Assessment of Human Islet Grafts in Frozen Sections of CD-1 Athymic nu/nu Mouse Liver for Molecular Analysis

Abstract Background Post-transplantation islet graft monitoring is hampered by a lack of efficient methods to locate and analyze islets in situ. We evaluated histologic methods to rapidly locate islets within the liver parenchyma post-transplantation, using several staining strategies, prior to analysis using laser capture microdissection. Methods Human islets were isolated (n = 8) from brain dead, multiorgan donor pancreases at the McGill University Health Centre Islet Transplant Laboratory. Mean yield was 247,609 ± 195,272 IE and 3172 ± 1645 IE/g (purity and viability, respectively, 84.5 ± 8.6% and 95 ± 5% average; mean ± SD). Diabetic athymic CD-1 nu/nu mice (streptozotocin intraperitoneal injection, 200 mg/kg) were maintained with sustained release insulin pellets until a suitable islet preparation was available for transplant. Intraportal islet transplantation of 2000 IE/mouse was performed via the ileocecal vein, as previously described. Frozen sections of liver containing human islets were prepared from specimens collected on days 0, 4, and 30 post-transplant. Every twentieth slide from serial sectioned liver was stained using a rapid protocol to determine if islets were present. Sections were fixed and stained for 5 minutes with either an anti-human insulin fluorescein isothiocyanate (FITC)-conjugated primary antibody (Ins-FITC), Newport Green (NG), or diphenylthiocarbazone (dithizone, DZ). Results Islets were readily localized using each technique, mostly toward the liver periphery. However, DZ had a faint appearance in 10-μm-thick sections and was best utilized to locate sections containing islets during sectioning. Conclusion Ins-FITC, NG, and DZ are all good candidates for a rapid islet staining protocol to evaluate human islet grafts in situ, with DZ being best for sectioning and Ins-FITC and NG being equal in locating islets during processing for laser capture microdissection.