A310 Background: Poor initial graft function remains a major clinical problem in human liver transplantation. Its incidence (15 to 25% of transplanted patients) appears to depend on the length of cold storage and has thus been ascribed to injury from harvesting, cold storage and warm reperfusion after blood vessel reconnection; i.e. cold ischemia/warm re-perfusion (CI-WR) injury. The underlying mechanism of CI-WR injury has been studied extensively over the last decade, but it remains poorly understood. Recently, it has been demonstrated that different time courses and mechanisms of cell death occur in rat livers after CI-WR, with early sinusoidal endothelial cell necrosis followed by delayed hepatocyte apoptosis. These processes did not appear to be of major importance in the mechanism of graft failure since both were similar under non-lethal and lethal conditions. By contrast, prolonged CI was associated with an alteration of the extracellular matrix (ECM), leading to extensive coagulative necrosis of hepatocytes as found in conditions ending in graft failure. Aims: Since it is known that warm ischemia/WR of livers induces an increased local synthesis of metalloproteinases (MMPs), proteases implicated in the degradation of the hepatic ECM, we looked if the same phenomenon was occurring during prolonged CI/WR of rat livers ensuing in ECM disappearance responsible for the graft malfunction. Methods: Rat livers were preserved for 0, 24 (viable livers) and 42 h (non-viable livers) in the UW solution (4°C), then reperfused for 0, 10, 20, 30, 40, 60, 90, and 120 minutes with a krebs buffer solution (37°C) containing 20% erythrocytes. The first 15 mLs of the liver rinsing solution were taken during early WR and frozen liver biopsies were obtained at the end of each period (n=4 in all groups). Total liver RNAs and proteins were extracted, then MMP and their natural inhibitor (TIMP) expression was analysed by Northern Blot quantitative analysis and MMP activity was nalyzed using a fluorescent substrate (MCA-MAT) or by zymography. Results: although most of these genes are not expressed in the normal rat liver, there was an induction of MMP-13 and 14 expression as well as of TIMP-1 and 2 expression in all livers, independently from CI or WR periods. There was also an increased MMP activity in all livers, as measured ding a in presence or absence of an inhibitor (BB94), whatever the experimental condition. An increased activity of MMPs (particularly gelatinases A and B) was found in the initial flush solution. Comments: These data are quite different from those observed following warm ischemia/WR, particularly when considering that, following CI, the ECM rapidly deteriorates with WR. However since this phenomenon occurs very early during WR, it might be due to the liberation of MMPs already present in sinusoidal endothelial cells and released by their necrosis during prolonged CI, as suggested by our data and others. Conclusion: Induction of hepatic MMP synthesis does not occurs following prolonged CI/WR of rat livers and thus has no major role in the alteration of the hepatic extracellular matrix observed under these conditions. This phenomenon might result from the local passive release of preformed MMPs from dead hepatic cells.