Corneal Nerves Promote Alkali Burn Repair by Modulating Macrophages and Neutrophils via Calcitonin Gene-Related Peptide.

All animal experiments were approved by the Institutional Animal Care and Use Committee of the General Hospital of Chinese PLA and were in keeping with the standards in the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. Six to 8-week-old male C57BL/6 mice were purchased from GemPharmatech (Nanjing, China) and raised in the Animal Experimental Center of the General Hospital of Chinese PLA (Beijing, China). The mice were anesthetized by intraperitoneal injection of 2.5% avertin (250 mg/kg; Sigma-Aldrich). The cornea was topical anesthetized with procaine eye drops before the alkali burn. A circular filter paper (2 mm in diameter) was soaked with 1.8 µL of 0.5 N NaOH and applied to the central cornea for 30 seconds. After removing the filter paper, the cornea was immediately rinsed with 50 mL of PBS for 1 minute. Subsequently, gatifloxacin eye gel was applied to the injured eye. After the experiment, mice were placed on a homeothermic blanket until full recovery from anesthesia. Mice in the control group did not receive the injury. The corneal denervation (CD) mouse model was performed as previously described. 29 Briefly, mice were anesthetized by intraperitoneal injection of avertin. The temporal eyelid margin was incised, followed by an incision into the temporal conjunctival sac. The peribulbar tissue was dissected to expose the optic nerve. The nerve bundle accompanying the optic nerve was clamped for 30 seconds. Next, the conjunctival sac and the temporal eyelid were sutured. Control mice underwent a sham surgery that involved only a conjunctival incision and dissection of the peribulbar tissue. After surgery, the corneal alkali burn was performed, and gatifloxacin eye gel was applied to the conjunctival sac. Resiniferatoxin (RTX; HY-N2333; MedChenExpress) is a potent and selective agonist of the transient receptor potential vanilloid 1 (TRPV1) receptor. In animal models, RTX is capable of inducing long-term ablation of sensory nerves. RTX was dissolved in a vehicle consisting of 10% DMSO, 40% PEG300, 5% Tween-80, and 45% saline, at a working concentration of 0.5 µg/µL. To ablate corneal sensory nerves, mice were treated with RTX eye drops 3 times daily for 5 consecutive days, after which a corneal alkali burn was induced to evaluate the role of sensory nerves in tissue repair. To observe corneal nerve damage, corneal nerves were stained with β-III tubulin, and corneal whole-mount images were analyzed using ImageJ software. Corneal nerve was quantified as the percentage area occupied by β-III tubulin-positive nerve fibers within the region of interest. We dissolved CGRP powder (HY-P0203; MedChenExpress) in sterile PBS to a final concentration of 50 µM, aliquoted it, and stored it at –80°C. We treated mice in the CGRP group with 5 µL eye drops 4 times per day for 14 days. The control group received topical administration of PBS, 4 times daily for 14 days. BIBN-4096 is a CGRP receptor antagonist. It binds to the CGRP heterodimeric receptor CALCRLl-RAMP1, blocking CGRP from binding and thereby inhibiting its physiological effects. 30 BIBN-4096 was purchased from MedChenExpress (HY-10095). For the animal experiments, the working concentration of BIBN-4096 was 1 mg/kg. Briefly, BIBN-4096 was first dissolved in pure dimethyl sulfoxide (DMSO) to a stock concentration of 2 mg/kg. This stock solution was then reconstituted with an equal volume of PBS to achieve a final concentration of 1 mg/kg. Mice in the BIBN-4096 group received a 10 µL subconjunctival injection (at both inferotemporal and superotemporal sites) 1 day before alkali burn induction. Subsequent subconjunctival injections were given every other day until the animals were euthanized. Control mice received subconjunctival injections of the vehicle solution, which consisted of a 50% DMSO and 50% PBS mixture. PLX5622, a highly selective CSF1R inhibitor, enables noninvasive and sustained macrophage depletion through oral administration and has been widely used in ophthalmic research. 31 – 33 To deplete macrophages, 6-week-old C57BL/6 mice were fed a diet of PLX5622 AIN-76A (1200 mg plx5622/KG; M1809060; Moldiets) or a control diet for 14 days. 33 On day 14, the mice underwent corneal alkali burn. Following the injury, the mice were treated topically with either CGRP solution or PBS. The PLX5622 or control diet was continued until the mice were euthanized for tissue collection. Following alkali burn, corneal healing was assessed using a slit-lamp microscope. The corneal epithelial defects were stained with 1% sodium fluorescein, and images were captured under cobalt blue light. The images were then analyzed using ImageJ software. Corneal opacity was evaluated under white light and scored according to a previously established grading system as follows: 0 (transparent cornea), 1 (slight opacity, iris and pupils clearly visible), 2 (moderate opacity, iris and pupils visible), 3 (severe opacity, iris and pupils hardly visible), and 4 (complete opacity, iris and pupils completely obscured). Two ophthalmologists, who were blinded to the experimental groups, independently performed the opacity grading. Anterior segment optical coherence tomography (OCT) of the cornea was performed in mice using an inverse spectroscopic OCT (IS-OCT; Optoprobe, Pontypridd, UK) system according to the manufacturer’s guidelines. Tibiae and femurs were isolated from 8 to 12-week-old mice. The epiphyses at both ends of the bones were removed, and the bone marrow cells were flushed out using a 23G needle. The cell suspension was then filtered through a 70-µm cell strainer. For the culture of neutrophils, the mouse bone marrow neutrophil isolation solution kit (Solarbio) was used to isolate neutrophils according to the manufacturer's instructions. We resuspend the neutrophils in RPMI-1640 medium containing 100 units/mL penicillin/streptomycin and 10% fetal bovine serum (FBS). For bone marrow-derived macrophages (BMDMs) culture, the harvested bone marrow cells were cultured in RPMI-1640 medium containing 10% FBS, 20 ng/mL M-CSF (315-02; PeproTech), and 100 units/mL penicillin/streptomycin for 7 days to induce differentiation. The medium was changed on day 3 and day 5. Mature BMDMs were obtained on day 7. Bone marrow-isolated neutrophils were incubated with CGRP (1 nM) for 10 minutes, followed by culturing with RPMI-1640 medium containing 5% FBS and 100 units/mL penicillin/streptomycin for 6 hours at 37°C with 5% CO 2 to induce cell death. After 6 hours, the cells were seeded onto poly-lysine-treated coverslips. Following a 30-minute incubation to allow for complete adhesion, TUNEL labeling was performed. For macrophages, BMDMs cultured for 7 days were seeded into 24-well plates containing coverslips. The cells were treated with 1 nM CGRP for 20 minutes, followed by incubation in pure RPMI-1640 medium for 24 hours to induce apoptosis. After 24 hours, the cells were washed with PBS and subjected to a TUNEL assay. SQ22536 is an adenylyl cyclase inhibitor that blocks cAMP production. To assess the effect of CGRP and SQ22536 co-incubation on BMDMs apoptosis, cells were treated with 1 nM CGRP and 30 µM SQ22536 for 20 minutes, followed by incubation in pure RPMI-1640 medium for 24 hours to induce apoptosis. For TSP-1-induced apoptosis, BMDMs were incubated with RPMI-1640 medium containing 100 nM TSP-1 (HY-P701325; MedChenExpress) for 24 hours. Apoptosis was assessed after the 24-hour treatment period. BMDMs cultured for 7 days were seeded into 6-well plates. The cells were first treated with 1 nM CGRP for 20 minutes, followed by incubation with RPMI-1640 medium containing IL-4 (3 ng/mL, PeproTech), IL-13 (3 ng/mL, PeproTech), 10% FBS, and 100 units/mL penicillin/streptomycin for 48 hours. For the effect of co-incubation with CGRP and SQ22536 on BMDMs polarization, BMDMs were treated with 1 nM CGRP and 30 µM SQ22536 for 20 minutes, followed by culture in complete RPMI-1640 medium containing 3 ng/mL IL-4 and IL-13 for 48 hours. For TSP-1 treatment, BMDMs were incubated with RPMI-1640 medium containing IL-4, IL-13, 10 nM TSP-1, 10% FBS, and 100 units/mL penicillin/streptomycin for 48 hours. The expression of polarization markers was then analyzed by qPCR. BMDMs cultured for 7 days were seeded in a 24-well plate containing coverslips. The culture medium used was RPMI-1640 medium containing 1 nM of CGRP, 10% FBS, and 100 units/mL of penicillin/streptomycin. For the SQ22536 inhibition experiment, complete RPMI-1640 medium containing 1 nM CGRP and 30 µM SQ22536 was used. For the TSP-1 experiment, complete RPMI-1640 medium containing 10 nM TSP-1 was used. The cells were cultured for 24 hours. Neutrophils were labeled with PKH26 (C2071S; Beyotime) according to the manufacturer's instructions, and cultured in RPMI with 2% serum for 12 hours to induce cell death. Pre-treated BMDMs were incubated with PKH26-labeled dead/dying neutrophils at 37°C for 30 minutes. The reaction was stopped by washing cells with ice-cold PBS, and the cells were fixed with 4% paraformaldehyde at room temperature for 30 minutes for the subsequent immunofluorescence staining. Freshly isolated BMDMs (1 × 10⁶ cells) were stimulated with 1 nM CGRP in complete RPMI-1640 medium (supplemented with 10% FBS and 100 units/mL penicillin/streptomycin) for 30 minutes under standard culture conditions (37°C, 5% CO₂). For the SQ22536 inhibition experiment, cells were treated with complete RPMI-1640 medium containing 1 nM CGRP and 30 µM SQ22536 for 30 minutes. Subsequently, intracellular cAMP concentrations were determined using a cAMP ELISA kit (QTJA20937; CZKEWEI), following the manufacturer’s established protocol. To investigate the effect of CGRP on TSP-1 expression, BMDMs cultured for 7 days were seeded on coverslips and stimulated with 1 nM CGRP alone or 1 nM CGRP in combination with 30 µM SQ22536 in complete RPMI-1640 medium for 24 hours. After washing with PBS, the cells were fixed with 4% paraformaldehyde and subjected to immunofluorescence staining to detect TSP-1 expression. Fluorescence intensity was quantified using ImageJ software. On day 14, the mice were euthanized, and the eyeballs were enucleated and fixed in 4% paraformaldehyde. Following dehydration, the eyeball tissues were embedded in paraffin and sectioned at a thickness of 4 µm. The sections were stained with hematoxylin and eosin (H&E) and imaged using an optical microscope (Olympus, Melville, NY, USA). According to the manufacturer’s instructions, frozen sections were fixed with 4% paraformaldehyde at room temperature for 30 minutes and permeabilized with 0.5% Triton X-100 for 5 minutes. The sections were then incubated with the TUNEL reaction mixture (C1088; Beyotime) at 37°C for 60 minutes. Finally, the nuclei were counterstained with DAPI, and the sections were mounted. For adherent cells, after fixation, the cells were permeabilized with 0.3% Triton X-100 for 5 minutes and then incubated with the TUNEL reaction mixture at 37°C for 60 minutes. Finally, the nuclei were counterstained with DAPI, and the coverslips were mounted. The mouse corneas were cut into small pieces and digested in 37°C PBS containing 2.5 mg/mL Liberase TM (Sigma-Aldrich, St. Louis, MO, USA) for 50 minutes. Then, they were filtered through a 70-µm cell strainer to obtain a single-cell suspension. The following antibodies were used: PerCP anti-mouse CD45 antibody (103130; Biolegend), PE/Cyanine7 anti-mouse/human CD11b antibody (101216; Biolegend), FITC anti-mouse CD206 antibody (141703; Biolegend), APC anti-mouse Ly-6G antibody (127614; Biolegend), PE anti-mouse CD11c antibody (117307; Biolegend), and CD3 antibody, and anti-mouse (130-119-798; Miltenyi). Single cell suspensions of corneas were stained with antibodies for 15 minutes at room temperature. Flow cytometry analysis was performed on Beckman DxFLEX (Beckman Coulter, USA), and the data were analyzed with FlowJo version 9.2 (FlowJo, LLC, Ashland, OR, USA). The proportion of macrophages in the cornea was calculated as: CD45+ cell percentage × percentage of F4/80+CD11b+ cells within the CD45+ population. The proportion of neutrophils in the cornea was calculated as: CD45+ cell percentage × percentage of F4/80−Ly6G+ cells within the CD45+ population. For corneal whole-mount preparation, the procedures were as follows: mouse corneal samples were first fixed in 4% paraformaldehyde at room temperature for 30 minutes, followed by permeabilization with 0.5% Triton X-100 for 20 minutes at room temperature. Nonspecific binding sites were blocked by incubation with goat serum at room temperature for 1 hour. For quantitative analysis and morphological observation, the corneal samples were incubated with CGRP antibody (1:200; 14959; Cell Signaling Technology) or neuron-specific beta-III tubulin NL557-conjugated antibody (1:20; NL1195R; R&D Systems) overnight at 4°C. For samples requiring a secondary antibody, they were washed and subsequently incubated with a goat anti-rabbit IgG H&L 488 (1:500, ab150077; Abcam) for 2 hours at room temperature. After final washes, the samples were mounted using DAPI-containing medium (Abcam). The images were captured using Olympus slideview VS200 (Olympus, Japan). For frozen section and cell climbing slices, sections were fixed with 4% paraformaldehyde for 10 minutes (frozen section) or 30 minutes (cell climbing slices), permeabilized with 0.3% Triton X-100 for 5 minutes, and blocked with goat serum for 30 minutes. The sections were then incubated with primary antibodies at 4°C overnight. The primary antibodies used were as follows: anti-CD68 antibody (1:200, ab283654; Abcam), anti-myeloperoxidase antibody (1:200, ab208670; Abcam), RAMP1 antibody (1:200, YA1818; MedChemExpress), rabbit anti-CALCRL polyclonal antibody (1:200, abs118554; Absin), and anti-thrombospondin 1 antibody (1:200, ab85762; Abcam). After washing, the samples were incubated for 2 hours at room temperature with secondary antibody (goat anti-rabbit 488 1:500, ab150077; Abcam, or goat anti-rabbit 568 1:500, A11011; Invitrogen). Following a final wash, the samples were mounted with a medium containing DAPI. The images were captured using a confocal laser scanning microscope (Olympus FluoView 3000; Olympus, Japan). Mouse corneal tissues were harvested and homogenized under low-temperature conditions. After centrifugation, the supernatant was collected, and CGRP concentrations were measured using an ELISA kit (Cloud Clone Corp.; CEA876Mu) following the manufacturer’s instructions. For the detection of intracellular cAMP levels, cells were collected and then subjected to ultrasonication. After centrifugation, the supernatant was collected for analysis using an ELISA kit (QTJA20937; CZKEWEI). Briefly, sample or standard was added to antibody-precoated 96-well plates and incubated at 37°C for 1 to 2 hours. After washing, biotinylated detection antibody was added and incubated at 37°C for 1 hour, followed by washing and incubation with horseradish peroxidase-conjugated streptavidin for 30 minutes at 37°C in the dark. After a final wash, chromogenic reagent was added and incubated at 37°C until color development. The reaction was stopped with stop solution, and absorbance was measured at 450 nm using a microplate reader. CGRP or cAMP concentrations were calculated by comparison with the standard curve. Total RNA was extracted from corneal and cellular samples using Trizol reagent (15596018; Invitrogen). Complementary DNA (cDNA) was synthesized with the HiScript II Q Select RT SuperMix (Vazyme). Gene expression was quantified using the ABI QuantStudio 5 (ThermoFisher, USA) with iTaq Universal SYBR Green Supermix (1725124; Bio-Rad). The relative expression of target genes was normalized to GAPDH and calculated using the 2−ΔΔCt method. The primer sequences for target genes are listed in Supplementary Table S1 . To investigate the key molecules involved in corneal injury repair, we referred to previous literature 34 , 35 and performed mRNA sequencing on the cornea on day 7 after alkali burn. Total RNA was isolated from the mouse cornea, and sequencing libraries were constructed following the standard Illumina protocol (VAHTS Universal V6 RNA-seq Library Prep Kit for Illumina). Before sequencing on the Illumina NovaSeq 6000, the cDNA libraries were evaluated for quality and fragment size using an Agilent 4200 Bioanalyzer. The generated raw reads were first filtered with Seqtk and then mapped to the reference genome employing Hisat2 (version 2.0.4). Following gene fragment counting by StringTie (version 1.3.3b), the data were normalized via the TMM method. The edgeR software package was applied to identify statistically significant differentially expressed genes (DEGs), defined as those with a log2 fold-change of ≥1 and an adjusted P value < 0.05. Volcano plots and heatmaps were plotted with VolcaNoseR and HeatmapR. The transcriptomic dataset GSE255049 was downloaded from the Gene Expression Omnibus (GEO) database. The GSE255049 dataset was constructed by Lu et al. and contains RNA-seq data of BMDMs under control conditions or CGRP stimulation. 20 Differential expression analysis was performed to identify genes with statistically significant changes between experimental conditions, applying a threshold of an adjusted P value 1. The results of this analysis were visualized using a volcano plot. Furthermore, a heatmap was generated to display the expression patterns of the significant DEGs across all samples, using z-score normalization for row scaling to facilitate comparison. To interpret the biological functions of the identified genes, Gene Ontology (GO) enrichment analysis was conducted for the Biological Process (BP) domain. Terms with a false discovery rate (FDR) <0.05 were considered statistically significant. Data are presented as mean ± standard deviation. Comparisons between the two groups were performed using an unpaired t -test, whereas comparisons among multiple groups were conducted by ANOVA with GraphPad Prism software (version 8.0; GraphPad Software, Inc., San Diego, CA, USA). A P value of < 0.05 was considered statistically significant.

To evaluate the dynamic changes in corneal innervation following alkali burn, we applied a 2-mm diameter filter paper disk saturated with 0.5 N NaOH to the mouse central cornea for 30 seconds ( Fig. 1 A). We observed a significant reduction in corneal nerve fibers within the injured area on days 1, 3, 7, and 14 post-injury ( Figs. 1 B,  1 C). Corneal nerve regeneration was observed, with the corneal nerve area increasing from 53.1 ± 4.818% on day 1 to 84.47 ± 3.609% on day 14 (see  Fig. 1 C). Corneal nerves modulate tissue healing in corneal alkali burn. ( A ) Schematic of the corneal alkali burn model. ( B ) Whole-mount corneal nerve staining results of blank, and day 1, day 3, day 7, and day 14 after injury. Red = beta-Ⅲ tubulin. ( C ) Quantification analysis of corneal nerve area ( n = 4). ( D ) Schedule of corneal denervation, alkali burn, and clinical evaluation. ( E ) Schematic of the corneal denervation (CD) model. ( F ) Corneal nerve of CD mice. ( G ) Representative image of corneal fluorescein staining. ( H ) Percentage of injured area to total corneal surface area quantitatively analyzed by ImageJ software ( n = 6). ( I ) Schedule of topical RTX treatment, alkali burn, and clinical evaluation. ( J ) Schematic of the topical RTX treatment. ( K ) Corneal nerve of RTX-treated mice. ( L ) Representative image of corneal fluorescein staining. ( M ) Percentage of injured area to total corneal surface area quantitatively analyzed by ImageJ software ( n = 6). Data are shown as mean ± SD. ns, not significant; * P < 0.05; ** P < 0.01; *** P < 0.001. To induce CD, mice in the experimental group underwent clamping of the ciliary nerve of the trigeminal nerve. The detailed surgical protocol is summarized in  Figure 1 E, and successful induction was verified by whole-mount corneal nerve staining ( Fig. 1 F). Corneal alkali burns were induced immediately after CD or sham surgery ( Fig. 1 D). Corneal epithelial healing was assessed on days 1, 3, 5, 7, and 14 post-injury. Delayed epithelial healing was observed in the CD group compared to the sham surgery group ( Figs. 1 G,  1 H). Corneal sensory nerves were ablated by topical application of RTX ( Figs. 1 J,  1 K). Five days post-RTX treatment, mice underwent a corneal alkali burn ( Fig. 1 I). Epithelial healing was assessed on days 1, 3, 5, 7, and 14 post-injury. The results showed that sensory denervation with RTX resulted in significantly delayed corneal epithelial healing ( Figs. 1 L,  1 M). To identify the molecules involved in the nerve-mediated repair of corneal alkali burns, we performed RNA sequencing of the cornea on day 7 post-injury. The results showed an increased expression of the two CGRP receptors, RAMP3 and CALCRL ( Fig. 2 A). Quantitative PCR was performed to measure the expression of CGRP receptors (CALCRL, RAMP1, RAMP2, and RAMP3) after alkali burn. The data showed that the receptor was upregulated at different time points after the injury ( Fig. 2 B). Immunofluorescence staining showed the co-localization of CGRP with beta-III tubulin-positive nerve fibers in both the cornea and trigeminal ganglion, and CGRP-specific green fluorescence was not detected in non-neuronal tissues ( Figs. 2 C,  2 D). ELISA showed that corneal CGRP levels were significantly increased on days 1, 3, and 7 post-injury ( Fig. 2 E). Analysis of corneal whole-mounts indicated that CGRP was released from nerves and diffused into the corneal tissue after alkali burn ( Fig. 2 F). Alkali burn causes an increase in CGRP levels in the cornea. ( A ) Volcano plot of differentially expressed genes (DEGs) for the CGRP receptor in corneas: blank group versus alkali burn group. ( B ) The qPCR analysis of CGRP receptors CALCRL, RAMP1, RAMP2, and RAMP3 ( n = 4 per group). ( C, D ) Confocal images of the mouse cornea whole mounts C ( n = 4) and trigeminal ganglion section D ( n = 4) stained for corneal nerves with beta-Ⅲ tubulin ( red ) and CGRP ( green ) obtained from blank mice. ( E ) ELISA shows the level of CGRP in cornea after alkali burn ( n = 4). ( F ) Whole-mount corneal staining results of blank, and day 1, day 3, day 7, and day 14 after alkali burn. Red = beta-Ⅲ tubulin, green = CGRP ( n = 4). Data are shown as mean ± SD. ns, not significant; * P < 0.05; ** P < 0.01; *** P < 0.001. To determine the role of CGRP in corneal tissue healing after alkali burn, a corneal alkali burn model was established and topically treated with either CGRP solution or PBS, with corneal healing assessed on days 1, 2, 3, 4, 5, 7, and 14 ( Fig. 3 A). Results showed that CGRP treatment accelerated corneal epithelial wound closure relative to the PBS control ( Figs. 3 B,  3 C). Additionally, reduced corneal opacity was observed in the CGRP group on days 7 and 14 ( Figs. 3 D,  3 E). However, when we used BIBN-4096 to inhibit the CGRP receptor ( Fig. 3 F), we found that epithelial healing was significantly delayed ( Figs. 3 G,  3 H). The degree of corneal opacity observed on days 7 and 14 was similar to the control ( Figs. 3 I,  3 J). Anterior segment OCT showed an increase in central corneal thickness following topical BIBN-4096 application ( Figs. 3 K,  3 L). The histological analysis by H&E staining of the corneas revealed a more organized corneal structure and reduced inflammatory cell infiltration in the CGRP-treated group. In contrast, the BIBN-4096-treated group exhibited disrupted corneal tissue and extensive inflammatory cell infiltration ( Fig. 3 M). Topical CGRP application promotes corneal tissue repair after alkali burn. ( A ) Schedule of corneal alkali burn, topical CGRP or vehicle treatment, clinical evaluation, and tissue harvesting. ( B ) Representative images of corneal fluorescein staining following treatment with CGRP or vehicle. ( C ) Percentage of injured corneal area in CGRP- or vehicle-treated mice quantitatively analyzed by ImageJ software ( n = 6). ( D ) Representative images of cornea appearance following treatment with CGRP or vehicle. ( E ) Quantification of the clinical score of corneal opacity ( n = 6). ( F ) Schedule of corneal alkali burn, topical BIBN-4096 or vehicle treatment, clinical evaluation, and tissue harvesting. ( G ) Representative images of corneal fluorescein staining following treatment with BIBN-4096 or vehicle. ( H ) Percentage of injured corneal area in BIBN-4096 or vehicle-treated mice quantitatively analyzed by ImageJ software ( n = 6). ( I ) Representative images of cornea appearance following treatment with BIBN-4096 or vehicle. ( J ) Quantification of the clinical score of corneal opacity ( n = 6). ( K ) Representative AS-OCT images showed central corneal thickness. ( L ) Quantification of the central corneal thickness on days 7 after injury ( n = 6). ( M ) Histological analysis of the cornea on day 14 was performed using H&E staining. Black arrows indicate inflammatory cells. Data are shown as mean ± SD. ns, not significant; * P < 0.05; ** P < 0.01; *** P < 0.001. To elucidate the neuro-immune mechanisms by which CGRP regulates corneal healing, we performed RNA sequencing on corneal tissues from CGRP-treated and control mice at day 7. Transcriptomic analysis revealed altered expression of multiple immune-related genes following CGRP treatment ( Figs. 4 A,  4 B). Quantitative PCR analysis confirmed that the expression of inflammation-related genes, including IL-1β, iNOS, MCP-1, MMP3, and MMP9, was upregulated after alkali burn, whereas their expression was downregulated following CGRP treatment ( Fig. 4 C). Flow cytometry demonstrated an increased infiltration of CD45+ cells into the cornea after alkali burn. CGRP treatment led to a reduction of this CD45+ cell infiltration ( Figs. 4 D,  4 E). Furthermore, a decrease in macrophage infiltration was observed via immunofluorescence ( Figs. 4 F,  4 G), whereas flow cytometric analysis demonstrated a reduction in macrophage number accompanied by an increased proportion of the anti-inflammatory macrophage subtype following local CGRP administration ( Figs. 4 H–J). Immunofluorescence staining ( Figs. 4 K,  4 L) and flow cytometry ( Fig. 4 M) confirmed that topical CGRP treatment reduced neutrophil infiltration in the cornea. CGRP suppresses corneal tissue inflammation following alkali burn. ( A ) The volcano plot displayed the upregulated and downregulated genes in the corneas of mice from the CGRP-treated group compared with the PBS-treated group at 7 days post-injury ( n = 3). ( B ) The heatmap showed the differentially expressed genes in the corneas of mice from the CGRP-treated group compared with the PBS-treated group at 7 days post-injury ( n = 3). ( C ) The mRNA expression of the inflammatory factors IL-1β, iNOS, MCP-1, MMP3, and MMP9 was detected by qPCR on day 7 post-injury following CGRP treatment ( n = 4). ( D, E ) Representative flow cytometry plots and bar charts of CD45+ cells in the cornea after 7 days of treatment with CGRP following injury. ( F, G ) Representative micrographs and bar charts of macrophages in the cornea after 7 days treatment with topical CGRP or PBS ( n = 4). ( H ) Corneal macrophages were determined by flow cytometry ( n = 4). ( I, J ) Representative flow cytometry plots and bar charts of CD206+ macrophages in the cornea of the CGRP treatment group ( n = 4). ( K, L ) Representative micrographs and bar charts of neutrophils in the cornea after 7 days treatment with topical CGRP or PBS ( n = 4). (M) Corneal neutrophils were determined by flow cytometry ( n = 4). Data are shown as mean ± SD. ns: not significant; * P < 0.05; ** P < 0.01; *** P < 0.001. To investigate the role of macrophages in the promotion of corneal alkali burn tissue repair by CGRP, we fed mice a diet containing PLX5622 to eliminate the macrophages. Following the injury, the mice were treated topically with either CGRP solution or PBS ( Fig. 5 A). Corneal immunofluorescence staining confirmed a significant reduction in macrophages after PLX5622 administration ( Supplementary Fig. S1 ). Corneal epithelial wound healing was assessed by fluorescein staining. Compared to the control diet group, epithelial healing was significantly delayed in the PLX5622 group. Furthermore, epithelial healing showed no significant difference between CGRP- and PBS-treated mice fed with PLX5622 ( Figs. 5 B,  5 C). Similarly, corneal opacity in the CGRP-treated group was comparable to the PBS group after PLX5622 administration ( Figs. 5 D,  5 E). These results indicate that the pro-healing effect of CGRP is dependent on macrophages. To investigate the underlying mechanisms, we performed a TUNEL assay and immunofluorescence staining on the corneas. The TUNEL assay revealed a significant increase in apoptotic cells in the PLX5622 group compared to the controls ( Figs. 5 F,  5 G). Furthermore, neutrophil staining demonstrated substantial accumulation of cells with a fragmented morphology in the corneas of PLX5622-fed mice ( Figs. 5 H,  5 I). Macrophage depletion results in delayed corneal repair after injury and abolishes the therapeutic effect of CGRP. ( A ) Experimental timeline showing the schedule of PLX5622 feeding, corneal alkali burn induction, topical CGRP or vehicle treatment, clinical evaluation, and tissue harvesting. ( B ) Representative images of corneal fluorescein staining in PLX5622-fed mice topically treated with CGRP or vehicle. ( C ) Percentage of injured corneal area in PLX5622-fed mice treated with CGRP or vehicle. Quantitatively analyzed by ImageJ software ( n = 6). ( D ) Representative images of cornea appearance in PLX5622-fed mice treated with CGRP or vehicle. ( E ) Quantification of the clinical score of corneal opacity ( n = 6). ( F, G ) Representative micrographs and quantification of TUNEL+ cells in corneas 7 days after alkali burn in mice fed with PLX5622 or control diet ( n = 4). ( H, I ) Representative micrographs and quantification of neutrophils in corneas 7 days after alkali burn in mice fed with PLX5622 or control diet ( n = 4). Data are shown as mean ± SD. ns, not significant; * P < 0.05; ** P < 0.01; *** P < 0.001. To investigate the mechanism of CGRP on macrophages, bone marrow-derived macrophages and neutrophils were isolated from mice and cultured in vitro ( Figs. 6 A,  6 B). Immunofluorescence staining confirmed that both cell types express CALCRL and RAMP1, the two subunits of the CGRP receptor ( Fig. 6 C). To further understand the molecular mechanisms by which CGRP regulates macrophage function, we analyzed the dataset GSE255049 . Heatmap and volcano plot analyses demonstrated that CGRP treatment upregulated numerous macrophage genes associated with anti-inflammatory responses, efferocytosis, and cell death ( Figs. 6 D,  6 E). These genes were clustered based on their functional roles, which are color-coded in the circles next to the heatmap. GO enrichment analysis demonstrated significant enrichment of the DEGs in biological processes critical to macrophage function, including “tissue remodeling,” “macrophage differentiation,” “apoptotic cell clearance,” “regulation of inflammatory response,” and “regulation of interleukin-1 production” ( Fig. 6 F). Expression of the CGRP receptor in neutrophils and macrophages, and transcriptomic analysis of macrophages following CGRP treatment in vitro. ( A, B ) Light microscope image of bone marrow-derived macrophages and bone marrow-derived neutrophils ( n = 4). ( C ) CALCRL and RAMP1 expression was detected in bone marrow-derived macrophages and neutrophils using immunostaining. Green = CALCRL, red = RAMP1 ( n = 4). ( D ) Heatmap of selected significantly upregulated and downregulated genes depicting standardized gene expression values in CGRP-treated cells compared to PBS-treated cells. The colored circles next to the heatmap denote gene functions. ( E ) Volcano plot displaying upregulated and downregulated genes in macrophages from the CGRP-treated group versus the PBS-treated group. ( F ) GO enrichment analysis of significantly upregulated ( red ) and downregulated ( blue ) genes in CGRP-treated and saline-treated macrophages. In vitro, CGRP stimulation increased apoptosis in both macrophages and neutrophils ( Figs. 7 A,  7 C,  7 E,  7 G). Correspondingly, topical CGRP application in vivo also enhanced the apoptosis of these immune cells in the cornea ( Figs. 7 B,  7 D,  7 F,  7 H). To assess the effect of CGRP on macrophage anti-inflammatory function, we pre-treated 7-day BMDMs with CGRP for 20 minutes before stimulating them with IL-4 and IL-13 (3 ng/mL) for 48 hours. The qPCR results demonstrated that CGRP significantly upregulated CD206 expression ( Fig. 7 I) and Arg-1 ( Fig. 7 J) expression, indicating that macrophages were polarized to the M2 type. To determine CGRP's effect on efferocytosis, CGRP-pretreated BMDMs were co-cultured with apoptotic neutrophils. Analysis demonstrated a significant increase in the proportion of macrophages that engulfed apoptotic cells after CGRP treatment ( Figs. 7 K,  7 L). CGRP enhances apoptotic, anti-inflammatory, and efferocytic functions of macrophages. ( A, E ) Representative micrographs and quantification of apoptosis in macrophages treated with CGRP in vitro ( n = 4). ( B, F ) Representative micrographs and quantification of apoptotic macrophages in the cornea following 7 days of topical CGRP treatment ( n = 4). ( C, G ) Representative micrographs and quantification of apoptosis in neutrophils treated with CGRP in vitro ( n = 4). ( D, H ) Representative micrographs and quantification of apoptotic neutrophils in the cornea following 7 days of topical CGRP treatment ( n = 4). ( I, J ) CGRP treatment promotes CD206 and Arg-1 mRNA expression in macrophages ( n = 4). ( K, L ) Representative micrographs and quantification of macrophage efferocytosis of apoptotic neutrophils ( n = 4). Data are shown as mean ± SD. ns, not significant; * P < 0.05; ** P < 0.01; *** P < 0.001. CGRP binds to the CALCRL-RAMP1 heterodimeric receptor, a G protein-coupled receptor complex, which activates adenylyl cyclase (AC) to produce the second messenger cAMP, thereby exerting its biological functions ( Fig. 8 A). To investigate whether CGRP regulates macrophages via the cAMP pathway, we performed in vitro experiments using the AC inhibitor SQ22536. The results showed that intracellular cAMP levels increased in macrophages following CGRP treatment, and this increase was attenuated by co-treatment with SQ22536 ( Fig. 8 B). Compared with CGRP treatment alone, co-treatment with SQ22536 significantly reduced CGRP-induced macrophage apoptosis ( Fig. 8 C), downregulated the expression of CD206 and Arg-1 ( Figs. 8 D,  8 E), and decreased the phagocytic capacity for apoptotic neutrophils ( Fig. 8 F), indicating that CGRP regulates macrophages through the cAMP pathway. Analysis of the GSE255049 database indicated that the Thbs1 gene, which encodes the multifunctional extracellular matrix protein TSP-1, was significantly upregulated in macrophages after CGRP stimulation ( Fig. 8 G). This protein is known to regulate multiple biological processes, including tissue repair. We subsequently confirmed that TSP-1 was upregulated in CGRP-treated BMDMs, and that TSP-1 expression was downregulated in BMDMs co-treated with CGRP and SQ22536 compared with CGRP treatment alone ( Figs. 8 H,  8 J). Furthermore, an increase in the proportion of TSP-1-positive macrophages was also detected in CGRP-treated corneas ( Figs. 8 I,  8 K). After treating BMDMs with TSP-1, we observed increased apoptosis ( Fig. 8 L) and upregulated CD206 expression ( Fig. 8 M), whereas Arg-1 expression showed no significant difference ( Fig. 8 N). In addition, the phagocytic capacity of BMDMs for neutrophils was enhanced ( Fig. 8 O). This confirmed that CGRP modulates macrophage function through TSP-1. CGRP modulates macrophage function via the cAMP/TSP-1 signaling pathway ( n = 4). (A) Schematic of the cAMP signaling pathway downstream of the CGRP-CALCRL/RAMP1 axis. Activation of adenylyl cyclase (AC) elevates the intracellular level of cAMP. (B) The concentration of cAMP in cell lysates was measured by a competitive ELISA after in vitro stimulation of macrophages with CGRP or co-stimulation of CGRP with SQ22536 for 30 minutes ( n = 4). (C) Bar graph showing quantification of apoptosis in macrophages following in vitro treatment with CGRP alone or in combination with SQ22536 ( n = 4). (D, E) The expression of CD206 and Arg-1 in macrophages after treatment with CGRP or CGRP combined with SQ22536 ( n = 4). (F) Macrophage efferocytosis after CGRP or CGRP + SQ22536 treatment ( n = 4). (G) Volcano plot showing upregulation of thbs1 gene expression in CGRP-treated macrophages. (H, J) Representative micrographs and quantification of increased TSP-1 expression in macrophages following CGRP or CGRP + SQ22536 stimulation in vitro ( n = 4). (I, K) Representative micrographs and quantification of TSP-1+ macrophages in the cornea following 7 days of topical CGRP treatment ( n = 4). (L) Quantification of apoptosis in macrophages treated with TSP-1 in vitro ( n = 4). (M, N) The expression of CD206 and Arg-1 in macrophages after treatment with TSP-1 ( n = 4). (O) Macrophage efferocytosis after CGRP or CGRP + SQ22536 treatment. Data are shown as mean ± SD. ns, not significant; * P < 0.05; ** P < 0.01; *** P < 0.001.

This study reveals an important neuro-immune regenerative axis following corneal alkali burn and elucidates the complex interactions among corneal nerves, CGRP, immune cells, and the tissue repair process ( Fig. 9 ). Initially, we found that corneal nerves promote tissue repair in alkali burns by releasing CGRP. Further investigation revealed that CGRP modulates corneal macrophages and neutrophils in alkali burn. Mechanistically, CGRP regulates macrophage function through the cAMP/TSP-1 pathway, including efferocytosis, apoptosis, and anti-inflammatory polarization, thereby facilitating tissue regeneration. Schematic illustration of corneal nerve-derived CGRP facilitating alkali burn repair through coordinated actions on neutrophils and macrophages. Corneal nerves are crucial for maintaining corneal integrity. Severe corneal alkali burns induce nerve damage, resulting in impaired corneal tissue healing. 36 , 37 The corneal nerves include sensory nerves, sympathetic nerves, and parasympathetic nerves. CD mechanically eliminates all the nerve components of the cornea, whereas local treatment using RTX specifically eliminate the sensory nerves of the cornea. Although corneal alkali burns damage some nerves, we used a mouse CD model and RTX-induced denervation to simulate NK, and observed a significant delay in corneal tissue healing. These findings provide direct evidence that corneal nerve function is a positive regulator of tissue repair. Previous studies have shown that corneal nerve regeneration can significantly promote wound healing. 38 – 40 In corneal alkali burns, we observed corneal nerve regeneration after injury. Although the CD model eliminated corneal nerves, corneal nerves could regenerate after CD. 29 In our study, we observed that the mouse epithelium in the CD group was completely healed at 14 days. These results indicate that corneal nerves play a key role in promoting tissue repair. CGRP is a neurotransmitter released from nerves, which is widely present in various tissues throughout the body, 41 – 43 and its levels change under different pathological conditions. 44 , 45 A previous study indicated that elevated levels of prostaglandin E2 (PGE2) in tissues can promote CGRP release from TRPV1-positive nerves. 46 We hypothesize that in corneal alkali burn, CGRP is released through this mechanism due to the intense inflammatory response in the cornea. Previous studies have shown that in the cornea, CGRP concentrations in tears and corneal tissue increase in dry eye disease. 47 Elevated tear CGRP levels have also been observed at 1 week, 1 month, and 3 months following corneal refractive surgery, 48 whereas mechanical injury to the cornea leads to a decrease in CGRP concentration. 27 However, no studies have yet investigated changes in corneal CGRP levels after alkali burn. Our study found that CGRP concentrations increased significantly from day 1 to day 7 post-alkali burn. Interestingly, although corneal nerves are damaged in alkali burns, the remaining nerves are still capable of releasing substantial amounts of CGRP. The cornea contains various neuropeptides, such as SP, vasoactive intestinal peptide (VIP), and PACAP. 5 Our study focused exclusively on the role of CGRP in injury repair. The functions of other neuropeptides in corneal alkali burn, and which neuropeptide plays a dominant role in corneal wound healing, remain to be elucidated through further research. Inflammation is the core pathological mechanism of corneal alkali burn. 49 Following an alkali burn to the cornea, various inflammatory cells infiltrate the corneal tissue. These cells, along with resident corneal cells, release large quantities of pro-inflammatory cytokines such as IL-1β, MCP-1, and iNOS and matrix metalloproteinases (MMPs). 50 , 51 Our study demonstrated that alkali burn induced the upregulation of inflammatory factors in the cornea, and topical CGRP application suppressed their expression. Persistent inflammation leads to corneal cell apoptosis and tissue destruction. This dysregulated inflammatory response is the primary cause of corneal opacity, scar formation, and visual impairment. 52 Consequently, controlling the inflammatory response after a corneal alkali burn is crucial. Although topical corticosteroids can suppress inflammation, their long-term use carries risks of corneal melting and cataract formation. 30 Numerous studies have confirmed that stem cell therapy is an effective approach for modulating inflammation 53 , 54 ; however, its therapeutic efficacy has limitations, and standardized production and treatment protocols remain underdeveloped. 55 Our research demonstrates that CGRP, a natural component present in corneal tissue, can inhibit the infiltration of inflammatory cells into the cornea and suppress the expression of various inflammatory factors in a corneal alkali burn model, which exhibits potent anti-inflammatory capabilities. These findings suggest that CGRP could potentially be developed as a novel anti-inflammatory agent for the treatment of alkali burns. Macrophages play a dual role in corneal alkali burns through distinct subtypes: M1 macrophages drive pro-inflammatory tissue destruction, whereas M2 macrophages are crucial for anti-inflammatory repair. Therapeutic strategies focused on modulating macrophage function may lead to improved treatment outcomes. 56 Previous studies have reported regulatory effects of CGRP on macrophages: in endometriosis, CGRP impairs macrophage phagocytic capacity, promoting disease progression 57 ; in skin injuries, it enhances macrophage efferocytosis and anti-inflammatory functions to facilitate tissue healing 20 ; in infectious arthritis, CGRP suppresses antimicrobial activity of macrophages, leading to joint damage 58 ; whereas in infectious keratitis, CGRP promotes macrophage M2 polarization and suppresses corneal inflammation. 26 Consistent with previous studies, we observed that CGRP treatment increased CD206 expression in corneal macrophages, as well as CD206 and Arg-1 expression in BMDMs. As CD206 and Arg-1 are markers for M2 macrophages, its upregulation indicates enhanced anti-inflammatory and tissue-repair functions. 59 This finding is consistent with the faster corneal tissue healing and lower inflammatory levels observed in the CGRP treatment group. Notably, due to limited samples, our in vivo analysis of M2 macrophages was confined to CD206 expression without assessing other markers, which may not fully capture macrophage heterogeneity. Future studies using single-cell sequencing could further characterize corneal CD206+ macrophage subsets. We found that CGRP promotes macrophage apoptosis both in vivo and in vitro, which could explain the decreased macrophage infiltration in the cornea. A recent study by Lu et al. found that CGRP promotes apoptosis in macrophages treated with pro-inflammatory signals (TNF and IL-1), whereas macrophages treated with anti-inflammatory cytokines (IL-4, IL-10, or IL-13) were unaffected. 20 Although our in vitro experiments did not specifically examine CGRP-induced apoptosis under these defined conditions, we did observe an overall increase in macrophage apoptosis. The simultaneous increase in macrophage apoptosis and efferocytosis is not contradictory. This is because CGRP selectively promotes apoptosis of M1 macrophages while driving polarization toward the M2 phenotype, and M2 macrophages are known to possess greater efferocytic capacity than M1 cells. 60 Besides, studies using CSF1R inhibitors have demonstrated that macrophages repopulate rapidly following depletion. 61 We speculate that a similar repopulation dynamic may occur after CGRP-induced apoptosis. Importantly, given CGRP's established role in promoting M2 polarization, these newly repopulated macrophages may be predisposed toward a pro-repair phenotype, further amplifying the therapeutic effect. Future studies using animal models are warranted to investigate the repopulation dynamics of macrophages following CGRP-induced apoptosis. Neutrophils act as the first line of defense against tissue damage and may have dual functions: after alkali burns, neutrophils will infiltrate the damaged tissue to resist pathogen invasion and remove debris. 62 Due to the short half-life of the infiltrated neutrophils, the dead neutrophils will release their intranuclear substances and granular components, hindering the wound healing process. 63 Chen et al.’s research found that neutrophils will rapidly undergo pyroptosis, and these pyroptotic neutrophils will release IL-1β, inhibiting corneal healing. 23 Our results showed that CGRP promoted neutrophil apoptosis both in vivo and in vitro. However, we did not observe the adverse effects on corneal tissue healing due to the substantial neutrophil death. This may be explained by a concomitant enhancement of macrophage efferocytosis. Macrophage efferocytosis, the process by which macrophages phagocytose apoptotic cells, is a fundamental mechanism for maintaining tissue homeostasis. Studies have shown that macrophages engulfing apoptotic cells upregulate M2 markers such as CD206, 64 thereby forming a positive feedback loop that promotes inflammation resolution and tissue repair. Proper efferocytosis contributes to the resolution of inflammation. 65 We observed that CGRP treatment enhanced the ability of macrophages to engulf apoptotic neutrophils. This indicates that even though CGRP increases neutrophil apoptosis, the enhanced efferocytic activity effectively clears these dying cells and avoids tissue damage and excessive inflammation caused by neutrophil death. After using PLX5622 to eliminate macrophages, we observed a delay in corneal tissue healing, as well as the accumulation of apoptotic cells and neutrophils in the cornea. These neutrophils exhibited a fragmented morphology. We speculate that in the absence of macrophages, the clearance of dead neutrophils is impaired, leading to their accumulation in the corneal tissue and consequent tissue damage. To investigate the molecular mechanisms by which CGRP modulates macrophage function, this study focused on the cAMP/TSP-1 signaling pathway. Our findings demonstrate that CGRP treatment significantly increased intracellular cAMP levels in macrophages, leading to upregulation of TSP-1 protein. To establish causality, we used the adenylyl cyclase inhibitor SQ22536. Co-treatment with SQ22536 not only blocked CGRP-induced cAMP elevation but also significantly attenuated TSP-1 upregulation, as well as downstream macrophage functional responses including apoptosis, M2 polarization, and efferocytosis. These results suggest that the cAMP/TSP-1 pathway may play a critical role in CGRP-mediated tissue repair following corneal alkali burn. As a known agonist of G protein-coupled receptors, CGRP classically signals through adenylate cyclase activation and subsequent cAMP elevation. 66 , 67 Our results provide new evidence for the involvement of this canonical pathway in CGRP-promoted corneal healing. TSP-1, a protein known for its immunomodulatory and pro-apoptotic functions, was identified as a key downstream effector in this cascade. By mediating the effects of CGRP stimulation, TSP-1 promotes macrophage apoptosis, efferocytosis, and M2 polarization, thereby accelerating corneal tissue healing. This study has the following limitation: first, the use of different administration routes for CGRP (eye drops) and BIBN-4096 (subconjunctival injection) may have introduced bias in the observed therapeutic effects. Second, PLX5622, as a CSF1R inhibitor, depletes not only macrophages but also Langerhans cells and other myeloid populations. Future studies using more selective macrophage-depleting agents or genetic mouse models are needed to achieve precise macrophage ablation. In summary, we demonstrate that corneal nerves and CGRP may play a regulatory role in the injury repair process following corneal alkali burns. Specifically, after injury, corneal nerves release CGRP, which binds to receptors on neutrophils and macrophages. This binding promotes neutrophil apoptosis and, via the cAMP-TSP-1 pathway, enhances macrophage apoptosis, efferocytosis, and anti-inflammatory functions, thereby facilitating tissue healing. Given these mechanisms, CGRP shows promise as an innovative therapeutic target for promoting tissue repair after corneal alkali burns.