Ly93 Inhibits Sphingomyelin Synthesis and Attenuates Inflammation and Injury in Dry Eye Conjunctival Organoids

Abstract

Purpose

We aimed to investigate the pathogenic role of sphingomyelin (SM) in dry eye disease (DED) and evaluate Ly93, an SM synthase inhibitor, as a potential therapy for DED-associated inflammation.

Methods

Tear samples from DED patients and healthy participants were analyzed for SM concentrations using liquid chromatography-tandem mass spectrometry. Correlations between SM levels and clinical parameters were evaluated. The effects of SM on normal conjunctival organoids and the therapeutic efficacy of Ly93 on hyperosmolarity-induced DED conjunctival organoids were investigated. SM levels after treatment with Ly93 were also measured. The inflammatory cytokines (IL-1β, IL-6, and MMP-9), signaling pathways (STAT1, nuclear factor-κB, and PI3K), apoptosis (caspase-3 and caspase-9), ferroptosis (Fe²⁺, GPX4, TFR, and 4-HNE) were analyzed.

Results

Tear SM concentrations were significantly elevated in DED patients and positively correlated with disease severity. Exogenous SM upregulated STAT1, IL-1β, IL-6, and inducible nitric oxide synthase expression in conjunctival organoids. Hyperosmolarity-induced DED conjunctival organoids exhibited elevated MMP-9 and IL-1β, activated signaling pathways (nuclear factor-κB, STAT1, and PI3K), increased apoptosis (caspase-3 and caspase-9), and ferroptosis (elevated Fe²⁺, TFR, and 4-HNE, and decreased GPX4). Ly93 treatment effectively reduced SM levels and attenuated inflammation, apoptosis, and ferroptosis in the DED organoid model.

Conclusions

Our study demonstrates that elevated SM promotes ocular surface inflammation and cellular injury. Inhibition of SM synthase by Ly93 alleviates DED-associated inflammation and injury, suggesting that targeting SM metabolism represents a promising therapeutic strategy.

Dry eye disease (DED) is the most prevalent disorder of the ocular surface, with a reported global prevalence ranging from 5% to 50%.¹ Although inflammation is considered the core of the vicious circle of DED,² the underlying molecular mechanisms remain incompletely clarified. Recent clinical investigations have revealed abnormalities in sphingolipid metabolism in DED,³ yet the contribution of sphingolipids to the initiation and progression of DED remain poorly understood.

Sphingomyelin (SM) is a major sphingolipid enriched in the plasma membrane, endocytic recycling compartment, and trans-Golgi network of mammalian cells.⁴ SM homeostasis is essential for maintaining cellular integrity and functions, whereas its dysregulation has been implicated in diverse pathological processes, including inflammation, atherosclerosis, cancer, diabetes, and Niemann–Pick disease.⁵ Elevated SM levels have been reported in both meibum and tears of patients with DED.^6,7 Pharmacological inhibition of SM synthase 2 (SMS2) using the compound 14I has been shown to reduce SM synthesis and alleviate DED signs, such as corneal fluorescein staining (CFS) and reduced tear secretion, as well as ocular surface inflammation (TNF-α, IL-1β, and MMP-9) in mouse models.⁸ However, the mechanistic role of SM in DED pathogenesis, particularly at the cellular and molecular levels, remains to be elucidated.

Human conjunctival organoids have emerged as a robust in vitro model to investigate ocular surface homeostasis and disease, recapitulating key structural and functional features of native conjunctival epithelium and enabling continuous expansion.⁹ Hyperosmotic stress, a hallmark of DED, has been shown to induce the release of inflammatory mediators (nuclear factor-κB [NF-κB], TNF-α, MMP-9, IL‐1β, IL‐6, and IL‐8) and epithelial cell injury in corneal epithelial organoid models,¹⁰ suggesting that hyperosmolarity-induced conjunctival organoids may serve as a relevant experimental platform for mechanistic and therapeutic studies in DED.

This study aimed to elucidate the mechanistic role of SM in the inflammatory response and cellular injury associated with DED. We first quantified tear SM levels in patients with DED and healthy controls and analyzed correlations with clinical parameters. Human conjunctival organoids were used as an in vitro model to evaluate the effect of SM on DED-associated inflammation and injury. Moreover, a hyperosmolarity-induced DED organoid model was then established to evaluate whether Ly93, a selective SMS2 inhibitor, could reduce SM and mitigate inflammation, apoptosis, and ferroptosis.

Methods

Participants and Ocular Examination

This study was approved by the Medical Ethics Committee of Beijing Tongren Hospital (TRECKY2019-130) and conducted in accordance with the Declaration of Helsinki. Ten DED patients and eight age-matched healthy controls were enrolled and provided informed consent.

Inclusion criteria for the DED group: age ≥18 years old, ocular surface disease index (OSDI) of ≥13, and tear break-up time (TBUT) of <10 seconds.¹¹ Inclusion criteria for the control group were age ≥18 years old, OSDI of <13, and TBUT of ≥10 seconds. Participants were excluded if they presented with other ocular surface diseases or systemic conditions known to cause DED.

Ocular symptoms were quantified using the OSDI questionnaire (range, 0–100).¹² Ocular examinations were then performed, including measurement of CFS, TBUT, meibum secretion and quality, and lid margin abnormality according to established protocols.^13,14 CFS was graded using the Oxford 5-point scale.¹⁵ Meibum quality was graded by a 0 to 4 scale (0, clear; 1, cloudy; 2, granular; 3, toothpaste; and 4, no meibum extracted).¹⁶ The lid margin was graded based on four parameters: irregularities, telangiectasia, orifice obstruction, and Marx line displacement, with each abnormality scored as 1 point.¹⁵ Nonstimulated tear samples (50 µL) were collected from each participant using sterile disposable microcaps (Drummond Scientific, Broomall, PA, USA), and stored at −80°C until lipidomic analysis.¹⁵

Establishment of Human Conjunctival Organoids

The human tissue experiments complied with the guidelines of the ARVO Best Practices for Using Human Eye Tissue in Research (Nov2021), were approved by the Institutional Review Board of Beijing Tongren Hospital (TRECKY2021-024) and were conducted in accordance with the tenets of the Declaration of Helsinki. Human conjunctival organoids were generated as previously described.⁹ Approximately 1 mm³ of deidentified healthy conjunctival tissue was obtained from pterygium surgery patients at the corneal department of Beijing Tongren Hospital and transported to the laboratory in Advanced DMEM/F12 (ADF) medium (Gibco, Grand Island, NY, USA). Within 4 hours of harvesting, tissues were dissected without removal of the fibrous layer and digested in 0.25% trypsin-EDTA at 37°C for 10 to 15 minutes, during which the suspension was gently pipetted every 5 minutes. Digestion was terminated with ADF medium, followed by twice centrifugations at 500×g for 5 minutes.

Cell pellets were resuspended in Matrigel (356231, Corning, Corning, NY, USA) and plated (25 µL/well) in 48-well plates, polymerized at 37°C for 20 to 30 minutes, and overlaid with expansion medium containing ADF medium, 10 mM HEPES, GlutaMAX (1×), 100 U/mL and 0.1 mg/mL penicillin/streptomycin, B27 Supplement (1×), 1.25 mM N-acetylcysteine, 0.1 µg/mL Noggin, 5% R-spondin 1 conditioned medium (produced as described by Pleguezuelos-Manzano et al¹⁷), 100 ng/mL FGF1, 100 ng/mL FGF10, 0.15 nM Wnt Surrogate, 3 µM A83-01, 10 µM ROCK inhibitor Y-27632, 1 µM Forskolin, and 100 µg/mL Primocin. The expansion medium was changed every 2 to 3 days. Established organoids (three-dimensional [3D] spheroids exhibiting a stable, differentiated morphology and are ready for passaging) were passaged at ratios of 1:2 to 1:4 every 9 to 11 days.

SM Treatment of Conjunctival Cells and Organoids

After 10 days of culture, conjunctival organoids were dissociated into single cells using 0.25% trypsin-EDTA. The cells were incubated at 37°C for 12 hours in expansion medium supplemented with a SM mixture (CAS No. 85187-10-6, HY-113498, MedChemExpress, Monmouth Junction, NJ, USA) at final concentrations of 20, 40, 80, or 160 µM. The SM solution was prepared in normal saline with 20% (w/v) sulfobutylether-β-cyclodextrin (SBE-β-CD). Vehicle-treated cells (expansion medium supplemented with an equivalent volume of the 20% SBE-β-CD solution) served as controls. Cell viability was assessed using a Viability/Cytotoxicity Assay Kit for Animal Live and Dead Cells (Biotium, Hayward, CA, USA) according to the manufacturer's protocol.¹⁸ Live cells were visualized by green fluorescence from the esterase substrate calcein-AM, and dead cells were identified by red fluorescence from the membrane-impermeable DNA dye ethidium homodimer III. Images were acquired using a fluorescence microscope (Olympus BX50, Tokyo, Japan).

For organoid experiments, conjunctival organoids were cultured for 8 days before SM treatment. The culture medium was then replaced with fresh medium containing SM at concentrations of 5, 10, 20, 40, 80, or 160 µM. Vehicle-only medium (expansion medium supplemented with an equivalent volume of the 20% SBE-β-CD) was used for controls. After 48 hours of incubation, the organoids were collected for subsequent analysis.

Establishment of Hyperosmolarity-Induced DED Organoid Model and Ly93 Intervention

A hyperosmolarity-induced DED organoid model was generated following a previously described protocol with minor modifications.¹⁰ In brief, conjunctival organoids cultured for 8 days were incubated in expansion medium supplemented with NaCl to achieve osmolarities of 510 mOsm/L (90 mM), 560 mOsm/L (120 mM), or 620 mOsm/L (150 mM). Vehicle-only medium (expansion medium supplemented with an equivalent volume of ddH₂O) served as controls. The morphological change induced by hyperosmotic stress was quantified by determining the percentage of intact organoids. Organoids were classified as disintegrated if they exhibited a loss of structural integrity, characterized by blurred and irregular boundaries, potentially accompanied by the shedding of cells into the surrounding Matrigel. Conversely, intact organoids were defined as those maintaining a complete, solid 3D architecture with a smooth boundary. The total number of intact and disintegrated organoids was quantified manually from 4 randomly selected high-magnification (200×) images per group using ImageJ software (National Institutes of Health, Bethesda, MD, USA). The intact percentage was calculated using the following formula: Intact organoids percentage = Intact organoids/(Intact organoids + Disintegrated organoids) × 100%.

The cytotoxicity of Ly93 was first assessed in both conjunctival cells and organoids. Conjunctival cells were incubated for 12 hours with Ly93 at concentrations of 12.5, 25.0, 50.0, or 100.0 µM,¹⁹ followed by viability testing using the Viability/Cytotoxicity Assay Kit. Organoids were split and cultured for 11 days in medium containing the varying Ly93 concentrations. Vehicle-treated cells (expansion medium supplemented with an equivalent volume of dimethyl sulfoxide) served as controls. The percentage of organoids formed from total conjunctival cells was quantified using ImageJ software.

Based on these results, a nontoxic concentration of Ly93 (25 µM) was determined for intervention studies. Conjunctival organoids were divided into three groups: vehicle group (expansion medium supplemented with equivalent volumes of vehicles only), DED group (expansion medium with 120 mM NaCl), and DED + Ly93 group (expansion medium with 120 mM NaCl and 25 µM Ly93). After 8 days of organoid culture, the assigned treatment medium was applied for 48 hours, after which organoids were harvested for downstream analyses.

Lipidomic Analysis

The samples (human tear sample, n = 10 of DED and n = 8 of control; organoids, n = 4 per group) were thawed on ice and extracted with 0.5 mL of methyl-tert-butyl ether/methanol (3:1, v/v) containing an internal standard mixture. After vortexing for 15 minutes, 100 µL of water was added, followed by vortexing for 1 minute and centrifugation at 12,000 rpm for 10 minutes. The upper organic phase (200 µL) was collected, evaporated to dryness under vacuum, and reconstituted in 200 µL of acetonitrile/isopropanol (1:1, v/v) for liquid chromatography-tandem mass spectrometry analysis.²⁰

The sample extracts were analyzed using a UPLC system (ExionLC) coupled to a QTRAP 6500+ mass spectrometer (Sciex) equipped with an electrospray ionization source (LC-ESI-MS/MS system). Chromatographic separation was achieved on a Thermo Accucore C30 column (2.6 µm, 2.1 mm × 100 mm) using a binary solvent system: A: acetonitrile/water (60/40, v/v) with 0.1% formic acid and 10 mmol/L ammonium formate; and B: acetonitrile/isopropanol (10/90, v/v) with 0.1% formic acid and 10 mmol/L ammonium formate. A multistep gradient elution was used: 80:20 (A: B) at 0 minute, 70:30 at 2.0 minutes, 40:60 at 4 minutes, 15:85 at 9 minutes, 10:90 at 14 minutes, 5:95 at 15.5 minutes, 5:95 at 17.3 minutes, then returned to 80:20 at 17.3 minutes and held until 20 minutes. The flow rate was 0.35 mL/min, column temperature 45°C, and injection volume of 2 µL.

The electrospray ionization source was operated in both positive and negative ion modes under the following conditions: source temperature 500°C; ion spray voltage +5500 V/−4500 V; gas 1, gas 2, and curtain gas pressures at 45, 55, and 35 psi, respectively; collision gas set to medium. Multiple reaction monitoring transitions were optimized for each lipid species. Data acquisition and processing were performed using Analyst 1.6.3 (Sciex, Framingham, MA, USA). Among these data, SM concentrations were normalized using z-score transformation and visualized via heatmaps.

RNA Extraction and RT-qPCR

Three independent samples from each group were used RT-qPCR. Total RNA was extracted using the RNA Easy Fast Tissue/Cell Kit (Tiangen, Beijing, China) and reverse-transcribed into cDNA with HiScript III All-in-One RT SuperMix (Vazyme, Nanjing, China). qPCR was performed on an ABI 7500 system (Applied Biosystems, Foster City, CA, USA) using Taq Pro Universal SYBR qPCR Master Mix (Vazyme). Cycling conditions were: 95°C for 30 seconds, followed by 40 cycles of 95°C for 10 seconds and 60°C for 30 seconds. Relative gene expression levels were calculated by the 2^^−ΔΔCt method. Primer sequences were as follows: IL-1β (forward 5′-AGCTACGAATCTCCGACCAC-3′; reverse 5′-CGTTATCCCATGTGTCGAAGAA-3′), IL-6 (forward 5′-ACTCACCTCTTCAGAACGAATTG-3′; reverse 5′-CCATCTTTGGAAGGTTCAGGTTG-3′), and STAT1 (forward 5′-CAGCTTGACTCAAAATTCCTGGA-3′, reverse 5′-TGAAGATTACGCTTGCTTTTCCT-3′).

Frozen Section Preparation and Periodic Acid-Schiff (PAS) Staining

Conjunctival organoid samples were embedded in optimal cutting temperature compound and stored at −80°C until use. Samples were cryosectioned at a thickness of 5 µm and stored at −80°C until further processing.

PAS staining was performed on frozen sections according to the manufacturer's instructions (PAS Stain Kit, Solarbio, Beijing, China). Briefly, sections were hydrated in distilled water. The sections were then treated with PAS oxidant for 2 minutes and rinsed twice with distilled water for 1 minute each. Subsequently, sections were incubated with Schiff reagent in the dark for 15 minutes. After rinsing with distilled water for 1 minute, the sections were counterstained with hematoxylin for 30 seconds.

Immunofluorescence Staining

Immunofluorescence staining was performed to evaluate the expression of K19 (1:200, Cell Signaling Technology, Danvers, MA, USA), K13 (1:200, Cell Signaling Technology), IL-1β (1:200, Abcam, Cambridge, UK), STAT1 (1:200, Proteintech, Rosemont, IL, USA), p63 (1:250, Abcam), Ki67 (1:200, Abcam), NF-κB p65 (1:200, Abcam), MUC1 (1:200, Abcam), MUC5AC(1:250, Abcam), caspase-9 (1:200, Cell Signaling Technology), ZO-1 (1:200, Abcam), and 4-hydroxynonenal (4-HNE) (1:100, Invitrogen, Carlsbad, CA, USA) in frozen sections of the conjunctival organoids. Sections were incubated overnight at 4°C with these primary antibodies, followed by 1 hour of incubation with Alexa Fluor 488-conjugated goat anti-rabbit (ab150077, 1:500, Abcam) or goat anti-mouse (ab150113, 1:500, Abcam) secondary antibodies. Fluorescence images were acquired using a fluorescence microscope (Olympus BX-51) and were quantified using ImageJ software.

Protein Extraction and Western Blotting

Conjunctival organoids were lysed in radioimmunoprecipitation assay buffer (Beyotime, Shanghai, China) supplemented with 1% phenylmethanesulfonyl fluoride, 1% protease inhibitor, and 1% phosphatase inhibitor. Lysates were sonicated for 2 minutes, incubated on ice for 20 minutes, and centrifuged at 13,000×g for 10 minutes to collect the supernatant. Protein concentration was determined using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA).

Equal amounts of protein were resolved by SDS-PAGE on 12% polyacrylamide gels and transferred to polyvinylidene fluoride membranes.²¹ Membranes were blocked with skim milk and incubated overnight at 4°C with primary antibodies against IL-1β (1:1000, Abcam), phospho-STAT1 (1:1000, Proteintech), STAT1 (1:1000, Proteintech), inducible nitric oxide synthase (iNOS) (1:1000, Proteintech), p63 (1:1000, Abcam), NF-κB p65 (1:1000, Abcam), Ki67 (1:1000, Abcam), MMP-9 (1:1000, Abcam), glutathione peroxidase 4 (GPX4, 1:1000, Abcam), PI3K (1:1000, Abcam), caspase-3 (1:1000, Cell Signaling Technology), caspase-9 (1:1000, Cell Signaling Technology), transferrin receptor (TFR, 1:1000, Abcam), IRE1 (1:1000, Abcam), AKT (1:1000, Abcam), phospho-AKT (1:1000, Abcam), and β-actin (AC026, 1:50000, Abconal). After washing, membranes were incubated with goat anti-rabbit (Ab205718, 1:5000, Abcam) or goat anti-mouse (Ab205719, 1:5000, Abcam) secondary antibodies. Protein bands were visualized using the Tanon 5200 chemiluminescent imaging system (Tanon, Shanghai, China).

Intracellular Fe²⁺ Analysis

Intracellular Fe²⁺ levels in conjunctival organoid frozen sections were assessed using the FerroOrange kit (Dojindo, Kumamoto, Japan) according to the manufacturer's Instructions. Briefly, sections were washed with Hank's Balanced Salt Solution and then incubated with 1 µmol/L FerroOrange solution at 37°C for 30 minutes in the dark. Red fluorescence intensity was immediately visualized using fluorescence microscopy.

Statistical Analysis

All statistical analyses were performed using GraphPad Prism software (version 9.5.1, La Jolla, CA, USA). Data normality was assessed using the Shapiro–Wilk test. For comparisons between two groups, the unpaired Student t test was applied for normally distributed data, and the Mann–Whitney U test was used for non-normally distributed data. For comparisons among multiple groups, one-way ANOVA followed by Tukey's post hoc test was used for normally distributed data, and the Kruskal–Wallis test followed by Dunn's post hoc test was applied for non-normally distributed data. All statistical tests were two tailed, and a P value of <0.05 was considered statistically significant.

Results

SM Levels in DED and Association With Clinical Parameters

Lipid profiling of tear samples revealed no significant difference in total lipid content between patients with DED and healthy controls (P = 0.6544) (Fig. 1A). In contrast, total SM level was significantly elevated in the DED group compared with the control group (P = 0.0155) (Fig. 1B). Twenty-five SM species were detected in tears, with these species exhibiting an increased trend in DED group (Fig. 1C). Ten SM species (d18:0/14:0, d18:0/16:0, d18:0/17:0, d18:0/18:0, d18:0/20:0, d18:1/14:0, d18:1/15:0, d18:1/16:0, d18:1/17:0, and d18:2/22:1) were significantly elevated in the DED group compared with the control group (all P < 0.05) (Fig. 1D). Furthermore, total SM and these 10 species of SM positively correlated with CFS, lid margin score, and meibum quality score, while showing negative correlations with TBUT and tear secretion (Fig. 1E).

Figure 1.

Tear SM levels in patients with DED and their association with clinical parameters. (A, B) Quantification of total tear lipids (A) and total SM level (B) in DED patients compared with healthy controls. (C) Heatmap displaying 25 SM species detected in tears. Color intensity represents the z-score (red, positive; blue, negative). (D) Analysis of the 10 SM species that were significantly elevated in the tears of DED patients. (E) Spearman's correlation between SM levels and clinical parameters (OSDI, CFS, lid margin score, meibum quality score, TBUT, and tear secretion). Color scale indicates the Spearman's correlation coefficient (range, −1 to 1). *P < 0.05; ns, not significant.

Establishment and Characterization of Human Conjunctival Organoids

Primary human bulbar conjunctival tissues were dissected, enzymatically dissociated, and embedded in Matrigel (Fig. 2A). The organoid expansion medium was supplemented with several key molecules, including B27, FGF1, FGF10, N-acetylcysteine, Wnt surrogate, Y-27632, A83-01, and forskolin (Fig. 2A). Within 3 to 4 days, the cells self-organized into dense organoid structures, characterized by tightly packed cells forming 3D solid spheroids (Fig. 2B). These organoids were fully established by around day 10, with passaging conducted every 9 to 11 days (Fig. 2B). PAS staining was positive in conjunctival organoids, showing the dense cellular structure and active secretory function (Fig. 2C). Immunofluorescence analysis confirmed the identity of the organoids, showing prominent expression of K19 and K13, along with p63⁺ basal cells and Ki67⁺ proliferating cells (Fig. 2D). Furthermore, ZO-1 was consistently localized at cell-cell junctions, demonstrating the formation of a robust and integrated epithelial barrier (Fig. 2D). The presence of MUC1- and MUC5AC-positive cells further confirmed that the organoids retained the mucin-secreting capacity characteristic of the native human conjunctiva (Fig. 2D).

Figure 2.

Establishment of human conjunctival organoids. (A) Schematic workflow for the generation of human conjunctival organoids from primary conjunctival tissue. (B) Representative images of organoids after 10 days of culture at passages 0, 1, and 2 (scale bar, 200 µm). (C) PAS staining in the conjunctival organoids (scale bar, 50 µm). (D) Immunofluorescence staining showing expression of K19, K13, p63, Ki67, ZO-1, MUC1, and MUC5AC in passage day 10 organoids (scale bar, 50 µm).

Hyperosmolarity Induced DED-Associated Injury and Inflammation

Treatment with hyperosmotic medium (560 and 610 mOsm/L) for 48 hours resulted in pronounced organoid disintegration (Fig. 3A) and a significant decrease in intact organoid percentage (Fig. 3B). Ki67 levels decreased in all hyperosmotic groups, whereas p63 expression increased at 510 mOsm/L, but decreased at higher osmolarities (Fig. 3C). MMP-9 and IL-1β were elevated across all hyperosmotic groups (Fig. 3C). NF-κB p65, STAT1, and PI3K activated at 560 mOsm/L, along with increased caspase-3 expression (Fig. 3C). TFR was elevated at 560 and 610 mOsm/L, and GPX4 decreased at 610 mOsm/L, suggesting ferroptotic involvement (Fig. 3C). Immunofluorescence analysis revealed that the 560 mOsm/L group exhibited a significant decrease in the percentage of Ki67⁺ cells (17.99% ± 3.18% vs. 54.79% ± 4.19%; P = 0.0003), although no statistically significant difference was observed in the proportion of p63⁺ cells (66.55% ± 17.58% vs. 63.76% ± 3.29%; P = 0.8001; Fig. 3D–F). Additionally, the protein expression of caspase-9, NF-κB p65, STAT1, and IL-1β within the organoid tissues showed an upward trend in the 560 mOsm/L group (Fig. 3G). Based on these findings, a hyperosmolarity of 560 mOsm/L was selected to establish the DED model for subsequent experiments.

Figure 3.

Hyperosmolarity induced DED-associated injury and inflammation of conjunctival organoids. (A) Representative images showing hyperosmolarity-induced organoid disintegration (scale bar, 100 µm). (B) Quantification of the percentage of intact organoids. (C) Western blotting analysis of proliferation markers (p63 and Ki67), inflammatory mediators (MMP-9 and IL-1β), signaling molecules (NF-κB p65, STAT1, and PI3K), ferroptosis markers (TFR and GPX4), and apoptosis marker (caspase-3). (D–F) Immunofluorescence staining of p63⁺ and Ki67⁺ cells in organoids (scale bar, 50 µm). (G) Immunofluorescence staining of caspase-9, NF-κB p65, STAT1, and IL-1β in organoids (scale bar, 50 µm). *P < 0.05; ***P < 0.001; ****P < 0.0001; ns, not significant.

Exogenous SM Induced Inflammation and Injury of Conjunctival Organoids

SM concentrations of ≥40 µM induced conjunctival cell death, whereas organoids exhibited no overt structural damage (Fig. 4A). Western blotting demonstrated that SM concentrations of ≥20 µM upregulated phospho-STAT1, STAT1, IL-1β, and iNOS protein levels (Figs. 4B, 4C). Consistently, RT-qPCR revealed significantly increased IL-1β, IL-6, and STAT1 mRNA expression in SM-treated organoids (all P < 0.05) (Figs. 4D–F). Immunofluorescence staining revealed a significant reduction in K13 expression in organoids treated with 160 µM SM, whereas no significant changes were observed at 40 and 80 µM concentrations (P = 0.0064, 0.9487, and 0.9858, respectively) (Figs. 4G, 4H). K19 protein levels showed no significant differences between the vehicle and any SM-treated groups (all P > 0.2268) (Figs. 4G, 4I). Notably, all SM-treated groups exhibited a significant decrease in the percentage of Ki67⁺ cells (all P < 0.001) (Figs. 4G, 4J). SM treatment at 80 µM significantly increased MUC1 levels (P = 0.0139) (Figs. 4G, 4K). Additionally, MUC5AC expression followed a nonlinear trend, peaking at 40 and 80 µM before declining significantly under 160 µM SM stress (all P < 0.001) (Figs. 4G, 4L).

Figure 4.

Exogenous SM induced inflammation and injury of conjunctival organoids. (A) Effects of SM treatment on the viability of conjunctival cells and morphology of organoids (scale bar, 100 µm for cells and 200 µm for organoids). (B, C) Western blotting analysis of inflammatory markers (phospho-STAT1, STAT1, IL-1β, and iNOS) in SM-treated organoids. (D–F) Relative mRNA expression of IL-1β, IL-6, and STAT1 after SM treatment. (G) Immunofluorescence staining of K13, K19, Ki67, MUC1, and MUC5AC in SM-treated organoids (scale bar, 50 µm). (H–L) Comparative analysis of immunofluorescent results: K13 (H), K19 (I), Ki67 (J), MUC1 (K), and MUC5AC (L). *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; ns, not significant.

Ly93 Reduced SM Levels and Alleviated DED-Associated Inflammation and Injury

Ly93 exhibited a dose-dependent cytotoxic profile, showing no adverse effects on conjunctival cell viability at 12.5 and 25.0 µM, but significant inhibition at 50.0 and 100.0 µM (Fig. 5A). This dose-dependent toxicity was consistently observed during organoid development (Figs. 5B, 5C). The organoid formation rate was significantly inhibited in the 50 and 100 µM groups compared with the control (both P < 0.0001) (Fig. 5C). Conversely, no significant impact on organoid formation rate was observed in both the 12.5 and 25.0 µM groups (P = 0.6029 and 0.2433, respectively) (Fig. 5C). Based on these safety profiles, 25 µM Ly93 was selected for subsequent experiments.

Figure 5.

Ly93 attenuated DED-associated injury in conjunctival organoids. (A) Live/dead cell staining of conjunctival cells treated with varying Ly93 concentrations. Green fluorescence presented live cells, and red fluorescence presented dead cells (scale bar, 100 µm). (B, C) Effects of Ly93 on conjunctival organoid formation (scale bar, 200 µm). (D) Damaged organoids in DED and DED + Ly93 groups. (E) Comparison of the percentage of intact organoids. (F–H) Immunofluorescence staining and quantification of p63⁺ and Ki67⁺ cells. *P < 0.05; **P < 0.01; ****P < 0.0001; ns, not significant.

In the hyperosmolarity-induced DED conjunctival organoids, 25 µM Ly93 treatment significantly rescued the structural damage of organoids, increasing the proportion of intact organoids from 85.25% ± 1.71% to 96.25% ± 2.63% (P < 0.0001) (Figs. 5D, 5E). Furthermore, Ly93 partially reversed the reduction in cell proliferation, as evidenced by an increased percentage of Ki67⁺ cells compared with the DED model group (35.79% ± 6.89% vs. 19.27% ± 3.17%; P = 0.029) (Figs. 5F, 5G). There was no significant difference in the percentage of p63⁺ cells among the three groups (all P > 0.05) (Figs. 5F, 5H).

Lipidomic analysis revealed that there was no significant difference in total lipid content between the DED conjunctival organoids with and without Ly93 treatment (P = 0.4649) (Fig. 6A). However, Ly93 reduced the total SM level in the DED conjunctival organoids (P = 0.0434) (Fig. 6B). A total of 24 SM species were detected in the conjunctival organoids, the majority of which exhibited decreased abundance following Ly93 treatment (Fig. 6C). Specifically, Ly93 significantly reduced 16 SM species (all P < 0.05) (Fig. 6D). Notably, five of these downregulated species (d18:0/16:0, d18:0/20:0, d18:1/14:0, d18:1/15:0, and d18:1/16:0) corresponded with those identified as elevated in the tears of patients with DED in our preceding clinical analysis (Figs. 1D, 6D).

Figure 6.

Ly93 reduced the SM level in hyperosmolarity-induced DED conjunctival organoids. (A) Total lipid content and (B) proportion of SM in the hyperosmolarity-induced DED conjunctival organoids with and without Ly93 treatment. (C) Heatmap showing relative abundance of 24 SM species in conjunctival organoids. (D) Sixteen SM species showed significant differences between the two groups. Color intensity represents the z-score (red, positive; blue, negative). *P < 0.05; **P < 0.01; ns, not significant.

Immunofluorescence staining revealed that Ly93 reduced tissue-level expression of caspase-9, IL-1β, STAT1, and NF-κB p65 (Fig. 7A). Ly93 also effectively inhibited the accumulation of intracellular Fe²⁺ and 4-HNE in the DED conjunctival organoids (Figs. 7B, 7C). Western blotting analysis demonstrated that Ly93 lowered TFR levels, but did not restore GPX4 expression (Figs. 7D, 7E). Furthermore, Ly93 suppressed the activation of signaling pathways, as evidenced by reduced levels of phospho-AKT, NF-κB p65, and phospho-STAT1/STAT1 (Figs. 7F–H). Consistent with these findings, Ly93 also diminished the protein levels of caspase-9 and IL-1β in DED conjunctival organoids (Figs. 7F, 7H).

Figure 7.

Ly93 attenuated DED-associated inflammation in conjunctival organoids. (A) Immunofluorescence staining of caspase-9, ZO-1, IL-1β, STAT1, and NF-κB p65 between vehicle control, DED, and DED + Ly93 groups (scale bar, 50 µm). (B) Intracellular Fe²⁺ analysis (scale bar, 50 µm). (C) Immunofluorescence staining of 4-HNE (scale bar, 50 µm). (D–H) Western blotting analysis of GPX4 (D), TFR (E), AKT pathway (F), caspase-9 and NF-κB p65 pathway (G), and IL-1β and STAT1 pathway (H) protein expression.

Discussion

In this study, we demonstrated that SM levels were elevated in the tears of patients with DED and were positively correlated with multiple clinical indicators of disease severity. Using the human conjunctival organoids, we demonstrated that SM can directly promote inflammation and cellular injury. We further established a hyperosmolality-induced DED organoid model and demonstrated that Ly93 treatment effectively reduced SM accumulation, thereby mitigating conjunctival injury, inflammation, apoptosis, and ferroptosis.

Our finding of elevated tear SM in DED patients is consistent with previous studies.^6,7 Similar SM enrichment has been observed in the meibum of patients with meibomian gland dysfunction, where it was associated with impaired meibum quality and meibomian gland atrophy.^22,23 Galor et al.²² reported that tear SM level was negatively correlated with tear secretion. Paranjpe et al.²⁴ also noted an association between SM and poor meibum quality in meibomian gland dysfunction patients. Ham et al.²⁵ demonstrated that SM levels negatively correlated with TBUT and tear secretion while positively correlating with symptom scores and the CFS. These findings are highly consistent with our results, collectively underscoring the potential of SM as a biomarker for DED severity.

We successfully established human conjunctival organoids expressing K13, K19, p63, Ki67, MUC1, and MUC5AC, recapitulating both the structural and secretory features of human conjunctival tissue, which aligned with the previous study.⁹ Furthermore, we developed a hyperosmolarity-induced organoid model to mimic the elevated tear osmolarity characteristic of DED.^26–28 Hyperosmotic stress damaged the conjunctival organoids, which aligned with the previous study using conjunctival epithelial cells.²⁹ The conjunctival organoids treated with hyperosmotic medium highly expressed DED-associated inflammation (IL-1β, MMP-9, and NF-κB), consistent with hyperosmolality-induced corneal epithelial organoids.¹⁰ Our DED model showed activated STAT1, NF-κB, and PI3K/AKT pathways, consistent with previously reported mechanisms.^30–32 Additionally, hyperosmotic stress reduced proliferation (Ki67), induced apoptosis (caspase-3 and caspase-9), and ferroptosis (TFR and GPX4), as reported in previous studies on the mechanism of DED.^33–36

Exogenous SM exacerbated inflammatory and injury responses in conjunctival organoids. SM homeostasis is critical for membrane integrity and signal transduction, and excessive SM can act as a proinflammatory mediator.^37–39 In our study, although exogenous SM generally induced inflammation and reduced the viability of conjunctival organoids, a paradoxical downregulation of iNOS protein and IL-1β mRNA, and MUC5AC expression was observed at the highest concentration (160 µM). The nonlinear response may be attributed to profound cytotoxicity at this concentration, leading to a collapse of cellular metabolic and transcriptional activity within the organoids. Mechanistically, the accumulation of SM can disrupt the turnover of endogenous SM, promoting their catabolism into bioactive metabolites such as ceramide, sphingosine-1-phosphate, and ceramide-1-phosphate. These metabolites act as signaling molecules that induce the expression of inflammatory mediators, including CAAT/enhancer binding proteins, cytosolic phospholipase A2, cyclo-oxygenase-2, and NF-κB.^3,40,41 Crucially, excess SM levels may also initiate signaling cascades by modulating lipid raft organization to trigger TLR4/NF-κB signaling pathway, release proinflammatory cytokines (such as IL-1β, TNF-α, and IL-6).^42,43 These cytokines may activate the JAK/STAT1 pathway, leading to an amplified inflammatory cascade.^44,45 Furthermore, the SM-mediated metabolic flux has been shown to drive iNOS overexpression and the subsequent release of nitric oxide, further exacerbating tissue damage and programmed cell death.^46,47

In the present study, we demonstrated that Ly93, a selective inhibitor of SMS2, significantly reduced SM levels and mitigated hyperosmolarity-induced inflammation and apoptosis in conjunctival organoids. This observation is consistent with the findings of Yang et al.,⁸ who reported that the SMS2 inhibitor 14I suppressed SM synthesis and alleviated clinical signs and inflammation in DED mouse model. The protective efficacy of Ly93 in our model is further supported by its established anti-inflammatory and regulatory roles in other systems. Li et al.¹⁹ reported that Ly93 suppresses SMS activity, reduces lipopolysaccharide-induced IL-6 production, and increases IκB (an inhibitor of NF-κB) and monocyte chemoattractant protein-1 in macrophages. Ly93 has been shown to ameliorate diet-induced insulin resistance through modulation of the IRS-1/Akt/GSK-3β signaling pathway⁴⁸ and to decrease SM levels and downregulate ceramide in HEK293 cells.⁴⁹ Ceramide plays a key role in meibomian gland dysfunction, and inhibition of ceramide de novo synthesis alleviates inflammation, apoptosis, and keratinization in vivo.⁵⁰ Collectively, these findings reinforce our conclusion that targeting SM metabolism with Ly93 represents a promising strategy for alleviating DED-related injury and inflammation.

Notably, our study reveals that Ly93 inhibits ferroptosis in the DED organoid model with pathway specificity. Ly93 significantly reduced TFR expression and the accumulation of intracellular Fe²⁺ and 4-HNE, but failed to reverse the downregulation of GPX4. These findings suggest that Ly93 may block the ferroptosis process primarily by modulating iron metabolic homeostasis rather than directly augmenting the GPX4-mediated antioxidant defense system. Ferrostatin-1, a ferroptosis inhibitor, and glutamate have been shown to lower SM levels in oligodendrocytes.⁵¹ Additionally, acid sphingomyelinase, a key enzyme in SM metabolism, has been implicated in ferroptosis in cancer, diabetic osteoporosis, and lipopolysaccharide-induced inflammation.^52–54 Together, these findings suggest that ferroptosis is an integral component of SM-mediated pathology in DED.

Although our study provides novel insights into the role of SM metabolism in DED, there were several limitations. First, the clinical tear SM analysis involved a relatively small sample size. Second, we used a mixture of SM species for exogenous intervention. Therefore, the specific biological effects of individual SM subtypes with varying carbon chain lengths require further investigation. Third, the high concentrations of exogenous SM used in our experiments primarily simulate acute injury, which may differ from the mechanisms underlying chronic, trace accumulation in vivo. Fourth, because the ocular surface is an integrated unit, our findings in the conjunctiva warrant further exploration in corneal organoid models to provide a more comprehensive perspective on how SM metabolism affects ocular surface homeostasis. Finally, our current findings were based on in vitro organoid models. Thus, the in vivo safety and efficacy of SM modulation remain to be validated in animal models and clinical trials.

Conclusions

Our study demonstrates that elevated SM promotes ocular surface inflammation and cellular injury. Inhibition of SM synthase by Ly93 reduces SM levels and attenuates inflammation, apoptosis, and ferroptosis in the DED organoid model. These findings reveal a novel pathogenic role for SM in DED, suggesting SM synthase as a potential therapeutic target for DED-associated inflammation and injury.