Abstract
Purpose :
Human pluripotent stem cell (hPSC)-based corneal cell therapies are in demand to reduce dependence on donor tissue and to increase patients’ access to curative treatments. We carried out serial single-cell RNA-sequencing (scRNAseq) to obtain in-depth knowledge of heterogenous populations arising from our feeder-free differentiation of hPSCs towards limbal stem cells (LSCs). Deep-phenotyping of cell heterogeneity helps to further define relevant hPSC-LSC subpopulations and may serve as a data-driven strategy to facilitate their progression towards clinical translation.
Methods :
Three genetically independent hPSC lines were subjected to corneal differentiation with a previously established protocol. scRNAseq was performed on samples collected from undifferentiated hPSCs (D0) and differentiating cells at day 10 (D10) and 24 (D24), using a modified SORT-seq-protocol. Raw sequencing data were preprocessed, followed by quality control filtering, yielding a dataset of 4737 high-quality cells. Standard analyses, including dimensionality reduction, Louvain clustering and detection of high variable genes, were utilized to identify cell clusters and differentially expressed marker genes between the clusters.
Results :
scRNAseq analysis identified eight distinct clusters which were annotated according to their temporal and transcriptomic identity. The identified main clusters confirmed our previously published findings, presenting a uniform cluster of undifferentiated hPSCs at D0, a transient epithelial cluster at D10, and an LSC-like cluster at D24. Additionally, among cells collected on different days, we detected heterogeneous cell populations, consisting of one transient and two persisting mesodermal clusters as well as persisting PSC-like and further differentiated LSC-like cells. These off-target populations were more pronounced within the induced versus embryonic lines. The identified cell populations were confirmed via detection of cluster-specific markers in immunofluorescence and flow cytometry.
Conclusions :
The molecular deep-phenotyping by scRNAseq provided further insight to the hPSC-LSC population heterogeneity. Further analyses on the cell identity and the drivers of cell heterogeneity act as a data-driven strategy for improving both the general and cell line-specific differentiation protocols.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.