Clinically compliant differentiation of human pluripotent stem cells into beta cells and enrichment of stem cell islets

Bahareh Rajaei1, Amadeo Muñoz García1, Juri Juksar1,  Jason Doppenberg2, Marten Engelse1, Eelco de Koning1, Françoise Carlotti1

1Department of Internal Medicine, Leiden University Medical Center

2Transplant Center, Leiden University Medical Center

 

Background: Pluripotent stem cell-derived pancreatic islets (SC-islets) hold great promise for β cell replacement therapy in patients with type 1 diabetes. Despite significant progress in the differentiation protocols, current SC-islet preparations contain non-target (non-islet) cells, which raises the safety risk and increases the transplant volume. Here, we describe a clinically-compliant, full 3D differentiation protocol that includes a purification step relying on the principle of isopycnic centrifugation (density gradient separation) for the generation of the final cell product enriched in SC-islet cell clusters.

Methods: We adapted a seven-stage (30-day) differentiation protocol to generate SCislets in a 3D suspension culture system. Differentiation efficiency was monitored by flow cytometry. Stage-7 cells were evaluated in vitro by single-cell RNA sequencing and in vivo by intraperitoneal glucose tolerance test after transplantation under the kidney capsule of immunodeficient mice. A linear continuous density gradient was created by mixing UW solution with Xenetix 350 to enrich for SC-islets in the final cell product generated from research grade HUES8 and clinical grade RC9 as well as 3 GMP-iPSC lines. Purity of each fraction was assessed by DTZ staining. The SC-islets (C-pep/GCG) enrichment was monitored by qPCR and immunostaining.

Results: Before density gradient separation, the cell clusters contained on average

52.5 ± 5.4% SC-β cells (C-peptide positive), 8.4 ± 2.7% SC-α cells (glucagon positive), and 4.0 ± 1.7% bihormonal cells. Importantly, these cells were organized in islet cellrich and endocrine cell-poor clusters, as assessed by dithizone (DTZ) staining. During density gradient separation, islet cell-rich clusters appeared at a density of 1.053– 1.088 g/ml, which is similar to primary human islets. Pooled enriched SC-islet fractions consisted of 93.9 ± 0.8% DTZ-positive tissue, while the tissue volume for transplantation was reduced by 22 ±13% after purification. Enriched SC-islets displayed improved functionality both in vitro and in vivo. In vivo, stimulated human Cpeptide was increased by 25.6 fold 6 months post transplantation in enriched SC-islet transplants, while non-purified SC-islet transplants showed a 9.5 fold increase as compared to day 14 post transplantation.

Conclusion: We propose density gradient separation as a GMP-compliant purification method that is fast, easily scalable, and cost-effective to enrich SC-islets in the final cell product. In contrast to antibody-based single-cell sorting approaches, this method does not destroy the islet cytoarchitecture, which is associated with alteration of islet function and cell loss. This method raises the safety of the final cell product and provides a smaller transplant volume without adverse effects on the biological activity of the cells.