Skip to main content
. 2021 Jun 7;12:3330. doi: 10.1038/s41467-021-23525-x

Fig. 1. Efficiently generating pancreatic progenitor (PP) 3D clusters.

Fig. 1

a A scheme showing the generation of the NKX6.1-NLS-GFP reporter hPSC line. Using a LoxP-flanked Puromycin (Puro) selection, a 2A-NLS-GFP was knocked in and fused in frame to the end of endogenous NKX6.1 coding sequence. The Puro cassette was later removed by CRE excised. The cells will show nucleus-localized GFP expression once endogenous NKX6.1 gene is activated. b Representative images of GFP and NKX6.1 immunofluorescence in NKX6.1-NLS-GFP hPSC-derived PPs showing the fidelity of the GFP reporter. c Representative images of GFP and nucleus for PPs derived from NKX6.1-NLS-GFP hPSCs using our improved protocol, showed the high induction efficiency of PPs. d Schematic graph of the strategy for assembling PPs into 3D clusters. The PPs were dissociated and re-aggregated by centrifuging in V-bottom 96-well plate to form cell pellets and then form 3D clusters after incubation. e Morphology of PP 3D clusters maintained on air–liquid interface. f Enlarged image of NKX6.1-NLS-GFP hPSC-derived PP 3D clusters showed the uniform expression of GFP (left panel); aggregated undifferentiated NKX6.1-NLS-GFP hPSCs showed no GFP expression (right panel). g Cell cytometry analysis showed the high percentage of PPs (NKX6.1+/PDX1+) existing in H1 cell-derived PP 3D clusters (n = 3). Scale bar, 100 µm. Images of b, c, e, and f are representative of three independent experiments, respectively.