Infusing pancreatic islets into the portal vein currently represents the preferred approach for islet transplantation, despite considerable loss of islet mass almost immediately after implantation. Therefore, approaches that obviate direct intravascular placement are urgently needed. A promising candidate for extrahepatic placement is the omentum. We aimed to develop an extracellular matrix skeleton from the native pancreas that could provide a microenvironment for islet survival in an omental flap. To that end, we compared different decellularization approaches, including perfusion through the pancreatic duct, gastric artery, portal vein, and a novel method through the splenic vein. Decellularized skeletons were compared for size, residual DNA content, protein composition, histology, electron microscopy, and MR imaging after repopulation with isolated islets. Compared to the other approaches, pancreatic perfusion via the splenic vein provided smaller extracellular matrix skeletons, which facilitated transplantation into the omentum, without compromising other requirements, such as the complete depletion of cellular components and the preservation of pancreatic extracellular proteins. Repeated MR imaging of iron-oxide-labeled pancreatic islets showed that islets maintained their position in vivo for 49 days. Advanced environmental scanning electron microscopy demonstrated that islets remained integrated with the pancreatic skeleton. This novel approach represents a proof-of-concept for long-term transplantation experiments.

We have recently developed a model of pancreatic islet transplantation into a decellularized pancreatic tail in rats. As the pancreatic skeletons completely lack endothelial cells, we investigated the effect of cotransplantation of mesenchymal stem cells and endothelial cells to promote revascularization. Decellularized matrix of the pancreatic tail was prepared by perfusion with Triton X-100, sodium dodecyl sulfate and DNase solution. Isolated pancreatic islets were infused into the skeletons via the splenic vein either alone, together with adipose tissue-derived mesenchymal stem cells (adMSCs), or with a combination of adMSCs and rat endothelial cells (rat ECs). Repopulated skeletons were transplanted into the subcutaneous tissue and explanted 9 days later for histological examination. Possible immunomodulatory effects of rat adMSCs on the survival of highly immunogenic green protein-expressing human ECs were also tested after their transplantation beneath the renal capsule. The immunomodulatory effects of adMSCs were also tested in vitro using the Invitrogen Click-iT EdU system. In the presence of adMSCs, the proliferation of splenocytes as a response to phytohaemagglutinin A was reduced by 47% (the stimulation index decreased from 1.7 to 0.9, P = 0.008) and the reaction to human ECs was reduced by 58% (the stimulation index decreased from 1.6 to 0.7, P = 0.03). Histological examination of the explanted skeletons seeded only with the islets showed their partial disintegration and only a rare presence of CD31-positive cells. However, skeletons seeded with a combination of islets and adMSCs showed preserved islet morphology and rich vascularity. In contrast, the addition of syngeneic rat ECs resulted in islet-cell necrosis with only few endothelial cells present. Live green fluorescence-positive endothelial cells transplanted either alone or with adMSCs were not detected beneath the renal capsule. Though the adMSCs significantly reduced in vitro proliferation stimulated by either phytohaemagglutinin A or by xenogeneic human ECs, in vivo co-transplanted adMSCs did not suppress the post-transplant immune response to xenogeneic ECs. Even in the syngeneic model, ECs co-transplantation did not lead to sufficient vascularization in the transplant area. In contrast, islet co-transplantation together with adMSCs successfully promoted the revascularization of extracellular matrix in the subcutaneous tissue.

Adult pancreatic β-cells are mostly locked in quiescence, limiting large-scale analysis of their cell cycle. To overcome this limitation, we used synthetic in vitro transcribed mRNAs encoding Cyclin D1 and CDK4 to induce proliferation in up to 70% of primary rat β-cells. Flow cytometry-based analysis of cell-cycle markers revealed substantial heterogeneity in β-cell cell-cycle progression, identifying five distinct groups of proliferating β-cells based on G1 entry timing. Total cell-cycle length ranged from approximately 26 to 34 hours and was primarily determined by variability in G1 duration (13-19 hours). In the fastest-dividing β-cells, G1, S, and G2M lasted approximately 13, 6, and 7 hours, respectively, whereas later-entering β-cells exhibited progressively longer G1 phase. Additionally, we identified a small β-cell subpopulation that failed to complete division and may have exited the cell cycle. These findings reveal cell-cycle entry and G1 progression as a major source of heterogeneity in β-cell proliferation and provide quantitative benchmark for developing strategies to enhance controlled β-cell regeneration in diabetes.