Treating diet-induced diabetes and obesity with human embryonic stem cell-derived pancreatic progenitor cells and antidiabetic drugs

Abstract

Human embryonic stem cell (hESC)-derived pancreatic progenitor cells effectively reverse hyperglycemia in rodent models of type 1 diabetes, but their capacity to treat type 2 diabetes has not been reported. An immunodeficient model of type 2 diabetes was generated by high-fat diet (HFD) feeding in SCID-beige mice. Exposure to HFDs did not impact the maturation of macroencapsulated pancreatic progenitor cells into glucose-responsive insulin-secreting cells following transplantation, and the cell therapy improved glucose tolerance in HFD-fed transplant recipients after 24 weeks. However, since diet-induced hyperglycemia and obesity were not fully ameliorated by transplantation alone, a second cohort of HFD-fed mice was treated with pancreatic progenitor cells combined with one of three antidiabetic drugs. All combination therapies rapidly improved body weight and co-treatment with either sitagliptin or metformin improved hyperglycemia after only 12 weeks. Therefore, a stem cell-based therapy may be effective for treating type 2 diabetes, particularly in combination with antidiabetic drugs.

Figure 1

SCID-Beige Mice Rapidly Develop Obesity, Fasting Hyperglycemia, Glucose Intolerance, and Insulin Resistance Following Exposure to HFDs (A and B) Body weight (A) and fasting blood glucose levels (B) were measured during a 14-day acclimation period on normal chow and for 51 days following administration of one of the following diets: 10% fat (black; n = 11 mice), 45% fat (red; n = 11 mice), 60% fat (blue; n = 11 mice), or Western diet (green; n = 11 mice). (C and D) Blood glucose (C, raw values and area under the curve) and plasma mouse insulin levels (D) were assessed during an oral glucose tolerance test (OGTT; n = 4–6 mice per group) on day 47. See Figure S1 for OGTTs at days 5 and 32. (E) An insulin tolerance test (ITT) was performed on day 42 (n = 9–11 mice per group). Glucose levels are presented as a percentage of basal glucose levels (at time 0) and the area above the curve was calculated using 100% as the baseline. (F) Adiposity (% fat) was assessed by DEXA at day 43 in a subset of mice (n = 3 mice per group). (G) Mouse leptin levels were measured on days 47–49 following a 4- to 6-hr morning fast (dashed line indicates the highest standard concentration in the leptin assay). See Figures S1E–S1G for circulating lipid levels. (H) Immunofluorescent staining of epididymal fat from immunocompromised SCID-beige mice fed either a 10% or 60% fat diet for 36 weeks, and from an immunocompetent adult ob/ob mouse. F4/80 and FGF21 are shown as macrophage and adipocyte markers, respectively. Crown-like structures, representing macrophage infiltration, are shown in insets. Scale bars, 200 μm. (A–C and E–G) ^∗p < 0.05 versus 10% controls (two-way ANOVA for comparison of multiple time points in line graphs, and one-way ANOVA for bar graphs); (D) ^∗p < 0.05 versus time 0 (one-way ANOVA). Data are represented as mean ± SEM (line graphs) or as box-and-whisker plots with individual mice shown as separate data points. See also Figure S6 and Table S2.

Figure 2

Exposure to HFDs Does Not Affect the Function of hESC-Derived Pancreatic Endocrine Cells In Vivo The development of hESC-derived progenitor cells into pancreatic endocrine cells was assessed in mice fed a 10% fat (black), 45% or 60% fat (purple), or Western (green) diet. See Figure S2 for characterization of the progenitor cells pre-transplantation. (A) Human C-peptide levels were measured after an overnight fast and 40 min following an oral mixed-meal challenge (“fed”) at 8, 12, 16, and 20 weeks post-transplantation. ^∗p < 0.05, paired t test (fast versus fed). (B and C) At 18 weeks post-transplantation, human C-peptide levels were measured during an i.p. glucose tolerance test (ipGTT). In (B) data are normalized to baseline levels, and in (C) raw levels (ng/ml) are presented for individual animals, with each diet group shown on a separate plot. ^∗p < 0.05, one-way repeated-measures ANOVA (versus time 0). (D–F) At 24 weeks post-transplantation, an i.p. arginine tolerance test (ipArgTT) was performed. Plasma was collected after a 4-hr fast and 15 min following arginine administration to measure human insulin (D) and glucagon (E and F) levels. (E) shows glucagon levels at 0 and 15 minutes in transplant recipients, and (F) shows glucagon levels in sham-treated mice (Sham, striped bars) and transplant recipients (Tx, solid bars) at 15 minutes only. The red line indicates the lower limit of detection for the glucagon assay. (D and E) ^∗p < 0.05, one-tailed paired t test (0 versus 15 min); (F) ^∗p < 0.05, two-tailed t test (sham versus Tx). Data points from individual mice are shown as box-and-whisker plots. See also Table S2.

Figure 3

Gene Expression Profiles Are Similar for hESC-Derived Graft Tissues Exposed to Different LFDs or HFDs In Vivo The gene expression of islet-related genes was assessed in macroencapsulated hESC-derived grafts at 29 weeks post-transplantation from mice fed a LFD (10% fat) or HFD (45% fat, 60% fat, and Western), as well as in adult human islet preparations. All data are presented as the fold change relative to undifferentiated hESCs (H1 cells) using a log scale. Data points from individual mice or different human islet donors are shown as box-and-whisker plots. ^∗p < 0.05 for hESC-derived grafts from each diet group versus human islets (one-way ANOVA). See also Tables S2 and S3.

Figure 4

The Morphology of Macroencapsulated hESC-Derived Pancreatic Endocrine Cells Is Similar in Grafts from Mice Fed a LFD or HFD (A, B, and D–I) Representative immunofluorescent images of Theracyte devices at 29 weeks post-transplantation from mice fed 10% fat or 60% fat diets. (A) The majority of hESC-derived cells within devices from both diet groups were endocrine cells. The expression of synaptophysin (endocrine marker, red), CK19 (ductal marker, green), trypsin (exocrine marker, blue), and DAPI (nuclear marker, white) is shown. Scale bars, 100 μm. (B) The endocrine compartment was mainly composed of cells expressing either insulin (red, guinea pig antibody), glucagon (green, rabbit antibody), or somatostatin (blue, Ms antibody); scale bars, 100 μm. Higher-magnification images are shown in (E); scale bars, 50 μm. (C) The percentage of cells (% of DAPI+ nuclei) within devices that were immunoreactive for insulin (insulin+), glucagon (glucagon+), or both hormones (ins+/gcg+) was quantified in grafts from mice fed 10% fat or 45%–60% fat diets. ^∗p < 0.05, two-tailed t test. Data are represented as mean ± SEM and data points from individual mice are also shown. (D) Example of graft tissue from the 60% fat diet group with a region of endocrine cells that expressed both insulin and glucagon (white arrows); scale bars, 25 μm. (F–I) The majority of insulin-positive cells (INS, red) show a mature beta cell phenotype, including co-expression of key beta cell transcription factors (green): (F) PDX1, (G) NKX2.2, (H) NKX6.1, and (I) MAFA; scale bars, 50 μm. See also Figure S2 and Tables S2 and S4.

Figure 5

Cell Transplant Recipients on HFDs Show Improved Glucose Homeostasis Compared with Sham-Treated Mice (A and B) HbA1C levels were measured at 12 (A) and 24 (B) weeks post-transplantation in sham mice (solid bars) and transplant recipients (Tx, striped bars). (C) At 20 weeks post-transplantation, blood glucose levels were measured after an overnight fast and 40 min following an oral meal challenge (“Fed”). (D and E) ipGTTs were performed at 18 (D) and 24 (E) weeks post-transplantation in sham mice (solid lines with closed symbols; solid bars) and transplant recipients (Tx, dashed lines with open symbols; striped bars) on 10% fat (gray; 18 weeks: Tx/Sham, n = 4 mice; 24 weeks: Tx, n = 5 and Sham, n = 4 mice), 45% or 60% fat (purple; 18 weeks: Tx, n = 12 and Sham, n = 7 mice; 24 weeks: Tx, n = 13 and Sham, n = 6 mice), or Western (green; 18 weeks: Tx, n = 5 and Sham, n = 4 mice; 24 weeks: Tx, n = 6 and Sham, n = 3 mice) diets. The area under the curve is shown to the right for each ipGTT. (F) An ITT was performed at 22 weeks post-transplantation in sham and transplant recipients from each diet group (10% Tx, n = 5 mice; 10% Sham, n = 4 mice; 45%–60% Tx, n = 12 mice; 45%–60% Sham, n = 6 mice; West Tx, n = 6 mice; West Sham, n = 4 mice). The area above the curve (right panel) was calculated using the fasting glucose level (100%) for each animal as the baseline. For clarity, all glucose curves (GTTs and ITTs) from sham and transplanted mice are shown separately for each diet group. See Figure S4 for glucose curves combined on the same plots. Data are represented as mean ± SEM (line graphs) or as box-and-whisker plots showing individual mice as separate data points. For all box-and-whisker plots: ^∗p < 0.05, one-way ANOVA versus 10% sham; #p < 0.05, two-tailed t test (Tx versus sham). For all line graphs: ^∗p < 0.05, two-way ANOVA, Tx versus sham. See Figure S3 for long-term body weight and blood glucose tracking, and Figure S5 for the effect of cell transplants on endogenous pancreas, liver, fat, and circulating leptin levels. See also Table S2.

Figure 6

Diet-Induced Obesity Was Reversed Following Progenitor Cell Transplantation Combined with an Antidiabetic Drug (A–F) Fasting body weight was assessed in mice fed a 10% fat diet without drug (black/gray; all panels; n = 8 mice), 60% fat diet without drug (A and B, blue; n = 7–8 mice per group), 60% fat diet plus metformin (A and C, purple; n = 7–8 mice per group), 60% fat diet plus sitagliptin (A and D, orange; n = 8 mice per group), or 60% fat diet plus rosiglitazone (A and E, green; n = 8 mice per group). Body weight tracking for sham mice from all treatment groups is shown in (A). Sham mice (solid lines, closed symbols) and transplant recipients (Tx; dashed lines, open symbols) from each treatment group are shown together with the LFD control as a reference (B–E). The change in body weight from day −2 to day 12 is shown in box-and-whisker plots to the right of each line graph, with each data point representing an individual mouse. Data on line graphs are represented as mean ± SEM. (F) Body weight pre-transplant (day −2) and post-transplantation (day 75). (G and H) Epididymal fat pad weight relative to body weight (G) and plasma mouse leptin levels (H) were assessed at 20 weeks post-transplantation. (A–E) ^∗p < 0.05, one-way ANOVA for comparisons of all groups (LFD versus HFD Sham, LFD versus HFD Tx, HFD Sham versus HFD Tx); (F–H) #p < 0.05, one-way ANOVA for comparison of each group to LFD control; ^∗p < 0.05, two-tailed t test, Tx versus Sham. See Figure S6 for the phenotype of mice on HFDs in cohort 2 (pre-transplant) and Figure S7 for long-term blood glucose tracking following administration of drugs with or without cell transplants.

Figure 7

The Combination of Either Sitagliptin or Metformin with a Cell Transplant Resulted in Improved Glucose Tolerance within 12 Weeks (A–E) OGTTs were performed at 12 weeks post-transplantation in mice fed a 10% fat diet without drug (black/gray; all panels; Sham: n = 5 mice), 60% fat diet without drug (A and B, blue; Tx/Sham: n = 7 mice), 60% fat diet plus metformin (A and C, purple; Tx/Sham: n = 5 mice), 60% fat diet plus sitagliptin (A and D, orange; Tx: n = 7 mice, Sham: n = 8 mice), or 60% fat diet plus rosiglitazone (A and E, green; Tx: n = 7 mice, Sham: n = 6 mice). Blood glucose levels for sham mice from all treatment groups are shown together in (A). Sham mice (solid lines, closed symbols) and transplant recipients (Tx; dashed lines, open symbols) from each treatment group are shown together with LFD controls as a reference (B–E). The area under the curve (AUC) is shown in box-and-whisker plots to the right of each line graph, with separate data points representing individual mice. Data on line graphs are represented as mean ± SEM. (A) ^∗p < 0.05 (one-way ANOVA versus LFD). (B–E) For glucose curves, two-way ANOVA was used to examine differences at each time point during the OGTT (^∗p < 0.05, HFD Sham versus LFD; #p < 0.05, HFD Tx versus LFD). For the AUC, one-way ANOVA was used compare all groups (LFD versus HFD Sham, LFD versus HFD Tx, HFD Sham versus HFD Tx; ^∗p < 0.05; ns, not significant). (F and G) Fasting mouse C-peptide levels at 4 weeks (F) or 16 weeks (G) post-transplantation. ^∗p < 0.05, two-tailed t test (Tx versus Sham). (H) Human C-peptide levels were measured after an overnight fast and 60 min following an intraperitoneal glucose challenge. ^∗p < 0.05, paired two-tailed t test (0 versus 60 min).