Abstract
Purpose :
Management of limbal stem cell deficiency (LSCD) is a clinical challenge and there is no established pre-clinical model representing different disease severity defined in man. This study aims to develop an easy-to-create, reproducible, and reliable mouse model of graded LSCD.
Methods :
To achieve mild, moderate, or severe LSCD, filter paper rings with a variety of central angles (90, 180, or 270 degrees) were utilized to deliver alkali burn to different sizes of limbal area (1, 2, or 3 quarters). Corneal opacity, corneal epithelial defects, and central corneal thickness (CCT) were evaluated on days 0, 2, 4, 6, and 8 after injury. Histopathological examinations, corneal stem cell (P63α) and nerve (β III tubulin) staining, and corneal protein level of substance P (SP) by ELISA were performed on day 8.
Results :
The 3-quarter Alkali Burn group showed significantly higher corneal opacity score than 2- or 1-quarter Burn groups on day 8 post-injury (P < 0.0001), and exhibited ~60% wound healing which was significantly lower than >75% wound healing in both 2- and 1-quarter Burn groups (P < 0.05 and < 0.01, respectively). CCT increased with the severity of injury, with the 3-quarter Burn group 50% and 75% thicker than 2- and 1-quarter Burn groups, respectively (P < 0.001 and < 0.0001). As injury severity increased, P63α-positive cells gradually decreased with few observed in the 3-quarter Burn group, along with reduced epithelial layers. Furthermore, the corneal nerves became progressively sparser with increased injury severity, as evidenced by ~25%, ~50%, and ~75% loss of subbasal nerves in 1-, 2-, and 3-quarter Burn groups, respectively (P < 0.01, < 0.0001 and < 0.0001). Consistently, the SP levels in the injured corneas were minimally or moderately decreased in the 1- or 2-quarter Burn group, but significantly reduced in the 3-quarter Burn group (P < 0.05).
Conclusions :
Our data demonstrate the successful development of mild, moderate, and severe LSCD of clinical signs and pathological manifestations in the injured animals that simulate the three stages of clinical LSCD, and thus provide new insights into distinct pathological features underlying different grades of LSCD and serve as a new tool for further exploring the disease mechanisms and developing new effective therapeutics for repairing damaged LSC.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.