ROCKII inhibition promotes the maturation of human pancreatic beta-like cells

Abstract

Diabetes is linked to loss of pancreatic beta-cells. Pluripotent stem cells offer a valuable source of human beta-cells for basic studies of their biology and translational applications. However, the signalling pathways that regulate beta-cell development and functional maturation are not fully understood. Here we report a high content chemical screen, revealing that H1152, a ROCK inhibitor, promotes the robust generation of insulin-expressing cells from multiple hPSC lines. The insulin expressing cells obtained after H1152 treatment show increased expression of mature beta cell markers and improved glucose stimulated insulin secretion. Moreover, the H1152-treated beta-like cells show enhanced glucose stimulated insulin secretion and increased capacity to maintain glucose homeostasis after transplantation. Conditional gene knockdown reveals that inhibition of ROCKII promotes the generation and maturation of glucose-responding cells. This study provides a strategy to promote human beta-cell maturation and identifies an unexpected role for the ROCKII pathway in the development and maturation of beta-like cells.Our incomplete understanding of how pancreatic beta cells form limits the generation of beta-like cells from human pluripotent stem cells (hPSC). Here, the authors identify a ROCKII inhibitor H1152 as increasing insulin secreting cells from hPSCs and improving beta-cell maturation on transplantation in vivo.

Fig. 1

High content screening to identify H1152 that promotes the generation of INS⁺ cells. a Scheme of the directed differentiation protocol and representative images of hESC derived definitive endoderm (DE) and pancreatic progenitors (PP). Scale bar = 50 µm. b Scheme of the chemical screen. c Primary screening result. d Chemical structure of H1152. e Representative images from primary screening. Red: insulin/INS, blue: DAPI. Scale bar = 50 μm. f Dose response curve for effects of H1152 on percentage of INS⁺ cells. g Intracellular FCM to detect the expression of insulin (upper graphs) and c-peptide (lower graphs) of DMSO or H1152-treated cells. h The increase of INS⁺ cells does not depend on the continued presence of H1152. N = 3 independent biological replicates. p-values were calculated by unpaired two-tailed Student’s t-test. ***p < 0.001. Error bar is SEM. i Immunofluorescent imaging of DMSO or H1152 treated cells stained with antibodies against insulin and Ki67. Scale bar = 50 µm. j Immunofluorescent imaging of DMSO or H1152 treated cells for c-peptide expression in 2 different iPSC lines. Scale bar = 50 µm. AA Activin A; RA Retinoic acid

Fig. 2

H1152 promotes the maturation of hESC-derived glucose-responding cells. a Scheme of the directed differentiation protocol. b Flow cytometry analysis, the percentage of INS-GFP⁺ cells and the mean signal of INS-GFP⁺ cells of DMSO and H1152 treated spheres. c–e Confocal imaging (c) intracellular FCM (d) and qRT-PCR (e) analysis of H1152-treated or DMSO-treated spheres. N = 3–6 independent biological replicates. Scale bar = 25 μm. Error bar is SEM. Primary human islets were used as a control in Fig. 2e. UCN3: urocortin3, SS: somatostatin, PP: pancreatic polypeptide. f Total c-peptide content of H1152-treated or DMSO-treated spheres, compared with human islets. g KCl-stimulated insulin secretion of H1152-treated or DMSO-treated spheres. h GSIS of H1152-treated or DMSO-treated spheres. N = 8–12 independent biological replicates. n.s. indicates non-significant difference. p-values were calculated by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01. AA Activin A; Ch Chir; Glc Glucose; RA Retinoic acid; KSIS KCl stimulated insulin secretion; GSIS Glucose stimulated insulin secretion. The bottom and top of the box represent the first and third quartiles, the band inside the box represents the median. The ends of the whiskers represent the minimum and maximum of all the data

Fig. 3

Global transcription analysis suggests that H1152 treatment upregulated genes involved in mature pancreatic beta cell function. a Heatmap and hierarchical clustering of transcriptional profiling in INS ^GFP/W HES3-derived INS-GFP⁺ cells treated with H1152 compared to DMSO treated cells. Fetal and adult primary human beta cells were used as controls (GPL11154). b Pathway enrichment analysis on up-regulated or down-regulated genes in INS ^w/GFP HES3-derived INS-GFP⁺ cells treated with H1152 and adult human islets. c Heatmap of genes in the upregulated pathways, including insulin secretion and type 2 diabetes mellitus. d Heatmap of the genes in the downregulated pathways, including focal adherent and cell cycle. e GSEA of DMSO-treated, H1152-treated INS-GFP⁺ cells and fetal and adult human primary beta cells. The dataset are the gene expression level of DMSO-treated or H1152-treated INS-GFP⁺ cells. The adult beta gene set includes the genes that are >50-fold higher expressed in adult beta cells compared to fetal beta cells. The fetal beta gene set includes the genes that are >50-fold higher expressed in fetal beta cells compared to adult beta cells

Fig. 4

H1152 promotes the generation and maturation of hESC-derived glucose-responding cells in the presence of a distinct protocol. a Scheme of the directed differentiation. b–d Confocal imaging (b) intracellular FCM (c) and qRT-PCR (d) analysis of H1152-treated or DMSO-treated cells. Scale bar: 100 µm. N = 6–8 independent biological replicates. UCN3: Urocortin3, P1: PDX1, N6.1: NKX6.1, SS: somatostatin, PP: pancreatic polypeptide. Error bar is SEM. e The percentage of c-peptide content and fold induction of H1152-treated or DMSO-treated cells after exposure to 30 mM KCl. Fold induction was calculated by dividing the percentage of c-peptide content of KCl treated condition by that of the untreated condition. N = 8–12 independent biological replicates. f The percentage of c-peptide content and fold induction of H1152-treated or DMSO-treated cells in the presence of 2 mM D-glucose (LG) and 20 mM D-glucose (HG). Fold induction was calculated by dividing the percentage of c-peptide content of HG treated condition by that of the LG treated condition. N = 8–12 independent biological replicates. p-values were calculated by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01. KSIS KCl stimulated insulin secretion, GSIS Glucose stimulated insulin secretion. The bottom and top of the box represent the first and third quartiles, the band inside the box represents the median. The ends of the whiskers represent the minimum and maximum of all the data

Fig. 5

H1152-treated cells show improved capacity to respond to glucose stimulation and maintain glucose homeostasis in vivo. a Blood glucose of mice transplanted with H1152-treated or DMSO-treated cells. The cells were transplanted at day −4. 180 mg/kg STZ was given at day 0. The kidneys were removed at day 134. Error bar is SEM. b GSIS of mice 5 weeks after transplantation with H1152-treated or DMSO-treated cells. The mice were fasted overnight and injected with 3 g/kg D-glucose. Red and blue arrows indicate positive GSIS in mice transplanted with H1152- or DMSO-treated cells, respectively. c IPGTT of mice 5 weeks after transplanted with H1152-treated or DMSO-treated cells. The mice were fasted overnight and injected with 2 g/kg D-glucose. The blood glucose level was measured every 15 min. N = 6 mice for each condition. Fold induction was calculated by dividing the blood glucose level at each time point by the baseline glucose level. Error bar is SEM. d Quantification of immunohistochemistry analysis of grafts of H1152-treated cells or DMSO-treated cells N = 3 independent biological replicates. Error bar is SEM. STZ Streptozotocin. p-values in a were calculated by two-way repeated measures ANOVA with a Bonferroni test. p-values in c were calculated by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01

Fig. 6

ROCKII inhibition promotes the generation and maturation of glucose-responding cells. a Scheme of the directed differentiation protocol. b–d Intracellular FCM (b), Confocal imaging (c) and qRT-PCR analysis (d) of INS ^w/GFP HES3-carrying vectors to express shROCKI, shROCKII or scrambled shRNA in the presence or absence of 10 ng/ml dox treatment. Primary human islets and H1152-treated cells were used as controls in d. Error bar is SEM. e The percentage of c-peptide content and fold induction of ROCKI-KD and ROCKII-KD cells after 30 mM KCl treatment. Fold induction was calculated by dividing the percentage of c-peptide content in the KCl treated condition by that of the untreated condition. N = 8–12 independent biological replicates. f The percentage of c-peptide content and fold induction of ROCKI-KD and ROCKII-KD cells in the presence of 2 mM D-glucose (LG) and 20 mM d-glucose (HG). Fold induction was calculated by dividing the percentage of c-peptide content in the HG treated condition by that of the LG treated condition. Scale bar: 25 µm. Primary human islets and H1152-treated cells were used as controls in d. H1152-treated cells were used as controls in e and f. N = 8–12 independent biological replicates. p-values were calculated by unpaired two-tailed Student’s t-test. *p < 0.05, **p < 0.01. N6.1, NKX6.1. The bottom and top of the box represent the first and third quartiles, the band inside the box represents the median. The ends of the whiskers represent the minimum and maximum of all the data