Increased Platelet-Activating Factor Receptor Gene Expression by Corneal Epithelial Wound Healing

Abstract

purpose. Platelet activating factor (PAF) is a potent inflammatory mediator the synthesis of which increases in the cornea after injury. The effects of PAF are mediated by receptors (PAF-R), which are present in target cells. This study was undertaken to investigate the effects of wound healing, PAF, and growth factors on modulating PAF-R mRNA levels in corneal epithelial cells.

methods. Cultures of rabbit corneal epithelial (RCE), rabbit limbal epithelial (RLE), rabbit corneal fibroblast (RCF), and rabbit corneal endothelial (RCEn) cells, as well as rabbit corneal keratocytes (RCKs) were used. For the in vivo wound-healing experiments, a 7-mm central corneal de-epithelialization was performed in anesthetized rabbits. For the in vitro experiments, wounded rabbit corneas were maintained in organ culture. Corneas were stimulated with 120 nM PAF or preincubated with PAF antagonists, cyclohexamide (CHX) or actinomycin D (AcD) before adding PAF. RCE cells were stimulated with transforming growth factor (TGF)-β1, -β2, and, -β3, basic fibroblast growth factor (bFGF), keratinocyte growth factor (KGF); and hepatocyte growth factor (HGF). Total RNA was isolated and PAF-R expression evaluated by reverse transcription–polymerase chain reaction (RT-PCR), Northern blot analysis, and quantitative RT-PCR.

results. PAF-R mRNA was expressed in RCE, RLE, and RCEn cells and RCKs, but not in RCFs. After epithelial injury, PAF-R expression increased from 2.5 to 4 times, both in vitro and in vivo. Addition of cPAF further stimulated PAF-R gene expression in epithelium, which was abolished by PAF antagonists. Quantitative RT-PCR revealed that PAF stimulated PAF-R mRNA threefold after injury. The induction of PAF-R by its agonist required previous injury and was inhibited by AcD but not by CHX. Treatment of RCE cells with TGF-β1, -β2, or -β3, HGF, and KGF increased mRNA in PAF-R; however, bFGF had no effect.

conclusions. Corneal injury produces changes in PAF-R mRNA expression. Whereas stroma fibroblastic cells lost the PAF-R gene expression found in keratocytes, corneal epithelial injury upregulated PAF-R mRNA. These results suggest that activation of selective growth factors and increases in PAF synthesis after injury stimulate PAF-R gene transcription and constitute important feedback mechanisms needed to maintain the inflammatory process and regulate epithelial wound healing.

Platelet activating factor (PAF) is a bioactive lipid that accumulates in the cornea after injury ¹ and produces several biochemical responses associated with inflammation and wound healing. It induces a rapid activation of phospholipase A₂, the release of arachidonic acid, and the selective synthesis of prostaglandins in corneal epithelium. ² It also activates mitogen-activated protein kinase ³ and increases intracellular calcium by stimulating calcium influx into the cells. ⁴ In addition, PAF activates the gene expression of selective metalloproteinases (MMPs) involved in tissue remodeling, such as MMP-1 and MMP-9, ^{5 6} as well as urokinase plasminogen activator (uPA) expression, ⁷ probably through the stimulation of an AP-1 transcription factor. ⁵ PAF also stimulates the expression of COX-2, the inducible isoform of the cyclooxygenase that synthesizes prostaglandins associated with the inflammatory response. ⁴ These events are receptor-mediated and can be inhibited using specific PAF antagonists. A PAF receptor (PAF-R) has been cloned from several species ⁸ and contains a seven-transmembrane domain typical of G-protein–coupled receptors. PAF-R mRNA in rat is expressed in many tissues, such as spleen, small intestine, kidney, liver, lung, and brain. ⁹ In the eye, PAF-R had been localized in iris-ciliary body, retina, and corneal epithelium. ^{10 11 12} Recently, we reported the presence of a single PAF binding site in the plasma membrane of corneal epithelial cells, as well as a partial sequence of a polymerase chain reaction (PCR) product in rabbit corneal epithelium with 87% sequence homology to the human PAF-R. ¹³ In addition to being produced by corneal cells, PAF is produced after stimuli by inflammatory cells, such as monocyte/macrophages, leukocytes (PMNs), and eosinophils ¹⁴ and acts through juxtacrine or paracrine mechanisms on target cells. ¹⁵ Depending on the insult (e.g., alkali burn, graft rejection), the cornea is infiltrated with these cells that release PAF, which, through its receptor, could elicit specific cellular responses. Regulation of PAF-R expression in corneal cells could therefore be one of the key determinants in PAF-mediated corneal inflammation and wound healing. Changes in PAF-R mRNA expression have been shown in some leukemic cell lines (monocytes and eosinophils), ^{16 17 18} but there are no studies on the regulation of PAF-R gene expression in the eye, nor are there in vivo studies, except for a report on the ileum, which shows an increase of PAF-R mRNA expression in a rat model of intestinal injury. ¹⁹ In the present study, we investigated the presence of PAF-R in different corneal cells, determined PAF-R gene expression using in vivo and in vitro models of corneal epithelial injury, examined the regulation of PAF-R mRNA expression by its agonist, and investigated the effects of exogenous added growth factors on PAF-R gene expression. 

Materials and Methods

cPAF (1-alkyl-2nmethylcarbamyl-sn-glycerol-3phosphorylcholine; an analog of PAF with 98% or more purity) was obtained from Cayman (Ann Arbor, MI). An 18.6-mM stock solution was prepared in ethanol and stored under N₂ at −20°C. Under these conditions, cPAF is stable for more than 1 year (Bazan, unpublished observations, 1990). Before use, the solution was diluted in Dulbecco’s phosphate-buffered saline (PBS). The PAF antagonists BN 50727 and BN 50730 were gifts of Pierre Braquet, (Institute Henri Beaufour, Le Plessis–Robinson, France). These antagonists were prepared in a 3.85-mM stock solution in dimethyl sulfoxide (DMSO) and diluted in PBS before use. Cycloheximide (CHX; Sigma, St Louis, MO) was dissolved in ethanol, and actinomycin D (AcD, Sigma) was prepared in a 10-mM stock solution in DMSO. Final concentrations of ethanol and DMSO in the medium were less than 0.01% and did not affect PAF-R expression. Hepatocyte growth factor (HGF; human recombinant double-chain) was a gift from Genentech (San Francisco, CA). Keratinocyte growth factor (KGF), human recombinant transforming growth factor (TGF)-β1, -β2, and -β3, and basic fibroblast growth factor (bFGF) were from Sigma. All the growth factors used in the experiments were 95% to 98% pure. Epidermal growth factor (EGF), insulin, antibiotic-antimycotic mixture, trypsin-EDTA and platelet-poor horse serum (PPHS) were from Sigma. Dispase II was from Boehringer–Mannheim (Indianapolis, IN). Collagenase was from Worthington Biochemical (Lakewood, NJ). Taq DNA polymerase was from Perkin–Elmer (Branchburg, NJ), and M-MLV reverse transcriptase, ethidium bromide, dNTP mix, DNA mass ladder, and fetal bovine serum (FBS) were from Gibco (Grand Island, NY). Hybond N⁺ membranes were from Amersham (Arlington Heights, IL), pHcGAP (human GAPDH cDNA–containing plasmid) was from American Type Culture Collection, (Rockville, MD), and the cDNA fragments were labeled withα -³²P-dCTP (Dupont–NEN, Boston, MA). 

Cell Cultures

Rabbit Corneal Epithelial (RCE) and Endothelial (RCEn) Cells.

Rabbit eyes were obtained from Pel-Freeze Biologicals (Rogers, AR), maintained in ice, and used within 24 hours of enucleation. Only corneas with intact epithelia and without visible abrasions were used. Corneas were collected in sterile Dulbecco’s modified Eagle’s medium (DMEM)-Ham’s F12 (1:1) containing gentamicin (50 μg/ml) and an antibiotic-antimycotic mixture (100 U/ml penicillin; 100 U/ml streptomycin; 0.25 μg/ml amphotericin). RCE cells were cultured, as described. ¹³ For RCEn cells, the endothelium was collected, treated with trypsin-EDTA (0.05%–0.02%) for 20 minutes at 37°C, transferred into six-well culture dishes (with each well corresponding to endothelium from 8–10 corneas), and fed with DMEM-F12 (1:1) supplemented with EGF (10 ng/ml), insulin (10 μg/ml), and 5% FBS. Media were changed three times weekly. Primary cultured RCEn cells were used. 

Rabbit Corneal Fibroblasts (RCFs) and Rabbit Limbal Epithelial (RLE) Cells.

Cells were obtained from explant cultures. After the removal of the corneal epithelium and endothelium, the cornea was separated from the limbus by a 10-mm trephine, cut in approximately 1 × 1 mm² portions, and placed in 100-mm tissue culture dishes, ensuring that, for RLE cells, the epithelium side was touching the surface of the culture dish. Cells were cultured, as previously described. ²⁰  

Rabbit Corneal Keratocytes (RCKs).

To obtain (RCKs), the procedure of Beales et al. ²¹ was followed, with minor modifications. Epithelium and endothelium were removed from the cornea, and the stroma was treated with 0.25% trypsin in Hank’s balanced salt solution (HBSS) for 4 hours at room temperature. After washing, the stroma was predigested with 1 mg/ml collagenase in HBSS for 10 minutes to remove any remaining epithelial and endothelial cells. Afterward, the stroma was washed three times with DMEM and incubated in the same medium with 3 mg/ml collagenase for 4 hours at 37°C. Cells were either dissolved in lysis buffer to obtain RNA or collected by centrifugation and resuspended in 10 ml DMEM, plated at 6 to 8 × 10⁵ cells/60-mm dish with DMEM-1% PPHS, and allowed to attach overnight. The next day, the medium was changed to DMEM without PPHS. Medium was changed every 2 to 3 days. The cells were treated with 0.05% trypsin-0.02% EDTA and subcultured in calcium-free DMEM at 70% to 80% confluence with a 1:2 split for another two passages. 

Corneal Epithelial Wound In Vivo and In Vitro

New Zealand White rabbits (2–3 kg) of both sexes were used. The animals were maintained with food and water ad libitum until used. The experiments were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the Louisiana State University Medical Center protocol for animal studies. The animals were anesthetized with an intramuscular injection of 1 ml/kg (body weight) containing a mixture of ketamine-xylazine (100 mg/ml each) in distilled water. Corneas were injured by gently scraping a 7-mm diameter central area of corneal epithelium. The right cornea was injured and the left used as control. Corneal epithelium from the entire cornea was collected 48 hours after injury, at which time the epithelium covered the injured area so that the collected tissue included all epithelium and no limbus. For the in vitro experiments, commercially obtained corneas were dissected, a center 7-mm diameter epithelial wound was made, and the corneas were incubated in DMEM, as described. ²² Epithelium was collected at different times after injury, according to the experiments. 

PAF-R Regulation Experiments

In the experiments with cells, cultures were shifted to DMEM-F12 serum-free medium without EGF or insulin overnight. Serum-starved cells were left untreated or treated with any of the following agents: cPAF (120 nM), HGF or KGF (each at 20 ng/ml concentration), bFGF (50 ng or 100 ng/ml) or TGF-β1, -β2, or -β3 (each at 1 or 10 ng/ml) and incubated for the times specified in each experiment. For the experiments with PAF-R antagonist, 4 μM BN50727 or BN50730 was added to the in vitro wounded corneas 30 minutes before adding cPAF. For protein synthesis or RNA synthesis inhibition experiments, wounded corneas in vitro were pretreated with 30 μM CHX or 10 μM AcD for 1 hour before adding cPAF. 

RNA Extraction and RT-PCR

For reverse transcription–polymerase chain reaction (RT-PCR) RNA was extracted, as previously described, ¹³ and a 2.5-μg sample denatured at 65°C for 10 minutes with 2.5 U RNase inhibitor and 0.5 μg Oligo-d(T) adjusted with DEPC (diethyl pyrocarbonate)-treated water to reach a final volume of 15μ l. First-strand cDNA was synthesized using 100 U M-MLV reverse transcriptase, 4 mM dNTP, and 10 mM dithiothreitol. The mixture was heated at 37°C for 60 minutes in 25 μl. PCR was performed using a 1:10 dilution of cDNA and 1.5 U Taq DNA polymerase in a 50-μl reaction mixture containing 1× PCR buffer (10 mM Tris-HCl, 50 mM KCl, and 0.01% Triton X-100), 2.5 mM MgCl₂, 0.2 mM of each dNTP, and 0.4 μM of each upstream and downstream primer, as described. ¹³ The sense and antisense primers for rabbit β-actin were: 5′-AAG-ATC-TGG-CAC-CAC-ACC-TT-3′ and 5′-CGA-ACA-TGA-TCT-GGG-TCA-TC-3′. The reaction was started at 95°C for 10 minutes, followed by 36 cycles at 95°C for 30 seconds, 60°C for 30 seconds, and 72°C for 45 seconds and a final cycle at 72°C for 7 minutes for PAF-R. For β-actin, 34 cycles were performed at 95°C for 30 seconds, 62°C for 30 seconds, and 72°C for 45 seconds and a final cycle at 72°C for 7 minutes. Amplified cDNA products were resolved in 1.8% agarose gel containing 1 μg/ml ethidium bromide. A 100-bp ladder was used to determine the size of the products. The expected sizes of the cDNA fragments were 487 bp for PAF-R and 125 bp for β-actin. The sequence of the PCR product is available in GenBank (accession number A1279940). 

Quantitative RT-PCR

Assays were performed using the RNA PCR kit (GeneAmp; Perkin–Elmer, Foster City, CA). PAF-R product was obtained and purified using a kit (Qiagen, Chatsworth, CA). A composite 3′ primer (primer 3: 5′-GGA-TGA-TGA-CCT-CAT-AAT-GCT-CAA-AGC-A-3′) with primer 1 was designed to produce a sequence approximately 100 bp shorter than the original PAF-R primer but with the same flanking primer site that primers 1 and 2 used to amplify PAF-R. ¹³ The PAF-R product was then amplified using the PAF-R primers 1 and 3 and reamplified using PAF-R primers 1 and 2, followed by purification to obtain the internal standard. For quantitative RT-PCR, 1 μl of DNA of each sample was mixed with a serial dilution of competitor DNA (internal standard) after first-strand cDNA synthesis, and PCR was performed in a 50-μl volume using PAF-R primers 1 and 2. ²³ Twenty-five microliters of the PCR products were run in a 2% agarose gel in 0.5% of Tris-borate electrophoresis (TBE) buffer and stained with ethidium bromide. The intensity of both target PAF-R and competitive bands were evaluated with a phosphorimager and analyzed by computer (Molecular Analyst; Bio-Rad, Richmond, CA). Standard curves were then constructed using this fixed amount of internal standard and variable amounts of total RNA from experimental samples, to determine the linear range of internal standard to target mRNA PCR product ratios. ²³  

Northern Hybridization

Rabbit PAF-R cDNA probe was obtained from purified RT-PCR products using a DNA purification kit (Wizard PCR Prep; Promega, Madison, WI). A GAPDH probe (0.75 kb) was used to determine the relative amount of RNA in each lane by rehybridizing the blots that had been probed with PAF-R. The cDNA fragments were labeled withα -³²P-dCTP by random primer extension with a final specific activity of 2 to 4 × 10 ⁶ cpm/ml. Fifteen micrograms per lane of total RNA was denatured, electrophoresed on a 1.2% formaldehyde agarose gel, and transferred to a membrane (N+, Hybond; Amersham). Hybridization was performed for 18 hours at 42°C in a buffer containing 5× SSC (1× SSC = 3 M NaCl, 0.3 M sodium citrate [pH 7.0]), 1% sodium dodecyl sulfate (SDS), 1× Denhardt’s solution, 50% formamide, 100 μg/ml denatured salmon sperm DNA, and 2 to 4 × 10⁶ cpm/3 to 8 ng/ml ³²P-labeled probe. After hybridization, the filters were successively washed in a solution containing 1× SSC and 0.1% SDS for 30 minutes at room temperature, 1× SSC and 0.1% SDS for 30 minutes at 60°C, and 0.1× SSC and 0.1% SDS for 30 minutes at 60°C. The filters were exposed to a phosphorimager, and the intensity of each mRNA band was normalized as a ratio to that of the GAPDH mRNA band, based on quantification by computer (Molecular Analyst; Bio-Rad). 

Results

Using primers that recognize a PCR product in rabbit tissue with 87% homology to the human PAF-R, ¹³ we analyzed the distribution of PAF-R mRNA in different rabbit corneal cells. Amplification products of the expected size were detected in fresh, as well as in first-passage, RCE cells (Fig. 1A ). Comparatively, PAF-R mRNA was higher in RLE cells than in RCE cells when expressed in relation to β-actin, and although the differences were small, they were consistent in all the experiments. PAF-R mRNA was also expressed in RCEn cells, but not in cultured RCFs. Interestingly, corneal keratocytes obtained from fresh tissue or after cells were cultured in serum-free medium expressed PAF-R mRNA (Fig. 1B) . The RCKs showed, under phase-contrast microscopy, a dendritic morphology similar to that already described, ^{21 24} and when 10% FBS was added to the cultures for 2 days, the cells changed to a fibroblastic appearance, and PAF-R gene expression was lower. After a second passage of the keratocytes in the presence of 10% FBS, the PAF-R gene expression was lost (data not shown). 

Previous studies have shown that the actions of PAF in corneal epithelium are blocked by PAF antagonists. To investigate the regulation of PAF-R mRNA in an in vivo model, corneas were injured and PAF-R mRNA induction determined during wound healing. By semiquantitative RT-PCR, it was found that 48 hours after in vivo injury, there was a fourfold increase in PAF-R gene expression compared with control. Similar results were observed when the corneas were injured in vitro and incubated for 48 hours without serum or growth factors (Fig. 2A ). To confirm these results, Northern blot analysis was performed using a rabbit cDNA probe. The PAF-R probe was hybridized to a corneal epithelium mRNA species of approximately 3.6 kb (Fig. 2B) and, after 48 hours of corneal epithelial injury, there was a 4.2-fold increase in vivo and a 2.5-fold increase using the organ culture model in PAF-R gene expression relative to GAPDH. 

The synthesis of PAF is activated after corneal injury ¹ and triggers several cellular responses mediated by the PAF-R; therefore, we investigated whether PAF actions could modulate PAF-R gene expression. Noninjured and de-epithelialized corneas were incubated in organ culture with 120 nM cPAF for 48 hours, and the expression of PAF-R mRNA was analyzed by quantitative RT-PCR. PAF further increased PAF-R gene expression by three times in injured corneas, an increase that was inhibited by the PAF antagonist BN50727 (Fig. 3) . Another PAF antagonist, BN50730, had similar effects (data not shown), implicating receptor-mediated mechanisms of PAF in its receptor’s expression. Increased induction was observed after 4 hours of incubating injured corneas with cPAF, and further increases were found at 8 hours and sustained at similar levels up to 48 hours. By 72 hours, there was no significant effect by PAF (data not shown). The induction of PAF-R mRNA by its agonist was not inhibited by CHX (Fig. 4) , suggesting that the expression of the PAF-R gene is independent of protein synthesis. Moreover, there was an increase in PAF-R mRNA after CHX treatment alone, implicating certain proteins as required for the suppression of PAF-R transcription or, alternatively, for the degradation of PAF-R mRNA. Injured corneas incubated with 10 μM AcD and then stimulated with PAF showed an inhibition of PAF-R mRNA expression (Fig. 4) , suggesting that PAF stimulates the expression of its receptor by increasing gene transcription. The upregulation of PAF-R gene expression by PAF is observed only after epithelial injury. Noninjured corneas incubated with PAF in organ culture for 24 or 48 hours showed no changes; RCE cell cultures stimulated with PAF for short (2, 4, and 6 hours; for 4 hours see Fig. 5A ) or for longer (8, 12, or 24 hours; data not shown) times also showed no changes, suggesting that during wound healing, there are other factors that trigger the PAF response. 

Corneal injury promotes the release of several growth factors, so we questioned their possible involvement in PAF-R gene expression. Our experiments using the isolated cornea in an organ culture model suggest that the candidate’s growth factors must be expressed in corneal cells and must have receptors expressed in epithelial cells. TGF-β1,-β2, and -β3 and bFGF and their receptors have been found in epithelial and stromal cells. ²⁵ KGF and HGF are synthesized in stromal cells, and their receptors are expressed in epithelial cells ^{25 26} ; therefore, we investigated these growth factors as possible candidates to increase PAF-R gene expression. RCE cells stimulated with both paracrine growth factors, HGF and KGF, increased PAF-R mRNA three times compared with controls (Fig. 5A) . When RCE cells were stimulated with 1 or 10 ng/ml of TGF-β1, -β2, or -β3 for 4 hours, there was a 2- to 2.5-fold increase in PAF-R mRNA compared with nonstimulated cells, whereas RCE cells stimulated with 50 or 100 ng/ml bFGF, another growth factor with similar epithelial–stromal interactions, had no effect on PAF-R induction (Fig. 5B) . 

Discussion

We have recently demonstrated in bovine and rabbit corneal epithelial cells, ¹³ as well as in human corneal epithelial cells, ²⁷ that PAF contains a plasma membrane receptor. The present studies refer to the regulation of the gene expression of PAF-R. At present, there are no commercially available antibodies for PAF-R that react specifically with rabbit tissue. For this reason, we were unable to examine the translation of the gene to protein. In this study we showed that PAF-R mRNA is expressed in different corneal cells, demonstrating for the first time that corneal endothelium expresses a PAF-R gene and suggesting that PAF can affect some functions of this corneal layer. In fact, it has been shown that PAF inhibits transendothelial fluid transport, causing edema, an effect inhibited by a PAF antagonist. ²⁸ Furthermore, after anterior segment inflammation (e.g., uveitis), PAF increases in aqueous humor. ²⁹ It is important to consider this potent inflammatory mediator in pathologic conditions in which the function of the endothelium is compromised. PAF antagonists could be of therapeutic importance in attenuating damage of the endothelium during prolonged inflammation. 

One interesting finding is that stromal cells express PAF-R mRNA under naive conditions or in cultures with media but without serum. It has been reported that stromal cells cultured in 10% FBS change their phenotype and mimic the fibroblastic type observed during corneal wound healing. ²¹ Under these conditions, we did not detect PAF-R gene expression. This could be a mechanism of protection after corneal injury to avoid the effects of PAF on fibroblasts. Previous studies have shown that PAF induces the synthesis of MMPs and uPA in corneal epithelium. ⁵ During inflammation, there is an increase in PAF synthesis ¹ ; therefore, downregulation of PAF-R expression when keratocytes are transformed into fibroblasts could avoid extracellular matrix degradation by PAF. These results also suggest that, after inflammation, PAF synthesized in stromal cells ³⁰ could act as a paracrine mediator in epithelial and endothelial cells to exert its effects. Further studies are necessary to investigate the action of PAF in stroma–epithelium interactions. 

Our results also show that PAF-R gene expression is upregulated in corneal epithelium during wound healing, and further increases were found after prolonged PAF treatment. This suggests that the increase in PAF after injury ¹ could constitute an important feedback mechanism to maintain a PAF response in epithelial cells and consequently sustain the inflammatory process. Increases in receptor gene expression during wound healing were observed after in vivo and in vitro injury, and the results were confirmed using quantitative RT-PCR and Northern blot analysis. Previous in vitro studies have demonstrated upregulation of PAF-R gene expression after cells were stimulated with lipopolysaccharide (LPS). ³¹ We are aware of only one in vivo study showing that injection of LPS with PAF increases the level of rat PAF-R mRNA after intestinal injury. ¹⁹ Although the detailed mechanisms by which PAF upregulates the gene expression of its receptor are unknown, our experiments with AcD and CHX suggest that transcriptional regulation is required and is independent of de novo protein synthesis. One interesting result is that the upregulation of PAF-R gene transcription by PAF requires previous injury of the epithelial cells, indicating that during wound healing, there are other mediators to which the cells are exposed that are necessary for the responsiveness to PAF. Mediators such as PAF, cytokines, and growth factors sensitize, or prime, cells to respond in an enhanced manner to agonist stimuli. ¹⁴ Corneal injury produces the release of several growth factors that control the process of wound healing. ³² Our results using corneal organ culture suggest the activators of PAF-R gene expression must be factors intrinsic to the corneal cells and are not synthesized by tears or inflammatory cells that arrive at the cornea after injury. Therefore, we choose these growth factors as possible candidates that express receptors in the epithelial cells and that can be synthesized by epithelial or stromal cells. 

In this study, the three TGF-β isoforms had similar stimulatory effects on PAF-R expression. These results are in agreement with a previous report that demonstrated upregulation at the transcriptional level of PAF-R in B lymphoblastoma cells after exposure to 10 ng/ml TGF-β2. ³³ TGF-β1, -β2, and -β3 had been reported in corneal epithelium, as well as their receptors. ²⁵ They delay the rate of re-epithelialization, inhibit cell proliferation promoted by other growth factors, ³⁴ modulate differentiation processes, and stimulate the expression of extracellular proteins. ^{35 36} Another growth factor with autocrine–paracrine interactions is bFGF. This growth factor stimulates proliferation of corneal and limbal epithelial cells ³⁷ and enhances angiogenesis, ³⁸ but has no effect on PAF-R gene expression in corneal epithelial cells, suggesting that the induction of PAF-R is selective to certain growth factors. The paracrine growth factors HGF and KGF also upregulate PAF-R gene expression and accelerate corneal epithelial wound healing. ^{26 39 40} HGF also stimulates cell migration and inhibits differentiation. ³⁹  

Previous data have shown that PAF can amplify the action of inflammatory agents in corneal epithelial cells. In this study injury increased the expression of PAF-R mRNA and PAF further stimulated PAF-R gene transcription in corneal epithelial cells. This response occurred only after injury, and the results suggest that growth factors are likely to play a role in this modulation. However, further studies are needed to determine a more detailed function of the individual growth factors. The in vivo increase in PAF-R gene expression after injury could be a combination of the stimulatory (primary) effect of a brief stimulation with growth factors and the long-term exposure to PAF that could be released from the cells after injury. ¹ These increased levels of PAF, as occur in severe inflammatory conditions, may selectively enhance the response to PAF. 

 Supported by National Institutes of Health Grant 04928 and Core Grant EY02377.

 Submitted for publication September 1, 1999; revised November 22, 1999 and January 11, 2000; accepted January 26, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

PAF-R gene expression in rabbit corneal cells. The PAF-R, detected by RT-PCR, revealed a 478-bp receptor product. (A) Fresh corneal epithelium (F-RCE; two corneas per sample) was scraped from rabbits, and primary cultures of RCE, RLE, RCEn, and RCF cells were used to extract total RNA. CHO-c cells that do not express PAF-R were used as negative control. Other negative controls without reverse transcriptase and without RNA showed no signals (data not shown). M, 100-bp DNA ladder. The lower gel shows expression ofβ -actin, a housekeeping gene. The gels are representative of four similar experiments. (B) Fresh corneal keratocytes (F-RCK; four corneas per sample), primary (RCK), and first-passage (RCKp₁) cells were obtained. When the RCKp₁ cells were cultured in the presence of 10% FBS for 48 hours, there was loss of PAF-R gene expression. The lower gel shows that the expression of β-actin does not change under the same conditions. This experiment was performed three times with similar results each time.

Figure 2.

 

Corneal epithelial wound healing activated the expression of PAF-R mRNA. Rabbit corneas in vivo or incubated in organ culture were injured. Forty-eight hours after injury, the epithelium derived from three corneas per sample was collected and RNA isolated. C, control; W, wound. (A) RT-PCR amplification. The bars represent the mean ± SEM of the ratio between the expression of PAF-R and β-actin for four to six samples. *Significant differences from control (in vitro, P < 0.02; in vivo, P < 0.001; Student’s t-test). (B) Northern blot analysis. In this case, four corneas from four different rabbits per sample were used. The blot was initially probed with a rabbit PAF-R cDNA fragment that recognizes a single mRNA species of approximately 3.6 kb. The GAPDH was used as an internal control for each lane. The experiment was repeated once with similar results.

Figure 3.

 

PAF induced PAF-R mRNA in injured rabbit corneal epithelium. Injured corneas (four corneas per sample) were incubated for 48 hours in the presence of 120 nM cPAF. BN50727 (4 μM) was added to the corneas 30 minutes before adding PAF. (A) Electrophoresis gel of a representative quantitative RT-PCR amplifying the target mRNA as well as the internal standard (IS). (B) Concentration of the product was determined by comparison of the intensity of the band of the product with the IS. The values correspond to the average of three experiments. *Significant differences compared with control (P < 0.0001; Student’s t-test).

Figure 4.

 

Cyclohexamide (CHX) potentiated and actinomycin D (AcD) inhibited PAF-R mRNA expression by PAF in wounded corneas incubated in organ culture for 24 hours. Northern blot analysis shows the effects of 30 μm CHX and 10 μm AcD added to the wounded corneas 1 hour before adding 120 nM cPAF. Each sample was composed of six corneas. Experiment repeated once with similar results.

Figure 5.

 

Effect of growth factors on PAF-R gene expression in RCE cells. Cells were serum starved overnight and then incubated for 4 hours in the presence of (A) 120 nM cPAF or 20 ng/ml HGF or KGF, and (B) 1 to 10 ng/ml TGF-β1, -β2, or -β3 or 50 to 100 ng/ml bFGF. Equal amounts of total RNA (15 μg/lane) were separated by electrophoresis and hybridized with rabbit PAF-R and GAPDH cDNA probes by Northern blot analysis. Experiment repeated once with similar results.

The authors thank Alexey V. Ershov for his advice in quantitative RT-PCR and Laurie Varner for technical assistance.