β-cell dysfunction due to increased ER stress in a stem cell model of Wolfram syndrome

Abstract

Wolfram syndrome is an autosomal recessive disorder caused by mutations in WFS1 and is characterized by insulin-dependent diabetes mellitus, optic atrophy, and deafness. To investigate the cause of β-cell failure, we used induced pluripotent stem cells to create insulin-producing cells from individuals with Wolfram syndrome. WFS1-deficient β-cells showed increased levels of endoplasmic reticulum (ER) stress molecules and decreased insulin content. Upon exposure to experimental ER stress, Wolfram β-cells showed impaired insulin processing and failed to increase insulin secretion in response to glucose and other secretagogues. Importantly, 4-phenyl butyric acid, a chemical protein folding and trafficking chaperone, restored normal insulin synthesis and the ability to upregulate insulin secretion. These studies show that ER stress plays a central role in β-cell failure in Wolfram syndrome and indicate that chemical chaperones might have therapeutic relevance under conditions of ER stress in Wolfram syndrome and other forms of diabetes.

Figures

Figure 1

iPS cells from Wolfram subjects differentiated into insulin-producing cells. (A) Diagram of WFS1 structure showing the mutation sites and Sanger sequencing profiles in the four Wolfram subjects described here. Arrows indicate the four deleted nucleotides (CTCT). (B) Immunostaining of Wolfram cultures differentiated into endoderm (SOX17), pancreatic endoderm (PDX1), and C-peptide–positive cells. Scale bar, 50 μm. (C) Differentiation efficiency in controls and WFS1 mutant cells determined by imaging (n = 10 for each of three independent experiments). (D) Immunostaining of WFS1, glucagon, and C-peptide in iPS-derived pancreatic cell cultures at d15 of differentiation. Scale bar, 20 μm.

Figure 2

UPR-dependent reduction of insulin content in Wolfram β-cells. (A) Insulin mRNA levels in control and Wolfram β-cells normalized to TBP (TATA-binding protein) mRNA levels and to the number of C-peptide–positive cells used for analysis. (B) Insulin protein content in the β-cells normalized to the number of C-peptide–positive cells. Error bars represent three independent experiments with three replicates in each experiment. *P < 0.05 for vehicle of WS vs. control cells. (C) Transmission electron microscope images of β-cells (scale bar, 2 μm). (D) Quantification of granule numbers per cell. Three independent differentiation experiments with n = 20 sections for each subject of each experiment. *P < 0.05 for WS-1 vs. control or WS-1 cells treated with 4PBA for 7 days or WS-1 cells carrying wild-type WFS1 expression vector. Ctrl, control.

Figure 3

Increased UPR pathway activity and increased sensitivity of the ER-to-ER stress in Wolfram β-cells. (A) Proinsulin-to-insulin ratio determined by ELISA; n = 6 for each of two independent experiments. (B–F) Quantification of components of the UPR displayed as fold change compared with vehicle-treated control cells (set to 1). Western blot analysis and quantification for (B) nuclear ATF6α (n = 3) and (C) phosphorylated eIF2α (n = 3) in β-cells. (D–F) Quantitative PCR for mRNA levels of sXBP-1, ATF4 in β-cells, and GRP78 in iPS cells (n = 3 for each experiment). *P < 0.05. (G) Transmission electron microscope image showing ER morphology in control and Wolfram cells after 12-h treatment with 10 nmol/L TG. Arrows point to the ER. Scale bar, 500 nm. β-Cells were treated with TG 10 nmol/L for 12 h, and 1 mmol/L 4PBA or 1 mmol/L TUDCA was added 1 h prior to and during TG treatment. 4PBA 7d refers to treatment from day 9 to day 15 of β-cell differentiation.

Figure 4

Differential effect of ER stress on insulin secretion in Wolfram and control cells. (A) Fold change of human C-peptide secretion in response to indicated secretagogues. Cells were incubated in 5.6 mmol/L glucose for 1 h followed by 16.9 mmol/L glucose, 15 mmol/L arginine, 30 mmol/L potassium, or 1 mmol/L DBcAMP + 16.9 mmol/L glucose for an additional hour. Fold change of human C-peptide secretion was calculated as the ratio of C-peptide concentration occurring in secretagogue-stimulated media over the concentration in 5.6 mmol/L glucose. Results represent three independent experiments, with n = 3 for each experiment. *P < 0.05 of TG vs. vehicle; #P < 0.05 of TG + 4PBA vs. TG. (B) Fold change of human C-peptide secretion to 16.9 mmol/L glucose stimulation; n = 3 for each of two independent experiments. (C) Fold change of human C-peptide secretion in response to 30 mmol/L potassium stimulation upon TM treatment. Results represent three independent experiments, with n = 3 for each experiment. (D) Fold change of human C-peptide and glucagon in control and Wolfram cells under indicated conditions; n = 3 for each of three independent experiments. Arg, arginine.

Figure 5

ER stress and reduced glucose-stimulated insulin secretion of Wolfram cells in vivo. (A) Immunohistochemistry for insulin and urocortin 3 in transplants. Scale bar, 20 μm. (B) Fold change of human C-peptide in the sera of mice transplanted with human islets and control and Wolfram cells after intraperitoneal glucose injection. Bars show the median. (C) Human C-peptide levels in the sera of transplanted mice over a 1-month time period. Error bars show SD. (D) Fold change of human C-peptide in the sera of transplanted mice over a 1-month time period. Error bars show SD. Immunohistochemistry for insulin and (E) ATF6α and (F) CHOP in transplants. Scale bar, 20 μm. C-PEP, C-peptide.