Preservative-free latanoprost induces meibomian gland dysfunction through inflammatory and oxidative stress pathways

Abstract

Purpose

This study aimed to investigate the effects of preservative-free latanoprost on meibomian gland function in mice and its possible mechanism.

Methods

Disproportionality analysis was conducted using the FAERS database to evaluate adverse reaction reports and epidemiological characteristics associated with preservative-free latanoprost. In mouse models, 0.005% preservative-free latanoprost or vehicle control was topically applied four times daily for 7, 14, or 28 days. Morphological changes of the meibomian gland in mice were detected by immunohistochemistry. Immunofluorescence staining, western blotting, and quantitative real-time PCR (qRT-PCR) were used to assess the expression of the prostaglandin F2α receptor (FP), inflammatory cells and mediators, oxidative stress-related markers, and signaling pathway–related factors in mouse meibomian gland tissues.

Results

Reports of adverse reactions caused by preservative-free latanoprost increased annually in the FAERS database. Locally applied preservative-free latanoprost in mice led to an escalation of mucous secretion at the eyelid margin, accompanied by meibomian gland duct obstruction, lipid accumulation in glandular acini, an elevation in the expression levels of FP and Slco2a, and a reduction in the expression levels of PGDH within the meibomian glands. Other inflammatory markers, including CCL2, IL-1β, TNF-α, IL-6, and CXCL5, were significantly upregulated in the latanoprost-treated group, showing approximately 1.3–1.5-fold increases in CCL2, IL-1β, TNF-α, and CXCL5, and a more pronounced ~4-fold increase in IL-6 (p < 0.05). Consistently, oxidative stress-related proteins NOX4, 3-nitrotyrosine (3-NT), and 4-hydroxynonenal (4-HNE) were significantly increased (~1.3–1.5-fold, p < 0.05), whereas the antioxidant proteins superoxide dismutase 2 (SOD2) and Keap-1 were significantly altered, with SOD2 decreased to ~0.6-fold and Keap-1 increased to ~1.2-fold compared with controls (p < 0.05). Additionally, the Erk and NF-κB signaling pathways were significantly activated.

Conclusion

0.005% preservative-free latanoprost induces meibomian gland dysfunction in mice by promoting inflammatory responses and oxidative stress.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40360-025-01078-9.

Introduction

Meibomian gland dysfunction (MGD) is a disease characterized by chronic diffuse damage to the meibomian glands, ductal obstruction, or abnormal quantity or quality of meibum secretion [1]. In recent years, numerous clinical trials have reported a high incidence of MGD in glaucoma patients [2–5]. The researchers reiterated that glaucoma patients had significantly worse meibum quality, lower meibomian gland secretion, and thinner lipid layer thickness compared to healthy control [5, 6], with a notable negative impact on the quality of life, compliance with treatment, and progression of glaucoma treatment in patients [7]. Moreover, previous studies have demonstrated that the high incidence of MGD in glaucoma patients is closely associated with the long-term use of anti-glaucoma eye drops, particularly prostaglandin (PG) analogs, which are widely prescribed as first-line therapy in clinical practice [3, 6]. In a clinical setting, Mocan et al. reported a significantly higher prevalence of MGD in glaucoma patients receiving PG analog monotherapy (92.0%) compared with those treated with non-PG analog regimens (58.3%) [3]. These findings raise important patient safety concerns, as PG analog-associated ocular surface alterations may compromise tear film stability, exacerbate ocular discomfort, and negatively affect treatment adherence and long-term therapeutic outcomes. In another study, Cho et al. compared patients receiving PG analog monotherapy with those treated with combination therapy and found no significant difference in meibomian gland dropout between the two groups, indicating that PG analogs are likely the primary contributors to MGD rather than adjunctive intraocular pressure (IOP)-lowering medications [8]. As the first PG analog used in the treatment of glaucoma [9], latanoprost also has been reported to induce significantly higher scores for lid margin abnormality, meiboscore, and meibum compared with untreated control eyes in clinical studies [10, 11].

With the continuous emergence of new drugs in recent years, reports of drug-induced diseases have become increasingly common. The FDAs Adverse Event Reporting System (FAERS) is a database of drug-related adverse reactions reported by groups including physicians, pharmacists, and drug manufacturers, which has been used to summarize drug-induced pancreatitis, drug-related interstitial lung disease, etc. in previous studies [12, 13]. Latanoprost represents a significant addition to the first-line treatment options for reducing IOP in patients with open-angle glaucoma or ocular hypertension and continues to be widely prescribed worldwide [14, 15]. Therefore, the potential adverse reactions of latanoprost will be evaluated using the FAERS database in this study. Meanwhile, it is important to examine its potential role in eye disorders.

With the continuous emergence of new drugs, reports of drug-induced diseases have increased substantially in recent years. The FDA Adverse Event Reporting System (FAERS), which collects drug-related adverse events reported by healthcare professionals, manufacturers, and patients, has been widely used to identify safety signals and characterize drug-associated adverse reactions, including drug-induced pancreatitis and interstitial lung disease [12, 13]. Latanoprost is a first-line IOP-lowering agent for patients with open-angle glaucoma or ocular hypertension and remains one of the most commonly prescribed prostaglandin analogs worldwide [14, 15]. However, systematic evaluations of its potential ocular adverse effects, particularly MGD, remain limited. Therefore, in the present study, we first utilized the FAERS database to identify and characterize clinical safety signals associated with latanoprost-related eye disorders, providing real-world evidence of its potential association with MGD. To complement the pharmacovigilance analysis and to further explore the underlying biological mechanisms, we subsequently employed a mouse model to investigate the effects of latanoprost on meibomian gland structure and function. By integrating FAERS-based signal detection with controlled experimental validation, this study aims to provide a comprehensive assessment of latanoprost-induced MGD from both clinical and mechanistic perspectives. In our previous study, we demonstrated that 0.005% preservative-free latanoprost can induce dry eye-like ocular surface damage by promoting inflammation in mice [16]. However, despite accumulating clinical evidence linking latanoprost use to MGD, the molecular and cellular mechanisms underlying latanoprost-induced MGD remain largely unknown, and direct experimental evidence elucidating these mechanisms is still lacking.

Latanoprost is an ester prodrug of prostaglandin F2α (PGF2α), it can also bind to the PGF2α receptor, interacting with the inflammatory factor PGF2α to perform biological function [17]. This mechanism may lead to various inflammation-related adverse reactions on the ocular surface [16]. Given that MGD is an inflammation-driven eyelid disorder, the local application of preservative-free latanoprost eye drops could potentially influence the pathogenesis of MGD. Understanding the role and mechanism of PG analogs in MGD would help preserve the structural and functional integrity of meibomian glands during eye drop usage, which could significantly enhance the glaucoma treatment effect. In this study, we aim to evaluate the epidemiological characteristics of adverse ocular reactions to preservative-free latanoprost, explore the effects of 0.005% preservative-free latanoprost eye drops on the mouse meibomian gland, and delve deeper into the potential mechanism.

Methods

Study design and data source

We conducted a retrospective pharmacovigilance analysis of latanoprost based on the FAERS databases (https://fis.fda.gov/extensions/FPD-QDE-FAERS/FPD-QDE-FAERS.html), which are publicly accessible repositories of adverse drug event reports. The FAERS database comprises voluntary reports from various sources, including healthcare professionals, patients, pharmacists, and pharmaceutical companies. According to FDA FAERS data use guidelines, the database is primarily intended for signal detection rather than for establishing causal relationships, as reports may be subject to underreporting, reporting bias, and incomplete clinical information. This database adheres to national safety reporting guidelines and encodes all adverse events using preferred terms (PTs) from the Medical Dictionary of Regulatory Activities. Due to the public accessibility and anonymized nature of patient records in both databases, our study did not involve informed consent or ethical approval. To ensure inclusion of the most comprehensive and consistently structured reports, we extracted all publicly available FAERS reports from the first quarter of 2004 through the fourth quarter of 2023, as the FAERS database is documented to provide quarterly data files covering this period. Specifically, our analysis focused on adverse reaction reports associated with Xelpros, Monoprost, and Iyuzeh, which are preservative-free latanoprost formulations.

Animals

In this study, male C57BL/6J mice aged 6–8 weeks [18], obtained from the Shanghai SLAC Laboratory Animal Center (Shanghai, China), were utilized. Mice were randomly assigned to experimental groups, with five animals per group (n = 5). C57BL/6J mice were selected because they are a well-characterized inbred strain widely used in ocular and inflammatory research, and their meibomian gland anatomy and immune responses are well documented, making them suitable for investigating MGD [19]. All experimental procedures were approved by the Experimental Animal Ethics Committee of Xiamen University (Approval No. XMULAC20220258) and adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. During all experimental procedures, mice were anesthetized by intraperitoneal injection of pentobarbital sodium (50 mg/kg). Within 3–5 minutes after injection, mice reached a deep anesthesia state, characterized by slow respiration and absence of pedal or righting reflexes. The depth of anesthesia was confirmed before any manipulation to ensure that animals were completely unconscious and free from pain. At the end of experiments, mice were humanely euthanized by cervical dislocation under deep anesthesia. This method was selected because pentobarbital sodium induces rapid and stable anesthesia, ensuring a painless and humane euthanasia process in accordance with the AVMA Guidelines for the Euthanasia of Animals (2020).

Topical administration of 0.005% latanoprost

In alignment with our previous study, 0.005% preservative-free latanoprost was prepared in phosphate-buffered saline (PBS), supplemented with 0.01% dimethyl sulfoxide (DMSO) [20]. The mice were randomly allocated into three groups (n = 5 per group): (1) the untreated control group (UT group), which received no eye drop administration; (2) the vehicle-treated group (Vehicle group), which received PBS; and (3) the group treated with 0.005% latanoprost eye drops (Latanoprost group). Topical application of eye drops was performed four times daily, with a volume of 5 μL administered to each eye per application, over periods of 7, 14, and 28 days.

Animal examination

The ocular surface phenotype of the mice was evaluated in a blinded manner by the same researcher on days 7, 14, and 28 using a slit-lamp microscope (Takagi Seiko Co., Ltd., Nagano, Japan). This evaluation included measuring corneal epithelial integrity using corneal fluorescein staining, assessing conjunctival sac secretion, and observing changes in the eyelid margins. For corneal fluorescein staining, the cornea was divided into four quadrants, and staining was scored on a scale of 0–3 for each quadrant (0 = no staining; 1 = mild punctate staining; 2 = moderate staining; 3 = severe or confluent staining), with a total score ranging from 0 to 12. Higher scores indicated greater corneal epithelial damage. Additionally, the meibomian gland (MG) structure was imaged using a stereoscopic microscope (Leica M165-FC, Germany), following the methodology described in a previous study [21].

LipidTOX staining

Frozen eyelid sections were fixed in 4% paraformaldehyde for 10 minutes and rinsed in PBS for 5 minutes. LipidTOX neutral lipid staining was employed to visualize neutral lipid accumulation within the meibomian glands, as alterations in lipid synthesis and secretion are key pathological features of meibomian gland dysfunction. The sections were then incubated with LipidTOX neutral lipid stain (H34475; Thermo Fisher Scientific, Waltham, MA, USA) diluted 1:100 in buffer solution for at least 30 minutes at room temperature in the dark. Negative control sections incubated without LipidTOX dye were included to validate staining specificity. Following incubation, nuclei were counterstained with DAPI. For each mouse, at least three nonadjacent eyelid sections were analyzed, and representative images were captured using a fluorescence microscope (DM2500; Leica Microsystems, Wetzlar, Germany).

Immunofluorescent staining

For immunofluorescence staining, cryosections were first fixed with cold acetone (−20 °C for 10 minutes), then permeabilized with 0.2% Triton X-100 for 20 min. After that, sections were blocked with 2% BSA for 1 h at room temperature, and incubated with CD-11b (1:250, ab8878; Abcam), and Ly6G (1:600) at 4 °C overnight. Secondary antibodies corresponding to the species of the primary antibodies were applied the next day, and the nucleus was stained with 4’,6-diamidino-2-phenylindole (DAPI; Catalog no. H-1200; Vector, Burlingame, CA, USA). All staining procedures involving fluorescent dyes were performed under light-protected conditions. Images were captured using an upright fluorescence microscope (DM2500; Leica Microsystems), and fluorescence intensity was quantified using NIS-Elements software.

Immunohistochemistry

Following deparaffinization and rehydration, paraffin-embedded sections were treated with 3% hydrogen peroxide for 10 minutes at room temperature to quench endogenous peroxidase activity. Subsequently, heat-induced antigen retrieval was performed using 0.01 M sodium citrate buffer (pH 6.0) at 95–100 °C for 8 minutes, followed by membrane permeabilization with 0.2% Triton X-100 for 20 minutes at room temperature. Sections were then blocked with 2% bovine serum albumin (BSA) for 1 hour at room temperature and incubated overnight at 4 °C with the following primary antibodies: prostaglandin F2α receptor (FP; rabbit polyclonal, 1:100, Abcam, ab133058), NOX4 (rabbit polyclonal, 1:100, Abcam, ab133303), 3-nitrotyrosine (3-NT; rabbit polyclonal, 1:100, Abcam, ab61392), and 4-hydroxynonenal (4-HNE; rabbit polyclonal, 1:100, Abcam, ab46545). On the following day, sections were incubated with horseradish peroxidase (HRP)-conjugated species-specific secondary antibodies at room temperature, and immunoreactivity was visualized using 3,3′-diaminobenzidine (DAB). Slides were subsequently counterstained with hematoxylin, dehydrated, and mounted. Images were captured using a Nikon microscope, and staining intensity was quantified by measuring the integrated optical density (IOD) using Image J software (National Institutes of Health, USA). For each sample, at least three randomly selected fields were analyzed, and the mean IOD value was used for statistical analysis.

Total RNA extraction, reverse transcription and qRT-PCR

The meibomian glands were carefully dissected from each group under a dissecting microscope after removal of the skin, subcutaneous tissue, muscle, and palpebral conjunctiva. Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, United States) according to the manufacturer’s protocol. RNA concentration was measured using a NanoDrop 2000 spectrophotometer. A relatively high amount of total RNA (10 μg) was used for cDNA synthesis to ensure sufficient cDNA yield and detection sensitivity, given the limited tissue volume and low RNA abundance of isolated meibomian glands. cDNA was synthesized using a reverse transcription kit (Catalog No. RR047A; TaKaRa, Shiga, Japan). Real-time PCR analysis was performed using Hieff™ qPCR SYBR® Green Master Mix (No Rox) (catalog no. 11201ES08; Yeasen, China) on a LightCycler® 96 Real-Time PCR System (catalog no0.05815916001, Roche, Switzerland), following a previously described amplification protocol [22]. All qRT-PCR reactions were performed in triplicate. mRNA expression levels were quantified using the comparative threshold cycle (Ct) method, with β-actin serving as the internal control. The sequences of the specific primers used are provided in Table 1.

Table 1.

Primers involved in this study

Gene	Sense Primer	Anti-sense Primer
FP	CTGGACTCATCGCAAACACAA	AGGAAGCCTTTGACTTCTGTCTA
SLCO2a	TGAAGCGTTTTGTTTTCCCTCT	CGGGTGTGGAACATCCCATAA
PGDH	GTGAACGGCAAAGTGGCTCT	TCCAATCCACCAATGCTACCT
CCL2	TTAAAAACCTGGATCGGAACCAA	GCATTAGCTTCAGATTTACGGGT
IL-1β	GGGCCTCAAAGGAAAGAATC	TACCAGTTGGGGAACTCTGC
TNF-α	TCTACTGAACTTCGGGGTGATCG	ACGTGGGCTACAGGCTTGTCA
IL-6	TAGTCCTTCCTACCCCAATTTCC	TTGGTCCTTAGCCACTCCTTC
IL-8	AAGAAACCACCGGAAGGAAC	ACTCCTTGGCAAAACTGCAC
β-actin	CCTAAGGCCAACCGTGAAAAG	AGGCATACAGGGACAGCACAG

Western blot

Proteins extracted from meibomian glands were dissolved in cold RIPA buffer containing protease and phosphatase inhibitors. Protein concentration was determined using a BCA Protein Assay Kit (23225; Thermo Fisher Scientific) and adjusted to an equal level. To denature the proteins, the supernatant was boiled at 100 °C for 10 minutes. Equal amounts of protein were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred electronically to polyvinylidene fluoride (PVDF) membranes (catalog no. IPVH00010; Millipore, Billerica, MA, USA). Membranes were blocked with 5% BSA (2 hours at room temperature) and then incubated overnight at 4 °C with primary antibodies targeting NOX-4, 3-NT, 4-HNE, Keap1, SOD2, p-NFκB, NFκB, p-Erk and Erk (all at 1:1000 dilution). Following incubation with HRP-conjugated goat anti-rabbit IgG or goat anti-mouse IgG, protein expression levels were visualized using an enhanced chemiluminescence reagent (NcmECL Ultra®, NCM Biotech, China), recorded with a transilluminator (ChemiDoc XRS System; BioRad, USA), and quantified using Quanty Software (Shanghai, China). β-actin was used as a reference antibody for data normalization.

Statistical analysis

Adverse event reporting signal detection was performed using the disproportionality method, employing reported odds ratio (ROR) [23], proportional reported ratio (PRR) [24], bayesian confidence propagation neural network (BCPNN) [25], and multi-item gamma poisson shrinker (MGPS) [26]. Detailed calculation procedures and screening criteria for all disproportionality methods are provided in Supplementary Tables 1 and 2 and described in our previous studies [27–37]. We further evaluated the distribution of adverse drug reaction reports for preservative-free Latanoprost. Statistical analyses were performed using SPSS (version 26.0; IBM, USA), GraphPad Prism (version 10.1.2), Microsoft Excel 2019, and R (version 4.2.2). Data are presented as mean ± standard deviation (SD). Comparisons between two groups were conducted using Student’s t-test, while comparisons among multiple groups were performed using one-way analysis of variance (ANOVA). A p value < 0.05 was considered statistically significant. R-based analyses employed packages including ggplot2 (version 3.4.4), ggrepel (version 0.9.4), dplyr (version 1.1.4), and DescTools (version 0.99.52).

Results

Adverse reactions and epidemiological characteristics of preservative-free latanoprost

Since 2019, reports of adverse reactions to preservative-free latanoprost have been on the rise, reaching a peak in 2023 (Fig. 1A). Disproportionality analysis revealed significant signals for adverse reactions related to MG, including eyelid irritation (ROR = 83.228), eyelid swelling (ROR = 79.2), eyelid pruritus (ROR = 66.216), blepharitis (ROR = 52.505), and eyelid erythema (ROR = 36.763) (Fig. 1B).

Fig. 1.

Adverse reaction reports and epidemiological distribution of preservative-free latanoprost in the FAERS database. Notes: Fig. 1A Apresents the epidemiological distribution of preservative-free latanoprost; Fig. 1B Bshows the positive signal values for eyelid-related adverse reactions caused by preservative-free latanoprost. Abbreviation: ROR, reporting odds ratio; PRR, proportional reporting ratio; BCPNN, Bayesian confidence propagation neural network; MGPS, multi-item gamma poisson shrinker

Topical application of preservative-free latanoprost induced changes in eyelid phenotype

Compared with vehicle-treated mice, preservative-free latanoprost–treated mice exhibited visibly increased mucinous secretions along the eyelid margins beginning on day 14, and more apparent plugging of the meibomian gland orifices by day 28, based on qualitative slit-lamp observations (Fig. 2A). A closer inspection indicated that no obvious difference in meibomian gland dropout grade was observed between the two groups during the observation period (Fig. 2B). These findings suggest that preservative-free latanoprost–induced MGD remains at an early stage.

Fig. 2.

Effects of topical application of latanoprost on eyelid phenotype in mice. Notes: Fig. 2A Adepicts increased mucinous secretions on the eyelid margins in mice treated with latanoprost eye drops; Fig. 2B Billustrates the differences in meibomian gland dropout grade among groups after latanoprost treatment, with no significant differences observed between groups. Abbreviation: UT, untreated control mice

Topical application of preservative-free latanoprost-induced functional MG changes

Immunofluorescence staining with LipidTOX revealed a significant increase in lipid deposition in the preservative-free latanoprost-treated group (2.58±0.28) compared with the vehicle group (1.20±0.04), representing an approximately 2.15-fold increase on day 28 (p < 0.001) (Fig. 3). Building upon the adverse drug reactions induced by 0.005% preservative-free latanoprost, including blepharitis, as well as our previous research findings [5, 16, 18], we further explored the inflammatory and oxidative stress levels associated with latanoprost-induced meibomian gland inflammation.

Fig. 3.

Immunofluorescence staining results of mouse meibomian glands among groups. Notes: Fig. 3 LipidTox immunofluorescence staining results indicate increased lipid deposition in the acinar units of the latanoprost-treated group on day 28. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Scale bar: 50 μm. abbreviation: UT, untreated control mice

Topical application of preservative-free latanoprost increased prostaglandin accumulation in MG

According to previous reports, FP, as the prostaglandin F receptor, binds prostaglandin F2α and mediates its physiological effects [38]. At day 28 of treatment, immunohistochemical and qRT-PCR analyses revealed a significant increase in FP expression in the meibomian glands of the latanoprost-treated group compared with both the vehicle and UT groups (Fig. 4A). Slco2a is a prostaglandin transporter, whereas PGDH functions as prostaglandin dehydrogenase [39]. Consistently, at day 28, qRT-PCR analysis showed that Slco2a mRNA expression in the latanoprost-treated group increased to approximately 2.70 ± 0.80-fold, compared with 1.10 ± 0.75-fold in the vehicle group (p < 0.05). In contrast, PGDH expression was markedly reduced to approximately 0.25 ± 0.20-fold in the latanoprost group, compared with 0.75 ± 0.30-fold in vehicle-treated mice (p < 0.05) (Fig. 4B). These findings indicate a cumulative effect of latanoprost on prostaglandin-related pathways in mouse meibomian glands at later stages of treatment.

Fig. 4.

Accumulation of prostaglandins in the meibomian glands of mice caused by preservative-free latanoprost. Notes: Fig. 4A Ashows the immunohistochemical results for FP among the different groups; Fig. 4B Bdisplays the changes in the expression levels of prostaglandin-related genes among the groups. After treatment with preservative-free latanoprost, the expression levels of FP and Slco2a were significantly increased, while the expression level of PGDH was significantly decreased. *p < 0.05, **p < 0.01, ***p < 0.001. Scale bar: 50 μm. abbreviation: UT, untreated control mice

Topical application of preservative-free latanoprost stimulated inflammation in MG

Myeloid-derived cells such as monocytes and neutrophils can be identified by the expression of CD11b and Ly6G on the cell surface. In this study, CD11b and Ly6G immunofluorescent staining showed more inflammatory cells infiltrating in MG from the preservative-free latanoprost group compared with both the vehicle group and UT group (Fig. 5A). At day 28, qRT-PCR analysis revealed significant upregulation of multiple inflammatory cytokines and chemokines in the latanoprost-treated group. Specifically, CCL2 mRNA levels increased to 1.45 ± 0.32-fold compared with the vehicle group (1.08 ± 0.30, p < 0.05). IL-1β expression was elevated to 1.52 ± 0.35-fold, relative to 1.05 ± 0.25 in the vehicle group (p < 0.01). Similarly, TNF-α mRNA levels increased to 1.30 ± 0.28-fold compared with 1.01 ± 0.20 in vehicle-treated mice (p < 0.05). Notably, IL-6 showed a marked increase to 4.20 ± 2.10-fold, whereas vehicle-treated mice exhibited levels of 1.20 ± 1.05 (p < 0.05). In addition, CXCL5 expression increased to 1.50 ± 0.40-fold compared with 1.05 ± 0.35 in the vehicle group (p < 0.01) (Fig. 5B). These results suggested that topical application of latanoprost induced chronic inflammation of the MG microenvironment.

Fig. 5.

Topical application of preservative-free latanoprost stimulated inflammation in the meibomian glands. Notes: Fig. 5A Ashows the immunofluorescent results of the CD11b and Ly6G myeloid cell markers. Fig. 5B Billustrates the differences in expression of inflammatory cytokines and chemokines among groups, including upregulated levels of CCL2, IL-1β, TNF-α, IL-6, and CXCL5 in the latanoprost-treated group. *p < 0.05, **p < 0.01, ***p < 0.001. Scale bar: 50 μm. abbreviation: UT, untreated control group

The intracellular signaling pathways involved in latanoprost regulation of cytokine and chemokine output in MG

In addition, we analyzed the activation of signaling pathways among the three groups. The results showed significant activation of the Erk and NF-κB signaling pathways in the Latanoprost group compared to the other two groups (Fig. 6). The marked activation of these two pathways suggests the potential presence of significant chronic inflammatory activation in the meibomian glands following Latanoprost administration in mice.

Fig. 6.

Changes in the expression levels of pathway proteins in the meibomian glands of mice treated with preservative-free latanoprost. Notes: local application of preservative-free latanoprost in mice significantly increased the expression levels of Erk and NF-κB proteins

Topical application of preservative-free latanoprost promoted oxidative stress in MG

Numerous studies have shown that inflammation and oxidative stress are closely linked and promote each other in the pathogenesis of many chronic diseases. We detected the levels of NADPH oxidase subunits and oxidative stress markers by immunohistochemical staining and western blot and found that topical application of latanoprost promoted the expression of NOX4, 4HNE, and 3-NT in MG (Fig. 7A–B). Meanwhile, we investigated the changes in the antioxidant defense system. Western blotting demonstrated that latanoprost treatment showed an elevated protein level of keap1 and decreased protein levels of SOD2 (Fig. 7C). These results, taken together, suggested that topical application of latanoprost causes oxidative stress in MG.

Fig. 7.

Changes in markers of oxidative stress induced by latanoprost in mouse meibomian glands. Notes: Figs. 7A and B depict significant increases in markers of oxidative stress induced by latanoprost in mouse meibomian glands, such as NOX4, 3-NT, and 4-HNE. Conversely, antioxidant stress-related indicators significantly decreased, including SOD2. Additionally, higher levels of Keap-1 protein indicate weaker antioxidant capacity, which significantly increased in the latanoprost group. *p < 0.05, **p < 0.01, ***p < 0.001. Scale bar: 50 μm. abbreviation: UT, untreated control group

Discussion

Previous studies have suggested that adverse reactions to latanoprost eye drops, particularly ocular surface reactions, are mainly attributable to preservatives [40]. However, based on our epidemiological analysis and adverse event reports, we observed that with the increasing use of preservative-free latanoprost, reports of eyelid-related adverse reactions-including irritation, swelling, itching, inflammation, and redness-have continued to rise over recent years. These symptoms are commonly associated with MGD. Clinical studies have also demonstrated an increased incidence of MGD in patients receiving long-term latanoprost treatment [41]. To further investigate this association, we conducted animal experiments to evaluate the effects of preservative-free latanoprost on meibomian gland function and to explore potential underlying mechanisms. Our results showed that topical application of preservative-free 0.005% latanoprost led to obstruction of meibomian gland orifices and lipid accumulation within the glands in mice. In addition, latanoprost treatment induced oxidative stress and inflammatory responses in meibomian gland tissues. These changes are characteristic features of clinical MGD. Collectively, these findings suggest that latanoprost-induced MGD may not be solely attributable to preservatives but may also be directly related to latanoprost itself. Latanoprost is an ester prodrug of PGF2α, a known pro-inflammatory mediator. Upon binding to the FP receptor, PGF2α can activate inflammatory signaling pathways. Consistent with this, our previous study demonstrated that latanoprost interacts with FP receptors on the ocular surface, triggering inflammation and dry eye–like changes in mice [16].

In this study, after the local application of latanoprost in mice, not only did we observe an increase in the FP expression in the meibomian gland tissue but interestingly, we also discovered significant upregulation Slco2a in the latanoprost subject group, meanwhile PGDH was considerably downregulated. Slco2a plays a role in the transmembrane transport of prostaglandins, facilitating their accumulation. Conversely, PGDH, which is responsible for the metabolism and degradation of prostaglandins, sees a reduction that diminishes prostaglandin breakdown. These changes collectively contribute to an increased prostaglandin presence [42–44]. Thus, it becomes evident that local use of latanoprost eye drops not only allows latanoprost itself to bind with FP receptors but also enables the meibomian gland to produce more PGF2α, creating a positive feedback loop. These findings suggest that even preservative-free latanoprost can still lead to the build-up of prostaglandins within the meibomian gland when locally applied to the mouse eye, thereby instigating relevant inflammatory responses.

Ocular surface inflammation plays a significant role in MGD, with the disease’s pathophysiological processes heavily involving inflammatory responses [45, 46]. Multiple inflammatory signaling pathways were activated in MGD, such as NF-κB, MAPK, and PI3K [47–49], leading to the release of various pro-inflammatory cytokines and chemokines like CCL2, IL-1β, TNF-α, CXCL5, and IL-6. NF-κB is a key transcriptional activator for many pro-inflammatory cytokines, including IL-1β and TNF-α [50]. Additionally, it has been shown that Latanoprost can trigger the P38-NF-κB signaling pathway in human fibroblasts, leading to a heightened release of inflammatory cytokines and chemokines [51]. In our study, treatment with preservative-free Latanoprost led to a significant activation of Erk, a key protein in the MAPK signaling pathway, and NF-κB in mouse meibomian glands. The significant upregulation of inflammation-related markers CD11b, Ly6G, CCL2, IL-1β, TNF-α, CXCL5, and IL-6 further validates the critical role of inflammation in the dysfunction of meibomian glands following Latanoprost treatment. Our findings suggest that the binding of Latanoprost to its FP receptors may induce inflammation in the meibomian gland, involving the activation of NF-κB and Erk signaling pathways and the production of inflammatory cytokines. Hence, we further substantiate that preservative-free latanoprost induces meibomian gland dysfunction in mice through the provocation of inflammatory responses within the meibomian gland. These results align with previous reports regarding the pathological mechanisms of MGD [52].

Oxidative stress is the imbalance between the oxidation and antioxidation systems of cells and tissues, resulting from excessive production of reactive oxygen species (ROS) [53]. Our previous research results demonstrated a significant increase in oxidative stress levels in MGD mice induced by a high-fat and high-sugar diet [18]. Additionally, NOX4, 3-NT, and 4HNE proteins play important roles in the oxidative stress process [54]. NOX4 is primarily involved in nuclear oxidative stress processes [55], and tyrosine may be oxidized to form 3-NT under conditions of oxidative stress, making 3-NT concentration a marker of cellular or tissue oxidative stress activity [56]. Our study revealed a significant increase in the levels of NOX4, 3-NT, and 4HNE proteins involved in oxidative stress processes in the meibomian gland of mice treated with preservative-free latanoprost. Conversely, the levels of key proteins involved in the anti-oxidative stress process, including SOD2 [57], were significantly decreased. Therefore, we speculate that one of the potential mechanisms underlying MGD induced by preservative-free latanoprost in mice is an increase in oxidative stress levels and a decrease in anti-oxidative stress levels.

Antioxidant enzymes such as peroxidases are expressed in the human meibomian gland and conjunctival tissues, and previous studies have already confirmed that oxidative stress plays a crucial role in the changes associated with corneal epithelial damage [58]. The production of ROS is related to the activity of nicotinamide adenine dinucleotide phosphate (NADPH oxidase), particularly in phagocytic immune cells and endothelial cells, with inflammation also playing a role in this process [59]. When ROS is produced chronically or over a long duration, accumulating beyond the body’s capacity to clear ROS, related oxidative metabolites become major mediators of inflammatory pathological responses, promoting the initiation and cascade of inflammatory responses [60]. The pro-inflammatory activity of these substances is partially related to cells of the immune system, such as polymorphonuclear neutrophils [61]. Therefore, the ROS metabolites that gradually accumulate during oxidative stress are drivers of inflammatory responses, further exacerbating local inflammation in the mouse meibomian gland. During the inflammatory process, activated phagocytes such as neutrophils and macrophages produce a large amount of ROS, thereby further aggravating oxidative stress damage [62]. Hence, the inflammatory process can induce oxidative stress, and oxidative stress can also induce inflammation through the activation of various pathways; oxidative stress and inflammation are closely related pathophysiological processes. Our research demonstrates that both factors play a pivotal role in inducing MGD in mice when preservative-free latanoprost is applied topically to the eye, and they may interact to form a vicious cycle.

Conclusion

In summary, according to our research findings, we have uncovered for the first time that preservative-free latanoprost also has the potential to induce MGD in mice. This matches epidemiological data reported in the adverse drug reaction database regarding the adverse reactions of preservative-free latanoprost. We have also verified that inflammatory damage and oxidative stress are significant mechanisms involved in this process. Our study validates the mechanisms from databases to animal experiments for pharmacovigilance and medication safety. Therefore, glaucoma clinicians need to consider the possible adverse effects of long-term use of latanoprost when prescribing glaucoma patients, especially those with pre-existing meibomian gland dysfunction or limbal inflammation.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Caihong Huang, Yiran Yang, and Shinan Wu conceived the research idea. Caihong Huang conducted data cleaning and a literature review. Caihong Huang, Yiran Yang, Shinan Wu, Zhaolin Liu, Lin Chen, Dan Yan and Mingyan Wei contributed to drafting and critically revising the work for intellectual content. Caihong Huang and Shinan Wu conducted the analysis and created the figures and tables. Ke Yan, Ruochen Wang, Jiaoyue Hu, Wei Li, and Zuguo Liu provided a critical review of the manuscript. All authors have read and approved the manuscript.

Funding

Supported by grants from the Natural Science Foundation of Fujian (2023J01012, CH), and National Natural Science Foundation of China (82271054, ZL).

Data availability

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: FAERS Publish Dashboard (https://www.fda.gov/drugs/questions-and-answers-fdas-adverse-event-reporting-system-faers/fda-adverse-event-reporting-system-faers-public-dashboard).

Declarations

Competing interest

The authors declare no competing interests.