Efficient induction of pancreatic alpha cells from human induced pluripotent stem cells by controlling the timing for BMP antagonism and activation of retinoic acid signaling

Abstract

Diabetes mellitus is caused by breakdown of blood glucose homeostasis, which is maintained by an exquisite balance between insulin and glucagon produced respectively by pancreatic beta cells and alpha cells. However, little is known about the mechanism of inducing glucagon secretion from human alpha cells. Many methods for generating pancreatic beta cells from human pluripotent stem cells (hPSCs) have been reported, but only two papers have reported generation of pancreatic alpha cells from hPSCs. Because NKX6.1 has been suggested as a very important gene for determining cell fate between pancreatic beta and alpha cells, we searched for the factors affecting expression of NKX6.1 in our beta cell differentiation protocols. We found that BMP antagonism and activation of retinoic acid signaling at stage 2 (from definitive endoderm to primitive gut tube) effectively suppressed NKX6.1 expression at later stages. Using two different hPSCs lines, treatment with BMP signaling inhibitor (LDN193189) and retinoic acid agonist (EC23) at Stage 2 reduced NKX6.1 expression and allowed differentiation of almost all cells into pancreatic alpha cells in vivo after transplantation under a kidney capsule. Our study demonstrated that the cell fate of pancreatic cells can be controlled by adjusting the expression level of NKX6.1 with proper timing of BMP antagonism and activation of retinoic acid signaling during the pancreatic differentiation process. Our method is useful for efficient induction of pancreatic alpha cells from hPSCs.

Fig 1. Scheme of the differentiation process and effects of LDN193189 and EC23.

(A) Scheme of the stepby-step differentiation protocol. (B, C) Analysis of NKX6.1 mRNA expression in cells treated with LDN193189 or EC23 by qPCR B; 4M, C; 15M63. N = 2 each. **P<0.01, ***P<0.001.(D, E) Analysis of NKX6.1 protein expression in cells treated with LDN193189 or EC23 by immunocytochemistry. D; 4M, E; 15M63.

Fig 2. Effects of LDN193189 or EC23 or both LDN193189 + EC23 on expressions of NKX6.1, insulin and glucagon.

(A, B) Analysis of mRNA expression in cells treated with LDN or EC or both LDN193189 + EC23 by qPCR. A; 4M, B; 15M63, N = 5 each. **P<0.01, ***P<0.001. ns: not significant. (C, D) Analysis of protein expression in cells treated with LDN19319 + EC23 by immunocytochemistry. C; 4M, D; 15M63. In the case of 4M, positive staining rates were 75.9 ± 8.3% and 4.8 ± 3.3% for NKX6.1 (P<0.001), 38.8 ± 2.6% and 14.8 ± 2.2% for C-peptide (P<0.001), 4.8 ± 3.3% and 33.4 ± 7.7% for GCG (P<0.001), in control and LDN + EC cells respectively. In the case of 15M63, positive staining rates were 46.0 ± 5.0% and 2.8 ± 0.7% for NKX6.1 (P<0.001), 25.9 ± 1.5% and 3.5 ± 0.6% for C-peptide (P<0.001), 14.8 ± 1.3% and 14.8 ± 1.3% for GCG (P>0.5), in control and LDN + EC cells respectively. N = 3 each.

Fig 3

Temporal gene expression patterns analyzed by qPCR (A, B) Temporal expression pattern of NKX6.1. A; 4M, B; 15M63 (C, D) Temporal expression pattern of insulin. C; 4M, D; 15M63 (E, F) Temporal expression pattern of glucagon. E; 4M, F: 15M63 (G, H) Temporal expression pattern of ARX. G; 4M, H: 15M63, Experiments were performed three times. *P<0.05, **P<0.01, ***P<0.001. ns: not significant.

Fig 4

Hormone secretion experiments (A, B) Analysis of insulin secretion from cells stimulated by low or high glucose. A; 4M, B; 15M63 (C, D) Analysis of glucagon secretion from the cells stimulated by low or high glucose. C; 4M, D; 15M63, Experiments were performed three times. *P<0.05, **P<0.01, ***P<0.001. ns: not significant.

Fig 5

Scheme of long-term maintenance of LDN + EC cells and GCG secretion assay (A) LDN+ EC cells were transplanted under the kidney capsule and retrieved at 3 or 6 weeks after transplantation. (B) LDN + EC cells embedded in alginate fiber were implanted into the peritoneal cavity and retrieved at the same timepoint. (C-F) Analysis of glucagon secretion from the cells stimulated with 1- and 6-mM glucose with/without arginine. (C, E) cells transplanted under the kidney capsule. D; 4M, E; 15M63 (D, F) cells embedded in alginate fiber implanted into the peritoneal cavity. G; 4M, E; 15M63 (G, K) Immunochemistry for LDN + EC cells transplanted under the kidney capsule. G; 4M (4 weeks), K; 15M63; (6weeks). (H, L) Immunochemistry for LDN + EC cells embedded alginate fiber implanted into the peritoneal cavity. H; 4M (4 weeks), L; 15M63 (4 weeks). Red; insulin c-peptide, green; glucagon. The rates of glucagon positive cells (88.3 ± 2.0%) and C-peptide positive cells (5.4 ± 0.04%) in alginate fiber for 4M. The rates of glucagon positive cells (91.2 ± 3.4%) and C-peptide positive cells (3.2 ± 0.5%) in alginate fiber for 15M63. (I, M) HE staining for LDN + EC cells transplanted under the kidney capsule. I; 4M (4 weeks), M; 15M63; (6weeks). (J, N) HE staining for LDN + EC cells embedded alginate fiber implanted into the peritoneal cavity. J; 4M (4 weeks), N; 15M63 (4 weeks). The rate of survived cells was 94.8 ± 2.6% in alginate fiber for 4M and 83.2 ± 11.4% for 15M63 respectively. N = 2 each.

Fig 6. Scheme of differentiation of LDN + EC cells.

Treatment with LDN193189 and EC23 at stage 3 suppressed NKX6.1 expression. Ultimately, these cells differentiated into pancreatic alpha cells.