Acetaminophen (AAP) overdose can cause severe liver injury and liver failure in experimental animals and humans. Recently, several authors proposed that apoptosis might be a major mechanism of cell death after AAP treatment. To address this controversial issue, we evaluated a detailed time course of liver injury after AAP (300 mg/kg) in fasted C3Heb/FeJ mice. Apoptotic hepatocytes were quantified in H&E-stained liver sections using morphologic criteria (cell shrinkage, chromatin condensation and margination, and apoptotic bodies). The number of apoptotic hepatocytes remained at baseline (0.2 +/- 0.1 cells/10 high-power fields [HPF]) up to 2 h after AAP administration. However, between 3 and 24 h, apoptotic cell death increased significantly, e.g., 6.3 +/- 0.8 cells/10 HPF at 6 h. Despite the increase in the number of hepatocytes meeting the morphological criteria of apoptosis, this cell fraction remained well below 1% of all parenchymal cells. No evidence for caspase-3 processing or increase in enzyme activity was detected at any time. These results were compared to the overall percent of necrotic cells in liver sections. Confluent areas of centrilobular necrosis were estimated to involve 40-60% of all hepatocytes between 3 and 24 h after AAP administration. These numbers correlated with the increase in plasma alanine aminotransferase activities, which reached a peak level of 5900 +/- 1350 U/l at 24 h. A similar result was obtained with higher doses of AAP and with the use of fed animals. Thus, oncotic necrosis and not apoptosis is the principal mechanism of liver-cell death after AAP overdose in vivo.

Obstruction of the common bile duct in a variety of clinical settings leads to cholestatic liver injury. An important aspect of this injury is hepatic inflammation, with neutrophils as the prominent cell type involved. However, the pathophysiologic role of the infiltrating neutrophils during cholestatic liver injury remains unclear. Therefore, we tested the hypothesis that neutrophils contribute to the overall pathophysiology by using bile duct-ligated (BDL) wild-type animals and mice deficient in the ␤ 2 integrin CD18. In wild-type animals, neutrophils were activated systemically as indicated by the increased expression of Mac-1 (CD11b/CD18) and L-selectin shedding 3 days after BDL. Histologic evaluation (48 ؎ 10% necrosis) and plasma transaminase levels showed severe liver injury. Compared with shamoperated controls (< 10 neutrophils per 20 high-power fields), large numbers of neutrophils were present in livers of BDL mice (425 ؎ 64). About 60% of these neutrophils had extravasated into the parenchyma. In addition, a substantial number of extravasated neutrophils were found in the portal tract. In contrast, Mac-1 was not up-regulated and plasma transaminase activities and the area of necrosis (21 ؎ 9%) were significantly reduced in CD18-deficient animals. These mice had overall 62% less neutrophils in the liver. In particular, extravasation from sinusoids and portal venules (PV) was reduced by 91% and 47%, respectively. Immunohistochemical staining for chlorotyrosine, a marker of neutrophilderived oxidant stress, was observed in the parenchyma of BDL wild-type but not CD18-deficient mice. In conclusion, neutrophils aggravated acute cholestatic liver injury after BDL. This inflammatory injury involves CD18-dependent extravasation of neutrophils from sinusoids and reactive oxygen formation. (HEPATOLOGY 2003;38:355-363.)

Reperfusion injury can cause liver dysfunction after cold storage and warm ischemia. Recently it has been suggested that more than 50% of hepatocytes and sinusoidal endothelial cells (SEC) are undergoing apoptosis during the first 24 hours of reperfusion. The aim of our study was to quantify apoptotic and necrotic hepatocytes and apoptotic SEC after 60 or 120 minutes of warm, partial no-flow ischemia and 0 to 24 hours reperfusion in male SD rats. Apoptotic cells were identified by TUNEL assay in combination with morphological criteria. After 60 minutes of ischemia and 1 hour of reperfusion there was a significant increase of apoptotic hepatocytes (0.7 ؎ 0.1% vs. 0.3 ؎ 0.1% in controls) and SEC (1.5 ؎ 0.6% vs. 0.3 ؎ 0.1% in controls). The number of apoptotic SEC and hepatocytes was not different from controls at 6 hours or 24 hours of reperfusion. In contrast, the number of necrotic hepatocytes was quantified as 12 ؎ 2% at 1 hour, 34 ؎ 6% at 6 hours, and 57 ؎ 11% at 24 hours. These results correlated with the increase in plasma ALT levels at these time points. Longer (120 min) ischemia times did not affect the number of apoptotic cells but increased hepatocellular necrosis to 58 ؎ 4% at 6 hours reperfusion. No significant increase in caspase-3 activity and processing was detectable in any of these livers. Moreover, the caspase inhibitor Z-Asp-cmk (2 mg/kg IV) had no significant effect on reperfusion injury. Our results suggest that only a small minority of SEC and hepatocytes undergo apoptosis after 60 to 120 minutes of warm ischemia followed by 0 to 24 hours of reperfusion. Oncotic necrosis appears to be the principal mechanism of cell death for both cell types. (HEPATOLOGY 2001;33:397-405.)Ischemia-reperfusion injury is responsible for primary liver dysfunction and failure after transplantation, 1-5 after liver resection (Pringle maneuver), 6 or after hemorrhagic shock. 7 Cold storage appears to cause injury mainly to sinusoidal endothelial cells (SEC), 8,9 whereas warm ischemia affects hepatocytes and endothelial cells. 10 In addition, Kupffer cells are activated and primed, 11-13 which causes a substantial aggravation of the injury during the early reperfusion period. At the same time, activation of complement 14 and the formation of cytokines 15 and chemokines 16,17 lead to hepatic neutrophil sequestration and a neutrophil-dependent injury 18 several hours later. Injury by Kupffer cells and neutrophils is largely dependent on reactive oxygen species 19-21 and proteases. [22][23][24] Although the postischemic oxidant stress is not sufficient to kill hepatocytes by lipid peroxidation, 25 there is evidence that reactive oxygen species and proteases cause hepatocellular necrosis. 22,26,27 In recent years, several laboratories reported evidence for apoptotic cell death during hepatic ischemia and reperfusion. [28][29][30][31] According to these studies, 50% to 70% of endothelial cells and 40% to 60% of hepatocytes undergo apoptosis during reperfusion. 30,31 A high percentage of apoptotic hepatocytes were also ide...

Cholestasis-induced liver injury during bile duct obstruction causes an acute inflammatory response. To further characterize the mechanisms underlying the neutrophil-induced cell damage in the bile duct ligation (BDL) model, we performed experiments using wild-type (WT) and ICAM-1-deficient mice. After BDL for 3 days, increased ICAM-1 expression was observed along sinusoids, along portal veins, and on hepatocytes in livers of WT animals. Neutrophils accumulated in sinusoids [358 +/- 44 neutrophils/20 high-power fields (HPF)] and >50% extravasated into the parenchymal tissue. Plasma alanine transaminase (ALT) levels increased by 23-fold, and severe liver cell necrosis (47 +/- 11% of total cells) was observed. Chlorotyrosine-protein adducts (a marker for neutrophil-derived hypochlorous acid) and 4-hydroxynonenal adducts (a lipid peroxidation product) were detected in these livers. Neutrophils also accumulated in the portal venules and extravasated into the portal tracts. However, no evidence for chlorotyrosine or 4-hydroxynonenal protein adducts was detected in portal tracts. ICAM-1-deficient mice showed 67% reduction in plasma ALT levels and 83% reduction in necrosis after BDL compared with WT animals. The total number of neutrophils in the liver was reduced (126 +/- 25/20 HPF), and 85% of these leukocytes remained in sinusoids. Moreover, these livers showed minimal staining for chlorotyrosine and 4-hydroxynonenal adducts, indicating a substantially reduced oxidant stress and a diminished cytokine response. Thus neutrophils relevant for the aggravation of acute cholestatic liver injury in BDL mice accumulate in hepatic sinusoids, extravasate into the tissue dependent on ICAM-1, and cause cell damage involving reactive oxygen formation.