In Vitro Scalable Production of Mature, Glucose-Responsive β Cells from Human Embryonic Stem Cells for Therapy and Drug Screening | AIChE

In Vitro Scalable Production of Mature, Glucose-Responsive β Cells from Human Embryonic Stem Cells for Therapy and Drug Screening


The generation of functional, insulin-secreting pancreatic β cells from stem cells in vitro would provide an unprecedented cell source for cell replacement therapy and drug screening in diabetes.  However, insulin-producing cells generated in vitro from human embryonic stem cells (hESC) lack many characteristics of bona fide β cells, including glucose-stimulated insulin secretion, appropriate marker expression, and rapid function after transplantation.  Furthermore, production at a scale for practical use is highly desirable and technically challenging.

We have developed an in vitro, suspension-based protocol that produces β cells from hESC (ES-β cells) in a 300-mL spinner flask using defined factors.  ES-β cells increased intracellular [Ca2+] and secreted insulin in response to multiple sequential high glucose challenges at amounts comparable to cadaveric β cells.  Insulin was packaged into secretary granules in a manner that was indistinguishable from cadaveric β cells when visualized by electron microscopy.  In addition to insulin, ES-β cells expressed the β cell markers PDX1, MAFA, and NKX6-1 and did not express other non-β cell markers, including glucagon, somatostatin, ghrelin, and pancreatic polypeptide.  This scalable protocol produced approximately 3x108 cells, of which 30-50% are ES-β cells.

Two weeks after transplantation into SCID/Beige mice, high amounts of human insulin was detected in the serum of mice that received ES-β cells, similar to that found in mice transplanted with cadaveric β cells.  The amount of human insulin increased after intraperitoneal injection of glucose, demonstrating the cells were functional in vivo.  No human insulin was detected in mice transplanted with hESC-derived pancreatic progenitors and non-functional insulin-producing endocrine, consistent with previous findings. 

To further demonstrate the utility of these ES-β cells, we treated ES-β cells with different categories of known anti-diabetic drugs that increase insulin secretion and with prolactin, which increases β cell proliferation.  Increasing β cell function and mass are highly sought after drug targets for treating diabetes.  The categories of anti-diabetic drugs tested were GCK activators, sulfonylureas, GLP-1 agonists, and meglitinide, and, as expected, treatment increased insulin release from ES-β cells.  In addition, treatment with prolactin increased proliferation of ES-β cells, as indicated by an increase in Ki67 staining.

We have developed a scalable process to generate functional ES-β cells in vitro that are indistinguishable from cadaveric β cells.  This protocol is able to produce >108 ES-β cells per batch, meaning it is the first scalable human source of functional β cells for therapy and drug discovery.  Given that the only alternative is the very limited and unreliable supply of cadaveric β cells, this discovery represents a major advance in stem cell engineering for treating diabetes.