Siqin Wu, Shivam Chandel, Galyna Bryzgalova, Paschalis Efstathopoulos, Kelly Blust, Cheng Zhao, Eda Erbil, Anna Falk, My Hedhammar, Per-Olof Berggren, Fredrik Lanner
26 min
Developing reliable cell therapies for type 1 diabetes requires the efficient differentiation of human pluripotent stem cells (hPSCs) into functional pancreatic islets. Existing protocols often struggle with two major issues: the presence of unwanted, proliferative non-endocrine cells that pose safety risks, and the production of immature islets with poor glucose responsiveness. The authors aimed to refine the differentiation process by optimizing the timing of developmental stages and the physical environment of the cells.
They focused on the transition from the pancreatic progenitor (PP) stage to the endocrine progenitor (EP) stage. By shortening the duration of the PP stage and utilizing laminin-521 (LN-521) as a substrate, they improved the efficiency of EP formation. Crucially, they introduced a step where EP cells are allowed to self-aggregate into 3D clusters, a process designed to mimic natural pancreatic development and selectively remove non-endocrine cells.
The refined protocol consistently produced functional stem cell-derived islets (SC-islets) across all eight tested hPSC lines. The 3D aggregation step was particularly effective, leading to a significant reduction in Ki-67+ proliferative cells and NEUROD1- non-endocrine populations. The resulting SC-islets displayed strong glucose-stimulated insulin secretion (GSIS) in vitro, with dynamic responses comparable to primary human islets.
When transplanted into the anterior chamber of the eye in diabetic mice, these SC-islets successfully matured and restored glycemic control. Single-cell RNA sequencing confirmed that the final product was composed primarily of mature beta and alpha cells, with a notable absence of non-endocrine or proliferative cell populations. Furthermore, the protocol overcame a previously identified metabolic bottleneck, as the islets showed robust insulin secretion in response to glyceraldehyde.
This study provides a scalable and robust method for generating high-quality pancreatic islets for potential clinical use. By addressing the common problems of cell-line variability and the presence of dangerous non-endocrine contaminants, this protocol offers a more reliable path toward safe and effective cell-based treatments for type 1 diabetes. The ability to produce functional, pure endocrine clusters from various stem cell sources is a significant step forward in the field of regenerative medicine.
The success of cell therapy for type 1 diabetes (T1D) depends on reliable differentiation of stem cells into functional pancreatic islets. Current protocols produce stem cell-derived islets (SC-islets) that contain non-endocrine cells and show limited maturity. We developed a robust protocol that generates functional SC-islets from all eight tested human pluripotent stem cell (hPSC) lines. Differentiation to the endocrine progenitor (EP) stage on 2D laminin-521 is improved by shortening the prior pancreatic progenitor (PP) stage. Notably, allowing EP cells to self-aggregate efficiently removes proliferative and non-endocrine cells. Subsequent suspension culture yields SC-islets with strong glucose responsiveness in vitro. After transplantation into the anterior chamber of the eye of diabetic mice, SC-islets further mature and restore normal glycemic control. Single-cell analyses show that the SC-islets are free of non-endocrine cell populations before and after transplantation. This protocol enables production of highly functional SC-islets suitable for T1D cell therapy.
Sam: The evidence suggests it does. The clusters responded to sugar the way a real pancreas would—detecting it and releasing insulin in response. The researchers also transplanted these clusters into diabetic mice, using the eye as a transplant site, which is useful because it lets scientists watch the cells through a microscope as they mature. The mice regained control over their blood sugar.
Alex: That's a meaningful result. But does the self-sorting process fully solve the tumor concern?
Sam: The study suggests the clusters are free of those problematic, rapidly dividing cells. Using a technique called single-cell analysis—essentially reading the identity of each individual cell in the final product—they confirmed the clusters were pure. That said, the researchers are clear that moving from mouse models to human patients is a long and complex process.
Alex: Are there other limitations worth flagging?
Sam: Two in particular. First, while the protocol works, the researchers don't yet fully understand the molecular reasons why shortening that middle stage improves the final quality. The mechanism is observed, but not entirely explained. Second, the long-term stability of these grafts—specifically beyond six months—still needs more investigation.
Alex: So it's a solid, functional model, but the underlying science of the timing and the question of long-term durability are still open.
Sam: Precisely. But the potential here is meaningful. This approach could eventually allow researchers to create personalized cell banks from a patient's own stem cells—transplants that the immune system wouldn't reject, and that are free of the tumor risk. If that holds up, it could reduce or even remove the need for the lifelong immune-suppressing medication that transplant patients currently depend on.
Alex: That's a significant potential shift in how type 1 diabetes might be managed. Thanks for walking me through the science, Sam.
Sam: It was a pleasure. It's a good example of how working with biology's own logic—rather than against it—can sometimes be the most effective path forward.
Alex: Thanks for listening to ResearchPod.