Sulfur Vacancy‐Rich MoS₂ Nanodot‐Based Hybrid Eye Drops as an Antibiotic‐Free Antifungus for Fungal Keratitis Therapy

ABSTRACT

Fungal keratitis is a highly blinding ocular infection that demands innovative treatment strategies, moving beyond conventional antifungal therapies. In this study, we explore the potential of MoS₂ nanodots engineered with sulfur vacancies, small sizes, hydrophilicity and surface positive charge as eye drops for effective fungal keratitis management. The tailored MoS₂ nanodot‐based hybrid nanomaterials (HNAF) exhibit remarkable antifungal efficacy through a synergistic interaction of their abundant sulfur vacancies and positive surface charges. The ultrasmall size and positive surface charge of HNAF allow the nanodots to pass through corneal epithelial barriers and electrostatically bind to the negatively charged cell membranes of Fusarium solani. The rich sulfur vacancies act as active catalytic centers that decompose local H₂O₂ into toxic hydroxyl radicals (•OH) and deplete intracellular GSH under dark physiological conditions, without requiring external stimuli in fungal cells. This cooperation between barrier penetration and targeted catalytic oxidation leads to significant antifungal activity in vitro and in vivo with excellent biosafety. This research underscores the potential of MoS₂ nanodot‐based hybrid nanomaterials as a step forward in managing fungal keratitis, offering an antibiotic‐free, safe, and potent alternative to current therapeutic modalities.

An engineered MoS₂ nanodots with sulfur vacancies as antibiotic‐free antifungal eye drops was successfully synthesized. Superior broad‐spectrum antifungal efficacy and multifaceted antifungal mechanism enable hybrid eye drops to intervene in fungal keratitis and effectively prevent the occurrence of fungal corneal blindness.

1. Introduction

Fungal keratitis (FK) is a highly blinding corneal infection characterized by stromal destruction, perforation, endophthalmitis and corneal scarring [1]. Its high misdiagnosis rate, the shortage of effective antifungal agents and the limited therapeutic outcomes are primary factors contributing to reduced visual acuity and blindness [2]. Fusarium is the most common pathogen, accounting for 37%–62% of the total pathogens [3]. The incidence and prevalence of FK are influenced by geographical location and climate fluctuations [4]. The cornea is especially vulnerable to fungal infections due to its susceptibility to injury and lack of blood vessels, which reduces its natural resistance. The primary pathogenic fungi responsible for FK are Fusarium species accounting for 53.2% of cases, followed by Aspergillus species at 15.9% in the past two decades in north China [5, 6]. FK tends to affect males and individuals in the middle‐aged population, who constitute a significant portion of the workforce [7]. This is attributed to the higher incidences of FK during the harvest seasons. FK is often treated with eye drops and lamellar/penetrating keratoplasty. As a noninvasive antibiotic administration method, eye drops account for 90% of ophthalmic clinical medication [8]. However, due to the lack of broad‐spectrum antifungal drugs, the types of specific antifungal drugs are limited, and the toxic and side effects of drugs aggravate corneal injury, even anterior chamber inflammation, and so forth, which increases the difficulty of FK treatment [9]. In addition, due to drug abuse, the problem of drug resistance is highlighted. Some fungal strains are resistant to commonly used drugs, and the therapeutic effect is worrying. More importantly, the corneal barrier can also hinder drug penetration, seriously affecting the efficacy.

Voriconazole and natamycin are commonly prescribed medications for FK treatment [10]. However, an ophthalmic formulation of voriconazole has not yet been approved. Although natamycin stands as the sole FDA‐approved therapeutic for FK, frequent dosing poses challenge to patient adherence, and the condition of FK worsens with the emergence of drug‐resistant fungal strains. Consequently, the development of novel functional nanomaterials characterized by efficient penetration and robust antifungal properties is of utmost urgency [11]. Recently, antimicrobial nanomaterials have ushered in a new era of hope for the treatment of ocular infections [12, 13, 14]. However, the majority of these nanomaterials rely on external stimulus, such as laser and ultrasound, or the incorporation of antifungal drugs and silver nanoparticles to achieve broad‐spectrum antifungal activity, with their therapeutic efficacy constrained by the limitations of treatment conditions and drug‐resistant fungi species [15, 16, 17, 18]. Even more unfortunately, recent research has discoveredmicrobial resistance to nanosilver [19]. To our knowledge, research on nanomaterials as broad‐spectrum antifungal agents is scarce, particularly in the context of utilizing molybdenum disulfide quantum dots for the treatment of FK, making this study a pioneering endeavor.

In this work, we report an innovative therapeutic approach for FK by employing meticulously designed MoS₂ nanodots, characterized by their unique sulfur vacancies, ultrasmall size, exceptional hydrophilicity, and surface positive charge, as antifungal eyedrops (Figure 1). This customized nano‐formulation leverages the synergistic antifungal effect of its sulfur vacancies and surface positive charge to enhance the treatment efficacy for FK. We first investigated the intrinsic catalytic performance of the sulfur vacancy‐rich MoS₂ without external stimuli. Based on this catalytic foundation, we systematically evaluated the antifungal activity of these eyedrops in vitro and delved into the underlying mechanism through comprehensive gene transcriptome analysis. Furthermore, this research thoroughly assesses the biosafety of the eyedrops both in vivo and in vitro, ensuring their safety and reliability during application. Additionally, we have established an FK infection model to investigate the eyedrops' ability to clear fungi in vivo and further explore the underlying mechanisms behind their excellent therapeutic effects in vivo. These series of explorations and discoveries will provide new insights and methodologies for the treatment of FK.

FIGURE 1.

Schematic illustration of the preparation of HNAF eye drops, the mechanism for overcoming corneal barrier and killing fungi, and the application in the treatment of fungal keratitis.

2. Results and Discussion

2.1. Synthesis and Characterization of MoS₂‐Based Hybrid Nano‐Antifungal (HNAF) Eye Drops

By utilizing the chemical Li‐intercalation method, ultrasmall single‐layer MoS₂ nanodots are derived from bulk MoS₂ powder, as outlined in our previously reported procedure (see details in the Experimental Section) [20]. Figure 2 illustrates the characterization of ultrasmall single‐layer MoS₂ nanodots. The transmission electron microscopy (TEM) image reveals that the as‐prepared sample is ultrasmall in size and exhibits uniformity (Figure 2A). The high‐resolution TEM (HRTEM) image in Figure 2B exhibits the continuous lattice fringes, which further confirms the single‐crystalline nature of MoS₂ nanodots. The MoS₂ nanodots have a thickness of approximately 1.1 nm as measured from their atomic force microscope (AFM) height image (Figure 2C), indicating that they are likely single‐layered. Atomic resolution scanning TEM (STEM) is employed to further characterize the crystal phase of the MoS₂ nanodots. Nanodots displaying distinct atomic patterns, each around 5 nm in size, are evident in the high‐angle annular dark‐field STEM (HAADF‐STEM) image shown in Figure 2D. In the filtered HAADF‐STEM image of a single nanodot, the triangular lattice fringes are formed by Mo atoms, while S atoms display minimal contrast (Figure 2E). When viewed perpendicularly from the basal plane, the corresponding line section analysis indicates that most of the two S atoms are dislocated between the two nearest Mo atoms, accompanied by individual S atom vacancies (Sv) (Figure 2F). The detailed analysis provides strong support for the metallic 1T phase structure of the as‐prepared MoS₂ nanodots, aligning with the findings in our previous report [20]. X‐ray photoelectron spectroscopy (XPS) is employed to further analyze the electronic structure of the metallic 1T phase MoS₂ nanodots. In the XPS Mo 3d spectrum (Figure 2G), there are two prominent peaks at 231.6 and 228.4 eV, suggesting that the MoS₂ nanodots are primarily composed of metallic 1T phases. Additionally, two shoulder peaks are observed at 233.2 and 229.4 eV, corresponding to the semiconductor 2H phase MoS₂. In the XPS S 2p spectrum (Figure 2H), we also observe two prominent peaks at 163.0 and 161.8 eV, which can be attributed to the 1T phase, as well as two smaller peaks at 163.3 and 164.4 eV, corresponding to the 2H phase. The XPS characterization reveals that the prepared MoS₂ nanodots consist mainly of metallic 1T phases, with an estimated 75% based on the Mo 3d spectrum.

FIGURE 2.

Characterization of MoS₂ nanodots. (A) TEM image, (B) HR‐TEM image, and (C) AFM height image of MoS₂ nanodots. (D) Atomic‐resolution HAADF‐STEM image of MoS₂ nanodots and (E) corresponding filtered image of the dashed box area in (D). (F) The cross‐sectional profiles along the white dashed lines in (E). High‐resolution (G) Mo 3d, and (H) S 2p spectra of MoS₂ nanodots.

The terminal functionalized copolymer mPEG‐b‐PLL‐TA (PEL‐TA) with disulfide bonds tailed was synthesized via multi‐step reactions outlined in Scheme S1. The chemical structures of the copolymers were confirmed by ¹H NMR, FTIR and Raman spectroscopy measurements. In Figure S1, the ¹H NMR spectra of the thioctic acid‐tailed copolymers PEL‐TA clearly showed the characteristic peaks of PEG and PLL units. The characteristic signals at 7.25 and 5.01 ppm of the benzyloxycarbonyl groups in polypeptide blocks disappeared completely upon deprotection. Moreover, the FT‐IR spectra in Figure S2 confirmed that the intensity of the ester bond absorption around 1720 cm⁻¹ from the benzyloxycarbonyl groups disappeared, indicating that the deprotection reaction was complete. The Raman spectrum of the copolymer PEL‐TA (Figure S3) clearly indicated the existence of disulfide bonds at 622 and 724 cm⁻¹ absorption peaks. These results confirmed the successful synthesis of PEL‐TA.

Generally, MoS₂ synthesized by the lithium intercalation method can be stably dispersed in water due to the negative charge on its surface caused by electron transfer during the intercalation process [21]. However, its stability under physiological conditions still faces significant challenges. Therefore, some amphiphilic or hydrophilic polymers (e.g., PEG, PVP, and PEI) are used to modify MoS₂, especially those containing thiol or disulfide bonds that can form stable covalent bonds with MoS₂ [22], exhibiting excellent biocompatibility. To regulate the surface charge and improve the biological stability of MoS₂, we utilized the positively charged (PEL‐TA) and electrically neutral (PEG‐TA) disulfide‐based polymers to covalently modify the surface of ultrasmall MoS₂ nanodots to obtain the corresponding PEL‐TA/MoS₂ and PEG‐TA/MoS₂ nanodots, respectively (see details in the Experimental Section). As shown in Figure S4 and Table S1, the blank MoS₂ nanodots exhibit a large negative potential, with a hydrated particle size of 6.48 nm, which is consistent with the results of TEM image (Figure 2d). After modification with PEL‐TA, the potential of MoS₂ was altered to +23.03 mV (Figure 3A), and its hydrated particle size increased to 27.23 nm (Figure 3A), suggesting its ability to traverse the corneal barrier. As the control group PEG‐TA/MoS₂, it has a size comparable to PEL‐TA/MoS₂ but has an opposite potential (−32.30 mV).

FIGURE 3.

Antifungal evaluation of HNAF in vitro. (A) size and zeta potential of PEG‐TA/MoS₂ and HNAF. MIC of HNAF of fungal strain NO. 211641 (B), MICs of HNAF of clinical fungal strains (C), and the corresponding growth‐inhibit curve (D). (E) The SEM images of fungal cells after treatment with 5 × MIC of HNAF. (F) The turbidity of fungal solution after treatment with different concentrations of HNAF, NAT and Vor for 24 h, and (G) the corresponding plate counting assays. (H) The protein leakage and DNA leakage of fungal cells with different treatments. (I) ROS generation level of different treatments in fungal cells. (J) GSH oxidation level in fungal cells under different treatments. The asterisks indicate significant differences (t‐test, *p < 0.05, **p < 0.01, ****p < 0.0001). Data are presented as mean ± SD (n = 3).

The PEL‐TA/MoS₂ was dispersed within PBS to formulate the HNAF eye drops. Subsequently, we employed the fluorescence labelling technique to visualize the penetration of these eye drops into the cornea. As depicted in Figure S5, fungal spores co‐cultured with HNAF/Cy5 exhibited distinct red fluorescence, confirming the successful synthesis of HNAF/Cy5 and paving the way for the subsequent step of real‐time in vivo observation of the penetration behavior of the eye drops. To characterize the chemical catalytic properties of the prepared nanodots before biological evaluations, we investigated the catalytic performance of HNAF in a cell‐free system. Electron paramagnetic resonance (EPR) spectroscopy was performed using DMPO (5,5‐dimethyl‐1‐pyrroline N‐oxide) as a spin‐trapping agent under dark conditions (Figure S11). Upon incubation with H₂O₂, the Sv‐rich HNAF reaction system exhibited significant characteristic signals of •OH, four peaks with relative intensities of 1:2:2:1 corresponding to the DMPO‐•OH adduct. These results clearly demonstrate that HNAF can mediate Fenton‐like catalytic reactions to generate cytotoxic •OH under dark physiological conditions, without requiring any external energy input such as light irradiation. Its excellent catalytic capability is attributed to the abundant sulfur vacancies within the structure, which are widely recognized as critical active sites for such reactions. This intrinsic, excitation‐free catalytic capability establishes a solid foundation for the subsequent therapeutic applications.

2.2. In Vitro Anti‐Fungal Activity

We first quantitatively evaluated the antifungal efficacy of HNAF eye drops by determining their minimum inhibitory concentration (MIC) values against clinically derived 20 strains of Fusarium solani. The MIC values were summarized in Figures 3C, S6, and Table S2. HNAF showed satisfactory antifungal activity against 20 clinical strains, and the highest MIC value was only 3.13 µg/mL (Table S3). The MIC of HNAF in strain 211641, which was 1.172 µg/mL (Table S2), was the lowest, showing strain 211641 is the most susceptible over the clinical strains. However, we examined the antifungal effects of strain 211567 to HNAF, which was the representative MIC concentration of 3.13 µg/mL for most strains. According to Figure 3D, HNAF can completely kill pathogens within 24 h at both 2× and 5 × MIC concentrations. The turbidity of fungal solution after treatment with different concentrations of HNAF, Natamycin eye drops (Nat) and Voriconazole eye drops (Vor) for 24 h is shown in Figure 3F. It can be seen that the fungal solution can still remain clear at 0.5 × MIC concentration of HNAF, Nat group starts to become cloudy at 1 × MIC concentration, and Vor can only remain clear at 5 × MIC concentration (MIC of Vor was 11.76 µg/mL and Nat was 5.16 µg/mL according to previous studies of our group) [23]. To further determine antifungal properties, the above fungal solution was spread on agar plate and cultured for another 48 h, it was found that HNAF could kill most pathogens at 1 × MIC concentration, some fungi cells could still survive with 5 × MIC concentration of Nat, while Vor could not kill fungi even at 5 × MIC concentration (Figure 3G). These results suggested that HNAF eye drops exhibited higher antifungal activity as compared to Vor and Nat.

2.3. Investigation of Antifungal Mechanism

In our study, we proposed a two‐step pathway leading to the synergistic anti‐fungal activity: (1) Membrane damage and induce leakage of intracellular contents. (2) Generation of reactive oxygen species (ROS) inside fungi cells.

2.3.1. Integrity Disruption of Fungal Membrane Induced by HNAF

To visualize the influence of HNAF on the morphology and surface of F. solani, the scanning electron microscope (SEM) was then employed to further investigate the morphology changes of fungi after treatment with HNAF. Prior to treatment, the spore and mycelia of fungi exhibited normal morphology with smooth cellular exteriors and intact cell wall/membrane. After being treated with HNAF, the shrinking, broken and collapse of the cell wall/membrane were obviously visible (Figure 3E). HNAF can damage fungi cell membrane structure, subsequently change its permeability, inducing important intracellular substances leaking out and dramatic change in the shape and size of cells. To assess the fungal membrane damage, DNA and protein loss were utilized to characterize the fungal membrane permeability. In general, the more severe fungal membrane was damaged, the stronger DNA and protein leakage out of fungi was achieved [24, 25]. As shown in Figure 3H, the DNA and protein depletion in fungi cells cultured with HNAF significantly increased in comparison with the control, suggesting that fungal membrane permeability was severely damaged.

2.3.2. ROS was the Lethal Factor for Antifungal Toxicity of HNAF

Oxidative stress has been considered as one of the most important mechanisms of toxicity related to nanoparticles exposure, they can damage DNA, cell membranes, and cellular proteins, and may lead to cell death [26]. To investigate the potential role of oxidative stress induced by HNAF, ROS generation in the presence of HNAF were monitored with oxidation‐sensitive fluorescent probe DCFH‐DA that passively diffuses through the cell membrane into the cell. The ROS level is in turn correlated with the anti‐fungal activity of the HNAF, due to its damaging effect to the cell membrane. As shown in Figure 3I, when cultivated with HNAF, strong fluorescence signals were detected after treatment, and with the increase concentration of HNAF, the fluorescence signal gets stronger both in spores and mycelia of the fungi cells. These results verify that HNAF induced intracellular ROS in fungus and it should be responsible for the antifungal effect.

One of the main functions of thiol‐dependent antioxidant systems is to scavenge ROS to keep cellular redox homeostasis and protect against oxidative stress. Glutathione (GSH) plays a vital role in the protective system of microbes, which protects the cellular contents from damage induced by oxidative stress [27]. The disruption of the GSH systems is also a possible reason for the upregulation of ROS. Therefore, typical Ellman's assay was performed to determine the GSH oxidation level under different treatments. H₂O₂ (1 mM) + GSH (0.8 mM) was used as positive control and the loss percentages in treatment of HNAF for 1 or 2 h were 75.06% and 83.8%, respectively. As shown in Figure 3J, the substrates with GSH solution displayed obvious incremental oxidation behavior as incubation time and incremental concentrations of HNAF. In HNAF with 0.25 × MIC concentration, the consumption of GSH at 1 and 2 h were 21.0% and 29.4%, respectively. Thus, HNAF could catalyze the oxidation of GSH of adherent fungal cells and result in the disruption of fungal antioxidant system.

Taken together with the above results, it could be concluded that the enhanced antifungal efficiency of HNAF was ascribed to the membrane damage and increase oxidative stress followed by cell death.

2.3.3. Potential Antifungal Mechanism of HNAF

The potential antifungal mechanism of HNAF was further explored through RNA sequencing to identify and analyze the differentially expressed genes (DEGs) in F. solani after HNAF‐treatment. Given that our cell‐free EPR results (Figure S11) demonstrated the intrinsic, stimuli‐free catalytic ability of HNAF to generate •OH radicals, we sought to clarify how this exogenous oxidative pressure alters the fungal transcriptional profile. As a result, a total of 6237 DEGs including 2433 up‐regulated genes and 3804 downregulated genes were detected compared to the control group (Figure 4A). The Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways analysis in Figure 4B showed top 25 enriched KEGG pathways, in which the pathway of metabolism was the most enriched (386 genes). Since the peroxisome is the most influenced cellular process and the excellent ROS generation ability of HNAF, we assume that the largest influence of HNAF on the KEGG pathway of metabolism is most likely due to the disruption of the redox metabolic process. Therefore, the part of DEGs with the dramatic change, especially the redox‐related genes, was further exhibited in the heatmap (Figure 4C). Figure 4D shows the significantly downregulated redox‐related genes in the peroxisome and glutathione pathways, suggesting the protection capacity against oxidative stress was broken after the treatment of HNAF.

FIGURE 4.

Gene transcriptome analysis. (A). Volcano plots of differentially expressed genes (DEGs) after being treated with HNAF (B). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways analysis of DEGs between the control group and HNAF‐treated group (n = 3). (C) Heatmap showed significant changed DEGs based on the expressed genes associated with redox metabolism. (D) Redox metabolism pathways and associated DEGs with dramatic change. (E) The mRNA expression levels of representative genes through the real‐time PCR (qRT‐PCR) after of the HNAF treatment (5 × MIC) and the PBS treatment as the control group. EnviIP: indicates environmental information process. CellP: indicates cell process. The asterisks indicate significant differences (*p < 0.05, **p  < 0.01). Data are presented as mean ± SD (n = 3).

Furthermore, we used quantitative real‐time PCR (qRT‐PCR) to confirm the results of transcriptomic analysis (Figure 4E), and the relative primer sequences are shown in Table S4. Peroxisomes are single‐layer membrane‐coated organelles that contain abundant enzymes and participate in multiple biochemical metabolic processes, such as β‐oxidation of fatty acid, glyoxylic acid cycle, generation and degradation of ROS [28]. We found that the relative mRNA expression of PEX14 and PEX19 were downregulated. PEX gene is mainly involved in peroxisome proliferation and plays an important role in peroxisome formation and fungal pathogenicity [29]. PEX14 is a dock‐complex protein, the absence of which leads to protein transport defects and mislocalization [30]. PEX19 is a membrane protein receptor involved in the transport of membrane proteins, recognizing and transporting them into peroxidase [29]. The key enzymes, antioxidative catalase (CAT) gene, Cu/Zn related superoxide dismutase (SOD1) gene and Mn related SOD2 gene in hydrogen peroxide metabolism in eukaryotes, are indispensable components of the whole peroxisomes antioxidant system [31]. Their downregulation suggested that the fungal antioxidant system was disrupted.

We also found that the relative mRNA expression of α‐methylacyl‐CoA racemase (AMACR) was downregulated. AMACR is found in peroxisome and is essential in β‐oxidation of branched‐chain fatty acids [32]. Glyoxylic acid cycle plays an important role in pathogen infection and in the growth and development of fungi. Isocitrate lyasc (ICL) and malate synthetase (MLS) are two key enzymes in the glyoxylate cycle. In fungi both enzymes are located in the peroxisome. The downregulation of the two genes led to the defect of appressorium formation, and the significantly decreased pathogenicity of fungi [33]. Compared with the control group, the relative mRNA expression of representative glutathione reductase (GSR) decreased significantly, which may indicate the catalytic conversion of GSSG into reduced GSH was affected, and the consumption of GSH was aggravated. Glutathione S‐transferase1 (GST1) was downregulated, indicating that GSH detoxification function was weakened. This finding suggested that the cellular redox balance changed and the cell's ability to maintain a high GSH/GSSG ratio reduced, thereby impaired the microbial antioxidant response [34].

In summary, integrating the potent catalytic capability of HNAF with fungal transcriptomic evidence suggests a clear link underlying its fungicidal mechanism. Specifically, the HNAF‐mediated catalytic ROS (•OH) generation acts as an initial trigger disrupting cellular redox homeostasis, which subsequently drives the downstream transcriptional response. HNAF‐induced downregulation of key peroxisome and glutathione genes, paired with direct intracellular GSH depletion, genetically and chemically compromises the fungal antioxidant defenses. This establishes a self‐amplifying feedback cascade, where the primary oxidative stress escalates into a secondary, substantial intracellular ROS accumulation, ultimately contributing to membrane peroxidation and cell death. Collectively, these findings successfully construct a robust framework bridging the physical‐chemical catalytic profiles of HNAF with its downstream molecular and phenotypic biological consequences. Nevertheless, the precise antifungal pathways of HNAF represent a highly complex and dynamic biological process, and more efforts will be required in future investigations to further explore and validate these molecular biological mechanisms.

2.4. Biocompatibility of HNAF

To evaluate the effects of HNAF on cell incubation, we examined the cells cytotoxicity of the HNAF by CCK‐8 assay. HCE‐2 cell, conjunctival epithelial cells and rabbit corneal stromal cells (CSCs) were incubated separately with different concentrations of HNAF for 24 or 48 h before assessment. As highlighted could be seen in Figure S7A,B, there was no significant difference in cell activity after human corneal epithelial cells were treated with different HNAF for 24 or 48 h. It was figured out from Figure S7C,D that there was no significant difference in cell activity after human conjunctival epithelial cells were treated with different HNAF for 24 or 48 h. As shown in Figure S7E,F, there was no significant difference in cell activity after CSCs were treated with different HNAF for 24 or 48 h. When the drug concentration reached 25 µg/mL, the cell survival rate decreased slightly, but was still greater than 80%. Therefore, the HNAF samples had good biocompatibility.

The viability status of the HCE‐2 cells was determined 24 h after HNAF treatment by Calcein‐AM staining (Figure 5A). As shown in Figure 5B, there was no significant difference in mean fluorescence intensity between control and HNAF groups. Therefore, the HNAF samples were not toxic to the cells and did not cause cell death.

FIGURE 5.

Corneal permeability in vitro and ex vivo. Representative images of HCE‐2 cell layers immunofluorescent stained with Ca‐AM (A), β‐catenin and ZO‐1 (C) after incubation with HNAF, alongside the corresponding statistical analysis results (B,D). (E) Representative micrographs of corneal sections obtained from each group stained with β‐catenin and ZO‐1 at 1 h after topical application, followed by (F) the statistical analysis results. (G) The time‐dependent corneal permeation behavior of HNAF/Cy5 eye drops by two‐photon microscopy in vivo. (H) Representative in vivo two‐photon microscopy 3D images of corneal at 150 min after administration HNAF/Cy5 in C57BL/6 mice. Data are presented as mean ± SD (n = 3). **p < 0.01, ***p < 0.001. MFI, mean fluorescence intensity.

2.5. HNAF Eye Drops Alters Tight Junction Protein Distribution: Implications for Permeability

ZO‐1 is a vital important component protein responsible for maintaining the mechanical barrier and permeability of the epithelium, playing a pivotal role in the tight junctions of the corneal epithelium [35]. β‐catenin, a multifunctional protein primarily situated on the cell membrane, participates in the regulation of intercellular adhesion. It is also found within tight junctions in the corneal epithelium [36]. To investigate the corneal permeability mechanism of HNAF, the expression of β‐catenin and ZO‐1 was detected by immunofluorescence staining in vitro and ex vivo. As depicted in Figure 5C, the tight junctions formed by β‐catenin and ZO‐1 proteins between cells were disrupted after 24 h of HNAF treatment in HCE‐2 cells. Consequently, the boundary of HCE‐2 cells was no longer continuous in the HNAF group. However, in the control group tight junctions were observed between HCE‐2 cells. The mean fluorescence intensity of β‐catenin and ZO‐1 was significantly lower in the HNAF group than in the control group (Figure 5D). The results indicated that HNAF effectively inhibited the formation of cell‐cell junctions in vitro. This disruption of intercellular tight junctions might have further enhanced the corneal penetration of HNAF. As illustrated in Figure 5E, both β‐catenin and ZO‐1 were predominantly located at the superficial layer of the corneal epithelium with a continuous distribution in the control group. However, upon treatment with HNAF, the fluorescence distribution became discontinuous, accompanied by a significant decline in the fluorescence intensity of the corneal epithelium (Figure 5F). This observation aligns with the in vitro findings, suggesting that HNAF facilitates its penetration into the corneal epithelium to reach the stromal layer by opening the tight junctions between corneal epithelial cells. These results collectively demonstrate that HNAF has the capability to open tight junctions both in vitro and in vivo. Thus, HNAF eye drops disrupt the continuous distribution of tight junction proteins ZO‐1 and β‐catenin, suggesting transient opening of intercellular junctions. Although the observed disruption of ZO‑1 and β‑catenin distribution suggests transient tight‑junction opening, functional assays (e.g., TEER or fluorescein‑dextran flux) and proper controls (e.g., EDTA and negatively charged MoS₂) are needed to confirm increased paracellular transport and the role of surface positive charge. Nonetheless, this morphological finding provides a plausible explanation for the deep stromal penetration of HNAF visualized by two‑photon microscopy (Figure 5G,H), and serves as a basis for future functional validation.

2.6. In Vivo Evaluation of HNAF Penetration and Platform Retention

Fungi enter the corneal stroma through epithelial defects for invasive growth, but in the later stage of FK treatment, the epithelium recovers completely and the drug needs to further overcome the corneal barrier to reach the stroma. Therefore, evaluation of corneal permeability of drugs is particularly important in the treatment of FK. We used fluorescence labeling technology to directly study and observe the HNAF penetration behavior of mice ocular surface in real time under a two‐photon laser scanning microscope. With the extension of administration time, HNAF/Cy5 drugs gradually spread from the corneal epithelium to the stroma, and the fluorescence in the stroma became stronger and stronger, showing a good time dependence (Figure 5G,H). It is worth noting that the corneal endothelium of mice showed fluorescence at 150 min after administration (Figure S8). The above results indicate that HNAF has the ability to overcome the corneal barrier and can reach the stromal layer to exert its efficacy.

2.7. In Vivo Antifungal Therapeutic Effects

The antifungal activity of HNAF eye drops was explored by establishing the FK mice model. Figure 6A showed the timeline of the FK model establishment. The FK model was established, with drug intervention starting 12 h post‐fungal inoculation (recorded as 0 h), using natamycin (Nat) as a positive control. The negatively charged PEG‐TA/MoS₂ was excluded as a control due to its lack of antifungal activity (Figure S6); instead, the FDA‐approved Nat served as the clinical standard, against which HNAF demonstrated superior efficacy and stromal penetration (Figure 5G,H). The mice corneas were observed, scored, and photographed daily for 7 days. the FK model was established as shown in the figure, and 12 h after fungal inoculation was recorded as 0 h. Drug intervention was given starting at 0 h, and Nat was selected as a positive control. The drug was administered twice daily for 7 days. The mice cornea was observed by slit lamp and scored and photographed at 1, 2, 3, 5 and 7 days of FK.

FIGURE 6.

In vivo antifungal therapeutic effects of HNAF. (A) Timeline of the establishment of fungal keratitis model and drug administration. (B) Representative images of cornea. (C) Slit‐lamp micrographs. (D) Corneal fluorescein stained at 1 d, 2 d, 3 d, 5 d, 7 d after FK under different treatments. (E) Clinical scores for FK after treatment with topical eye drop solutions (n = 6). (F) The corneal thickness and (G) Relative epithelial defect area at 7 d after FK in normal, control, NAT and HNAF groups (n = 6). (H) Representative histological sections stained with hexamine silver of mice cornea after treatment with the topical eye drop solutions on the second day in normal, control, NAT and HNAF groups. (I) Photographic images showing F. solani colonies after corneal processing and subsequent culture for day 7 in normal, control, NAT and HNAF groups. (J) Analysis of fungal numbers (n = 3). ND, not detected. CFU, colony forming unit. Data are presented as the mean ± SD; *p < 0.05, **p < 0.01, ***p < 0.001.

As shown in Figure 6B, the control cornea and the Nat cornea showed more severe corneal opacity, edema and infiltration than the HNAF cornea at day 1. With the passage of time, the corneal opacity and edema infiltration in the control group gradually aggravated, and the whole cornea ulcer appeared on the seventh day (Figure 6B,C). There were no serious changes in corneal opacity and edema infiltration in the Nat group, and the cornea was still cloudy on the seventh day (Figure 6B,C). In the HNAF group, the corneal opacity and edema were gradually improved, and the cornea was basically transparent on the seventh day, showing the phenomenon of being cured (Figure 6B,C). The condition of the Nat group was more severe than that of the HNAF group at 1, 2, 3, 5 and 7 days, which was confirmed by fluorescein sodium photographs (Figure 6D). As depicted in Figure 6D, fluorescein sodium staining of the cornea in the control group persisted from 1 to 7 days, suggesting compromised corneal integrity. In the Nat group, fluorescein sodium staining was most severe on the third day and persisted until the seventh day without returning to normal. Conversely, in the HNAF group, no significant corneal staining was observed throughout the course of the disease, indicating successful corneal healing and complete recovery of the corneal epithelium in mice. Corneal clinical scores were recorded for the control, Nat, and HNAF eye drop groups in mice with FK after 1, 2, 3, 5 and 7 days of treatment. As depicted in Figure 6E, the total corneal score in the control group exhibited a steady increase over time, indicating ongoing development and deterioration of corneal disease. Conversely, both the HNAF and Nat groups displayed lower scores on the first day, with significantly lower scores observed in both groups compared to the control group on the second, third, fifth, and seventh days.

From day 2 to day 7, the Nat group exhibited significantly lower scores compared to the control group, indicating stable disease with no significant aggravation but also no notable improvement. Conversely, the scores of HNAF eye drop groups gradually decreased, from day 1 to day 7, with significantly lower clinical scores than both the Nat group and the control group. This downward trend in scores suggests a gradual cure of FK with HNAF treatment. Compared to Nat eye drops, HNAF demonstrated superior efficacy with lower clinical scores, effectively shorten the course of FK.

Comparison of corneal thickness and the area of corneal epithelial defect provided insights into corneal injury repair. Corneal thickness was significantly reduced in both the HNAF and Nat groups compared to the control group (Figure 6F). Particularly, corneal thickness in the HNAF group approached normal levels. Similarly, at 7 days post‐model creation, the relative corneal epithelial defect area was notably lower in both the HNAF (32.88) group and the Nat group (49.12) compared to the control group (104.28) (Figure 6G). Remarkably, the HNAF group demonstrated the most significant recovery, indicating that HNAF may have a more pronounced beneficial effect on accelerating corneal epithelial healing. This finding suggests that HNAF, a nanomaterial with antimicrobial and anti‐inflammatory properties, may facilitate quicker wound closure in the corneal epithelium by promoting cell proliferation and migration, potentially through its ability to combat fungi and support the healing process. Hexamine silver staining revealed a substantial invasion of fungal hyphae into the corneal stroma in the control group, indicative of severe fungal infection (Figure 6H). Conversely, the Nat group exhibited a smaller amount of fungal infiltration in the cornea. Interestingly, no significant specific fungal staining was observed in the corneas of the HNAF group compared to the control group. These images directly visualize established fungal infection (control) and its clearance after HNAF treatment (HNAF group), confirming the in vivo antifungal efficacy. These findings highlight the excellent antifungal effect of HNAF eye drops on FK. The antifungal efficacy of HNAF eye drops was further evaluated by re‐culturing corneal tissue on the seventh day of the FK model (Figure 6I). As depicted in Figure 6J, significant colony growth was observed in the control group, with a markedly higher number of colonies compared to the HNAF and Nat groups. While some colony growth was noted in the Nat group, minimal fungal growth was observed in the HNAF group, with significantly fewer colonies than the Nat group. Remarkably, no viable fungi were detected in the corneas of the HNAF group upon re‐culturing, consistent with the results of scoring and slit‐lamp photographs. Compared with the other two groups, the corneal structural integrity of the HNAF group was restored. These findings underscore the excellent therapeutic efficacy of HNAF eye drops in vivo.

The pathological conditions of mice cornea, such as the infiltration of inflammatory cells in the cornea, hyphal invasion and the integrity of corneal structure, were studied by tissue staining. On the seventh day, HE staining showed that the number of inflammatory cells in the cornea of the control group was larger, and the integrity of the corneal structure was destroyed (Figure 7A). The corneal structure of the Nat group and the HNAF group was relatively complete. Only mild inflammatory reaction was observed in the Nat group, and almost no obvious inflammatory reaction was observed in the HNAF group (Figure 7A). These findings underscore the ability of HNAF eye drops to mitigate the inflammatory response in FK and effectively cure the condition.

FIGURE 7.

(A) HE staining of cornea at 7 d after FK in normal, control, NAT and HNAF groups. (B) Fluorescence staining images of NF‐κB, IL‐6, and TNF‐α of the control, NAT and HNAF groups on the seventh day. (C) Mean fluorescence intensity (MFI) of NF‐κB, IL‐6, and TNF‐α. Data are presented as mean ± SD, n = 3 per group. *p < 0.05, **p < 0.01, ***p < 0.001. (D) The tissue safety assessment of HNAF at 7 d after FK, HE staining including lirs, ciliary body, lens and retina in normal, control, NAT and HNAF groups.

Furthermore, inflammatory factors are increased in FK, which is an inflammatory condition of the cornea caused by fungal infections. FK is characterized by the invasion of the cornea by pathogenic fungi, which triggers a host immune response, leading to the release of various inflammatory factors [37]. The production of pro‐inflammatory cytokines such as NF‐κB, IL‐6, and TNF‐α is increased in response to fungal infection [38, 39]. These cytokines play a crucial role in eliciting the migration of immune cells and facilitating inflammation at the location of infection. Therefore, the secretions of proinflammatory cytokines (NF‐κB, IL‐6, and TNF‐α) were detected through an immunofluorescence test, and the statistical analysis was performed and is shown in Figure 7B. The HNAF and Nat groups had significantly lower levels of inflammatory factors (NF‐κB, IL‐6, and TNF‐α) than the control group, and the HNAF group had significantly lower levels than the Nat groups (Figure 7C). Our findings highlight the significant reduction in inflammatory response achieved by HNAF in FK. These multiple effects, namely fungicidal activity, anti‐inflammatory action, and promotion of epithelial repair, are not independent but likely synergistic. The elimination of fungal pathogens reduces the antigenic drive for NF‐κB activation, thereby lowering IL‐6 and TNF‐α levels. This dampened inflammation relieves the suppression of corneal epithelial regeneration, accelerating wound closure. Concurrently, the transient opening of tight junctions enhances HNAF penetration into the stroma, improving drug delivery to the site of infection without exacerbating inflammation because the anti‐inflammatory effect dominates. This multi‑target, single‑agent strategy uniquely addresses the concurrent infection, inflammation, and tissue damage in FK, representing a distinct advantage over conventional therapies that target only one aspect of the disease. While inflammation serves as a crucial defense mechanism against fungal infections in FK, excessive or unregulated inflammation can lead to tissue damage and vision impairment [40, 41]. Hence, effective management of inflammation is imperative in FK treatment, typically involving the use of anti‐fungal medications and anti‐inflammatory drugs to combat both infection and inflammation. The outcomes of our study underscore that HNAF not only exhibits remarkable anti‐fungal properties but also delivers a substantial reduction in inflammation.

2.8. Biosafety of HNAF In Vivo

The potential ocular irritation of HNAF was evaluated according to a slightly modified Draize test [42]. As illustrated in Figure S9A, slit‐lamp photographs revealed no conjunctival discharge, conjunctival hyperemia, or corneal epithelial damage in rabbit eyes for up to 72 h following administration of HNAF eye drops. This suggests that rabbit eyes can tolerate HNAF well. Furthermore, HE staining demonstrated that the structure and integrity of various ocular tissues including the cornea, conjunctiva, iris, ciliary body, and retina remained unaffected (Figure S9B), indicating that HNAF is safe to the eyes [43].

HE staining was used to observe the iris, ciliary body, ciliary body and retina of the mice eyes except the cornea after 7 days of FK (Figure 7D), the tissue structure was not affected, indicating that HNAF did not cause damage to other ocular structures [44]. HE staining of the heart, liver, spleen, lung and kidney showed that the structure of each tissue was not affected (Figure S10), indicating that HNAF did not cause damage to other organs and tissues of the whole body. These findings indicate that HNAF exhibits no ocular or systemic toxicity. The results suggest that HNAF does not induce corneal irritation and possesses a high level of safety, thus sparking potential clinical interest [45].

Several nano‐based antifungal approaches have recently been explored for FK, including photodynamic therapy (PDT), enzymatic nanoparticles (e.g., lyticase), silver nanoparticles (AgNPs), and drug‐loaded nanocarriers. PDT, despite in vitro efficacy, is clinically impractical for ophthalmic use because light cannot adequately penetrate the opaque or ulcerated cornea, and the required equipment is not suitable for routine eye care. Natural enzymatic systems are costly, unstable, and potentially immunogenic. Silver nanoparticles can cause dose‐dependent corneal toxicity and emerging resistance, while drug‐loaded carriers still rely on conventional antifungals. In contrast, HNAF functions as a defect‑engineered nanozyme. As recently reviewed by Li et al., engineering defects such as sulfur vacancies can unlock superior biocatalytic activities in nanozymes, including enhanced ROS generation and glutathione depletion [46]. Leveraging this property, HNAF not only acts as a self‑driven antifungal nanozyme but also achieves excellent corneal penetration owing to its small size (∼27 nm) and positive surface charge (+23.03 mV), enabling deep stromal delivery without external activation. Overall, this sulfur‐vacancy‐based and antibiotic‐free MoS₂ nanodot formulation provides a novel strategy for the treatment of FK.

3. Conclusions

Our findings indicate that HNAF eye drops effectively enhance corneal permeability by opening tight junctions of corneal epithelial cells, thereby improving drug permeability. The broad‐spectrum antifungal performance of this material is excellent against more than 20 common clinical fungal species, outperforming even conventional clinical drugs. Furthermore, HNAF eye drops exhibit no toxicity to ocular surface‐associated cells and are not irritating to rabbit eyes. Additionally, HNAF eye drops demonstrate enhanced anti‐fungal activity both in vitro and in vivo while reducing dosage and frequency requirements, local ocular irritation, and side effects. Notably, HNAF exhibits dual properties, not only demonstrating anti‐fungal and anti‐inflammatory effects, but also restoring corneal structural integrity. In summary, our study underscores the promising potential of HNAF eye drops for the intervention of FK.

Author Contributions

Ruixing Liu: writing – original draft, methodology, formal analysis, data curation, conceptualization. Xiwen Geng: writing – original draft, methodology, formal analysis, data curation. Huiying Chen: writing – original draft, methodology, formal analysis, data curation. Boyuan Cheng: investigation, formal analysis, data curation. Huayang Feng: investigation, formal analysis, data curation. Haohao Cui: investigation, formal analysis, data curation. Xiangrong Pan: investigation, formal analysis, data curation. Jingguo Li: funding acquisition, project administration, resources, writing – review & editing. Zhan Zhou: validation, supervision. Chaoliang Tan: writing – review & editing, resources, project administration, funding acquisition. Zhanrong Li: funding acquisition, project administration, resources, writing – review & editing.

Funding

This research was supported by the Natural Science Foundation of Henan Province of China (242300421018, 252300421044 and 252300421024), the National Natural Science Foundation of China (82371108, 52173143 and 82101090), Henan Province Science and Technology Research and Development Plan Joint Fund Project (242301420010), Henan Province Youth Health Science and Technology Innovation Talent Training Project (LJRC2024003), the Science and Technology Innovation Talent Program of University in Henan Province (23HASTIT016), and the Basic Science Key Project of Henan Eye Hospital (23JCZD001 and 24JCZD003). C.T. thanks the funding support from the Start‐Up Grant (Project No. 9610710) from City University of Hong Kong and ITC via Hong Kong Branch of National Precious Metals Material Engineering Research Center (NPMM).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: adhm71459‐sup‐0001‐SuppMat.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.