OBJECTIVETo compare the effect of continuous and discontinuous density gradient centrifugation for purification of human pancreatic islets with COBE 2991 cell processor.

METHODSHuman pancreases were obtained from brain-dead donors and stored in cold UW solution. The connective tissues were removed from the pancreases, and the pancreatic ducts were perfused with a cold enzyme (Liberase). The islets were then separated by gentle mechanical dissociation and purified with discontinuous (10 pancreases) or continuous (8 pancreases) gradients of HCA-Ficoll in COBE 2991 cell processor. Samples were collected in duplicate for determination of the quantity of islets, islet equivalents (IEQ), and the purity.

RESULTSThe weights of the pancreases before and after connective tissue removal and pancreas duct perfusion, and the quantity of islets obtained (including islets quantity of different diameters and total IEQ) after dissociation were not significantly different. Continuous gradient of HCA-Ficoll, compared with discontinuous gradient, resulted in significantly greater final islet quantity (55,000 IEQ vs 206,000 IEQ, P=0.000) and islet purity (58.0%-/+8.0% vs 33.5%-/+10.3%, P=0.000) and also greater number of islets with a diameter lager than 200 microm (P<0.01).

CONCLUSIONContinuous density gradient centrifugation can be more effective than discontinuous gradient in islet purification.