This work was funded by an unrestricted grant from Horus Pharma. Carole Gard is employed by Horus Pharma. Christophe Baudouin: Financial support and consultant (Alcon; Allergan; Santen; Laboratoires Théa). There are no patents, products in development or marketed products to declare. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.
To investigate the pro-inflammatory intracellular mechanisms induced by an in vitro model of dry eye disease (DED) on a Hela-modified conjunctiva-derived cells in hyperosmolarity (HO) stress conditions. This study focused on CCL2 induction and explored the implications of the nuclear factor of activated T-cells 5 (NFAT5) as well as mitogen-activated protein kinases (MAPK) and nuclear factor kappa B (NFĸB). This work was completed by an analysis of the effects of cyclosporine A (CsA), dexamethasone (Dex) and doxycycline (Dox) on HO-induced CCL2 and NFAT5 induction.
A human HeLa-modified conjunctiva-derived cell line was cultured in NaCl-hyperosmolar medium for various exposure times. Cellular viability, CCL2 secretion, NFAT5 and CCL2 gene expression, and intracytoplasmic NFAT5 were assessed using the Cell Titer Blue® assay, enzyme-linked immunosorbent assay (ELISA), RT-qPCR and immunostaining, respectively. In selected experiments, inhibitors of MAPKs or NFκB, therapeutic agents or NFAT5 siRNAs were added before the hyperosmolar stimulations.
HO induced CCL2 secretion and expression as well as NFAT5 gene expression and translocation. Adding NFAT5-siRNA before hyperosmolar stimulation led to a complete inhibition of CCL2 induction and to a decrease in cellular viability. p38 MAPK (p38), c-Jun NH2-terminal kinase (JNK) and NFĸB inhibitors, CsA and Dex induced a partial inhibition of HO-induced CCL2, while Dox and extracellular signal-regulated kinase (ERK) inhibitor did not. Dex also induced a partial inhibition of HO-induced NFAT5 gene expression but not CsA or Dox.
These in vitro results suggest a potential role of CCL2 in DED and highlight the crucial role of NFAT5 in the pro-inflammatory effect of HO on HeLa-modified conjunctiva-derived cells, a rarely studied cellular type. This inflammatory pathway involving NFAT5 and CCL2 could offer a promising target for developing new therapies to treat DED, warranting further investigations to fully grasp the complete intracellular mechanisms.
DED) is one of the most common ocular pathologies in the world, with a prevalence of 3–15% [
The molecular mechanism that regulates the transcription and secretion of these pro-inflammatory actors under hyperosmolar conditions is poorly understood. Among the actors involved, CCL2, a potent chemoattractant protein that attracts monocytes to the inflammation site [
Topical CsA is actually the only drug specifically formulated to treat dry eye syndrome and approved by regulatory medical agencies (FDA, EMA). Restasis® (Allergan) and Ikervis® (Santen) provide an alternative to artificial tears, Dex and Dox. CsA has immunosuppressive and anti-inflammatory effects, but its exact therapeutic mechanisms on the ocular surface of dry eye patients remains unclear.
The number of patients suffering from DED is set to grow as the world population ages. A better understanding of the pathophysiology of DED and the mechanism of CsA could help find new targets for treating this pathology and relieve millions of patients from permanent and painful discomfort.
Therefore, our aim was to analyze and characterize in vitro conjunctival CCL2 induction in a hyperosmolar model of dry eye and to determine the relationship between NFAT5 and CCL2 on HeLa-modified conjunctiva-derived cells. Moreover, we investigated intracellular signaling pathways such as MAPKs (p38, JNK and ERK) and the NFĸB transcription factor. We also investigated the effects of the three major anti-inflammatory therapeutic agents: CsA, Dex and Dox on CCL2 and NFAT5 induction by HO (
1) Is HO able to stimulate CCL2 or NFAT5? 2) Is NFAT5 involved in the CCL2 stimulation? 3) Are MAPKs or NFĸB implicated in the CCL2 and NFAT5 stimulations? 4) What are the effects of CsA, Dex or Dox on these stimulations?
The Wong Kilbourne derivative of the Chang (WKD) HeLa-modified conjunctiva-derived epithelial cell line (clone 1-5c-4, American Type Culture Collection [ATCC] certified cell line [CCL], 20.2) was cultured under classic conditions (moist atmosphere, 5% CO2, 37°C) in Dulbecco minimum essential medium supplemented with 10% fetal bovine serum, 1% glutamine (200 mM), 1% penicillin (10,000 units/mL) and streptomycin (10,000 μg/mL) for 24 h to reach either 80% confluence before hyperosmolar stimulations and to reach 50% confluence before transfection for siRNA experiments. All reagents for cellular culture were purchased from Gibco (Gibco, Life technologies, Carlsbad, CA, USA). This cell line has previously been used for toxicological in vitro studies and was shown to respond similarly to the IOBA conjunctival cell line [
Each hyperosmolar medium was prepared by adding sodium chloride (Sigma-Aldrich, Saint Louis, MO, USA) to supplemented medium. All osmolarity values were assessed using an osmometer (Roebling 13DR, Berlin, Germany) including the supplemented medium, which was found at 340 mOsM (mosm.L-1) as expected from information provided by the manufacturer. We decided not to change this osmolarity level in order to keep the cell line in its regular medium even if this value is above the expected values of normal tear film, i.e., 302.2±8.3 mOsM, and in order to avoid any hypo-osmotic regulatory effects. In dry eye subjects, tear osmolarity is generally found at 336.4±22.3 mOsM [
For experiments that analyzed a range of hyperosmolar conditions, cells were treated with supplemented medium (340 mOsM) or with hyperosmolar medium ranging from 400 mOsM to 600 mOsM for 4 h before collecting cell lysates for real-time quantitative PCR (RT-qPCR), or 24 h before collecting supernatants for ELISA, performing the cell viability test or fixing cells for immunostaining. For the time-course experiments, cells were treated with medium (340 mOsM) or with hyperosmolar medium (500 mOsM) for durations ranging from 30 min to 24 h before collecting cell lysates for RT-qPCR, supernatants for ELISA or performing the cell viability test. This kind of protocol using a high osmolarity such as 500 mOsM during long duration as 24h on ocular surface cells is a common protocol often reported to study DED [
For the siRNA experiments, cells were treated with a mix of lipofectamine (Life Technologies, Carlsbad, CA, USA) and negative control siRNA (Life Technologies) or NFAT5 siRNA (Life Technologies) for 24 h, as recommended by the manufacturer. After this period of transfection, the cell medium was replaced by supplemented medium (340 mOsM) or hyperosmolar medium (500 mOsM) for 4 h before collecting cell lysates for RT-qPCR or 24 h before collecting supernatants for ELISA or performing the cell viability test.
For experiments testing the effects of therapeutic agents or cellular pathway inhibitors, cells were treated for 1 h with various dilutions of therapeutic agents. CsA was tested at 0.1, 1 and 10 μg/mL; the highest concentration of 10 μg/mL corresponding to 1/100 and 1/50 of Ikervis® and Restasis®, respectively. Dex was tested at 10−10, 10−8 and 10-6M and Dox at 10 μg/mL. A mix of CsA 10 μg/mL and Dex 10–8 M was also tested. The inhibitors of p38 (SB203580), JNK (SP600125) and MEK/ERK (UO126) (inhibitor of MEK1 and hence of its downstream target ERK) were tested at 10 μM while the NFĸB inhibitor (PDTC) was tested at 50 μM. The concentrations tested were chosen regarding previous published studies [
For all these culture assays, we used one well for each condition concerning immunostaining (6-wells plate) or RTqPCR assays (24-wells plate) and at least two wells for ELISA and viability test (96-wells plate).
Cellular viability was analyzed using resazurin dye to measure the metabolic capacity of cells, which is an indicator of cell viability, using the CellTiter-Blue®assay (Promega). After an incubation period under the above-described stress conditions, CellTiter-Blue® reagent was added to the cell culture following the protocol described by the manufacturer. Fluorescence intensity was then quantified using a microplate reader with an excitation wavelength of 560 nm and an emission wavelength of 590 nm (Infinite M1000, Tecan, Lyon, France). Fluorescence values were then normalized with respect to control cells considered as 100% viable.
Enzyme-linked immunosorbent assays (ELISAs) for human CCL-2 were performed using the commercial DuoSet ELISA Development kit (R&D Systems, Minneapolis, MN, USA). After the incubation period under different stress conditions, cellular supernatants were collected, centrifuged to remove potential cellular fragments, and were stored at −80°C. ELISAs were performed on supernatants according to the manufacturer’s protocols. Absorbance was read at 450 nm using a microplate reader (Infinite M1000, Tecan) with a reference wavelength of 570 nm
Cells were grown on round sterile cover glasses (diameter, 14 mm; Menzel GmbH, Braunschweig, Germany). After the incubation period under different stress conditions, the cells were washed and fixed in 4% paraformaldehyde-PBS (Sigma-Aldrich). They were then permeabilized in a 0.3% Triton (Triton X-100, Sigma-Aldrich) solution for intracellular staining, followed by a 1% bovine serum albumin (Calbiochem, Merck Millipore, Darmstadt, Germany) incubation for 30 min and an overnight period with the primary antibody anti-NFAT5 at a final concentration of 2 μg.mL-1 (goat polyclonal antibody SC5499; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Cells were then incubated for 1 h with the secondary antibody Alexa Fluor 488 rabbit anti-goat (Invitrogen, Life Technologies). Cover glasses were then mounted with Mountant PermaFluor® (Thermo Fisher Scientific, Courtaboeuf, France) before observation with an epifluorescence microscope (Leica DM6000B, Rueil-Malmaison, France).
After the incubation period under different stress conditions, the cells were washed and lysed, and their total RNA was extracted using a NucleoSpin RNA II extraction kit (Macherey-Nagel, Düren, Germany). RNA content was measured using a NanoDrop detector (ND-1000 spectrophotometer) and cDNA was synthesized from equal amounts of RNA using Multiscribe reverse transcriptase (TaqMan Reverse Transcription Reagents, Applied Biosystems, Life Technologies). Concentrations of each sample were adjusted to 5 ng/μL of cDNA. The reaction mixture containing 25 ng of cDNA per well was preheated at 95°C for 10 min, followed by 40 cycles (95°C/15 s and 60°C/1 min). Each assay was normalized by amplifying the housekeeping cDNA GAPDH (ID Hs9999905). Target cDNA was amplified using the 7300 Real-Time PCR system (Applied Biosystems, Life Technologies) with assays-on-demand primers for human CCL2 (Hs00234140) and NFAT5 (Hs00232437) (Applied Biosystems, Life Technologies). Changes in mRNA expression were calculated according to the 2-ΔΔCT method (CT, cycle threshold), with ΔCT = CTtarget gene-CTgapdh and ΔΔCT = ΔCTstimulated-ΔCTcontrol.
All experiments were performed at least three times: immunostainings and RTqPCR assays were performed on one well per condition, in three separated experiments. Viability tests and ELISA were performed at least in two wells per condition, in three separated experiments. Their conditions were compared using one-way analysis of variance (ANOVA) followed by Dunnett’s multiple comparison test, or two-way ANOVA followed by Sidak’s or Tukey’s multiple comparison test (GraphPad, GraphPad Software, La Jolla, CA, USA). Only
Cells exposed to supplemented medium (340 mOsM) or different hyperosmolar conditions from 400 to 600 mOsM for 24 h were tested for cellular viability (A) with CellTiterBlue® assay and CCL2 secretion (B) by ELISA, and cells treated for 4 h were lysed for mRNA expression by RT-qPCR to determine the CCL2 and NFAT5 gene expression (C, D). Data show mean ± SEM,
After a 24-h treatment, conjunctival cellular viability significantly decreased in the 500-mOsM hyperosmolar condition, with 16% of cell viability loss compared to control (
Hyperosmolar conditions (500, 550 or 600 mOsM) applied for 24 h induced a significant secretion of CCL2 in supernatants by HeLa-modified conjunctiva-derived cells (235, 220, 148 pg/mL, respectively) (
CCL2 mRNA expression significantly increased in HO 450 mOsM, reaching a relative expression of 12.6 compared to the 340-mOsM control after 4 h (
Four hours in hypertonic conditions at 450 and 500 mOsM induced an increase in NFAT5 gene expression, 2.7- and 3.1-fold respectively, compared to the 340-mOsM condition (
Cells exposed to supplemented medium (340 mOsM) or the hyperosmolar condition (500 mOsM) for different times from 30 min to 24 h were tested for cellular viability (A) with CellTiterBlue® assay, CCL2 secretion (B) with ELISA and were lysed for mRNA expression using RT-qPCR to determine CCL2 and NFAT5 gene expression (C, D). Data showing mean ± SEM,
HeLa-modified conjunctiva-derived cells in a 500-mOsM medium started to significantly decrease their viability only after 9 h of stress (88% viability) (
Similarly, CCL2 secretion significantly increased after 9 h in 500 mOsM (1 pg/mL vs 52 pg/mL) (
CCL2 gene expression was also enhanced by the 500-mOsM medium with CCL2 mRNA expression increasing 14-fold compared to the 340-mOsM control after only 4 h in hyperosmolar conditions (
NFAT5 gene expression increased 3.1-fold compared to control cells after 4 h in hyperosmolar conditions (
Cells treated for 24 h with negative control siRNA or NFAT5 siRNA were then exposed to supplemented medium (340 mOsM) or hyperosmolar condition (500 mOsM). Cells exposed for 24 h were tested for cellular viability (A) with CellTiterBlue® assay, CCL2 secretion (B) by ELISA and cells exposed for 4 h were lysed for mRNA expression using RT-qPCR to determine the CCL2 and NFAT5 gene expression (C, D). Data showing mean ± SEM,
The cell viability test did not reveal any toxic effect when using NFAT5 siRNA versus control siRNA in basal conditions of 340 mOsM (
Concerning the CCL2 secretion in 340 mOsM, control and NFAT5 siRNAs did not induce significantly different effects, with a low level of CCL2 in supernatant (37 and 16 pg/mL, respectively) (
The same inhibition profile was observed with CCL2 gene expression. The CCL2 mRNA expression with control siRNA increased 14.3-fold in 500-mOsM conditions compared to 340-msOsM conditions (
The specificity of the NFAT5 siRNA used was confirmed by RT-qPCR of NFAT5 (
The lack of toxicity of siRNAs was also verified using the CellTiter-Blue viability test. It showed that control siRNA and NFAT5 siRNA had no effect on the viability of HeLa-modified conjunctiva-derived cells compared to cells without siRNA in classic medium (data not shown).
Cells were treated for 1 h with DMSO 0.1%, SB203580 10 μM, SP600125 10 μM, U0126 10 μM or PDTC 50 μM before adding supplemented medium 340 mOsM or hyperosmolar medium to reach 500 mOsM. Cells exposed to medium or HO for 24 h were tested for CCL2 secretion (A) by ELISA and cells exposed for 4 h were lysed for mRNA expression using RT-qPCR to determine the CCL2 and NFAT5 gene expression (B, C). Data showing mean ± SEM,
In hyperosmolar 500-mOsM conditions for 24 h, the control DMSO 0.1% induced CCL2 secretion in supernatants (257 pg/mL) (
The same CCL2 mRNA inhibition profile was observed in hyperosmolar conditions following addition of the inhibitors SB203580, SP600125 and PDTC with a 7.4-, 11.3- and 3.5-fold mRNA increase, respectively, compared to the 18.4-fold increase for DMSO (
Under 340-mOsM basal conditions, the compounds tested had no effect on either the production of CCL2 by HeLa-modified conjunctiva-derived cells or its gene expression.
Treating cells with PDTC increased NFAT5 gene expression 1.9-fold under basal 340-mOsM conditions (
None of the compounds tested had a negative effect on cellular viability compared to their respective control conditions: DMSO 0.1% in 340 mOsM or in 500 mOsM (data not shown).
Cells were treated for 1 h with DMSO 0.1%, CsA (0.1, 1 and 10 μg/mL), Dex (10−10, 10−8 and 10-6M), Dox 10 μg/mL or a mix of CsA 10 μg/mL and Dex 10-8M before adding supplemented medium 340 mOsM or HO to reach 500 mOsM. Cells exposed to medium or HO for 24 h were tested for CCL2 secretion (A) by ELISA and cells exposed for 4 h were lysed for mRNA expression using RT-qPCR to determine the CCL2 and NFAT5 gene expression (B, C). Data showing mean ± SEM,
After 24 h in 500-mOsM conditions, CCL2 secretion with control DMSO 0.1% reached 250 pg/mL, whereas after 4 h, CCL2 mRNA expression increased 12.2-fold (
HO conditions increased NFAT5 mRNA expression 3.2-fold. This HO-induced increase was significantly inhibited by Dex 10−8 and 10-6M (2.2-fold and 2.0-fold, respectively) (
Mixing Dex and CsA induced no effect that differed significantly from each of the conditions alone (statistics not shown; two-way ANOVA followed by Tukey’s multiple comparison test). Under 340-mOsM conditions, none of the compounds tested had any effect on the production and gene expression of CCL2, or on NFAT5 gene expression by HeLa-modified conjunctiva-derived cells. None of the compounds tested affected cellular viability compared to DMSO 0.1% conditions under 340 mOsM or 500 mOsM, respectively (data not shown).
Despite its worldwide high prevalence and its significant impact on everyday life for patients, DED actually has few symptomatic and curative treatments. It is therefore important to improve the understanding of the pathogenesis mechanisms with the ultimate goal of developing new targeted therapeutic strategies. Conjunctiva could be an important tissue to study as it covers the major part of the ocular surface and is the site where inflammatory reactions take place. Moreover, conjunctival cells themselves are able to secrete inflammatory cytokines and to participate in inflammatory processes [
We demonstrated that HeLa-modified conjunctiva-derived cells respond to hyperosmotic stress by inducing gene expression and secretion of CCL2, just as corneal cells or other cell types do [
CsA displayed a concentration-dependent inhibitory effect on the HO-induced CCL2 production in our model. This was also observed with Dex, an effect previously reported in a study on corneal cells [
The three MAPKs, p38, JNK and ERK, as well as the transcription factor NFκB are activated by HO in several types of mammalian cells [
HO is known to activate NFAT5, a transcription factor that plays a key role in osmoprotection by driving osmoprotective gene expression, thus promoting cell survival and protecting cells from the deleterious effects of shrinkage. In addition to this fundamental role, it has a wider range of functions, such as inflammatory cytokine production [
All these results obtained with HeLa-modified-WKD cells should be compared and confirmed with primary human conjunctival cells. Nevertheless, our results highlight the potentially important role of conjunctival cells in the response of the ocular surface to a modification of the tear film. They are able to secrete pro-inflammatory mediators that could participate in the vicious circle of DED by interfering with the inflammatory process stimulated by tear film HO. The inflammatory cellular pathway of NFAT5 translocation, a typical feature of hyperosmolar stress, leading to CCL2 secretion through other factors such as MAPKs and NFκB, constitutes a biological cascade with potentially promising new targets of interest for DED therapy.
(XLSX)
(XLSX)
(XLSX)
(XLSX)
(XLSX)
We thank Benedict O'Donnell and Linda Northrup for editing advice and Georges Kalouche for providing technical assistance for in vitro experiments. We also thank the Institut National de la Santé Et de la Recherche Médicale.