Estimating residual undifferentiated cells in human chemically induced pluripotent stem cell derived islets using lncRNA as biomarkers

Abstract

Abstract Human pluripotent stem cells (hPSCs) can generate insulin-producing beta cells for diabetes treatment, but residual undifferentiated cells may cause tumors. We developed a highly sensitive assay to detect these cells in islet cells derived from human chemically induced pluripotent stem cells (hCiPSCs), which are transgene-free and safer. We used RNA-seq data to find protein-coding and non-coding RNAs that were only expressed in hCiPSCs, not in islet cells. We confirmed these biomarkers by RT-qPCR and ddPCR. We chose long non-coding RNA (lncRNA) markers, which performed better than protein-coding RNA markers. We found that LNCPRESS2, LINC00678 and LOC105370482 could detect 1, 1 and 3 hCiPSCs in 106 islet cells by ddPCR, respectively. We tested our method on several hCiPSC lines, which could quantify 0.0001% undifferentiated cell in 106 islet cells by targeting hCiPSCs-specific lncRNA transcripts, ensuring the safety and quality of hCiPSC-derived islet cells for clinical use.