Pluripotent stem cell-derived corneal endothelial cells as an alternative to donor corneal endothelium in keratoplasty

Here, we evaluate the efficacy of cryopreserved human embryonic stem cell (hESC)-derived corneal endothelial cells (CECs) to form a functional monolayer of corneal endothelium (CE) in mammals (rabbits) and non-human primates (monkeys). We injected cryopreserved hESC-derived CECs in rabbits and monkeys either immediately after removing 8 mm of the central portion of the CE or a few days later when corneal edema developed. All clinical models developed deturgesced and clear corneas 2-3 weeks following the CEC injection and remained comparable to the CE of the untreated eye. Confocal scanning microscopy confirmed an intact structure of hexagonal/polygonal cells and immunohistochemical analysis illustrated a monolayer expressing barrier and pump function proteins in the regenerated CE. The necropsy examination confirmed no remarkable change in multiple tissues examined for teratoma formation. In conclusion, our data demonstrate the efficacy of cryopreserved hESC-derived CECs to form a functional CE on the denuded Descemet's membrane. Overall design: Next-generation RNA sequecning (RNA-Seq) of hESC-derived CECs at day 20 (non-cryopreserved) and cryopreserved hESC-derived CECs (cryopreserved for 40 days in liquid nitrogen) was performed. Briefly, three biological replicates for each of non-cryopreserved hESC-derived CECs and cryopreserved hESC-derived CECs were used for RNA-Seq library preparation. Total RNA was subjected to RNA-Seq library preparation using NEBNext Ultra II RNA Library Prep Kit for Illumina. The RNA-Seq bar-coded pooled libraries were sequenced (2?×?100?bp) using a HiSeq 2500 (rapid run mode) genome analyzer. The paired-end reads were assembled with SeqMan NGen (Ver. 12), using default parameters and aligned to the human reference genome (GRCh38.p11). The ArrayStar (Ver., 12; DNASTAR) was used for normalization, differential gene expression, and statistical analysis of mapped paired-end reads using the default parameters. FPKM was used for the normalization and quantification of gene expression.