**Background:** Dry eye disease (DED) is a multifactorial condition driven by a self-perpetuating vicious cycle involving tear hyperosmolarity, oxidative stress, inflammation, and apoptosis. Reactive oxygen species (ROS) activate the NLRP3 inflammasome, leading to IL-1β and IL-18 secretion, which exacerbates ocular surface inflammation. Current anti-inflammatory treatments (e.g., corticosteroids, cyclosporine A) have variable efficacy and side effects. Nanozymes, particularly dual-atom nanozymes (DAN), offer a novel approach to scavenge excess ROS and break this cycle. This study aimed to develop FeMn-DAN eye drops and evaluate their antioxidant, anti-inflammatory, and therapeutic effects in DED.
**Methods:** FeMn-DAN was synthesized by embedding Fe and Mn single atoms in N-doped carbon (FeMn-DA/NC) via a four-step method (FeMnZIF-8 synthesis, SiOₓ coating, calcination at 900°C, and acid etching). The nanozyme was characterized by TEM, ac-HAADF-STEM, XRD, XPS, XANES, and ICP-MS. In vitro, human corneal epithelial cells (HCE-2) were exposed to hypertonic medium (500 mOsM) to induce oxidative stress. Cells were pre-treated with DAN (1, 2, 4, 8 μg mL⁻¹) or NAC (10 mM). ROS levels were measured using DCFH-DA and MitoSOX Red. Antioxidant enzyme activities (SOD, CAT, GPX) were assayed, and expression of SOD1, CAT, GPX1, HO-1 was assessed by immunofluorescence and western blot. Cell viability was measured by CCK-8. Apoptosis was evaluated via 8-OHdG and JC-1 staining. NLRP3 inflammasome activation (NLRP3, ASC, Caspase-1, IL-1β, IL-18, NF-κB P65, P-NF-κB P65, IL-6) was examined by immunofluorescence and western blot. In vivo, DED was induced in C57BL/6 mice using 0.2% benzalkonium chloride (BAC) twice daily for 7 days. Mice then received 5 μL of 0.9% saline, DAN (100 μg mL⁻¹), or 0.05% CsA twice daily for 7 days (n=6/group). Tear volume, corneal opacity (0-4 scale), and fluorescein staining (0-4 per area) were assessed at days 0, 4, and 7. Histological analysis included H&E staining (corneal epithelial thickness), PAS staining (goblet cell count), TUNEL (apoptosis), DHE (ROS), and immunofluorescence for 8-OHdG, NLRP3, ASC, Caspase-1, IL-1β, IL-18. Safety was evaluated by H&E of ocular tissues and major organs.
**Key Results:** FeMn-DA/NC had an average size of 150 nm, with Fe and Mn content of 0.23 wt% and 0.02 wt%, respectively. XANES confirmed Fe valence between 0 and +3, Mn between 0 and +4, with Fe-N and Mn-N coordination. DAN showed low cytotoxicity (>80% viability at 0.98-31.25 μg mL⁻¹). In vitro, DAN (2 μg mL⁻¹) significantly reduced hypertonicity-induced ROS (DCF fluorescence) and mitochondrial ROS (MitoSOX), comparable to NAC. DAN increased SOD, CAT, and GPX activities and upregulated SOD1, CAT, GPX1, and HO-1 expression. Cell viability improved from ~50% (hypertonic) to 62% (DAN 2 μg mL⁻¹). DAN suppressed 8-OHdG and restored mitochondrial membrane potential (JC-1). DAN inhibited NLRP3/ASC/Caspase-1 activation and reduced IL-1β, IL-18, IL-6, NF-κB P65, and P-NF-κB P65 expression (immunofluorescence and western blot). In vivo, DAN-treated mice showed reduced corneal opacity scores (e.g., day 7: DAN vs. control), fluorescein staining scores (negligible micropunctate vs. plaque staining), and increased tear volume (5.18±0.29 mm vs. 3.02±0.60 mm in control; CsA: 5.38±0.48 mm). H&E revealed preserved corneal epithelial thickness and reduced stromal infiltration. PAS staining showed significantly restored goblet cell numbers (p<0.01 vs. control). DHE staining confirmed reduced ROS in DAN and CsA groups. TUNEL and 8-OHdG staining were decreased. NLRP3, ASC, Caspase-1, IL-1β, and IL-18 expression in corneal epithelium were markedly reduced by DAN, with superior IL-1β suppression compared to CsA. No toxicity was observed in ocular or systemic tissues.
**Clinical Implications:** FeMn-DAN eye drops effectively scavenge ROS, restore antioxidant defenses, inhibit NLRP3 inflammasome activation, and break the vicious cycle of DED. In a mouse model, DAN improved tear secretion, corneal integrity, and goblet cell preservation, comparable to or better than CsA, with excellent biocompatibility. This nanozyme-based strategy offers a promising, safer alternative for DED treatment by targeting oxidative stress and inflammation simultaneously. Further studies are needed to optimize biosafety and stability for clinical translation.