Comparison of top and bottom loading of a dextran gradient for rat pancreatic islet purification.

Rat pancreatic islet yields obtained with dextran gradient purification were compared after suspending the digest into either the top or the bottom layer of the gradient. A 5-layer discontinuous gradient was used, which consisted of 16 ml 31% dextran as bottom layer, overlayered with 25%, 23%, 20% and 11% dextran (4 ml each). When the digest of 1 rat pancreas was suspended into the top layer of the gradient, the total number of islets obtained from the 11-20, 20-23 and 23-25% interfaces was 862 +/- 38, 240 +/- 39 and 54 +/- 5, respectively. From this gradient, also 1409 +/- 81 islets were retrieved from the bottom layer (i.e., exocrine pellet). In contrast, when the pancreas digest was suspended into the bottom layer of the gradient, 1964 +/- 63, 435 +/- 42, and 177 +/- 34 islets were obtained from the successive interfaces, and only 50 +/- 20 islets from the exocrine pellet. The total islet volume obtained from the two uppermost interfaces was 3.46 +/- 0.31 microliters after top-loading, and 4.93 +/- 0.16 microliters after bottom-loading (n = 7, p < 0.01). When the islets retrieved from one bottom loaded gradient were transplanted into either 1 (n = 6) or 2 (n = 9) diabetic recipients, glucose levels normalized in all instances. We therefore conclude that a bottom-loaded dextran gradient separates islets from exocrine tissue effectively, resulting in significantly higher islet yields than obtained with a top-loaded dextran gradient.