Materials

Gallic acid (purity ≥ 99%) was purchased from J&K Chemicals CO (Beijing, China). LPS and epidermal growth factor were obtained from Sigma-Aldrich (St. Louis, MO, USA). HCECs and RAW264.7 macrophages were obtained from the American Type Culture Collection (Manassas, VA, USA). Insulin was obtained from Gibco (St. Louis, MO, USA). The 2',7'-Dichlorodihydrofluorescein diacetate (DCFH-DA) and cell counting kit-8 were acquired from Beyotime Biotechnology (Nanjing, China). The IL-6 (DY406) and TNF-α (DY410) DuoSet enzyme-linked immunosorbent assay (ELISA) kits were purchased from R&D Systems (Minneapolis, MN, USA). Phosphatase inhibitor was obtained from EpiZyme (Shanghai, China). Antibodies against p-P65 (#3033S, 1:1000), P65 (#8242S, 1:1000), phospho-IκB-α (#2859S, 1:1000), IκB-α (#4814S, 1:1000), β-actin (#3700, 1:5000), HO-1 (#48768, 1:1000), and NQO-1 (#48768, 1:1000) were purchased from Cell Signaling Technology (Danvers, MA, USA). Nrf2 (ab137550, 1:1000) and phospho-Nrf2 (ab76026, 1:1000) antibodies were obtained from Abcam (Cambridge, UK). Antibodies against NF-κB P65 (66535, 1:200), Nrf2 (16396, 1:200), and fluorescein (FITC) (SA00003, 1:100) were purchased from Proteintech (Chicago, IL, USA). Hematoxylin–eosin (H&E) and periodic acid-Schiff (PAS) staining kits were obtained from Solarbio Life Sciences (Beijing, China). An in situ cell death detection kit was obtained from Roche (Mannhein, Germany). The BCA kit, Griess reaction assay kit, antifade mounting medium with DAPI and protease inhibitor cocktail were obtained from Beyotime Biotechnology (Shanghai, China).

Cell culture

HCECs were cultured in Dulbecco’s modified Eagle’s Medium/F12 (DMEM/F12) supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin, 10 ng/mL epidermal growth factor, and 100 ng/mL insulin. RAW264.7 macrophages were cultured in DMEM supplemented with 1% penicillin/streptomycin and 10% FBS.

In vitro hyperosmotic stress model and inflammation model

To test drug efficacy in vitro, we established two cell models related to the pathogenesis of DED. First, we simulated the state of HCECs in a hypertonic environment by adding sodium chloride (NaCl) to the culture medium. After stable subculture, HCECs were seeded onto a 6-well plate and switched to serum-free medium (DMEM/F12 without FBS) for 24 h. We cultured the cells in the medium for another hour by increasing the osmolality to 500 mOsm by adding an additional 90 mM NaCl to a serum-free medium with an osmolality of 320 mOsm. In addition, cells cultured in 500 mOsm medium were pre-treated with or without 100 μM GA (dissolved directly in the serum-free medium), which was added 23 h prior to NaCl supplementation.

We then stimulated macrophages with LPS to establish a classical model of inflammation. The RAW264.7 cells were seeded onto a 6- or 24-well plate and cultured in DMEM containing 10% FBS to achieve cell monolayers that were approximately 60–70% confluent after stable subculture. Subsequently, the cells were treated with 100 μM GA (dissolved directly in DMEM containing 10% FBS) for 1 h (ELISA and Griess reaction assay) or 16 h (fluorescence assay and Western blot analysis), followed by stimulation with 1 μg/mL LPS. Finally, P65 and p-P65 protein expression, phosphorylated-IkB-α and IkB-α activities were determined after 8 h of intracellular incubation by fluorescence assay and Western blot analysis. Release of IL-6, TNF-α and NO were measured in cell culture medium supernatant after 24 h incubation with ELISA and Griess reaction assay.

In vitro cytotoxicity test

A cytotoxicity assay was performed to verify the safety of the drug in vitro. Briefly, after stable subculture, HCECs and RAW264.7 cells were seeded in a 96-well plate at a density of 5 × 104 /mL and 1 × 105 /mL in each well, respectively, and cell medium (100 μL per well) was added. The cells were then treated with a series of GA at concentrations ranging from 0 to 200 μM in a complete medium for 24 h. Thereafter, 10 μL of CCK-8 reagent was added to each well and cells were left to incubate at 37 ℃ for 1 h. Finally, the absorbance (A) was measured at 450 nm. Six parallel wells were used in each group, and the average value was obtained. At the same time, the control group included untreated cells containing complete media. No cells were set as the blank wells. The cell survival rate was calculated as follows: cell survival rate (%) = [(Aexperimental group − Ablank group) / (Acontrol group − Ablank group)] × 100%. The experiment was repeated three times independently.

Wound healing assay

We used a wound healing assay to demonstrate the effects of GA on cell migration in vitro. First, HCECs were cultured in DMEM/F12 containing 10% FBS. The cells were then plated in a 6-well plate and grown to 90% confluence. Thereafter, artificial wounds were created using a sterile 200 μL plastic pipette tip to scratch across the cell surface. Dissociated cells were removed by washing with phosphate buffered saline (PBS) (all PBS mentioned has a concentration of 0.01 M, pH = 7.4). The indicated amount of GA, at a final drug concentration of 0–100 μM was added for co-incubation, and images of the same area of the wound were taken at 0, 6, 18, and 24 h to determine wound closure. The scratch area was measured using Image J. The cell migration rate was expressed using the following formula: a = (1 − b/c) × 100%, where a is the migration rate, b is the area of scratch at the indicated time, and c is the initial area of the scratch.

ELISA and detection of NO

In animal experiments, one sample consisted of the protein from the conjunctiva of two eyes and cornea of four eyes. In cell experiments, one sample consisted of the protein from the cell supernatant. The supernatant from each group was collected, and the TNF-α and IL-6 levels were quantified in accordance with the manufacturer’s protocols for the ELISA kits. The nitrite level was detected using the Griess reaction assay according to the manufacturer’s instructions.

Measurement of cellular production of ROS

Intracellular ROS was detected by an oxidation-sensitive fluorescent probe (DCFH-DA). To detect the inhibitory effect of drugs on intracellular ROS production, RAW264.7 cells or HCECs were seeded onto 6-well plates at a density of 1 × 105 /well and were incubated overnight. After drug addition, the cells were incubated for 16 h at 37 ℃, then LPS was added to each well and incubated for 8 h, hyperosmotic stress was added to each well and incubated for 1 h. Cells were then cultured in a serum-free DMEM/F12 or DMEM medium containing 10 μL DCFH-DA at 37 °C for 30 min and washed three times with PBS. DCFH-DA can be deacetylated intracellularly by nonspecific esterases and further oxidized by ROS to the fluorescent compound 2,7-dichlorofluorescein (DCF). Finally, DCF fluorescence was observed under a fluorescence microscope (Leica Microsystems, Mannheim, Germany), and the intensity of fluorescence was measured by flow cytometry (Becton Dickinson).

Cellular immunofluorescence

Cells were attached to cell slides in 24-well plates. After 24 h of cell growth, GA was added to the medium to achieve a final concentration of 100 μM for 16 h. Next, LPS was added to each well, and incubated for 8 h. Then, the cell slides were fixed in 4% PFA for 20 min at room temperature and permeabilized with 0.1% Triton X-100 for 10 min at 4 ℃. Next, the cell slides were incubated with primary antibodies against NF-κB P65 (1:200 dilution) and Nrf2 (1:200 dilution) overnight at 4 ℃. After washing with PBS, cell slides were incubated with fluorescent secondary antibodies (1:400 dilution). Finally, after PBS washing, an antifade mounting medium containing DAPI (no washing required) was used and the slides sealed. Fluorescence images were obtained using a fluorescence microscope (Leica Microsystems).

Western blot analysis

Briefly, the cells were lysed in Radio Immunoprecipitation Assay (RIPA) buffer containing a protease inhibitor cocktail and phosphatase inhibitor. Protein levels were quantified using a Bicinchoninic Acid Assay (BCA) kit, and then the loading buffer was added to the sample, which was boiled at 95 °C for 5 min. Equal amounts of protein were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride (PVDF) membranes. Then, the blots were washed with TBST (10 mM Tris–HCl, 150 mM NaCl, 0.1% Tween-20, pH 7.6), blocked with 5% skimmed milk for 2 h, and incubated at 4 °C overnight with primary antibodies at the dilutions recommended by the supplier. β-actin acted as the loading control. Membranes were washed with TBST and incubated with secondary antibodies for 2 h at room temperature. Protein bands were detected using enhanced chemiluminescence (ECL) chemiluminescence reagents (Millipore) and visualized using a luminescent image analyzer. Finally, band analyses were performed using Image J [30].

Experimental dry eye murine model and the experimental group

The animal research protocol was approved by the Laboratory Animal Ethics Committee of Wenzhou Medical University (ID No. wydw2022-0064). All procedures were performed in accordance with the Association of Research and Vision in Ophthalmology (ARVO) statement. One hundred and twenty female C57BL/6 mice, aged 6–8 weeks, were used in the following experiments. We created the EDE animal model by subcutaneous injection of 0.5 mg/0.2 mL scopolamine three times a day (8 a.m., 1 p.m., 6 p.m.) and exposed the animals to an intelligently controlled environmental system (ICES) with ventilation and 20% humidity as previously described [31, 32]. Food, water, and animal behavior were not restricted during the experiment.

To explore the concentration of GA in animal experiments, we reviewed the literature and found that the concentration of GA in the Tibetan Medicine Formula Jikan Mingmu Drops for db/db mice was 13.8830 mg/mL [28]. Therefore, we explored the protective effect of 1 mg/mL, 5 mg/mL and 10 mg/mL GA on corneal epithelial cells in EDE, as shown in Additional file 1: Figs. S5a and b, 5 mg/mL and 10 mg/mL GA can significantly reduce the spotting of corneal fluorescein sodium in mice, while 1 mg/mL GA has no effect. Hence, the lowest concentration of 5 mg/mL was chosen with good effect in subsequent experiments.

In order to explore the therapeutic effects of GA, after the preliminary screening to exclude existing ocular surface diseases, we randomly assigned a control group, named (1) Treatment group of the normal control group (T-NC), mice that did not receive EDE and were not given eye drops. The remaining mice were subjected to EDE for 10 days. After 10 days of dry eye modelling, the mice were randomized into three groups and continued with induction of EDE or treatment with eye drops for 5 days. The groupings were as follows: (2) Treatment group of the EDE group (T-EDE), mice that continued to receive EDE but were not given eye drops; (3) Treatment group of the EDE + PBS group (T-EDE + PBS), mice that continued to receive EDE and were treated with PBS eye drops; (4) Treatment group of the EDE + GA group (T-EDE + GA), mice that continued to receive EDE and were treated with eye drops made of 5 mg/mL GA dissolved in PBS. Eye drops (2 μL) were applied topically to both eyes of the mice three times a day (8 a.m., 1 p.m., and 6 p.m.) until they were euthanized by pentobarbital injection after 5 days.

To explore the preventive effect of GA on dry eye, after the preliminary screening to exclude existing ocular surface diseases, the mice were randomized into four groups for EDE-induced or eye drop treatment for five days. The groupings were as follows: (1) Prevention group of the normal control group (P-NC), mice that were not EDE-induced and not given eye drops; (2) Prevention group of the EDE group (P-EDE), mice that were EDE-induced but were not given eye drops; (3) Prevention group of the EDE + PBS group (P-EDE + PBS), mice that were EDE-induced and treated with PBS eye drops; (4) Prevention group of the EDE + GA group (P-EDE + GA), mice that were EDE-induced and were treated with 5 mg/mL GA dissolved in PBS eye drops. Eye drops (2 μL) were applied topically to both eyes of the mice three times a day (8 a.m., 1 p.m., and 6 p.m.) until they were euthanized by pentobarbital injection after five days.

Corneal fluorescein staining

Corneal fluorescein staining score measurements were performed as previously described [33]. Sodium fluorescein (1%, 1 μL) was instilled into the inferior conjunctival sac using a micropipette. After 90 s, punctate staining of the corneal surface was performed in a double-blind fashion. The intensity of corneal fluorescein staining was calculated using a 4-point system: 0 point, no staining; 1 point, superficial stippling micropunctate staining with less than 30 spots; 2 points, punctate staining with 30 or more spots, but no diffuse staining; 3 points, severe diffuse staining, but no positive plaque or patch; and 4 points, positive fluorescein plaque or patch. The scores in the top, bottom, left, and right areas were totaled to generate a final score, which ranged from 0 to 16.

Terminal deoxynucleotidyl transferase mediated dUTP nick end labeling (TUNEL) assay

After euthanasia, the eyeballs of the mice were removed and placed in optimum cutting temperature (OCT) glue at − 80℃ overnight. Frozen sections (sagittal plane, thickness of 10 μm) were obtained and placed at room temperature for approximately 1 h. Cell apoptosis in these tissue sections was examined by TUNEL staining using the in situ cell death detection kit according to the manufacturer’s instructions. Corneal sections were fixed with 4% paraformaldehyde at room temperature for 10 min. After fixation, they were permeabilized with Triton-X (0.1% in PBS) for 10 min, then 50 μL (5 μL enzyme solution in 45 μL solution) TUNEL reaction mixture was applied and incubated for 1 h at 37 °C in a humidified chamber. Counterstaining with DAPI (1:1000 dilution) was performed for 30 min. The sections were mounted with an anti-fading mounting medium and sealed with cover glass for microscopic observation.

H&E staining and PAS staining

After 5 days of treatment, the entire lacrimal eyeball, with eyelids attached, and conjunctiva were fixed in 10% formalin for 24 h. After dehydration, the specimens were embedded in paraffin, cross-sectioned, and stained with H&E reagent and PAS for histological examination. Each section was observed under a microscope (Imager.z1; Germany). To prevent experimental bias, all images were captured randomly and assessed by two independent researchers in a blinded manner. Goblet cells in the superior conjunctiva were counted using three images taken from three mice at a magnification of 10 × the actual size.

In vivo chronic ocular irritancy test and intraocular pressure monitoring

The chronic ocular irritancy of GA was assessed using ocular irritancy tests adapted and modified from previous studies [30]. Three female New Zealand albino rabbits (weight: ~ 2.5 kg) were acquired from Wenzhou Medical University Animal Center and housed individually in cages on a standard laboratory diet. Briefly, 50 μL of GA (5 mg/mL) was instilled into the lower conjunctival sac of the right eye of each rabbit, while 50 μL of PBS was instilled in the opposite eye as a reference; these two eye drops were applied three times a day, one drop each time. The eye drops were applied for seven days. The eyes were evaluated for clinical signs, intraocular pressure, and sodium fluorescein staining by an experienced doctor using a slit-lamp (Kang Hua®, Chongqing, China) before instilling eye drops every day. After 7 days and 24 h, or 48 h, or 72 h later, fluorescein sodium staining, intraocular pressure measurement, and evaluation of clinical signs were also done, and the number of infiltrates were scored from 0 (no sign) to 3 (severe) according to the Draize test under the supervision of trained optometrists who have experience working with animals. The score was based on the following criteria: 0 points, no redness, inflammation, or excessive tearing; 1 point, slight redness, slight inflammation, and slight tearing; 2 points, moderate redness, moderate inflammation, and excessive tearing; 3 points, severe redness, inflammation, and excessive tearing. After 10 days, histopathological changes in the cornea were observed by H&E staining at 24 h. The schematic of the treatment plan and observation plan can be seen in Figure S3a in the supporting information.

Statistical analysis

All biological experiments were repeated three times independently. Data are expressed as the mean ± standard deviation. Statistical analyses were performed with SPSS 25 (SPSS Inc., Chicago, IL) using one-way analysis of variance (ANOVA) followed by post-hoc Tukey's test. A P value of 0.05 or less was considered statistically significant. *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.001 compared to the LPS group or EDE group.