Role of p27^Kip1 in cAMP- and TGF-β2–Mediated Antiproliferation in Rabbit Corneal Endothelial Cells

Abstract

purpose. To determine whether p27^Kip1 plays a role in antiproliferation mediated by antimitogens (cAMP and TGF-β2) in rabbit corneal endothelial cells (CECs).

methods. Cell proliferation was assayed using a colorimetric method to determine the number of viable cells. Subcellular localization of cell cycle–regulatory proteins was determined by immunofluorescent staining, and expression of the proteins was analyzed by immunoblot analysis.

results. When cells were treated with cAMP or TGF-β2, serum-mediated cell proliferation was inhibited in a dose-dependent manner. Simultaneous treatment of the two antimitogens did not show a synergistic effect on inhibition of cell growth. Expression of cell cycle–regulatory proteins, such as cyclin-D1, cyclin-E, cdk2, cdk4, p21^Cip1, and p27^Kip1 was determined using immunofluorescent staining. A strong nuclear staining was observed for p27^Kip1. The other proteins were not stained or were only very faintly stained. Treatment of cells with either cAMP or TGF-β2 did not change the staining potential of any proteins other than p27^Kip1, but all cells were positive for nuclear p27^Kip1 when treated with either TGF-β2 or cAMP. In contrast, mitogen (FGF-2)-containing medium decreased the number of p27^Kip1-positive cells. When the expression level of p27^Kip1 was determined using immunoblot analysis in the cells treated either with FGF-2 alone or with a concomitant treatment with FGF-2 and cAMP for 24 hours, FGF-2 markedly decreased the p27^Kip1 level, and cAMP prevented the decrease in p27^Kip1 level induced by FGF-2. No such phenomenon occurred during a short-term exposure of cells to either FGF-2 or cAMP or to a combination of the two. When cells were stained for phosphorylated p27^Kip1, FGF-2 markedly enhanced the staining of phosphorylated p27^Kip1 in nuclei, whereas both cAMP and TGF-β2 prevented the phosphorylation of p27^Kip1.

conclusions. These findings suggest that both antimitogens upregulate the expression of p27^Kip1 as they prevent the decrease of the p27^Kip1 level mediated by mitogen. Furthermore, cAMP and TGF-β2 may inhibit the G₁-to-S transition by blocking phosphorylation of p27^Kip1, which is a prerequisite for nuclear export of the inhibitor molecule for degradation.

The corneal endothelium, a monolayer of differentiated cells located in the posterior portion of the cornea, is essential for maintaining corneal transparency. Maintenance of corneal transparency requires an intact endothelial layer. If too many cells are lost, as may happen not only with aging but through disease or injury, a decline in corneal transparency ensues. ^{1 2 3 4} It has long been believed that the capacity for regeneration of corneal endothelium after injury is severely limited in humans. Thus, corneal endothelium is considered a nonreplicating tissue. ^{5 6} However, recent studies by Joyce et al., ⁷ Senoo and Joyce, ⁸ and Senoo et al. ⁹ show that corneal endothelial cells (CECs) in vivo are arrested in the G₁ phase of the cell cycle, suggesting that these cells possess proliferative potential. When G₁ phase arrest is overcome in cultured human CECs by transfection with viral oncoproteins, the cells resume their proliferative capacity for several generations before the onset of replicative senescence. ^{10 11 12} In some cases, epidermal growth factor or fibroblast growth factor (FGF)-2 markedly stimulates the proliferative potentials of CECs in culture. ^{13 14} Furthermore, FGF-2 has been proposed to be the direct mediator of endothelial mesenchymal transformation observed in ectopic fibrosis present in the corneal endothelium–Descemet membrane complex. ^{14 15} Nonetheless, mitosis of CECs is seldom observed in adult human eyes, even during the wound repair process. The underlying mechanisms that keep the endothelial cells from moving out of the G₁ phase are only partially understood. Transforming growth factor (TGF)-β2, present in the aqueous humor of the anterior chamber, has been proposed to suppress mitotic activity of the cells. ¹⁶  

The proliferation of all cells and their progression through the cell cycle are regulated by the sequential activity of various cyclin-dependent kinases (cdks). ^{17 18 19 20} The enzyme activity of cdks is dependent on physical interactions with one of the cyclin proteins, which are the regulatory subunits of these complexes. In addition, cdk activity can be negatively regulated by a group of proteins collectively termed cdk inhibitors (CKIs). CKI levels, similar to cyclin levels, vary during the cell cycle, thus contributing to the timing of cyclin–cdk activation. One family of CKIs includes p21^Cip1, p27^Kip1 (hereafter abbreviated as p21 and p27), and p57^Kip1. The N termini of these CKIs share homology and can bind to and inhibit cdks. ^{20 21 22} Overexpression of these inhibitors can attenuate the proliferative response, whereas a reduction in their expression increases proliferation. 

The CKI p27 was initially found to be induced by an extracellular antimitogenic signal. ²³ It accumulates in many situations in which cells are arrested in the G₀/G₁ phase. Its expression is elevated in contact-inhibited or mitogen-deprived cells, and it can negatively regulate G₁ phase progression in response to antimitogenic signals. ^{23 24 25 26} For example, TGF-β exerts antimitogenic effects through p27 that can inhibit both cyclin-D-cdk4 and cyclin-E-cdk2. In proliferating cells, p27 is expressed at a threshold level, much of it bound in a complex with cyclin-D-cdk4. In TGF-β–treated cells, cdk4 synthesis is inhibited, and p27 is mobilized into complexes with cyclin-E-cdk2, resulting in the loss of activity of both kinases and concomitant G₁ arrest. ^{19 24 26} TGF-β2 is the major TGF-β isoform in aqueous humor. ^{27 28} It has been proposed that the growth factor in aqueous humor plays a key role in maintaining CECs in a G₁ phase–arrested state in vivo. ¹⁶ Another antimitogen, cAMP, has long been recognized to inhibit the growth of certain cells. ^{29 30 31 32 33} A large increase in cAMP concentration inside the cell is generally growth inhibitory, because it induces an elevation of p27 levels in most cell lines of mesenchymal origin. ^{31 32} We therefore examined whether cAMP and TGF-β2 induce the accumulation of p27, preventing cdk activation and ultimately G₁ progression. The concentration of p27 is thought to be regulated predominantly by a posttranslational mechanism, ^{34 35} by which p27 is degraded by both the ubiquitin-proteasome pathway and ubiquitin-independent proteolytic cleavage. ³⁶ Regulation of ubiquitin-mediated proteolysis is often achieved by phosphorylation of the target proteins; thus, phosphorylation of Thr¹⁸⁷ of p27 leads to ubiquitination and degradation of p27. ³⁷ We, therefore, explored whether cAMP and TGF-β2 influence the phosphorylation of p27 in comparison to the effect of FGF-2, a potent mitogen of rabbit CECs. ^{14 15}  

Materials and Methods

Cell Cultures

Isolation and establishment of rabbit CECs were performed as previously described. ³⁸ Briefly, the Descemet membrane–corneal endothelium complex was treated with 0.2% collagenase and 0.05% hyaluronidase (Worthington Biochemical, Lakewood, NJ) for 60 minutes at 37°C. Cultured cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco-BRL, Grand Island, NY) supplemented with 10% fetal bovine serum (Irvine Scientific, Santa Ana, CA) and 50 μg/ml of gentamicin (DMEM-10) in a 5% CO₂ incubator. This method has been shown to promote cell proliferation during the early phase of culture and to maintain the culture as a contact-inhibited monolayer when the cells reach confluence. First-passage CECs were used for all experiments. For subculture, confluent cultures were treated with 0.2% trypsin and 5 mM EDTA for 3 to 5 minutes. TGF-β2 (R&D Systems, Minneapolis, MN), or the membrane-permeable cAMP analogue, 8-bromo-cAMP (Sigma, St. Louis, MO), or a combination of the two was used to impair cell proliferation mediated by serum. When cells were treated with FGF-2 (Intergen, Purchase, NY), heparin (10 μg/ml) was added to the cultures, because our previous study showed that CECs require supplemental heparin for FGF-2 activity to occur. ^{14 15} It is important to note that rabbit CECs maintain the following characteristics in culture: The primary cultures grow readily, permitting examination of the division cycle; they lose their proliferative potential as they are serially passaged, and thus, the second passage cells no longer divide. As the cell number decreases, the cultures become attenuated. They maintain type IV collagen expression (the physiologic collagen phenotype of Descemet membrane) and deposit Descemet membrane–like extracellular matrix. ³⁹ This behavior of rabbit CECs markedly differs from that of bovine CECs in culture, ⁴⁰ which assume proliferation beyond the life span of rabbit CECs or human CECs in culture. 

Cell Proliferation Assay

Rabbit CECs (4 × 10³/well) were plated in 96-well tissue culture plates. When cells reached approximately 60% confluence, the medium was removed and replaced with serum-free medium for 24 hours. Cells were brought back to DMEM-10 and simultaneously treated with TGF-β2, 8-bromo cAMP, or a combination of the two for 36 hours. At the end of the incubation period, 15 μl of reagent solution (Cell Titer 96^RAqueous One; Promega, Madison, WI) was added to the wells. The plates were incubated for 1 hour at 37°C in a humidified 5% CO₂ atmosphere, after which the absorbency was read at 490 nm, using a 96-well plate reader (MR700 Microplate Reader; Dynatech Laboratories, Chantilly, VA). 

Protein Preparation and Determination of Protein Concentration

Cells were washed with ice cold phosphate-buffered saline (PBS) and then lysed with lysis buffer (20 mM HEPES [pH 7.2], 10% glycerol, 10 mM Na₃VO₄, 50 mM NaF, 1 mM phenylmethylsulfonyl fluoride, 0.1 mM dithiothreitol [DTT], 1 μg/ml leupeptin, 1 μg/ml pepstatin, and 1% Triton X-100) on ice for 30 minutes with occasional rocking. The lysate was subjected to sonification, and the cell homogenates were then centrifuged at 14000g for 10 minutes. The concentration of the resultant supernatant was assessed by Bradford assay, using bovine serum albumin (BSA) as a standard, as previously described. ⁴¹  

SDS-Polyacrylamide Gel Electrophoresis

The conditions of electrophoresis were as described by Laemmli, ⁴² using the discontinuous Tris-glycine buffer systems. Twenty-five micrograms protein was loaded on a 12.5% SDS-polyacrylamide gel and separated under the reduced condition. After gels were exposed to enhanced chemiluminescence film (ECL; Amersham Life Science, Buckinghamshire, UK), the relative density of the bands was estimated using a one-dimensional image analyzer (LKB Ultrascan XI; Pharmacia LKB Biotechnology, Pleasant Hill, CA). 

Immunoblot Analysis

The proteins separated by SDS-PAGE were transferred to a polyvinylidene difluoride (PVDF) membrane at 0.22 A for 16 hours in a semidry transfer system (transfer buffer: 48 mM Tris-HCl [pH 8.3], 39 mM glycine, 0.037% SDS, and 20% methanol). Immunoblot analysis was performed as described previously, using an avidin-biotin complex staining kit (ABC Vectastain; Vector Laboratories Inc, Burlingame, CA). ^{14 15} All washes and incubations were performed at room temperature in TTBS (0.9% NaCl, 100 mM Tris-HCl [pH 7.5], 0.1% Tween 20). Briefly, PVDF membrane was immediately placed in the blocking buffer (5% nonfat milk in TTBS) and kept for 2 hours. The incubations were performed with primary antibodies (1:5000 dilution) for 2 hours, with biotinylated secondary antibodies (1:5000 dilution) for 1 hour, and with ABC reagent for 30 minutes. The membrane was treated with ECL reagent for 1 minute, and the ECL-treated membrane was exposed to ECL film. 

Immunofluorescent Staining

Rabbit CECs (4 × 10⁴/chamber) were plated on four-well chamber slides. We chose to stain cultures that were approximately 80% to 90% confluent for cell cycle–regulatory proteins so that the culture would closely mimic the in vivo situation. Cells were washed with PBS and fixed with 4% paraformaldehyde in PBS at room temperature for 15 minutes for the staining of p21, p27, and phosphorylated p27. Cells stained for cyclins and cdks were fixed with methanol at −20°C for 10 minutes After fixation and extensive washing with PBS, cells were permeabilized and blocked with buffer A (0.1% Triton X-100, 1% BSA in PBS) for 15 minutes at room temperature. The subsequent incubation was performed with buffer A, and all washes were performed in PBS at room temperature. Cells were incubated with the primary antibodies (1:200 dilution) at 37°C for 1 hour and then incubated with fluorescein isothiocyanate (FITC)–conjugated secondary antibodies (1:200 dilution) in the dark at room temperature for 45 minutes. After extensive washing, the slides were mounted in a drop of mounting medium (Vectashield; Vector Laboratories Inc.) to reduce photobleaching. Antibody labeling was examined using a laser scanning confocal microscope (LSM-510; Carl Zeiss, Thornwood, NY). For fluorescein examination, an argon laser at 488 nm was used in combination with a 505- to 530-nm emission filter for detection. A ×25 oil immersion objective (numeric aperture, 1.3; Plan-Neofluar; Zeiss) was used to acquire the images. Image analysis was performed using the standard system operating software provided with the microscope. 

Antibodies

Mouse monoclonal antibodies against cyclin-D1, cyclin-E, and cdk2 and rabbit polyclonal antibody against p21 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse monoclonal antibodies against p27, cdk4, and β-actin were purchased from Sigma, and rabbit polyclonal anti-phosphorylated p27 antibody was purchased from Zymed Laboratories, Inc. (San Francisco, CA). FITC-conjugated goat anti-mouse IgG and -rabbit IgG antibodies were purchased from Jackson ImmunoResearch Laboratories (West Grove, PA), and biotinylated goat anti-mouse IgG antibodies were purchased from Vector Laboratories, Inc. 

Results

Effect of 8-Bromo cAMP and TGF-β2 on Cell Proliferation in Rabbit CECs

The inhibitory effect of the antimitogens on serum-mediated cell proliferation in CECs was investigated. 8-Bromo cAMP inhibited cell proliferation in a dose-dependent manner (Fig. 1A) , albeit to a low degree. At 0.1 mM, 8-bromo cAMP showed no inhibitory effect, and cells treated with 0.3 mM 8-bromo cAMP showed a very slight decrease in cell proliferation. 8-Bromo cAMP at 1.0 mM inhibited cell proliferation by approximately 16%, and 8-bromo cAMP at 3.0 mM showed a maximal inhibition of serum-mediated cell proliferation by 22%. This finding differs from the effect of cAMP on cell proliferation on Schwann cells, in which a higher concentration of 8-bromo cAMP (1.0 mM) decreases cell proliferation, whereas a lower concentration (0.1 mM) stimulates cell proliferation. ⁴³ The inhibitory activity of TGF-β2 on serum-mediated cell proliferation is similar to that of 8-bromo cAMP. TGF-β2 at 1.0 ng/ml showed almost 16% inhibition on cell proliferation and TGF-β2 at 10 ng/ml demonstrated a maximal inhibition of approximately 20% (Fig. 1B) . Such low levels of inhibitory activity of TGF-β2 were consistent with our previous result. ⁴¹ Of interest, when the two antimitogens were simultaneously added to the culture, no synergistic effect was observed (Fig. 1C) . TGF-β2 at two concentrations, 1.0 ng/ml and 10 ng/ml, was used in combination with 8-bromo cAMP concentrations ranging from 0.1 mM to 3.0 mM. When the concentration of the two antimitogens that respectively showed a maximal inhibition of serum-mediated cell proliferation was used, inhibition of cell proliferation of CECs by the concomitant treatment never exceeded 20%. 

Expression of Cell Cycle–Regulatory Proteins in Rabbit CECs

Expression of cell cycle proteins, including cyclin-D1, cyclin-E, cdk2, cdk4, p21, and p27, was first determined using immunofluorescent staining (Fig. 2) . Cyclin-D1, cyclin-E, and cdk4 showed a very faint diffuse cytoplasmic staining, whereas the staining of cdk2 and p21 was undetectable. In contrast, p27 showed strong positive staining in the nuclei in almost half the population of cells maintained in DMEM-10. Control experiments performed in parallel in the absence of each of the primary antibodies showed negative staining profiles (data not shown). When cells were treated with 8-bromo cAMP, no detectable staining difference was observed in any tested proteins other than p27 (Fig. 3) . Treatment of cells with 8-bromo cAMP induced p27 expression in the nuclei of all cells. Treatment with TGF-β2 had a similar effect on p27 expression, with all cells staining strongly for nuclear p27 (Fig. 4) . The staining of cdk2 and cdk4 appears to be slightly enhanced at the perinuclear location in some of the cells treated with TGF-β2, and the staining profiles of cyclin-D1 and -E and p21 were not altered by TGF-β2 treatment. It should be noted that different fixation methods result in different staining potentials when cell cycle–regulatory proteins are tested using indirect immunofluorescent staining. Cyclins and cdks stained much better with methanol fixation than with paraformaldehyde fixation, whereas the reverse preference was observed in the staining potential of p21 and p27. Such different fixation methods were therefore adapted throughout the experiments. 

Effect of 8-Bromo cAMP and TGF-β2 on p27 Expression in Rabbit CECs

The expression of p27 was further determined in various growth conditions (Fig. 5) . When cells reached approximately 80% to 90% confluence, they were deprived of serum for 24 hours, and then one of the following treatments was applied: DMEM-10, FGF-2 in serum-free medium, 8-bromo cAMP in DMEM-10, TGF-β2 in DMEM-10, or a combination of 8-bromo cAMP and TGF-β2 in DMEM-10. Growth medium (DMEM-10) showed that approximately half the population of the cells stained for nuclear p27, similar to those shown in Figure 2 . Both 8-bromo cAMP and TGF-β2 stimulated the expression of p27 similar to that shown in Figures 3 and 4 . All cells were strongly stained for p27 in the nuclei. Simultaneous treatment of cells with 8-bromo cAMP and TGF-β2 showed a strong nuclear staining of p27 in all cells. When cells were treated with FGF-2, more cells stained for nuclear p27 than with DMEM-10 treatment. The increased p27 expression induced by FGF-2 was probably caused by contact inhibition mediated by the growth-stimulatory activity of FGF-2. The control, stained in the absence of anti-p27 antibody, showed negative staining for the protein. Because the cultures used for these studies were 80% to 90% confluent and thus were near the contact-inhibited confluence, we examined p27 expression in actively growing cells to eliminate the possibility of contact-inhibition–mediated elevation of p27. The actively cycling cells treated with FGF-2 demonstrated negligible staining for p27, whereas both mitogen-deprived cells (D0) and cells treated with antimitogens (TGF-β2 or cAMP) showed strong nuclear p27 staining in all cells (Fig. 6) . 

To determine whether antimitogen had an effect at the level of p27 synthesis, the p27 expression level was further determined using immunoblot analysis as a function of exposure time (Fig. 7A) . Cells were treated with either FGF-2 or FGF-2 plus 8-bromo cAMP for 1 hour, 6 hours, or 24 hours. Cells maintained in serum-free medium served as the control. One-hour treatment of cells with either FGF-2 or FGF-2 plus 8-bromo cAMP showed a slight decrease in the p27 level, when compared with the p27 level in cells maintained in the serum-free condition. When either treatment was applied to the cells for 6 hours, neither FGF-2 nor FGF-2 plus 8-bromo cAMP significantly altered the level of p27 when compared with that of the control. An interesting finding was that the p27 level was significantly increased when cells were maintained in the serum-free condition for 24 hours, whereas FGF-2 markedly decreased the p27 level, and 8-bromo cAMP reversed the effect of FGF-2. Relative density of the peptide bands was estimated and compared with the p27 level of cells maintained in the serum-free condition for 1 hour (control). As stated, depletion of serum for 24 hours significantly stimulated the p27 level. There was an approximate 75% increase in the p27 level when compared with the level of the control. When cells were treated with FGF-2, the p27 level was markedly decreased by 50%. The addition of 8-bromo cAMP prevented the decrease in the p27 level induced by FGF-2. 8-Bromo cAMP stimulated the expression of p27 by 45% when compared with the p27 level achieved by FGF-2 (Fig. 7B) . 

Nuclear p27 is phosphorylated at the residues of threonine (Thr) and serine (Ser) before nuclear export into the cytoplasm, in which the phosphorylated p27 is subjected to degradation by either the ubiquitin-proteasome pathway or ubiquitin-independent proteolytic cleavage. ³⁶ We, therefore, determined whether 8-bromo cAMP or TGF-β2 influences the phosphorylation of p27 (Fig. 8) . When cells were stained with anti-phosphorylated p27 antibody (phosphorylated at Thr¹⁸⁷), a few cells were stained with the antibody against the phosphorylated p27 in the cells treated with either antimitogen. Approximately 20% of cells were stained for the phosphorylated form of the protein. On the contrary, FGF-2 tripled the number of cells with nuclear staining of the phosphorylated p27, compared with those treated with either 8-bromo cAMP or TGF-β2. When compared with the staining profiles of total p27 as shown in Figures 3 and 4 , these data suggest that both 8-bromo cAMP and TGF-β2 prevent the phosphorylation process of p27, at least at Thr¹⁸⁷. Thus, the unphosphorylated p27 is retained in the nuclei as the active form to perform as a cdk inhibitor. These data further suggest that FGF-2 may be involved in the p27 degradation pathway by inducing phosphorylation of the molecules to facilitate the nuclear export process. This finding is consistent with our previous study (Kay, manuscript submitted). 

Discussion

It has long been believed that the capacity for regeneration of corneal endothelium after injury is severely limited in humans. Thus, the corneal endothelium is considered to be a nonreplicating tissue. ^{5 6} Joyce et al., ⁷ Senoo and Joyce, ⁸ and Senoo et al. ⁹ were the first to show that CECs in vivo are arrested in the G₁ phase of the cell cycle. Although this information suggests that CECs possess proliferative potential under physiologic conditions, mitosis of CECs is seldom observed in adult humans, even during the wound repair process. The underlying mechanisms that keep endothelial cells from moving out of the G₁ phase have been partially identified. Adult CECs may markedly decrease in their response to the growth-stimulatory factors. TGF-β2, present in aqueous humor of the anterior chamber, may suppress mitogenic activity of the cells, resulting in the maintenance of CECs in a G₁ phase–arrested state in vivo; and contact inhibition may block mitosis to avoid improper cell proliferation. ^{9 16 44} Inhibition of cell proliferation is central to the TGF-β response in many cell types, including endothelial cells. ^{19 45} TGF-β can induce antiproliferative responses at many points during the division cycle. However, these responses effectively inhibit cell cycle progression only during the G₁ phase. Among the known antiproliferative actions of TGF-β, this antimitogen inhibits cdk4 synthesis and cdk enzyme activity through the action of p15^INK4b and mobilizes p27 from the cyclin-D–cdk4 complex to the cyclin-E–cdk2 complex, resulting in the loss of activity of both kinases and concomitant arrest of G₁. ^{19 24 26} Although the action of TGF-β is well known, the mechanism by which cAMP induces cell cycle arrest is less understood, despite the fact that cAMP has long been recognized to inhibit the growth of certain cells. ^{29 30 31 32 33} In some cells cAMP blocks the mitogenic effects of growth stimulatory factors by upregulating p27 expression and preventing cdk4 activation. ^{31 32 33} Recently, cAMP has been reported to inhibit proliferation of aortic vascular smooth muscle cells by inducing p53 and p21. ⁴⁶  

Because TGF-β2 is a major TGF-β isoform in aqueous humor that constantly bathes CECs, and because cAMP, also present in aqueous humor, ⁴⁷ is readily activated by prostaglandin E₂ in CECs, ⁴⁸ we investigated the antiproliferative actions of these antimitogens. We initially examined the inhibitory activity of cAMP and TGF-β2 on serum-mediated endothelial proliferation. Both antimitogens were able to inhibit proliferation of rabbit CECs, but the maximal inhibition induced by either antimitogen only reached slightly more than 20%. Such a low level of inhibitory action is consistent with our previous study in which TGF-β2 showed a similar inhibition of endothelial proliferation ⁴¹ ; however, the level is far less than the previous data reported by Chen et al. ¹⁶ The differences between our study and that of Chen et al., may be a result of the different species investigated (rabbit cells in our study versus rat cells in theirs) and the different growth conditions used (10% serum only in our study versus 10% serum plus 25 ng/ml FGF-2). Nevertheless, these findings suggest that endothelial cells respond to the antimitogens, because their proliferation is impeded by the exposure to the antimitogen. The low level of inhibitory action of the two antimitogens may be beneficial to CECs that are constantly exposed to the growth-modulating factors in the aqueous humor. 

We then examined the expression pattern of cell cycle–regulatory proteins, including cyclins, cdks, and CKIs, using indirect immunofluorescent staining. Of particular interest, rabbit CECs under growth-supporting conditions demonstrated strong staining of p27 in nuclei in half of the cell population, whereas cyclins (cyclin-D1 and -E) and cdks (cdk2 and -4) showed very faint if any staining with their respective antibodies. These observations suggest that rabbit CECs reaching confluence may downregulate cyclins and cdks, whereas they stimulate the expression of p27. cAMP and TGF-β2 significantly enhanced the expression of nuclear p27, regardless of the cell growth stages, whereas neither cAMP nor TGF-β2 altered the staining potentials of cyclins and cdks. The stimulatory action of antimitogens on p27 expression was further examined using immunoblot analysis. Data shown in Figure 7 suggest that prolonged exposure of cells to antimitogens and extended mitogen-deprived conditions elevates the p27 level in rabbit CECs. They also indicate that antimitogens are able to prevent the decrease of the p27 level mediated by mitogen. 

The concentration of p27 is regulated predominantly by a posttranslational mechanism. ^{34 35} Both the ubiquitin-proteasome pathway and ubiquitin-independent proteolytic cleavage degrade p27. ³⁶ It has been known that phosphorylation of Thr¹⁸⁷ of p27 leads to ubiquitination and degradation of the molecule. ³⁷ Recent data have also suggested that the F-box protein, which functions as the receptor component of the ubiquitin–ligase complex, binds to p27 only when Thr¹⁸⁷ is phosphorylated. Such binding then results in the ubiquitination and degradation of p27. ^{37 49 50} We, therefore, examined whether phosphorylation of Thr¹⁸⁷ of p27 was influenced by the two antimitogens. Our data show that both cAMP and TGF-β2 prevented the phosphorylation process of p27, at least at Thr¹⁸⁷, thus blocking exportation of the inhibitor from the nuclei, suggesting that these antimitogens are able to maintain p27 in an active form in the nuclei. Phosphorylation at Ser¹⁰ was recently reported to be a major phosphorylation site of p27 that influences the protein stability. ⁵¹ It is unknown whether Ser¹⁰ is involved in ubiquitin-mediated proteolysis, as was the Thr¹⁸⁷ site of p27. When cells are given a mitogen such as FGF-2, the number of cells containing the nuclear phosphorylated p27 was markedly increased. These cells may be ready for nuclear export of p27 for ubiquitination. The mechanisms by which p27 is phosphorylated and antimitogen prevents phosphorylation are yet to be elucidated. Characterization of these mechanisms should shed light on fundamental issues, such as how the cell cycle is arrested in the G₁ phase in CECs under physiologic conditions. 

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

 Submitted for publication April 9, 2001; revised August 8, 2001; accepted August 14, 2001.

 Commercial relationships policy: N.

 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 TGF-β2, 8-bromo cAMP on cell proliferation in rabbit CECs. When cells reached approximately 60% confluence, they were starved of serum for 24 hours. Cells were then treated for 36 hours with one of the following: (A) DMEM-10 with 8-bromo cAMP in concentrations ranging from 0.1 mM to 3.0 mM; (B) DMEM-10 with TGF-β2 in concentrations ranging from 0.1 ng/ml to 10.0 ng/ml; (C) DMEM-10 with a combination of TGF-β2 (filled bars: 1.0 ng/ml; open bars: 10.0 ng/ml) and 8-bromo cAMP (0.1 mM, 0.3 mM, 1.0 mM, or 3.0 mM). Cells treated with TGF-β2 at 1 ng/ml or 10 ng/ml in the absence of cAMP served as controls to their counterparts simultaneously treated with cAMP. Data are representative of five experiments.

Figure 2.

 

Subcellular localization of cell cycle–regulatory proteins in rabbit CECs. Cells maintained in DMEM-10 were fixed in methanol for staining of cyclins and cdks and in paraformaldehyde for staining of p21 and p27. Cells were then respectively stained with anti-cyclin-D1, anti-cyclin-E, anti-cdk2, anti-cdk4, anti-p21, or anti-p27 antibodies, followed by staining with FITC-conjugated secondary antibody. Data are representative of five experiments. Bar, 20 μm.

Figure 3.

 

Subcellular localization of cell cycle–regulatory proteins in rabbit CECs treated with 8-bromo cAMP. When cells reached approximately 80% to 90% confluence, they were treated with 1.0 mM 8-bromo cAMP in DMEM-10 for 36 hours, fixed, and stained as described as in Figure 2 . Data are representative of four experiments. Bar, 20 μm.

Figure 4.

 

Subcellular localization of cell cycle–regulatory proteins in rabbit CECs treated with TGF-β2. When cells reached approximately 80% to 90% confluence, they were treated with 1.0 ng/ml TGF-β2 in DMEM-10 for 36 hours, fixed, and stained with respective antibodies as described as Figure 2 . Data are representative of four experiments. Bar, 20 μm.

Figure 5.

 

Expression of p27 in rabbit CECs treated with 8-bromo cAMP, TGF-β2, or FGF-2. When cells reached approximately 80% to 90% confluence, they were deprived of serum for 24 hours and then treated with 8-bromo cAMP (1 mM), TGF-β2 (1 ng/ml), a combination of TGF-β2 (1 ng/ml) and 8-bromo cAMP (1 mM), or FGF-2 (10 ng/ml). After 36 hours, cells were fixed in paraformaldehyde and stained for p27. Antimitogen treatment was performed in DMEM-10, whereas FGF-2 treatment was performed in the absence of serum. Data are representative of four experiments. Bar, 20 μm.

Figure 6.

 

Expression of p27 in actively growing rabbit CECs treated with 8-bromo cAMP, TGF-β2, or FGF-2. When cells reached approximately 50% to 60% confluence, they were deprived of serum for 24 hours (D0) and then treated with 8-bromo cAMP (1 mM), TGF-β2 (1 ng/ml), or FGF-2 (10 ng/ml). Treatment of cells with antimitogen and FGF-2 was performed as described in Figure 5 . After 24 hours, cells were fixed in paraformaldehyde and stained for p27. Data are representative of three experiments. Bar, 20 μm.

Figure 7.

 

Effect of 8-bromo cAMP on the expression of p27 in rabbit CECs. When cells reached 80% to 90% confluence, they were placed in serum-free medium for 24 hours and then treated with FGF-2 (10 ng/ml) in the presence or absence of 8-bromo cAMP (1.0 mM) for 24 hours. Proteins were extracted and 25 μg of each sample was subjected to SDS-PAGE on a 12.5% gel under reduced conditions, and immunoblot analysis was performed. (A) Lane 1: cells maintained in serum-free medium (D0); lane 2: cells treated with FGF-2 in D0; and lane 3: cells simultaneously treated with FGF-2 and 8-bromo cAMP in D0. (B) After gels were exposed on ECL film, the relative density of the p27 bands was estimated by using a one-dimensional image analyzer. Cells maintained in serum-free conditions for 1 hour served as the controls (100%). To control for loading, β-actin was immunoblotted in parallel. Open bars: cells maintained in D0; hatched bars: cells treated in FGF-2; filled bars: cells treated with FGF-2 and cAMP. Data are representative of three experiments.

Figure 8.

 

Subcellular localization of phosphorylated p27 in rabbit CECs. When cells reached 80% to 90% confluence, they were treated for 24 hours with 8-bromo cAMP (1 mM), TGF-β2 (1 ng/ml), or FGF-2 (10 ng/ml). Treatment with 8-bromo cAMP and TGF-β2 was performed in DMEM-10; treatment with FGF-2 was performed in D0. After fixation with paraformaldehyde, cells were stained with anti-phosphorylated p27 Thr¹⁸⁷ antibody and FITC-conjugated secondary antibody. Data are representative of three experiments. Bar, 20 μm.