The negative role of cyclin G in ATM-dependent p53 activation

Abstract

Cyclin G is one of the earliest p53 target genes to be identified, but its function in the p53 pathway has been elusive. Although the precise mechanisms of cyclin G in this novel network have not been explored, recent studies have demonstrated that cyclin G is a key regulator of the p53-Mdm2 network. Here we present evidence that cyclin G-mediated p53 regulation is dependent upon the status of ataxia-telangiectasia mutated (ATM) protein, which activates p53 in response to DNA damage. Abrogation of cyclin G enhances p53 accumulation and phosphorylation of p53 at the Ser-15 residue, resulting in cell cycle arrest. Ectopically expressed cyclin G significantly reduces the steady-state levels of p53 as well as that of phosphorylated p53 at Ser-15 after DNA damage in normal human dermal fibroblasts containing normal ATM. However, cyclin G does not cause similar reductions in p53 levels in ATM-mutated cells. We also show that translocation of cyclin G to the nucleus requires functional ATM. Thus, our findings identify a new role of cyclin G in ATM-dependent p53 regulation and in cell cycle regulation during DNA damage.

Main

Cyclin G is a member of the cyclin family and contains a well-conserved cyclin box, but lacks a PEST sequence or destruction box (Tamura et al., 1993). Compared to the other members of the cyclin family, the role of cyclin G in cell cycle check points is not well understood. Several reports show that cyclin G enhances cellular growth (Skotzko et al., 1995; Smith et al., 1997; Shimizu et al., 1998), and due to cyclin G overexpression in human cancer cells (Reimer et al., 1999) it is suspected that cyclin G functions as an oncogenic protein unlike other p53 target genes. In addition, cyclin G can increase the sensitivity of cancer cells to drug-induced apoptosis (Okamoto and Prives, 1999). Cyclin G is also a transcriptional target gene of the tumor-suppressor p53, and several studies have investigated its roles in a p53-dependent manner (Okamoto and Beach, 1995; Zauberman et al., 1995). Recent studies have demonstrated that cyclin G dephosphorylates Mdm2 by recruiting protein phosphatase 2A (PP2A), resulting in p53 stabilization (Okamoto et al., 1996, 2002). Compared to cyclin G+/+ cells, cyclin G-deficient cells exhibited a decrease in Mdm2 dephosphorylation and an increase in p53 protein levels (Kimura et al., 2001; Okamoto et al., 2002). It is also observed that cyclin G-mediated p53 stabilization is dependent upon Mdm2, yet, ubiquitination does not seem to be a main pathway of cyclin G-mediated p53 degradation (Ohtsuka et al., 2003). These data strongly suggest that cyclin G serves as a negative regulator of the p53 activation pathway, although the precise mechanisms of the cyclin G-p53 network have been obscure. All of these recent observations argue that cyclin G expression is more likely associated with growth promotion rather than arrest. Several types of DNA damage including ionizing radiation stimulate the ability of the ataxia-telangiectasia-mutated (ATM) kinase to phosphorylate p53 at serine-15 (Ser-15), thereby increasing the transcriptional activity of p53 (Siliciano et al., 1997; Canman et al., 1998; Alarcon-Vargas and Ronai, 2002). Various forms of cellular stress result in the phosphorylation of p53 at a number of key residues in the p53 protein that are essential for p53 stabilization.

In order to investigate the effects of cyclin G on the status of p53 phosphorylation, cyclin G+/+ and cyclin G−/− MEFs were treated with DNA-damaging agents. Cells were harvested either 8 h after exposure to γ-irradiation (5 Gy) or 24 h after MMC treatment (2.5 μg/ml). Western blotting was then used to analyse the total levels of p53 as well as phosphorylated p53. As shown in Figure 1, both γ-irradiation and MMC treatment significantly enhanced p53 accumulation and p21Waf1 induction in cyclin G−/− cells, in comparison to the levels of these proteins in cyclin G+/+ cells. Among the three N-terminus phosphorylation sites of p53 (Ser-15, -20 and -46) tested in this study, only the Ser-15 site was significantly phosphorylated by these DNA-damaging agents. The phosphorylation levels of p53 at Ser-15 were also higher in cyclin G−/− cells than those of cyclin G+/+ cells. However, we failed to detect expression of phosphorylated p53 protein on serine-20 or -46 (data not shown). These results suggest that p53 accumulation is positively affected by the absence of cyclin G through the phosphorylation of p53 at the Ser-15 site.

Figure 1
figure1

Abrogation of cyclin G increases the levels of p53 and phosphorylation at Ser-15 in response to DNA damage. MEF were isolated from mouse embryos using established procedures (Jensen et al., 2003). Cyclin G+/+ MEF and cyclin G−/− MEFs were exposed to γ-irradiation (5 Gy) and treated with MMC (2.5 μg/ml) for 24 h. Cells were harvested at the indicated times and lysed for Western blot analysis using antibodies against p53, phospho-p53 (Ser-15, -20, and -46), p21, cyclin G and β-actin

To examine whether cyclin G affects the control of cell cycle checkpoint in response to DNA damage, we next subjected cyclin G+/+ and cyclin G−/− MEFs to either 5 Gy of γ-irradiation or 2.5 μg/ml of MMC, and then performed a cell cycle analysis by FACS. As presented in Figure 2, the S-phase population of cyclin G−/− cells was lower than that of cyclin G+/+ cells even in unstressed condition. Moreover, exposure to genotoxic stresses reduced the S-phase population of cyclin G−/− cells more significantly than in that of cyclin G+/+ cells (7.6 vs 14.5% after MMC treatment; and 7.1 vs 14% after γ-irradiation). These results imply that loss of cyclin G leads to cell cycle arrest in response to DNA damage.

Figure 2
figure2

The effect of cyclin G loss on cell cycle arrest in response to DNA damage.Cyclin G+/+ and cyclin G−/− MEFs were treated with either MMC (2.5 μg/ml) or γ-irradiation (5 Gy), respectively. Subconfluent cells were pulse-labeled for 30 min with 10 μ M 5-bromo-2′-deoxyuridine (BrdU). Both adherent and floating cells were harvested, fixed in 80% ethanol, and consequently double stained with FITC-conjugated anti-BrdU antibody and propidium iodide (5 μg/ml). Cell cycle analysis was performed by a FACScan, and the data were processed using the CellQuest program. BrdU uptake, as measured by FITC fluorescence, is depicted on the y-axis. DNA content, as measured by propidium iodide fluorescence, is depicted on the x-axis. Populations of cells in different phases of the cell cycle are gated: G0/G1 population (a), S-phase population (b) and G2/M population (c). The percentage of cells in each gate per each sample is indicated in the table

It is well established that ATM is important in promoting p53 stabilization in response to genotoxic stress by phosphorylating p53 at Ser-15. We have demonstrated that loss of cyclin G also causes a significant increase in the level of phosphorylated p53 at Ser-15. We then examined whether cyclin G was involved in ATM-mediated p53 stabilization in response to DNA damage. The levels of total p53, phosphorylated p53 (Ser-15) and p21Waf1 were assayed to determine the effects of cyclin G expression in normal and ATM-mutated human dermal fibroblasts. Patient-derived AT cells and normal human dermal fibroblasts were infected with either adenovirus-expressing GFP (Ad-GFP) or cyclin G (Ad-cyclin G). At 48 h after infection, these cells were exposed to either γ-irradiation (5 Gy) or MMC (2.5 μg/ml). At 8 h after γ-irradiation or 24 h after MMC treatment, cells were lysed, and the levels of expression were analysed by Western blotting. As shown in Figure 3a and b, ectopically expressed cyclin G caused a reduction in the levels of total p53 and phosphorylated p53 protein (Ser-15) in response to DNA damage. The expression level of p21Waf1, a p53 target, was also significantly decreased in normal HDFs. In contrast, cyclin G-mediated p53 degradation was not observed in ATM-mutated cells (Figure 3a and b). These results indicate that cyclin G requires functional ATM in order to negatively regulate p53 under normal conditions.

Figure 3
figure3

Cyclin G-mediated degradation of p53 protein requires wild-type ATM. Normal and ATM-mutated human dermal fibroblasts were infected with either Ad-GFP or Ad-cyclin G at a multiplicity of infection (moi) of 50. At 48 h after infection, cells were exposed to either 5 Gy of γ-irradiation (a) or 2.5 μg/ml of MMC (b). Cells were harvested 8 h after γ-irradiation (IR) or 24 h after MMC treatment and lysed for Western blot analysis using antibodies against p53 (Ab-6, Oncogene), phospho-p53 (Ser-15, Signaling), p21 (Ab-1 and Ab-6, Oncogene), cyclin G (Santa Cruz and Upstate) and β-actin (AC-15, Sigma)

To further support the hypothesis that the negative role of cyclin G on p53 can be affected by functional ATM, we investigated whether normal ATM is also required for nuclear localization of cyclin G. Normal and ATM-mutated dermal fibroblasts obtained from ataxia-telangiectasia patients were transfected with either vector alone (pEGFP) or pEGFP-cyclin G (GFP-cyclin G). At 48 h after transfection, cells were exposed to 5 Gy of γ-irradiation. After 8 h, cells were fixed as described in Materials and methods section and counterstained with 4,6-diamidino-2-phenylinddole (DAPI) to visualize the nucleus. Figure 4 shows that, in normal HDFs, green fluorescence proteins (GFP) are diffusely located in both the cytoplasm and nucleus, while GFP–cyclin G fusion proteins are predominantly localized only in the nucleus. ATM-mutated HDFs, in contrast, show that both GFP and GFP–cyclin G are diffusely located in both cytoplasm and the nucleus. We also tested whether ATM directly recruited cyclin G into the nucleus, but immunoprecipitation failed to show a direct interaction between these two proteins (data not shown). Together, we provide evidence that cyclin G negatively regulates ATM-dependent p53 activation and plays an important role in cell cycle checkpoint control during DNA damage.

Figure 4
figure4

Aberrant cellular localization of cyclin G in AT cells. Normal and AT cells were transfected with pEGFP (vector control) or pEGFP–cyclin G (GFP–cyclin G fusion plasmid) using the Nucleofector electrophoration system (AMAXA, program p-22). At 48 h after transfection, cells were exposed to 5 Gy of γ-irradiation and incubated for an additional 8 h. Cells were fixed with 3.7% formaldehyde for 20 min, followed by permeabilization in 0.2% Triton X-100 for 5 min and counterstained with DAPI to visualize the nucleus. Localization of GFP or GFP–cyclin G was determined using a fluorescence microscope. Nuclei were stained with DAPI. Two images (green and blue) are overlaid (Merge, light-blue). All images were processed by the use of Adobe Photoshop software

Although the structural and biochemical aspects of cyclin G have been well established, the biological and functional relevance of cyclin G in a cellular context remains elusive. A number of studies have speculated on the possible cellular roles of the cyclin G protein, and these studies imply that cyclin G plays an essential role in cell cycle checkpoint control in response to DNA damage through the p53-Mdm2 network, and cyclin G may also have the potential to exert oncogenic properties (Kimura et al., 2001; Chen, 2002; Okamoto et al., 2002; Jensen et al., 2003; Ohtsuka et al., 2003). Our present study demonstrates that cyclin G plays an important role in regulating p53 function and in modulating the cellular response to DNA damage in an ATM kinase-dependent manner. Here, we identify a novel ATM-p53 network involving cyclin G and propose that cyclin G functions as a negative regulator of p53 via the ATM pathway. It is well established that ATM belongs to the protein kinase family encoding a phosphoinositide 3-kinase-related domain. Through phosphorylation of p53 at Ser-15, ATM plays a crucial role in activating p53 and allowing p53 to escape the negative regulation of Mdm2. ATM kinase also phosphorylates Mdm2 at S395, thereby counteracting the ability of Mdm2 to target p53 for degradation in vivo (Siliciano et al., 1997; Canman et al., 1998; Giaccia and Kastan, 1998; Maya et al., 2001; Alarcon-Vargas and Ronai, 2002). It has been speculated that the activated p53 and deactivated Mdm2 are the more phosphorylated forms of each protein (Maya et al., 2001; Alarcon-Vargas and Ronai, 2002). Cyclin G-deficient MEFs were shown to contain hyperphosphorylated Mdm2 and higher p53 levels when compared to cyclin G+/+ MEFs (Kimura et al., 2001; Chen, 2002; Okamoto et al., 2002). Although the direct effects of cyclin G on the status of p53 phosphorylation have been unclear, recent studies demonstrated that cyclin G negatively regulates p53 stabilization by recruiting PP2A to dephosphorylate Mdm2 (Kimura et al., 2001; Chen, 2002; Okamoto et al., 2002). Consistent with previous studies, we found that loss of cyclin G enhanced phosphorylation of p53 at Ser-15 and promoted p53 stabilization. Cyclin G-mediated modulation of total p53 and phosphorylated p53 is absent in AT cells derived from ataxia-telangiectasia patients, while cyclin G significantly represses p53 expression in normal counterparts. While there are some reports that p53 and its target genes are not activated by ionizing radiation in AT cells, our results indicate that p53 and its target gene p21 are inducible, and that phosphorylation of p53 on Ser-15 is not affected in AT cells used in this study as reported earlier elsewhere (Siliciano et al., 1997). These findings suggest that the negative effect of cyclin G on p53 regulation requires the presence of normal ATM. Cyclin G is found in the nucleus (Smith et al., 1997; Shimizu et al., 1998), and it binds to p53 and Mdm2 (Kimura et al., 2001; Chen, 2002; Okamoto et al., 2002; Ohtsuka et al., 2003). However, this nuclear localization of cyclin G is apparent only in normal fibroblast cells. In ATM-mutated cells, cyclin G fails to locate in the nucleus and instead is diffused throughout the nucleus and cytoplasm. These results also support our hypothesis that cyclin G is involved in regulating p53 function in the ATM-p53 network. From this, we suggest that cyclin G participates in ATM downstream signaling by interacting with Mdm2 and p53 and hindering p53 activation via its interaction with PP2A. Whether cyclin G plays a role in growth promotion or suppression like other p53 target genes has been controversial (Skotzko et al., 1995; Smith et al., 1997; Shimizu et al., 1998; Reimer et al., 1999; Zhao et al., 2003). Our results strongly imply that cyclin G expression is associated with growth promotion rather than tumor suppression. Moreover, a recent report clearly shows that cyclin G-null mice exhibit a significant decrease in tumor incidence, mass and malignancy after treatment with potent carcinogens (Jensen et al., 2003).

Taken together, our results demonstrate that cyclin G can negatively influence p53 function in the presence of functional ATM and predict that cyclin G may target proteins to form a complex with the p53 protein. Identification of additional component(s) in this cellular context should provide important insights into the regulation of p53 and its target genes. Thus, ATM-dependent regulation of cyclin G on p53 expression may be important for the efficient execution of cell cycle transitions during DNA damage.

References

  1. Alarcon-Vargas D and Ronai Z . (2002). Carcinogenesis, 23, 541–547.

  2. Canman CE, Lim D, Cimprich KA, Taya Y, Tamai K, Sakaguchi K, Appella E, Kastan MB and Siliciano JD . (1998). Science, 281, 1677–1679.

  3. Chen X . (2002). Dev. Cell, 2, 518–519.

  4. Giaccia AJ and Kastan MB . (1998). Genes Dev., 12, 2973–2983.

  5. Jensen MR, Factor VM, Fantozzi A, Helin K, Huh CG and Thorgeirsson SS . (2003). Hepatology, 37, 862–870.

  6. Kimura SH, Ikawa M, Ito A, Okabe M and Nojima H . (2001). Oncogene, 20, 3290–3300.

  7. Maya R, Balass M, Kim ST, Shkedy D, Leal JF, Shifman O, Moas M, Buschmann T, Ronai Z, Shiloh Y, Kastan MB, Katzir E and Oren M . (2001). Genes Dev., 15, 1067–1077.

  8. Ohtsuka T, Ryu H, Minamishima YA, Ryo A and Lee SW . (2003). Oncogene, 22, 1678–1687.

  9. Okamoto K and Beach D . (1995). EMBO J., 13, 4816–4822.

  10. Okamoto K, Kamibayashi C, Serrano M, Prives C, Mumby MC and Beach D . (1996). Mol. Cell. Biol., 16, 6593–6602.

  11. Okamoto K, Li H, Jensen MR, Zhang T, Taya Y, Thorgeirsson SS and Prives C . (2002). Mol. Cell, 9, 761–771.

  12. Okamoto K and Prives C . (1999). Oncogene, 18, 4606–4615.

  13. Reimer CL, Boras AM, Kurdistanni SK, Garreau JR, Chung M, Aaronson SA and Lee SW . (1999). J. Biol. Chem., 274, 11022–11029.

  14. Shimizu A, Nishida J, Ueoka Y, Kato K, Hachiya T, Kuriaki Y and Wake N . (1998). Biochem. Biophys. Res. Commun., 242, 529–533.

  15. Siliciano JD, Canman CE, Taya Y, Sakaguchi K, Appella E and Kastan MB . (1997). Genes Dev., 11, 3471–3481.

  16. Skotzko M, Wu L, Anderson WF, Gordon EM and Hall FL . (1995). Cancer Res., 55, 5493–5498.

  17. Smith ML, Kontny HU, Bortnick R and Fornace Jr. AJ . (1997). Exp. Cell Res., 230, 61–68.

  18. Tamura K, Kanaoka Y, Jinno S, Nagata A, Ogiso Y, Shimizu K, Hayakawa TT, Nojima H and Okayama H . (1993). Oncogene, 8, 2113–2118.

  19. Zauberman A, Lupo A and Oren M . (1995). Oncogene, 10, 2361–2366.

  20. Zhao L, Samuels T, Winckler S, Korgaonkar C, Tompkins V, Horne MC and Quelle D.E . (2003). Mol. Cancer Res., 1, 195–206.

Download references

Acknowledgements

We thank T Ouchi for helpful discussion, the technical help of K-T Kim, and W Wong and M Meyer for proof-reading the manuscript. This work was supported in part by Grants CA85681 and CA78356.

Author information

Correspondence to Sam W Lee.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Ohtsuka, T., Jensen, M., Kim, H. et al. The negative role of cyclin G in ATM-dependent p53 activation. Oncogene 23, 5405–5408 (2004). https://doi.org/10.1038/sj.onc.1207693

Download citation

Keywords

  • p53
  • cyclin G
  • ATM
  • cell cycle
  • DNA damage

Further reading