Regulated genes and signaling pathways during the expansion of PP-containing cells (PP cells) under the initial condition (CINI)

(A) Growth curves of two samples showing exponential expansion of PP cells for 32 d.

(B) Expression of pancreatic (PDX1, NKX6-1, SOX9) as well as liver and gut markers (AFP, CDX2, respectively) at the PP stage of representative differentiations before expansion. Expression at each sample is shown as a fold difference from the average expression level. In red are shown samples that could not be expanded and in green samples that could be expanded.

(C) MDS plot representing the Euclidian distance of samples at p0 (n=4), p5 (n=3) and p10 (n=3).

(D) Enrichment plots for regulated genes (p0 vs p10) of the TGFβ signaling pathway, E2F target genes and DNA replication show a negative correlation of the TGFβ pathway but a positive correlation of E2F target genes and DNA replication with expansion.

(D-H) Transcript levels expressed in normalized RNA-Seq counts of p0 (n=4), p5 (n=3) and p10 (n=3) PP cells for genes encoding signals or receptors of the TGFβ (D), FGF (E) and PDGF (F) signaling pathways, the RA producing enzyme ALDH1A1 (G) as well as components of the NOTCH (H) signaling pathway.

Horizontal lines represent the mean ± SD. Statistical tests were one-way ANOVA using p0 as the control condition for the comparison with p ≤ 0.033 (*), p ≤ 0.002 (**), ≤ 0.0002 (***) and ≤ 0.0001 (****).

Reproducible expansion of PP cells under condition 5 (C5)

(A) Growth curves and regression analysis for PP cells expanded under C0, C1 and C5 for at least 10 passages. The doubling time (Td) of C5-expanded cells (n=7) was 2.3 days with a 95% confidence interval (CI) of 2.13-2.51 days. This was clearly increased as compared to C0-(n=2, Td=3.92 days, 95% CI = 3.22-4.98 days) and C1-expanded cells (n=2, Td=3.55 days, 95% CI = 2.88-4.62 days). The translucent shading represents the 95% CI of the growth rate at the different conditions.

(B-D) Gene expression profile of C0-, C1- and C5-expanded cells as shown by qPCR for expression of the key pancreas progenitor markers PDX1 (B), NKX6.1 (C) and SOX9 (D) during the expansion. Expression is normalized against the expression of each marker at p0.

(E-G) Representative images of immunofluorescent staining of p0 PP cells (E) as well as C5-expanded cells at p5 (F) and p10 (G) for the PP transcription factors PDX1, NKX6.1 and SOX9. (H-J) Flow cytometry analysis of p0 PP cells (H), as well as C5-expanded cells at p5 (I) and P10 (J) for PDX1, NKX6.1 and SOX9.

(K) Cumulative results of the flow cytometry analyses for PDX1+/SOX9+ and PDX1+/SOX9+/NKX6.1+ C5-expanded PP cells at p0, p5 and p10.

Horizontal lines represent the mean ± SD. Statistical tests were two-way ANOVA with Tukey’s test, using p0 as the control condition for the comparison with p ≤ 0.033 (*), p ≤ 0.002 (**), ≤ 0.0002 (***) and ≤ 0.0001 (****). Scale bar corresponds to 50 μm.

Expansion of PP cells promotes primarily their proliferation rather than their survival.

(A, B) Histogram plots showing the % of PP cells that had incorporated EdU during expansion under CINI (A) and C5 (B).

(C) Summary of flow cytometry data comparing proliferation, measured by EdU incorporation (n=4), and cell death, measured by Annexin V/7-AAD staining (n=6), of PP cells expanded under CINI and C5.

Horizontal lines represent mean ± SD. Means were compared with multiple t-tests and significance is p ≤ 0.033 (*), p ≤ 0.0021 (**), p ≤ 0.0002 (***) or p ≤ 0.0001 (****).

Reproducible expansion under condition 6 promotes PP identity

(A) Growth curves and regression analysis for PP cells expanded under C5 and C6 for ten passages. The regression line for C5 showed a doubling time of 2.3 d (n=7) as compared to 2.5 d (n=10) for C6.

(B-D) Representative images of immunofluorescent staining of p0 PP cells (B) as well as C6-expanded cells at p5 (C) and p10 (D) for the PP transcription factors PDX1, NKX6.1 and SOX9. (E-G) Flow cytometry analysis of non-expanded p0 PP cells (E) and C6-expanded cells at p5 (F) and p10 for PDX1+/SOX9+/ NKX6.1+ cells (G).

(H, I) Cumulative results of the flow cytometry analyses for PDX1+/SOX9+ and PDX1+/SOX9+/NKX6.1+ C6-expanded cells at p0, p5 and p10 (H) and comparison of the % of C5- and C6-expanded PDX1+/SOX9+/NKX6.1+ cells at p5 and p10 (I).

(J) Survival rates of PP cells before freezing and after thawing at p0 or during expansion.

Horizontal lines represent the mean ± SD. Statistical tests were two-way ANOVA with Tukey’s test, using p0 as the control condition for the comparison with p ≤ 0.033 (*), p ≤ 0.002 (**), ≤ 0.0002 (***) and ≤ 0.0001 (****). Scale bar corresponds to 50 μm.

Expansion under C6 stabilizes PP cell identity by repressing differentiation and alternative cell fates

(A, B) Correlation analyses of the transcriptome profiles of non-expanded (p0) and p10 expanded PP cells (A) and the transcriptome profiles of p5 and p10 expanded cells (B). The numbers of upregulated and downregulated genes (normalized counts ≥ 200 and 0.5 ≥ FC ≥ 2) are shown in red and blue, respectively and r is the correlation coefficient.

(C) Mostly affected biological processes between p0 and p10.

(D) PCA of feeder expanded cells and corresponding p0 cells (shades of green), fibronectin (FN) expanded cells and corresponding p0 cells (shades of blue), as well as vitronectin-N (VTN-N) expanded cells and corresponding p0 cells (shades of red). Darker shades correspond to earlier passages.

(E) Comparative expression levels of GP2 in normalized RNA-Seq counts.

(F) GP2 immunofluorescence of p12 expanded PP cells and FC of p10 ePP cells.

(G-J) Comparative expression levels of progenitor (G), MPC (H), endocrine (I) and liver (J) markers in normalized RNA-Seq counts of ePP cells and their corresponding p0 cells expanded on feeders (green), FN (blue) and VTN-N (red).

Expansion of H9-derived and CRTD1-derived PP cells under C6

(A, B) Growth curve and regression analysis of the expansion of H9-derived PP cells (A) and CRTD1-derived PP cells (B).

(C, D) Flow cytometry analysis for PDX1+/SOX9+ and PDX1+/SOX9+/NKX6.1+ cells during the expansion under C6 of H9-derived PP cells (C) and CRTD1-derived PP cells (D) at p0, p5 and p10.

Differentiation of ePP cells into SC-islets containing functional β-cells.

(A, B) Immunofluorescence analysis of SC-islets derived from p0 PP cells (dPP) or expanded PP cells for at least ten passages (ePP) for INS and GCG expression (A) or INS and SST expression (B).

(C) Percentages of INS+, INS+/GCG+ as well as GCG+ cells in SC-islets derived from dPP or ePP cells as determined by flow cytometry.

(D) Expression levels of NKX6-1, PDX1 and SLC30A8 as determined by qPCR and expressed as fold induction relative to expression levels in hPS cells.

(E) Secretion of C-peptide following sequential stimulation by 16.7 mM glucose and 16.7 mM glucose / 30 mM KCl (30 mM KCl) after exposure in basal conditions with 2.8 mM glucose. Stimulation index is the ratio of secretion under these conditions to secretion in basal conditions.

Horizontal lines represent the mean ± SD. Statistical tests were two-way ANOVA with Tukey’s test, using p0 as the control condition for the comparison with p ≤ 0.033 (*), p ≤ 0.002 (**), ≤ 0.0002 (***) and ≤ 0.0001 (****). Scale bar corresponds to 50 μm. Scale bar corresponds to 100 μm (E, F).