Fosfenopril Attenuates Inflammatory Response in Diabetic Dry Eye Models by Inhibiting the TLR4/NF-κB/NLRP3 Signaling Pathway

Abstract

Purpose: The purpose of this study was to investigate the involvement of the TLR4/NF-κB/NLRP3 signaling pathway and its underlying mechanism in diabetic dry eye.

Methods: Two models of diabetic dry eye were established in high glucose-induced human corneal epithelial (HCE-T) cells and streptozotocin (STZ)-induced C57BL/6 mice, and the TLR4 inhibitor fosfenopril (FOS) was utilized to suppress the TLR4/NF-κB/NLRP3 signaling pathway. The expression changes in TLR4, NF-κB, NLRP3, and IL-1β, and other factors were detected by Western blot and RT‒qPCR, the wound healing rate was evaluated by cell scratch assay, and the symptoms of diabetic mice were evaluated by corneal sodium fluorescein staining and tear secretion assay.

Results: In the diabetic dry eye model, the transcript levels of TLR4, NF-κB, NLRP3, and IL-1β were raised, and further application of FOS, a TLR4 inhibitor, downregulated the levels of these pathway factors. In addition, FOS was found to be effective in increasing the wound healing rate of high glucose-induced HCE-T cells, increasing tear production, and decreasing corneal fluorescence staining scores in diabetic mice, as measured by cell scratch assay, corneal sodium fluorescein staining assay, and tear production.

Conclusions: The current study found that the TLR4/NF-κB/NLRP3 signaling pathway regulates diabetic dry eye in an in vitro and in vivo model, and that FOS reduces the signs of dry eye in diabetic mice, providing a new treatment option for diabetic dry eye.

Dry eye disease (DED), which is caused by multiple factors, is one of the most frequent chronic ocular surface diseases globally, leading to ocular discomfort and even visual impairment, seriously affecting the quality of life of patients.¹ Diabetes mellitus (DM) is a disease of the endocrine system in which blood glucose levels are abnormally elevated.² It is one of the most common diseases globally, and the number of people with DM is projected to reach 693 million people by 2045, especially in developing countries.^3,4 The Tear Film and Ocular Surface Society (TFOS) just published a report that DM was categorized as a possible risk factor for DED.⁵ 

DM can affect all ocular tissues,⁶ and diabetic ophthalmopathy is a group of diseases that affect several parts of the eye caused by diabetes, including diabetic retinopathy, macular edema,⁷ cataract,⁸ glaucoma,⁹ keratoconus,¹⁰ and dry eye.¹¹ However, early on, these complications have no obvious symptoms and are often overlooked. In particular, the visual impact of diabetes complicating dry eye is often underestimated.¹² Clinically, DED is more common than various diabetic corneal symptoms.^13,14 It has been shown that the prevalence of DED is significantly higher in patients with diabetes than in normal control subjects^15,16 by approximately 54%,^17,18 and DED is significantly correlated with blood glucose, glycosylated hemoglobin, and duration of diabetes.^13,19 In diabetic or non-diabetic patients, eye symptoms may be the same in patients with dry eye, such as a gritty sensation, soreness, decreased visual acuity, photophobia, itching, decreased corneal sensitivity, and associated tear film break-up time (TBUT), Schirmer's test, and corneal staining abnormalities.¹⁹ More severe cases may be complicated by keratoconus, delayed epithelial wound healing, persistent epithelial defects, and corneal ulcers.²⁰ The disease greatly reduces the standard of living of the population. 

It has been demonstrated that inflammation is a major factor in the pathogenesis of DED, and hyperglycemia triggers both innate and adaptive immune responses in the functional units of tears. Toll-like receptors (TLRs) play key roles in triggering innate immune responses and inflammation.²¹ TLRs have been associated with many autoimmune diseases.²² Toll-like receptor 4 (TLR4) is the most widely studied TLR, and its expression is increased in the cornea when ocular surface barrier function is disrupted, which is a factor in the ocular surface inflammatory response.²³ Nod-like receptor protein 3 (NLRP3) expression requires activation of the TLR4/NF-κB pathway. Made up of NLRP3, apoptosis-associated speck-like protein (ASC), and cysteinyl aspartate specific proteinase-1 (caspase-1), the NLRP3 inflammasome is a crucial aspect of the innate immune system,²⁴ and it has a significant impact on multiple diseases, such as diabetes,²⁵ inflammatory bowel disease,²⁶ atherosclerosis,²⁷ and Alzheimer's disease.²⁸ 

Current research has indicated that the TLR4/NF-κB/NLRP3 signaling pathway regulates several diseases across the body, such as dry eye,²⁹ diabetic cardiomyopathy,³⁰ and autoimmune hepatitis.³¹ However, the function of the TLR4/NF-κB/NLRP3 signaling pathway in diabetes-associated dry eyes has not been reported. Consequently, the aim of this work was to explore the expression of the TLR4/NF-κB/NLRP3 signaling pathway in an in vitro and an in vivo model of diabetic dry eye in order to generate novel concepts for the creation of medications for diabetic dry eye. 

Materials and Methods

Establishment of Animal Model

The animal studies in this research were approved by the Yantai Yuhuanding Hospital's Animal Protection and Utilization Committee and were carried out in compliance with the ARVO Statement on the Use of Animals in Ophthalmology and Vision Research. The 8-week-old, 25 to 30 g male C57BL/6 mice were provided by Jinan Pengyue Experimental Animal Breeding Co. The mice were housed in an SPF-rated animal house at temperatures of 20°C to 26°C. The ocular surfaces of all animals were assessed before grouping to exclude individuals with ocular abnormalities. 

Eighty male C57BL/6 mice aged 8 weeks, weighing 25 to 30 g, were divided into 2 groups: 20 mice were in the normal control group (NC group) and 60 mice were in the diabetes mellitus dry eye disease (DDE group). The DDE group was injected intraperitoneally with streptozotocin (STZ) at a dose of 50 mg/kg³² once a day, and it was administered at the same time for 5 days. 

• (1) Blood glucose was measured and recorded on the fourth and 10th weeks after intraperitoneal injection. Blood was collected from the mice by tail vein incision, and blood glucose was measured by a fast glucose meter. For the purpose of imitating type 1 diabetes in mice, blood glucose values above 300 mg/dL (16.7 mmol/L) were deemed effective, and mice having blood glucose readings below this cutoff were not used in the studies.
• (2) After the mice were injected intraperitoneally with STZ for 8 weeks, the original DDE mice were randomly divided into 3 groups: (1) the DDE group remained untreated; (2) the DDE + fosfenopril (FOS) group was treated with 1 mM FOS drops; and (3) the DDE + PBS group was treated with 5 µL of PBS drops twice a day for 7 consecutive days. Only the right eye of each mouse was used in all experiments.

Establishment of the Cell Model

This study used an immortalized human corneal epithelial cell line, HCE-T, with cell generations of 5 to 10, which was cultured in HCE-T-specific complete medium (Shanghai Fuheng Biotechnology Co., Ltd.). 

HCE-T cells were inoculated at 1.5 × 10⁶/mL, cultured for 24 hours, affixed to the plate and then split into groups, and each group of cells was starved for 12 hours. The cell experiment was divided into three parts: the first part was divided into the normal control group (NC group), high glucose modeling group (HG group), and the hyperosmotic control mannose group (MAN group). The NC group remained untreated. The final glucose concentration in the HG group was 25 mM, and the final concentration in the MAN group was also 25 mM; the second part of the experiment was divided into the high glucose modeling group (HG group), the TLR4 inhibitor FOS group (HG + FOS group), and the inhibitor negative control PBS group (HG + PBS group). The HG group was the same as before, the HG + FOS group was the HG group adding 1 mM FOS, and the HG + PBS group was the HG group adding the same volume of PBS. The third part of the experiment was divided into the HG group, the TLR4 agonist lipopolysaccharides (LPS) group (HG + LPS group), and the agonist negative control PBS group (HG + PBS group). The HG + LPS group was the HG group adding LPS at a concentration of 10 ug/L. 

Corneal Fluorescein Staining

The degree of injury to mice corneal epithelial cells was assessed by corneal fluorescein staining. The eyes of STZ mice were stained with 0.25% sodium fluorescein solution at 4 weeks, 6 weeks, and 8 weeks. FOS and PBS were applied to the eyes at week 8. Corneal staining was observed and scored under a slit lamp microscope after 7 days. After the mice were fully anesthetized, 1 µL of 0.25% fluorescein sodium solution was dropped into the conjunctival sac of the mice and rinsed with sterile saline after 1 minute. Then, it was examined using a slit lamp microscope and captured on camera using cobalt blue light. 

Corneal Fluorescein Staining Scoring Scale 

Mouse corneas were divided equally into 4 quadrants, with a total score of 16 as the sum of the scores of the 4 quadrants, with each quadrant scored from 0 to 4. The following was the scoring scale: 4 points = fluorescein-positive plaques; 3 points = very dense fluorescence; 2 points = dense patchy pattern; and 1 point = similar sparse punctate weak fluorescence; 0 points = no fluorescence.³³ All fluorescein staining was scored by a single blinded observer. 

Tear Production

The measurement time for this experiment was the same as that described in the Corneal fluorescein staining section, the tear volume in each eye was measured with phenol red saturated cotton wool (produced by Tianjin Jingming New Technology Development Co., Ltd.). Cotton threads were placed in the lower lid vault of mice one third from the lateral corner of the eye, and the length of the threads wetted was measured after 15 seconds (counting unit is mm). The experiment was performed at the same period (15:00 PM). 

Cell Morphology

The morphology of human corneal epithelial cells was observed under a 10-fold inverted microscope (Shanghai Caikang Optical Instrument Co., Ltd. Shanghai, China). After 24 hours of culture in different glucose concentrations, the cell morphology was observed by adjusting the microscope magnification and photographed. The operation was completed by the same experimenter. 

Cell Activity Experiment

Cell Counting Kit-8 (CCK-8 with WST-8: (2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2, 4-disulfophenyl)-2H-tetrazole, monosodium salt) was used to assay the cellular activity of the high glucose. Human corneal epithelial cell suspensions (100 µL/well) were added to 96-well plates. After 24 hours of glucose treatment, each well received 10 µL of CCK-8 (Dojindo, Japan) solution for an additional 2 hours. 

Cell Scratching Experiment

HCE-T cells were grown in 6-well plates (1 × 10⁶ cells/well) after cell adherence using a pipette tip to scratch the cells. Twenty-four hours later, each well was uniformly scratched with a 1000 µL pipette tip. The original medium was replaced with HG, mannose, HG + PBS, HG + FOS, and HG + LPS medium, respectively. Using a fluorescent inverted microscope, the scratched areas were examined and captured on camera at 0, 6, 12, and 24 hours, respectively. 

Real-Time qPCR

Cells were treated according to the Establishment of the cell model section and cultured for 24 hours. Total cellular RNA was extracted by the TRIZOL (TRIcom Reagent, Tianmu Biologicals, Beijing, China) method, RNA purity and concentration were determined by a spectrophotometer (Nanodrop 2000/2000C) and reverse transcription products were obtained using G592 (ABM, Vancouver, Canada) reverse transcription reagent cDNA. Polymerase chain reaction was performed using G891 (ABM, Vancouver, Canada) qPCR Master Mix (10.0 µL), 10 µM forward and reverse primers (0.5 µL each), 2 µL of template cDNA and H2O (7 µL). Three replicate wells were set up for each group. The relative expression of TLR4, NLRP3, NF-κB, and IL-1β in the samples was analyzed by the 2^−ΔΔCT method using β-actin as an internal reference gene. Each treatment was tested in triplicate in each experiment, and each experiment was repeated three times. The sequences of the RT–qPCR primers synthesized by Sangon Biotech (Shanghai, China) are shown in the Table. 

Table.

 

List of Primers

 

Western Blotting

The grouping and treatment of animals in this experiment were the same as in the Establishment of the cell model section. One week after eye spotting, the mice were anesthetized and euthanized by cervical dislocation, and the corneal tissues were extracted. The grouping of cells in this experiment was the same as in the Method section Cell morphology. RIPA buffer (Solarbio, Beijing, China) and phenylmethane sulphonyl fluoride (PMSF) were added to extracted corneal tissues and cells, respectively, at a ratio of 100:1. The corneas were added with steel beads to a tissue homogenizer, lysed well on ice for 2 hours, and then centrifuged at 4°C and 12,000 × g for 20 minutes, and the supernatant was obtained. The protein concentration was measured by a BCA assay kit, and each sample was diluted to the same concentration. Then, we applied one fourth of the volume of the 5 × super sample buffer and mixed well. Each treatment was tested in triplicate in each experiment, and each experiment was repeated three times. Western blot experiments were then performed using the following antibodies: TLR4 (1:1000, Shanghai, Abmart, China), NLRP3 (1:1000, Shanghai, Abmart, China), NF-κB (1:1000, Shanghai, Abmart, China), IL-1β (1:1000, Affinity, USA), and β-Tubulin (1:10 000, Affinity, USA). 

Data Analysis

Data were analyzed using ImageJ (Bethesda, MD, USA), Adobe Illustrator (San Jose, CA, USA), and GraphPad Prism version 8.0 (San Diego, CA, USA) software. For multiple comparisons of non-normally distributed data, the Kruskal-Wallis test was used. To test more than two sets of data, 1-way ANOVA was used. A 2-way ANOVA was used to compare the experimental results of the different intervention groups over time. It was deemed statistically significant when P < 0.05. 

Results

Alteration of HCE-T Cell Morphology and Activity Under High Glucose Stimulation

Inverted microscopy was applied for the observation of cell morphology, and after culturing at different glucose concentrations (25, 30, and 35 mM) for 24 hours, it was found that the number of dead cells increased and the cell size decreased as the glucose concentration increased (Fig. 1A). CCK8 experiments showed that with the increase in glucose concentration, cell activity decreased, the difference was statistically significant (P < 0.05; Fig. 1B). 

Figure 1.

 

Results of cell morphology and cell activity. Alterations in HCE-T cells stimulated with high glucose. In this study, we observed the morphology and size of cells treated with different glucose concentrations (25, 30, and 35 mM) using inverted microscopy (A). Effects of different concentrations (25, 30, and 35 mM) of glucose on cell survival (B). Each experiment was repeated three times (mean ± SD; *P < 0.05 versus the NC group).

The TLR4/NF-κB/NLRP3 Signaling Pathway Is Highly Expressed in High Glucose-Induced HCE-T Cells

High glucose environments can cause a variety of inflammatory responses on the ocular surface. In this experiment, an NC group, an HG group and a hyperosmotic control MAN group were established, and RT‒qPCR and Western blot experiments showed that the expression levels of TLR4, NF-κB, NLRP3, and IL-1β-related proteins and mRNAs were significantly higher than the HCE-T cells of the HG group stimulated with high glucose (P < 0.05), whereas the difference between the NC and MAN groups was not statistically significant (P > 0.05; Fig. 2). Therefore, in this experiment, we found that the TLR4/NF-κB/NLRP3 signaling pathway was highly expressed in high glucose-induced HCE-T cells. 

Figure 2.

 

Changes in TLR4, NLRP3, NF-κB , and IL-1β in high glucose-induced HCE-T cells. HCE-T cells were stimulated with glucose at a final concentration of 25 mM. RT‒qPCR was used to determine the mRNA expression of the TLR4/NF-κB/NLRP3 signaling pathway (A); Western blotting analysis was used to determine the protein expression of the TLR4/NF-κB/NLRP3 signaling pathway (B, C), and the grayscale of the bands was analyzed with β-tubulin as an internal reference gene. Each treatment was tested in triplicate in each experiment, and each experiment was repeated three times (mean ± SD; ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

Expression of TLR4, NLRP3, NF-κB, and IL-1β-Related Factors in High Glucose-Induced HCE-T Cells can be Reduced in Response to FOS

Fosfenopril (FOS) inhibits TLR/NF-κB signaling to attenuate inflammatory responses.³⁴ To confirm whether the TLR4 inhibitor FOS can effectively inhibit the expression of the TLR4/NF-κB/NLRP3 signaling pathway in high-glucose-induced HCE-T cells and to ensure the rigor of the experiments, we established an HG group, TLR4 inhibitor FOS group (HG + FOS group), and inhibitor negative control PBS group (HG + PBS group). The epithelial cells were collected after 24 hours of culture. The RT-qPCR results showed that the mRNA levels in HG + FOS-treated HCE-T cells were significantly lower than those in the other two groups (P < 0.05; Fig. 3A). The protein levels of TLR4, NLRP3, NF-κB, and IL-1β were displayed in the Western blot results, which showed significantly lower levels in the HG + FOS group treated HCE-T cells compared to those in the HG group and HG + PBS group (P < 0.001; Figs. 3B, 3C). Therefore, the TLR4/NF-κB/NLRP3 signaling pathway may play a role in diabetic dry eyes. 

Figure 3.

 

Effects of the TLR4 inhibitor FOS on TLR4, NLRP3, NF-κB , and IL-1β-related factors in high glucose-induced HCE-T cells. High glucose stimulation of HCE-T cells with FOS. RT‒qPCR was used to determine the mRNA expression of the TLR4/NF-κB/NLRP3 signaling pathway (A); Western blotting analysis was used to determine the protein expression of the TLR4/NF-κB/NLRP3 signaling pathway (B, C), and the grayscale of the bands was analyzed with β-tubulin as an internal reference gene. Each treatment was tested in triplicate in each experiment, and each experiment was repeated three times (mean ± SD; ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

LPS Increased the Expression of TLR4, NLRP3, NF-κB, and IL-1β-Related Factors in High Glucose-Induced HCE-T Cells

To further confirm the expression of TLR4/NF-κB/NLRP3 signaling pathway in high glucose-induced HCE-T cells, this experiment set up an HG group, a TLR4 agonist LPS group (HG + LPS group); and an agonist-negative control group, PBS group (HG + PBS group). The epithelial cells were collected after 24 hours of culture. RT-qPCR showed that the mRNA levels in the HG + LPS group were significantly higher than those in the other two groups (P < 0.05; Fig. 4A). Western blot results showed that the protein transcription levels of TLR4, NLRP3, NF-κB, and IL-1β in the HG + LPS group were significantly higher than those in the other 2 groups (P < 0.05; Figs. 4B, 4C). Thus, we demonstrated here that the TLR4/NF-κB/NLRP3 signaling pathway may play a role in diabetic dry eye. 

Figure 4.

 

LPS increases the expression of inflammatory mediators in high glucose-stimulated HCE-T cells. RT‒qPCR was used to determine the mRNA expression of the TLR4/NF-κB/NLRP3 signaling pathway (A); Western blotting analysis was used to determine the protein expression of the TLR4/NF-κB/NLRP3 signaling pathway (B, C), and the grayscale of the bands was analyzed with β-tubulin as an internal reference gene. Each treatment was tested in triplicate in each experiment, and each experiment was repeated three times (mean ± SD; ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

FOS can Accelerate the Healing of High Glucose-Induced Damage in HCE-T Cells

Wound healing is delayed in patients with diabetes compared to healthy individuals. To investigate whether high glucose affects the rate of injury healing in corneal epithelium, the cells were made with a single scratch by using a pipette tip from the bottom to the top of the well after cell adherence. Fluorescence inverted microscopy was used to observe the damage repair of HCE-T cells. The healing rate of HCE-T cells was significantly decreased in the HG group than those in the NC group and the MAN-treated group at 6 hours and 12 hours scratching (P < 0.001; Figs. 5A, 5B). The healing rate of HCE-T cells increased in the HG + FOS group compared with those in the HG group and HG + PBS group at 6 hours and 12 hours of scratching (P < 0.05; Figs. 5C, 5D).The healing rate of HCE-T cells was significantly decreased in the HG + LPS group than those in the HG group and HG + PBS group at 6 hours and 12 hours of scratching (P < 0.05; Figs. 5E, 5F). This suggests that high glucose affects the healing rate of corneal epithelial injury, and LPS slows the healing of high glucose-induced injury in HCE-T cells. Whereas the rate of corneal healing becomes faster after the application of FOS, a TLR4 inhibitor. 

Figure 5.

 

Results of cell scratching experiments. HCE-T cells were grown in media for 24 hours. Vertical lines were made on six-well plates using a 1000 µL pipette tip. Then, different media were used for incubation, and the healing of HCE-T cells was observed after 0, 6, 12 hours (A, C, E). ImageJ was used for calculating the repair areas (B, D, F). The identical experiment was repeated three times (mean ± SD; ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

STZ Induces Significant Dry Eye Signs in Diabetic Mice

STZ injection induced maintained high blood glucose levels at 8 w (Fig. 6B), demonstrating successful modeling. To verify the progression of dry eye in STZ-induced diabetic mice, corneal fluorescence staining images of mouse eyes were taken and scored under a slit-lamp microscope, which showed an increase in corneal fluorescence staining scores at 6 weeks and 8 weeks (P < 0.001; Figs. 6A, 6C), and tear secretion was determined by the phenol-red cotton wool method, which started to decrease at 6 weeks and then decreased markedly at 8 weeks (P < 0.01; Fig. 6D). Therefore, diabetic mice showed obvious dry eye signs after 8 weeks of STZ induction, and 8 week old diabetic mice were selected for subsequent experiments. 

Figure 6.

 

Progression of dry eye in diabetic mice. Streptozotocin was injected intraperitoneally to induce type 1 diabetes in mice, and blood glucose was measured at 10 days and 4 weeks after the last injection (B) (n = 5). Corneal fluorescein sodium staining (A and C) (n = 5) and tear secretion (D) (n = 5) were compared between STZ-induced diabetic mice at 4 weeks, 6 weeks, and 8 weeks (ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

FOS Reduces the Protein Expression of TLR4, NLRP3, NF-κB, and IL-1β-Related Factors in the Corneas of Diabetic Mice

Western blot experiments were conducted to measure the protein levels of TLR4, NF-κB, NLRP3, and IL-1β in corneal tissues of the 4 groups in order to determine if the TLR4 inhibitor FOS affected the TLR4/NF-κB /NLRP3 signaling pathway in STZ-induced diabetic mice. TLR4, NF-κB, NLRP3, and IL-1β-related proteins were expressed at higher levels in the DDE group than in the NC group (P < 0.001); the expression levels of related proteins in the DDE + FOS group were lower than those in the DDE group (P < 0.001) , and there was no statistically significant difference in protein expression levels between the DDE + PBS group and the DDE group (P > 0.05; Fig. 7). Therefore, the TLR4/NF-κB/NLRP3 signaling pathway was highly expressed in the corneas of diabetic mice, and the administration of the TLR4 inhibitor FOS reduced the expression level of related factors. 

Figure 7.

 

TLR4 inhibitor FOS reduces TLR4, NLRP3, NF-κB, and IL-1β protein expression in the corneas of diabetic mice. The expression of the TLR4/NF-κB/NLRP3 signaling pathway was determined by Western blotting analysis (A, B), analyzing the grayscale of the bands with β-tubulin as an internal reference gene (C–F). Each treatment was tested in triplicate in each experiment, and each experiment was repeated three times (mean ± SD; ns P > 0.05, *P < 0.05, **P < 0.01, ***P < 0.001).

Dry Eye Signs in Diabetic Mice Are Relieved After Application of FOS Eye Spotting

The corneal fluorescence staining score of the DDE + FOS group decreased after 1 week of eye-dotting compared with that of the DDE group and the DDE + PBS group (P < 0.01; Figs. 8A, 8B), the length of the phenol red cotton staining wool test became longer (P < 0.01; Fig. 8C), and there was no statistically significant difference between the DDE and DDE + PBS groups. Therefore, dry eye signs in diabetic mice could be alleviated after administration of the TLR4 inhibitor FOS. 

Figure 8.

 

Effect of the TLR4 inhibitor FOS on dry eye in diabetic mice. Corneal sodium fluorescein staining (A, B) (n = 5) and tear secretion assessment (C) (n = 5) were performed in diabetic mice with topical administration of 1 mL FOS for 1 week, and PBS was used as a mediator control (ns P > 0.05, **P < 0.01, ***P < 0.001).

Discussion

DED is a prevalent ocular surface problem that is intimately linked to diabetes mellitus.¹³ Despite the rising trend of diabetic dry eye complications, the pathogenesis has still not been fully elucidated. It is important to explore the pathogenesis of diabetic dry eye to alleviate symptoms. By investigating the role of the TLR4/NF-κB/NLRP3 signaling pathway in diabetic mice induced with STZ and in an in vitro model of high glucose-induced HCE-T cells, the present experiments were designed to investigate the possible mechanisms underlying the pathogenesis of diabetic dry eye. The diabetic dry eye model exhibited dramatically higher expression levels of TLR4, NF-κB, NLRP3, and IL-1β, as demonstrated by both in vitro and in vivo experimental data. Administration of the TLR4 inhibitor FOS reduced the expression levels of related factors in HCE-T cells and corneas of diabetic mice, and accelerated healing of cellular damage could be seen by scratch experiments. Eight weeks after intraperitoneal injection of STZ in normal C57BL/6 mice, we found decreased tear secretion and increased corneal sodium fluorescein staining scores in C57BL/6 mice and improved dry eye signs in diabetic mice after administration of FOS by eye spotting for 1 week. Thus, these data imply that the TLR4/NF-κB/NLRP3 signaling pathway plays a regulatory role in an in vitro and an in vivo model of diabetic dry eye and that FOS attenuates dry eye signs in diabetic mice. Similar studies have not been reported in the literature. 

The innate immune system's first line of defense against microbial invasion of the eye is the corneal epithelium. Corneal epithelial cells in patients with diabetes are exposed to high glucose levels over a long period of time, leading to a weakening of the barrier function of the corneal epithelium,^35–37 and, accordingly, the diabetic cornea is more susceptible to infections than those of healthy individuals.^38,39 In the HCE-T cell scratch experiment, when comparing the healing rate of HCE-T cells in the HG group to those in the NC and MAN groups, we discovered a statistically significant decrease, which corresponds to the phenomenon of delayed wound healing in patients with diabetes,⁴⁰ whereas the healing rate of HCE-T cells increased after application of the TLR4 inhibitor FOS. The corneal epithelium of the HG group revealed that factors, such as TLR4, NLRP3, NF-κB, and IL-1β, were highly expressed in the HG group compared to the NC and MAN groups, whereas there was no statistically significant difference between the MAN and NC groups. This suggests that the change in osmolality has little effect on the corneal epithelium, and the effect of high glucose on the corneal epithelium is not related to the hyperosmolality produced by high glucose. This is in line with what earlier research has shown.⁴¹ 

Patients with diabetes mellitus frequently present with symptoms of dry eye, and clinically, patients with diabetes mellitus, either type 1 or type 2, commonly report symptoms of dry eyes, including faster tear film breakup time and decreased tear volume.¹⁷ However, due to the limited number of clinically prediabetic patients, most studies on diabetic ocular surface complications have focused on patients with diagnosed diabetes, and the early signs and etiology of diabetic ocular surface complications are unknown. Corneal sodium fluorescein staining and changes in tear secretion are key indicators for evaluating dry eye. Referring to the specification for the preparation of dry eye animal models (draft),⁴² we chose the phenol red cotton thread test and corneal fluorescein sodium staining as clinical indicators for detecting dry eye in our experiment. In this study, we used the type I diabetic mouse model produced by STZ, and by observing the corneal fluorescein sodium staining scores and tear production experiments in diabetic mice at 4, 6, and 8 weeks, we found that the scores increased, and the tear production decreased at 6 weeks and 8 weeks, which was consistent with the findings of QIAN et al.⁴³ Moreover, in several clinical studies,^44–46 tear production decreased as the duration of diabetes increased. All these results indicate the presence of typical dry eye manifestations in diabetic mice. 

Recent research has shown that inflammation plays an important role in the etiology of DED,^1,47,48 as evidenced by increased levels of proinflammatory cytokines, and that low-grade chronic inflammation has also been implicated as an important mechanism in the development of DM and its complications.¹⁴ TLR4, when activated by specific exogenous substances, such as lipopolysaccharide (LPS), recruits myeloid differentiation factor (MyD88) receptor and triggers nuclear factor κB (NF-κB) to mediate a series of inflammatory responses,⁴⁹ which includes the regulation of the expression of the NLRP3 inflammasome, which is also closely associated with a variety of ocular diseases. Niu and others⁵⁰ found that the mRNA and protein expression of the NLRP3 inflammasome was upregulated in dry eye; Zheng and others⁵¹ demonstrated that NLRP3 gene expression was increased in ocular surface samples from patients with DED, suggesting that NLRP3 plays an important role in the progression of DED. In this experiment, we hypothesized that one of the pathogeneses of diabetic dry eye occurs based on the TLR4/NF-κB/NLRP3 signaling pathway and that blocking the TLR4 pathway can inhibit the inflammatory response and downregulate the mRNA and protein levels of this pathway and downstream factors as an effective way of treating diabetes-related dry eye. The outcomes of both in vivo and in vitro studies in this experiment showed that factors, such as TLR4, NLRP3, NF-κB, and IL-1β, were highly expressed at the protein and mRNA levels in the diabetes-associated dry eye model, and the application of the TLR4 inhibitor FOS downregulated the expression of these factors. The corneal sodium fluorescein staining score of diabetic mice decreased, tear secretion increased, and HCE-T cell healing was accelerated after administration of the TLR4 inhibitor FOS. These findings imply that this pathway is involved in the inflammatory response in diabetic dry eye. This is in line with recent observations that NF-κB is also expressed in the lacrimal gland of STZ-induced diabetic rats,⁵² and that NLRP3 and IL-1β are also overexpressed in the cornea and conjunctiva of diabetic mice.^32,53 

In summary, the TLR4/NF-κB/NLRP3 signaling pathway was found to play a regulatory role in the pathogenesis of diabetic dry eye, and the TLR4 inhibitor FOS downregulated the expression of factors related to this pathway and improved the ocular surface of diabetic mice induced with STZ. This finding provides new ideas for future drug development and mechanism of action studies in diabetic dry eye. 

Acknowledgments

Funded by a grant from the Yantai Science and Technology Development Plan, China (Grant No. 2017WS109 and Grant No. 2021YD025). 

Disclosure: K. Jiang, None; F. Zhang, None; Y. Chen, None; X. Li, None; X. Zhao, None; P. Jiang, None; Y. Li, None