Background: The current technique of human pancreas digestion for islet isolation relies on selective distribution of collagenase delivered via the pancreatic duct to produce digestion and removal of peri-acinar fibrous tissue. However, the collagenase has relatively little effect on the interlobular fibrous tissue, which must therefore be broken down by mechanical means within the digestion chamber so as to release the contained acini and islets. The current way of achieving this in the Ricordi chamber is to place five or six stainless steel balls within the chamber and shake vigorously. The shaking presumably breaks down the interlobular fibrous tissue by a combination of shear force induced by the movement of tissue through the shaking process, assisted by numerous blows from the steel balls. Intuitively, one would expect some islets would be destroyed rather than released by such a battering.
Methods: In an attempt to improve the efficiency of islet isolation we have designed a new digestion/filtration chamber that consists of a glass cylinder, sealed with Teflon plates holding in mesh filters at each end, secured in place by a central threaded tie-rod and external knurled nuts. A ring-shaped piston within the cylinder can be pushed up and down the travel by two rods passing out through sealed ports in the Teflon disk at one end and connected to an external handle. The handle is used to gently push the piston up and down the travel of the cylinder, which pushes the fluid and tissue through the central lumen of the ring-piston. A series of hooks attached to the central tie-rod catch the fibrous strands of the passing tissue; the shearing forces produced cause disruption by a process thought to be similar to teasing the tissue apart with fine forceps.
Results: A series of initial experiments with human pancreas showed the prototype to be too large, causing temperature control problems, and a redesigned smaller chamber was produced, maintaining the crucial design features. Experience processing five human pancreata has now demonstrated that in three of five pancreata the new chamber produced a good yield (>200,000 I.E.) of remarkably well separated and intact islets, the entire dispersion process being under 1 hour. However, in two isolations the collagenase digestion was poor, with few free islets. A copy of the new chamber (reserved for porcine work only) has been produced, as well as a copy of the Ricordi chamber. We have confirmed that the new chamber can isolate porcine islets in large numbers (>5000 islets/g pancreas [n = 2], but note that pig islets are small).
Conclusion: These preliminary studies are sufficiently encouraging to justify further direct comparison with the Ricordi chamber for the purpose of animal and human islet isolation.