Background: Primary nonfunction resulting in immediate graft loss is responsible for the failure of a large number of islet transplants. Evidence is accumulating to single out endotoxin contamination of the various reagents needed for islet isolation as a major cause of early graft loss.
Methods: Islets isolated with endotoxin-containing (400 endotoxin units/ml) collagenase type V and "endotoxin-free" (3.1 endotoxin units/ml) Liberase were compared. Graft function was assessed using a syngeneic murine model of marginal islet mass transplantation. Pro-inflammatory cytokine production by islets was measured by ELISA in culture supernatants, and quantitative reverse transcriptase-PCR. Islet cell apoptosis was measured using the annexin assay.
Results: Graft function was significantly delayed when islets were isolated with endotoxin-containing collagenase. Addition of endotoxin to the Liberase solution similarly delayed graft function. After 18 hr in culture, collagenase-isolated islets released higher amounts of proinflammatory cytokines compared with Liberase-isolated islets (interleukin-6: 2,185+/-1,174 pg/ml vs. 520+/-201 pg/ml; tumor necrosis factor-alpha: 304+/-298 pg/ml vs. 0; IL-1beta: 12.5 pg/ml+/-12.5 vs. 0). This observation correlated with higher cytokine mRNA expression in collagenase-isolated islets. The percentage of apoptotic islet cells immediately after isolation was 17.2%+/-9.4 in collagenase-isolated islets and 7.1%+/-2.1 in Liberase-isolated islets.
Conclusions: We propose that endotoxin contamination is the primum movens of a chain of events that involves intra-islet cytokine production and release and islet cell apoptosis, and endotoxin contamination can ultimately lead to primary nonfunction in vivo. This emphasizes the fact that using endotoxin-free reagents during isolation is a key factor for successful islet transplantation.