Logo of molcellbPermissionsJournals.ASM.orgJournalMCB ArticleJournal InfoAuthorsReviewers
Mol Cell Biol. 2004 Jun; 24(12): 5606–5619.
PMCID: PMC419867

Docking-Dependent Regulation of the Rb Tumor Suppressor Protein by Cdk4


Phosphorylation of target proteins by cyclin D1-Cdk4 requires both substrate docking and kinase activity. In addition to the ability of cyclin D1-Cdk4 to catalyze the phosphorylation of consensus sites within the primary amino acid sequence of a substrate, maximum catalytic activity requires the enzyme complex to anchor at a site remote from the phospho-acceptor site. A novel Cdk4 docking motif has been defined within a stretch of 19 amino acids from the C-terminal domain of the Rb protein that are essential for Cdk4 binding. Mutation or deletion of the docking motif prevents Cdk4-dependent phosphorylation of full-length Rb protein or C-terminal Rb fragments in vitro and in cells, while a peptide encompassing the Cdk4 docking motif specifically inhibits Cdk4-dependent phosphorylation of Rb. Cyclin D1-Cdk4 can overcome the growth-suppressive activity of Rb in both cell cycle progression and colony formation assays; however, while mutants of Rb in which the Cdk4 docking site has been either deleted or mutated retain growth suppressor activity, they are resistant to inactivation by cyclin D1-Cdk4. Finally, binding of Cdk4 to its docking site can inhibit cleavage of exogenous and endogenous Rb in response to distinct apoptotic signals. The Cdk4 docking motif in Rb gives insight into the mechanism by which enzyme specificity is ensured and highlights a role for Cdk4 docking in maintaining the Rb protein in a form that favors cell survival rather than apoptosis.

The mechanisms employed by the protein kinase family of enzymes to ensure substrate and phospho-acceptor site specificity are beginning to receive increased interest as it becomes apparent that, in general, the information contained within the consensus phosphorylation motif is not sufficient to explain the fidelity of a given kinase in vivo. The broad consensus motif at the phosphorylation site targeted by the cyclin-cyclin-dependent kinases (cyclin-CDKs) is T/SPXR/K (23, 52); however, in order to gain specificity, these enzymes participate in a direct interaction with their substrate at a site distinct from the phospho-acceptor site (1, 2, 35, 48). In addition, at least some cyclin-CDKs are able to phosphorylate a given substrate through an interaction with an associated binding protein. For example, E2F1 can function as a targeting factor for cyclin A-Cdk2-dependent phosphorylation of its heterodimeric partner DP1 (14, 30, 58). Thus, the cyclin-CDKs belong to a growing class of kinases that require a specific docking interaction for efficient site-specific phosphorylation of a substrate (5).

To date, the G1-associated kinase cyclin D-Cdk4 has been shown to phosphorylate only a limited subset of possible CDK target proteins (20, 28), suggesting that it has a strict requirement for determinants outside the phosphorylation motif and that these determinants are distinct from those required by other family members. There is good evidence that as well as providing increased specificity, cyclin D-Cdk4 substrate docking is required for the enzyme's maximal catalytic activity. This is illustrated by the fact that Cdk4 has very low specific activity against peptides containing consensus phosphorylation sites from the retinoblastoma (Rb) protein (41). The Rb protein, which is phosphorylated in an orchestrated manner as cells exit G0 and proceed through the cell division cycle, is the best-characterized substrate for Cdk4 (39). There are a total of 16 possible phospho-acceptor sites that fit the consensus for the cyclin-CDK family within the primary amino acid sequence of Rb. However, which sites are phosphorylated when and how distinct CDKs acquire specificity for a given site are still largely undefined. The initial phosphorylation of Rb coincides with activation of Cdk4 (45, 50). Evidence suggests that Cdk4-dependent phosphorylation displaces histone deacetylases from the Rb pocket and that this phosphorylation event is also required for subsequent modification by Cdk2-containing complexes (22, 36). The precise phospho-acceptor site specificity of Cdk4-containing complexes is contentious; however, there is a broad consensus that Cdk4 is primarily responsible for phosphorylation at a number of sites within the C-terminal domain of Rb (10, 20, 28, 41, 60).

Cyclin D1, like cyclin E and a number of other proteins, contains an LXCXE motif that can interact with the LXCXE binding domain within the Rb pocket (13, 27, 33, 42, 59). Early studies on the role of the LXCXE binding region of Rb suggested that this region was important for the regulation of Rb by cyclin D1-containing complexes (13, 16). However, subsequent studies have shown that a cyclin D1 mutant lacking the LXCXE motif retains the ability to phosphorylate Rb both in vitro and in cells and that this mutant is able to reverse the growth-inhibitory properties of Rb in intact cells (10, 24). Furthermore, expression of Rb mutant constructs which are unable to interact with LXCXE motif proteins induces cell cycle arrest that is overcome by cyclin D1-Cdk4 (11). When Dick et al. (11) looked at the pattern of cyclin D1-Cdk4-dependent phosphorylation in cells, they found no difference when the Rb pocket mutant was compared with wild-type Rb. These studies suggest that the LXCXE binding pocket plays a relatively minor role in the regulation of Rb by cyclin D1-Cdk4 or that its function can be replaced by other determinants within the sequence of the Rb protein. In fact, in vitro studies have suggested that a region within the C terminus of Rb is required for efficient cyclin D1-Cdk4-dependent phosphorylation. Thus, while an Rb protein containing Cdk4-specific phospho-acceptor sites but missing the LXCXE binding domain (amino acids [aa] 792 to 928) is efficiently phosphorylated by cyclin D1-Cdk4, an Rb mutant protein which has the LXCXE binding domain but not the extreme C terminus of Rb is a poor Cdk4 substrate (1).

In the present study, we define a novel Cdk4 docking motif within the C-terminal domain of the Rb protein. Mutation of the docking motif prevents stable binding of Cdk4 to the Rb protein and inhibits Cdk4-dependent modification of Rb at critical regulatory phospho-acceptor sites. Although cyclin D1-Cdk4 can overcome growth inhibition imposed by wild-type Rb, it cannot inactivate the growth suppressor activity of Rb which has mutations within the Cdk4 docking site, confirming that the C-terminal region of this tumor suppressor protein contains a functional docking site for the kinase. Furthermore, evidence is presented that Cdk4 binding can protect Rb from caspase-mediated cleavage. Together the data suggest that docking of Cdk4 to the C terminus of Rb may be critical for maintaining Rb in a form that favors cell growth rather than apoptosis.


Immunochemicals, peptides, and plasmids.

The following antibodies were used for immunoblot analysis: anti-Rb G3245 (BD Pharmingen), anti-Rb C15 sc-050-R (Santa Cruz), anti-Cdk4 sc-260 (Santa Cruz), anti-cyclin D1 sc-718 (Santa Cruz), anti-Rb IF8, and the PhosphoPlus Rb (Ser780, Ser795, and Ser807/Ser811) antibody kit (New England Biolabs). Peptides were supplied by Chiron Mimitopes and dissolved at 10 mg/ml in dimethyl sulfoxide (DMSO).

pCMV-hemagglutinin (HA)/cyclin D1 and pCMV-HA/Cdk4 were gifts from Xin Lu, and pCMV-CD20 was from David Lane. Alanine mutations were introduced into the C terminus of human Rb with a QuikChange site-directed mutagenesis kit (Stratagene) by following the manufacturer's instructions. In the plasmid pGEX-Rb/773-928 (carrying the C-terminal Rb construct comprising aa 773 to 928 [Rb/773-928]), mutations were introduced at Lys889 and Leu891 (mutant 1), Phe897, Gln889, and Lys900 (mutant 2), Lys900, Leu901, and Met904 (mutant 3), and Arg908 and Arg910 (mutant 4) and a stop codon was introduced at the site corresponding to Asp886 (mutant 5). In the plasmid pcDNA3.1-Rb, mutations were introduced at Leu891 and Phe897 (mutant Rb-LF) and Phe897 (mutant Rb-F) and a stop codon was introduced after the site corresponding to Asp886 [mutant Rb(1-886)].


cDNA from a kinase-dead Cdk4 construct (Cdk4dn) was subcloned from the pCMV-neo-Bamcdk4dn plasmid (a gift from S. van den Heuvel) into the BamHI site of the pFastBAC vector (Gibco Invitrogen Corporation). The Bac-to-Bac baculovirus expression system (Gibco Invitrogen Corporation) was used to generate a baculovirus stock by following the manufacturer's instructions. Baculovirus constructs for cyclin D1 and Cdk4 (gifts from C. Sherr), cyclin E and cdk2 (gifts from D. Morgan), cyclin D1 and Cdk4dn, or Cdk4dn alone were expressed in Sf9 cells. Cells were harvested after 48 h, pelleted by centrifugation, and resuspended in 2 volumes of lysis buffer (10 mM HEPES [pH 7.4], 10 mM NaCl, 2 mM dithiothreitol [DTT], 10 μg of leupeptin/ml, 4 μg of aprotinin/ml, 2 μg of pepstatin/ml, 10 μg of soybean trypsin inhibitor/ml, 400 μg of Pefabloc/ml, 1 mM EDTA, 1.2 mM benzamidine). Following incubation on ice for 15 min, the lysates were centrifuged at 13,000 × g for 15 min at 4°C and the supernatant was removed to form the enzyme preparation, frozen in liquid nitrogen, and stored at −70°C.

Rb, wild type or mutant, was translated in vitro using the TNT T7 Coupled Reticulocyte Lysate system (Promega) in the presence of 0.8 MBq of [35S]methionine (Amersham Pharmacia) and 1 μg of DNA template (pcDNA3.1-Rb) per reaction mixture, according to the manufacturer's guidelines.

Human Rb was expressed in Escherichia coli using pET9Rb, and cells were harvested by low-speed centrifugation, gently resuspended in 50 mM HEPES (pH 7.4) containing 10% (wt/vol) sucrose, and subjected to a freeze-thaw cycle in liquid nitrogen. The lysate was brought to the following composition: 1 M KCl, 2 mM DTT, 0.5 mg of lysozyme/ml, 10 μg of leupeptin/ml, 4 μg of aprotinin/ml, 2 μg of pepstatin/ml, 10 μg of soybean trypsin inhibitor/ml, 400 μg of Pefabloc/ml, 1 mM EDTA, and 1.2 mM benzamidine. The lysate was then incubated on ice for 15 min. Following centrifugation at 13,000 × g for 15 min, the supernatant was applied to an HQ-Porus 20 column equilibrated in column buffer (50 mM HEPES, pH 7.4, containing 5% [vol/vol] glycerol, 2 mM DTT, 0.1 mM EDTA, 0.01% [vol/vol] Triton X-100, 40 μg of Pefabloc/ml, and 1 mM benzamidine) plus 50 mM NaCl. The column was washed with 10 column volumes of column buffer plus 50 mM NaCl, and elution was performed with a linear gradient of 0.05 to 1 M NaCl in column buffer over 20 column volumes. Fractions containing Rb protein, as determined by immunoblotting, were pooled, concentrated, and dialyzed against column buffer containing 50 mM NaCl and then applied to an HS-Porus column equilibrated in column buffer containing 100 mM NaCl. The column was washed as described above and developed with a linear gradient of 0.1 to 1 M NaCl in column buffer. Rb-containing fractions were pooled, concentrated, and dialyzed against column buffer containing 50 mM NaCl.

Glutathione S-transferase (GST)-Rb was expressed in E. coli by using pGEX-Rb/773-928 (a gift from D. P. Lane). The protein was purified on glutathione-Sepharose beads (Pharmacia) by following the manufacturer's instructions. His-Cdk4 was expressed in E. coli and purified on nickel-agarose beads (Novagen) by following the manufacturer's instructions.

Kinase assays.

35S-labeled Rb protein, wild type or mutant, was incubated with Sf9 cell lysate (2 μg of total protein) expressing cyclin D1-Cdk4 or cyclin E-Cdk2 as indicated in the figure legends. Kinase buffer (100 mM HEPES [pH 7.4], 20 mM MgCl2, 5 mM EGTA, 20 mM β-glycerophosphate, 2 mM DTT, 20 μM protein kinase inhibitor, 2 mM NaF) with 100 μM ATP was added to a final volume of 10 μl, and the reaction mixture was incubated for 10 min at 30°C. The reaction was stopped by the addition of sodium dodecyl sulfate (SDS)-sample buffer, and results were analyzed using SDS-7% polyacrylamide gel electrophoresis (SDS-7% PAGE) and autoradiography.

GST-Rb/773-928, wild type or mutant, was incubated with Sf9 cell lysate (2 μg of total protein) expressing cyclin D1-Cdk4 or cyclin E-Cdk2, as indicated in the figure legends, in a final volume of 10 μl of kinase buffer with 100 μM ATP containing [γ-32P]ATP (250 cpm/pmol). The reaction mixture was incubated for 10 min at 30°C, the reaction was stopped by the addition of SDS-sample buffer, and results were analyzed by SDS-12% PAGE and autoradiography.

Transfection and cell cycle analysis.

Saos2 (human osteosarcoma cell line) and MDA MB231 (human breast cancer cell line) cells were grown in Dulbecco's modified Eagle's medium (DMEM; Invitrogen) with 10% (vol/vol) fetal bovine serum (Invitrogen) and 1% (vol/vol) penicillin-streptomycin in 10% (vol/vol) CO2. Cells were transfected at 70% confluency with DNA as indicated in the figure legends by using Lipofectamine 2000 (Invitrogen) and harvested after 24 h. Cells required for immunoblot analysis were lysed in 2 volumes of 25 mM HEPES, pH 7.6, containing 1% (vol/vol) NP-40, 150 mM KCl, 5 mM DTT, 50 mM NaF, 10 μg of leupeptin/ml, 4 μg of aprotinin/ml, 2 μg of pepstatin/ml, 10 μg of soybean trypsin inhibitor/ml, 400 μg of Pefabloc/ml, 1 mM EDTA, and 1.2 mM benzamidine for 15 min at 4°C. Following centrifugation at 14,000 × g for 15 min, the supernatant was removed, frozen in liquid nitrogen, and stored at −70°C. Samples were analyzed by immunoblot analysis (55). Cells for cell cycle analysis were washed in phosphate-buffered saline (PBS) containing 3 mM EDTA and detached with PBS containing 3 mM EDTA at 37°C for 5 min. Cells were collected by centrifugation, resuspended in growth medium with 20 μl of fluorescein isothiocyanate-CD20 (Becton Dickinson), and incubated at 4°C for 30 min. Cells were washed twice with PBS containing 1% fetal bovine serum and resuspended in 100 μl of PBS and 900 μl of ice-cold ethanol added dropwise and then incubated at 4°C for >2 h. Following centrifugation, 75 × 104 to 100 × 104 cells/ml were resuspended in a total volume of 2 ml of PBS containing 100 μl of propidium iodide (1 mg/ml) plus 100 μl of RNase (400 μg/ml) and incubated in the dark at 37°C for 30 min. Cell cycle analysis was carried out by counting a minimum of 10,000 transfected cells by using a FACScan (Becton Dickinson).

Colony formation assay.

Saos2 cells were transfected at 90% confluency. After 48 h, cells were trypsin treated and replated at different dilutions onto 10-cm-diameter dishes containing DMEM with 10% fetal bovine serum and 1.5 mg of Geneticin (Invitrogen)/ml added. The medium was changed after 3 days and then weekly with DMEM plus Geneticin. Colonies were counted after 15 days by fixing in methanol for 30 min and staining with 10% (vol/vol) Giemsa (Sigma).

Caspase 3 assay.

35S-labeled Rb (5 μl), generated by in vitro translation, or 1 μg of unlabeled GST-Rb/773-928 was incubated with 40 ng of purified recombinant human caspase 3 (Pharmingen) in a 10-μl reaction mixture (20 mM HEPES, pH 7.5, containing 10% [vol/vol] glycerol, 2.5 mM DTT, and 5 mM NaF) at 37°C for 30 min. The reaction was stopped by the addition of SDS-sample buffer, and results were analyzed by SDS-PAGE and autoradiography or immunoblot analysis with phospho-specific antibodies (49). The caspase inhibitor DEVD-CHO (Calbiochem) and Sf9 cell lysate (2 μg of total protein) expressing CDKs were added as indicated in the figure legends.

GST pull-down assay.

A kinase assay was carried out using GST-Rb/773-928 as described above. The GST-Rb/773-928 protein was pulled down by the addition of glutathione-Sepharose (Pharmingen) beads. After centrifugation and washing, the beads were resuspended in caspase buffer (20 mM HEPES [pH 7.6], 10% [vol/vol] glycerol, 2.5 mM DTT, 5 mM NaF) containing 40 ng of caspase 3 in the presence or absence of the caspase inhibitor DEVD-CHO (10 μM) for 30 min at 37°C. The reaction was stopped by the addition of SDS-sample buffer, and results were analyzed by SDS-12% PAGE and autoradiography.

Cdk4 binding assays.

Glutathione-Sepharose beads (30 μl of a 50% slurry washed three times in buffer containing 0.5% NP-40, 1 mM DTT, and 1× protease inhibitor [PI] mix) were added to GST-Rb/773-928 (1 μg) and 70 μl of pull-down buffer (0.5% NP-40, 1 mM DTT, 1× PI mix) for 30 min at 4°C. Following extensive washing, the beads were collected and resuspended in 150 μl of pull-down buffer containing purified His-Cdk4 (2 μg) and incubated for 2 h at 4°C. The beads were washed four times in pull-down buffer. Alternatively, in vitro-translated Rb, full-length or mutant (15 μl), and anti-Rb G3245 (1 μg) were incubated in 70 μl of pull-down buffer plus 3% bovine serum albumin (BSA) for 30 min at 4°C. Protein G beads (30 μl of a 50% slurry washed three times in buffer containing 3% BSA, 0.5% NP-40, 1 mM DTT, and 1× PI mix) were added for 1 h at 4°C. Following extensive washing in pull-down buffer plus 3% BSA, beads were collected and resuspended in 150 μl of pull-down buffer with Cdk4 (2 μg) and incubated for 2 h at 4°C. Beads were washed three times in pull-down buffer plus 3% BSA and one time in buffer. All beads were then resuspended in 10 μl of SDS-sample buffer and analyzed by SDS-12 or 7% PAGE and immunoblotting. Immunoblots were developed with anti-Cdk4 and anti-Rb IF8.


The C terminus of Rb is required for the phosphorylation of full-length Rb by Cdk4 in vitro and in cells.

Stoichiometric phosphorylation of the Rb protein by cyclin D-Cdk4 shows a strict requirement for substrate docking, and this finding is illustrated by the fact that isolated peptides containing consensus Cdk4 phosphorylation sites are very poor substrates for the holoenzyme complex (41). Furthermore, Rb constructs in which a portion of the extreme C terminus has been deleted are phosphorylated inefficiently despite containing target phosphorylation motifs (1). While the recognition sites for cyclin E- and cyclin A-Cdk2 on Rb are relatively well defined (1), it is not clear how the specificity of Cdk4 for a subset of possible CDK phospho-acceptor sites is maintained.

The cysteine protease caspase 3 cleaves Rb, removing a 5-kDa fragment (aa 887 to 928) from the C terminus (53). We therefore employed caspase-cleaved Rb to explore the importance of the C-terminal 42 aa for Cdk4-dependent phosphorylation. Full-length untagged Rb protein was translated in the presence of [35S]methionine by using reticulocyte lysate, and phosphorylation and/or caspase cleavage was detected by changes in the electrophoretic mobility of the labeled protein (Fig. (Fig.1A).1A). Full-length Rb was readily phosphorylated by cyclin D1-Cdk4, giving an increase in the apparent molecular mass of the substrate; however, caspase-cleaved Rb was not phosphorylated as no change in mobility was observed in the presence of cyclin D1-Cdk4 (Fig. (Fig.1A).1A). Lack of phosphorylation was not due to inhibition of cyclin D1 or Cdk4 by caspase 3 as the protease was inactivated using the specific caspase inhibitor DEVD-CHO prior to the addition of the kinase. DEVD-CHO does not affect the catalytic activity of Cdk4, as the presence of DEVD-CHO in the kinase assay did not prevent the phosphorylation of full-length Rb (Fig. (Fig.1A).1A). As cyclin E, like cyclin D1, contains an LXCXE motif (13, 16), we also looked at the ability of cyclin E-Cdk2 to phosphorylate C-terminally cleaved Rb. In this case, both full-length Rb and the caspase 3-cleaved protein were efficiently phosphorylated by cyclin E-Cdk2 (Fig. (Fig.1A).1A). The incorporation of 32P into a GST-C-terminal Rb fragment (Rb/773-928) fusion protein was used as an independent assay to verify that the C-terminal 42 aa are required for cyclin D1-Cdk4-dependent phosphorylation. Rb phosphorylation assays were carried out using cyclin D1-Cdk4 or cyclin E-Cdk2 normalized to specific activity against Rb/773-928. Under these conditions, cyclin E-Cdk2 phosphorylated both Rb/773-928 and caspase-cleaved Rb/773-928 (Fig. (Fig.1B)1B) to the same extent. However, phosphorylation of the caspase-cleaved fragment by cyclin D1-Cdk4 was barely detectable in this assay (Fig. (Fig.1B),1B), confirming the results obtained with a full-length Rb substrate. To determine the effect of the C terminus on the ability of cyclin D1-Cdk4 to target specific sites in the C-terminal domain of Rb, phospho-specific antibodies to the Ser795, Ser780, and Ser807or Ser811 sites were employed (Fig. (Fig.1C).1C). Full-length untagged Rb purified from E. coli was left untreated or cleaved using caspase 3 and then phosphorylated using cyclin D1-Cdk4. Cdk4-dependent phosphorylation of full-length Rb was detected with all the antibodies; however, no measurable phosphorylation at the Ser795 or Ser807/Ser811 site was seen using C-terminally cleaved Rb and phosphorylation at the Ser780 site was severely impaired.

FIG. 1.
Caspase cleavage inhibits phosphorylation of Rb by cyclin D1-Cdk4. (A) Full-length 35S-labeled Rb was left untreated or incubated with caspase 3 (40 ng) in the presence (+) or absence (−) of the caspase inhibitor DEVD-CHO (10 μM). ...

The results presented above demonstrate that in order to get efficient phosphorylation of Rb both in general (Fig. 1A and B) and at specific phospho-acceptor sites (Fig. (Fig.1C),1C), Cdk4 has a strict requirement for the C-terminal 42 aa of the protein, suggesting that this region has Cdk4-specific determinants. To determine whether there was similar dependence on the extreme C terminus of Rb for Cdk4-dependent phosphorylation in vivo, an Rb construct [Rb(1-886)] was produced which mimicked caspase cleavage of the full-length protein. When this construct was expressed in reticulocyte lysates, as expected, it was not modified by cyclin D1-Cdk4 and showed no change in mobility, whereas it was efficiently phosphorylated by cyclin E-Cdk2 (Fig. (Fig.1D).1D). Similarly, expression of Rb(1-886) in the Rb null CDK-deficient cell line Saos2, together with cyclin D1 and Cdk4, gave no pronounced change in mobility (Fig. (Fig.1E).1E). The majority of wild-type Rb was phosphorylated by cyclin D1-Cdk4 as determined by the increase in molecular mass. Together the data lend strong support for the existence of a cyclin D1-Cdk4 recognition and docking site within the C-terminal 42 aa of Rb that is essential for the phosphorylation of full-length Rb, and a C-terminal fragment, both in vitro and in cells.

Identification of a C-terminal motif required for Cdk4-dependent phosphorylation of Rb.

A synthetic peptide library based on the C terminus of Rb (Fig. (Fig.2A)2A) was used to explore the relationship between this region and phosphorylation at distinct sites. The library was screened to determine whether any of the peptides were able to block Cdk4-dependent phosphorylation of Rb. Figure Figure2B2B shows that peptide 7 inhibited the incorporation of 32P into Rb/773-928 by cyclin D1-Cdk4 but had no effect on the ability of cyclin E-Cdk2 to phosphorylate this substrate. In contrast, peptide 6, which contains a putative cyclin binding KXL motif starting at Lys889 (1), gave only modest inhibition of cyclin D1-Cdk4 activity. Subsequent studies have shown that the effect of this peptide may be nonspecific as neither mutation of the KXL motif nor that of any other amino acid affects its activity (data not shown). Peptides 1 and 2, which contain consensus CDK phosphorylation motifs, inhibited Cdk2-dependent phosphorylation of Rb/773-928, whereas, consistent with Cdk4's having very low specific activity against peptide substrates, these peptides did not compete with Rb/773-928 for phosphorylation by Cdk4. Titration of peptide 7 over an extended range of concentrations (Fig. (Fig.2C)2C) showed that cyclin E-Cdk2 was completely insensitive to the inhibitory activity of this peptide. As peptide 7 does not contain any putative cyclin-CDK phosphorylation sites, we reasoned that this peptide was likely inhibiting Cdk4 activity by blocking substrate recognition.

FIG. 2.
Mapping the Cdk4 docking motif. (A) A series of overlapping 20-aa peptides spanning the C terminus of Rb (aa 797 to 928). Putative consensus residues for phosphorylation by cyclin-CDKs are highlighted in italics. (B) Incorporation of 32P into GST-Rb/773-928 ...

In order to identify key residues required for activity, we synthesized derivatives of peptide 7 in which each residue was sequentially mutated to alanine (Fig. (Fig.2D).2D). Based on the relative ability of the alanine-mutated peptides to inhibit Cdk4-dependent phosphorylation, we were able to define the recognition motif for cyclin D1-Cdk4 on Rb as lying between aa 893 and 910. With this information (Fig. (Fig.2D,2D, lower panel), the essential motif has been defined as GESKFQQKLAEMTSTRTR, where boldface letters represent residues that are important for binding and underlined boldface letters represent residues that are critical.

To confirm that the recognition motif is critical for Cdk4 phosphorylation of Rb, a series of mutants was generated using the Rb/773-928 construct (Fig. (Fig.3A).3A). Following expression in E. coli and purification on glutathione beads, equal amounts of wild-type and mutant Rb/773-928 protein were added to the cyclin D1-Cdk4 assay mixture. The recognition motif mutants (mutants 2 to 4) were all poor substrates for Cdk4 (Fig. (Fig.3B,3B, upper panel). In fact, mutations within the recognition motif had an effect on phosphorylation similar to that of a truncation mutation that mimicked caspase cleavage (mutant 5). Consistent with the results obtained using peptide 6 (Fig. (Fig.2B),2B), mutation of the putative cyclin binding site at Lys889 (mutant 1) had a negligible effect on Cdk4-dependent incorporation into Rb/773-928. None of the mutations had a significant effect on the ability of cyclin E-Cdk2 to phosphorylate Rb/773-928 (Fig. (Fig.3B,3B, lower panel).

FIG. 3.
Mutation of critical residues within the Cdk4 docking motif. (A) The Cdk4-docking domain is shown with critical residues underlined. Residues in italics were mutated to Ala within Rb/773-928. Mutations were introduced into the putative cyclin binding ...

When point mutations within the Lys889 putative cyclin binding site (Lue891 → Ala) and the Cdk4 recognition motif (Phe897 → Ala) were introduced into full-length Rb (Rb-LF mutant) (Fig. (Fig.3C,3C, left panel), the 35S-labeled protein, like the Rb(1-886) fragment, was refractory to phosphorylation by Cdk4. Furthermore, a single point mutation within the Cdk4 recognition motif (Phe897 → Ala; Rb-F mutant) also abolished the Cdk4-dependent shift in Rb mobility (Fig. (Fig.3C,3C, right panel). When introduced into Saos2 cells together with cyclin D1 and Cdk4, the Rb-LF and Rb-F mutants were inefficient Cdk4 substrates (Fig. (Fig.3D)3D) as judged by a lack of change in Rb mobility, whereas wild-type Rb protein was readily phosphorylated, resulting in a change in mobility.

Cdk4 docks at the C-terminal recognition site of Rb.

The motif identified within the C terminus of Rb as being important for Cdk4-dependent phosphorylation (Fig. (Fig.2D2D and and3)3) bears no similarity to the previously defined motif for cyclin binding. We therefore speculated that the recognition motif might represent a Cdk4 docking site. To verify first whether Cdk4 was able to bind Rb and whether this was independent of cyclin D1 and second whether binding required residues crucial for recognition and phosphorylation by cyclin D1-Cdk4 (Fig. (Fig.3B),3B), the following assays were carried out. Wild-type Rb/773-928, recognition site mutants (mutants 2 to 4), or the truncation mutant (mutant 5) was captured on glutathione-Sepharose beads (Fig. (Fig.4A,4A, lower panel) and incubated with Cdk4 purified from E. coli (Fig. (Fig.4A,4A, upper panel). Following extensive washing, Cdk4 was found in association with beads bearing Rb/773-928, whereas no binding was detected in a bead-alone control (Fig. (Fig.4A,4A, upper panel, compare lanes labeled wt and beads). Binding of Cdk4 to all the docking site mutants (Fig. (Fig.4A,4A, upper panel, lanes 2 to 4) and, as expected, to the truncation mutant (Fig. (Fig.4A,4A, upper panel, lane 5) was significantly reduced under conditions where equal amounts of protein were captured (Fig. (Fig.4A,4A, lower panel). The ability of Cdk4 to bind full-length Rb was then determined using immunoprecipitation assays. Wild-type Rb, the Phe897 mutant (Rb-F), and the truncation mutant [Rb(1-886)] were translated in reticulocyte lysates, captured using anti-Rb immunoglobulin G, and mixed with purified Cdk4. Rb capture and Cdk4 binding were verified by immunoblotting. Although Cdk4 protein was detected in the immunocomplex containing wild-type Rb protein, the levels detected in complex with the Phe897 mutant (Rb-F) or the truncation mutant [Rb(1-886)] were identical to that seen in the bead-alone control (Fig. (Fig.4B4B).

FIG. 4.
Cdk4 binds with lower affinity to mutant Rb. (A) One microgram of purified recombinant wild-type GST-Rb/773-928 (wt) or mutant 2, 3, 4, or 5 (Fig. (Fig.3)3) adsorbed to glutathione-Sepharose beads was incubated with His-Cdk4 (2 μg) for ...

The data show that Cdk4 can bind, in the absence of cyclin D1, to both full-length Rb and the isolated Rb C terminus in a manner that depends on the integrity of the recognition motif. This demonstrates that the C-terminal cyclin D1-Cdk4 recognition site constitutes a Cdk4 docking site. The fact that the docking site is essential for correct cyclin D1-Cdk4-dependent phosphorylation of Rb suggests that Cdk4 docking plays a key role in both substrate specificity and maximal catalytic activity.

The Cdk4 docking site of Rb is required for growth regulation.

The data presented above show that the Cdk4 docking site is essential for Cdk4-dependent phosphorylation of Rb both in vitro and in cells; however, they do not address whether the docking site is essential for Cdk4 to exert its control over cell growth. In order to address these issues, we analyzed Rb-dependent cell cycle progression and colony growth in Saos2 cells. Previous studies have suggested that the expression of cyclin D-Cdk4 in Saos2 cells can overcome G1 arrest imposed by Rb expression (16, 24). We therefore hypothesized that the ability of cyclin D1-Cdk4 to overcome Rb-imposed G1 arrest should be reduced in Rb docking site mutants if this site is critical for Cdk4-dependent regulation. We first verified that C-terminally truncated Rb [Rb(1-886)] retained the ability to inhibit cell growth. When Saos2 cells were transfected with Rb(1-886) and cell cycle distribution was determined using flow cytometry, we found an increase of between 14 and 23% in the population of G1-phase cells in three independent experiments (Fig. 5A and B). This increase is comparable to the 15 to 20% increase seen in the presence of wild-type Rb. In addition, colony formation assays demonstrated that Rb(1-886) reduced the number of Saos2 cell colonies by an extent similar to that found with the wild-type protein (Fig. (Fig.5C).5C). Together these data show that the C-terminal truncation mutant of Rb retains growth-inhibitory activity similar to that of the wild-type protein. When cyclin D1-Cdk4 was titrated in the absence of Rb, the enzyme itself increased the number of cells with a >2N DNA content while it decreased the number of cells in G1 phase compared to the number of such cells in the vector control (Fig. 5A and B). This is consistent with recent observations that Cdk4 can promote cell cycle progression in the absence of Rb (34). Whereas Rb expression in the absence of cyclin D1-Cdk4 led to G1-phase arrest, producing an increase in the number of cells with a 2N DNA content (Fig. 5A and B), coexpression of cyclin D1-Cdk4 prevented accumulation of G1-phase cells, suggestive of the fact that Rb was inactivated in the presence of the enzyme. On the other hand, the increased population of G1-phase cells seen in the presence of Rb(1-886) (Fig. 5A and B) remained unaltered in cells coexpressing cyclin D1-Cdk4, suggesting that C-terminally truncated Rb is refractory to Cdk4-dependent inactivation. In contrast, neither Rb nor Rb(1-886) was able to generate G1 arrest when expressed in the presence of cyclin E-Cdk2, suggesting that the ability of Cdk2 to inactivate Rb occurs independently of the Cdk4 docking site.

FIG. 5.FIG. 5.
Growth-suppressing ability of Rb mutants. (A and B) Cell cycle distribution of Saos2 cells following transfection with full-length wild-type Rb (Rb) or truncated Rb [Rb (1-886)]. Saos2 cells were transfected with 10 μg of Rb or Rb(1-886) and 2 ...

Assays were employed to confirm that the C-terminal docking site on Rb is critical for Cdk4-dependent regulation of cell growth and colony formation (Fig. (Fig.5C).5C). Colony growth was suppressed in Saos2 cells expressing wild-type Rb and Rb(1-886) compared to that in vector-alone controls. In addition, cells expressing an Rb construct with a single F → A mutation within the Cdk4 docking site (Rb-F) also demonstrated suppressed colony growth, showing that this mutant retains its growth suppressor function. When cyclin D1-Cdk4 was coexpressed with wild-type Rb, the number of colonies increased from 60 to 145. However, there was no increase in the number of colonies detected when cyclin D1-Cdk4 was expressed with either Rb(1-886) or Rb-F. Thus, colony growth suppression imposed by Rb mutants in which the Cdk4 docking site is absent or mutated cannot be overcome by cyclin D1-Cdk4.

The data presented above demonstrate a strong relationship between the ability of Cdk4 to bind to a docking site within the extreme C terminus of the Rb tumor suppressor protein and the ability of the enzyme to overcome Rb-imposed suppression of cell cycle progression and cell growth.

Cyclin D1-Cdk4 can protect Rb from caspase 3-dependent cleavage in cells.

As the Cdk4 docking motif is adjacent to the caspase cleavage site, we sought to discover whether Cdk4 docking could play a role in regulating caspase-dependent cleavage of Rb. We first determined whether Cdk4 up-regulation could modulate the susceptibility of Rb to cleavage in cells induced to undergo apoptosis. Saos2 cells were transfected with Rb alone or in combination with cyclin D1 and Cdk4. Whereas Rb expressed by itself was found almost exclusively in a hypophosphorylated form (Fig. (Fig.6A),6A), the coexpression of cyclin D1-Cdk4 led to the generation of a substantial proportion of hyperphosphorylated protein as detected by changes in electrophoretic mobility (Fig. (Fig.6A).6A). Caspase-dependent Rb cleavage is induced by tumor necrosis factor (TNF) treatment (26). In the present study, TNF induced cleavage of hypophosphorylated Rb to give ΔRb and further processing of Rb into lower-molecular-weight fragments (Fig. (Fig.6A).6A). However, in the presence of cyclin D1-Cdk4, background levels of both ΔRb and TNF-induced ΔRb, as well as lower-molecular-weight Rb fragments, were absent. Studies with MDA MB231 cells have shown that tamoxifen-induced apoptosis is accompanied by Rb cleavage (17). To determine the effect of cyclin D1-Cdk4 overexpression on the cleavage of endogenous Rb, we therefore monitored the appearance of the 5-kDa cleavage product in tamoxifen-treated cells. Figure Figure6B6B shows the appearance of the 5-kDa Rb cleavage product in tamoxifen-treated cells that had been previously transfected with either vector alone (left panel) or cyclin D1 and Cdk4 (right panel). Expression of cyclin D1-Cdk4 prior to tamoxifen treatment suppressed the cleavage of Rb (Fig. (Fig.6B,6B, right panel). Hence, up-regulation of cyclin D1-Cdk4 levels in cells is sufficient to inhibit initial cleavage and subsequent degradation of Rb.

FIG. 6.
Overexpression of cyclin D1-Cdk4 prevents cleavage of Rb. (A) Saos2 cells were transfected with 1 μg of Rb alone or in combination with 1 μg of cyclin D1 and 1 μg of Cdk4. The transfections were normalized for DNA by using empty ...

Uncoupling Cdk4 docking from its catalytic activity.

To determine whether the ability of cyclin D1-Cdk4 to overcome Rb cleavage in cells (Fig. (Fig.6)6) is direct and specific, we compared the effects of Cdk4 and Cdk2 on caspase 3-dependent cleavage in vitro. When 35S-labeled full-length Rb was phosphorylated with cyclin D1-Cdk4, the characteristic cleavage product (ΔRb) seen in the presence of caspase 3 was no longer observed (Fig. (Fig.7A,7A, compare lanes labeled −Enzyme and Cyclin D1-Cdk4) whereas Rb phosphorylated by cyclin E-Cdk2 was still a good substrate for caspase 3 as demonstrated by the appearance of a faster-migrating ΔRb form (Fig. (Fig.7A).7A). Similarly, with Rb/773-928, cyclin D1-Cdk4-phosphorylated protein was resistant to cleavage compared to Rb that had been phosphorylated by cyclin E-Cdk2 (Fig. (Fig.7B).7B). These experiments were both carried out under conditions in which caspase 3 was rate limiting. One possible explanation for the above-described results is that cyclin D1-Cdk4 can act as a competitive substrate for caspase 3. Cdk4 has not been described as a caspase 3 substrate, and we were unable to identify any potential cleavage sites. On the other hand, cyclin D1 from Xenopus laevis is cleaved by caspase 3 at a DEVD278 site (18). However, we found that human cyclin D1 (which has the sequence EEVD) was not a substrate for caspase 3 under conditions in which known substrates were efficiently cleaved (data not shown). The data presented in this section suggest that cyclin D1-Cdk4 protects Rb from cleavage (Fig. (Fig.7)7) through a direct mechanism.

FIG. 7.
Cyclin D1-Cdk4 inhibits caspase cleavage of Rb. (A) 35S-labeled full-length Rb was left unphosphorylated (−Enzyme) or phosphorylated using Sf9 lysate containing cyclin D1-Cdk4 or cyclin E-Cdk2 and subsequently incubated in the presence (+) ...

As cyclin D1-Cdk4 can both bind to and phosphorylate the C-terminal domain of Rb, we investigated which of these activities was primarily responsible for the protection from cleavage offered to Rb by Cdk4. Rb/773-928 was phosphorylated in the presence of [32P]ATP by either cyclin D1-Cdk4 or cyclin E-Cdk2, and the substrate was captured on glutathione beads, washed extensively to remove associated CDK complexes, and added to a caspase cleavage assay mixture (Fig. (Fig.7C,7C, left panels). Although we were unable to completely deplete cyclin D1-Cdk4 from the GST-Rb pull-down mixture (Fig. (Fig.7C,7C, right panel), the ability of caspase 3 to cleave Rb was partially restored by removal of ~70% of the added Cdk4 protein. These data suggest that Cdk4-dependent phosphorylation, by itself, may not be sufficient to prevent caspase-dependent cleavage of Rb; however, we cannot rule out the possibility that phosphorylation contributes to this protective mechanism in a complex cellular environment. Two approaches were taken to determine the effects of cyclin D1-Cdk4 binding to Rb in the absence of phosphorylation. First, cyclin D1-Cdk4 was incubated with Rb in the absence of ATP, and second, a kinase-dead Cdk4 construct (Cdk4dn) was employed. Figure Figure7D7D shows that while 35S-labeled Rb incubated with cyclin E-Cdk2 in the absence of ATP (right panel) was cleaved by caspase 3, under the same conditions cyclin D1-Cdk4 prevented cleavage (left panel). When wild-type Cdk4 was replaced with Cdk4dn, preincubation of 35S-labeled Rb with Sf9 cell lysate expressing cyclin D1-Cdk4dn or Cdk4dn alone was sufficient to prevent caspase cleavage whereas preincubation with uninfected Sf9 cell lysate had no protective effect (Fig. (Fig.7E).7E). Finally, while addition of cyclin D1-Cdk4 to wild-type Rb, in the absence of phosphorylation, protected the protein from caspase 3 cleavage, the enzyme was not able to prevent the cleavage of Rb-F, which was cleaved efficiently in both the absence and presence of cyclin D1-Cdk4 (Fig. (Fig.7F).7F). The above data demonstrate that docking of Cdk4 to, rather than phosphorylation of, Rb is critical for inhibition of caspase 3-dependent cleavage.


One mechanism by which protein kinases can specifically phosphorylate a given phospho-acceptor site within a target substrate or discriminate between potential phosphorylation motifs in distinct substrates is through a direct docking interaction at a site distinct from the phospho-acceptor sequence (5). Efficient phosphorylation of target proteins by the cyclin-CDKs requires both substrate binding and kinase activity, putting them into the above-mentioned category. Interestingly, both the cyclin regulatory subunit (2, 3, 35, 47, 63) and the CDK catalytic subunit (47, 54) have been reported to participate in specific docking interactions. The cyclin binding sites from several proteins have been identified, and these contain a core R/KXL motif (2, 3). Although the CDK subunits have been shown to bind substrates, as well as regulatory factors, the requirement for CDK substrate docking is not well defined. We have mapped and characterized a Cdk4 docking motif within the C terminus of Rb that is a major determinant of cyclin D1-Cdk4-dependent phosphorylation of full-length Rb in vitro and in cell systems. In addition, the interaction of Cdk4 with its docking site is required for the enzyme to overcome growth inhibition imposed by Rb and to modulate caspase-dependent cleavage of the Rb protein.

The Cdk4 docking site is critical for Rb phosphorylation.

Previous studies have suggested that both the Rb protein B domain, which can interact with the LXCXE motif present in some cyclins, and certain R/KXL motifs within the C terminus of Rb can independently interact with cyclin D to direct Cdk4-dependent phosphorylation (1, 13). In the present study, we have defined a site within the C-terminal domain of Rb (aa 893 to 910) that is required for cyclin D1-Cdk4-dependent phosphorylation of Rb both in vitro and in cells (Fig. (Fig.11 to to3).3). Furthermore, peptides based on this domain specifically inhibit the activity of Cdk4 presumably by competing for substrate binding (Fig. (Fig.2C).2C). The amino acids from positions 893 to 910 of Rb display no characteristics of a cyclin binding site but are required for cyclin D1-independent binding of Cdk4 to both full-length and C-terminal constructs of Rb (Fig. (Fig.4).4). Our data therefore demonstrate that there is a docking site for Cdk4 in the C terminus of Rb and that binding to this site is required for the correct phosphorylation of Rb by this kinase. During the course of our study, Pan et al. (40) confirmed the importance of the C-terminal region of Rb for Cdk4-dependent regulation by reporting on the isolation of a spontaneous Rb mutant (Leu901 → Gln) that suppresses cyclin D-Cdk4-dependent phosphorylation. The present study explains this observation, as Leu901 lies within the Cdk4 docking motif and mutation of the equivalent residue within our docking site peptide (Fig. (Fig.2D)2D) significantly reduced its inhibitory activity, suggesting that this is an important contact residue for the interaction between Rb and Cdk4.

Using full-length Rb, we found that loss of the Cdk4 docking site, or point mutations that inhibited enzyme binding, essentially prevented Cdk4-dependent modification in terms of both total phosphorylation and modification at specific phospho-acceptor sites. This finding is in contrast to results for the LXCXE binding domain and the R/KXL motif both of which are dispensable for Cdk4-dependent phosphorylation (1, 11, 41). We would therefore propose a model in which the critical step required for Rb phosphorylation is the interaction of Cdk4 with its docking site. This interaction would then facilitate subsequent binding of cyclin D1 to R/KXL motifs located toward the N terminus of the caspase cleavage site (1) or, perhaps under some conditions with the B domain, further stimulating of site-directed phosphorylation of the Rb protein.

Relationship between Cdk4, Rb, and growth control.

Early studies on the relationship between Cdk4 activity and the Rb protein suggested that the ability of Cdk4 to regulate cell growth was dependent on the presence of Rb (16, 24). Thus, cyclin D1 or Cdk4 up-regulation could promote oncogenic transformation by inactivating the tumor suppressor function of Rb. More recent reports have suggested that cyclin D1-Cdk4 can promote cell growth independently of the Rb protein and that this may contribute to the enzyme's oncogenic activity (34). In their study, Leng et al. (34) found that cyclin D1-Cdk4 promoted entry into S phase regardless of Rb status. In the present study, we also found that expression of cyclin D1-Cdk4 in the absence of Rb was sufficient to promote S-phase entry (Fig. 5A and B). However, cyclin D1-Cdk4 did not promote significant S-phase entry in the presence of Rb as the cell cycle profile was similar to that seen in the vector-alone control. Cyclin D1-Cdk4 did, however, overcome Rb-imposed G1 arrest (Fig. 5A and B). Interestingly, when cyclin D1-Cdk4 levels were titrated (data not shown), we found that higher amounts of enzyme (≥2 μg of transfected DNA) were able to promote S-phase entry irrespective of Rb status. This suggests that in the presence of functional Rb, cyclin D1-Cdk4 levels are rate limiting for cell cycle progression. However, when the ratio of cyclin D1-Cdk4 to Rb is increased or when Rb function is absent, the kinase can promote S-phase entry through an alternative strategy. Thus, in nontransformed cells we would expect cyclin D1-Cdk4 to be dependent on Rb for its activity whereas in certain transformed cells the enzyme is likely to promote growth through both Rb-dependent and -independent mechanisms.

Although cyclin D1-Cdk4 can overcome the growth-suppressive activity of Rb as determined by both cell cycle progression and colony formation assays (Fig. (Fig.5),5), the enzyme is unable to rescue cells from the growth-suppressive effects of Rb in which the Cdk4 docking site has been either deleted or mutated. The introduction of a single point mutation at a residue critical for Cdk4 binding to Rb (Fig. (Fig.4)4) renders the protein completely resistant to Cdk4-dependent inactivation (Fig. (Fig.5C).5C). This result is all the more striking when compared to those of studies showing that cyclin D1-Cdk4 can still promote S-phase entry in the presence of an Rb phosphorylation site mutant in which all the consensus CDK sites have been mutated to alanine (34). Thus, Cdk4 docking may play a role in regulating Rb activity that goes beyond substrate recognition and phospho-site specificity. Indeed, we have found that binding of Cdk4 to Rb can initiate conformational changes in the Rb protein in the absence of phosphorylation (data not shown), suggesting that docking may contribute to the conformational changes required for Cdk4 to prime Rb for subsequent Cdk2-dependent phosphorylation (22).

A number of studies have suggested that the binding activity of Cdk4, rather than its catalytic activity, is sufficient to explain some of the biological functions of the kinase. Cdk4-dependent inhibition of MyoD-regulated transcription and muscle cell differentiation can occur in the absence of Cdk4 catalytic activity and requires the enzyme to bind a 15-aa sequence in the C-terminal domain of MyoD (61, 62). More recently, Cdk4 has been linked, both genetically and biochemically, to the regulation of STAT signaling pathways in Drosophila (8, 51). Cdk4 binds to STAT92E and regulates its steady-state levels independent of its catalytic activity. This demonstrates a role for Cdk4 docking activity in cell fate determination (8). In the present study, we found evidence to support additional functions for Cdk4 binding as docking of the enzyme to Rb was sufficient to inhibit caspase cleavage (Fig. (Fig.66 and and77).

The role of cyclin D1-Cdk4 in apoptosis has been controversial. Several studies have linked induction of cyclin D1-Cdk4 to cell death in postmitotic neurons and senescent fibroblasts (19, 29). In addition, cyclin D1 and/or Cdk4 overexpression in specific tumor cell lines leads to apoptosis in response to selected stimuli (21). However, these studies are at odds with the fact that both cyclin D1 and Cdk4 are frequently up-regulated in human cancers (46). Cyclin D-Cdk4 activity can also be up-regulated by loss of the Cdk4- and Cdk6-specific inhibitor and tumor suppressor p16INK4a (4, 46) or by the generation of a Cdk4 mutant enzyme that is refractory to p16-mediated inhibition (56, 64). Thus, up-regulation of cyclin D1-Cdk4 activity is well tolerated by many tumor cells. Studies with transgenic cells lend strong support for the role of cyclin D1-Cdk4 as a positive growth regulator that promotes proliferation and suppresses apoptosis. For example, mice expressing Cdk4 that is refractory to p16 develop tumors in various tissue types and are highly susceptible to carcinogen-induced papillomas (43, 44). The photoreceptor cells of cyclin D1 knockout mice exhibit increased rates of cell death, leading to retinal degeneration (37), and the loss of cyclin D1 in DT40 cells increases their sensitivity to radiation-induced cell death (31). Consistent with the results of the latter study, stable expression of cyclin D1 increases the survival of some cell types when the cells are exposed to ionizing radiation (15, 57). The above-cited studies suggest that inappropriate overexpression or loss of cyclin D1 or Cdk4 activity can lead to cell death dependent on cell type.

Specific cleavage and degradation of the Rb protein in some cell types undergoing apoptosis suggests that the loss of the Rb protein may favor cell death. This is supported by the fact that the cells of several developing tissues of Rb−/− mice have increased rates of ectopic apoptosis (9, 25, 32). In addition, there is evidence that the cleaved form of Rb may actually promote cell death by retaining some activities while losing others (7, 26). Perhaps the most convincing study linking caspase-dependent cleavage of Rb to apoptosis found that transgenic mice carrying a caspase-resistant Rb construct showed tissue- and signal-specific roles for Rb in the regulation of apoptosis (6). We and others have proposed that the phosphorylation status of the Rb protein may be an important factor in determining its apoptotic function (12, 55). Most recently it has been proposed that phosphorylation of Rb within the C terminus by Cdk4 favors cell growth over apoptosis by relieving Rb-mediated inhibition of cell cycle progression while preserving Rb antiapoptotic activity (38). On the other hand, phosphorylation of Rb at a single site (Ser567) within domain A, which occurs inefficiently and only in the presence of high levels of Cdk2 activity, favors cell death (38). Here we have demonstrated that binding of Cdk4 to its docking site within the C terminus may contribute to the ability of the enzyme to inhibit caspase-dependent cleavage of Rb. This raises the possibility that Cdk4 may act in a docking site-dependent manner to preserve Rb in an antiapoptotic form by promoting its phosphorylation and inactivation while preventing its C-terminal cleavage.


K.L.B. has a Programme Grant from CRUK that also supports M.W.

We thank Alison Sparks for help with cell cycle analysis.


1. Adams, P. D., X. Li, W. R. Sellers, K. B. Baker, X. Leng, J. W. Harper, Y. Taya, and W. G. Kaelin, Jr. 1999. Retinoblastoma protein contains a C-terminal motif that targets it for phosphorylation by cyclin-cdk complexes. Mol. Cell. Biol. 19:1068-1080. [PMC free article] [PubMed]
2. Adams, P. D., W. R. Sellers, S. K. Sharma, A. D. Wu, C. M. Nalin, and W. G. Kaelin, Jr. 1996. Identification of a cyclin-cdk2 recognition motif present in substrates and p21-like cyclin-dependent kinase inhibitors. Mol. Cell. Biol. 16:6623-6633. [PMC free article] [PubMed]
3. Ball, K. L. 1997. p21: structure and functions associated with cyclin-CDK binding. Prog. Cell Cycle Res. 3:125-134. [PubMed]
4. Bartkova, J., J. Lukas, P. Guldberg, J. Alsner, A. F. Kirkin, J. Zeuthen, and J. Bartek. 1996. The p16-cyclin D/Cdk4-pRb pathway as a functional unit frequently altered in melanoma pathogenesis. Cancer Res. 56:5475-5483. [PubMed]
5. Biondi, R. M., and A. R. Nebreda. 2003. Signalling specificity of Ser/Thr protein kinases through docking-site-mediated interactions. Biochem. J. 372:1-13. [PMC free article] [PubMed]
6. Chau, B. N., H. L. Borges, T. T. Chen, A. Masselli, I. C. Hunton, and J. Y. Wang. 2002. Signal-dependent protection from apoptosis in mice expressing caspase-resistant Rb. Nat. Cell Biol. 4:757-765. [PubMed]
7. Chen, W. D., J. Geradts, M. M. Keane, S. Lipkowitz, M. Zajac-Kaye, and F. J. Kaye. 1999. The 100-kDa proteolytic fragment of RB is retained predominantly within the nuclear compartment of apoptotic cells. Mol. Cell Biol. Res. Commun. 1:216-220. [PubMed]
8. Chen, X., S. W. Oh, Z. Zheng, H. W. Chen, H. H. Shin, and S. X. Hou. 2003. Cyclin D-Cdk4 and cyclin E-Cdk2 regulate the Jak/STAT signal transduction pathway in Drosophila. Dev. Cell 4:179-190. [PubMed]
9. Clarke, A. R., E. R. Maandag, M. van Roon, N. M. van der Lugt, M. van der Valk, M. L. Hooper, A. Berns, and H. te Riele. 1992. Requirement for a functional Rb-1 gene in murine development. Nature 359:328-330. [PubMed]
10. Connell-Crowley, L., J. W. Harper, and D. W. Goodrich. 1997. Cyclin D1/Cdk4 regulates retinoblastoma protein-mediated cell cycle arrest by site-specific phosphorylation. Mol. Biol. Cell 8:287-301. [PMC free article] [PubMed]
11. Dick, F. A., E. Sailhamer, and N. J. Dyson. 2000. Mutagenesis of the pRB pocket reveals that cell cycle arrest functions are separable from binding to viral oncoproteins. Mol. Cell. Biol. 20:3715-3727. [PMC free article] [PubMed]
12. Dou, Q. P., and V. W. Lui. 1995. Failure to dephosphorylate retinoblastoma protein in drug-resistant cells. Cancer Res. 55:5222-5225. [PubMed]
13. Dowdy, S. F., P. W. Hinds, K. Louie, S. I. Reed, A. Arnold, and R. A. Weinberg. 1993. Physical interaction of the retinoblastoma protein with human D cyclins. Cell 73:499-511. [PubMed]
14. Dynlacht, B. D., O. Flores, J. A. Lees, and E. Harlow. 1994. Differential regulation of E2F transactivation by cyclin/cdk2 complexes. Genes Dev. 8:1772-1786. [PubMed]
15. Epperly, M., L. Berry, A. Halloran, and J. S. Greenberger. 1995. Inhibition of G1-phase arrest induced by ionizing radiation in hematopoietic cells by overexpression of genes involved in the G1/S-phase transition. Radiat. Res. 143:245-254. [PubMed]
16. Ewen, M. E., H. K. Sluss, C. J. Sherr, H. Matsushime, J. Kato, and D. M. Livingston. 1993. Functional interactions of the retinoblastoma protein with mammalian D-type cyclins. Cell 73:487-497. [PubMed]
17. Fattman, C. L., B. An, L. Sussman, and Q. P. Dou. 1998. p53-independent dephosphorylation and cleavage of retinoblastoma protein during tamoxifen-induced apoptosis in human breast carcinoma cells. Cancer Lett. 130:103-113. [PubMed]
18. Finkielstein, C. V., L. G. Chen, and J. L. Maller. 2002. A role for G1/S cyclin-dependent protein kinases in the apoptotic response to ionizing radiation. J. Biol. Chem. 277:38476-38485. [PubMed]
19. Freeman, R. S., S. Estus, and E. M. Johnson, Jr. 1994. Analysis of cell cycle-related gene expression in postmitotic neurons: selective induction of Cyclin D1 during programmed cell death. Neuron 12:343-355. [PubMed]
20. Grafstrom, R. H., W. Pan, and R. H. Hoess. 1999. Defining the substrate specificity of cdk4 kinase-cyclin D1 complex. Carcinogenesis 20:193-198. [PubMed]
21. Han, E. K., M. Begemann, A. Sgambato, J. W. Soh, Y. Doki, W. Q. Xing, W. Liu, and I. B. Weinstein. 1996. Increased expression of cyclin D1 in a murine mammary epithelial cell line induces p27kip1, inhibits growth, and enhances apoptosis. Cell Growth Differ. 7:699-710. [PubMed]
22. Harbour, J. W., R. X. Luo, A. Dei Santi, A. A. Postigo, and D. C. Dean. 1999. Cdk phosphorylation triggers sequential intramolecular interactions that progressively block Rb functions as cells move through G1. Cell 98:859-869. [PubMed]
23. Holmes, J. K., and M. J. Solomon. 1996. A predictive scale for evaluating cyclin-dependent kinase substrates. A comparison of p34cdc2 and p33cdk2. J. Biol. Chem. 271:25240-25246. [PubMed]
24. Horton, L. E., Y. Qian, and D. J. Templeton. 1995. G1 cyclins control the retinoblastoma gene product growth regulation activity via upstream mechanisms. Cell Growth Differ. 6:395-407. [PubMed]
25. Jacks, T., A. Fazeli, E. M. Schmitt, R. T. Bronson, M. A. Goodell, and R. A. Weinberg. 1992. Effects of an Rb mutation in the mouse. Nature 359:295-300. [PubMed]
26. Janicke, R. U., P. A. Walker, X. Y. Lin, and A. G. Porter. 1996. Specific cleavage of the retinoblastoma protein by an ICE-like protease in apoptosis. EMBO J. 15:6969-6978. [PMC free article] [PubMed]
27. Kelly, B. L., K. G. Wolfe, and J. M. Roberts. 1998. Identification of a substrate-targeting domain in cyclin E necessary for phosphorylation of the retinoblastoma protein. Proc. Natl. Acad. Sci. USA 95:2535-2540. [PMC free article] [PubMed]
28. Kitagawa, M., H. Higashi, H. K. Jung, I. Suzuki-Takahashi, M. Ikeda, K. Tamai, J. Kato, K. Segawa, E. Yoshida, S. Nishimura, and Y. Taya. 1996. The consensus motif for phosphorylation by cyclin D1-Cdk4 is different from that for phosphorylation by cyclin A/E-Cdk2. EMBO J. 15:7060-7069. [PMC free article] [PubMed]
29. Kranenburg, O., A. J. van der Eb, and A. Zantema. 1996. Cyclin D1 is an essential mediator of apoptotic neuronal cell death. EMBO J. 15:46-54. [PMC free article] [PubMed]
30. Krek, W., M. E. Ewen, S. Shirodkar, Z. Arany, W. G. Kaelin, Jr., and D. M. Livingston. 1994. Negative regulation of the growth-promoting transcription factor E2F-1 by a stably bound cyclin A-dependent protein kinase. Cell 78:161-172. [PubMed]
31. Lahti, J. M., H. Li, and V. J. Kidd. 1997. Elimination of cyclin D1 in vertebrate cells leads to an altered cell cycle phenotype, which is rescued by overexpression of murine cyclins D1, D2, or D3 but not by a mutant cyclin D1. J. Biol. Chem. 272:10859-10869. [PubMed]
32. Lee, E. Y., C. Y. Chang, N. Hu, Y. C. Wang, C. C. Lai, K. Herrup, W. H. Lee, and A. Bradley. 1992. Mice deficient for Rb are nonviable and show defects in neurogenesis and haematopoiesis. Nature 359:288-294. [PubMed]
33. Lee, J. O., A. A. Russo, and N. P. Pavletich. 1998. Structure of the retinoblastoma tumour-suppressor pocket domain bound to a peptide from HPV E7. Nature 391:859-865. [PubMed]
34. Leng, X., L. Connell-Crowley, D. Goodrich, and J. W. Harper. 1997. S-phase entry upon ectopic expression of G1 cyclin-dependent kinases in the absence of retinoblastoma protein phosphorylation. Curr. Biol. 7:709-712. [PubMed]
35. Luciani, M. G., J. R. Hutchins, D. Zheleva, and T. R. Hupp. 2000. The C-terminal regulatory domain of p53 contains a functional docking site for cyclin A. J. Mol. Biol. 300:503-518. [PubMed]
36. Lundberg, A. S., and R. A. Weinberg. 1998. Functional inactivation of the retinoblastoma protein requires sequential modification by at least two distinct cyclin-cdk complexes. Mol. Cell. Biol. 18:753-761. [PMC free article] [PubMed]
37. Ma, C., D. Papermaster, and C. L. Cepko. 1998. A unique pattern of photoreceptor degeneration in cyclin D1 mutant mice. Proc. Natl. Acad. Sci. USA 95:9938-9943. [PMC free article] [PubMed]
38. Ma, D., P. Zhou, and J. W. Harbour. 2003. Distinct mechanisms for regulating the tumor suppressor and antiapoptotic functions of rb. J. Biol. Chem. 278:19358-19366. [PubMed]
39. Mittnacht, S. 1998. Control of pRB phosphorylation. Curr. Opin. Genet. Dev. 8:21-27. [PubMed]
40. Pan, W., S. Cox, R. H. Hoess, and R. H. Grafstrom. 2001. A cyclin D1/cyclin-dependent kinase 4 binding site within the C domain of the retinoblastoma protein. Cancer Res. 61:2885-2891. [PubMed]
41. Pan, W., T. Sun, R. Hoess, and R. Grafstrom. 1998. Defining the minimal portion of the retinoblastoma protein that serves as an efficient substrate for cdk4 kinase/cyclin D1 complex. Carcinogenesis 19:765-769. [PubMed]
42. Qian, Y., C. Luckey, L. Horton, M. Esser, and D. J. Templeton. 1992. Biological function of the retinoblastoma protein requires distinct domains for hyperphosphorylation and transcription factor binding. Mol. Cell. Biol. 12:5363-5372. [PMC free article] [PubMed]
43. Rane, S. G., S. C. Cosenza, R. V. Mettus, and E. P. Reddy. 2002. Germ line transmission of the Cdk4(R24C) mutation facilitates tumorigenesis and escape from cellular senescence. Mol. Cell. Biol. 22:644-656. [PMC free article] [PubMed]
44. Rane, S. G., P. Dubus, R. V. Mettus, E. J. Galbreath, G. Boden, E. P. Reddy, and M. Barbacid. 1999. Loss of Cdk4 expression causes insulin-deficient diabetes and Cdk4 activation results in beta-islet cell hyperplasia. Nat. Genet. 22:44-52. [PubMed]
45. Reed, S. I. 1997. Control of the G1/S transition. Cancer Surv. 29:7-23. [PubMed]
46. Ruas, M., and G. Peters. 1998. The p16INK4a/CDKN2A tumor suppressor and its relatives. Biochim. Biophys. Acta 1378:F115-F177. [PubMed]
47. Russo, A. A., P. D. Jeffrey, A. K. Patten, J. Massague, and N. P. Pavletich. 1996. Crystal structure of the p27Kip1 cyclin-dependent-kinase inhibitor bound to the cyclin A-Cdk2 complex. Nature 382:325-331. [PubMed]
48. Schulman, B. A., D. L. Lindstrom, and E. Harlow. 1998. Substrate recruitment to cyclin-dependent kinase 2 by a multipurpose docking site on cyclin A. Proc. Natl. Acad. Sci. USA 95:10453-10458. [PMC free article] [PubMed]
49. Scott, M. T., N. Morrice, and K. L. Ball. 2000. Reversible phosphorylation at the C-terminal regulatory domain of p21(Waf1/Cip1) modulates proliferating cell nuclear antigen binding. J. Biol. Chem. 275:11529-11537. [PubMed]
50. Sherr, C. J. 2000. The Pezcoller lecture: cancer cell cycles revisited. Cancer Res. 60:3689-3695. [PubMed]
51. Silver, D. L., and D. J. Montell. 2003. A new trick for Cyclin-Cdk: activation of STAT. Dev. Cell 4:148-149. [PubMed]
52. Songyang, Z., S. Blechner, N. Hoagland, M. F. Hoekstra, H. Piwnica-Worms, and L. C. Cantley. 1994. Use of an oriented peptide library to determine the optimal substrates of protein kinases. Curr. Biol. 4:973-982. [PubMed]
53. Tan, X., and J. Y. Wang. 1998. The caspase-RB connection in cell death. Trends Cell Biol. 8:116-120. [PubMed]
54. Wagner, P., A. Fuchs, C. Gotz, W. Nastainczyk, and M. Montenarh. 1998. Fine mapping and regulation of the association of p53 with p34cdc2. Oncogene 16:105-111. [PubMed]
55. Wallace, M., P. J. Coates, E. G. Wright, and K. L. Ball. 2001. Differential post-translational modification of the tumour suppressor proteins Rb and p53 modulate the rates of radiation-induced apoptosis in vivo. Oncogene 20:3597-3608. [PubMed]
56. Wolfel, T., M. Hauer, J. Schneider, M. Serrano, C. Wolfel, E. Klehmann-Hieb, E. De Plaen, T. Hankeln, K. H. Meyer zum Buschenfelde, and D. Beach. 1995. A p16INK4a-insensitive CDK4 mutant targeted by cytolytic T lymphocytes in a human melanoma. Science 269:1281-1284. [PubMed]
57. Xia, F., and S. N. Powell. 2002. The molecular basis of radiosensitivity and chemosensitivity in the treatment of breast cancer. Semin. Radiat. Oncol. 12:296-304. [PubMed]
58. Xu, M., K. A. Sheppard, C. Y. Peng, A. S. Yee, and H. Piwnica-Worms. 1994. Cyclin A/CDK2 binds directly to E2F-1 and inhibits the DNA-binding activity of E2F-1/DP-1 by phosphorylation. Mol. Cell. Biol. 14:8420-8431. [PMC free article] [PubMed]
59. Zalvide, J., and J. A. DeCaprio. 1995. Role of pRb-related proteins in simian virus 40 large-T-antigen-mediated transformation. Mol. Cell. Biol. 15:5800-5810. [PMC free article] [PubMed]
60. Zarkowska, T., and S. Mittnacht. 1997. Differential phosphorylation of the retinoblastoma protein by G1/S cyclin-dependent kinases. J. Biol. Chem. 272:12738-12746. [PubMed]
61. Zhang, J. M., Q. Wei, X. Zhao, and B. M. Paterson. 1999. Coupling of the cell cycle and myogenesis through the cyclin D1-dependent interaction of MyoD with cdk4. EMBO J. 18:926-933. [PMC free article] [PubMed]
62. Zhang, J. M., X. Zhao, Q. Wei, and B. M. Paterson. 1999. Direct inhibition of G(1) cdk kinase activity by MyoD promotes myoblast cell cycle withdrawal and terminal differentiation. EMBO J. 18:6983-6993. [PMC free article] [PubMed]
63. Zhu, L., E. Harlow, and B. D. Dynlacht. 1995. p107 uses a p21CIP1-related domain to bind cyclin/cdk2 and regulate interactions with E2F. Genes Dev. 9:1740-1752. [PubMed]
64. Zuo, L., J. Weger, Q. Yang, A. M. Goldstein, M. A. Tucker, G. J. Walker, N. Hayward, and N. C. Dracopoli. 1996. Germline mutations in the p16INK4a binding domain of CDK4 in familial melanoma. Nat. Genet. 12:97-99. [PubMed]

Articles from Molecular and Cellular Biology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...