Clinical islet transplantation achieves insulin independence in selected patients, yet current methods for extracting islets from their surrounding pancreatic matrix are suboptimal. The islet basement membrane (BM) influences islet function and survival, and is a critical marker of islet integrity following rodent islet isolation. No studies have investigated the impact of islet isolation on BM integrity in human islets, which have a unique duplex structure. To address this, samples were taken from 27 clinical human islet isolations (donor age 41-59, BMI 26-38, CIT <10h). Collagen IV, panlaminin, perlecan and laminin-α5 in the islet BM were significantly digested by enzyme treatment. In isolated islets, laminin-α5 (found in both layers of the duplex BM) and perlecan were lost entirely, with no restoration evident during culture. Collagen IV and pan-laminin were present in the disorganised BM of isolated islets, yet a significant reduction in pan-laminin was seen during the initial 24h culture period. Islet cytotoxicity increased during culture. Therefore, the human islet BM is substantially disrupted during the islet isolation procedure. Islet function and survival may be compromised as a consequence of an incomplete islet BM, which has implications for islet survival and transplanted graft longevity.

Despite huge advances in the field of islet transplantation over the last two decades, current islet isolation methods remain suboptimal, with transplantable yields obtained in less than half of all pancreases processed worldwide. Successful islet isolation is dependent on the ability of collagenase-based enzyme blends to digest extracellular matrix components at the islet–exocrine interface. The limited availability of donor pancreases hinders the use of full-scale islet isolations to characterize pancreas digestion by different enzyme components or blends, or allow the influence of inter-pancreatic variability between donors to be explored. We have developed a method that allows multiple enzyme components to be tested on any one pancreas. Biopsies of 0.5 cm3 were taken from seven standard (age ≥45) and eight young (age ≤35) pancreases. Serial cryosections were treated with Serva collagenase, neutral protease (NP), or the two enzymes together at clinically relevant concentrations. Following digestion, insulin and either collagen IV or laminin-α5 were detected by immunofluorescent labeling. Protein loss at the islet–exocrine interface was semi-quantified morphometrically, with reference to a control section. Differential digestion of the two proteins based on the enzyme components used was seen, with protein digestion significantly influenced by donor age. Treatment with collagenase and NP alone was significantly more effective at digesting collagen IV in the standard donor group, as was the NP mediated digestion of laminin-α5. Collagenase alone was not capable of significantly digesting laminin-α5 in either donor group. Combining the two enzymes ameliorated the age-related differences in the digestion of both proteins. No significant differences in protein loss were detected by the method when analyzed by two independent operators, demonstrating the reproducibility of the assay. The development of this simple yet reproducible assay has implications for both enzyme batch testing and identifying inter-donor digestion variability, while utilizing small amounts of both enzyme and human tissue.