The cyclin-dependent kinase inhibitor p27Kip1 is degraded in late G1 phase by the ubiquitin-proteasome pathway, allowing cells to enter S phase. Due to accelerated degradation of p27Kip1, various human cancers express low levels of p27Kip1 associated with poor prognosis. S-phase kinase–associated protein 2, the F-box protein component of an SCF ubiquitin ligase complex, is implicated in degradation of p27Kip1 during S-G2 phases. Recently, Kip1 ubiquitination–promoting complex has been reported as another ubiquitin ligase that targets cytoplasmic p27Kip1 exported from the nucleus in G0-G1 phases. Here, we identified a RING-H2–type ubiquitin ligase, Pirh2, as a p27Kip1-interacting protein. Endogenous Pirh2 physically interacted with endogenous p27Kip1 in mammalian cells. Pirh2 directly ubiquitinated p27Kip1 in an intact RING finger domain-dependent manner in vivo, as well as in vitro. Ablation of endogenous Pirh2 by small interfering RNA increased the steady-state level of p27Kip1 and decelerated p27Kip1 turnover. Depletion of Pirh2 induced accumulation of p27Kip1 in both the nucleus and cytoplasm. Pirh2 expression was induced from late G1-S phase, whereas p27Kip1 was decreased in synchronization with accumulation of Pirh2. Furthermore, reduction of Pirh2 resulted in an impairment of p27Kip1 degradation and an inhibition of cell cycle progression at G1-S transition in a p53-independent manner. Overall, the results indicate that Pirh2 acts as a negative regulator of p27Kip1 function by promoting ubiquitin-dependent proteasomal degradation. [Cancer Res 2007;67(22):10789–95]
- ubiquitin ligase
- cell cycle
Cell cycle progression is controlled by cyclin-dependent kinases (CDK; ref. 1). The CDK activities are negatively regulated by CDK inhibitor proteins (CKI; ref. 2). In the quiescent and early G1 phase, p27Kip1, one of the Cip/Kip-type CKIs, exists in abundance in the cell nucleus to suppress cell cycle progression ( 3). Although the level of p27Kip1 mRNA does not vary during the cell cycle, p27Kip1 protein is degraded via the ubiquitin-proteasome system at late G1 phase, and thereby, cyclin D–CDK4 is activated to phosphorylate retinoblastoma protein. S-phase kinase–associated protein 2 (Skp2), an F-box substrate recognition subunit of the SCF ubiquitin ligase complex, recognizes Thr187-phosphorylated p27Kip1 in collaboration with cdc kinase subunit 1 (Cks1) to promote ubiquitination of p27Kip1 ( 4– 7). p27Kip1 is accumulated in Skp2-knockout mouse tissues and embryonic fibroblast cells ( 8). Skp2 is therefore thought to be one of the ubiquitin ligases for p27Kip1.
Many studies have shown that a low expression level of p27Kip1 is associated with poor prognosis in many types of human cancers, such as colon, breast, lung, and stomach cancer ( 9– 13). The low expression level of p27Kip1 is due to accelerated degradation of p27Kip1 in cancer cells. Moreover, some studies have shown high expression levels of Skp2 protein in human cancers, such as oral squamous cell carcinomas, colorectal carcinomas, non–Hodgkin's lymphomas, and non–small cell lung carcinomas ( 14, 15). Because it was shown in these studies that a high expression level of Skp2 is correlated with a low expression level of p27Kip1, Skp2/Cks1 is thought to be involved in degradation of p27Kip1 in human cancers.
However, expression of Skp2 begins in S phase, except for Skp2-overexpressing tumors ( 16, 17). Although Skp2 knockout cells are impaired in degradation of p27Kip1 in S-G2 phases, Hara et al. showed that Skp2-independent degradation of p27Kip1 was observed in Skp2-knockout lymphocytes in G1 phase ( 8, 18). It is therefore assumed that Skp2-independent mechanisms exist for stability control of p27Kip1. The Skp2-independent ubiquitination activity is unrelated to Thr187 phosphorylation of p27Kip1. Stability of p27Kip1 is affected by not only phosphorylation of Thr187 but also by phosphorylation of Ser10 ( 19– 22). Furthermore, phosphorylation of Ser10 regulates nuclear export of p27Kip1 for degradation via the ubiquitin-proteasome pathway in an Skp2-independent manner ( 21, 23). Recently, the cytoplasmic RING finger-type E3 complex Kip1 ubiquitination–promoting complex (KPC) was identified as a ubiquitin ligase for p27Kip1 ( 24). In the cytoplasm, KPC mediates ubiquitination of p27Kip1 exported from the nucleus in G0 and early G1 phase in a Ser10 phosphorylation–dependent manner. However, nuclear ubiquitin ligase for p27Kip1 that ubiquitinates p27Kip1 at G1-S transition is still unknown.
In the present study, we did yeast two-hybrid screening using the amino-terminal region of p27Kip1 as bait to identify novel p27Kip1-binding proteins that regulate p27Kip1 stability. We found that the RING finger protein Pirh2, a ubiquitin ligase for p53 ( 25), interacted with p27Kip1. Pirh2 ubiquitinated p27Kip1 in vivo, as well as in vitro, and thereby, p27Kip1 was degraded via the ubiquitin-proteasome pathway. Down-regulation of endogenous Pirh2 by small interfering RNA (siRNA) resulted in deceleration of p27Kip1 turnover and inhibition of cell cycle progression at G1-S transition in a p53-independent manner. Our results suggest that Pirh2 regulates p27Kip1 stability and cell cycle progression at G1-S transition.
Materials and Methods
Cell culture, reagents, and antibodies. All cells were maintained in DMEM supplemented with 10% heat-inactivated fetal bovine serum and antibiotics. T98G cells were synchronized at G0 phase by serum deprivation as described previously ( 26). MG132 (Peptide Institute) and cycloheximide (Wako) were purchased. p27Kip1 was immunoprecipitated by using polyclonal antibody C-19 (Santa Cruz Biotechnology) and detected by monoclonal antibody (57, BD Biosciences). Polyclonal antibodies against human Pirh2 (Calbiochem), mouse Pirh2 (T-18, Santa Cruz Biotechnology), luciferase (Sigma), Skp2 (H-435, Santa Cruz Biotechnology), Cks1 (FL-79, Santa Cruz Biotechnology), and cyclin A (H-432, Santa Cruz Biotechnology) were purchased. Monoclonal antibodies against FLAG-epitope (M2, Sigma), HA-epitope (12CA5, Roche), Xpress-epitope (Invitrogen), β-actin (AC-15, Sigma), Hsp90 (BD Biosciences), p130 (BD Biosciences), α-tubulin (DM1A, Sigma), and p53 (DO-1, Santa Cruz Biotechnology) were used for immunoprecipitation and/or immunoblotting. Anti–cAMP-responsive element binding protein (CREB) polyclonal antibody (from Dr. M. Montminy, Salk Institute) and anti-KPC1 antiserum (from Dr. K.I. Nakayama, Kyushu University) were kindly provided.
Mammalian expression vectors and transfection. Human Pirh2 cDNA amplified by reverse transcription–PCR (RT-PCR) from HCT116 cell RNA was cloned into the pcDNA4-HisMax expression vector (Invitrogen). The resultant pcDNA4-HisMax-hPirh2 encodes 6×His and the Xpress-epitope tag fused with the amino terminus of human Pirh2. PCR-based site-directed mutagenesis was used to obtain cDNA fragments encoding RING finger mutant hPirh2 (C148A). pcDNA3-FLAG-p27 and pCGN-HA-ubiquitin were described previously ( 19, 27). 293 and HepG2 cells were transfected by the calcium phosphate method.
Immunoprecipitation, Ni2+-resin precipitation, and immunoblotting. Cells were lysed in immunoprecipitation lysis buffer [0.5% Triton X-100, 150 mmol/L NaCl, 20 mmol/L Tris-HCl (pH 7.6), and protease inhibitor mix] or radioimmunoprecipitation assay buffer [50 mmol/L Tris-HCl (pH 8.0), 150 mmol/L NaCl, 0.1% SDS, 0.5% deoxycholate, 1% Triton X-100, and protease inhibitors mix]. Immunoprecipitation, Ni2+-resin precipitation, and immunoblotting were done as described previously with some modifications ( 27, 28).
In vitro ubiquitination assay. Glutathione S-transferase (GST)–tagged hp27Kip1, hPirh2, and hPirh2 (C148A) protein were expressed in Escherichia coli DH5α and affinity-purified with glutathione–Sepharose 4B (GE Healthcare), and then the GST tag was removed by cleavage with PreScission protease (GE Healthcare). p27Kip1 protein was incubated with or without Pirh2 protein at 30°C for 30 min in a 20-μL ubiquitination mixture supplemented with 50 mmol/L Tris-HCl (pH 8.3), 5 mmol/L MgCl2, 2 mmol/L DTT, 10 mmol/L phosphocreatinine, 0.2 units/mL phosphocreatinine kinase, 5 mmol/L adenosine-5′-triphosphate, 2 μL GST-ubiquitin, 50 μg/mL ubiquitin aldehyde, 250 μmol/L MG132, protease inhibitor mix, E1; (120 ng, Boston Biochem), and UbcH5b (50-260 ng, Boston Biochem). After incubation, the reactants were subjected to immunoblotting with anti-p27Kip1 antibody.
RNA interference. hPirh2 siRNA1 (5′-CAUGCCCAACAGACUUGUG-3′), hPirh2 siRNA2 (5′-GUAGCACAGACUCCUAUGC-3′), hPirh2 siRNA3 (5′-UGGACGAUCCACUGUUCAG-3′), mPirh2 siRNA1 (5′-AUUUAUGCCUAACCACGAA-3′), mPirh2 siRNA2 (5′-CCAUGCCAUCCGAAUACCA-3′), Skp2 siRNA1 (5′-CUGCGGCUUUCGGAUCCCA-3′), Skp2 siRNA2 (5′-GCAUGUACAGGUGGCUGUU-3′), KPC1 siRNA1 (5′-UCAACAGCGUCCUCAAUCA-3′), and KPC1 siRNA2 (5′-CCACCAUCGUGUCUGUAGA-3′). siRNA oligonucleotides with 3′ dTdT overhangs were synthesized by Greiner Bio-One. Nonsilencing control siRNA (Qiagen) was purchased. The cells were transfected with siRNA oligonucleotides by using Oligofectamine or Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's protocol.
Subcellular fractionation. Nuclear and cytoplasmic extracts were prepared by using NE-PER nuclear and cytoplasmic extraction reagents (Pierce) according to the supplier's protocol.
Cell cycle analysis. Cells were fixed with 70% ethanol in PBS (−) at −20°C overnight. After treatment with 0.5 mg/mL of RNase, the cells were stained with 50 μg/mL of propidium iodide. The DNA contents were examined by flow cytometry (Epics XL, Beckman Coulter).
Identification of Pirh2 as a p27Kip1-interacting protein. To identify novel proteins interacting with p27Kip1, we did yeast two-hybrid screening of 3.2 × 106 clones of a rat brain cDNA library using the amino-terminal portion of p27Kip1 (amino acids 2-26 residues) as a bait. After second screenings, four distinct interacting candidates were identified after DNA sequencing. Among these clones, clone 9 encoded an unidentified protein that possessed a typical RING-H2 domain. During this study, two groups reported a RING-H2 protein with E3 ubiquitin ligase activity that was identical to clone 9. Beitel et al. reported the protein as an androgen receptor–interacting protein, whereas Leng et al. reported the protein as a p53-inducible and p53-interacting protein (Pirh2; p53-inducible protein with RING-H2 domain) which promotes ubiquitination of p53 ( 25, 29). Clone 9 is called Pirh2 hereafter. Recently, Logan et al. reported Pirh2 as a ubiquitin ligase for histone deacetylase 1 ( 30).
To verify physical interaction between p27Kip1 and Pirh2 in mammalian cells, expression plasmids for His6-Xpress-Pirh2 and FLAG-p27Kip1 were cotransfected into HEK293 cells. When His-tagged Pirh2 was precipitated with Ni2+ resin and immunoblotted with anti-p27Kip1 antibody, p27Kip1 was only coprecipitated in the presence of transfected Pirh2 ( Fig. 1A ). To further confirm this interaction, we examined whether endogenous Pirh2 associated with endogenous p27Kip1. COLO320DM cell extracts were subjected to immunoprecipitation with anti-Pirh2 or antiluciferase as an isotype control and immunoblotted with anti-p27Kip1 antibody. As shown in Fig. 1B, p27Kip1 was coprecipitated with Pirh2 in the case of precipitation with anti-Pirh2 but not in an isotype control. Taken together, the results suggest that Pirh2 and p27Kip1 physically interact with each other in mammalian cells.
Pirh2 promotes ubiquitination of p27Kip1 in vivo as well as in vitro. To determine whether Pirh2 can promote ubiquitination of p27Kip1 in vivo, FLAG-p27Kip1 and HA-ubiquitin expression plasmids were cotransfected into HEK293 cells with Xpress-Pirh2 or Xpress-Pirh2 mutants (in which residue Cys148 within the RING finger domain was replaced with alanine, C148A) expression plasmid. Cell extracts were subjected to immunoblotting with anti-p27Kip1 antibody. Cotransfection with Pirh2 enhanced ubiquitination of p27Kip1, which was not seen in the absence of Pirh2 or in the case of cotransfection with RING mutant (C148A) Pirh2 ( Fig. 2A ). Pirh2-mediated ubiquitination of p27Kip1 was also observed in HepG2 cells ( Fig. 2B). In this experiment, only HA-ubiquitinated p27Kip1 was detectable because p27Kip1 was immunoprecipitated twice as described in Fig. 2. Immunoblots with anti-HA and anti-p27Kip1 antibodies gave ladders with similar mobility shifts identical to the molecular mass of a ubiquitin chain ( Fig. 2B; note that Ub1-p27 or Ub2-p27 equally migrate in all panels). These results suggested that the high–molecular mass species recognized by anti-p27Kip1 antibody were indeed ubiquitin-modified p27Kip1. Pirh2 could not induce ubiquitination of another Cip/Kip family CKI member, p57Kip2, which is a close relative of p27Kip1, indicating that Pirh2 specifically ubiquitinated p27Kip1 (data not shown). Pirh2 did not discriminate between phosphorylated and unphosphorylated forms of Ser10, because wild-type p27Kip1 and Ser10 mutants (S10A and S10E) were ubiquitinated by Pirh2 to similar extents (data not shown). We then investigated whether ubiquitination of p27Kip1 could be reconstituted in a cell-free system with purified components. Pirh2-mediated polyubiquitination was detected only with all of the reaction components and depended on the presence of an intact RING finger domain ( Fig. 2C). In this in vitro ubiquitination assay, the high–molecular mass p27Kip1 bands disappeared when lysine-free GST-ubiquitin was used, indicating that p27Kip1 ubiquitination induced by Pirh2 consists of polymerization reaction of the ubiquitin chain ( Fig. 2D). These observations suggest that Pirh2 promotes ubiquitination of p27Kip1 in vivo, as well as in vitro, and that p27Kip1 is a direct substrate for Pirh2.
Ablation of cellular Pirh2 causes an accumulation and stabilization of p27Kip1 protein. To clarify the role of Pirh2 in a more physiologic setting, endogenous Pirh2 mRNA was targeted with siRNA, and the effect on steady-state level of endogenous p27Kip1 was assessed. Pirh2 targeting siRNA, but not control siRNA, induced a decrease in intracellular Pirh2 in COLO320DM cells ( Fig. 3A ). Concomitant with the decrease in Pirh2 level, p27Kip1 protein accumulated. p53 protein level was not altered by siRNA treatment ( Fig. 3A, top), presumably because of its mutation in COLO320DM cells ( 31). Accumulation of p27Kip1 by Pirh2 ablation was also observed in T98G ( Fig. 3D), HeLa, U2OS, Saos2, and MCF7 cells (data not shown). Because the expression levels of Skp2, Cks1, an essential cofactor for p27Kip1 ubiquitination by SCFSkp2 complex ( 4, 6), and p130, one of the targets for SCFSkp2 ( 32), were not altered even in Pirh2-depleted cells ( Fig. 3A, top), accumulation of p27Kip1 was not due to unexpected inactivation of SCFSkp2 by Pirh2 depletion. Treatment with MG132 increased p27Kip1 protein levels in control cells, whereas p27Kip1 levels in Pirh2-depleted cells were hardly affected, indicating involvement of Pirh2 function in proteasome-dependent destabilization of p27Kip1 ( Fig. 3A, bottom). To analyze the effect of Pirh2 depletion on p27Kip1 protein turnover, we did a cycloheximide assay. As shown in Fig. 3B, the results indicated that ablation of Pirh2 prolonged the half-life of p27Kip1 protein in COLO320DM cells. These data indicated that Pirh2 specifically promotes turnover of p27Kip1 protein at the physiologic level.
Next, we examined the effect of Skp2 or KPC1, a component of KPC ubiquitin ligase complex with RING domain that catalyzes p27Kip1 ubiquitination, on the steady-state level of endogenous p27Kip1. In Skp2−/− MEFs, p27Kip1 protein was accumulated compared with wild-type MEFs. Accumulation of p27Kip1 protein was also observed even in Skp2−/− MEFs transfected with Pirh2 siRNA ( Fig. 3C). Depletion of Pirh2 or KPC1 alone induced accumulation of p27Kip1 protein ( Fig. 3D). Ablation of Pirh2 together with KPC1 resulted in an additive increment of p27Kip1 protein level. Because KPC1 expression was not affected by Pirh2 siRNA transfection, accumulation of p27Kip1 in Pirh2-depleted cells was not due to unexpected inactivation of KPC complex by Pirh2 depletion ( Fig. 3D). These data indicated that not only SCFSkp2 and KPC complexes but also Pirh2 regulate p27Kip1 protein levels and that Pirh2 regulates p27Kip1 protein stability in a manner independent of SCFSkp2 and KPC complex.
Pirh2 destabilizes p27Kip1 in both the nucleus and cytoplasm. Next, we examined the subcellular localization of endogenous Pirh2 and p27Kip1 by immunoblot analysis of nuclear and cytoplasmic fractions prepared from COLO320DM cells treated with control or Pirh2 siRNA. As shown in Fig. 4 , Pirh2 and p27Kip1 were detected in both nuclear and cytoplasmic fractions. In cells depleted of Pirh2 by siRNA, p27Kip1 was accumulated in both nuclear and cytoplasmic fractions compared with cells transfected with control siRNA. On the other hand, nuclear p27Kip1 was predominantly accumulated in Skp2-depleted cells compared with cytoplasmic p27Kip, and only cytoplasmic p27Kip1 was accumulated in KPC1-depleted cells. These results indicated that Pirh2 promoted ubiquitination of p27Kip1 and, subsequently, degradation of it in both the nucleus and cytoplasm.
Pirh2 is induced from late G1 to S phase. To determine whether the expression of Pirh2 is dependent on cell cycle progression, we investigated the expression levels of Pirh2 in various cell cycle phases. T98G cells were synchronized at G0 phase by serum deprivation, and then cell cycle progression was stimulated by a serum-containing medium. The expression level of Pirh2 was low in G0 and early G1 phases compared with that in an asynchronous cell extract and gradually increased toward S phase (Supplementary Fig. S1). On the other hand, expression of Skp2 began from S phase. The expression level of p27Kip1 was inversely diminished in synchronization with an increase in the expression level of Pirh2. These observations indicated that Pirh2 was induced from late G1 phase toward S phase.
Pirh2 contributes to the degradation of p27Kip1 and cell cycle progression at G1-S transition in a p53-independent manner. Next, we examined the effects of depletion of endogenous Pirh2 on p27Kip1 protein degradation and cell cycle progression at G1-S transition. In this experiment, we used a glioblastoma cell line T98G, in which the p53 gene is mutated ( 33) to exclude any effects of p53 on cell cycle progression. Moreover, T98G cells are easily arrested at G0 phase by serum depletion and then synchronously reenter into G1 phase by serum stimulation ( 26). We transfected T98G cells with control or Pirh2 siRNA and then synchronized the cells in G0 phase by serum withdrawal. After serum stimulation, cells were harvested at the indicated times and analyzed expression levels of p27Kip1 and Pirh2 by Western blotting. The reduction of p27Kip1 protein in late G1 phase was markedly inhibited by treatment with Pirh2 siRNA ( Fig. 5A ). Concomitant with impairment of p27Kip1 protein degradation, ablation of Pirh2 resulted in an arrest in cell cycle progression at G1-S transition. These results indicate that Pirh2 activity is required for p27Kip1 degradation and cell cycle progression at G1-S transition in a p53-independent manner. Finally, we compared the effects on p27Kip1 protein abundance and cell cycle progression of depletion of three ubiquitin ligases for p27Kip1. T98G cells transfected with control, Pirh2, Skp2, or KPC1 siRNA were synchronized in G0 phase and released as described above. Knockdown efficiencies of Pirh2, Skp2, and KPC1 at 18 h after release were about 60%, 75%, and 50%, respectively. As shown in Fig. 5B, reduction of Pirh2 most strongly induced both an impairment of p27Kip1 protein degradation and G1-S arrest when compared with Skp2 and KPC1. Taken together, these data indicate that Pirh2 regulates p27Kip1 stability and cell cycle progression at G1-S transition.
It has been shown that the cellular amount of p27Kip1 is mainly regulated at the posttranslational level and is tightly controlled in synchronization with the cell division cycle ( 34– 36). It has also been shown that Skp2−/− cells are impaired in disruption of p27Kip1 in S-G2 phases but that p27Kip1 still undergoes ubiquitination and is degraded in late G1 phase even in Skp2-null cells ( 8, 18). These studies suggest that another ubiquitin ligase promotes proteasomal degradation of p27Kip1 in late G1 phase.
In this study, we identified RING finger type ubiquitin ligase Pirh2 as a new p27Kip1-interacting protein. Pirh2 certainly associated with p27Kip1 in mammalian cells ( Fig. 1). Pirh2 directly ubiquitinated p27Kip1 in intact RING finger domain-dependent manner in vivo, as well as in vitro ( Fig. 2). The expression level of Pirh2 was low in G0 and early G1 phases and increased toward S phase, whereas the expression level of p27Kip1 was inversely diminished in synchronization with an increase in the expression level of Pirh2 (Supplementary Fig. S1). Furthermore, depletion of Pirh2 in late G1 phase markedly impaired degradation of p27Kip1 and cell cycle progression at G1-S transition ( Fig. 5). These results indicated that Pirh2, which is distinct from Skp2, functions as a ubiquitin ligase for p27Kip1 at G1-S transition.
In addition to p27Kip1, Pirh2 has been shown to associate with other molecules, including p53, androgen receptor, TIP60, and HDAC1 ( 25, 29, 30, 37). We showed that the block in the cell cycle progression due to Pirh2 depletion does not require p53 activity. To show that Pirh2 depletion–induced cell cycle arrest does not occur when p27Kip1 is suppressed, we tried to perform the same experiment, as shown in Fig. 5, by using Pirh2 and/or p27Kip1 siRNA. However, Pirh2 blocked cell cycle progression even in the T98G cells simultaneously treated with p27Kip1 siRNA (data not shown). This is because siRNA-mediated depletion of p27Kip1 was inefficient in T98G cells, and thereby, remaining p27Kip1 was stabilized and accumulated by Pirh2 depletion. Therefore, we cannot rule out the participation of other components than p53 and p27Kip1, which might also be stabilized in the absence of Pirh2.
Here, we evaluate the effects of depletion of three E3s, Pirh2, Skp2, and KPC1, on the cell cycle progression at G1-S phase with regard to changes in the DNA contents and expression levels of cyclin A and p27Kip1 ( Fig. 5B). Although cyclin A was induced at 18 and 24 h after release in these three E3-depleted cells, the expression levels were different. Careful comparison of protein levels of cyclin A and p27Kip1 at 18 and 24 h with these three-ligase–depleted cells revealed that the cyclin A induction was most impaired in Pirh2-depleted cells, and in particular, the reduction in p27Kip1 level was most greatly impaired in Pirh2-depleted cells. We think that the relatively higher amount of p27Kip1 may more effectively impair the cell cycle progression and that the low cyclin A levels in the extract from Pirh2-depleted cells reflect the existence of low number of S phase cells expressing cyclin A at 18-h to 24-h position. Taken together, differences in the cell cycle profiles in Fig. 5B were closely related to differences in degree of impairment of p27Kip1 degradation at G1-S transition. These results suggest that Pirh2 is a unique ubiquitin ligase in regulating p27Kip1 stability and cell cycle progression at G1-S transition.
The cytoplasmic RING finger-type E3 complex KPC mediates ubiquitination of p27Kip1 exported from the nucleus in G0 and early G1 phase in a Ser10 phosphorylation–dependent manner ( 24). On the other hand, Pirh2 exists in both the nucleus and cytoplasm, and Pirh2 destabilizes p27Kip1 in both fractions ( Fig. 4). Relatively low levels of Pirh2 were present in early G1 phase, and the increment of Pirh2 began from 9 h after serum addition (mid-late G1 phase) to be apparent at 12 to 15 h (G1-S transition) after serum stimulation, whereas KPC1 level did not change during the cell cycle ( Fig. 5B). It is possible that Pirh2 promotes ubiquitin-dependent degradation of p27Kip1 in the nucleus and cytoplasm at late G1 phase, whereas KPC and Pirh2 cooperatively degrade p27Kip1 exported from the nucleus to cytoplasm in early-mid G1 phase. On the other hand, Skp2 contributes degradation of nuclear p27Kip1 in S-G2 phase. In S phase, Skp2 expression is induced, and a low level of p27Kip1 is maintained in the nucleus by collaboration of Skp2 and Pirh2, in which expression level is also high in S phase. Therefore, our study indicated that Pirh2 has different characters from Skp2 and KPC, and it is thought that Pirh2, Skp2, and KPC ubiquitin ligases coordinately control the abundance of the whole cellular p27Kip1 protein level and the progression of the cell cycle.
In Fig. 3D, depletion of KPC1 causes a large increase in p27Kip1 protein levels in T98G cells, comparable with that seen when Pirh2 is depleted. Yet in Fig. 5B, the same cells treated with KPC1 siRNA can clearly progress through the cell cycle, whereas the Pirh2-depleted cells remain blocked. We think that these are the very data showing the difference in the role at G1-S transition from Pirh2 to KPC1. Note that the experiment shown in Fig. 3D was done in asynchronous cells, whereas the experiment in Fig. 5B was done in synchronization/release conditions. The experiments in asynchronous conditions indicate requirement of both Pirh2 and KPC1 in regulation of p27Kip1 stability in steady-state cells. On the other hand, the synchronization/release experiments give us the evidence for the cell cycle phase at which activity of Pirh2 or KPC1 is required. As discussed above, data shown in Fig. 5B suggest that Pirh2 rather than KPC1 negatively regulates p27Kip1 stability and positively regulates cell cycle progression at G1-S transition.
The original report concerning Pirh2 showed it to be also a p53 target gene, as well as being able to regulate p53 levels ( 25). Pirh2-mediated degradation of p27Kip1 may participate in facilitating execution of p53-induced apoptosis. p27Kip1 is cleaved by caspase-3–like activity in apoptotic cells ( 38), leading to a dramatic induction of cdk2 activity. Induction of apoptosis is accompanied by up-regulation of cyclin A–associated kinase activity, and inhibition of this activity suppresses apoptosis ( 39– 42). Considering that p53 enhances transcription of Pirh2 and given that disappearance of p27Kip1 is required for executing the apoptotic process, Pirh2 may participate in p53-induced apoptosis by destructing p27Kip1 to activate cyclin A/cyclin E–Cdk2 complexes.
The expression of p27Kip1 is reduced in many types of human cancers, and it is inversely related to tumor aggressiveness and directly associated with prognosis in individuals with such cancer ( 9– 13). Because Duan et al. reported high expression of Pirh2 in human lung cancer ( 43), Pirh2, as well as Skp2, may be involved in poor prognosis of human cancers to promote accelerated degradation of p27Kip1. Because depletion of endogenous Pirh2 by siRNA treatment resulted in a marked inhibition of cell cycle progression of glioblastoma cells ( Fig. 5), a Pirh2 inhibitor may be useful in cancer therapies for Pirh2-overexpressing tumors.
Grants support: Ministry of Education, Science, Sports, Culture, and Technology of Japan grants-in-aid 16021220 (M. Kitagawa), 17590056 (K. Kitagawa), and 17770110 (T. Hattori), Ministry of Health, Labor, and Welfare of Japan Third Term Comprehensive Control Research for Cancer grant-in-aid (M. Kitagawa), and Ministry of Education, Science, Sports, Culture, and Technology of Japan Hamamatsu University School of Medicine COE program (M. Kitagawa).
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank Drs. Kei-ichi Nakayama, Takumi Kamura, Noriko Ishida, and Mark Montminy for kindly providing the materials, Sayuri Suzuki and Tomomi Iida for technical assistance, and our laboratory members for their helpful discussions.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received May 31, 2007.
- Revision received August 28, 2007.
- Accepted September 21, 2007.
- ©2007 American Association for Cancer Research.