Involvement of Two Distinct Ubiquitin E3 Ligase Systems for p27 Degradation in Corneal Endothelial Cells

Abstract

purpose. p27^Kip1 (p27) is an important regulator of G₁ progression. For cells to proliferate, p27 must undergo proteolysis. FGF-2 enables phosphorylation of p27 at both the Thr-187 and Ser-10 sites, an event that is prerequisite for polyubiquitination. This study was undertaken to determine whether degradation of the two phosphorylated p27s is mediated by a distinct ubiquitin E3 ligase complex at different subcellular locations.

methods. Expression of p27, KPC1, KPC2, Skp1, Skp2, and Cul1 was analyzed by immunoblot analysis. Association of p27 with ubiquitin E3 ligase was determined with coimmunoprecipitation followed by immunoblot analysis. Inhibitors were used to inhibit proteasomal degradation and nuclear export of the phosphorylated p27. DNA synthesis was measured by BrdU incorporation into DNA.

results. Among ubiquitin ligase complex proteins, Cul1, KPC1, and KPC2 were constitutively expressed, whereas expression of Skp1 and Skp2 was temporally induced by FGF-2. Skp1, Skp2, and Cul1 were involved in polyubiquitination of phosphorylated p27 at Thr-187 (pp27Thr187) in nuclei. Maximum association of pp27Thr187 with the ubiquitin E3 ligase occurred 24 hours after FGF-2 stimulation. pp27Ser10 used the cytoplasmic ubiquitin E3 ligases KPC1 and KPC2, with maximum protein interaction observed at 8 hours. MG132 effectively blocked degradation of both pp27Thr187 and pp27Ser10, whereas leptomycin B blocked the nuclear export of pp27Ser10. Both inhibitors blocked BrdU incorporation into DNA.

conclusions. The findings demonstrate distinct polyubiquitination pathways for pp27Thr187 and pp27Ser10; the former is ubiquitinated through the nuclear ubiquitin E3 ligase system during late G₁ phase; the latter by cytosolic ubiquitin E3 ligase during early G₁ phase.

Cell-cycle regulation is a hallmark of many biological activities that encompass physiological processes, such as development, and pathophysiological processes, such as cancer and wound healing. The Cdk inhibitor p27^Kip1 (p27) is an important regulator of G₁ progression. It is highly expressed in the G₀ phase of the cell cycle. ¹ However, in the mitogenic pathway, p27 is degraded through the ubiquitin-proteasome machinery, leading cell cycle progression from the G₁ to the S phase. ^{2 3 4} For p27 to be degraded, phosphorylation of p27 is a prerequisite; there are at least four known phosphorylation sites: Thr-187, Ser-10, Thr-157, and Thr-198. The Cdk2-cyclin-E complex is responsible for phosphorylation of p27 at Thr-187, ^{5 6} whereas Akt has been shown to phosphorylate p27 directly on Ser-10, Thr-157, and Thr-198. ^{7 8 9} The Ser-10 site is also phosphorylated by human kinase-interacting stathmin (hKIS), ¹⁰ a nuclear serine-threonine kinase. 

Although phosphorylation as a major posttranslational modification of p27 has been studied extensively, the means by which p27 phosphorylation is coordinated at multiple sites was not fully understood. Our recent study ¹¹ demonstrated the differential kinetics of p27 phosphorylation at Ser-10 and Thr-187 and showed that phosphorylated p27 at Thr-187 (pp27Thr187) and at Ser-10 (pp27Ser10) represent two distinct populations of p27 in the G₁ phase of the cell cycle. We further showed that these two populations of p27 function at different stages of the G₁ phase of the cell cycle in response to mitogenic signals and that phosphorylation of p27 at Ser-10 is the major mechanism for the G₁/S transition in corneal endothelial cells (CECs) in response to FGF-2 stimulation. ¹¹ The differential kinetics involved in phosphorylation, ubiquitination, and degradation of p27 clearly demonstrate that phosphorylation of p27 at Ser-10 followed by ubiquitination and degradation is an early event as opposed to the late event observed with pp27Thr187. In CECs, phosphorylation of p27 at Ser-10 takes place within 1 hour of FGF-2 stimulation, followed immediately by nuclear export of the phosphorylated molecule to the cytosol, where pp27Ser10 is subsequently ubiquitinated and degraded at the proteasome. The entire event, from phosphorylation in the nuclei to degradation of pp27Ser10 in the cytoplasm, occurs between 1 and 16 hours after FGF-2 stimulation. In contrast, phosphorylation of p27 at Thr-187 begins 8 hours after FGF-2 stimulation and reaches its maximum level at 16 hours, during which time the pp27Thr187 is ubiquitinated. The ubiquitinated pp27Thr187 is then degraded at the proteasome during the next 4 to 8 hours in the nuclei. Such differential kinetics of phosphorylation and ubiquitination and the different subcellular localizations for degradation of the two pp27Thr187 and pp27Ser10 clearly demonstrate that the two phosphorylated p27s indeed represent two different populations of p27. 

The polyubiquitination and destruction of pp27Thr187 has been extensively studied: pp27Thr187 is recognized by the Skp1 (S-phase kinase-associated protein 1)-cullin-F-box (SCF) ubiquitin ligase complex SCF^Skp2 that binds the phosphorylated p27 through the F-box protein Skp2 ^{12 13 14} and its cofactor Cks1 (Cdc kinase subunit 1). ^{15 16} Kamura et al. ¹⁷ have reported the existence of a Skp2-independent pathway for the degradation of p27 at the G₁ phase of the cell cycle: KPC (Kip1 ubiquitination-promoting complex), consisting of KPC1 and -2, controls the degradation of p27 in the G₁ phase after the phosphorylated p27 is exported from the nucleus. Thus, the Skp2-independent degradation of p27 may represent the pp27Ser10 population, because of the cytoplasmic subcellular localization of the event. Kamura et al. proposed that p27 degradation is regulated by two distinct mechanisms: translocation-coupled cytoplasmic ubiquitination by KPC at the G₁ phase of the cell cycle and nuclear ubiquitination by Skp2 at the S and G₂ phases. These findings, taken together, indicate that translocation-coupled cytoplasmic ubiquitination is probably used by pp27Ser10, whereas the nuclear ubiquitination by Skp2 is used by pp27Thr187. 

Because it has not been determined whether the two mechanisms for the degradation of p27 are operated at a single cell level, the present study was undertaken to investigate whether degradation of pp27Thr187 and pp27Ser10 is mediated by different ubiquitin ligase complexes and whether ubiquitination and degradation of the two phosphorylated p27s occurs in different subcellular localizations in response to FGF-2 stimulation. This study demonstrates that the two phosphorylated p27s (pp27Thr187 and pp27Ser10) use different ubiquitin ligase complexes: pp27Thr187 uses the SCF^Skp2 pathway in the nucleus for ubiquitination and subsequent degradation at proteasome, whereas ppSer10 uses the KPC complex in the cytoplasm. 

Materials and Methods

Anti-p27, Skp1, Skp2, and Cul1 antibodies and anti-lamin B antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-pp27Ser10 and anti-pp27Thr187 were obtained from Zymed Laboratories, Inc. (South San Francisco, CA); leptomycin B (LMB) and anti-ubiquitin antibody from Biomol (Plymouth Meeting, PA); FGF-2 from Calbiochem (San Diego, CA); anti-α-tubulin and peroxidase-conjugated secondary antibodies from Sigma-Aldrich (St. Louis, MO); and MG132 from Chemicon (Temecula, CA). Anti-KPC1 and -2 antibodies were obtained through the courtesy of Keiichi I. Nakayama (Kyushu University, Japan). ¹⁷  

Cell Culture

Rabbit eyes were purchased from Pel Freez Biologicals (Rogers, AR). Isolation and establishment of rabbit CECs were performed as previously described. ¹⁸ Briefly, the corneal endothelium-Descemet’s membrane complex was treated with 0.2% collagenase and 0.05% hyaluronidase (Worthington Biochemical, Lakewood, NJ) for 90 minutes at 37°C with gentle shaking. Primary cultured cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen-Gibco, Grand Island, NY) supplemented with 15% fetal bovine serum (Omega Scientific, Tarzana, CA) and 50 μg/mL gentamicin (DMEM-15) in a 5% CO₂ incubator. For subculture, confluent primary cultures were treated with 0.05% trypsin and 5 mM EDTA in phosphate-buffered saline (PBS) for 5 minutes. First-passage CECs maintained in DMEM-15 were used for all experiments. Heparin (10 μg/mL) was added to cell cultures treated with FGF-2 (10 ng/mL), because our previous study showed that the initiation of FGF-2 activity in CECs requires the addition of supplemental heparin. ¹⁹ In some experiments, pharmacologic inhibitors were used in the presence of FGF-2 stimulation: MG132 (10 μM), ²⁰ or leptomycin B (10 ng/mL). ¹¹  

Cell Proliferation Assays

Cell proliferation was assayed by the bromodeoxyuridine (BrdU) cell proliferation assay kit (Roche Applied Sciences, Indianapolis, IN) as described previously. ¹¹ Briefly, the cells were seeded in six-well plates at a concentration of 1 × 10⁵ cells/well. After 1 day, the medium was changed from DMEM-15 to DMEM, and the cells were incubated for 30 hours. The serum starved cells were then subjected to their respective experimental conditions for 24 hours, with each culture containing BrdU (10 ng/mL). Cells were then fixed and permeabilized, and the DNA was denatured by treatment with fixative-denaturing solution. Detector peroxidase-conjugated anti-BrdU monoclonal antibody was then used, and the color reaction was developed by using the chromogenic substrate tetramethylbenzidine. The product was quantified with a spectrophotometric plate reader at dual wavelengths of 370 and 492 nm. 

Cytoplasmic and Nuclear Protein Extractions

CECs cultured in each culture condition on 100-mm tissue culture dishes were washed twice with sterile PBS, then incubated in an enzyme-free cell dissociation solution (Chemicon) for 5 minutes at room temperature. Cells were detached by scraping, transferred to microcentrifuge tubes, and pelleted at 5000g for 1 minute. Nuclear and cytoplasmic proteins were extracted with a nuclear extraction kit containing cytoplasmic lysis buffer and nuclear lysis buffer (Chemicon), according to the manufacturer’s instructions. Briefly, the harvested cells were resuspended in two cell pellet volumes of cold cytoplasmic lysis buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 μg/mL aprotinin, 1 μg/mL leupeptin, 1 μg/mL pepstatin, and 0.5 mM dithiothreitol and allowed to swell on ice for 15 minutes. Igepal CA-630 (Sigma-Aldrich) was then added to a final concentration of 0.1%, and the swollen cells were incubated an additional 5 minutes on ice and then homogenized with a 27-gauge needle. Nuclei were pelleted at 8000g for 20 minutes at 4°C. The supernatant containing the cytosolic portion was transferred to a fresh tube and stored at −80°C until further use. The remaining pellet containing the nuclear portion was washed by centrifugation with cold nuclear extraction buffer containing 1 mM PMSF, 1 μg/mL aprotinin, 1 μg/mL leupeptin, 1 μg/mL pepstatin, and 0.5 mM dithiothreitol and then resuspended in two thirds of the original cell pellet volume of cold nuclear extraction buffer. Nuclei were disrupted by drawing and ejecting 10 times with a syringe attached to a 27-gauge needle. Nuclear proteins were extracted at 4°C for 60 minutes with gentle agitation, using an orbital shaker. The nuclear protein extract was clarified by centrifugation at 16,000g for 5 minutes at 4°C and then dialyzed (Slide-A-Lyzer MINI Dialysis Unit, 7K MWCO; Pierce, Rockford, IL). Nuclear and cytoplasmic protein concentrations were determined using the Bradford reagent (Bio-Rad Laboratories, Inc., Hercules, CA), with bovine serum albumin as a standard. To verify the purity of the fractions, 15 μg of nuclear or cytoplasmic proteins were immunoblotted with lamin B and α-tubulin antibodies. These verified nuclear and cytoplasmic proteins were used for immunoprecipitation and immunoblot analysis. 

Immunoprecipitation, Gel Electrophoresis, and Western Blot Analysis

For immunoprecipitation, equal amounts of protein (750 μg) of the nuclear and cytoplasmic extracts in lysis buffer were incubated with protein G Sepharose beads (Invitrogen, Carlsbad, CA) for 1 hour. After the beads were discarded by centrifugation, the supernatant was incubated with rabbit polyclonal anti-pp27Thr187 or pp27Ser20 antibodies (2.0 μg IgG/mg protein) overnight at 4°C. An equal volume of protein G Sepharose beads was added, and the incubation continued for 3 hours, followed by centrifugation to obtain the immune complex. Precipitated proteins were washed in PBS containing protease inhibitors (1 μg/mL leupeptin, 1 μg/mL pepstatin, 1 μg/mL aprotinin, and 1 mM PMSF). After the immune complex was washed, a protein sample buffer was added to the complex, which was then boiled for 5 minutes for SDS-PAGE on preformed tricine polyacrylamide gels. The proteins separated by SDS-PAGE ²¹ were transferred to a 0.22-μm nitrocellulose membrane (Whatman Inc., Florham Park, NJ), and nonspecific binding sites of nitrocellulose membrane were blocked by 5% nonfat milk in PBS containing 0.1% Tween-20. The incubations were performed with primary antibodies (1:1000 dilution for ubiquitin, 1:2500 for p27, pp27Ser10, or pp27Thr187, and 1:3000 for KPC1 and -2) and peroxidase-conjugated secondary antibody (1:5000 dilution). Membranes were treated with enhanced chemiluminescence (ECL) reagent (GE Healthcare, Piscataway, NJ) and exposed to ECL film. The relative density of the polypeptide bands detected on ECL film was determined with commercial software (Gel-doc; Bio-Rad Laboratories, Inc.). 

Results

We first determined the protein expression pattern of p27, KPC1, KPC2, Skp1, Skp2, and Cul1 and their subcellular localization. After cells were stimulated with FGF-2 for 2 to 36 hours, nuclear fractions and cytoplasmic fractions were prepared as described earlier. ¹¹ The purity of each fraction was measured by using lamin B (nuclear protein) and α-tubulin (cytoplasmic protein). We initially used a wide range of treatment times with FGF-2 to determine the whole spectrum of expression of each protein, since the two phosphorylated p27s have very different kinetics: Phosphorylation is an early event for pp27Ser10 and a late event of pp27Thr187. When total p27 was examined, the protein level of nuclear p27 gradually decreased in response to FGF-2 stimulation. Samples obtained from the cells maintained without FGF-2 showed the maximum level of p27, whereas the nuclear p27 level in cells treated with FGF-2 for 36 hours reached only 18% of the maximum level (Fig. 1) . In contrast, the cytoplasmic p27 was increased after 2 hours of stimulation with FGF-2, reaching the maximum level after 8 hours of stimulation; thereafter, the protein level decreased dramatically, until the protein was barely detectable in the cytosol by 24 hours (Fig. 1) . The ubiquitin E3 ligase complex proteins involved in ubiquitination of pp27Thr187, such as Skp1, Skp2, and Cul1, were only observed in the nuclear fraction. Skp2 was induced by FGF-2 stimulation and reached its maximum expression within 24 hours, whereas Skp1 reached maximum level 12 hours after FGF-2 stimulation, much faster than Skp2. Cul1 expression was constitutive (Fig. 1) . When similar experiments were performed for the other ubiquitin E3 ligase system (KPC1 and -2), constitutive expression of cytoplasmic KPC1 and -2 was observed in response to FGF-2 stimulation, whereas the absence of both proteins in the nuclear fraction (Fig. 2)confirms the previous findings. ¹⁷ These findings clearly demonstrate that two distinct ubiquitin E3 ligase systems have differential subcellular localizations; thus, confirming the previous finding that KPC is a cytoplasmic protein, and Skp2 is a nuclear protein. 

Physical association between the phosphorylated substrate (pp27 in this study) and the ubiquitin E3 ligase complex is required for polyubiquitination of the substrate. Immunoprecipitation was performed with the specific antibody to either pp27Thr187 or pp27Ser10, to determine whether the two phosphorylated p27s employ distinct ubiquitin E3 ligase systems to be polyubiquitinated. Two different time courses for FGF-2 stimulation were chosen based on the findings shown in Figures 1 and 2and our previous study ¹¹ : The pp27Ser10 study was performed at 4, 8, 12, 16, and 20 hours after FGF-2 stimulation, whereas the pp27Thr187 experiment was performed at much later time points, at 8-hour intervals for up to 36 hours. Figure 3shows that the immune complex precipitated with anti-pp27Thr187 antibody contained proteins composing SCF^Skp2 (Skp2, Skp1, and Cul1). The association of pp27Thr187 with these ubiquitin E3 ligase proteins reached its maximum level 24 hours after FGF-2 stimulation (Fig. 3) . No such association was observed in cells that had been stimulated with FGF-2 for 36 hours. When the immune complex was blotted with anti-ubiquitin antibody, maximum polyubiquitination of pp27Thr187 was also observed 24 hours after FGF-2 stimulation, after which the extent of ubiquitination was greatly reduced (Fig. 3) . Of interest is that Cul1 involved in ubiquitin ligase complex formation shows similar kinetics with Skp1 and Skp2, even though Cul1 expression is constitutive. The immune-complex of the cytoplasmic fraction precipitated with anti-pp27Thr187 demonstrated the absence of Skp2, Skp1, and Cul1 and the absence of polyubiquitination of pp27Thr187 (Fig. 3) . These results were expected, since pp27Thr187 ¹¹ and the nuclear ubiquitin ligase E3 complex proteins (Skp1, Skp2, and Cul1) are not present in the cytoplasm (Fig. 1) . 

Translocation-coupled cytoplasmic ubiquitination of p27 is mediated, not by SCF^Skp2 but by KPC. The cytoplasmic fraction, therefore, was immunoprecipitated with anti-pp27Ser10 antibody. Figure 4demonstrates that the immune-complex precipitated with anti-pp27Ser10 antibody contained KPC1 and -2. The amount of KPC protein in the complex was at its maximum 8 hours after FGF-2 stimulation, after which the amount of KPC proteins in the immune-complex was gradually reduced. It is likely that the limited amount of pp27Ser10 is the rate-limiting step, as we reported that pp27Ser10 is greatly decreased within 16 hours after FGF-2 stimulation. ¹¹ When the immune-complex precipitated with anti-pp27Ser10 antibody was further analyzed for ubiquitination profiles, polyubiquitination of pp27Ser10 was observed; the maximum ubiquitination was also observed 8 hours after FGF-2 stimulation. The formation of ubiquitin E3 ligase complexes and polyubiquitination of pp27Ser10 was simultaneous. On the other hand, the nuclear fraction demonstrated that the immune-complex precipitated with anti-pp27Ser10 antibody did not contain either KPC1 or -2, the cytoplasmic ligase proteins. Thus, although pp27Ser10 in the nuclei was being exported to the cytoplasm, it could not be normally ubiquitinated by the nuclear ubiquitin ligase complex (i.e., Skp1, Skp2, and Cul1). 

Proteasomal degradation of the ubiquitinated pp27Thr187 and pp27Ser10 was further confirmed using MG132, which inhibits the chymotrypsin-like activity of 26S proteasome and cysteine protease (Fig. 5) . The steady state level of total p27 was first determined in CECs simultaneously treated with FGF-2 and MG-132. Both nuclear and cytoplasmic p27 were maintained at very high levels, as opposed to the protein levels observed in the absence of MG132 (Fig. 1) . In the presence of MG132, the nuclear p27 level 36 hours after FGF-2 stimulation was 55% of the maximum p27 level observed in the control cultures, and the amount of cytoplasmic p27 was very high, even 20 hours after FGF-2 stimulation. When the steady state level of the two phosphorylated p27s was further analyzed, nuclear pp27Thr187 was maintained at high levels until 36 hours after FGF-2 stimulation, and the cytoplasmic pp27Ser10 was also maintained at very high levels for up to 20 hours. The inhibitor did not alter the subcellular localization of the two phosphorylated p27s: pp27Thr187 was present only in the nucleus, whereas pp27Ser10 was present in the cytoplasm. 

Although phosphorylation of p27 at both the Thr-187 and Ser-10 sites took place in the nuclei, the phosphorylated p27 at Ser-10 has to be exported out of the nuclei. LMB was used to block the CRM-1-mediated nuclear export of pp27Ser10. ^{22 23} When cytoplasmic and nuclear fractions were subjected to immunoblot with anti-pp27Ser10 antibody, pp27Ser10 was markedly retained in the nuclei for up to 36 hours after FGF-2 stimulation (Fig. 6) , whereas the level of cytoplasmic pp27Ser10 was negligible. This finding clearly demonstrates that nuclear export of pp27Ser10 was greatly hampered by blocking CRM-1-mediated nuclear export. However, LMB did not alter the phosphorylation profile of the nuclear pp27Thr187. The maximum phosphorylation of p27 at Thr-187 was observed 16 hours after FGF-2 stimulation, similar to that observed in the absence of the inhibitor, ¹¹ suggesting that the inhibitor did not influence pp27Thr187. Of interest is that total p27 level in the nuclei showed a biphasic phenomenon: A very high p27 level maintained in LMB-treated cells for up to 16 hours and the gradual decrease in nuclear p27 thereafter. The high p27 levels observed 16 hours after treatment were probably achieved by the combined accumulation of pp27Ser10 (LMB-dependent) and pp27Thr187. Since LMB acts on the proteins destined for nuclear export, the inhibitor did not modulate the level of nuclear pp27Thr187. The reduced level of nuclear p27 in cells treated with LMB for 32 and 36 hours was mostly achieved by proteolysis of pp27Thr187, which was degraded by an LMB-independent pathway. On the other hand, the total cytosolic protein of p27 was barely detectable throughout the time period tested. 

Finally, we investigated whether the retarded degradation of the two phosphorylated p27s mediated by MG132 or LMB impedes cell cycle progression. CECs were treated with FGF-2 for 8, 16, or 24 hours in the presence or absence of the inhibitors. Incorporation of BrdU into DNA was used to determine the amount of DNA synthesis during the S-phase of the cell cycle. The BrdU incorporation into DNA was gradually elevated in a time-dependent manner (Fig. 7) . When proteasomal degradation of the phosphorylated p27s was hampered, DNA synthesis in the presence of MG132 was reduced in the range of 36% (8-hour period) to 51% (24-hour period), whereas LMB demonstrated less inhibition of DNA synthesis. The CECs simultaneously treated with FGF-2 and LMB for either 8 or 16 hours showed an 18% to 22% reduction, whereas cells treated for 24 hours had a 36% reduction in BrdU incorporation. These data confirmed that retarded degradation of both pp27Thr187 and pp27Ser10 blocked cell cycle progression. The inhibitory level observed with MG-132 may be caused by the inhibited degradation of both pp27Thr187 and pp27Ser10, whereas the inhibitory level achieved by LMB is solely caused by phosphorylation of p27 at the Ser-10 site. Therefore, the difference between MG-132- and LMB-mediated inhibition of the FGF-2-stimulated cell cycle progression at 24 hours may have been caused by the phosphorylation of p27 at Thr-187. It is critical to compare the values 24 hours after FGF-2 stimulation, because phosphorylation and polyubiquitination of p27 at Thr-187 reach maximum levels 24 hours after FGF-2 stimulation. These findings further confirmed our previous data ¹¹ that pp27Ser10 constitutes the major population of p27s during the G₁/S transition in the cell cycle. 

Discussion

Progression through the cell cycle is coordinated by Cdks and their activating cyclins and inhibitory CKI proteins. The CKI protein p27 negatively regulates the transition from the G₁ to S phases of the cell cycle by inhibiting the Cdk2-cyclin-E complex. Mitogenic signals cause a rapid decrease in p27 by inducing the polyubiquitination and proteasome-mediated degradation of the molecule. Phosphorylation of p27 is a known prerequisite for polyubiquitination and proteolysis of the molecule. ^{1 2 3 4 5} There are at least four known phosphorylation sites in the p27 molecules. Multiple studies have documented how p27 is phosphorylated, ubiquitinated, and degraded at the Thr-187 site ^{12 13 14 15 16} ; the Thr-187 site of p27 is phosphorylated by the Cdk2-cyclin E complex, and the phosphorylated p27 is then recognized by an SCF ubiquitin ligase complex that contains Skp2 as the specific substrate-binding F box protein. Skp2 is unique among known mammalian F box proteins, in that its levels fluctuate in the cell cycle, and such fluctuation also occurs in CECs in response to FGF-2 stimulation, as observed in this study (Fig. 1) . The polyubiquitinated pp27Thr187 is degraded by proteasome machinery in the nucleus. ^{11 12 13 14 15 16} Unlike this well-defined pathway, the mechanism by which the pp27Ser10 population is ubiquitinated, leading to proteasomal degradation in the cytoplasm, is not fully understood although earlier studies have demonstrated that Ser-10 is the major phosphorylation site of p27 ²⁴ and that pp27Ser10 is the major phosphorylated population of p27 involved in the G₁/S progression of the cell cycle. ¹¹ Phosphorylation of Ser-10 is required for the binding of p27 to CRM1, a carrier protein for nuclear export. ^{22 23} Kamura et al. ¹⁷ and colleagues reported that cytoplasmic ubiquitin ligase KPC regulates proteolysis of p27 in a Skp2-independent pathway at the early G₁ phase of the cell cycle. We recently reported that phosphorylation of p27 at Ser-10 facilitates export of the molecule from the nuclei to the cytoplasm and that ubiquitination and subsequent degradation of pp27Ser10 takes place during the early G₁ phase of the cell cycle. ¹¹ Taken together, these findings indicate that the KPC pathway, a Skp2-independent pathway, is likely used for polyubiquitination of pp27Ser10, while pp27Thr187 is ubiquitinated through the conventional SCF^Skp2 ligase pathway. Therefore, we decided to confirm that the KPC pathway is used to ubiquitinate pp27Ser10, whereas the SCF^Skp2 pathway is used by pp27Thr187. We further determined whether the two distinct ubiquitin E3 ligase systems operate at different subcellular localizations at a single cell level using differential kinetics. 

In this study, we demonstrated that the two phosphorylated p27s employ the distinct ubiquitin E3 ligase complexes. pp27Thr187 uses the SCF^Skp2 ligase complex present in the nuclei, thus confirming the previously published data, ^{12 13 14} whereas pp27Ser10 uses the cytoplasmic KPC complex, as reported earlier. ¹⁷ We also showed that KPC1 and -2 are constitutively expressed only in the cytoplasm, whereas the expression of the nuclear Skp1 and Skp2 is induced by FGF-2 stimulation, and neither protein is present in the cytoplasm. Events of polyubiquitination of pp27Thr187 progress sequentially after FGF-2 stimulation; first, phosphorylation of p27 at Thr-187 takes place, followed by expression of Skp2, complex formation of the phosphorylated p27 with SCF^Skp2, and finally polyubiquitination of the substrate (pp27Thr187) by the ubiquitin E3 ligase. Except for phosphorylation of p27 at Thr-187, which reaches its maximum level after 16 hours, the subsequent events beginning with Skp2 expression, ubiquitin E3 ligase complex formation, and polyubiquitination of the substrate reach their maximum levels at 24 hours. Thus, it is likely that Skp2 expression is the rate-limiting step for polyubiquitination of pp27Thr187. Unlike the late event of complex formation of pp27Thr187 with SCF^Skp2, it took less than 4 hours for pp27Ser10 to be associated with KPC 1 and -2 after FGF-2 stimulation, and only 8 hours to reach its maximum association. By 20 hours, the association among the proteins was greatly decreased. The present study also demonstrated that the two distinct ubiquitin E3 ligase systems were present in the different subcellular localizations. Thus, the distinct kinetics of ubiquitin ligase complex formation in differential subcellular location clearly confirm that the two phosphorylated p27s are different populations of p27 at a single-cell level. 

p27 plays an important role as the G₁ phase inhibitor in CECs. ^{25 26} Recent work has shown that human CECs remain arrested in the G₁ phase of the cell cycle throughout their lifespan. ^{27 28} Such characteristic behavior of cell proliferation dictates most of the wound-healing processes occurring in the corneal endothelium: CECs do not replace lost cells by cell division, instead using migration and attenuation to cover the denuded area. In contrast, in a small fraction of wound-healing processes, CECs are transformed into mesenchymal cells that subsequently produce fibrillar extracellular matrix (ECM) in the basement membrane environment. Thus, corneal fibrosis represents a significant pathophysiological problem—one that causes blindness by physically blocking light transmittance. One clinical example of corneal fibrosis observed in corneal endothelium is the development of a retrocorneal fibrous membrane (RCFM) in Descemet’s membrane. ^{29 30} In RCFM, CECs are converted to fibroblast-like cells: the contact-inhibited monolayers of CECs are lost, resulting in the development of multilayers of fibroblast-like cells. ^{30 31} These morphologically altered cells simultaneously resume their proliferation ability and deposit a fibrillar ECM in Descemet’s membrane. Our in vitro model elucidated the molecular mechanism of RCFM formation and demonstrated that FGF-2 directly mediates the endothelial mesenchymal transformation (EMT) observed in CECs. ^{6 31 32} We reported that, among the phenotypes altered during EMT, FGF-2 directly regulates cell cycle progression through the action of phosphatidylinositol 3-kinase. ^{6 33}  

In response to FGF-2 stimulation, CECs use two differential proteasomal pathways largely to remove p27. Thus, during the pathologic wound repair process observed in the RCFM, FGF-2 significantly facilitates phosphorylation of p27, producing major pp27Ser10 and minor pp27Thr187 populations and leading to a complete removal of the potent G₁ inhibitor of the cell cycle. As a consequence, the tissue undergoing wound healing produces a sufficient number of cells to participate in the repair process during the pathophysiologic conditions. Whether this differential regulatory mechanism for p27 degradation is applied to other cell systems remains to be determined. 

 Supported by National Eye Institute Grants EY06431 and EY03040 and Research to Prevent Blindness.

 Submitted for publication July 9, 2007; revised September 26, 2007; accepted November 26, 2007.

 Disclosure: J.G. Lee, None; E.P. Kay, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Effect of FGF-2 on expression and subcellular localization of p27, Skp1, Skp2, and Cul1. When cells reached approximately 70% confluence, they were starved of serum for 30 hours. The serum-starved cells were treated with FGF-2 for the designated time and then maintained in DMEM for up to 36 hours. At the end of treatment, the cells were lysed with cytoplasmic lysis buffer and the nuclei harvested by centrifugation. The supernatant was saved as the cytosolic fraction, and the pellets were treated with nuclear extraction buffer to rupture the nuclei. Each fraction was immunoblotted with the respective antibody. The relative density of immunoblot bands was determined with a gel documentation system. Data were normalized to lamin B for nuclear fractions and α-tubulin for cytoplasmic fractions (a loading control). The purity of fractions was also controlled with these two proteins. The relative differences were then compared with the values of unstimulated CECs (0 hours). The results represent data obtained in three independent experiments. N, nuclear fraction; C, cytoplasmic fraction.

Figure 2.

 

Effect of FGF-2 on expression and subcellular localization of KPC1 and -2. Subcellular fractions obtained from Figure 1were also analyzed by immunoblot analysis with anti-KPC1 and -KPC2 antibodies. Relative density of immunoblot bands was determined using a gel documentation system. Data were normalized to lamin B for nuclear fractions and α-tubulin for cytoplasmic fractions (a loading control). The purity of fractions was also controlled with these proteins. The relative differences were then compared with the values of unstimulated CECs (0 hours). Results of one of three similar experiments are presented. N, nuclear fraction; C, cytoplasmic fraction.

Figure 3.

 

Ubiquitination and physical association of pp27Thr187 with the nuclear ubiquitin E3 ligase in response to FGF-2 stimulation. Cytosolic and nuclear fractions were prepared as described in Figure 1 . Subcellular fractions were respectively immunoprecipitated with the anti-pp27Thr187 antibody. The immunoprecipitated complexes were then immunoblotted with the designated antibodies. Total proteins of nuclear and cytoplasmic fractions were loaded for detection of lamin B and α-tubulin, which were used to control the purity of fractions. Data were analyzed as described in Figure 1and represent the results of four independent experiments.

Figure 4.

 

The ubiquitination and physical association of pp27Ser10 with the cytoplasmic ubiquitin E3 ligase in response to FGF-2 stimulation. Cytosolic and nuclear fractions were prepared as described in Figure 1 . Subcellular fractions were respectively immunoprecipitated with the anti-pp27Ser10 antibody. The immunoprecipitated complexes were then immunoblotted with the designated antibodies. Total proteins of nuclear and cytoplasmic fractions were loaded for detection of lamin B and α-tubulin, which were used to control the purity of fractions. Data were normalized as described in Figure 1and represent results obtained in four independent experiments.

Figure 5.

 

Inhibitory effects of MG132 on the degradation of p27 mediated by FGF-2. When cells reached approximately 70% confluence, they were starved of serum for 30 hours. The serum-starved cells were treated with FGF-2 and MG132 for the designated times and maintained in DMEM with MG132 for up to 36 hours. At the end of the treatment, the cytosolic and nuclear fractions were prepared as described in Figure 1 . Each fraction was immunoblotted with the respective antibody. The relative density of the immunoblot bands was determined with a gel documentation system. Lamin B and α-tubulin were used to control the purity of fractions. Data were normalized as described in Figure 1and represent results obtained in three independent experiments.

Figure 6.

 

Effect of LMB on the degradation of p27 induced by FGF-2 stimulation. When cells reached approximately 70% confluence, they were starved of serum for 30 hours and then treated with FGF-2 and LMB for the designated time and maintained in DMEM with LMB for up to 36 hours. At the end of the treatment, the cytosolic and nuclear fractions were prepared as described in Figure 1 . Each fraction was immunoblotted with the respective antibody. The relative density of the immunoblot bands was determined with a gel documentation system. Lamin B and α-tubulin were used to control the purity of fractions. Data were normalized as described in Figure 1and represent the results obtained in three independent experiments.

Figure 7.

 

Effect of MG132 and LMB on cell proliferation stimulated by FGF-2. A BrdU-incorporation cell proliferation assay was performed. The serum-starved cells were stimulated with FGF-2 for the indicated times, with or without MG132 or LMB. BrdU was added for the last 16 hours. At the end of incubation, the measurement of BrdU incorporation into newly synthesized DNA of proliferating cells was performed by immunostaining against BrdU. The BrdU-incorporating nuclei were visualized with the chromogenic substrate tetramethylbenzidine, and the absorbance was measured in a spectrophotometric plate reader with 370 and 492 nm. *P < 0.01 compared with cells stimulated by FGF-2 for 16 hours; **P < 0.01 compared with cells stimulated by FGF-2 for 24 hours; ***P < 0.01 compared with cells treated with FGF-2 and LMB for 24 hours. The data shown were representative of three independent experiments.