Current Eye Research, Early Online, 1–10, 2014 ! Informa Healthcare USA, Inc. ISSN: 0271-3683 print / 1460-2202 online DOI: 10.3109/02713683.2014.964417

ORIGINAL ARTICLE

Different Effects of Thrombin on VEGF Secretion, Proliferation, and Permeability in Polarized and Non-polarized Retinal Pigment Epithelial Cells Hiroto Terasaki1, Makoto Shirasawa1, Hiroki Otsuka1, Takehiro Yamashita1, Eisuke Uchino1, Toshio Hisatomi2, Shozo Sonoda1 and Taiji Sakamoto1 Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

1

Department of Ophthalmology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan and 2Department of Ophthalmology, Kyushu University, Fukuoka, Japan

ABSTRACT We investigated the effect of thrombin on the secretion of vascular endothelial growth factor (VEGF), on cellular proliferation, and on the integrity of the barrier function of polarized retinal pigment epithelial (RPE) cells. In addition, we compared the responses of polarized to that of non-polarized RPE cells. Porcine polarized RPE cells were established using Transwell membranes. The polarization of the RPE cells was determined by their high transepithelial electrical resistance (TER4200 cm2) and by their differential secretion of VEGF (basal direction 4apical direction by 2.5). RPE cells were incubated with thrombin (5–20 U/ml) for 24 h. The concentration of VEGF in the culture medium was measured by enzyme-linked immunosorbent assay, and the TER was measured. Cellular proliferation was assessed by Ki-67 immunostaining. The area of laser-induced choroidal naovascularization (CNV) was measured in rat eyes and compare to that of controls with or without thrombin. Our results showed that thrombin significantly increased VEGF secretion both in polarized and non-polarized RPE cells in a dose-dependent way. Thrombin did not significantly affect the TER or the expression of tight-junctional proteins in polarized RPE cells, but decreased it in non-polarized RPE cells by inducing intercellular gaps. Ki-67-positive cells were observed in non-polarized RPE cells but not in polarized RPE cells as controls. After thrombin exposure, the number of Ki-67-positive cells increased significantly in non-polarized RPE cells but not in polarized RPE cells. The area of CNV was larger in thrombin-injected eye than control eyes. Although thrombin increased VEGF secretion regardless of cell polarity, its effects on proliferation and barrier integrity were dependent upon cell polarity. Cell polarization is an important factor for determining the response of RPE cells to thrombin, and the different responsive patterns to thrombin upon cell polarity might explain the complicated pathology of such diseases as age-related macular degeneration. Keywords: Barrier function, cell polarity, RPE, thrombin, VEGF

INTRODUCTION

seen in ocular diseases such as age-related macular degeneration (AMD), and damaged RPE cells lead to a breakdown of the blood–retinal barrier. The breakdown is followed by angiogenesis and scar formation which progresses to irreversible vision loss. During this process, subretinal hemorrhages or sub-RPE hemorrhages are the most damaging complications.5,6 Thrombin is a key enzyme in the coagulation cascade.7 Catalytically activated thrombin converts fibrinogen to fibrin and is thus often found at the site of

Under physiological conditions, the retinal pigment epithelial (RPE) cells are quiescent and form a monolayered, sheet-like structure with tight junctions between the RPE cells. This forms a barrier between the choroid and the retina.1,2 In addition, RPE cells secrete vascular endothelial growth factor (VEGF) which is essential for the maintenance of the retina and choroid.3,4 A dysfunction of the RPE cells is often

Received 17 August 2013; revised 30 June 2014; accepted 7 September 2014; published online 8 October 2014 Correspondence: Taiji Sakamoto, Department of Ophthalmology, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1, Sakuragaoka, Kagoshima 890-8520, Japan. Tel: 81-99-275-5402. Fax: 81-99-265-4894. E-mail: [email protected]

1

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

2

H. Terasaki et al.

hemorrhages. High-resolution optical coherence tomography has shown that fibrin is a common component of the subretinal space in such disease as diabetic macular edema, AMD and serous retinal detachment.8 Thus, it may be possible that thrombin is not only associated with the hemorrhage, but also plays other roles in different ocular diseases. In addition, thrombin can have different effects on various cell types such as cellular proliferation, release of growth factors and disruptions of the barrier function.9–11 Thrombin has also been reported to stimulate cultured RPE cells to proliferate, intercellular gap junctions to develop and RPE cells to up-regulate their secretion of VEGF.12–14 However, these data are not compatible with clinical findings. We recently developed a new RPE cell culture system that produced RPE cells with properties of RPE cells in situ, namely, they secreted VEGF asymmetrically and have functional tight junctions.15,16 We called these cells polarized cells and we showed that their reaction pattern to inflammatory cytokines differed significantly from non-polarized RPE cells.17 Most studies on RPE cells have been performed on RPE cells in culture, and the results have contributed to the understanding of how RPE cells are involved in the pathophysiology of retinochoroidal diseases. However, it is not easy to interpret these in vitro date because RPE cells are very plastic, and their properties, for example, polarization and differentiation, change easily depending on the culture conditions.18,19 Thus, the results obtained from studies of cultured RPE cells that are not polarized might not necessarily represent the results obtained from RPE cells in situ. Above all, it is necessary to know the differences between non-polarized and polarized RPE cells to clarify this issue. Thus, the purpose of this study was to compare the effects of thrombin on polarized RPE cells to that in non-polarized RPE cells.

METHODS RPE Cell Cultures Eyes of 5- to 6-month-old pigs were obtained from a local abbatoir, and the RPE cells were isolated as described in detail.17,20 The culture medium was alpha modified Eagle’s medium with 2 mM L-glutamine, 100 U/mL penicillin, 100 mg/mL streptomycin (Sigma-Aldrich, St Louis, MO) and 10% fetal bovine serum (FBS; Omega, Tarzana, CA) at 37  C under 5% CO2. The FBS was used for one week. The anterior segment of the eye was resected leaving an eye cup. Then, RPE cells were separated from Bruch’s membrane with dispase (GIBCO, Grand Island, NY). The RPE cells were seeded into T75 flasks (Falcon Plastics, Los Angeles, CA) with the above medium. After reaching confluence, RPE cells were

used for the following experiments. Therefore, the cells used in the following experiments were first passage cells. To prepare polarized RPE cells, approximately 1.0  105 RPE cells/well were seeded onto fibronectin-coated Transwell filters (12 mm internal diameter; 0.4 mm pore size; Transwell; Corning Costar, Corning, NY), and the cells were cultured in RPE maintenance medium containing 10% FBS for 2 days and in 2% FBS thereafter for more than 2 weeks. We determined that the RPE cells were non-polarized RPE cells on three days even though these cells were confluent, that is, the cells fully covered the filter and there were no intercellular spaces.

Measurement of Transepithelial Electrical Resistance The transepithelial electrical resistance (TER) of the RPE cells on the Transwells filters was measured with the Epithelial Voltohmmeter 2 (EVOM2; World Precision Instrument, Sarasota, FL) as described in detail.16,17

Enzyme-Linked Immunosorbent Assay The culture medium of the RPE cell was switched to FBS-free medium overnight to minimize the effect of the serum on the different agents. After stimulation with thrombin which was added to both the upper and lower chambers (5–20 U/ml; Calbiochem, San Diego, CA) for 24 h, the medium was collected and stored at 80  C until analyses by enzyme-linked immunosorbent assay. The concentration of VEGF-A in the upper and lower chambers was measured separately according to the manufacturer’s protocol (Quantikine; R&D Systems, Minneapolis, MN).

Terminal dUTP Nick-End Labeling Staining RPE cells were stained by the terminal dUTP nick-end labeling (TUNEL) method, and the number of TUNELpositive cells was determined by the ApopTag fluorescein direct in situ apoptosis detection kit (Chemicon International, Temecula, CA) as described by the manufacturer.17 The number of TUNEL-positive cells in 10 randomly selected microscopic fields (200) was counted in a masked fashion. H2O2 at a concentration of 250 mM was used to kill RPE cells for a positive control.

Proliferation Assay To differentiate between the effects of thrombin on cellular proliferation and cellular cycling, the cells Current Eye Research

Thrombin and Polarized RPE Cells were exposed to 20 U/ml thrombin for 6 h, and they were immunostained with anti-ki-67 antibody (Abcam Japan, Tokyo, Japan) as was done in our previously described method.15 Ki-67 is highly expressed in dividing cells,15 and the number of Ki-67-positive cells was counted in microscopic fields in a masked fashion.

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

Western Blot Analyses Western blotting was performed as we described in detail.17,20 Briefly, cell lysates were placed under reducing conditions on a 10% Tris-HCl gel (MiniProtein Tris-Glycine extended precast gels; Bio-Rad, Hercules, CA) and then transferred to PVDF membranes (Thermo Scientific, Rockford, IL). The primary antibody was then applied. After incubation with horseradish peroxidase-conjugated secondary antibody, the protein bands were made visible by ECL plus chemifluorescent reagent (GE Healthcare, Piscataway, NJ).

Immunohistochemistry and Transmission Electron Microscopy The presence of ZO-1 and F-actin, a tight junctionassociated molecule,16,17 was determined immunohistochemically in RPE cells as described in detail.16,17 The RPE cells were first treated with phosphatebuffered saline (PBS) containing 0.2% Triton X for 30 min followed by fixation in ice cold methanol for 15 min at 4  C. The specimens were blocked in 5% BSA before incubating with each of the primary antibodies; 1:100 ZO-1 (Invitrogen, Carlsbad, CA) and 1:200 Alexa fluor 594 phalloidin (Invitrogen) for F-actin. An anti-rabbit secondary antibody (Alexa Fluor 488; Molecular Probes, Eugene, OR) was used for 30 min in the dark at room temperature. After nuclear staining with DAPI (Vector Labs, Burlingame, CA), the slides with the cells were examined with a fluorescence microscope (Olympus, Tokyo, Japan). For transmission electron microscopy (TEM), the RPE cells were fixed in half-strength Karnovsky’s fixative for 24 h at 4  C. The fixed cells were processed as described in detail.16,17 Sections were examined with a JEOL JEM2100 (Peabody, MA) electron microscope and photographed with an Orius SC1000B Gatan (Pleasanton, CA) digital camera.

Definition of Polarization and Differentiation Because there is no clear definition of ‘‘polarization’’ and ‘‘differentiation’’ for RPE cells in vitro, we used !

2014 Informa Healthcare USA, Inc.

3

the following definition. ‘‘Polarized cells’’ were those cells with high TER (200 V cm2) or cells with a basal/apical secretion of VEGF ratio of 2.5. ‘‘Differentiated cells’’ were those cells showing a positive reaction for RPE65 protein, but a negative reaction for a-smooth muscle actin (a-SMA) protein in western blot assays. ‘‘Well-differentiated cells’’ were cells that met all of these criteria.15–17,20 The classification of the RPE cells was based on the direct measurement of the TER or basal/apical secretion of VEGF ratio or parallel experiments on the same batch of RPE cells.

Animal Experiments All animals were treated in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. The protocol was approved by the Committee on the Ethics of Animal Experiments of the Kagoshima University. A rat choroidal naovascularization (CNV) model was made according to our described methods.21 Briefly, rats were anesthetized with a 0.2 ml of a mixture of 100 mg/ml ketamine and 20 mg/ml xylazine. Pupils were dilated with a topical 5.0% phenylephrine and 0.8% tropicamide. CNV was experimentally induced by photocoagulation with an argon dye pulsed laser (Novus Varia, Lumenis, Salt Lake City, UT). The spot size was 100 mm, duration was 0.05 s, and intensity was 200 mW. Four laser photocoagulations were created in each eye between the major retinal vessels around the optic disk. On the same day, the rats were given an intravitreal injection of 2 U/ml of thrombin or 2 ml of PBS alone. After 7 days, the rats were anesthetized and perfused with 1 ml PBS containing 50 mg/ml fluorescein-labeled dextran (Sigma Aldrich). After the eyes were enucleated and briefly fixed in 4% PFA, the anterior segment was removed and the retina was carefully dissected from the eyecup. The retina was flat-mounted and was viewed with an Olympus fluorescence microscope. Images were photographed for each section with a CCD camera, and the size of CNV lesions was measured using the ImageJ software (National Institutes of Health, Bethesda, MD).

Statistical Analyses All values are expressed as the means ± standard error of the means. Student’s t-tests were used to determine whether differences were significant. Differences were considered significant at p50.05. Statistical analyses were performed using SPSS 19.0 (Chicago, IL).

4

H. Terasaki et al.

RESULTS

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

Morphological Characteristics of Polarized and Non-Polarized RPE Cells Fourteen days after seeding the RPE cells on Transwell filters, the cells were well pigmented and were arranged in a hexagonal mosaic. TEM showed that the nuclei of the RPE cells were basally located, and the pigment granules were congregated close to the apical membrane as in RPE cells in situ. The cells also had well-defined tight-junctional complexes (Figure 1A–B). RPE65, a marker for differentiated RPE cells, was also detected by western blot analysis on day 14 (Figure 1E). Thus, our morphological and western blot analyses showed that RPE cells grown on Transwell filters were polarized on day 14.

On the other hand, the RPE cells seeded on Transwell filters for only three days did not have the classic honeycomb pattern and were larger than the polarized RPE cells (Figure 1C). Microvilli and tightjunctional complexes were not found in these cells (Figure 1D). The expression of RPE65 was weak and a-SMA was expressed in these cells. These findings indicated that the RPE cells on this day were not differentiated (Figure 1E). The TER of the RPE cells on Transwell filters was measured on day 3 and weekly for 4 weeks thereafter. The TER gradual increased from 179.3 ± 4.8 V cm2 on day 3 to 345.5 ± 10.4 V cm2 on day 14. The TER reached a plateau at one month at 267 ± 5.6 V cm2 (Figure 1F). These findings indicate that the RPE cells grown on Transwell filters acquired a barrier function sometime between day 3 and day 14. Thereafter, RPE cells cultured on Transwell filters for 3 days were used as non-polarized RPE, and those grown for 14 days were used as polarized RPE cells for the following studies. The number of RPE cells on day 3 was 2.43 ± 0.15  105 cells per well which is enough to consider the cells as being confluent (supplementary Figure 1A).

TUNEL Assay Next, we determined whether thrombin was cytotoxic to polarized RPE cells by TUNEL staining. The number of TUNEL-positive polarized RPE cells/field after exposure to 50 U/ml of thrombin did not differ significantly from that of controls (Figure 2). Thus, 5–20 ng/ml of thrombin was used in the following experiments to minimize the effect of cellular toxicity.

VEGF Up-Regulation by Thrombin in Polarized and Non-Polarized RPE

FIGURE 1. Morphological characteristics of polarized and nonpolarized RPE cells. (A) and (B) Polarized RPE cells. RPE cells are well pigmented (arrows) and are arranged in a hexagonal mosaic (A). The bright spots are pores of the Transwell filter. TEM shows that the RPE cells have basally located nuclei and contain pigment granules that congregate close to the apical membrane. Well-defined tight-junctional complexes are present between adjacent cells. Foot processes are present on the basal side and numerous microvilli are present on apical surface (B). (C) and (D) Non-polarized RPE cells. Cells appear spindle- to cobblestone-shaped (C). TEM shows flat-shaped cells with short microvilli (D). (E) RPE65 is detected by western blot analysis in differentiated RPE cells on day 14 but not in nondifferentiated cells on day 3. a-SMA is detected in nondifferentiated RPE cells on day 3 but not in differentiated RPE cells on day 14. (F) TER gradually increases over time and reaches a plateau at one month. Original magnifications: (A) and (C) 100; (B) and (D) 1000.

In polarized RPE cells, a larger amount of VEGF was secreted on the basolateral than apical side as reported (Figure 3A).4,15,16 On the other hand, the secretion of VEGF from non-polarized RPE was not significantly different on the two sides (Figure 3B). Thrombin significantly up-regulated VEGF secretion in a dose-dependent manner in both polarized and non-polarized RPE cells. Hirudin, a thrombin inhibitor, blocked the thrombininduced VEGF up-regulation suggesting that the increase in VEGF was thrombin specific.

Thrombin Promotes Cellular Proliferation in Non-Polarized RPE but Not in Polarized RPE Next, we investigated the effects of thrombin on cell proliferation in non-polarized and polarized RPE Current Eye Research

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

Thrombin and Polarized RPE Cells

5

FIGURE 2. TUNEL assays. (A) and (B) Number of TUNEL-positive cells/field after exposure to 20–50 U/ml of thrombin is not significantly different from that in controls. H2O2 was used to produce dead RPE cells as positive controls (p50.01). Original magnification: (a) 200 **p50.01, Student’s t-tests.

cells. To determine whether the cultured RPE cells were proliferating, we used Ki-67, a nuclear protein that is a marker for cellular proliferation. In controls, Ki-67-positive cells were observed among the nonpolarized RPE cells but were not detected among polarized RPE cells (Figure 4A). Thus, non-polarized RPE cells were in active cell cycling but polarized RPE cells were quiescent similar to RPE cells in situ. Moreover, Ki-67-positive cells were observed in confluent ARPE-19 and porcine RPE cell without asymmetrical VEGF secretion (supplementary Figure 1B). After exposure to thrombin, the number of Ki67-positive cells was significantly increased in nonpolarized RPE (p50.01, Figure 4B). However, this increase was not observed in polarized RPE cells even with thrombin stimulation.

Thrombin Increases Permeability of Non-Polarized but Not Polarized RPE Cells

FIGURE 3. VEGF up-regulation by thrombin in polarized and non-polarized RPE cells. Secretion of VEGF from non-polarized RPE was not assymetrical (A). On the other hand, polarized RPE cells secreted larger amounts of VEGF on the basolateral side (B). Thrombin up-regulated VEGF secretion significantly in a dose-dependent manner in both polarized and non-polarized RPE cells (A,B). **p50.01, Student’s t-tests. !

2014 Informa Healthcare USA, Inc.

Earlier studies showed that the thrombin-induced intercellular gap junction formation affected the distribution of actin filaments in RPE cells.14,22 Therefore, we investigated whether thrombin affected the proteins associated with tight junctions and barrier function in both types of cells. Immunofluorescent studies showed that ZO-1 and F-actin were strongly positive at the borders of polarized RPE cells. In non-polarized RPE cells, the expressions of ZO-1 and F-actin were relatively weak and diffusely expressed at the cell

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

6

H. Terasaki et al.

FIGURE 4. Thrombin promotes cellular proliferation in non-polarized RPE but not in polarized RPE. In controls, Ki-67-positive cells were observed in non-polarized RPE cells but not in polarized RPE (A). After exposure to thrombin, Ki-67-positive cells were significantly increased in non-polarized RPE (p50.01), but not in polarized RPE cells even with thrombin stimulation (B). The effect of thrombin on cellular proliferation differed depending on cell polarity. **p50.01, Student’s t-tests. Original magnification: (a) 200.

borders (Figure 5A). Although thrombin did not affect the expressions in polarized RPE, the expression of F-actin had a diffuse pattern and the expression of ZO-1 was displaced in non-polarized RPE (Figure 5A). In addition, the TER of polarized RPE was not changed by 1–20 U/ml of thrombin, however the TER of non-polarized RPE was significantly decreased by 20 U/ml thrombin (p50.05; Figure 5B–C).

Intravitreal Injection of Thrombin Exacerbates Laser-Induced CNV in Rats Argon laser photocoagulation was applied to rat retinas, and on day 7 the area of the CNVs in the thrombin-injected eye was more than 2 times larger than that of non-thrombin-injected controls (p50.01, Figure 6A–B). However, these changes were limited to the laser-treated areas. There was no pathologic inflammation or degeneration of RPE cells in other areas (Figure 6C). These results suggest that thrombin exacerbated the progression of a CNV.

DISCUSSION Our results showed that the effect of thrombin on RPE cells differed depending on whether the cells

were or were not polarized. Thrombin increased VEGF secretion but did not induce cellular proliferation or barrier disruption in polarized RPE. In nonpolarized RPE cells, on the other hand, thrombin also increased VEGF secretion but also induced cell proliferation and disrupted barrier function. The reaction patterns of non-polarized RPE cells are consistent with that in earlier studies did not use polarized RPE cells.10,12–14

Altered Reactions Dependent on Cell Polarization The effect of thrombin on cellular proliferation and the barrier function of the RPE cells differed depending on the polarization. Both our results and previous results showed that polarized RPE cells with no stimulation were in the G0 phase with Ki-67-negative and p27-positive cells.15 On the other hand, nonpolarized RPE cells were in the interphase or cell division phase with Ki-67-positive and p27-negative cells. As thrombin induces cyclin D1, an important factor for the G1 to S transition, the proliferation of non-polarized RPE cells could be easily induced by thrombin.13 However, other factors such as the activation of the Rb family or activation of cyclin C are required for the transition from the G0 phase to the Current Eye Research

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

Thrombin and Polarized RPE Cells

7

FIGURE 5. Thrombin disrupts barrier property in non-polarized RPE but not in polarized RPE. Immunohistochemistry shows that ZO-1 and F-actin are strongly expressed and outline the polarized RPE cells. In contrast, the expression of ZO-1 was relatively weak and F-actin was diffusely expressed in non-polarized RPE cells (A). Although thrombin did not affect these expressions of ZO-1 and Factin in polarized RPE, it stimulated the expression of F-actin and displaced the expression of ZO-1 in non-polarized RPE. In addition, the TER of polarized RPE is not changed by thrombin (5–20 U/ml), but the TER of non-polarized RPE is significantly decreased by 20 U/ml thrombin ((B)–(C) *p50.05, Student’s t-tests.). Original magnification: (A) 200.

proliferating phase.23 These might cause the different response patterns of proliferation after exposure to thrombin. The barrier function of the RPE cells also depended on cellular polarization. TEM showed that there were in vivo-like mature tight junctions in polarized RPE cells. Furthermore, immunofluorescence showed strong expression of ZO-1 and F-actin, both components of tight-junction complexes, at the cell borders in polarized RPE cells in contrast to non-polarized RPE cells with ZO-1 being expressed weakly or not present. Although the mechanism for these differences is unclear, there are several reports on thrombin and barrier function using vascular endothelial cells or epithelial cells. In both cell types, F-actin filaments were displaced to the cell–cell contact zone with maturation of the tight junctions.24 ZO-1 was also accumulated in the cell borders with maturation.25,26 In endothelial cells, thrombin causes central stress formation and TER reduction associated with a dispersion of ZO-1.27,28 These findings were also reported for non-polarized RPE cells in vitro.14,22 Thus, non-polarized RPE cells might have characteristics similar to other endothelial cells, where both gap junctions and tight junctions are enlarged by !

2014 Informa Healthcare USA, Inc.

thrombin. However, there are several contradictory reports on the response of epithelial cells. Thrombin stimulated the expression of ZO-1 and increased the barrier function in lung epithelial cells.27 On the other hand, thrombin impaired the barrier function of human airway cell lines.29 It is likely that the reaction patterns may be different with the degree of maturation and with the cell type. Additionally, cell lines occasionally lose their original characteristics. From the findings of F-actin and ZO-1, polarized RPE cells may have characteristics of mature epithelial cells which might account for the different reaction patterns to thrombin.

Unaltered Reactions Dependent of Cell Polarization Thrombin stimulated the expression of VEGF in RPE cells, and this effect was not different for polarized and non-polarized RPE cells. In earlier studies, VEGF was reported to be up-regulated by thrombin in only non-polarized RPE cells.12 The signal transduction mechanisms of thrombin-induced VEGF expression are not mediated through a simple or single pathway,

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

8

H. Terasaki et al. results in a further disruption of RPE barrier and growth of the CNV membrane. Our findings have clinical implications. In the healthy and polarized conditions, RPE cells are resistant to stimulants such as thrombin. Thus, CNV is not a frequent complication in eyes with uveitis or acute central serous chorioretinopathy. However, in the unhealthy or non-polarized condition, such as with aging, substantial number of RPE cells would lose their polarization. As a result, the CNV progresses aggressively after hemorrhage/fibrin formation in exudative AMD, and they are more prominent at the area of damaged or aging RPE cell zones. Additionally, the presence of subretinal fibrin might be a warning sign of degeneration in AMD because thrombin is there. Fibrin itself can be a scaffold for cell proliferation, fibrosis and angiogenesis. Thus, the pathological role of thrombin might be more profound than is supposed.

Limitations FIGURE 6. Intravitreal injection of thrombin exacerbates laserinduced CNV in rat. The area of CNV in thrombin-injected eye is more than two times larger than that of controls ((A) (B) *p50.05, Student’s t-tests). However, these changes were highly limited to the laser-treated area. There is no pathologic inflammation or degeneration of RPE cells in other areas (C). Original magnifications: (B) 200; (C) 100.

but through complex and mixed pathways.12,30 For example, there are several input sequences affecting thrombin-induced VEGF secretion, such as MAPKJNK, ERK, p38, PKC and NF-kB interacting with each other. Additionally, the secondary inputs of transforming growth factor-beta are up-regulated by thrombin and also stimulate VEGF expression.30 On the other hand, the VEGF secretion stimulated by TNF-a is significantly altered in polarized RPE cells probably because the TNF-a-mediated VEGF secretion is mediated through comparatively simple pathways; the JNK, NF-kB and p38 MAPK pathways.31 This might be related to the consistent reaction pattern of TNF-a regardless of cellular polarization.

In vivo Effects In our rat CNV model, thrombin exacerbated the laser-induced CNV formation, but did not cause any apparent adverse or degenerative effect on the areas with no laser burn. In normal or quiescent RPE cells, thrombin does not disrupt the barrier or increase proliferation, while in the laser-burned areas, RPE cells were already damaged and supposedly depolarized for wound healing. These non-polarized RPE cells should be more responsive to thrombin, which

We used RPE cells on day 3 as non-polarized RPE cells and those on day 14 as polarized RPE cells. Although the upper well was fully covered by RPE cells, that is, confluent, on day 3, proliferating cells were detected by Ki-67 (Figure 4). Therefore, there may be some question whether the cells are confluent on day 3. However, even in RPE cells on day 14 with definite confluence, cell proliferation was detected in a small number of dishes without asymmetrical secretion of VEGF, which is usually not detected in polarized RPE cells with asymmetrical secretion (supplementary Figure 1). It is possible that the appearance of confluency (a culture well is fully covered by cells) is not necessarily a sign of maturation. Thus, the present results are likely to show the differences between non-polarized and polarized RPE cells, rather than between confluent and non-confluent cells. Of note, cellular proliferating activity is supposedly negatively correlated with polarization, but care must be taken to interpret this result. The present results show that polarized RPE cells and nonpolarized RPE cells have different response patterns to thrombin. Although it is certain that non-polarized RPE cells and fibroblastic RPE cells are often seen in neovascular membranes excised from AMD patients,32 it should be noted that there is no definite evidence that the present non-polarized RPE cells are equal to RPE cells in that lesion. In summary, thrombin up-regulates VEGF secretion but does not affect cellular proliferation or the barrier property in polarized RPE. On the other hand, thrombin stimulates proliferation and disrupts the barrier function in non-polarized cells. RPE cells are polarized, and their polarity is gradually lost by aging or degeneration. RPE cells are exposed to blood and Current Eye Research

Thrombin and Polarized RPE Cells thrombin in retinal diseases or trauma. If RPE cells are well polarized and healthy, the adverse effects of thrombin would be limited. If the polarity is lost, RPE might easily enhance the deterioration by CNV formation and/or fibrosis. Our results provide support for the role of preventive care to keep RPE cells healthy.

DECLARATION OF INTEREST

14.

15.

16.

The authors report no conflicts of interest. 17.

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

REFERENCES 18. 1. Rizzolo LJ. Polarity and the development of the outer blood-retinal barrier. Histol Histopathol 1997;12:1057–1067. 2. Marmor MF, Abdul-Rahim AS, Cohen DS. The effect of metabolic inhibitors on retinal adhesion and subretinal fluid resorption. Invest Ophthalmol Vis Sci 1980;19: 893–903. 3. Blaauwgeers HG, Holtkamp GM, Rutten H, Witmer AN, Koolwijk P, Partanen TA, et al. Polarized vascular endothelial growth factor secretion by human retinal pigment epithelium and localization of vascular endothelial growth factor receptors on the inner choriocapillaris. Evidence for a trophic paracrine relation. Am J Pathol 1999; 155:421–428. 4. Marneros AG, Fan J, Yokoyama Y, Gerber HP, Ferrara N, Crouch RK, et al. Vascular endothelial growth factor expression in the retinal pigment epithelium is essential for choriocapillaris development and visual function. Am J Pathol 2005;167:1451–1459. 5. Avery RL, Fekrat S, Hawkins BS, Bressler NM. Natural history of subfoveal subretinal hemorrhage in age-related macular degeneration. Retina 1996;16:183–189. 6. Scupola A, Coscas G, Soubrane G, Balestrazzi E. Natural history of macular subretinal hemorrhage in age-related macular degeneration. Ophthalmologica 1999;213:97–102. 7. Lundblad RL, Bradshaw RA, Gabriel D, Ortel TL, Lawson J, Mann KG. A review of the therapeutic uses of thrombin. Thromb Haemost 2004;91:851–860. 8. Yasuno Y, Miura M, Kawana K, Makita S, Sato M, Okamoto F, et al. Visualization of sub-retinal pigment epithelium morphologies of exudative macular diseases by high-penetration optical coherence tomography. Invest Ophthalmol Vis Sci 2009;50:405–413. 9. Laposata M, Dovnarsky DK, Shin HS. Thrombin-induced gap formation in confluent endothelial cell monolayers in vitro. Blood 1983;62:549–556. 10. Liu J, Schuff-Werner P, Steiner M. Thrombin/thrombin receptor (PAR-1)-mediated induction of IL-8 and VEGF expression in prostate cancer cells. Biochem Biophys Res Commun 2006;343:183–189. 11. McNamara CA, Sarembock IJ, Gimple LW, Fenton 2nd JW, Coughlin SR, Owens GK. Thrombin stimulates proliferation of cultured rat aortic smooth muscle cells by a proteolytically activated receptor. J Clin Invest 1993;91: 94–98. 12. Bian ZM, Elner SG, Elner VM. Thrombin-induced VEGF expression in human retinal pigment epithelial cells. Invest Ophthalmol Vis Sci 2007;48:2738–2746. 13. Parrales A, Lopez E, Lopez-Colome AM. Thrombin activation of PI3K/PDK1/Akt signaling promotes cyclin D1 !

2014 Informa Healthcare USA, Inc.

19.

20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

30.

31.

9

upregulation and RPE cell proliferation. Biochim Biophys Acta 2011;1813:1758–1766. Sakamoto T, Sakamoto H, Sheu SJ, Gabrielian K, Ryan SJ, Hinton DR. Intercellular gap formation induced by thrombin in confluent cultured bovine retinal pigment epithelial cells. Invest Ophthalmol Vis Sci 1994;35:720–729. Sonoda S, Sreekumar PG, Kase S, Spee C, Ryan SJ, Kannan R, et al. Attainment of polarity promotes growth factor secretion by retinal pigment epithelial cells: relevance to age-related macular degeneration. Aging (Albany NY) 2010;2:28–42. Sonoda S, Spee C, Barron E, Ryan SJ, Kannan R, Hinton DR. A protocol for the culture and differentiation of highly polarized human retinal pigment epithelial cells. Nat Protoc 2009;4:662–673. Shirasawa M, Sonoda S, Terasaki H, Arimura N, Otsuka H, Yamashita T, et al. TNF-alpha disrupts morphologic and functional barrier properties of polarized retinal pigment epithelium. Exp Eye Res 2013;110:59–69. Bok D. The retinal pigment epithelium: a versatile partner in vision. J Cell Sci Suppl 1993;17:189–195. Burke JM. Epithelial phenotype and the RPE: is the answer blowing in the Wnt? Prog Retin Eye Res 2008;27: 579–595. Terasaki H, Kase S, Shirasawa M, Otsuka H, Hisatomi T, Sonoda S, et al. TNF-alpha decreases VEGF secretion in highly polarized RPE cells but increases it in non-polarized RPE cells related to crosstalk between JNK and NF-kappaB pathways. PLoS One 2013;8:e69994. Yoshinaga N, Arimura N, Otsuka H, Kawahara K, Hashiguchi T, Maruyama I, et al. NSAIDs inhibit neovascularization of choroid through HO-1-dependent pathway. Lab Invest 2011;91:1277–1290. Ruiz-Loredo AY, Lopez E, Lopez-Colome AM. Thrombin promotes actin stress fiber formation in RPE through Rho/ROCK-mediated MLC phosphorylation. J Cell Physiol 2011;226:414–423. Giacinti C, Giordano A. RB and cell cycle progression. Oncogene 2006;25:5220–5227. Vasioukhin V, Bauer C, Yin M, Fuchs E. Directed actin polymerization is the driving force for epithelial cell-cell adhesion. Cell 2000;100:209–219. Rajasekaran AK, Hojo M, Huima T, Rodriguez-Boulan E. Catenins and zonula occludens-1 form a complex during early stages in the assembly of tight junctions. J Cell Biol 1996;132:451–463. Ando-Akatsuka Y, Yonemura S, Itoh M, Furuse M, Tsukita S. Differential behavior of E-cadherin and occludin in their colocalization with ZO-1 during the establishment of epithelial cell polarity. J Cell Physiol 1999;179:115–125. Kawkitinarong K, Linz-McGillem L, Birukov KG, Garcia JG. Differential regulation of human lung epithelial and endothelial barrier function by thrombin. Am J Respir Cell Mol Biol 2004;31:517–527. Baker SM, Kim N, Gumpert AM, Segretain D, Falk MM. Acute internalization of gap junctions in vascular endothelial cells in response to inflammatory mediator-induced G-protein coupled receptor activation. FEBS Lett 2008;582: 4039–4046. Hayashi S, Takeuchi K, Suzuki S, Tsunoda T, Tanaka C, Majima Y. Effect of thrombin on permeability of human epithelial cell monolayers. Pharmacology 2006;76:46–52. Hollborn M, Petto C, Steffen A, Trettner S, Bendig A, Wiedemann P, et al. Effects of thrombin on RPE cells are mediated by transactivation of growth factor receptors. Invest Ophthalmol Vis Sci 2009;50:4452–4459. Russo C, Polosa R. TNF-alpha as a promising therapeutic target in chronic asthma: a lesson from rheumatoid arthritis. Clin Sci (Lond) 2005;109:135–142.

10 H. Terasaki et al. 32. Lopez PF, Sippy BD, Lambert HM, Thach AB, Hinton DR. Transdifferentiated retinal pigment epithelial cells are immunoreactive for vascular endothelial growth factor in

surgically excised age-related macular degenerationrelated choroidal neovascular membranes. Invest Ophthalmol Vis Sci 1996;37:855–868.

SUPPLEMENTARY MATERIAL AVAILABLE ONLINE Supplementary Figure 1.

Curr Eye Res Downloaded from informahealthcare.com by Kainan University on 04/03/15 For personal use only.

Supplementary material can be viewed and downloaded at http://informahealthcare.com/cey

Current Eye Research

Different Effects of Thrombin on VEGF Secretion, Proliferation, and Permeability in Polarized and Non-polarized Retinal Pigment Epithelial Cells.

We investigated the effect of thrombin on the secretion of vascular endothelial growth factor (VEGF), on cellular proliferation, and on the integrity ...
532KB Sizes 0 Downloads 6 Views