Protein Kinase C Regulation of Corneal Endothelial Cell Proliferation and Cell Cycle

Abstract

purpose. The purpose of this study was to determine the role of protein kinase C (PKC) in corneal endothelial cell proliferation.

methods. Immunocytochemistry and Western blotting were used to define the PKC isoforms expressed in primary cultures of rat corneal endothelial cells. For proliferation studies, primary cultures of rat corneal endothelial cells were serum-starved for 48 hours and incubated for 2 hours with the PKC inhibitors staurosporine (10⁻⁹ to 10⁻⁷ M), chelerythrine (10⁻⁹ to 5 × 10⁻⁸ M), or calphostin C (10⁻⁹ to 10⁻⁷ M). Individual PKC isoforms were inhibited using PKCα antisense oligonucleotide transfection or exposure for 1 hour to myristoylated, pseudosubstrate-derived peptide inhibitors against PKCα, -αβγ, -ε, and -δ (10⁻⁸ to 10⁻⁶ M). Cells were then stimulated with 2.5% serum for 24 hours. Cell proliferation was measured with bromodeoxyuridine (BrDU) and Ki67 immunocytochemistry. Protein level of cyclin E was determined by Western blotting.

results. PKCα, -βΙΙ, -δ, -ε, -ι, -η, -γ, and -θ were detected in corneal endothelial cells. Maximum inhibition of PKC with staurosporine, calphostin C, and chelerythrine reduced cell proliferation to 7%, 31%, and 48% of control, respectively. Myristoylated peptide inhibition of PKCα and -ε reduced cell proliferation to 57% and 59% of control, respectively. PKCα antisense oligonucleotide reduced cell proliferation to 35% of control. Cyclin E protein level was decreased to 70%, 38%, 57%, and 43% of control in cells treated with calphostin C, staurosporine, chelerythrine, and PKCα antisense, respectively.

conclusions. PKC activity, in particular PKCα and -ε activity, is important in promoting corneal endothelial cell proliferation. Inhibition of PKC activity prohibits G1/S-phase progression and reduces cyclin E protein levels.

The corneal endothelium is a cell monolayer lining the posterior corneal surface that is critical in preventing corneal edema and maintaining corneal transparency. ¹ The corneal endothelium is considered to be a nonreplicating tissue, and consequently cell loss through disease or injury, such as from trauma or surgery, ^{2 3 4 5} can result in a permanent decline in cell density and an inability to maintain corneal transparency. ⁶ In addition to congenital hereditary endothelial dystrophy, ^{7 8 9 10} Fuchs’ dystrophy, ^{11 12} and inflammation, ¹³ age alone can account for corneal endothelial cell density decline. ⁶  

Although the corneal endothelium is considered a nonreplicating tissue, these cells possess proliferative capacity. Previous studies have revealed that corneal endothelial cells in vivo resemble limbal basal cells in their cell cycle marker profile and are arrested in G1 phase of the cell cycle. ^{14 15} Corneal endothelial cells can also overcome G1-phase arrest in vivo in the iridocorneal endothelial syndromes, characterized by uncontrolled proliferation of endothelial cells, supporting the proliferative capacity of endothelial cells. 

Protein kinase C (PKC) is a well-known regulator of cell proliferation. In particular, PKC can mediate G1-phase progression in the cell cycle. ^{16 17 18 19} PKC comprises a family of serine/threonine kinases important in intracellular signaling. ^{20 21 22} In addition to cellular proliferation, PKC has been implicated in the regulation of differentiation, migration, and apoptosis and in tumor promotion. ^{18 19 23 24} Eleven PKC isoforms have been identified in mammalian cells. These isoforms have distinct cellular location and function and are categorized as classical, novel, and atypical according to their activation sites. ^{22 25} The effect of PKC on the cell cycle has been reported to be both stimulatory and inhibitory, influenced by the PKC isoform and cell type studied as well as by the timing and duration of PKC activation or inhibition. ^{16 23 26} PKC has not to our knowledge been extensively studied in the corneal endothelial cell cycle. 

Our studies investigate the effect of PKC on corneal endothelial cell proliferation. We identified the PKC isoforms present in the corneal endothelium. We used three nonisoform selective PKC inhibitors—staurosporine, chelerythrine, and calphostin C—to determine the role of PKC activation in serum-stimulated proliferation. These three agents target different sites of the PKC molecule, lending specificity to their inhibition. The roles of individual PKC isoforms were studied with myristoylated pseudosubstrate-derived inhibitory peptides against PKCα, -ε, and -δ. ^{27 28 29} These peptides bind specifically to the catalytic domain of the targeted PKC isoform, inhibiting enzyme activation. As a second method to inhibit PKC isoforms specifically, cells were transfected with antisense oligonucleotides. Antisense oligonucleotides previously have been used to successfully modulate the activity of individual PKC isoforms. ^{30 31 32 33 34} The effect of PKC on G1 phase of the corneal endothelial cell was studied by evaluating cyclin E protein levels in cells treated with inhibitors of PKC activity. 

Materials and Methods

Corneal Tissue

Corneas were obtained from adult male Sprague–Dawley rats maintained in accordance with the ARVO Statement on the Use of Animals in Ophthalmic and Vision Research. Endothelial cell explant cultures were prepared according to Chen et al. ³⁵ Primary cultures were grown to confluency in Medium 199 (M199; GIBCO/BRL, Life Technologies, Grand Island, NY) supplemented with 50 μg/ml gentamicin (GIBCO/BRL, Life Technologies), 10% fetal bovine serum (HyClone, Logan, UT), and 25 ng/ml fibroblast growth factor (Biomedical Technologies, Stoughton, MA). Confluent cells were subcultured at low cell density onto sterile two- or eight-chamber slides (Laboratory Tek, Naperville, IL) and T75 flasks. For serum stimulation and PKC inhibition experiments, subcultured cells were serum-starved in M199 supplemented with 50 μg/ml gentamicin (M199-0) for 48 hours to synchronize cells in G0 phase of the cell cycle before treatment. All incubations were carried out in a 5% CO₂-95% air, humidified atmosphere at 37°C. 

Antibodies

Polyclonal rabbit antibodies to cyclin E and PKC isoforms -α, -βΙ, -βΙΙ, -δ, -ε, -γ, -ι, -η, -μ, -ζ, and -θ as well as fluorescein isothiocyanate (FITC)-conjugated and horseradish peroxidase (HRP)-conjugated anti-mouse IgG and anti-rabbit IgG antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Bromodeoxyuridine (BrDU) reagents, including mouse monoclonal anti-BrDU, were obtained from Amersham Pharmacia Biotech (Little Chalfont, Buckinghamshire, UK). BrDU-labeling reagents were used according to manufacturer’s protocol. Mouse monoclonal antibody to Ki67 was purchased from Novocastra (Newcastle, UK). FITC-conjugated streptavidin was obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). Nonmuscle myosin antibody was obtained from Biomedical Technologies (Stoughton, MA). 

Immunocytochemistry

Endothelial cells subcultured on two- or eight-chamber slides were rinsed with phosphate-buffered saline (PBS; GIBCO/BRL, Life Technologies) and fixed with 70% methanol for 30 minutes at −20°C. All further incubations were at room temperature. Slides were rinsed with PBS, and cells permeabilized for 10 minutes with 0.1% Triton X-100 in PBS. Nonspecific sites were blocked for 12 minutes using 4% bovine serum albumin (BSA) in PBS. Cells were incubated with primary antibody (BrDU: prepared per manufacturer’s protocol; dilutions in PBS: Ki67 1:100, streptavidin 1:200, PKCα 1:500, PKCβΙ 1:100, PKCβΙΙ 1:200, PKCδ 1:100, PKCη 1:100, PKCι 1:200, PKCθ 1:100, and PKCζ 1:50) for 2 hours. Slides were rinsed with PBS and incubated with secondary antibody (1:100 dilution in PBS) for 2 hours. Coverslips were mounted in Vectashield containing propidium iodide (PI) or DAPI (4′6 diamidino-2-phenylindole), which stain cell nuclei (Vector Laboratories, Inc., Burlingame, CA). Negative controls for the PKC antibodies have been described earlier. ³⁶  

Quantification of Proliferating Cells

For proliferation studies, cells were counted in a masked fashion, using the Nikon Eclipse E800 microscope (Garden City, NJ). At least five random 20× fields per specimen were examined. The total number of nuclei (PI or DAPI positive) was counted using the rhodamine or ultraviolet channel, respectively. BrDU- or Ki67-positive cells (FITC-positive) were counted using the FITC channel. Percentage of cells proliferating (% proliferation) was calculated by dividing the number of BrDU or Ki67 positive cells by the total number of cell nuclei. All experiments were conducted in duplicate and repeated at least three times for statistical evaluation unless otherwise noted. Statistical analysis was performed using Jandel Sigma Stat version 2.0 (Jandel Scientific Software, San Rafael, CA) to calculate significance using the paired t-test. 

PKC Inhibition

Synchronized serum-starved cells were incubated for 2 hours in M199-0 with or without the PKC inhibitors staurosporine (10⁻⁹, 10⁻⁸, and 10⁻⁷ M), chelerythrine (10⁻⁹, 10⁻⁸, and 5 × 10⁻⁸ M), or calphostin C (10⁻⁹, 10⁻⁸, and 10⁻⁷ M; BIOMOL, Biomolecules for Research Success, Plymouth Meeting, PA). Then 2.5% serum was added for 22 hours to stimulate the cells to enter the cell cycle. BrDU immunocytochemistry was used to indicate S-phase entry and Western blotting techniques evaluated cyclin E protein level. 

N-myristoylated, pseudosubstrate-derived inhibitory nonapeptides against PKCα, -δ, -ε, -αβγ, and -ζ were a generous gift of Driss Zoukhri (Schepens Eye Research Institute, Boston, MA) and Christian Sergheraert (Institut Pasteur de Lille, Lille, France). Synchronized serum-starved cells were exposed for 1 hour to myristoylated, pseudosubstrate-derived peptide inhibitors (10⁻⁸, 10⁻⁷, and 10⁻⁶ M) and then stimulated with 2.5% serum for 22 hours. Control samples were incubated with no peptide inhibitor. Cell proliferation was determined by Ki67 immunocytochemistry and counterstaining with DAPI. 

Antisense and sense oligonucleotides for PKCα were obtained from Midland Certified Reagent Company (Midland, TX). (Antisense sequence: 5′ (PS)GGTAAACGTCAGCCATGGTC3′; sense sequence: 5′(PS)GACCATGGCTGACGTTTACC3′.) Sense oligonucleotide was biotinylated. These sequences do not correspond to sequences of other PKC isoforms and correspond to oligonucleotides used successfully in previously published studies. ³³ Synchronized serum-starved cells were transfected with PKCα antisense oligonucleotide during an 18-hour period using the Qiagen Effectene nonliposomal lipid transfection reagents kit (Qiagen, Inc., Valencia, CA). A DNA:Effectene ratio of 1:26 was used. Cells were then stimulated with 2.5% serum for 24 hours. For controls, cells were transfected with PKCα sense oligonucleotide or with no oligonucleotide before serum stimulation. Viability was assessed using the Live/Dead assay kit (Molecular Probes, Eugene, OR). Transfection efficiency was determined by streptavidin staining of a biotinylated PKCα sense oligonucleotide. S-phase entry was detected by BrDU immunocytochemistry. Western blotting techniques were used to evaluate cyclin E, PKCα, and PKCβΙΙ protein levels. 

Western Blotting

Proteins were isolated by incubating cells for 30 minutes at 4°C in buffer composed of 50 mM Tris, pH 7.4, 250 mM NaCl, 2.5 mM EDTA, 1% NP-40–Triton X-100, 50 mM sodium azide, 0.1 mM sodium orthovanadate, 1 mM phenylmethyl sulfonylfluoride, 10 μg/ml aprotinin, and 10 μg/ml leupeptin (Sigma), homogenizing with a pellet pestle for 1 minute, and incubating overnight at 4°C. Protein content was quantified by spectrophotometry. Samples (with equal protein content) were electrophoresed on 10% polyacrylamide gels and electrophoretically transferred to PVDF membranes (Millipore, Bedford, MA). Membranes were incubated overnight at 4°C with 5% milk in PBS to block nonspecific binding sites, rinsed in 0.1% Triton X-100 in PBS (PBST) and incubated for 2 hours with primary antibody diluted in 5% milk. Antibody dilutions in 5% milk were as follows: cyclin E 1:200, myosin 1:200, PKCα 1:400, PKCβΙΙ 1:200, PKCβΙ 1:400, PKCγ 1:100, PKCδ 1:100, PKCε 1:1000, PKC ζ 1:50, PKCη 1:100, PKCθ 1:50, PKCμ 1:100, PKCι 1:250 in 5% milk. Blots were then rinsed 10 minutes three times with PBST and incubated in secondary antibody (1:10,000 in 5% milk) for 1 hour. After washing the membranes again three times in PBST for 10 minutes, antibody binding was visualized using the enhanced chemiluminescence method (Pierce, Rockford, IL). For quantification, films were digitally scanned using BDS-Image (Biological Detection System, Pittsburgh, PA), scans were analyzed with NIH Image software version 1.58, and protein content was normalized according to myosin protein content. In experiments with cyclin E antibody, the blots were reanalyzed using the myosin antibody. The bands were quantified as described above and the amount of cyclin E was standardized to the amount of myosin present in the same sample. 

Results

PKC Isoforms in the Corneal Endothelium

Eleven PKC isoforms have been identified in mammalian cells, but many tissues do not express all 11 isoforms. Also, individual isoforms may have tissue-specific and opposing activities. ²² To improve our understanding of the role of PKC in corneal endothelial cell proliferation and to facilitate the study of individual PKC isoforms, we defined which PKC isoforms are expressed in rat corneal endothelial cells. Rat corneal endothelial cells grown in primary culture were subcultured at low cell density, protein was extracted and evaluated for the expression of PKC isoforms using Western blot analyses. Cells were found to express PKCα, -βΙΙ, -δ, -ε, -θ, -η and -γ (Fig. 1) but not PKCβΙ, -μ, -ζ (data not shown). In addition, corneal endothelial cells were evaluated for expression of PKC isoforms using immunocytochemistry. PKCα, -βΙΙ, -δ, -θ, -η, and -ι were detectable. PKCβΙ, and -ζ were not detectable (Table 1) . 

Serum Dependence of Endothelial Cell Proliferation

PKC activity is isoform and cellular specific. When our studies began, it was not known whether PKC inhibition would have proliferative or antiproliferative effects on the corneal endothelial cell. To allow detection of either increased or decreased cell proliferation from PKC inhibition, we defined the amount of serum required to promote a moderate proliferative response. Primary cultures of rat corneal endothelial cells were subcultured at low cell density, serum-starved for 48 hours to synchronize the cells in G0 phase, and incubated for 24 hours with media containing 0% to 10% serum. Cell proliferation was quantified using BrDU staining (a marker of S-phase entry) and PI counterstaining. We found a concentration-dependent proliferative response that saturated at approximately 10% serum concentration (Fig. 2) . Cell proliferation (number of cells BrDU positive divided by the total number of cells indicated by PI staining) was measured at 3%, 15%, 41%, 54%, and 77% for cultures incubated with 0%, 1%, 2.5%, 5%, and 10% serum, respectively. On the basis of these results, 2.5% serum stimulation was used to study the effects of PKC inhibitors as a suboptimal concentration of serum to allow the detection of either proliferative or antiproliferative effects of PKC inhibition. 

Effect of PKC Inhibition on Cell Proliferation

Staurosporine can mediate G1-phase arrest and is a well-known inhibitor of PKC but it also inhibits other protein kinases. ¹⁷ To investigate a specific role of PKC, we studied three pharmacologic inhibitors of PKC (staurosporine, calphostin C, and chelerythrine) that act at different sites on the PKC molecule. Primary explants of rat corneal endothelial cells were subcultured at low cell density and serum-starved for 48 hours before a 2-hour incubation with a PKC inhibitor followed by 24 hours of serum stimulation. Cell proliferation was quantified using BrDU immunocytochemistry and PI counterstaining. Increasing concentrations of PKC inhibitor decreased the number of BrDU-labeled cells compared with control, whereas the total number of cells indicated by PI remained constant. A representative example of the immunocytochemical results is shown in Figure 3 . Using all three pharmacologic inhibitors, we found a concentration-dependent inhibition of cell proliferation. The percentage of proliferating cells was 43%, 24%, and 7% of control value (100%) for staurosporine (10⁻⁹, 10⁻⁸, and 10⁻⁷ M; Fig. 4A ), 70%, 44%, and 31% for calphostin C (10⁻⁹, 10⁻⁸, and 10⁻⁷ M; Fig. 4B ), and 89%, 80%, and 48% for chelerythrine (10⁻⁹, 10⁻⁸, and 5 × 10⁻⁸ M; Fig. 4C ), respectively. From this data we concluded that PKC inhibition reduces corneal endothelial cell proliferation. 

Effect of Inhibition of PKCα and -ε with Myristoylated, Pseudosubstrate-derived Peptides on Cell Proliferation

Of the PKC isoforms that are expressed in corneal endothelial cells, PKCα, -βΙΙ, -δ, and -ε have been linked most extensively to cell cycle regulation. ^{16 18 19} To determine which PKC isoforms are important in corneal endothelial cell proliferation, we inhibited these PKC isoforms using N-myristoylated, pseudosubstrate-derived peptides. Primary explants of rat corneal endothelial cells were subcultured at low cell density and serum-starved for 48 hours before a 1-hour incubation with myristoylated, pseudosubstrate-derived peptides against PKCα, -αβγ, -δ, and -ε. (The peptide against PKCαβγ indiscriminately inhibits PKCα, -βΙ, -βΙΙ, and -γ. A peptide inhibitor against PKCβΙΙ alone has not yet been synthesized.) Cells were then serum-stimulated for 24 hours followed by quantification of cell proliferation using Ki67 immunocytochemistry (a marker of actively cycling cells) and DAPI counterstaining. The percentage of proliferating cells was 57% and 59% of control value (100%) with PKCα and -ε inhibition (10⁻⁸ M peptide), respectively (Fig. 5) . Treatment with 10⁻⁶ M peptide led to cell death/apoptosis, as observed by decreased cell number and nuclear pyknosis on cell microscopy. Inhibition with 10⁻⁷ M peptide was less marked (data not shown), which was attributed to the behavior of the myristoyl moiety not dissolving homogeneously in serial dilutions, a phenomenon observed in previous reports. ²⁹ Cells treated with PKCδ and PKCαβγ peptide showed no detectable change in proliferation status. As a negative control to demonstrate the specificity of the peptide inhibitors, no inhibition was noted in cells treated with PKCζ peptide, because PKCζ was not expressed in endothelial cells by our immunocytochemistry and Western blot studies. 

From these findings we concluded that PKCα and -ε activity are important in promoting cell proliferation. Also, the antiproliferative effect of PKCα inhibition is masked with PKCαβγ inhibition, suggesting that PKCβΙΙ or -γ activity opposes PKCα activity and may inhibit proliferation. 

Effect of Inhibition of PKCα with PKCα Antisense Oligonucleotide Transfection on Cell Proliferation

To confirm the antiproliferative effect of PKCα inhibition, we used an alternative method of inhibiting PKCα using antisense oligonucleotide transfection. Synchronized serum-starved subcultures of rat corneal endothelial cells were transfected with PKCα sense or antisense oligonucleotide and serum-stimulated for 24 hours followed by quantification of transfection efficiency, viability, PKCα protein level, and cell proliferation. After optimizing transfection efficiency using guidelines in the Qiagen Effectene manufacturer’s protocol, we were able to demonstrate a transfection efficiency of more than 90%. Figure 6 illustrates streptavidin staining of biotinylated PKCα sense oligonucleotide, present in nearly every cell. A viability assay demonstrated that the majority of transfected cells remained viable, as indicated by positive esterase activity with the enzymatic conversion of calcein AM to fluorescent calcein in live cells and ethidium bromide staining of dead cells (data not shown). Western blot analysis of protein, extracted from cells treated as described above, demonstrated that transfection with PKCα antisense oligonucleotide reduced the amount of PKCα protein to 5% of nontransfected control (100%; Fig. 7 ), and 12% of sense oligonucleotide transfection control (100%; data not shown). PKCα antisense oligonucleotide transfection did not reduce PKC βΙΙ protein level (data not shown), indicating the isoform specificity of the antisense oligonucleotide inhibition. Transfection with PKCα antisense oligonucleotide significantly inhibited proliferation, but PKCα sense oligonucleotide transfection did not. The percentage of proliferating cells was measured as 35% and 69% of serum (100%), with antisense and sense oligonucleotide transfection, respectively (Fig. 8) . 

These results demonstrate that PKCα antisense oligonucleotide transfection is efficient, does not produce marked cell death, and specifically reduces PKCα protein level in corneal endothelial cells. Also, PKCα inhibition with transfection of PKCα antisense oligonucleotide results in reduced cell proliferation, supporting the role of PKCα in promoting cell proliferation. 

Effect of PKC Inhibition on Cyclin E Protein Level

Although human corneal endothelial cells do not replicate normally in vivo, they are arrested in G1 phase and maintain proliferative capacity. ^{14 15} We determined that PKC inhibition reduces cell proliferation and questioned whether PKC activity is important in G1-phase arrest. To study the effects of PKC inhibition on the corneal endothelial cell cycle, in particular on G1 phase, we evaluated protein levels of cyclin E, a cell cycle protein synthesized late in G1 phase that is necessary for progression to S phase and proliferation. Synchronized serum-starved subcultures of rat corneal endothelial cells were treated with PKC inhibitors (staurosporine, chelerythrine, calphostin C, PKCα antisense oligonucleotide transfection) as described above and serum-stimulated for 24 hours before protein extraction and analysis of cyclin E level by Western blotting techniques (Fig. 9A ). These blots were also analyzed using a nonmuscle myosin antibody, and the amount of cyclin E was standardized to the amount of myosin present in each sample. Figure 9B plots the results of three separate experiments, demonstrating a reduction in cyclin E to 71%, 58%, 38%, 47%, and 43% of control (100%) in cells treated with calphostin C (10⁻⁸ M), staurosporine (10⁻⁸ and 10⁻⁷ M), chelerythrine (10⁻⁸ M), and PKCα antisense oligonucleotide transfection, respectively. From this data we concluded that PKC inhibition reduces corneal endothelial cell proliferation by promoting G1-phase arrest and a concomitant reduction in cyclin E protein level. PKC inhibitors were not as effective in reducing cyclin E protein level compared with proliferation, suggesting that another cell cycle component important in G1-phase progression is also a target of PKC. 

Discussion

We determined for the first time that at least eight PKC isoforms are expressed in the corneal endothelium. Our studies demonstrated an inhibition of cell proliferation using three pharmacologic inhibitors of PKC as well as myristoylated, pseudosubstrate-derived peptides against PKCα and -ε, and antisense oligonucleotides to inhibit PKCα. All three methods of PKC inhibition (pharmacologic, pseudosubstrate-derived peptide inhibitors, and antisense oligonucleotide transfection) demonstrated a stimulatory role of PKC in endothelial cell proliferation. We also found that inhibition of PKC, particularly by staurosporine, affects protein levels of cyclin E, suggesting that PKC activity is important in G1-phase progression and that inhibition of PKC arrests cells in G1 phase before stimulation of cyclin E synthesis. 

Pharmacologic inhibition was chosen as a means to modulate PKC activity because of the lack of consistent findings noted in preliminary studies in which phorbol esters were used to stimulate PKC activity. This inconsistency may be due to the broad nature of phorbol ester stimulation, activating all but the atypical PKC isoforms, and the critical timing necessary for influencing cell cycle dynamics. Prolonged activation of PKC with phorbol esters leads to downregulation of PKC activity. Hence, phorbol ester treatment may lead to early activation and late inhibition of numerous PKC isoforms, some with possibly opposing activities. Our findings are supported by a study that demonstrated both stimulatory and inhibitory effects of phorbol esters on cell proliferation influenced by the phase of the cell cycle during the time of phorbol ester treatment. ¹⁶ Our findings are also supported by our data indicating that inhibition with a pseudosubstrate-derived peptide inhibitor against PKCαβγ overcame the antiproliferative effect of inhibition of PKCα alone. 

PKCα, -βΙΙ, -ε, and -δ were chosen for evaluation because they have been shown to be important in cell cycle events in other cell systems. ^{18 19} PKCα increased cell proliferation in smooth muscle cells and decreased proliferation in breast cancer cells. ¹⁶ PKCα and -ε increased cell proliferation in NIH3T3 cells, whereas PKCδ inhibited proliferation in these cells. ³⁷ PKCβΙΙ has been demonstrated to play a role in G2/M-phase transition. ¹⁶ In the present studies, both PKCα and -ε appear to be important for stimulation of corneal endothelial cell proliferation. 

We evaluated the effect of PKC inhibition on cyclin E protein expression to determine whether PKC is important in G1-phase arrest of corneal endothelial cells. Cyclin E positively regulates the activity of cyclin-dependent kinase 2 (cdk2). Activity of cyclinE/cdk2 complex is required for S-phase entry. ³⁸ Future studies, evaluating cyclin E activity and its interaction with cdk2, as well as cyclin D, p27, and pRb, will further characterize the cell cycle profile of cells treated with PKC modulators. This will help to better define where in the cycle (between G0 and S phase) cells are affected by PKC activity. As human corneal endothelial cells are arrested in G1 phase, understanding the mechanisms of G1-phase arrest in corneal endothelial cells may facilitate identification of factors that may allow endothelial cells to transiently overcome G1-phase arrest and proliferate in a controlled fashion. Factors, such as specific activators of PKCα and -ε may provide treatment for diseases characterized by reduced corneal endothelial cell density and corneal edema. 

In conclusion, PKC appears to participate in a signal transduction cascade leading to stimulation of corneal endothelial cell proliferation. The PKC isoforms -α and -ε, in particular, demonstrate important activity in cell cycle progression in these cells. Specific PKC isoform activity can be effectively modulated with myristoylated, pseudosubstrate-derived inhibitory peptides and antisense oligonucleotides, and this intervention may be useful in the treatment of disorders with abnormalities of endothelial cell density. 

 Supported by National Institutes of Health Grants R01 EY05767 (NCJ) and R01 EY09057 (DAD) and the Schepens Eye Research Institute.

 Submitted for publication March 6, 2000; revised July 24, 2000; accepted August 17, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Identification of PKC isoforms in corneal endothelial cells. Rat corneal endothelial cells were subcultured at low density and evaluated for presence of PKC isoforms via Western blot analysis. Arrow, band of interest. The apparent molecular weights of the PKC isoforms are: PKCα, 77 kDa; -βII, 77 kDa; -δ, 75 kDa; -ε, 90 kDa; -γ, 78 kDa; -η, 78 kDa; -θ, 80 kDa.

Table 1.

 

PKC Isoform Expression in the Corneal Endothelium

PKC Isoform	α	βI	βII	δ	ε	γ	η	ι	μ	θ	ζ
Immunocytochemistry	+	−	+	+	∗	N	+	+	N	+	−

 +, detected; −, not detected;

∗  , equivocal due to poor antibody labeling; N, not tested.

Figure 2.

 

Serum of 2.5% stimulates a moderate proliferative response in corneal endothelial cells. Serum-starved rat corneal endothelial cells synchronized in G0 phase were stimulated to enter the cell cycle by incubation for 24 hours with media containing 0% to 10% serum. Cell proliferation was quantified using BrDU incorporation and PI counterstaining. Results are mean ± SE of three independent experiments.

Figure 3.

 

Staurosporine inhibition of PKC reduces cell proliferation. Serum-starved rat corneal endothelial cells treated with staurosporine demonstrated a concentration-dependent reduction in S-phase entry indicated by BrDU-positive staining. Left: BrDU-positive staining in cells treated with staurosporine at concentrations 0, 10⁻⁹, 10⁻⁸, and 10⁻⁷ M; right: total number of cells from the same microscopic field, determined by PI staining of nuclei. Magnification, ×20.

Figure 4.

 

Staurosporine, calphostin C, and chelerythrine concentration-dependent inhibition of cell proliferation. Percent proliferation (determined by dividing the number of BrDU-positive cells by the total number of PI positive cells × 100) was measured in serum-starved endothelial cells treated for 2 hours with staurosporine (A), calphostin C (B), and chelerythrine (C) before serum stimulation for 24 hours. Results are mean ± SE of three independent experiments prepared in duplicate for (A) and (B) and two independent experiments prepared in duplicate for (C). *P < 0.05.

Figure 5.

 

Myristoylated, pseudosubstrate-derived peptide inhibition of PKCα and -ε result in reduced cell proliferation. Synchronized serum-starved rat corneal endothelial cells were incubated for 1 hour with myristoylated, pseudosubstrate-derived peptides (10⁻⁸ M) against PKCα, -ε, -δ, -αβγ, and -ζ before serum stimulation for 24 hours and quantification of cell proliferation using Ki67 immunocytochemistry and DAPI counterstaining. Control cells were exposed to no peptide or to PKCζ (not present in corneal endothelial cells). Results are mean of two independent experiments prepared in duplicate.

Figure 6.

 

PKCα oligonucleotide transfection is more than 90% efficient. Synchronized, serum-starved corneal endothelial cells were transfected with biotinylated PKCα oligonucleotide before serum stimulation for 24 hours. Streptavidin-staining revealed the presence of biotinylated sense oligonucleotide in nearly all cells (A). The total number of cells is shown with PI nuclear staining (B). Magnification, ×20.

Figure 7.

 

PKCα antisense oligonucleotide transfection reduces PKCα protein level. Synchronized, serum-starved corneal endothelial cells were transfected with PKCα antisense and sense oligonucleotides and serum-starved for 24 hours followed by protein extraction and analysis by Western blotting techniques. Control cells were not transfected.

Figure 8.

 

Inhibition of PKCα using antisense oligonucleotide transfection reduces cell proliferation. Synchronized serum-starved corneal endothelial cells were transfected with PKCα antisense oligonucleotide and serum-stimulated for 24 hours followed by quantification of cell proliferation with BrDU immunocytochemistry and PI counterstaining. Control cells were not transfected or were transfected with sense oligonucleotide. Results are mean ± SE of three independent experiments prepared in duplicate.* P < 0.05.

Figure 9.

 

Cyclin E protein level is decreased with PKC inhibition. Synchronized serum-starved subcultures of corneal endothelial cells were treated for 2 hours with calphostin C (10⁻⁸ M), staurosporine (10⁻⁸ and 10⁻⁷ M), or chelerythrine (10⁻⁸ M) or transfected with PKCα antisense oligonucleotide, before serum stimulation for 24 hours and protein extraction for analysis of cyclin E protein level by Western blotting techniques (A). (B) Graph is the mean of cyclin E protein level standardized to the amount of nonmuscle myosin present in the same samples ± SE of three independent experiments.* P < 0.05.

The authors thank David Mello, Deshea Harris, and Robin Hodges for their outstanding technical assistance in various phases of this project.