According to these observations, 50 to 70% of endothelial cells and 40 to 60% of hepatocytes appear to undergo apoptosis during reperfusion.12,59 Further, a high percentage of apoptotic hepatocytes were identified in human liver allografts. However, in

all these studies, there was an undue reliance on the TUNEL (terminal deoxynucleotidyl transferase-mediated dUTP nick end labelling) assay to characterise “apoptosis.” Of note, TUNEL stains any cell with DNA strand breaks, irrespective of whether the mode of cell injury is apoptosis (in which abundant DNA strand breaks are a feature) or necrosis. Jaeschke et al. reasoned that if apoptosis was the primary mechanism of cell death during hepatic IR injury, inhibition of caspases should be highly effective at reducing injury, yet this manoeuvre conferred only 50% protection.18 They then applied strict morphological criteria for apoptosis MAPK inhibitor in their histological analyses, which allowed them to demonstrate conclusively that hepatocytes actually appear to undergo necrosis during IR, while only a small minority of sinusoidal endothelial cells and hepatocytes (< 2%) apoptose.18 While it is well known that necrotic

cell death causes inflammation with concomitant hepatic inflammation, it has been unclear what initiates this response. Early activation of the complement cascade by release of cell content by damaged hepatocytes by ischemia can trigger KC activation. Complement can also stimulate formation of ROS by KCs and in turn, directly activate, MI-503 in vivo recruit neutrophils to the hepatic sinusoid. A mediator specifically released by necrotic, but not apoptotic cells, is high mobility group box 1 (HMGB1), which is a nuclear factor bound to chromatin.60 HMGB1 has been implicated to be an early mediator of injury and inflammation 上海皓元 in warm IR injury; it has been reported to bind to TLR4 on KCs and stimulate pro-inflammatory cytokine release.60

A neutralizing antibody to HMGB1 was shown to be protective only in wildtype mice, in contrast to TLR4 knockout animals.60,61 Further, exogenous administration of HMGB1 aggravated injury in TLR4-intact animals, but not in TLR4 null mice. This TLR4 response was thought to be modulated by HO-1 and signal transducer and activator of transcription (STAT)-4.57,60,61 Thus, inhibition of HMGB1 attenuated pro-inflammatory cytokine release, reduced neutrophil infiltration and protected against liver IR injury. Receptor for advanced glycation end products (RAGE) is a family of pattern recognition receptors recently reported to be important in hepatic IR injury. Within the liver, RAGE is expressed on dendritic cells and to a lesser extent on KCs.62 Hepatic IR increases expression of RAGE as well as engagement of receptors by HMGB1 and DAMPs with subsequent injury62 (Fig. 3).