Immortalized SV40 human CEC line (p4) was obtained from Dr. Hitoshi Watanabe (Osaka University, Osaka, Japan).33 Cells were cultured in Dulbeco's modified ragle's media (DMEM)/F12 (Invitrogen, Carlsbad, CA, USA) supplemented with human recombinant EGF (10 ng/mL) (Upstate Biotechnology, Lake Placid, NY, USA) and 10% fetal bovine serum. Cells were incubated in a CO₂-regulated incubator and the medium was renewed every two days.

Human conjunctival epithelial cells were given as a gift by Dr. Tae-im Kim.34 Human conjunctival tissues were donated by the YONSEI eye bank (Severance EYE & ENT Hospital, Seoul, Korea). Conjunctiva was dissected 2-mm lateral to the limbus of the cornea and incubated with 1.4 U dispase (Sigma-Aldrich Corp., St. Louis, MO, USA) for 1 hour in a 37℃ incubator. Conjunctival epithelial cells were then removed with a cell scraper and seeded on 60 mm plate. Cells were cultured in KGM-Gold medium (Lonza, Walkersville, MD, USA). The culture medium was changed every other day until the cells reached 60% to 70% confluence (~5–6 days). After reaching confluence, epithelial cells (1×10⁵ cells) were used to analyze gene expression.

Human corneal tissue was harvested from healthy corneas from the eye bank after penetrating or lamellar keratoplasty. Donor confidentiality was maintained in accordance with the Declaration of Helsinki and was approved by the Severance Hospital IRB Committee (CR04124), Yonsei University.

Corneas (described below) and epithelial tissue were collected by scraping and then stored at -80℃ until RNA extraction and analysis of mucin expression by quantitative reverse transcription PCR (qRT-PCR).

Primary CECs were isolated from corneas using a method described by Maldonado and Furcht.35 Briefly, the cornea was placed in a 35 mm plate containing dispase II (1.2 U/mL) and incubated for 3-hour at 37℃. The epithelial layer was gently scraped free with a no. 15 scalpel blade. The cell suspension was transferred to a 50-mL tube and centrifuged at 1000×g for 4 minutes. Collected cells were transferred to a flask coated with a mixture of fibronectin and collagen (Sigma-Aldrich Corp.), containing 5 mL of serum-free CnT-20 medium.

Cell migration was assayed using the Transwell system (Corning Costar, Acton, MA, USA) with 6.5 mm-diameter polycarbonate filters (8-µm pore size). Briefly, the lower surface of the filter was coated with 10 µg/mL fibronectin (Sigma-Aldrich Corp.), 10 µg/mL recombinant human TGFBIp (rhTGFBIp) (Sino Biological Inc., Beijing, China), or bovine serum albumin at a concentration of 3% (w/v) as a control for nonspecific binding. SV40-CECs or primary CECs (10⁵) were seeded onto chemotaxis filters in DMEM plus 1% fetal bovine serum. After the 5-hour migration period, non-migrating cells were removed and migrating cells were stained with hematoxylin and eosin (H&E) and quantified using Kodak 1D software (Eastman Kodak, Rochester, NY, USA). Results are representative of three different experiments in duplicate.

The wound-healing assay was performed by scratching confluent SV40-CECs or primary CECs on 35-mm dishes with micropipette tips. The cells were then washed to remove debris and treated with rhTGFBIp (10 µg/mL) for 16 hours. Images were captured at 0 and 16 hours after wounding. For quantitative analysis, five fields per plate were photographed, and distances between the front lines were measured using ImageJ software (National Institutes of Health). Each assay was repeated three times.

Cell-matrix adhesion assays were performed as described previously.36 The 96-well plates were coated with 10 µg/mL human fibronectin (Sigma-Aldrich Corp.) or 2.5–10 µg/mL rhTGFBIp. SV40-CECs or primary CECs in adhesion buffer (serum-free media: DMEM or CnT-20) were seeded at 10⁵ cells/well in 100 µL volume and incubated for 30 minutes at 37℃. After two washes to remove non-adherent cells, adherent cells were stained with H&E and quantified in triplicate by counting adherent cells in five randomly selected fields per well (Axiovert 100; Carl Zeiss Micro-Imaging, Thornwood, NY, USA). Results are representative of three different experiments in duplicate.

CEC proliferation was assessed using MTT assays (3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide) according to the manufacturer's protocol. Briefly, SV40-CECs or primary CECs were seeded onto 96-well plates and incubated with different concentrations of rhTGFBIp (2.5–10 µg/mL) for 72 hours. Fifteen microliters of MTT solution and 200 µL dimethyl sulfoxide were successively added to each well and the optical density measurement (490 nm) was performed. All groups of experiments were performed in quintuplicate.

SV40-CECs, primary CECs or conjunctival epithelial cells (2×10⁵ cells/well) were seeded onto 60 mm plates and treated with TGFBIp (10 µg/mL) for 24 hours. Total RNA was isolated from cells by extraction in TRIZOL reagent (Invitrogen) at each time point (3, 6, 9, 12, and 24 hours). SV40-CECs (2×10⁵ cells/well) were seeded onto 60 mm plates, After 24 hours, cells were pre-incubated for 40 minutes with or without PD98059 (10 µM) or Wortmannin (100 nM), then stimulated with TGFBIp (10 µg/mL) for 6 hours. Total RNA was isolated from cells by extraction in TRIZOL reagent (Invitrogen). Using the Power SYBR Green RNA-to-CT™ 1-Step kit (Applied Biosystems, Foster City, CA, USA) and StepOnePlus™ (Applied Biosystems), mRNA expression of the human glyceraldehyde 3-phosphate dehydrogenase (GAPDH), MUCIN1, MUCIN4, MUCIN5AC, MUCIN7, and MUCIN16 genes was measured according to the manufacturer's instructions. The qRT-PCR conditions for all genes were as follows: 48℃ for 30 minutes, 95℃ for 10 minutes, then 40 cycles of 95℃ for 15 seconds, and 60℃ for 1 minute. The results are based on cycle threshold (Ct) values. We calculated differences between the Ct values for experimental and reference genes (GAPDH) and graphed the results as the ratio of each RNA to the calibrator sample. Primers used for gene amplification were the following: MUCIN1 5′-AGCGTGAGTGATGTGCCATT-3′ (sense) and 5′-AGCGCA ACCAGAACACAGAC-3′ (antisense); MUCIN4 5′-GGTGGTG GAGGCGTTCTTAT-3′ (sense) and 5′-CTCACGTTCAGGGCT GTCAC-3′ (antisense); MUCIN5AC 5′-CGCTCAGCTGT TCTCTGGAC-3′ (sense) and 5′-GCACAGGTCGACTGGTTCT G-3′ (antisense); MUCIN7 5′-TGCCCCAATTACCACACCTA-3′ (sense) and 5′-TATTTTGGCCAGGAGCTGAA-3′ (antisense); MUCIN16 5′-CCAACTCTTCCGAAACAGCA-3′ (sense) and 5′-GCCAGTGGCGAGAAGTTACA-3′ (antisense); GAPDH 5′-A TGGGGAAGGTGAAGGTCG-3′ (sense), and 5′-GGGGTCATT GATGGCAACAATA-3′ (antisense). Three independent experiments were performed and statistical analysis was carried out using Newman-Keuls multiple comparison tests.

SV40-CECs (2×10⁵ cells/well) were seeded onto 60 mm plates. After 24 hours, cells were pre-incubated for 40 minutes with or without PD98059 (10 µM) or Wortmannin (100 nM), and then stimulated with TGFBIp (10 µg/mL) for 8 hours. Growth medium was then removed, and the cells were rinsed twice with phosphate buffered saline prior to lysis with a radioimmunoprecipitation assay (RIPA) buffer. Cell lysates were electrophoresed on SDS-polyacrylamide gel electrophoresis (PAGE) and proteins were transferred onto polyvinylidene difluoride membranes. The blocked membranes were incubated with the appropriate antibody [anti-human MUCIN1, MUCIN4, or ACTIN (1:1000 dilution, Abcam Inc., Cambridge, MA, USA)], and the immunoreactive bands were visualized with a chemiluminescent reagent, as recommended by Amersham Biosciences, Inc. (Piscataway, NJ, USA).

SV40-CECs (2×10⁵ cells/well) were seeded onto 60 mm plates. After 24 hours, cells were treated with TGFBIp (10 µg/mL) for the indicated period of time (5, 10, 15, 30, and 60 minutes). Cells were harvested with RIPA buffer and cell lysates were electrophoresed on SDS-PAGE and proteins were transferred onto polyvinylidene difluoride membranes. The blocked membranes were incubated with the appropriate antibody [anti-human pFAK, pSRC, pERK, pAKT, FAK, SRC, ERK, and AKT (1:1000 dilution, Cell Signaling Technology Inc., Danvers, MA, USA)], and the immunoreactive bands were visualized with an enhanced chemiluminescence immunoblotting system (GE Healthcare, Buckinghamshire, UK).

Data are presented as mean±standard error and statistical comparisons between groups were performed by one-way ANOVA followed by Tukey's test or Newman-Keuls multiple-comparison test when appropriate. All experiments were repeated at least three times.

To assess the effect of TGFBIp on CEC activity, we first examined the effect of TGFBIp on cell migration of CECs. As shown in Fig. 1A and B, TGFBIp increased CEC migration in a dose-dependent manner. The magnitude of TGFBIp-induced CEC migration was higher than that induced by fibronectin, which is a known CEC migration-inducing factor (Fig. 1B).

Next, we analyzed the effect of TGFBIp on CEC adhesion and found that TGFBIp increased SV40-CEC adhesion in a dose-dependent manner (Fig. 1C and D). In addition, CECs stimulated with TGFBIp exhibited enhanced proliferation activity compared with the activity of unstimulated cells (Fig. 1E). This increased activity was replicated in the wound-healing assay, which showed that SV40-CECs migrated more into the scratch wound after TGFBIp treatment (Fig. 1F and G). We cultured primary CECs from human corneal epithelial tissue and performed migration, adhesion, proliferation, and wound healing assays. As shown in Fig. 2, TGFBIp increased primary CEC migration, adhesion, proliferation, and wound healing in a dose-dependent manner. Taken together, these results suggest that TGFBIp promotes migration, adhesion, and proliferation of CECs, which may enhance wound healing in these cells.

Next, we investigated whether TGFBIp promotes the expression of mucin genes in the corneal and conjunctival epithelial cells. As shown in Figs. 3A and 4A, TGFBIp treatment increased the expression of MUC1, -4, -5AC, and -16 in SV40-CECs and primary CECs, and maximal expression was observed after 6 hours, whereas the mRNA levels of MUC7 were unchanged. Furthermore, TGFBIp also increased the expression of MUC1, -4, -5AC, -7, and -16 genes in conjunctival epithelial cells (Supplementary Fig. 1A, only online). Basal expression levels of MUC16 among the mucin genes were highest in SV40 immortalized CECs, primary cultured human CECs, human corneal epithelial tissue, and conjunctival epithelial cells (Figs. 3B and 4B) (Supplementary Figs. 1B and 2, only online). These observations provide evidence that TGFBIp may have a protective effect on ocular surfaces by inducing the expression of mucin genes in corneal and conjunctival epithelial cells.

To determine which signaling pathways are involved in modulating expression of mucin genes in TGFBIp-stimulated SV40-CECs, we analyzed the activity of intracellular signaling molecules. As shown in Fig. 5A and B, phosphorylation of ERK and AKT signaling molecules was increased by TGFBIp stimulation in a time-dependent manner, whereas TGFBIp had no effect on FAK or SRC phosphorylation (Fig. 5A). To further investigate the significance of the ERK and AKT signaling pathway in regulating mucin expression in TGFBIp-stimulated CECs, CECs were pre-treated with PD98059, an ERK inhibitor, or Wortmannin, an AKT inhibitor, and then stimulated with TGFBIp. As shown in Fig. 5C, TGFBIp-stimulated CECs showed increased expression of MUC1, -4, and -16 mRNA, and this effect was effectively blocked by PD98059 or Wortmannin. Western blot analysis confirmed the inhibition of TGFBIp-induced ERK and AKT phosphorylation and of TGFBIp-induced MUC1 and -4 expression by PD98059 or Wortmannin (Fig. 5D, E, and F). These results suggest that ERK and AKT signaling are important in regulating the expression of mucin genes in TGFBIp-treated CECs.