Collagen Mimetic Peptides Promote Corneal Epithelial Cell Regeneration

Abstract

The cornea of the eye is at risk for injury through constant exposure to the extraocular environment. A highly collagenous structure, the cornea contains several different types distributed across multiple layers. The anterior-most layer contains non-keratinized epithelial cells that serve as a barrier to environmental, microbial, and other insults. Renewal and migration of basal epithelial cells from the limbus involve critical interactions between secreted basement membranes, composed primarily of type IV collagen, and underlying Bowman's and stromal layers, which contain primarily type I collagen. This process is challenged in many diseases and conditions that insult the ocular surface and damage underlying collagen. We investigated the capacity of a collagen mimetic peptide (CMP), representing a fraction of a single strand of the damaged triple helix human type I collagen, to promote epithelial healing following an acute corneal wound. In vitro, the collagen mimetic peptide promoted the realignment of collagen damaged by enzymic digestion. In an in vivo mouse model, topical application of a CMP-containing formulation following a 360° lamellar keratectomy targeting the corneal epithelial layer accelerated wound closure during a 24 h period, compared to vehicle. We found that the CMP increased adherence of the basal epithelium to the underlying substrate and enhanced density of epithelial cells, while reducing variability in the regenerating layer. These results suggest that CMPs may represent a novel therapeutic to heal corneal tissue by repairing underlying collagen in conditions that damage the ocular surface.

Keywords: anti-inflammation; collagen mimetic peptides; collagen molecular activity; collagen reparative; corneal collagen damage; ocular surface disease; regeneration.

FIGURE 1

CMP realigns collagen fibers damaged by collagenase in vitro. (A) Edge-enhanced differential interference contrast images of applied type I collagen damaged by collagenase and treated with a vehicle (top) or CMP (bottom). Example vectors (dashed lines) assigned by algorithm show a higher degree of parallelism in CMP-treated preparations. (B) Vectors in preparations treated with CMP have a higher average tendency for parallel alignment compared to a vehicle (*: p < 0.001). Data: mean ± SEM; n = 7.

FIGURE 2

CMP promotes healing of corneal epithelium in mice. (A) Representative images showing residual corneal epithelial wounds visualized by fluorescein staining in vivo at the time of the initial wound (0 h) and 16 and 24 h later. Treatment with CMP (25 nM shown) appeared to accelerate the regeneration of the epithelium, as seen by smaller wound areas at 16 and 24 h. (B) Representative differential interference contrast images of H&E-stained histological sections through naïve and wounded cornea either left unhealed (by immediate sacrifice, arrows) or treated with either a vehicle or CMP and sacrificed 24 h later.

FIGURE 3

CMP accelerates the rate of epithelium wound closure. (A) For vehicle-treated eyes, the ratio of residual wound area at 24 h to initial (0 h) size decreases with a diminishing ratio of the area at 16 h to initial. The slope of the best-fitting regression line differs significantly from 0 (*, p = 0.03). For CMP-treated eyes (25 nM), accelerated closure clusters both ratios at smaller values, yielding an insignificant regression (p = 0.13). The cluster of ratios for a vehicle vs. CMP cohorts differed significantly (dashed lines, p < 0.001), as shown by multivariable analysis of variance to compare the means (mean ± SEM) (B, inset). Relative wound size calculated as the ratio of residual wound area at 16 h to initial and at 24 compared to 16 h for vehicle- and CMP-treated eyes (mean ± SEM). Wound area diminished significantly between 24 and 16 h compared to the initial 16 h period for CMP eyes (p = 0.03) but not for a vehicle (p = 0.10). Importantly, initial wound size for CMP-treated eyes (1.41 ± 0.13 mm) did not differ from vehicle-treated eyes (1.54 ± 0.17 mm, p = 0.38; n = 7 for each cohort).

FIGURE 4

CMP enhances the structure of healing corneal epithelium. Representative differential interference contrast images of histological sections with hematoxylin and eosin staining through regenerated epithelium (A, B) and the proliferative edge (C) 24 h following induced corneal injury. While vehicle-treated eyes demonstrated frequent gaps beneath the epithelium (A, left; dashed line), treatment with CMP enhanced adherence of the basal layer (brackets) to the underlying anterior stromal surface (arrows) for both 25 (A, right) and 250 nM (B, at higher magnification) concentrations of CMP. At the proliferative edge of the wound (C), compared to a vehicle, CMP increases the number of new epithelial cells adhering to the corneal stroma (circles). (D) Length of contiguous segments of adherence between basal epithelium and underlying stromal layer calculated from random samples of histological sections through cornea 24 h following epithelial removal (n = 6–11 each). Compared to naïve eyes, vehicle-treated corneas demonstrated a 36% reduction in the average length of adherent surface (#: p < 0.001). Treatment with CMP significantly increased adherence compared to a vehicle (*: p ≤ 0.02), comparable to adherence in naïve corneas (p ≥ 0.37). (E) The number of epithelial cell nuclei in basal layer quantified in random samples (n = 29–100 each). Treatment with CMP increased the number of cells per sample significantly compared to a vehicle (*: p ≤ 0.04); the number of cells following 250 nM CMP treatment exceeded naïve (#: p = 0.008) (F). Deviation in the number of epithelial nuclei in each random sample from (A) compared to the mean number for each cohort. The vehicle group was significantly more variable than naïve (#: p < 0.001); treatment with 250 nM CMP reduced deviation by 54% compared to a vehicle (*: p < 0.001) but remained higher than naïve (p = 0.04).