A new method to valued efficiency of enzyme blends for pancreas tissue digestion

Abstract

One of the best successful examples of cell therapy is represented by islet transplantation since 1990. However islet isolation methods are not completely standardized yet. More than half of isolation procedures failed to isolate adequate islets for transplantation, due to variable pancreas conditions and to unpredictable enzymatic blend efficiency. Enzymes used for pancreas digestion include collagenases and neutral proteases: their composition and activity are largely variable between different batches. We set up a new in vitro method to better in vitro characterize enzymatic blend before its use in human pancreas. In our experimental approach human immortalized cells (ECV-304) or human islets were coltured within a 3-D type-I collagen gel in 96 wells plate. After one culture day, cells and/or islets were treated with different commercial enzymes (Liberase, Serva NB1 premium grade, Collagenase type P, Thermolysin, Neutral proteases) from different batches at different concentrations and for different times. Digestion of 3-D type-I collagen fibril gels were monitored by optical dense absorption to a fix l; morphology of released cells and/or islets were valued by confocal microscopy analyses. Cells were immunostained about expression of some adhesion molecules, like: integrins, cadherins and associated molecules, catenins, to appreciate cells morphology and islets aggregation modifications. Cell viability was assessed by SYTO 13 and ethidium bromide. We found that Neutral proteases is less pure and more toxic than Thermolysin: it has collagenase activity and it significantly decreases cell viability. Even viable cells after neutral proteases showed an alterated morphology with an impairment of cell to cell communications. We, also, observed an higher efficiency in extraction by Serva NB1 compared to Liberase and Collagenase type P, used in the same experimental conditions. By this in vitro method we were able to compare the minimal active enzyme concentration of different enzyme blends: we found that Liberase has its best value as number of extracted cell/alive cells in 130 mg/ml; while, in actual protocols for human pancreas is used to 1.3 mg/ml. Preliminary results showed a correlation between data achieved in cell lines with those of human islets, thus confirming the potential predictive role of this method for the selection of enzymes for human pancreas digestion purpose.