Human iPS cell-derived insulin producing cells form vascularized organoids under the kidney capsules of diabetic mice

Abstract

Type 1 diabetes (T1D) is caused by autoimmune disease that leads to the destruction of pancreatic β-cells. Transplantation of cadaveric pancreatic organs or pancreatic islets can restore normal physiology. However, there is a chronic shortage of cadaveric organs, limiting the treatment of the majority of patients on the pancreas transplantation waiting list. Here, we hypothesized that human iPS cells can be directly differentiated into insulin producing cells (IPCs) capable of secreting insulin. Using a series of pancreatic growth factors, we successfully generated iPS cells derived IPCs. Furthermore, to investigate the capability of these cells to secrete insulin in vivo, the differentiated cells were transplanted under the kidney capsules of diabetic immunodeficient mice. Serum glucose levels gradually declined to either normal or near normal levels over 150 days, suggesting that the IPCs were secreting insulin. In addition, using MRI, a 3D organoid appeared as a white patch on the transplanted kidneys but not on the control kidneys. These organoids showed neo-vascularization and stained positive for insulin and glucagon. All together, these data show that a pancreatic organ can be created in vivo providing evidence that iPS cells might be a novel option for the treatment of T1D.

Figure 1. Generation of IPCs.

A) Stepwise differentiation of human iPS cells leads to the generation of definitive endodermal (DE) cells, pancreatic endodermal (PE) cells, pro-endocrine progenitors (S3), pre-IPCs (Pre) and finally insulin producing cells (IPCs). B) At the endodermal stage, the cells were stained for CXCR4 and sorted using immunomagnetic beads. These cells were further used to differentiate IPCs. C) Precursor cells were stained for Pdx1 to determine whether they had adopted the pancreatic lineage. Almost all the cells were expressing Pdx1.

Figure 2. Imaging of IPCs using Transmission Electron Microscopy.

A) Human iPS cell derived-IPCs were positive for C-peptide as well as Marf A. B) Human iPS cell-derived IPCs as well as human pancreatic islets were analyzed by transmission electron microscopy to identify insulin secretory granules. In human iPS cell-derived IPCs a few (∼50–70) insulin granules with and without characteristic halo around them were detected. In contrast, the human pancreatic islets had >100 insulin secretory granules with a characteristic halo around them (C). D) The number of granules in both human islets and the IPCs show that IPCs display fewer granules than those counted in the islets.

Figure 3. IPCs respond to high glucose levels as detected by the mitochondria stress test.

Human iPS cell-derived IPCs were tested for oxygen consumption using the mitochondria stress test. In high glucose (A), oxygen consumption significantly went up and could be blocked by Nifedipine. As expected, in low glucose (B) oxygen consumption could not be increased by IBMX (3-isobutyl-1-methylxanthine).

Figure 4. IPCs correct hyperglycemia in diabetic mice.

A) Eight weeks old Rag2^−/−γc^−/− mice were made diabetic following streptozotocin treatment. The pre-transplant blood levels were >400 mg/dl. Approximately 5×10⁶ human iPS cell-derived IPCs were transplanted under the kidney capsule of each mouse. The blood glucose levels were monitored for >100 days. Blood glucose levels with peak and trough kinetics were observed throughout the duration of the study. The normalization of blood glucose levels was observed in 3 out of 5 mice. The remaining 2 were borderline diabetic. B) To study whether the transplanted mice could control high sugar levels, we performed the glucose tolerance test. Mice that had become normoglycemic were subjected to intra-peritoneal glucose tolerance test. The normal healthy control mice displayed a faster blood glucose clearance while the mice transplanted with human iPS cell-derived IPCs had a slightly impaired glucose tolerance. C) To rule out that the pancreata of streptozotocin-treated mice did not regenerate, we sacrificed the transplanted mice and histologically examined the pancreas. Streptozotocin-treated mice that had received HPCs did not have any pancreatic islets left.

Figure 5. IPCs form organoids in vivo under the kidney capsule.

A) The Rag2^−/−γc^−/− mice transplanted with human iPS cell-derived IPCs were subjected to MRI to monitor the fate of the transplanted IPCs in real-time. MRI revealed that the transplanted IPCs were present as a white mass on the surface of the kidney they were transplanted into. Kidneys were imaged both in the axial and coronal projections. B) To further study the tissue on the kidneys, the mice were sacrificed. The transplanted IPCs were observed as a white vascularized organoid with triangular shape. There was no evidence of any teratoma formation or abnormal growth. The IPC transplanted kidney showed signs of tissue adhesion at the site of IPC injection (dotted line). C) To examine the organoids further, histological sections were stained by H&E and studied under the microscope. The organoids appeared triangular in shape and separate from the kidneys. More interestingly they appeared to have developed neovascularization (insert, arrow heads).

Figure 6. Histology and Immunostaining of the transplanted IPCs.

To confirm that the organoids were pancreatic in nature, histological sections were stained for C-peptide (A), insulin (B, E), isotype control (C) and glucagon (D). The overlay (F) shows that the cells are positive for both hormones suggesting that they are bihormonal.