Topical application of calcitonin gene-related peptide as a regenerative, antifibrotic, and immunomodulatory therapy for corneal injury

Calcitonin gene-related peptide (CGRP) is a multifunctional neuropeptide abundantly expressed by corneal nerves. Using a murine model of corneal mechanical injury, we found CGRP levels in the cornea to be significantly reduced after injury. Topical application of CGRP as an eye drop three times daily accelerates corneal epithelial wound closure, reduces corneal opacification, and prevents corneal edema after injury in vivo. We then used a series of in vitro and in vivo techniques to investigate the mechanisms underlying CGRP’s functions. CGRP promotes corneal epithelial cell migration, proliferation, and the secretion of laminin. It reduces TGF-β1 signaling and prevents TGF-β1-mediated stromal fibroblast activation and tissue fibrosis. CGRP reduces corneal endothelial cell apoptosis and death, preserves cell density and morphology, and promotes their pump function, thus reducing edema. Lastly, CGRP reduces neutrophil infiltration, macrophage maturation, and the production of inflammatory cytokines in the cornea. Taken together, our results show that corneal nerve-derived CGRP plays a cyto-protective, pro-regenerative, anti-fibrotic, and anti-inflammatory role in corneal wound healing. Given that current treatment options for corneal injury and opacity are scarce, CGRP has significant therapeutic potential in this area of unmet medical needs. In addition, our results highlight the critical role of sensory nerves in ocular surface homeostasis and injury repair.

pump function, thus reducing edema. Lastly, CGRP reduces neutrophil infiltration, macrophage 23 maturation, and the production of inflammatory cytokines in the cornea. Taken together, our results 24 show that corneal nerve-derived CGRP plays a cyto-protective, pro-regenerative, anti-fibrotic, and 25 anti-inflammatory role in corneal wound healing. Given that current treatment options for corneal 26 injury and opacity are scarce, CGRP has significant therapeutic potential in this area of unmet 27 medical needs. In addition, our results highlight the critical role of sensory nerves in ocular surface 28 homeostasis and injury repair.  It has been reported that CGRP levels in the tear film are changed after corneal surgeries 73 and trauma 32,33 , but its level in the cornea after injury has not been examined. More importantly, 74 while CGRP has been shown to promote corneal re-epithelialization in vitro in a single study 34 ,

81
Mechanical injury causes a decrease in CGRP levels in the cornea. 82 A 2mm-wide mechanical injury was induced at the center of the cornea by removing the epithelium 83 and superficial stroma (Fig. 1A). We observed that the CGRP levels in the cornea were 84 significantly lower on days 1, 3 and 7 post-injury (Fig. 1B). We also assessed levels of CGRP 85 receptors CLR, RAMP1, and RAMP2 in the cornea and found a significant upregulation in 86 RAMP1 gene expression level, which peaked at day 3 post-injury and returned to pre-injury level 87 by day 7 (Fig. 1C). The expression levels of CLR and RAMP2 were unchanged post-injury.

89
Topical application of CGRP accelerates corneal epithelial wound closure and reduces 90 corneal opacity. 91 We applied CGRP (5l of a 50M CGRP solution diluted in PBS as an eye drop) topically three 92 times daily immediately after injury for 14 days (Fig. 2A). The control animals were treated with 93 an equal volume of PBS eye drops. Corneal epithelial wound closure was highlighted with 94 fluorescein staining (Fig. 2B&C) CGRP decreases corneal thickness, scar formation and endothelial cell loss after injury. 104 In addition to slit lamp photography, the animals underwent live imaging with optical coherence 105 tomography (OCT) and confocal microscopy to visualize corneal microstructure in vivo. Shown 106 in Fig. 3A, anterior segment-OCT demonstrates the normal mouse eye anatomy. In the control 107 eyes post-injury, we observed hyperreflectivity of the corneal stroma, consistent with increased 108 corneal opacity on slit lamp exam; in addition, the central corneal thickness (CCT) increased from 109 90.5±1.5 at baseline to 159.6±16.8 μm on day 14 (Fig. 3A&B). The CGRP-treated corneas showed 110 lower CCT compared to the PBS-treated injured mice on days 7 (100.5±8.5 vs 166.8±11.7μm, 111 p<0.0001) and 14 (97.6±6.7μm vs 159.6±16.8μm, p=0.0008) (Fig. 3B). 112 In vivo confocal microcopy (IVCM) reveals cellular structure of the cornea; and naïve animals 113 have regular and densely packed polygonal epithelial and endothelial cells, in addition to uniform 114 reflectivity in the stroma (Fig. 3C). The control corneas demonstrated reduced uniformity and 115 density of corneal epithelial and stromal cells, as well as hyperreflectivity and dense scars in the 116 stroma, on day 14 post-injury. The CGRP-treated corneas showed comparable micro-anatomy to 117 the uninjured corneas, except mild scar formation in the stroma (Fig. 3C). Objective assessment 118 of the stromal hyperreflectivity (Fig. 3D) confirmed these observations and the depth of the scar 119 in the CGRP group is less deep compared with the controls (Fig. 3E). The corneal endothelial cell 120 density and morphology (demonstrated by coefficient of variation of cell size and the percentage 121 of normal hexagonal cells) were better maintained in the CGRP-treated eyes, compared to the 122 controls (Fig. 3F, G, and SI Appendix, Fig. S1A). 123 Similarly, histological assessment with H&E staining demonstrated that mechanical injury led to 124 overall thickening of the cornea, thinner epithelium, separation between the epithelium and 125 anterior stroma, inflammatory cell infiltration, disorganized stroma, and attenuated endothelium 126 in the control animals. On the contrary, the corneal structure and integrity were much better 127 maintained in the CGRP-treated mice (Fig. 3H). 128 To delineate the mechanisms underlying the regenerative function of CGRP at cellular and 129 molecular levels, we next sought to determine its effects on corneal epithelial, stromal, endothelial,   CGRP preserves corneal endothelial cell density and function in vitro and in vivo. 168 We cultured human corneal endothelial cells (hCEnC) in the presence of pro-inflammatory 169 cytokine TNF-α to mimic the post-injury inflammatory milieu in the corneal tissue in vivo.

170
Apoptosis and cell death of hCEnC were analyzed by measuring annexin V and propidium iodide 171 (PI) levels via flowcytometry, respectively. We observed that TNF-α resulted in a significant 172 increase in apoptotic and dead cells and that CGRP treatment reversed their levels to baseline (Fig.   173   6A&B). Consistently, we found that CGRP treatment significantly reduced TNF-α-induced 174 activation of Caspase 3 (Cas3) and Bax, two genes critical for apoptosis (Fig. 6C). The most 175 critical function of the CEnC lies in its Na + /K + -ATPase pumps. These pumps are located on the  CGRP dampens tissue inflammation after injury. 195 As CGRP has been shown to modulate the innate immune response 40-42 , we sought to determine 196 its effect on corneal inflammation in vivo. On day 3 post-injury, there was an increased infiltration 197 of CD45 + cells in the cornea in the injured PBS-treated controls (2.13±0.3% of all corneal cells), 198 compared to the uninjured naïve corneas (0.09±0.02%). Topical CGRP treatment reduced CD45 + 199 cell infiltration to 1.22±0.19% (Fig. 7A& B). Furthermore, the increased expression of pro- In this research report, we evaluated the effect of CGRP on corneal wound healing 216 following mechanical injury (Fig. 8). Corneal injury leads to nerve damage and depletion of 217 neuropeptides including CGRP. Topical application of CGRP as an eye drop accelerates corneal 218 epithelial closure, preserves corneal transparency, and prevents scar formation and edema.  with an initial increase during the healing phase (on day 2) followed by decline on day 7 32,33 . To 235 date, there is no report on the level of CGRP in the cornea after injury or surgery. We found that 236 removal of the epithelial and upper stromal layer of cornea leads to a marked decrease in CGRP 237 concentration in the cornea up to 7 days post-injury. Since CGRP is exclusively expressed by the 238 nerves in the cornea 18 , we attribute its reduced level to the loss of CGRP expressing corneal 239 nerves. Notably, we also observed an upregulation of CGRP receptor component RAMP1 in the 240 cornea during the same period (Fig. 1C). We speculate that this is due to either the infiltration of in EBM integrity (Fig. 4). Similarly in vivo CGRP treatment promotes laminin expression, thereby 265 restoring the EBM integrity and interrupting the TGF-β1 induced differentiation of fibroblasts.

266
Consistent with this, we observed much lower TGF-β1 levels in the CGRP-treated corneas in vivo.  The expression of CXCL1, which plays a predominant role in neutrophils recruitment, is 283 decreased by CGRP treatment in vivo, leading to reduced neutrophils infiltration (Fig. 7). These

307
In conclusion, corneal wound healing is a complex process that requires an integrated 308 response to restore its transparency. We found that topical treatment with CGRP promotes tissue 309 repair and limits injury-induced corneal opacity by enhancing epithelial cell proliferation and 310 migration, restoring epithelial basement membrane integrity, suppressing TGF-1-mediated tissue 311 fibrosis, decreasing neutrophil infiltration and macrophage activation, inhibiting expression of pro-312 inflammatory mediators, and preserves corneal endothelial cell density and function (Fig.8).

313
Collectively this proof-of-concept murine study demonstrates the efficacy of CGRP in treating   Corneal mechanical injury model 326 We used a well-established murine model of corneal mechanical injury as described previously 67 .

327
Briefly, the central area of the cornea was marked with a 2-mm trephine. The epithelial layer and 328 the superficial stroma were removed using a hand-held Algerbrush-II (Alger Inc., Lago Vista, TX).

329
The injured corneas were washed with sterile phosphate-buffered saline (PBS) to remove any 330 tissue debris. Post-injury induction, mice were randomly divided into 2 groups; the first group was    Germany) was used to perform RT-PCR cycle. The same protocol was followed with requisite 362 modifications for extracting RNA from mouse corneal tissues and cultured cells. Please see SI 363 Appendix, Table S1 for the detailed list of the primers used for RT-PCR experiments. All assays 364 were performed in duplicates, and data were normalized to the house keeping gene  The cells were fixed using 4% paraformaldehyde (PFA) for 10 minutes, followed by PBS washing  To assess the effect of CGRP on ERK signaling in CEC, the cells were cultured in a GFF medium 391 supplemented with 1000 nM CGRP for up to 3 hours. The cells were fixed, permeabilized and 392 stained as described above with ERK1/2 antibody. All antibodies used in this study are reported 393 in SI Appendix, Table S2. (Thermofisher Scientific Inc, Grand Island NY). Finally, the images were analyzed with ImageJ.

409
Please see SI Appendix, Table S2 for the full list of the antibodies.       Human corneal endothelial cells (CEnC) were cultured and exposed to 20ng/ml TNF-α for 24 hours. A. The representative flow cytometry plots evaluating CEnC stained with Annexin V (apoptotic cells) and Propidium Iodide (PI, dead cells) show that supplementing the media with 1M CGRP significantly reduced TNF-α -mediated CEnC loss. B. Reduced frequency of apoptotic and dead CEnC by CGRP. C. TNF-α-mediated increase in the apoptotic genes caspase 3 (Cas3) and Bax was decreased by CGRP treatment. D. CGRP induces higher expression of Na + /K + ATPase pumps compared to controls assessed with flow cytometry (B-D, n=3 per group). E. Mouse corneas were collected on Day 14 post-injury and CGRP treatment led to preserved zonula occludens-1 (ZO-1) and Na + /K + ATPase staining, compared to the PBS-treated controls (scale = 20μm). Analysis of immunohistochemical images showed higher endothelial cell density (F) and   Injury leads to nerve damage and a decrease in CGRP level in the cornea. Topical application of CGRP promotes corneal epithelial cell regeneration and res tores the epithelial basement membrane, thus reducing the release of pro-inflammatory and pro-fibrotic mediators including TNF-α, TGF-β, IL-1, and CXCL1 into the stroma. This leads to reduced keratocyte activation and stromal fibrosis. In addition, CGRP reduces neutrophil infiltration, macrophage maturation, and the production of inflammatory cytokines. It reduces corneal endothelial cell apopto sis and maintains its pump function. Clinically, topical application of CGRP as an eye drop accelerates epithelial clo sure, preserves transparency, and prevents scar formation and edema after corneal injury.