-cycle analysis by EdU incorporation and DAPI staining (data are mean ± SEM; n = 2) (C), cell size in G1 (representative experiments shown, out of three independent experiments) (D), and amount of chromatin-bound Rb in G1 (representative experiments shown, out of three independent experiments) (E) in HMECs expressing Clover-3FLAG alone (empty), wild type Rb, or the indicated mutant Rb following a 48-h induction with doxycycline (500 ng/mL). Rb^{LxCxE cleft mut.} lacks the LxCxE docking cleft. Rb^{Hellx mut.} denotes an Rb variant where the predicted docking interface residues F897, L901, and R908 are substituted with alanines. Rb^{LxCxE cleft+Helix mut.} contains both indicated mutations. Rb^ΔCdk lacks the 14 Cdk phosphorylation sites. Listed in the upper right of each histogram are median FSC for (D) and Clover-3FLAG-Rb for (E). (F) Schematic of Rb phosphorylation time course experiment. T98G cells were arrested by serum starvation (−FBS), Rb was induced (+dox), and samples were collected at 0, 4, 8, 12, and 24 h after release (+10% FBS, +dox). (G) Immunoblot analysis of lysates from Rb phosphorylation time course described in (F) with the denoted antibodies. Representative experiments shown (out of three independent experiments). (H and I) Quantification of (H) phospho-Rb(807/811)over total Clover-3FLAG-Rband (I) E2F1 from Rb phosphorylation time course in (G). Data are mean ± SEM; n = 3.

Figure 6.. Disruption of the Cyclin D-Cdk4,6-Rb Interaction Slows Tumor Growth

(A) Schematic of mouse experiment. PiggyBac integration constructs containing doxycycline-inducible mouse Rb1 fused to fluorescent Clover and 3FLAG affinity tag sequences were transfected into Kras^+/G12D; Trp53^−/− mouse pancreatic ductal adenocarcinoma cells (CKP-2167). Approximately 1 million CKP-2167 cells expressing variants of doxycycline-inducible mouse Rb were allografted by subcutaneous implantation. After 5 days of engraftment and growth, mice were given water supplemented with doxycycline (2 mg/mL) for 2 weeks. (B) Fold change in tumor volume compared to day 0 were calculated from caliper measurements. Data are mean ± SEM. A table of p values for all fold-change comparisons can be found in Figure S7B. n = 12 for empty, Rb^WT, Rb^{Helix mut.}, and Rb^ΔCdk. (C) Median tumor weight and interquartile ranges 15 days after doxycycline induction of either empty vector, wild type Rb, or Rb^{Helix mut.} n.s., *p > 0.05, **p ≤ 0.05, and ***p ≤ 0.001. n = 4 for empty, n = 10 for Rb^WT, and n = 10 for Rb^{Helix mut.}

Figure 7.. Cyclin D-Cdk4,6 Inactivates Rb through Multiple, Specific Docking Interactions to Drive Cell-Cycle Progression

(A) Schematic of the multiple docking interactions between cyclin D and Rb. The LxCxE cleft in the Rb pocket interacts with an LxCxE motif at the N terminus of cyclin D. The RxL motifs on the C terminus of Rb interact with the hydrophobic patch (HP) of cyclin D. The C-terminal helix of Rb interacts with an unknown part of cyclin D. Together, these interactions contribute to how the Cdk4 and Cdk6 active sites target the 14 accessible Cdk sites on Rb. (B) A model for the major functions of cyclin D-Cdk4,6 complexes. Cyclin D-Cdk4,6 docks, phosphorylates, and inactivates Rb to promote S Phase. Cyclin D-Cdk4,6 also promotes cell proliferation and survival through phosphorylation of other substrate proteins.