In vitro quality screening of human corneal stromal stem cells for cell-based therapy of corneal scarring

Abstract

Purpose : Corneal scarring is a major type of corneal blindness, a leading cause of global vision loss. Conventional treatments for corneal scarring include penetrating and lamellar keratoplasties which require donor corneal tissues. Due to the global limited supply of transplantable donor corneas, cell therapy is becoming an attractive strategy. Corneal stromal stem cells (CSSC) have been reported to prevent or remediate corneal scarring, however not every donor CSSC batches show good healing effects. This study has determined that CSSC effectiveness was affected by stem cell stability and other measurable features.

Methods : The expression of human ABCG2 and nestin in donor CSSC (n=18 cultures) was determined by qPCR. Anti-inflammatory potency was examined using a pro-inflammatory macrophage-osteoclastogenesis (OC) assay. Mouse RAW264.7 cells pre-incubated with CSSC conditioned medium concentrates (CMC) (naïve versus heat-denatured) were treated with RANKL peptide (50 ng/ml) and ConA (20 μg/ml) for 5 days to induce osteoclast differentiation. The expression of mouse alkaline phosphatase 5 (ACP5), matrix metalloproteinase 9 (MMP9), and cathepsin K (CTSK) was assayed by qPCR. A novel formula integrating these parameters generated a numerical Scarring Index (SI), which was compared with the in vivo scarring outcome (% scar area) after these 18 CSSC batches were applied to the anterior stromal wound created by Algerbrush ablation in a mouse model.

Results : Among these 18 donor CSSC, the expression of ABCG2 and nestin showed a consistent trend of DCT changes (normalized to housekeeping 18S). In the OC assay, naive CMC of most CSSC suppressed all 3 osteoclast genes, when compared with denatured CMC, giving low rates of inflammation (RInflam = expression folds of naive versus denatured). CMC from remaining CSSC lacked this global suppressive effect and not all 3 genes were reduced, hence giving higher RInflam. Using a novel formula: SI = [2^△CT_ABCG2 + 2^△CT_NES]/m + 2^{ΣRI(ACP5+MMP9+CTSK)}/n (m, n are constants), CSSC with SI <10 were correlated to an in vivo ~50% reduction of % scar area in the treatment of anterior stromal wound in mice and SI >10 was ineffective in scar inhibition/suppression and should be excluded for treatment.

Conclusions : We established SI calculation for donor CSSC and it can be used as a release criterion of CSSC for treatment purposes.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.