Cyclosporine A Ameliorates Inflammatory Responses in Dry Eye by Inactivating cGAS–STING Signaling Via the Initiation of Autophagy

Abstract

Purpose

The purpose of this research was to examine the roles and regulatory mechanisms of autophagy and the cyclic GMP-AMP synthase (cGAS)–stimulator of interferon gene (STING) pathway in dry eye (DE) inflammation and the mechanism by which cyclosporine A (CsA) attenuates the inflammatory response in dry eye disease (DED) by modulating autophagy and the cGAS–STING pathway.

Methods

Two DE models, hypertonicity-stimulated human corneal epithelial (HCE-T) cells and benzalkonium chloride (BAC)-treated C57BL/6 mice, were established to investigate the modulation of autophagy and the cGAS–STING pathway by CsA in DE. Changes in light chain 3 (LC3), P62, γ-H2AX, cGAS, and STING levels were detected via western blotting. Reverse-transcription quantitative PCR was used to detect changes in IL-6 and IL-1β mRNA levels after drug administration, and autophagosomes were observed via electron microscopy. The autophagy inhibitor 3-methyladenine (3-MA) was used to further clarify the effects of autophagy on the cGAS–STING pathway. Hematoxylin and eosin staining and corneal fluorescence staining were used to assess the signs of DE in the mice.

Results

Autophagy and the cGAS–STING pathway were significantly activated in hypertonicity-stimulated HCE-T cells and in the BAC-induced DE model. CsA activated autophagy and decreased the levels of γ-H2AX, cGAS, STING, IL-6, and IL-1β in the DE models. The autophagy inhibitor 3-MA reversed the inhibitory effects of CsA on the cGAS–STING pathway and the inflammatory response. Similarly, in vivo experiments revealed that CsA activated autophagy, inhibited the expression of components of the cGAS–STING pathway, promoted the restoration of corneal epithelial cell morphology, and reduced corneal fluorescence staining scores in mice with DE.

Conclusions

CsA inhibited the cGAS–STING pathway by activating autophagy, further reducing inflammation in DE models and attenuating the signs of DE in mice.

The percentage of the population suffering from dry eye disease (DED) worldwide ranges from 5% to 34%.¹ The development of DED is closely associated with modifications in the structure of the ocular surface, and changes in the homeostasis of the ocular surface ultimately lead to the development of DED.^2,3 Hyperosmolarity is considered a key factor in the physiopathology of severe DED.⁴ The homeostasis of corneal epithelial cells is disrupted under hyperosmotic conditions, which ultimately leads to intracellular DNA double-strand breaks.⁵ According to the results of clinical studies, the average tear film osmolality in the normal population is approximately 302 mOsm, whereas in patients with DED the mean tear film osmolality is approximately 322 mOsm.⁶ Severe DED can contribute to corneal damage and even loss of vision.^7,8

The cyclic GMP-AMP synthase (cGAS)–stimulator of interferon gene (STING) signaling pathway is an intracellular immune response-related pathway that is closely related to pathological processes.^9,10 cGAS recognizes DNA fragments and induces the generation of cyclic guanosine monophosphate–adenosine monophosphate (cGAMP), which binds to STING on the endoplasmic reticulum, leading to a conformational change in the STING protein that drives interferon expression in the nucleus.¹¹ In addition, STING can also activate the transcription factor nuclear factor kappa B (NF-κB), causing a series of downstream inflammatory responses and potentially multiple complications.¹² A large amount of DNA derived mainly from detached corneal and conjunctival epithelial cells appears on the surface of the eyes of individuals who have severe dry eye.¹³ The cGAS–STING pathway also plays a significant role in the induction of environmental dry eye inflammation.¹⁴ Abnormal activation of the cGAS–STING pathway induces pathological cellular damage and triggers chronic or systemic inflammation.^15,16 Therefore, drugs that target the cGAS–STING signaling pathway have good potential for clinical application and may be treatments for ocular surface inflammatory diseases.

Autophagy plays a significant role in regulating the homeostasis of organisms through the degradation and recycling of nonessential components.^17,18 Microtubule-associated protein light chain 3 (LC3) lipidation is essential for any type of autophagy; LC3-II, a lipidated form of LC3-I, has been shown to be a marker of autophagosomes in mammals.^19,20 The convertor protein P62 (sequestosome-1; SQSTMl) is also closely related to autophagic activity, and the P62 level reflects the extent of P62 degradation.²¹ The commonly used autophagy inhibitor 3-methyladenine (3-MA), which acts on phosphoinositide 3-kinase (PI3K), inhibits autophagosome formation.²² Cytosolic autophagy is closely related to inflammation, and a loss of autophagy can lead to cell death, which can trigger inflammatory responses in phagocytes via cytoplasmic surface receptors.²³ Autophagy, inflammation, and immune responses regulate each other and act in various ways to influence the stability of biological functions.^24,25 In addition to DED, autophagy has been shown to be involved in pathophysiological responses and repair in age-related macular degeneration,²⁶ keratitis,²⁷ and glaucoma.²⁸ However, the role of autophagy in the regulation of the cGAS–STING pathway in DED has remained unknown. Here, we performed a series of numerous experiments to evaluate the changes in and functional mechanisms of autophagy and the cGAS–STING pathway in DED.

Cyclosporine A (CsA) has some therapeutic effects on dry eye (DE). Our previous studies revealed that 0.05% CsA could contribute to the suppression of inflammation to treat DED by inhibiting the HMGB1/TLR4/NF-κB signaling pathway; however, the levels of inflammatory factors increased when the cells were stimulated with an excessively high concentration of CsA, resulting in a negative effect that promoted the development of inflammation.²⁹ Although the pharmacological effects of CsA have been continuously explored, the underlying mechanisms remain incompletely understood.³⁰ The U.S. Food and Drug Administration (FDA) has approved 0.05% CsA eye drops for the management of DED.⁴² In the present study, we first investigated the changes in autophagic activity and expression of the cGAS–STING signaling pathway components in HCE-T cells exposed to hyperosmolarity and then explored the potential mechanisms by which CsA affects autophagic activity in DED. In the present study, we further elucidated the anti-inflammatory effect of CsA on DED and its therapeutic mechanism. The development of drugs targeting the autophagy/cGAS–STING signaling pathway may lead to the identification of new treatments for DED and could help treat other inflammatory illnesses.

Materials and Methods

Reagents and Antibodies

CsA (Solarbio, Beijing, China) was dissolved in dimethyl sulfoxide (DMSO; Solarbio) at a concentration of 10 mM, and the cells were stimulated with isotonic medium diluted to the desired concentration. A total of 500 mg of 3-MA (MedChemExpress, Monmouth Junction, NJ, USA) was dissolved in 67.05 mL of sterilized phosphate-buffered saline (PBS) and frozen. A 0.2% benzalkonium chloride (BAC) solution was prepared by diluting 50% solubilized BAC (Sigma-Aldrich, St. Louis, MO, USA) with pure sterile water. The concentration of DMSO in the final prepared solution was less than 0.1%. Primary rabbit antibodies against SQSTM1/P62 (1:1000, #5114; Cell Signaling Technology, Danvers, MA, USA), LC3B (1:1000, #43566; Cell Signaling Technology), cGAS (1:1000, 15102; Cell Signaling Technology), STING (1:1000, #13647; Cell Signaling Technology), γ-H2AX (1:10,000, ab81299; Abcam, Cambridge, UK), and β-actin (1:10,000, AF7018; Affinity Biosciences, Cincinnati, OH, USA) were used.

Animal Experiments

The animal experiments were approved by the Animal Care and Utilization Committee of Yantai Yuhuanding Hospital and were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. For this project, Jinan Pangyue Laboratory Animal Breeding Co., Ltd. (Jinan, China) provided C57BL/6J female mice. All of the mice were healthy, weighed from 25 to 30 grams, had normal eyes, were 6 to 8 weeks old, and were of specific-pathogen-free (SPF) grade. The mice were housed in a standard environment at 21°C to 25°C with a 12-hour/12-hour light/dark cycle and 60% ± 10% humidity throughout the study period. Topical administration of BAC solution has been recognized as a commonly used method for establishing DE models.³¹ A pipettor (Eppendorf, Germany) was used to transfer 2 µL of 0.2% BAC solution to the conjunctival sac of each mouse twice daily for 10 days to construct the mouse DE model.

Following establishment of the DE model, the mice were split into two groups of 10 mice each at random. In the DE group, 2 µL of PBS (containing 0.01% DMSO) was applied topically three times daily for 14 days. In the DE + CsA group, 2 µL of 0.05% CsA (containing 0.01% DMSO) was applied topically three times daily for 14 days. The mice in the negative control (NC) group were administered 2 µL of PBS only twice daily for the first 10 days and then 2 µL of PBS containing 0.01% DMSO three times daily for 14 days. The mice were injected intraperitoneally with 300 µL of a tribromoethanol solution (Aibei Biotechnology, Nanjing, China) for anesthesia.

Cell Experiments

In this study, an immortalized human corneal epithelial cell line, HCE-T (Shanghai Fuheng Biotechnology Co., Ltd., Shanghai, China), was used at passages 6 to 10. The cells were cultured in HCE-T–specific complete medium. Different masses of sodium chloride powder were added to the isotonic media to prepare the hypertonic media (450, 500, and 550 mOsm/L). HCE-T cells were cultured in hypertonic medium to construct an in vitro model of DE.^32,33 The cells were incubated in hypertonic medium with or without CsA and prestimulated with 3-MA (10 mM) 2 hours before the in vitro experiments were performed.

Western Blotting

Mouse corneal tissues were thoroughly ground and placed in Eppendorf (EP) tubes (Crystalgen, Commack, NY, USA). Radioimmunoprecipitation buffer (RIPA; Solarbio) containing phenylmethylsulfonyl fluoride (PMSF; RIPA:PMSF = 100:1) was added, and the contents were mixed with a vortex mixer. Adherent cells were digested with trypsin in EP tubes, centrifuged for 15 minutes at 4°C and 12,000g,⁵² and lysed with the same ice-cold lysis solution. The EP tubes were placed in an ice box with continuous shaking for 30 minutes. Finally, the protein samples were centrifuged for 15 minutes at 4°C and 12,000g. Afterward, the protein concentration was measured via a Bicinchoninic Acid Assay Kit (Beyotime, Shanghai, China). The protein samples were transferred electrophoretically onto nitrocellulose (NC) membranes (MilliporeSigma, Billerica, MA, USA) under standard conditions.⁵² The diluted primary antibodies were incubated with the NC membranes overnight. The membrane was rinsed with Tris-buffered saline with Tween-20 (TBST) three times for 10 minutes each time and then incubated with 5% skim milk powder and 1:5000 diluted secondary antibody. Goat anti-rabbit IgG (H+L) (Affinity Biosciences). An enhanced chemiluminescence (ECL) solution (Affinity Biosciences) was evenly added dropwise to the NC membranes, which were then photographed for analysis. The relative expression of each protein was calculated via ImageJ (National Institutes of Health, Bethesda, MD, USA).

Reverse-Transcription Quantitative PCR

Corneal tissues or HCE-T cells were transferred to EP tubes and lysed on ice after the addition of TRIzol Reagent (Tianmu Biologicals, Beijing, China). RNA purity and concentration were assessed with a NanoDrop 2000/2000C spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). For each well, the required amount of RNA was calculated, and the remaining volume was brought up using ribonuclease-free purified water, followed by the addition of the G592 reverse transcription reagent from Applied Biological Materials (Vancouver, BC, Canada). The reaction reagents were added and placed in a reverse-transcription PCR (RT-PCR) instrument. After three independent experiments, the PCR data were analyzed using the 2^–∆∆Ct method.⁵² The primer sequences utilized in this investigation are presented in the Table.

Table.

Primer Sequences for Human Genes

Gene Name	Forward	Reverse
IL-1β	5′-GCCAGTGAAATGATGGCTTATT-3′	5′-AGGAGCACTTCATCTGTTTAGG-3′
IL-6	5′-AAATCACCATGCACCTCATCC-3′	5′-AGAGGATTGTGCCCGAACTAAA-3′
GAPDH	5′-CAACGTGTCAGTGGTGGACCTG-3′	5′-GTGTCGCTGTTGAAGTCAGAGGAG-3′
Primer sequences for mouse genes
IL-1β	5′-GGCAACTGTTCCTGAACTCAACT-3′	5’-CCATTGAGGTGGAGAGCTTTCAG-3′
IL-6	5′-ATGAAGTTCCTCTCTGCAAG-3′	5’-TTAGCCACTCCTTCTGTGAC-3′
GAPDH	5′-AAGAAGGTGGTGAAGCAGG-3′	5’-GAAGGTGGAAGAGTGGGAGT-3′

Corneal Fluorescein Staining

Corneal fluorescein staining was performed to assess the extent of damage to mouse corneal epithelial cells. The mice were injected intraperitoneally with 300 µL of a tribromoethanol solution. Following complete anesthesia, 2 µL of a 0.25% fluorescein sodium solution (Macklin, Shanghai, China) was dropped into the conjunctival spaces of the mice, and the sac was washed with sterile saline 1 minute later. Afterward, the samples were observed under a slit-lamp microscope, and images were captured with a camera using cobalt blue light (Carl Zeiss Meditec, Oberkochen, Germany).

Each of the four quadrants constructed from the mouse corneas was assigned a score between 0 and 4. Corneal fluorescence staining was graded by one observer in a blinded manner, and the total score was estimated for each quadrant. Each eye in each group received a staining score as follows: 4, fluorescein-positive plaques; 3, highly dense fluorescence staining; 2, a dense speckled pattern; 1, comparatively sparse staining; and 0, no fluorescence.³⁴

Cell Activity Assay

A Cell Counting Kit-8 (CCK-8; DOJINDO Laboratores, Kumamoto, Japan) was used to assess cell viability following stimulation with different concentrations of hypertonic media and the toxic effects of various CsA concentrations on cells. The cell inhibition rate was calculated as follows: Cell inhibition rate = [(Ac – As)/(Ac – Ab)] × 100%, where Ac represents control wells (medium containing cells, CCK-8, and no CsA), As represents experimental wells (medium containing cells, CCK-8, CsA), and Ab represents blank wells.

Corneal Hematoxylin and Eosin Staining

After anesthesia, the mice were sacrificed via cervical dislocation, and the eyeballs were removed. The extracted tissues were rinsed with PBS and fixed with 4% paraformaldehyde. The dehydrated and transparent eye tissues were transferred to liquid paraffin, and a small drop of paraffin was added to the mold as a base layer, which was slightly solidified and set aside. Preheated tweezers were used to remove the eyeball from the wax dip box and quickly place it into the center of the mold. The melted paraffin was slowly poured into the mold (Citotest Scientific, Jiangsu, China), covering the eyeball completely. When the paraffin had completely solidified, the wax block was removed. The eyeball was routinely cut into 4-µm-thick paraffin sections using a slicer (RM2125 RTS; Leica Biosytems, Nussloch, Germany). After staining, the sections were observed under an inverted microscope (AxiovertA1; ZEISS Microscopy, Jena, Germany) and photographed.

Electron Microscopy Analysis of Cells

HCE-T cells were cultured in hypertonic medium (500 mOsm/L) and treated with or without 6 CsA for 24 hours, after which the cells were digested with trypsin (Thermo Fisher Scientific) and collected in EP tubes. Electron microscopy fixative (Servicebio, Wuhan, China) was added, and the samples were fixed at 4°C for 2 hours. Postfixation with 1% osmium acid (Ted Pella, Redding, CA, USA) was performed. The samples were stained with 2% uranyl acetate (SPI, Tianjin, China) at room temperature, and images were collected using a transmission electron microscope (HT7800/HT7700; Hitachi, Tokyo, Japan) after sectioning.

Data Analysis

Prism 8.0 (GraphPad, Boston, MA, USA), Illustrator (Adobe, San Jose, CA, USA), and ImageJ were used to analyze the data. The Kruskal–Wallis test was used for multiple analyses of data that were not normally distributed. One-way ANOVAs with the Bonferroni correction for multiple comparisons were used to compare more than two groups. P < 0.05 was considered to indicate statistical significance.

Results

Analysis of HCE-T Cell Viability and the Toxic Effects of Different Concentrations of CsA on Cells Cultured in Media With Different Osmotic Pressures Via the CCK-8 Assay

Cell viability in hypertonic medium and the toxic effects of CsA on cells were measured via CCK-8 assay. Compared with that of the 312-mOsm/L isotonic stimulation group, the apparent (by inspection) slight decrease in cell viability was not statistically significant (P > 0.05) (Fig. 1A); however, the decrease in cell viability in the 500-mOsm/L and 550-mOsm/L hypertonic stimulation groups was statistically significant (P < 0.05) (Fig. 1A). We also found that the cell viability of the 550 mOsm/L hyperosmotic stimulation group decreased by more than 40%, indicating that cell viability was strongly affected by stimulation with 550-mOsm/L medium.

Figure 1.

A CCK-8 assay was performed to measure the viability of HCE-T cells cultured in media with different osmolarities and evaluate the toxic effects of different concentrations of CsA on HCE-T cells. (A) Effects of media with different osmolarities (312, 450, 500, or 550 mOsm/L) on HCE-T-cell activity after 24 hours of culture (n = 3/group). (B) Inhibitory effects of different concentrations (1, 3, 6, and 9 µM) of CsA on HCE-T cell proliferation (n = 3/group). The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

In addition, cells maintained at 312 mOsm/L were exposed to various CsA concentrations, and no significant difference in the cytotoxic effects was observed between 1 and 3 µM CsA (P > 0.05) (Fig. 1B). However, a decrease in the proliferative capacity of cells treated with 6- and 9-µM CsA was detected, indicating a significant difference in the cytotoxic effects of these concentrations of CsA (P < 0.05) (Fig. 1B). We observed a significant decrease in cell proliferation after treatment with 9-µM CsA.

Autophagy and the cGAS–STING Pathway Are Activated in HCE-T Cells Upon Hyperosmotic Stimulation

Autophagic activity is often assessed by measuring the expression of LC3-II and sequestosome-1 (SQSTM1)/P62.⁵³ Elevated levels of P62 are generally considered an indicator of inhibited autophagy activity, and elevated levels of LC3-II are generally considered to indicate the activation of autophagy. The ratio of LC3-II expression to the expression of an internal reference is commonly used to assess autophagic activity in current studies.³³

Compared with those in the 312-mOsm/L isotonic stimulation group, HCE-T cells in the 450-mOsm/L, 500-mOsm/L, and 550-mOsm/L hypertonic stimulation groups presented increased levels of LC3-II and decreased levels of P62 expression (P < 0.05) (Figs. 2A–C), and the differences were statistically significant. We noted that the protein expression levels of LC3 and P62 varied in an osmotic pressure–dependent manner. These results indicate that the autophagic activity of cells gradually increases with increasing osmotic pressure.

Figure 2.

Levels of the LC3 and P62 proteins in HCE-T cells cultured in hypertonic media. (A–C) The expression levels of the P62 and LC3 proteins in the 450-mOsm/L, 500-mOsm/L, and 550-mOsm/L hypertonic stimulation groups and in the 312-mOsm/L isotonic stimulation group were assessed by western blotting (n = 3/group). (D–F) Western blotting was performed to quantify the expression of the LC3 and P62 proteins in each hypertonic stimulation group, with β-actin used as an internal reference protein (n = 3/group). The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

According to the results of the CCK-8 assay and western blot analysis presented above, autophagy was significantly increased in human corneal epithelial cells in the 500-mOsm/L hyperosmotic stimulation group. Given that the survival rate of cells in the 500-mOsm/L hyperosmotic stimulation group was greater than 80%, we cultured HCE-T cells in 500-mOsm/L hyperosmotic medium to construct an in vitro experimental model for subsequent drug experiments.

Next, we explored the effect of time on the protein expression levels of LC3 and P62. HCE-T cells were stimulated for 0, 8, 16, or 24 hours (the last time point measured was 24 hours) in 500-mOsm/L medium. The results showed that the protein level of LC3-II increased over time, reaching the highest level at 24 hours of stimulation with the hypertonic solution, whereas the expression of P62 decreased over time, reaching the lowest level after 24 hours of stimulation (P < 0.05) (Figs. 2D–F).

The cGAS–STING pathway, an initial signaling pathway involved in the immune/inflammatory response, has garnered much attention in recent years.¹⁵ We explored in detail the changes in the expression of cGAS–STING pathway components in cells after culture in media with different osmotic pressures at different time points. The results revealed that the protein expression levels of cGAS, STING, and γ-H2AX were significantly increased following hyperosmotic stimulation for 24 hours (P < 0.05) (Figs. 3A, 3C–E). Notably, the increase in STING expression appeared to peak at 500 mOsm/L. We guess that STING activity may be influenced by factors that are released during cell death. Considering that excessive damage in cells exposed to 550 mOsm/L could alter the expression of these factors, we excluded this osmolarity from subsequent drug experiments. Compared with the levels in the isotonic stimulation group, the cGAS, STING, and γ-H2AX protein expression levels in the 500-mOsm/L hypertonic stimulation group were significantly higher at 8 hours, 16 hours, and 24 hours (P < 0.05 (Figs. 3B, 3F–H). In conclusion, the experimental results suggest that hypertonic stimulation promotes autophagic activity in HCE-T cells and simultaneously activates the cGAS–STING pathway.

Figure 3.

cGAS, STING, and γ-H2AX protein expression levels in HCE-T cells cultured in hypertonic medium. (A, C–E) Expression levels of the cGAS, STING, and γ-H2AX proteins in the 312-mOsm/L isotonic stimulation group and the 450-mOsm/L, 500-mOsm/L, and 550-mOsm/L hypertonic stimulation groups (n = 3/group). (B, F–H) Western blotting was performed to evaluate the levels of the cGAS, STING, and γ-H2AX proteins in each hypertonic stimulation group (n = 3/group). The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

CsA Activates Autophagy in HCE-T Cells

The cells were divided into five groups to investigate the effect of CsA on autophagy in hyperosmolarity-stimulated HCE-T cells: normal control (NC), DE, DE + 3-µM CsA, DE + 6-µM CsA, and DE + 9-µM CsA groups. Compared with the DE group, the DE + 6-µM CsA group presented significantly higher levels of the LC3 protein and significantly lower levels of the P62 protein (P < 0.05) (Figs. 4A–C). Notably, after treatment with 9-µM CSA, this effect appeared to be reversed. Therefore, 6 µM was ultimately selected as the drug concentration of CsA for subsequent experiments to improve its efficacy. At present, the gold standard for measuring autophagy levels is to observe the autophagy–lysosome system. Using electron microscopy, we found that the cytoplasm of the CsA-treated HCE-T cells appeared to contain more autophagosomes than the cytoplasm of the control HCE-T cells (Fig. 4D). These results indicate that, at a concentration of 6 µM, CsA activates autophagy in HCE-T cells. However, at 3 and 9 µM, the activation was not significantly different from that at baseline in the DE model.

Figure 4.

LC3 and P62 protein expression levels in CsA-pretreated HCE-T cells. (A–C) Expression levels of LC3 and P62 in HCE-T cells stimulated with different concentrations of CsA (3, 6, or 9 µM) (n = 3/group). (D) Numbers of autophagosomes and autophagolysosomes in HCE-T cells observed via transmission electron microscopy. Red arrows represent autophagosomes, and red circles represent autophagic lysosomes. The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

CsA Inhibits the Expression of cGAS–STING Pathway Components and Inflammatory Proteins in HCE-T Cells

The western blot results showed that, when HCE-T cells were exposed to the hypertonic solution at 500 mOsm/L, the levels of cGAS, STING, and γ-H2AX increased, whereas the activation of autophagy by 6-µM CsA inhibited this effect of hypertonicity, suggesting that intracellular inflammatory pathways were inhibited and that the degree of cellular damage was decreased. In addition, when HCE-T cells were pretreated with the autophagy inhibitor 3-MA at 10 mM for 2 hours prior to being pretreated with 6-µM CsA, pretreatment with 10-mM 3-MA for 2 hours significantly inhibited the promotion of autophagy by CsA, resulting in attenuation of the inhibitory effect of CsA on the cGAS–STING pathway (P < 0.05) (Figs. 5A–G).

Figure 5.

CsA inhibits the expression of cGAS–STING pathway components and inflammatory factors by activating autophagy. (A–G) LC3, P62, STING, cGAS, and γ-H2AX protein levels were assessed by western blot analysis. (H, I) IL-1β and IL-6 mRNA expression levels in the NC, DE, CsA + DE, and CsA + 3-MA + DE groups (n = 3/group). The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

According to the RT-PCR results, the mRNA expression levels of IL-1β and IL-6 in the DE group of HCE-T cells were significantly higher than those in the NC group. Compared with the DE group, the DE + CsA group presented considerably lower IL-1β and IL-6 mRNA levels. The DE + CsA + 3-MA pretreatment group presented considerably higher mRNA levels of IL-1β and IL-6 than the DE + CsA group (P < 0.05) (Figs. 5H–I) and higher mRNA levels than the 500-mOsm/L–treated group that was not exposed to CSA. Our experiments reveal that HCE-T cells exposed to hypertonic stimulation express multiple inflammatory factors, which provides evidence for the success of our modeling process. CsA may inhibit the expression of the inflammatory factors IL-1β and IL-6, possibly by increasing autophagic activity, and the inhibitory effect of CsA on the expression of inflammatory factors in HCE-T cells after hypertonic stimulation can be reduced by a pretreatment with the autophagy inhibitor 3-MA. Thus, we speculate that some concentrations (including 6 µM) of CsA exert the protective effect of CsA by activating autophagy to inhibit the cGAS–STING pathway and ultimately reduce the inflammatory response.

CsA Improves the Ocular Surface of Mice With DED

We examined the effects of CsA on the ocular surface of mice with DE by topically applying 0.05% CsA eye drops on a continuous basis. Fluorescence images of the corneas of the three groups of mice were captured under a slit lamp with cobalt blue light (Fig. 6A). The corneal fluorescein staining score was noticeably lower in the DE + CsA group than in the DE group (P < 0.0001) (Fig. 6B). Hematoxylin and eosin (H&E) staining revealed that the corneal epithelial cells of the normal mouse cornea were tightly and neatly arranged. The corneal epithelium in the DE group of mice exhibited significant thinning compared to the NC group, accompanied by corneal epithelial cell desquamation, and the DE + CsA group demonstrated markedly thickened corneal epithelium relative to the DE group (Fig. 6C). The experimental results showed that CsA greatly alleviated but did not completely reverse the ocular signs of DED in the mice.

Figure 6.

Corneal fluorescence scores and H&E staining of corneal tissues from DE model mice. (A) Photographs of corneal fluorescence staining in mice from the NC group, DE group, and DE + CsA group. (B) Comparison of the corneal fluorescence staining scores among the NC, DE, and DE + CsA groups (n = 10/group). (C) H&E staining of mouse corneal tissues. The arrow (→) indicates detached epithelial cells. The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

CsA Affects the Expression of Autophagy/cGAS–STING Pathway–Related Proteins in Mice With DE

Following the successful generation of a mouse model of DE, corneal tissues were collected from the three groups of mice, and the protein levels of LC3, P62, cGAS, STING, and γ-H2AX were assessed in corneal tissues from each group for comparison. The mRNA expression levels of the inflammatory factors IL-6 and IL-1β in mouse corneal tissues were assessed via RT-PCR. Western blotting showed that, compared with the levels in the corneal tissues of the mice in the DE group, autophagy was significantly increased (P < 0.05) (Figs. 7A–C), and the protein levels of cGAS, STING, and γ-H2AX were markedly reduced (P < 0.05) (Figs. 7D, 7E) in the corneal tissues of the mice from the 6-µM CsA–treated group. According to the RT-PCR data, the corneal mRNA expression levels of IL-6 and IL-1β in the mice from the DE + CsA group were significantly lower than those in the mice from the DE group (P < 0.001) (Figs. 7F, 7G). CsA appears to attenuate ocular surface damage in mice with DE via the autophagy/cGAS–STING pathway.

Figure 7.

Effects of 6-µM CsA on the expression of autophagy/cGAS–STING pathway–associated factors in the corneal cells of DE model mice. (A–E) LC3, P62, cGAS, STING, and γ-H2AX protein levels were measured in each group by western blotting. (F, G) RT-PCR was used to quantify the mRNA expression of IL-6 and IL-1β in each group (n = 3/group). The data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Discussion

DED is a very common chronic disease affecting the ocular surface, and many studies have confirmed that the pathophysiological mechanisms of DED are complex and varied.³⁵ The pathogenesis of DED needs to be investigated in depth to identify more effective methods to prevent its occurrence. The objective of this study was to examine the functions of autophagy and the cGAS–STING pathway in DED.

We noted that the expression of the autophagy-related factor LC3 was dramatically increased in both in vitro and in vivo DED models. These changes were followed by a large decrease in P62 expression and significant increases in cGAS, STNG, and γ-H2AX levels. CsA decreased the levels of cGAS, STING, and γ-H2AX, as well as expression of the inflammatory factors IL-1β and IL-6, and enhanced autophagy in HCE-T cells and the corneas of mice with DE. CsA improved the symptoms of DE in a mouse model of BAC-induced DE by reducing the corneal staining scores and corneal epithelial damage. Thus, these data suggest that autophagy and the cGAS–STING signaling pathway play regulatory roles in both in vitro and in vivo models of DE and that CsA may alleviate DE symptoms by increasing autophagy and inhibiting the cGAS–STING pathway.

Studies have shown that the synergistic regulation of autophagy and the inflammatory response is a pathogenic mechanism of DED. A previous study revealed that the expression of the autophagy-related proteins autophagy-related protein 5 (ATG5) and LC3-II/I is increased in the tears of patients with Sjögren syndrome related dry eye disease (SSDED).³⁶ Ma et al.²⁵ observed many autophagosomes in DE model mice. Increased levels of the autophagy markers LC3 (autophagosome marker) and lysosomal-associated membrane protein 1 (LAMP1; lysosome marker) were detected in human corneal epithelial cells under drying stress.³⁷ Elevated expression levels of LC3-II, SQSTM1, and DNA damage-inducible transcript 4 (DDIT4) were observed in mouse corneal epithelial cells in vitro and in a mouse model of DE.³⁸ These results suggest that the activation of autophagy is an adaptive response to DED. Our experimental study revealed that, compared with those in HCE-T cells in the 312-mOsm/L isotonic stimulation group, the LC3-II expression levels in HCE-T cells in the hypertonic stimulation groups increased, and the P62 expression level decreased, indicating that autophagy activity gradually increased as osmolarity increased. In addition, cellular autophagy increased with time and reached the highest level at 24 hours, the last time point evaluated during hypertonic solution stimulation. In addition, autophagy was significantly increased in BAC-induced DE model mice compared with normal mice. This result is consistent with previous research findings.³³

cGAS–STING signaling in DE has been relatively rarely studied and should be explored in depth. The histone H2AX plays an important role in DNA damage repair.^39,40 Following DNA damage and breakage, cGAS is converted into γ-H2AX, and the degree of DNA damage is positively correlated with the level of γ-H2AX, a marker of DNA damage.⁴¹ When stress occurs, cGAS binds to damaged DNA fragments and eventually initiates an immune–inflammatory response in cells.⁴¹ Our results revealed that hyperosmotic stimulation upregulated the expression of γ-H2AX, cGAS, and STING in a time-dependent manner, suggesting that hyperosmotic stimulation can lead to cellular DNA damage and that the more persistent (up to 24 hours) the osmotic stress is, the more severe the damage.

At lower doses, CsA has immunomodulatory properties, and the topical administration of low concentrations of CsA is commonly used to treat DE and other ocular surface diseases⁴²; however, the mechanism underlying the anti-inflammatory effect of CSA has not been precisely elucidated to date. Studies have shown that CsA induces endoplasmic reticulum stress and thus activates autophagy in malignant glioma models.⁴³ Thus, we stimulated HCE-T cells with various concentrations of CsA and discovered that, when the concentration of CsA reached 9 µM, the rate of HCE-T cell inhibition greatly increased. As its concentration increased, CsA had a strong effect on HCE-T cell proliferation. Moreover, we discovered that, when the cells were stimulated with varying concentrations of CSA (3, 6, or 9 µM), the effect of CSA on the activation of autophagy in the hyperosmotic environment-induced DE model appeared to be greatest at a concentration of 6 µM, whereas the ability of 9-µM CsA to activate autophagy was diminished, which might have been related to the toxicity of the drug. These experimental results support those of a recent study that showed that higher levels of CsA had toxic effects on cells.⁴⁴ Hypothesizing that 0.05% CsA eye drops would maximize the effect of CsA, 0.05% CsA eye drops were applied to the conjunctival sac of mice with DE.⁴⁵ In our study, topical 0.05% CSA decreased DE-induced corneal fluorescein staining.

Next, we explored how autophagy plays a role in DED. cGAS–STING signaling is closely related to autophagy signaling. cGAS–STING signaling not only induces autophagy as a cell-autonomous defense mechanism but is also regulated by autophagy-related proteins. cGAS–STING–induced autophagy limits viral propagation,⁴⁶ which can drive autophagic cell death, thereby preventing tumor cell growth.⁴⁷ Autophagy also regulates STING activity by assisting in its intracellular transport and lysosomal degradation.^48–50 On the one hand, STING promotes the release of IFN and proinflammatory factors through the cGAS–STING pathway⁵⁴; on the other hand, activated STING interacts with LC3 to initiate autophagy, which results in the degradation of STING and prevents the sustained activation of immune responses.⁵⁵ Recent studies have shown that signaling molecules in the cGAS–STING pathway interact with ATG and that autophagy regulates the cGAS–STING pathway by degrading key proteins in this pathway.⁴⁹

In this study, we used CsA to activate autophagy and decrease the expression of γ-H2AX, cGAS, and STING. The inhibitory effects of CsA on the levels of γ-H2AX, cGAS, and STING were reversed by the 3-MA–mediated suppression of autophagy, suggesting that autophagy controls the expression of cGAS and STING. A recent study demonstrated that the activation of autophagy protects HCE-T cells from hypertonic stress-induced inflammation.⁵¹ Specifically, autophagy induction reduces the levels of inflammatory mediators and promotes cell survival in an experimental model of DE, and the present study supports these findings.

Overall, this study shows that autophagy and the cGAS–STING signaling pathway may be involved in the pathogenic process of DED caused by a hypertonic solution. Both in vitro and in vivo models of dry eye involve the autophagy/cGAS–STING signaling pathway. CsA stimulated autophagy in hypertonically stressed HCE-T cells, and BAC induced DE in the mouse cornea to reduce the expression of cGAS–STING pathway components and alleviate DE signs. We acknowledge that our study has certain limitations. Specifically, the complex interplay between autophagy and the cGAS–STING pathway, as well as their broader network interactions with other signaling pathways, warrant further systematic investigation. The results of this investigation could provide a theoretical foundation for the clinical application of medications that target the autophagy/cGAS–STING pathway in the treatment of DED. This study also provides new insights and justification for further investigations into the etiology and management of DED.