Benzalkonium chloride-induced myofibroblastic transdifferentiation of Tenon’s capsule fibroblasts is inhibited by coculture with corneal epithelial cells or by interleukin-10

Benzalkonium chloride (BAC) is used as a preservative in eyedrops but induces subconjunctival fibrosis that can result in failure of glaucoma surgery. Tenon’s capsule fibroblasts in subconjunctival tissue interact with the corneal epithelium through tear fluid. With the use of a coculture system, we have now investigated the effect of human corneal epithelial (HCE) cells on myofibroblastic transdifferentiation of human Tenon fibroblasts (HTFs) induced by BAC (5 × 10−6%). Immunofluorescence and immunoblot analyses revealed that the BAC-induced expression of α smooth muscle actin (αSMA) in HTFs was suppressed by coculture of these cells with HCE cells (p < 0.01). The concentration of interleukin-10 (IL-10) in culture supernatants of BAC-treated HTFs was increased by coculture with HCE cells (17.26-fold, vs. coculure, p < 0.001). Immunofluorescence and immunoblot analyses also showed that exogenous IL-10 (300 pg/ml) suppressed the BAC-induced expression of αSMA by 43.65% (p < 0.05) as well as the nuclear translocation of myocardin-related transcription factor-A (MRTF-A) by 39.32% (p < 0.01) in HTFs cultured alone. Our findings suggest that corneal epithelial cells may protect against subconjunctival fibrosis by maintaining IL-10 levels and preventing the MRTF-A-dependent transdifferentiation of HTFs into myofibroblasts.

www.nature.com/scientificreports/ response to various stimuli including injury, infection, inflammation, and stress. Myocardin-related transcription factor (MRTF) signaling plays an important role in the differentiation and activation of myofibroblasts 11,12 . We have previously shown that MRTF signaling is associated with epithelial-mesenchymal transition (EMT) of retinal pigment epithelial cells 13 . MRTF-A binds to the promoter region and thereby regulates the expression of myofibroblast-related genes such as those for αSMA, fibronectin, and connective tissue growth factor (CTGF) 14 .
The immune system contributes to the regulation of myofibroblasts 15 , and the anti-inflammatory cytokine interleukin (IL)-10 is a potential therapeutic target for modulation of tissue remodeling 10 .
Histological specimens of failed filtration structures after glaucoma surgery show an increased number of αSMA-positive cells and high-density deposition of collagen types I or III compared with corresponding normal tissue 16,17 . BAC was shown to shorten bleb survival time after rabbit glaucoma filtration surgery as well as to increase the expression of αSMA in the filtration bleb 18 .
The cornea, conjunctiva, and eyelids maintain ocular surface homeostasis and interact in pathological changes that affect each tissue 19 . The corneal epithelium plays an important role as a barrier at the surface of the eye. Disruption of this barrier allows the spread of inflammation from the conjunctiva or eyelid into the cornea 19 . The corneal epithelium and subconjunctival tissue also communicate with each other via tear fluid 19 . However, the possible effect of the corneal epithelium on the BAC-induced transdifferentiation of Tenon's capsule fibroblasts into myofibroblasts has been unknown.
To address this question, we have now examined the effect of BAC on the transdifferentiation of human Tenon's capsule fibroblasts (HTFs) into myofibroblasts in a coculture system with simian virus 40 (SV40)immortalized human corneal epithelial (HCE) cells. We also examined the possible protective effect of IL-10 released into the coculture supernatant on the observed deleterious action of BAC on HTFs in this system.

Cytotoxicity of BAC for HTFs and HCE cells.
We first examined the effect of BAC on the viability of HTFs and HCE cells by measuring the release of lactate dehydrogenase (LDH) into the culture medium. The cells were deprived of serum for 24 h and then exposed to various concentrations of BAC for 24 h. We found that the amount of LDH in the medium of HTFs (p = 0.014) or HCE (p < 0.001) cells was increased by exposure to BAC in a concentration-dependent manner, with this effect being significant at BAC concentrations of ≥ 10 × 10 −6 % ( Fig. 1). We therefore performed subsequent experiments with BAC at the nontoxic concentration of 5 × 10 −6 %.

Effect of BAC on HTF transdifferentiation into myofibroblasts.
We next examined the effect of BAC at the nontoxic concentration on the phenotype of HTFs. Immunofluorescence analysis revealed that BAC induced marked expression of αSMA, a key marker of myofibroblast differentiation 9,10 , in these cells (Fig. 2a). Immunoblot analysis confirmed that BAC induced a significant increase in the abundance of αSMA in HTFs (2.02-fold, vs. BAC-untreated, p = 0.011, Fig. 2b).

Effect of HCE cells on the BAC-induced transdifferentiation of HTFs into myofibroblasts.
Given that the corneal epithelium and subconjunctival tissue including Tenon's capsule fibroblasts interact with each other through tear fluid 19 , we examined the effect of HCE cells on the BAC-induced myofibroblastic transdifferentiation of HTFs in a coculture system. HCE cells were thus cultured on the bottom of six-well plates, and HTFs were cultured in hanging mesh inserts. The cells were deprived of serum for 24 h and then exposed to BAC for 24 h. Immunofluorescence analysis revealed that HTFs cultured in the inserts in the absence of HCE cells manifested BAC-induced expression of αSMA, whereas this effect was greatly suppressed in HTFs cocultured with HCE cells (Fig. 3a). Again, these results were confirmed by immunoblot analysis (Fig. 3b).
Effects of BAC on cytokine release in the HTF/HCE cell coculture system. To explore the mechanism by which HCE cells might suppress BAC-induced expression of αSMA in HTFs, we measured the concentrations of IL-6, monocyte chemoattractant protein-1 (MCP-1), and IL-10 released into the culture medium. The levels of IL-6 and MCP-1 in culture supernatants of BAC-treated HTFs cultured alone were significantly greater than those for BAC-treated cocultures of HTFs and HCE cells (p < 0.01, Fig. 4a). The levels of IL-6 and MCP-1 did not differ significantly between HTF/HCE cell cocultures maintained with or without BAC. In contrast, the concentration of IL-10 in the medium of BAC-treated HTFs cultured alone was only 5.79% of that for BACtreated cocultures of HTFs and HCE cells (p < 0.001, Fig. 4a). Again, the level of IL-10 did not differ significantly between HTF/HCE cell cocultures maintained with or without BAC. To examine which of the two cell types might be responsible for the production of IL-10, we also determined the abundance of IL-10 mRNA by reverse transcription (RT) and quantitative polymerase chain reaction (qPCR) analysis in HTFs and HCE cells after coculture. Whereas the amount of IL-10 mRNA in HTFs did not differ significantly between cocultures maintained with or without BAC, that in HCE cells was significantly increased by BAC (1.32-fold, vs. HCE without BAC, p = 0.033, Fig. 4b).
Effect of IL-10 on the BAC-induced transdifferentiation of HTFs into myofibroblasts. Given that the level of IL-10 in culture supernatants of HTF/HCE cell cocultures was increased compared with that for HTFs alone in the presence of BAC, we hypothesized that IL-10 might contribute to the attenuation by HCE cells of the BAC-induced transdifferentiation of HTFs into myofibroblasts. We therefore examined whether exogenous IL-10 might inhibit this effect of BAC. HTFs were incubated for 24 h with IL-10 (300 pg/ml) in serum-free medium and then exposed to BAC for an additional 24 h. Immunofluorescence staining (Fig. 5a) and immunoblot analysis (Fig. 5b) indeed showed that the BAC-induced expression of αSMA in HTFs was essentially abolished in the presence of IL-10. www.nature.com/scientificreports/

Effects of BAC and IL-10 on the subcellular localization of MRTF-A in HTFs.
We previously showed that EMT is mediated through MRTF-A signaling in retinal pigment epithelial cells 13 . We therefore investigated whether MRTF-A signaling might also play a role in the BAC-induced transdifferentiation of HTFs into myofibroblasts. Immunofluorescence analysis revealed that MRTF-A was localized predominantly to the cytoplasm of HTFs in the absence of BAC, whereas it had translocated to the nucleus in cells exposed to BAC (Fig. 6a). Importantly, this effect of BAC on MRTF-A localization was inhibited IL-10. The BAC-induced translocation of MRTF-A to the nucleus and its significant attenuation by 39.32% in cells exposed to IL-10 were confirmed by immunoblot analysis (p < 0.001, Fig. 6b).

Discussion
Fibroblasts transdifferentiate into myofibroblasts in response to various stimuli. Myofibroblasts mediate ECM secretion, accumulation, and contraction and thereby contribute to fibrosis 9,20 . Activation of HTFs also initiates myofibroblast transdifferentiation and mediates subconjunctival fibrosis 9,16,17,21 . BAC was previously shown to induce conjunctival and subconjunctival fibrosis 1,4,22,23 . Furthermore, αSMA expression was found to be increased in the filtration surgery flap of a rabbit model of dry eye treated with BAC 18 . We have now shown that BAC www.nature.com/scientificreports/ increased the expression of αSMA in primary HTFs, suggesting that BAC triggers the transdifferentiation of HTFs into myofibroblasts and might thereby promote subconjunctival fibrosis. BAC has toxic effects on the ocular surface and can be allergenic 1 . It induces ocular surface disorders including tear film instability by eliciting the loss of corneal epithelial cells, conjunctival cells, and subconjunctival cells 4,24 . The concentration of BAC as a preservative in eyedrops ranges from 0.004 to 0.02% 25,26 . We found that BAC induced the transdifferentiation of HTFs into myofibroblasts at a concentration of 5 × 10 −6 %, a concentration much lower than that present in eyedrops.
Components of the ocular surface include the cornea, conjunctiva, lacrimal glands, and eyelids, and they are interconnected functionally 27 . The corneal epithelium acts as a physical barrier and plays an important role in the ocular immune response 19 . In the present study, activation of HTFs by exposure to BAC was suppressed www.nature.com/scientificreports/ by the additional presence of HCE cells in a coculture system, indicating that corneal epithelial cells and HTFs may interact with each other in vivo and that such interaction contributes to homeostasis of the ocular surface. Nuclear translocation of MRTF-A mediates activation of the expression of genes related to the myofibroblastic phenotype and fibrosis 9,11,12 . The MRTF-A pathway thus promotes αSMA expression in and scar tissue formation by conjunctival fibroblasts 28,29 . We have also previously shown that MRTF-A activation contributes to subretinal fibrosis 13 . MRTF-A forms a complex with G-actin and is located predominantly in the cytoplasm of cells under resting conditions 30,31 . Cell stimulation results in the dissociation of MRTF-A from G-actin and its transient translocation to the nucleus, where it interacts with serum response factor (SRF) and induces the expression of genes including that for αSMA 30,31 . The nuclear translocation of MRTF-A thus shifts the balance in the distribution of MRTF-A between the cytoplasm and the nucleus 32 . Our immunostaining and immunoblot data together showed that MRTF-A was essentially restricted to the cytoplasm, being absent from the nucleus, in HTFs under the control condition, whereas a proportion of the protein translocated to the nucleus in response to BAC exposure. Our present data thus show that BAC induces the nuclear translocation of MRTF-A as well as the expression of αSMA in HTFs, suggesting that BAC activates the MRTF-A-αSMA axis in these cells and thereby induces their transdifferentiation into myofibroblasts.
The corneal epithelium and subconjunctival tissue are connected to each other by tear fluid and the various factors that it contains 19 . Our results now indicate that the levels of IL-6 and MCP-1 in the culture supernatant . Effect of HCE cells on the BAC-induced transdifferentiation of HTFs into myofibroblasts in a coculture system. (a) HTFs cultured in hanging mesh inserts of six-well plates containing (or not) HCE cells were deprived of serum for 24 h, incubated in the absence or presence of BAC (5 × 10 −6 %) for 24 h, and then subjected to immunofluorescence staining with antibodies to αSMA (green). Nuclei were counterstained with DAPI (blue). Scale bars, 50 µm. (b) HTFs treated as in (a) were subjected to immunoblot analysis with antibodies to αSMA. Blots for duplicates of one representative experiment as well as quantitative data (means + s.e.m.) for the relative αSMA/α-tubulin band intensity ratio from five independent experiments are shown. **p < 0.01 (Holm-Sidak post hoc test). Full-length blots are presented in Supplementary Fig. 2   www.nature.com/scientificreports/ of HTFs were increased by exposure to BAC, and that the BAC-induced increases in the concentrations of these inflammatory cytokines were significantly inhibited by coculture of HTFs with HCE cells. In contrast, the production of IL-10 appeared to be decreased in HTF cultures exposed to BAC, and this effect was prevented by coculture with HCE cells. IL-10 is an anti-inflammatory cytokine and maintains balance of the immune response 33 . It also inhibits the proliferation of and collagen synthesis by dermal, lung, and kidney cells 10 . Various studies have examined the possible roles of cytokines in ocular surface tissue fibrosis. The levels of IL-10 and IL-6 in tear fluid were found to be increased in individuals subjected to long-term treatment with eyedrops containing BAC 3 , suggesting a possible role for IL-10 in BAC-induced reactions at the ocular surface. Stevens-Johnson syndrome (SJS) and ocular cicatricial pemphigoid (OCP) are rare but severe autoimmune diseases that are associated with chronic inflammation of the ocular surface, stromal scarring, and symblepharon. Pseudo-OCP is induced by long-term  ) with the exception that they were exposed (or not) to BAC for 24 h were subjected to subcellular fractionation, and the nuclear fraction was subjected to immunoblot analysis with antibodies to MRTF-A and to lamin A/C (nuclear marker). Blots for two independent experiments as well as quantitative data (means + s.e.m.) for the relative αSMA/lamin band intensity ratio from four independent experiments are shown. *p < 0.05, n.s. (Holm-Sidak post hoc test). Full-length blots are presented in Supplementary Fig. 4. www.nature.com/scientificreports/ use of topical ocular agents, especially that of antiglaucoma eyedrops containing preservatives. The conjunctiva of individuals with OCP was found to contain αSMA-expressing fibroblasts 34 or to manifest overexpression of inflammatory cytokines 35 . The level of IL-10 in tear fluid was also shown to be decreased in patients with SJS compared with healthy individuals 34 . Subconjunctival injection of mesenchymal stem cells (MSCs) reduced the incidence of corneal allograft rejection in rats, and the abundance of IL-10 mRNA in the allografts was significantly higher in animals treated with MSCs than in those not so treated 36 . Furthermore, subconjunctival injection of IL-10-overexpresing bone marrow-derived MSCs into rat corneal allografts prolonged allograft survival 37 .

Scientific Reports
There is evidence that conjunctival cells produces intracellular cytokines including TNF-α and IL-6 3 , which are involved in transdifferentiation of human Tenon's fibroblasts to myofibroblasts 38 , and excrete them in tears 3 . Their expression was also up-regulated in conjunctival scarring specimens obtained from patients with ocular surface inflammatory diseases 39,40 . In contrast, corneal epithelial cells secrete anti-inflammatory cytokines including IL-10 by BAC 41 . The relation between corneal epithelial cells and subconjunctival fibrosis has not previously been examined in vitro or in vivo. Our present results implicate cytokines in maintenance of homeostasis of ocular surface tissues exposed to BAC. The amount of IL-10 mRNA in HCE cells was significantly increased by exposure to BAC, suggesting that corneal epithelial cells may increase their production of IL-10 in response to treatment with eyedrops containing BAC. The amount of IL-10 mRNA in HTFs cocultured with HCE cells was also not decreased by exposure to BAC, suggesting that HCE cells might protect HTFs not only by producing IL-10 themselves but also by attenuating inhibition of IL-10 secretion from HTFs. In addition, we found that exogenous IL-10 inhibited the BAC-induced release of IL-6 and MCP-1 by HTFs cultured alone (p < 0.01). The concentrations of IL-6 and MCP-1 in culture supernatant of HTFs were thus increased 1.718 ± 0.241-fold and 1.940 ± 0.187-fold (means ± s.e.m. from three independent experiments), respectively, by exposure to BAC, whereas these values were reduced to 1.269 ± 0.101-fold and 1.560 ± 0.240-fold, respectively, in the additional presence of exogenous IL-10. These results thus indicate that IL-10 might ameliorate inflammatory and immune responses at the ocular surface including those in subconjunctival tissue. We further found that exogenous IL-10 inhibited the BAC-induced expression of αSMA and nuclear translocation of MRTF-A in HTFs, suggesting that IL-10 attenuates BAC-induced myofibroblastic transdifferentiation of HTFs by suppressing a BAC-MRTF-A-αSMA signaling pathway. IL-10 is therefore a potential therapeutic agent for tissue remodeling.
Our results suggest that corneal epithelial cells might play a key role in protection of Tenon fibroblasts from BAC-induced transdifferentiation into myofibroblasts. Furthermore, they implicate IL-10 as a key biological mediator in maintenance of homeostasis of ocular surface tissues, and they suggest that administration of IL-10 might protect Tenon fibroblasts from myofibroblastic transdifferentiation induced by BAC exposure. Our observations may provide the basis for a new strategy to prevent failure of glaucoma surgery due to BAC. The presence of drug-induced corneal epithelial damage might be a prognostic factor for glaucoma filtration surgery, with local application of IL-10 being a potential new approach to the control of subconjunctival fibrosis after such surgery. Moreover, IL-10 might also prove beneficial for topical ocular treatment in individuals with SJS or with OCP or pseudo-OCP.
In summary, we have examined for the first time the interaction of HTFs and HCE cells in a coculture system. We found that HCE cells inhibited the BAC-induced transdifferentiation of HTFs into myofibroblasts, and that this effect of HCE cells was mimicked by IL-10. Although the precise mechanism of this interaction between the two cell types is unclear, it may be mediated at the level of IL-10 production, given that the concentration of IL-10 in culture supernatants of HTFs was increased in the presence of HCE cells. BAC also induced the nuclear translocation of MRTF-A in HTFs, and this effect was again inhibited by IL-10, suggesting that MRTF-A mediates the effect of BAC on HTF transdifferentiation. Further studies are warranted to identify proteins that act upstream of MRTF-A in this pathway as well as to examine whether IL-10 might inhibit BAC-induced subconjunctival fibrosis in vivo.
Cell culture. Given that primary human corneal epithelial cells have a short life span in culture, we used an SV40-immortalized human corneal epithelial cell line that retains the phenotypic morphology and function of the parent cells 42 in this study. HCE cells were obtained from RIKEN Biosource Center (Tsukuba, Japan) and were maintained under a humidified atmosphere of 5% CO 2 and 95% air at 37 °C in DMEM-F12 supplemented with 10% FBS, insulin (5 μg/ml), cholera toxin (0.1 μg/ml), epidermal growth factor (10 ng/ml), and gentamicin  43 . HTFs were isolated as described previously 44 from individuals undergoing strabismus surgery who had no history of conjunctival disease or use of topical ocular medication. The use of these cells complied with the tenets of the Declaration of Helsinki and was approved by the human experimentation committee of Yamaguchi University Graduate School of Medicine, and informed consent was obtained from each donor. HTFs were maintained under an atmosphere of 95% air and 5% CO 2 at 37 °C in MEM supplemented with 10% FBS and antibiotic-antimycotic mixture, and they were used for the present study after three to eight passages. For coculture of HTFs and HCE cells, HTFs cultured in hanging mesh inserts to 70-80% confluence were transferred to six-well plates containing HCE cells also at 70-80% confluence, and both cell types were maintained in the same medium (MEM).
Assay of LDH. LDH  were fractionated by SDS-polyacrylamide gel electrophoresis on a 5-20% gradient gel (SuperSep Ace; Wako, Osaka, Japan) at a constant current of 20 mA per gel, after which the separated proteins were transferred to an Immobilon-P membrane (Millipore, Billerica, MA, USA). The membrane was exposed to 3% BSA in Tris- Statistical analysis. Quantitative data are presented as mean + s.e.m. values and were analyzed with the two-tailed Student's t test or by analysis of variance followed by the Holm-Sidak post hoc test for comparisons between two or among more than two groups, respectively. The analysis was performed with SigmaStat 13.0 software (Systat Software, San Jose, CA, USA). A P value of < 0.05 was considered statistically significant.

Data availability
The data sets generated and analyzed during the current study are available from the corresponding author on reasonable request.