That which does not kill us makes us stronger--does Nietzsche's quote apply to islets? A re-evaluation of the passenger leukocyte theory, free radicals, and glucose toxicity in islet cell transplantation.

In clinical islet transplantation, isolated islets are embolized into the liver via the portal vein (PV); however, up to 70% of the islets are lost in the first few days after transplantation (i.e., too quickly to be mediated by the adaptive immune system). Part of early loss is due to instant blood-mediated inflammatory reaction, an immune/thrombotic process caused by islets interacting with complement. We have shown that glucose toxicity (GT) also plays a critical role based upon the observation that islets embolized into the PVs of diabetic athymic mice are rapidly lost but, if recipients are not diabetic, the islet grafts persist. Using donor islets resistant to the β-cell toxin streptozotocin, we have shown that intraportal islets engrafted in non-diabetic athymic mice for as little as 3 days will maintain normoglycemia when streptozotocin is administered destroying the recipient's native pancreas β-cells. What is the mechanism of GT in β-cells? Chronic exposure to hyperglycemia over-exerts β-cells and their electron transport chains leak superoxide radicals during aerobic metabolism. Here we reinterpret old data and present some compelling new data supporting a new model of early intraportal islet graft loss. We hypothesize that diabetes stimulates overproduction of superoxide in both the β-cells of the islet grafts and the endothelial cells lining the intraportal microvasculature adjacent to where the embolized islets become lodged. This double dose of oxidant damage stresses both the islets, which are highly susceptible to free radicals because of inherent low levels of scavenging enzymes, and the adjacent hepatic endothelial cells. This, superimposed upon localized endothelial damage caused by embolization, precipitates inflammation and coagulation which further damages islet grafts. Based upon this model, we predict that pre-exposing islets to sub-lethal hyperoxia should up-regulate islet free radical scavenging enzyme levels and promote initial engraftment; reinterpretation of 30 years old "passenger leukocyte" data and preliminary new data support this. Other data suggests that pre-exposure of recipients to hyperoxia could up-regulate antioxidant enzymes in the hepatic endothelium. The combination of both effects could markedly enhance early intraportal islet graft survival and engraftment. Finally, if our model is correct, current in vitro and in vivo tests used to test batches of harvested islets for viability and function prior to transplantation are poorly conceived (n.b., it is already well-known that results using these tests often do not predict clinical islet transplantation success) and a different testing paradigm is suggested.