Lipoic Acid Modulates Nrf2/HO-1 Pathway to Suppress Inflammation in Fungal Keratitis Through Host-Directed Cytoprotective Mechanisms

Fungal keratitis (FK) is now recognized as a pervasive ophthalmic emergency, responsible for more than 1 million culture-proven cases annually and a disproportionate share of corneal blindness in tropical and subtropical regions^1,2. Incidence estimates approaching 6-8 per 100,000 in the United States and even higher in agrarian economies underscore a mounting global threat that preferentially strikes working-age populations after vegetal trauma or contact-lens wear^3,4. Filamentous fungi, most notably species of Fusarium and Aspergillus, dominate the etiological landscape, thriving in warm, humid climates and exploiting ocular micro-injuries to initiate infection⁵.

Despite half a century of clinical use, natamycin 5% remains the only topical agent approved specifically for FK; however, its polyene chemistry confers poor corneal penetration, requires frequent dosing, and is not universally available^6,7. Alarmingly, a successive surveillance study from South Asia has revealed incremental increases in the minimal inhibitory concentrations of both natamycin and triazoles, with Fusarium isolates exhibiting a ≥6% yearly increase in natamycin resistance and a parallel trend for voriconazole⁸. These data align with clinical reports of therapeutic failure and reinforce forecasts that the market for FK therapeutics will expand from USD 0.9 billion in 2023 to USD 1.48 billion by 2033, driven largely by the inadequacies of current pharmacotherapy. Ad-hoc surgical measures, such as conjunctival flaps or penetrating keratoplasty, salvage anatomy in recalcitrant ulcers, but are resource-intensive and vision-limiting⁹. Collectively, these shortcomings underscore the need for host-directed or multitarget strategies that extend beyond fungistatic action alone.

Oxidative stress and the resulting inflammatory cascade are now understood to be pivotal determinants of fungal pathogenesis and corneal scarring. Upon fungal invasion, corneal epithelial cells and resident immune populations, including conjunctival mast cells, limbal dendritic cells, and stromal keratocytes, initiate innate immune responses through pattern recognition receptors, such as Dectin-1 and Toll-like receptors¹⁰. Neutrophil-derived reactive oxygen species (ROS) accumulate in infected stroma as part of the antimicrobial oxidative burst, while infiltrating macrophages and T lymphocytes perpetuate chronic inflammation through cytokine amplification¹¹. Paradoxically, filamentous fungi upregulate their own antioxidant circuits (e.g., Yap1-regulated superoxide dismutases) to survive this hostile milieu. Excess ROS in turn activates NF-κB-dependent transcription of IL-1β, TNF-α, and other pro-inflammatory mediators that drive tissue destruction and corneal opacification¹². At the center of endogenous cytoprotection lies nuclear factor erythroid 2-related factor 2 (Nrf2), whose translocation to the nucleus orchestrates expression of heme oxygenase-1 (HO-1) and an array of detoxifying enzymes across multiple corneal cell types, including epithelial cells, keratocytes, and endothelial cells^13,14. Aging, diabetes, and chronic inflammation downregulate this axis, rendering ocular tissues vulnerable to oxidative insults, whereas genetic or pharmacological enhancement of Nrf2 expedites epithelial wound closure and suppresses cytokine surge in diverse corneal injuries¹⁵.

Proof-of-concept studies have begun to exploit this pathway in FK. Natural small molecules such as perillaldehyde¹⁶, gallic acid¹⁷, hydroxytyrosol¹⁸ , and the organoselenium compound ebselen¹⁹ achieve dual antifungal and anti-inflammatory effects by potentiating Nrf2-HO-1 signaling while directly impairing fungal viability. Nanomedicine platforms that co-deliver antifungals and antioxidants further illustrate the therapeutic promise of combining pathogen clearance with host defense reinforcement^20,21. Yet many of these candidates face translational hurdles related to formulation complexity or unproven ocular safety.

Lipoic acid (LA) is a ubiquitous dithiol antioxidant with a well-established systemic safety record and exceptional pharmacodynamic versatility. Based on studies in hepatic and neural tissues^10,12 LA is proposed to directly modify cysteine residues (particularly Cys151, Cys273, and Cys288) on Kelch-like ECH-associated protein-1 (Keap1), disrupting its interaction with Nrf2 and thereby preventing Nrf2 ubiquitination and proteasomal degradation, which results in increased Nrf2 protein levels, enhanced nuclear translocation, and augmented HO-1 transcription. However, direct demonstration of Keap1-LA interaction in corneal epithelial cells warrants future investigation. LA has demonstrated cytoprotective benefits in hepatic, neural, and ocular models of oxidative injury^13,15. Its amphipathic nature facilitates tissue penetration without the membrane toxicity observed with polyenes, and clinical formulations already exist for topical and systemic indications. Nevertheless, LA's capacity to modulate the corneal inflammatory micro-environment and thereby ameliorate FK has not been rigorously interrogated.

The present study, therefore, investigated whether LA could mitigate A. fumigatus-induced keratopathy by activating the Nrf2/HO-1 axis while suppressing pro-inflammatory cytokines. Using human corneal epithelial cells and a validated murine FK model, we delineated the dose-response characteristics, dissected mechanistic dependencies with selective inhibitors, and benchmarked therapeutic outcomes against uninfected controls and infected vehicle-treated groups. To our knowledge, this represents the first systematic investigation to establish lipoic acid's therapeutic efficacy in fungal keratitis through comprehensive dose-response characterization, mechanistic dissection using selective pharmacological inhibitors of both Nrf2 and HO-1, and validation in both cellular and animal models. This work uniquely integrates molecular, cellular, and translational readouts to provide definitive evidence of the Nrf2/HO-1 pathway dependency underlying LA's anti-inflammatory effects. By integrating molecular, cellular, and in vivo readouts, we sought to provide a comprehensive appraisal of LA as a readily translatable, host-directed therapy that addresses the twin imperatives of pathogen eradication and inflammatory restraint in FK.

All animal procedures were approved by the Institutional Animal Care and Use Committee (Protocol number: IACUC-2024-FK-007) and conducted in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. Male C57BL/6 mice, aged 8-10 weeks and weighing 20-25 g, were housed in individually ventilated cages under controlled environmental conditions of 22 ± 2 °C and 50%-60% relative humidity with a 12:12 h light-dark cycle. These specific pathogen-free animals were obtained from Charles River Laboratories (see Table of Materials). All mice underwent a 1 week acclimatization period to adapt to the new environment before experimental procedures, with unrestricted access to standard rodent chow and water.

CAUTION: Aspergillus fumigatus is a pathogenic organism requiring biosafety level 2 containment. All fungal cultures must be handled in a certified biological safety cabinet using appropriate personal protective equipment, including gloves, laboratory coats, and eye protection. Dimethyl sulfoxide (DMSO) is flammable and can penetrate skin; handle in well-ventilated areas and avoid direct contact. Ketamine and xylazine are controlled substances requiring secure storage and proper documentation per institutional regulations.

Experimental preparation
All required reagents, drugs, and materials were prepared according to the experimental workflow (see Table of Materials and Supplementary Figure 1 for schematic). Racemic (±)-α-Lipoic acid (LA) stock solutions were prepared using racemic α-lipoic acid (1,2-dithiolane-3-pentanoic acid) dissolved in DMSO to a concentration of 60 mM and stored in single-use aliquots at -80 °C to prevent oxidative degradation. The Nrf2 inhibitor ML385 and HO-1 inhibitor SnPPIX were similarly prepared and aliquoted for single-use to avoid freeze-thaw cycles. All cell culture media and supplements were prepared under sterile conditions and pre-warmed to 37 °C before use.

All materials contaminated with fungal cultures, including agar plates, culture tubes, and disposable plastics, were autoclaved at 121 °C for 30 min before disposal in biohazard waste containers. Chemical waste containing organic solvents was collected separately and disposed of through institutional environmental health and safety protocols. Animal tissues and carcasses were incinerated through approved biomedical waste services.

Fungal culture and infection preparation
Aspergillus fumigatus was cultured on Sabouraud dextrose agar at 37 °C for 5-7 days until mature conidial formation was observed. Conidia were harvested by flooding mature cultures with sterile phosphate-buffered saline (PBS) containing 0.05% Tween-80, and conidial suspensions were filtered through sterile gauze to remove hyphal fragments. Conidial concentration was determined using a hemocytometer and adjusted to established concentrations based on published fungal keratitis models. For in vitro experiments, 1 x 10⁶ conidia/mL (multiplicity of infection of 10:1) was selected based on previous studies demonstrating that this concentration induces robust inflammatory responses in corneal epithelial cells while maintaining adequate cell viability for downstream analyses^16,17. For in vivo studies, 1 x 10⁷ conidia/mL was chosen as this inoculum produces consistent, moderate-severity keratitis with reproducible disease scores by day 5 post-infection, providing sufficient dynamic range to assess therapeutic efficacy without excessive mortality^16,19.

Animal treatment and corneal infection
Under general anesthesia (intraperitoneal injection of ketamine 100 mg/kg and xylazine 10 mg/kg), fungal keratitis was induced by creating three 1 mm linear scratches on the central cornea using a sterile 25G needle, followed by topical application of 5 µL of A. fumigatus conidial suspension. Mice were randomly assigned to treatment groups (n=8 per group): control (uninfected), infected control, infected + LA treatment (60 µmol/L topically 4x times daily), and infected + LA + SnPPIX treatment (LA plus 5 mg/kg SnPPIX intraperitoneally daily). Clinical assessment and tissue collection were performed on day 5 post-infection.

Animal handling precautions: All procedures involving anesthetized animals should be performed under aseptic conditions. Monitor animals continuously during anesthesia for respiratory depression. Dispose of contaminated bedding, surgical waste, and infected tissues by autoclaving followed by incineration according to institutional biohazard waste management protocols.

Cell culture and treatment
Human corneal epithelial cells (HCECs) were obtained from ATCC and maintained in DMEM/F-12 medium supplemented with 10% fetal bovine serum, 1% penicillin-streptomycin, and 1% L-glutamine at equimolar concentrations at 37 °C in a humidified 5% CO₂ atmosphere. Cells were subcultured every 3-4 days upon reaching 80%-90% confluence and used between passages 3-8 to ensure consistency. For experimental treatments, cells were seeded at 5 x 10⁴ cells per well in appropriate plates and allowed to adhere for 24 h before treatment. For all in vitro mechanistic experiments, cells were first exposed to Aspergillus fumigatus conidia at a multiplicity of infection (MOI) of 10:1 for 4 h to establish infection, followed by the addition of LA (60 µM), ML385 (5 µM), or SnPPIX (1 µM) alone or in combination for an additional 24 h before sample collection. This sequential design allows assessment of LA's therapeutic effects on established fungal challenge rather than prophylactic treatment.

Cell viability assessment (CCK-8 analysis)
Cell viability was assessed using the Cell Counting Kit-8 (CCK-8). Cells were seeded in 96-well plates at 5 x 10⁴ cells per well and incubated for 24 h before treatment. Following drug exposure, 10 µL of CCK-8 solution was added to each well, and plates were incubated for 2 h at 37 °C. Absorbance was measured at 450 nm using a Multi-Mode Microplate Reader. Each experimental condition was performed in octuplicate, and experiments were repeated three times independently. Vehicle control (DMSO) concentration was maintained below 0.1% to avoid cytotoxic effects.

RNA extraction and quantitative real-time PCR
Total RNA was isolated from cultured cells and corneal tissues using the RNeasy Mini Kit according to the manufacturer's protocol. Tissue samples were homogenized in RLT buffer using a tissue homogenizer for 2 min at 50 Hz. RNA quality and concentration were assessed using a spectrophotometer, with samples having 260/280 ratios of 1.8-2.0 used for downstream analysis. First-strand cDNA synthesis was performed using the High-Capacity cDNA Reverse Transcription Kit with 1 µg total RNA per reaction. Quantitative PCR was conducted using SYBR Green-based qPCR master mix on a real-time PCR system with initial denaturation at 95 °C for 2 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Melting curve analysis was performed to verify amplicon specificity. All primers were validated for efficiency (90%-110%) and specificity prior to use. Relative mRNA expression was calculated using the 2^(-ΔΔCT) method with GAPDH as the reference gene.

Protein extraction and Western blot analysis
For total protein extraction, cells and tissues were lysed in RIPA buffer supplemented with protease and phosphatase inhibitor cocktails. Nuclear and cytoplasmic protein fractions were prepared using the NE-PER Nuclear and Cytoplasmic Extraction Reagents according to the manufacturer's protocol. Protein concentrations were determined using the Bradford assay. Equal amounts of protein (30-50 µg) were separated by SDS-PAGE and transferred to PVDF membranes. After blocking with 5% non-fat milk in Tris-buffered saline containing 0.1% Tween-20 for 1 h, membranes were incubated with primary antibodies overnight at 4 °C, followed by appropriate HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using enhanced chemiluminescence reagent with exposure times of 30-180 s depending on signal intensity. Bands were quantified using ImageJ software, with β-actin serving as the loading control for total proteins and lamin B1 for nuclear proteins.

Clinical assessment and histological analysis
Corneal disease severity was evaluated using a standardized clinical scoring system based on established protocols for murine fungal keratitis²². The scoring system assesses three parameters on a 0-4 scale: corneal opacity (0 = clear, 4 = completely opaque), surface irregularity (0 = smooth, 4 = severe ulceration with perforation), and inflammatory response (0 = no infiltrate, 4 = dense infiltration with hypopyon), yielding total scores from 0 to 12. All clinical assessments were performed by two independent observers who were masked to treatment group assignments, with inter-observer agreement exceeding 90%. Corneal photographs were captured using a slit-lamp biomicroscope equipped with a digital camera. For histological analysis, eyes were enucleated, fixed in 4% paraformaldehyde, paraffin-embedded, and sectioned at 5 µm thickness. Sections were stained with hematoxylin and eosin for general morphology assessment.

Enzyme-linked immunosorbent assay (ELISA)
Secreted levels of IL-1β and TNF-α in cell culture supernatants and corneal tissue homogenates were quantified using commercial ELISA kits according to the manufacturer's instructions. Cell culture supernatants were collected after 24 h of treatment, while corneal tissues were homogenized in PBS containing protease inhibitors. Samples were centrifuged at 12,000 x g for 10 min at 4 °C, and supernatants were used for ELISA analysis. All samples were analyzed in duplicate, and cytokine concentrations were calculated from standard curves.

Statistical analysis
All experiments were performed with appropriate biological replicates (n ≥6 for in vitro studies, n =8 for animal studies) and repeated at least 3x independently. Data are presented as mean ± standard error of the mean (SEM). Data analysis was performed using GraphPad Prism version 9.0. Specifically, we used the normality test function (Shapiro-Wilk) to assess data distribution, one-way ANOVA with Tukey's multiple comparisons test for experiments with more than two groups, and unpaired Student's t-test for two-group comparisons. Graphs were generated as grouped bar charts with individual data points overlaid as scatter plots, with error bars representing the standard error of the mean.

The experimental workflow demonstrated that lipoic acid treatment effectively modulates the Nrf2/HO-1 signaling pathway to suppress inflammation in fungal keratitis models.

Cytotoxicity assessment of Lipoic acid and pathway-specific inhibitors
Initial dose-response studies were conducted to establish optimal concentrations for each compound used in this investigation. LA demonstrated excellent biocompatibility with HCECs, maintaining cell viability at concentrations up to 60 µM after 24 h exposure, establishing this as the optimal LA concentration for subsequent mechanistic studies (Figure 1A). In contrast, the Nrf2 inhibitor ML385 exhibited a more restricted safety profile, with preliminary dose-response experiments determining 5 µM as the optimal concentration. The HO-1 inhibitor SnPPIX displayed the narrowest therapeutic window among the tested compounds, with SnPPIX at 2.5 µM and 5 µM significantly reducing cell viability compared to the control group, establishing 1 µM SnPPIX as the optimal concentration for mechanistic studies (Figure 1C).

LA activates Nrf2/HO-1 signaling in fungal-challenged corneal epithelial cells
To evaluate LA's effects on the Nrf2/HO-1 pathway, HCECs were challenged with A. fumigatus and treated with LA. LA treatment increased both Nrf2 and HO-1 mRNA expression in HCECs exposed to Aspergillus fumigatus, demonstrating activation of this cytoprotective signaling cascade at the transcriptional level (Figure 2A). The LA-mediated upregulation of both Nrf2 and HO-1 mRNA was substantially attenuated when cells were co-treated with the Nrf2 inhibitor ML385, confirming the specificity and hierarchical relationship of the Nrf2/HO-1 signaling pathway. Western blot analysis revealed that LA treatment enhanced both total and nuclear Nrf2 protein levels, with enhanced nuclear translocation indicating activation of this master regulator of antioxidant responses (Figure 2B, left panels). Quantification of Western blot bands confirmed significant increases in Nrf2 protein in both cytoplasmic (total) and nuclear fractions following LA treatment (Figure 2B, upper right graphs). The downstream effects of Nrf2 activation were clearly evident in HO-1 protein expression patterns, with LA treatment dramatically enhancing HO-1 protein expression as demonstrated by Western blot analysis and quantification (Figure 2B, lower right graph).

HO-1-dependent anti-inflammatory effects of LA in corneal epithelial cells
Having established LA's ability to activate the Nrf2/HO-1 pathway, the investigation next examined whether this activation translates into meaningful anti-inflammatory effects. A. fumigatus challenge induced robust inflammatory responses in HCECs, as evidenced by significant upregulation of key pro-inflammatory cytokines IL-1β and TNF-α at the mRNA level. LA treatment markedly attenuated these inflammatory responses, reducing both IL-1β and TNF-α mRNA expression (Figure 3A). The mechanistic dependence of these anti-inflammatory effects on HO-1 activity was established through parallel experiments using the specific HO-1 inhibitor SnPPIX. Co-treatment with SnPPIX substantially reversed LA's anti-inflammatory effects at the transcriptional level, with cytokine mRNA expression levels returning to near those observed in infected controls. These transcriptional changes were reflected at the protein level, with ELISA quantification demonstrating that LA treatment significantly reduced secreted IL-1β and TNF-α protein levels in cell culture supernatants following A. fumigatus challenge (Figure 3B). Again, co-administration of SnPPIX abolished these protective effects, with cytokine protein concentrations reverting to levels comparable to infected controls. This provides compelling evidence that HO-1 enzymatic activity is essential for mediating LA's therapeutic benefits.

Therapeutic efficacy of LA in the murine fungal keratitis model
Translation of the in vitro findings to a physiologically relevant disease model was accomplished using a well-established murine fungal keratitis model. Clinical assessment revealed that LA treatment (60 µmol/L applied topically 4x daily) significantly reduced corneal clinical scores compared to untreated infected controls, with visible improvements in corneal opacity and inflammatory signs. Representative corneal photographs on day 5 post-infection demonstrate marked corneal opacity and edema in infected control eyes, substantial improvement with LA treatment, and loss of protection when LA is co-administered with SnPPIX (Figure 4A). Quantitative clinical scoring confirmed these observations, with LA treatment producing approximately 47% reduction in disease scores compared to infected controls, and SnPPIX co-administration negating this therapeutic benefit (Figure 4B).

Molecular analysis of corneal tissues confirmed that the therapeutic effects observed clinically were associated with activation of the same Nrf2/HO-1 pathway identified in the in vitro studies. LA treatment significantly increased corneal HO-1 and Nrf2 mRNA expression in infected eyes (Figure 4C). Western blot analysis of corneal tissue lysates corroborated these findings at the protein level, showing enhanced HO-1 and Nrf2 protein expression in LA-treated corneas (Figure 4D, left panels). Quantification of the Western blot data confirmed significant increases in HO-1, total Nrf2, and nuclear Nrf2 protein levels following LA treatment, with these effects abolished by SnPPIX co-administration (Figure 4D, right graphs). Immunohistochemical analysis revealed enhanced nuclear localization of Nrf2 in corneal epithelial cells from LA-treated animals, further supporting pathway activation in vivo. Consistent with the in vitro findings, LA treatment substantially reduced corneal expression of inflammatory cytokines IL-1β and TNF-α at both mRNA and protein levels. qPCR analysis demonstrated a significant reduction in IL-1β and TNF-α mRNA transcripts in corneal tissues from LA-treated mice compared to infected controls (Figure 4E). ELISA quantification of corneal tissue homogenates confirmed these anti-inflammatory effects at the protein level, with marked reductions in both IL-1β and TNF-α protein concentrations in LA-treated corneas (Figure 4F). The critical role of HO-1 in mediating LA's therapeutic effects was definitively demonstrated through co-administration studies with SnPPIX. Mice receiving both LA and SnPPIX showed a marked reversal of therapeutic benefits, with clinical scores, Nrf2/HO-1 expression levels, and inflammatory cytokine concentrations returning to levels statistically indistinguishable from those of infected controls.

Collectively, these findings establish a coherent mechanistic pathway wherein lipoic acid activates the Nrf2/HO-1 cytoprotective axis at both cellular and tissue levels, resulting in substantial suppression of pro-inflammatory cytokine production and clinical improvement in fungal keratitis. The selective reversal of these effects by pathway-specific inhibitors confirms that the therapeutic benefits are strictly dependent on Nrf2 activation and downstream HO-1 enzymatic activity, validating the proposed mechanistic hierarchy and supporting the potential of LA as a host-directed therapy for fungal keratitis.

Data Availability:
Due to institutional policies at Jinan 2nd People's Hospital, we are unable to deposit primary research data in public repositories. Our institution requires that data from animal studies remain under institutional custody for regulatory and ethical oversight. However, all data supporting this study are available from the corresponding author (sarah880521@yeah.net) upon reasonable request. Requests will be processed within 2-4 weeks following institutional approval procedures. The raw Western blots are provided in Supplementary File 1.

Figure 1: Cytotoxicity assessment of lipoic acid and pathway inhibitors in human corneal epithelial cells. Cell viability was assessed using the CCK-8 assay after a 24 h treatment. (A) LA (12-60 µM) maintained >94% cell viability at all concentrations; 60 µM was selected for subsequent experiments. (B) ML385 (Nrf2 inhibitor, 1-20 µM) showed significant toxicity at ≥10 µM; 5 µM was selected. (C) SnPPIX (HO-1 inhibitor, 0.2-5 µM) showed significant toxicity at ≥2.5 µM; 1 µM was selected. Data are mean ± SEM (n=8). One-way ANOVA with Tukey's post-hoc test. ***P < 0.001 versus control. Please click here to view a larger version of this figure.

Figure 2: LA activates Nrf2/HO-1 signaling in A. fumigatus-challenged HCECs. Cells were challenged with A. fumigatus (MOI 10:1, 4 h), then treated with LA (60 µM) ± ML385 (5 µM) for 24 h. (A) qRT-PCR analysis of Nrf2 (left) and HO-1 (right) mRNA expression normalized to GAPDH. LA increased both Nrf2 and HO-1 mRNA; ML385 blocked these increases. (B) Western blot analysis (left panels) and quantification (right panels) of total Nrf2, nuclear Nrf2, and HO-1 protein. β-actin and lamin B1 served as loading controls. LA enhanced Nrf2 nuclear translocation and HO-1 expression; ML385 abolished these effects. Data are mean ± SEM (n=6). One-way ANOVA with Tukey's post-hoc test. *p < 0.05, **p < 0.01, ***p < 0.001 versus control; #p < 0.05, ##p < 0.01, ###p < 0.001 versus AF; Δp < 0.05, ΔΔp < 0.01, ΔΔΔp < 0.001 versus AF+LA. Please click here to view a larger version of this figure.

Figure 3: LA suppresses inflammatory cytokines through HO-1-dependent mechanisms. Cells were challenged with A. fumigatus (MOI 10:1, 4 h), then treated with LA (60 µM) ± SnPPIX (1 µM) for 24 h. (A) qRT-PCR analysis of IL-1β (left) and TNF-α (right) mRNA normalized to GAPDH. LA reduced cytokine mRNA; SnPPIX reversed this effect. (B) ELISA quantification of secreted IL-1β (left) and TNF-α (right) proteins. LA decreased cytokine secretion; SnPPIX abolished this suppression. Data are mean ± SEM (n=6). One-way ANOVA with Tukey's post-hoc test. *p < 0.05, **p < 0.01, ***p < 0.001 versus control; #p < 0.05, ##p < 0.01, ###p < 0.001 versus AF; Δp < 0.05, ΔΔ p< 0.01, ΔΔΔp < 0.001 versus AF+LA. Please click here to view a larger version of this figure.

Figure 4: Topical LA reduces disease severity in murine A. fumigatus keratitis by HO-1-dependent mechanisms. Mice received corneal scarification and A. fumigatus inoculation, followed by topical LA (60 µmol/L, 4x/day) ± intraperitoneal SnPPIX (5 mg/kg/day) for 5 days. (A) Representative corneal photographs at day 5. LA treatment reduced opacity and inflammation; SnPPIX co-treatment negated protection. (B) Clinical disease scores (0-12 scale). LA reduced scores by ~47%; SnPPIX reversed this benefit. (C) qRT-PCR of corneal HO-1 (left) and Nrf2 (right) mRNA normalized to GAPDH. LA increased expression; SnPPIX reduced it. (D) Western blot (left) and quantification (right) of corneal HO-1, total Nrf2, and nuclear Nrf2. β-actin and lamin B1 served as loading controls. Note: LA+ZnPP label refers to LA+SnPPIX treatment. LA increased protein expression; SnPPIX reversed these effects. (E) qRT-PCR of corneal IL-1β (left) and TNF-α (right) mRNA. LA reduced cytokine expression; SnPPIX abolished suppression. (F) ELISA of corneal IL-1β (left) and TNF-α (right) protein. LA decreased cytokine levels; SnPPIX reversed this effect. Data are mean ± SEM (n=8). Observers were masked to treatment groups. One-way ANOVA with Tukey's post-hoc test. *p < 0.05, **p < 0.01, ***p < 0.001 versus control;#p < 0.05, ##p < 0.01, ###p < 0.001 versus FK; Δp < 0.05, ΔΔp < 0.01, ΔΔΔp < 0.001 versus LA. Please click here to view a larger version of this figure.

Supplementary Figure 1: Workflow schematic. Please click here to download this File.

Supplementary File 1: Raw Western blots. Please click here to download this File.

The present findings confirm lipoic acid (LA) as a potent inducer of the Nrf2/HO-1 axis that mitigates fungal keratitis (FK) by curbing pro-inflammatory cytokine expression and preserving corneal architecture. Consistent with recent work showing robust Nrf2 activation by dithiolane compounds in oxidative disorders¹⁵, LA produced a five-fold rise in HO-1 transcripts and halved IL-1β/TNF-α secretion, outperforming several phytochemical activators reported for ocular inflammation^2,23. Recent studies with perillaldehyde¹⁶, gallic acid¹⁷, hydroxytyrosol¹⁵, and ebselen¹⁹ have demonstrated similar Nrf2/HO-1 activation in FK models, yet LA distinguishes itself through established clinical safety profiles, existing ophthalmic formulations, and superior pharmaceutical characteristics that facilitate immediate translation. The tight pharmacologic hierarchy demonstrated that LA-mediated increased Nrf2 protein levels and enhanced nuclear translocation, downstream HO-1 induction, and subsequent cytokine repression -- highlights an endogenous cytoprotective circuit that can be therapeutically leveraged instead of conventional fungistatic regimens that often fail against recalcitrant moulds⁸.

Crucially, the murine model validated the cellular data: topical LA reduced clinical scores by 47%, mirrored by heightened corneal Nrf2 nuclear localization and HO-1 immunoreactivity. The reversal of these benefits by SnPPIX definitively positions HO-1 catalytic activity as the indispensable effector of LA's protection. This mechanistic clarity aligns with broader observations that HO-1-derived carbon monoxide and biliverdin dampen NF-κB signaling and foster tissue repair in infectious corneal disease²⁴. The systematic use of ML385 to demonstrate upstream Nrf2 requirement and SnPPIX to prove downstream HO-1 dependence establishes the complete pathway hierarchy, with SnPPIX selected for in vivo validation based on its favorable systemic pharmacokinetics.

Several practical advantages emerge. First, LA's maximal non-cytotoxic dose (60 µM) is well below systemic safety thresholds documented in metabolic trials, providing a favorable therapeutic margin. Second, targeting host defenses circumvents the escalating problem of azole and echinocandin resistance reported in contemporary FK cohorts²⁵. Third, LA's dual action-antioxidant priming and immunomodulation offers a single-molecule alternative to multidrug antifungal-anti-inflammatory cocktails that pose additive toxicity risks²⁶.

That said, important caveats temper immediate clinical extrapolation. The in vitro work uses only an HCEC line; infiltrating immune cells central to keratitis pathogenesis are not modeled, and while the murine model incorporates full immune complexity, co-culture systems or ex vivo human corneal models would provide more complete mechanistic insights. The study examined only a single A. fumigatus strain; virulence factors and drug sensitivity vary widely among Fusarium, Curvularia, or emerging Candida auris keratitides, and future studies must evaluate LA against diverse fungal species to establish therapeutic breadth^23,25. Moreover, the acute dosing paradigm does not address long-term safety, potential tachyphylaxis from chronic Nrf2 stimulation, or the oncogenic concerns raised by sustained pathway activation in other tissues^15,27. Although SnPPIX elegantly confirmed HO-1 dependence, pharmacological inhibitors can exert off-target heme sequestration; CRISPR-mediated HO-1 knockdown in corneal epithelium would provide orthogonal validation. In vivo pharmacokinetics were inferred from clinical response rather than direct corneal drug concentration, and the contribution of stromal penetration remains undefined. While our data demonstrate increased Nrf2 protein levels and nuclear translocation, whether LA prevents Keap1-mediated degradation or enhances Nrf2 synthesis requires protein turnover assays or direct Keap1-Nrf2 binding studies, though LA's dithiol chemistry strongly suggests Keap1 cysteine modification as the likely mechanism. Finally, murine corneal thickness and immune milieu differ from the human eye, potentially influencing LA bioavailability and Nrf2 responsiveness.

Future investigations should therefore pursue three parallel trajectories. At the mechanistic level, single-cell RNA-seq of LA-treated corneas could map cell-type-specific transcriptomic shifts, clarifying whether keratocytes, infiltrating neutrophils, or resident dendritic cells initiate the protective cascade. Advanced imaging mass spectrometry may simultaneously track LA distribution and heme catabolite gradients, refining dosing schemes. Second, translational studies must include diverse filamentous and yeast pathogens, coupled with ex vivo human corneal buttons or organoids, to confirm spectrum-wide efficacy and rule out pathogen-specific escape pathways documented for natamycin and voriconazole^2,23. Third, formulation science should exploit nanocarrier or in-situ gelling systems to prolong precorneal residence, an approach that has already enhanced antifungal pharmacodynamics in preclinical ocular models. Such vehicles may permit LA dose-sparing while minimizing systemic exposure.

A broader strategic implication arises: adjunctive host-directed therapy could synergize with forthcoming systemic antifungals now in late-phase trials, such as olorofim and fosmanogepix, whose spectra do not fully cover resistant moulds²⁶. By dampening host-driven tissue damage, LA might lower the fungal burden threshold that new agents must eradicate, ultimately shortening treatment courses. Equally, precision-medicine efforts that integrate tear-film proteomics and genetic susceptibility markers may identify FK subpopulations with blunted Nrf2 signaling who stand to benefit most from LA-based regimens^28,29. Such biomarker-guided positioning echoes recent personalized frameworks proposed for glaucoma and uveitis management.

In conclusion, the study positions lipoic acid as a safe, mechanistically defined activator of the Nrf2/HO-1 pathway that confers substantial anti-inflammatory and clinical benefit in experimental fungal keratitis. By addressing the limitations noted -- pathogen diversity, long-term safety, and optimized ocular delivery -- and by pursuing the outlined future directions, LA-centred host-directed therapy could augment, and in selected cases replace, conventional antifungal strategies, meeting the urgent need for innovative interventions in sight-threatening corneal infections.