CK1alpha plays a central role in mediating MDM2 control of p53 and E2F-1 protein stability

Abstract

The ubiquitin ligase murine double minute clone 2 (MDM2) mediates ubiquitination and degradation of the tumor suppressor p53. The activation and stabilization of p53 by contrast is maintained by enzymes catalyzing p53 phosphorylation and acetylation. Casein kinase 1 (CK1) is one such enzyme; it stimulates p53 after transforming growth factor-beta treatment, irradiation, or DNA virus infection. We analyzed whether CK1 regulates p53 protein stability in unstressed conditions. Depletion of CK1 using small interfering RNA or inhibition of CK1 using the kinase inhibitor (D4476) activated p53 and destabilized E2F-1, indicating that steady-state levels of these proteins are controlled by CK1. Co-immunoprecipitation of endogenous CK1 with MDM2 occurred in undamaged cells, indicating the existence of a stable multiprotein complex, and as such, we evaluated whether the MDM2 Nutlin had similar pharmacological properties to the CK1 inhibitor D4476. Indeed, D4476 or Nutlin treatments resulted in the same p53 and E2F-1 steady-state protein level changes, indicating that the MDM2 x CK1 complex is both a negative regulator of p53 and a positive regulator of E2F-1 in undamaged cells. Although the treatment of cells with D4476 resulted in a partial p53-dependent growth arrest, the induction of p53-independent apoptosis by D4476 suggested a critical role for the MDM2 x CK1 complex in maintaining E2F-1 anti-apoptotic signaling. These data highlighting a pharmacological similarity between MDM2 and CK1 small molecule inhibitors and the fact that CK1 and MDM2 form a stable complex suggest that the MDM2 x CK1 complex is a component of a genetic pathway that co-regulates the stability of the p53 and E2F-1 transcription factors.

FIGURE 1.

Effects of CK1α depletion using siRNA on p53 protein stability. A375 cells were transfected with control siRNA (40 nm; lane 3), CK1α-specific siRNA (40 nm; lane 4), or VRK1-specific siRNA (40 nm; lane 5) for 72 h. A mock transfected control (lane 2) and an untreated control (DMEM only; lane 1) were included. The protein levels were determined by Western blotting with antibodies against the indicated proteins. RNAi, RNA interference.

FIGURE 2.

A CK1 inhibitor leads to a dose-dependent increase of p53 function and a decrease of E2F-1 protein levels. A375 cells were transfected with increasing concentrations (10–60 μm) of the CK1 inhibitor D4476 for 48 h (lanes 4–7). A DMSO solvent control (lane 2), a mock transfected control (lane 3), and an untreated control (DMEM only; lane 1) were included. A, cell lysates were examined by Western blotting with antibodies against the indicated proteins. B, ratio of hyperphosphorylated pRB (upper band on the blot)/hypophosphorylated pRB (lower band) is defined by changes in mobility in SDS-PAGE (26) and was quantified using Scion Image software.

FIGURE 3.

Lack of effect of CK1δ depletion on p53 protein stability. A375 cells were transfected with control siRNA (40 nm; lane 3) or CK1δ-specific siRNA (40 nm; lane 4) for 72 h. A mock transfected control (lane 2) and an untreated control (DMEM only; lane 1) were included. Protein levels were determined by Western blotting with antibodies against the indicated proteins.

FIGURE 4.

CK1α depletion using siRNA displays a synergistic effect with ionizing radiation on E2F-1 destabilization. A375 cells were transfected with control or CK1α-specific siRNA (40 nm; lanes 5–8) for 72 h and treated with (even-numbered lanes) or without (odd-numbered lanes) 5-gray x-ray and cultured for a further 24 h. A mock transfected control (lanes 3 and 4) and an untreated control (DMEM only; lanes 1 and 2) were included. A, protein levels were determined by Western blotting with antibodies against the indicated proteins. B, quantification of different protein levels were determined with Scion Image software and normalized against β-actin protein level measured for each different blot. RNAi, RNA interference.

FIGURE 5.

E2F-1 protein decrease following CK1α depletion using siRNA or D4476 treatment is p53-independent. A, A375 cells were transfected with control siRNA (80 nm; lane 3), CK1α-specific siRNA (40 and 40 nm of siRNA control; lane 4), p53-specific siRNA (40 and 40 nm of siRNA control; lane 5), or both CK1α and p53 siRNA (40 nm each; lane 6) for 72 h. A mock transfected control (lane 2) control and an untreated control (DMEM only; lane 1) were included. Protein levels were determined by Western blotting with antibodies against the indicated proteins. B and C, HCT116 wt (B) and HCT116 p53^−/− (C) cells were transfected with increasing concentrations (5–40 μm) of the CK1 inhibitor D4476 for 72 h. A DMSO solvent control, an untreated control (DMEM only), and a mock transfected control were included. The protein levels were determined by Western blotting with antibodies against the indicated proteins. RNAi, RNA interference.

FIGURE 6.

A protein-protein interaction in cells between CK1α and MDM2. A–C, A375 cells analyzed by immunoprecipitation using antibodies to MDM2 (A), p53 DO1 (B), or E2F-1 (C). Co-immunoprecipitation was determined by Western blotting with anti-CK1α (A–C) and antibodies against the immunoprecipitated proteins (MDM2 (A), p53 CM1 (B), and E2F-1 (C)). No antibody immunoprecipitation controls (No Ab IP, lanes 1 and 2 (A and B) and lanes 3 and 4 (C)) and immunoprecipitation controls without lysate (lanes 3 and 4 (A and B) and lanes 5 and 6 (C)) were included. The flow-through (FT) and the final eluate (E) were analyzed, along with the original lysate and the pre-cleared lysate (lanes 1 and 2 (C)).

FIGURE 7.

Nutlin-3 modulates the levels of MDM2-binding proteins p53 and E2F-1. A375 (A), HCT116 wt (B, lanes 1–4), and HCT116 p53^−/− (B, lanes 5–8) cells were treated for 24 h with 5 or 10 μm Nutlin-3 (lanes 3 and 4, respectively). A DMSO solvent control (lane 1) and an untreated control (DMEM only; lane 2) were included. The cell lysates were examined by Western blotting with antibodies against the indicated proteins.

FIGURE 8.

Analysis of the effects of the CK1-specific inhibitor D4476 and MDM2 inhibitor Nutlin-3 on cell cycle distribution of A375 melanoma cells. A, A375 cells were transfected with increasing concentrations (10–40 μm) of the CK1 inhibitor D4476 for 72 h. A DMSO solvent control, an untreated control (DMEM only), and a mock transfected control (only the last one is shown here) were included. B, A375 cells were treated for 24 h with 5 or 10 μm Nutlin-3. A DMSO solvent control and an untreated control were included. The cells were fixed in ethanol and then stained with propidium iodide. Cell DNA contents were determined by FACS and analyzed with FlowJo7 software.

FIGURE 9.

Analysis of the effects of the CK1-specific inhibitor D4476 on cell cycle distribution of HCT116 wt and p53^−/− cells. HCT116 wt (A) and HCT116 p53^−/− (B) cells were transfected with increasing concentrations (10–40 μm) of the CK1 inhibitor D4476 for 72 h. A DMSO solvent control, an untreated control (DMEM only), and a mock transfected control (only the last one is shown here) were included. The cells were fixed in ethanol then stained with propidium iodide. Cell DNA contents were determined by FACS and analyzed with FlowJo7 software.

FIGURE 10.

Analysis of the effects of the MDM2 inhibitor Nutlin-3 on cell cycle distribution of HCT116 wt and p53^−/− cells. HCT116 wt (A) and HCT116 p53^−/− (B) cells were treated for 24 h with 5 or 10 μm Nutlin-3. A DMSO solvent control and an untreated control (DMEM only) were included. The cells were fixed in ethanol then stained with propidium iodide. The cell DNA contents were determined by FACS and analyzed with FlowJo7 software.

FIGURE 11.

The MDM2 and CK1α complex form a genetic signal that regulates p53 and E2F-1 protein stability. Under normal conditions, CK1α interacts with MDM2 (condition 1), which promotes its binding to p53 and leads to p53 ubiquitination and degradation (condition 2) but also to inhibition of transactivation of p53 targets, p21 for example (condition 3). This is supported by our data: Nutlin-3 or CK1 depletion/inhibition stabilize p53 and p21, suggesting a pharmacologically similar interaction between p53 and the MDM2·CK1α complex. The interaction between MDM2 and CK1α should also allow the binding between E2F-1 and MDM2, an interaction that has been shown to prevent the degradation of E2F-1 by other E3-ligases (condition 4). This is also supported by the same data, but it is yet to be determined whether the MDM2-CK1α interaction and in turn whether CK1 binding to E2F-1 is phosphorylation-dependent or kinase docking-dependent. Under normal conditions, these phosphorylation or docking events mediated by CK1α therefore promote cell cycle progression. After DNA damage, the loss of the complex between MDM2 and CK1α could explain the destabilization of E2F-1. Alternatively, the MDM2·CK1α complex might be inactivated post-translationally, for example by changes in phosphorylation status. In addition, the destabilization of E2F-1 might be partly due to a loss of binding between CK1α and E2F-1.