Lacritin Salvages Human Corneal Epithelial Cells from Lipopolysaccharide Induced Cell Death

Innate immunity of the corneal epithelium is conferred by proteinaceous secretions from the epithelium and associated lacrimal and meibomian glands. Lacritin, an eye-specific protein with anti-microbial, cytoprotective and wound-healing properties, predominantly secreted by lacrimal glands, is absent in conditions such as Dry eye and Keratitis. In view of the biological significance of lacritin in human eye, we investigated its role in human corneal epithelial (HCE) cells during lipopolysaccharide (LPS)-induced infection. LPS-challenged HCE cells demonstrated apoptosis-mediated cell death and elevated lacritin levels. The LPS-induced cell death is alleviated with exogenous supplementation of recombinant lacritin. This cytoprotective effect of lacritin is mediated through Cyclooxygenase-2 (COX-2). This study is the first to highlight the protective role of lacritin and mechanism of its action during bacterial infection of cornea in vitro.

Scientific RepoRts | 5:18362 | DOI: 10.1038/srep18362 mitochondrial membrane spaces into the cytoplasm upon LPS treatment further confirms apoptosis-mediated cell death (Fig. 1B iii). Change in mitochondrial membrane potential, which is also a measure of apoptotic cell death, was studied by flow cytometry analysis using Rhodamine 123. LPS-induced mitochondrial membrane depolarization was depicted by the shift in cell population towards the lower scale of the FL1-H. In comparison to control cells (Fig. 1C i), 42% LPS-treated cells were depolarized (Fig. 1C ii).
Corneal epithelial cell signaling is altered by LPS. The cellular dynamics between nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κ B) and nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor-alpha (I-κ B α ) is altered in LPS-treated corneal epithelial cells. LPS-treated nuclear lysates showed time-dependent decrease in the nuclear translocation of NF-κ B ( Fig. 2A i). In addition, elevated I-kB α levels indicate inhibition of its degradation in the whole cell extracts ( Fig. 2A ii). Furthermore, the expression of the cyclin-dependent kinase inhibitor, p27, a key regulator of cell proliferation and apoptosis was also increased with LPS treatment ( Fig. 2A iii). Furthermore, the effect of LPS treatment on COX-2 expression at both mRNA (Fig. 2B i) and protein level was (Fig. 2B ii) monitored using real-time RT-PCR and western blot analysis respectively. The COX-2 expression was elevated at 3 h and diminished at later time points up to 24 h.

Lacritin recovers corneal epithelial cells from the deleterious effects of LPS. LPS-treated corneal
epithelial cells showed temporal regulation of lacritin. High dose of LPS (10 μ g/ml) induced early expression of lacritin (i.e. at 1.5 h) (Fig. 3A), whereas a lower dose (100 ng/ml) induced late expression (Fig. 3B). Lacritin treatment (0, 0.1, 1, 10, 100, 1000 nM) increases the HCE cell number in biphasic manner with the most effective dose at 10 nM. As lacritin shows maximum effect at 10 nM and viability of HCE cells decreases to 50% at (10 μ g/ ml) of LPS, all the experiments were performed at these doses. Addition of recombinant lacritin (10 nM) to the LPS-treated cells increased the cell viability from 54% to 72%. Lacritin alone increased total cell number by ~15% (Fig. 4A). However, lacritin was unable to restore the cell viability of LPS treated HCE cells depleted of COX-2 (Fig. 4B). Cyt c release during LPS-mediated apoptosis was reduced (Fig. 4C i). Further, NF-κ B translocation to  (ii) I-κ Bα and (iii) p27 proteins (iv) β -actin was used as an internal control. HCE cells were treated with LPS (10 μ g/ml) for different time periods (0, 1.5, 3, 6, 12 and 24 h). Equal amounts of total protein (for I-κ B α , p27 and β -actin) and nuclear protein (for NF-κ B-p65) was analysed by SDS-PAGE (10-12%), and after electrophoresis under same experimental conditions, proteins on the gel were transferred on to nitrocellulose membrane and probed with specific antibodies. The representative images of three independent experiments shown were cropped. (B) Effect of LPS on the expression of COX-2. (i) Quantitative analysis of COX-2 mRNA expression in HCE cells relative to 18S rRNA expression at different time points (0, 1.5, 3, 6, 12 and 24 h) after LPS treatment by qRT-PCR is reported as fold change relative to 0 h calculated using 2 −ΔΔCT . X-axis represents exposure time in hours. Values are mean ± SEM for three independent experiments. *Indicates significance at p < 0.05. (ii) HCE cells were treated with LPS (10 μ g/ml) for different time periods (0, 1.5, 3, 6, 12 and 24 h). Equal amounts of protein was analyzed by SDS-PAGE (12%), proteins on the gel were transferred on to nitrocellulose membrane and probed with COX-2 specific antibody under same experimental conditions. The representative images of three independent experiments shown were cropped. the nucleus (Fig. 4C ii) and I-kB α degradation (Fig. 4C iii) was enhanced. Furthermore, the declined levels of COX-2 were elevated in the presence of lacritin (Fig. 4C iv).

Discussion
Corneal epithelium in synergy with tear fluid and its soluble factors plays a crucial role in innate immunity during infection 26 . LPS, the bacterial virulence factor, is recognized by TLR4 on the epithelial cells; thereafter via downstream signaling regulate the proteins responsible for innate immunity 27 . Functional aspects of most tear proteins are uncharacterized in the eye pathophysiology 11 . Lacritin is one such eye-specific bactericidal protein 11,18,19 whose role in corneal infection remains largely unexplored. For the first time, we demonstrated that lacritin salvages the corneal epithelium against bacterial LPS and that this effect is mediated through COX-2.
Prominent corneal epithelial cell death observed at high LPS doses in our study corroborates most of the previous studies depicting apoptosis of cornea during infection by LPS 3,28 . Together, these studies indicate that the LPS action on the human corneal epithelium is dose-dependent.
LPS-mediated apoptosis involves the alteration of Bcl2 family proteins 29 . The decrease in the Bcl2/Bax protein ratio is a pro-apoptotic signal which shifts the homeostasis of the cells towards death by depolarizing and permeabilizing mitochondria resulting in release of Cyt c into the cytoplasm which in turn activates further steps for the execution of cell death. Similar to our study, Bcl2/Bax expression pattern was also observed in HCE cells treated with allethrin 30 and during hyper-osmolarity 31 . (ii) Untransfected or NC1 transfected or COX-2 depleted HCE cells were treated with either LPS (10 μ g/ml) alone or lacritin alone (10 nM) or both for 24 h. Viability of untransfected control cells was 100% and data is expressed as mean percent of untreated control ± SEM for three independent experiments. *Indicates significant difference at p < 0.05 when compared to untreated control. #Indicates significant difference at p < 0.05 when compared to their respective LPS alone treated cells. (C) Effect of recombinant lacritin on the levels of (i) Cyt c (ii) NF-κ B-p65 (iii) I-κ B α and (iv) COX-2 in LPS treated HCE cells. HCE cells were treated with either LPS (10 μ g/ml) alone or lacritin (10 nM) alone or both for 24 h. Equal amounts of total protein (for I-κ Bα and β -actin), nuclear protein (for NF-κ B-p65) and cytosolic protein (for Cyt c) was analysed by SDS-PAGE (10-12%), and after electrophoresis under same experimental conditions, proteins on the gel were transferred on to nitrocellulose membrane and probed with protein specific antibodies. The representative images of three independent experiments shown were cropped.
Scientific RepoRts | 5:18362 | DOI: 10.1038/srep18362 NF-κ B activation observed in the present study as well as in Pseudomonas aeruginosa infected corneal epithelial cells 32 reiterates the role of NF-κ B as a first-line of defense during inflammation and infection 32,33 . NF-κ B is rapidly activated by a large spectrum of chemically diverse agents including bacterial LPS, microbial and viral pathogens. Inactive NF-κ B sequestrated in the cytoplasm, is bound by I-κ B family proteins. Diverse stimuli phosphorylate I-κ B leading to its ubiquitination and subsequent degradation thus exposing nuclear localization signals (NLS) on NF-κ B subunits. This results in translocation of NF-κ B and subsequent activation of gene transcription 34 .
CDK inhibitor, P27, is a protein which plays an important role in progression of cell cycle in corneal epithelial cells 35 . Elevated levels of p27 were detected in LPS-induced corneal cells similar to that shown in staurosporine-induced apoptosis of corneal epithelial cells 36 .
In the present study, the ability of the cell to temporally regulate lacritin expression based on the LPS concentration indicates that LPS evokes innate immunity in addition to causing damage to the corneal epithelial cells as demonstrated elsewhere 37 . Whether LPS-induced lacritin confers innate immunity remains to be investigated further; however, lacritin was shown to offer cytoprotection to the human corneal epithelial cells under different stress conditions 23,24 .
LPS modulates COX-2 expression in a similar pattern as that of lacritin. Further, it is reported that lacritin's potential to induce proliferation is mediated by COX-2 17 . This indicates that the enhanced expression of COX-2 at 3 h could be a result of lacritin induction at an earlier time point (i.e., at 1.5 h) in corneal cells. However, normal levels of COX-2 were restored when LPS was administered in the presence of lacritin. This restoration offers partial but significant cytoprotection. Lacritin was unable to play a cytoprotective role in the LPS treated HCE cells depleted of COX-2, suggesting that the cytoprotective role of lacritin is mediated through COX-2.
The cytoprotective role of lacritin could be mediated by COX-2, in addition to lacritin's ability to reverse LPS-induced apoptotic signaling as shown in the present study and based on previous reports 17,38 . Few studies have shown the role of autophagy in cytoprotective function of lacritin 23,24 . Additional studies in this direction are warranted to elucidate the true therapeutic potential of this molecule in eye infections. Cell viability assay. Cell viability was determined by MTT assay. HCE cells were seeded in a 96-well culture plate at density of 6 × 10 3 cells/well overnight and grown in the presence or absence of LPS (0.01, 0.1, 1, 10, 100, and 1000 μ g/ml) for 24 h in a final volume of 100 μ l medium. The medium was aspirated and 20 μ l of MTT (5 mg/ml) was added to the fresh medium. After 3 h incubation at 37 °C, the formazan crystals were dissolved in 100 μ l of DMSO. Absorbance was read at 570 nm on a multi-well plate reader. The percentage inhibition of cell viability was calculated as a fraction of control. To elucidate the cytoprotective function of lacritin and role of COX-2, lacritin (10 nM) was pre-incubated for 2 h before the addition of LPS (10 μ g/ml) with HCE cells or HCE cells depleted of COX-2. For knockdown of COX-2, HCE cells were transfected with COX-2 siRNA at final concentration of 10 nM with Lipofectamine 2000 at a ratio of 1:2 for 24, 48 and 72 h. Untransfected and Universal negative control 1 siRNA transfected cells served as controls.

Materials
Western blotting analysis. Nuclear, cytoplasmic and whole extracts were prepared from control and treated cells as described previously 39 . The protein content was estimated by the Bradford assay 40 . Proteins were resolved on 7-12% SDS-PAGE gels and transferred onto nitrocellulose or PVDF membranes. The membranes were blocked with 3% (w/v) BSA and incubated with the primary antibodies [Bax, Bcl-2, Cyt C, p27, NFκ B-p65, I-κ B α (1:1000 dilution) and β -actin (1:500 dilution)] in 10 ml of antibody-dilution buffer (1× Tris-buffered saline and 0.05% Tween 20 with 1% BSA) with gentle shaking at 4 °C for 8-12 h and incubated with horseradish peroxidase or alkaline phosphatase conjugated secondary antibodies. The signals were detected by using specific substrates.
Mitochondrial membrane potential estimation by flow cytometry analysis. After treatment with LPS (10 μ g/ml), cells were incubated with Rhodamine 123 (10 μ g/ml) for 20 min. Data were collected using the program Cell Quest of FACS Calibur (Becton Dickinson, CA, USA) and fluorescence was measured using a FL-1 detector.
Real-time PCR studies. Total RNA was isolated from cells using supplier's protocol. Superscript TM III first strand synthesis system was used to synthesize cDNAs. Primers were designed based on the gene sequences available in NLM nucleotide database. For each gene, sequences of the forward and reverse primers used in respective PCR, are as follows: Lacritin Fwd: 5′ CTCTGACTCGACGGGTGCTG3′ Rev: 5′ CCGAAGTCTCCTGGGCTGTT3′ ; COX-2 Fwd: 5′ AACAGGAGCATCCTGAATGG3′ , Rev: 5′ GGTCAATGGAAGCCTG TGATG3′ ; 18S Fwd: 5′ GCTACCACATCCAAGGAAGGCAGC3′ , Rev: 5′ CGGCTGCTGG CACCAGACTTG3′ . Real-time studies were performed on an ABI Prism H7500 fast thermal cycler (Applied Biosystems, CA, USA). Each sample was run in triplicate in a final volume of 25 μ l containing 1 μ l of template (1:5 dilution), 10 pmol of each primer and 12 μ l of Power SYBR Green PCR master mix (Applied Biosystems). The real-time PCR results were presented as a Scientific RepoRts | 5:18362 | DOI: 10.1038/srep18362 change in expression relative to control using target gene C t values normalized to that of 18S gene C t values based on the comparative C t method 41 .

Statistical analyses.
The data for all the studies were obtained from three independent experiments and average was calculated for all the variables. Differences between groups were analyzed by ANOVA followed by Student Newman-Keuls test using Sigma Stat software. Values were considered significant at p < 0.05.