When incubated at 4°C, cultured rat hepatocytes or liver endothelial cells exhibit pronounced injury and, during earlier rewarming, marked apoptosis. Both processes are mediated by reactive oxygen species, and marked protective effects of iron chelators as well as the protection provided by various other antioxidants suggest that hydroxyl radicals, formed by classical Fenton chemistry, are involved. However, when we measured the Fenton chemistry educt hydrogen peroxide and its precursor, the superoxide anion radical, formation of both had markedly decreased and steady‐state levels of hydrogen peroxide did not alter during cold incubation of either liver endothelial cells or hepatocytes. Similarly, there was no evidence of an increase in O ₂ ⁻ /H ₂ O ₂ release contributing to cold‐induced apoptosis oc‐curring on rewarming. In contrast to the release/ level of O ₂ ⁻ and H ₂ O ₂ , cellular homeostasis of the transition metal iron is likely to play a key role during cold incubation of cultured hepatocytes: the hepatocellular pool of chelatable iron, measured on a single‐cell level using laser scanning microscopy and the fluorescent indicator phen green, increased from 3.1 ± 2.3 μ M (before cold incubation) to 7.7 ± 2.4 μ M within 90 min after initiation of cold incubation. This increase in the cellular chelatable iron pool was reversible on rewarming after short periods of cold incubation. The cold‐induced increase in the hepatocellular chelatable iron pool was confirmed using the calcein method. These data suggest that free radicalmediated hypothermia injury/cold‐induced apoptosis is primarily evoked by alterations in the cellular iron homeostasis/a rapid increase in the cellular chelatable iron pool and not by increased formation of O ₂ ⁻ /H ₂ O ₂ .—Rauen, U., Petrat, F., Li, T., de Groot, H. Hypothermia injury/coldinduced apoptosis—evidence of an increase in chelatable iron causing oxidative injury in spite of low O ₂ ⁻ 2/H ₂ O ₂ formation. FASEB J. 14, 1953–1964 (2000)

Although University of Wisconsin (UW) solution aims at the prevention of cold-induced cell injury, it failed to protect against cold-induced apoptosis of hepatocytes and liver endothelial cells: when incubated in UW solution at 4°C for 24 hours and subsequently rewarmed at 37°C, 72% ± 8% of rat hepatocytes and 81% ± 5% of liver endothelial cells lost viability. In both cell types, the observed cell damage occurred under an apoptotic morphology; it appeared to be mediated by a rapid increase in the cellular chelatable iron pool by a factor ≥2 (as determined in hepatocytes) and subsequent formation of reactive oxygen species (ROS). Consequently, this cell injury was decreased by iron chelators to 6 to 25% (hepatocytes) and 4% ± 2% (liver endothelial cells). Deferoxamine nearly completely inhibited the occurrence of apoptotic morphology in both cell types. In liver endothelial cells, cold-induced apoptosis occurring during rewarming after 24 hours of cold incubation in UW solution was far more pronounced than in cell culture medium (loss of viability: 81% ± 5% vs. 28% ± 13%), but viability could even be maintained for 2 weeks of cold incubation by use of deferoxamine. In conclusion, this pathological mechanism might be an explanation for the strong endothelial cell injury known to occur after cold preservation. With regard to the extent of this iron-mediated injury, addition of a suitable iron chelator to UW solution might markedly improve the outcome of liver preservation.

The molecular correlate of hypertonicity-induced cation channels (HICCs) and their role in proliferation vs. apoptosis is a matter of debate. We report in this paper that, in whole-cell patch-clamp recordings, hypertonic stress (340→450 mosM) reversibly increased the Na⁺ conductance of HepG2 cells from 0.8 to 5.8 nS. The effect was dose-dependently inhibited by flufenamate and amiloride, known blockers of HICCs, with some 50% efficiency at 300 μM. In parallel, both drugs decreased HepG2 cell proliferation [in 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays and with automatic cell counting]. Small interfering RNA (siRNA) silencing of the α-subunit of the epithelial Na⁺ channel (ENaC) reduced hypertonicity-induced Na⁺ currents to 60%, whereas the rate of HepG2 cell proliferation was approximately half of that of the control. Moreover, α-ENaC siRNA inhibited the regulatory volume increase of HepG2 cells (measured with scanning acoustic microscopy) by 60%. In florescence-activated cell sorting measurements, silencing of α-ENaC led to a significant decrease in the G1 and an increase in the G2/M phase of the cell cycle, whereas the S phase was not changing. Finally (determined by a caspase 3/7 assay), HICC inhibition by flufenamate and silencing of α-ENaC increased the rate of apoptosis in HepG2 cells. It is concluded that α-ENaC is one functional element of the HICC in HepG2 cells and that the channel is an important mediator of cell proliferation; likewise, HICC blockage shifts the system from a proliferative into a rather apoptotic one. This is the first report of a role of α-ENaC in cell proliferation.

Nitric oxide (NO) and hydrogen peroxide (H₂O₂) show cooperativity in their cytotoxic action. The present study was performed to decipher the mechanisms underlying this phenomenon. In cultured liver endothelial cells and in cultured, glutathione-depleted hepatocytes, the combined exposure to NO (released by spermine NONOate, 1 mM) and H₂O₂ (released by glucose oxidase) induced cell injury that was far higher than the injury elicited by NO or H₂O₂ alone. In both cell types, the addition of the NO donor increased H₂O₂ steady-state levels, although with different kinetics: in hepatocytes, the increase in H₂O₂ levels was already evident at early time points while in liver endothelial cells it became evident after ≥2 h of incubation. NO exposure inhibited H₂O₂ degradation, assessed after addition of 50 μM, 200 μM, or 4 mM authentic H₂O₂, significantly in both cell types. However, again, early and delayed inhibition was observed. The late inhibition of H₂O₂ degradation in endothelial cells was paralleled by a decrease in glutathione peroxidase activity. Glutathione peroxidase inactivation was prevented by hypoxia or by ascorbate, suggesting inactivation by reactive nitrogen oxide species (NO_x). Early inhibition of H₂O₂ degradation by NO, in contrast, could be mimicked by the catalase inhibitor azide. Together, these results suggest that the cooperative effect of NO and H₂O₂ is due to inhibition of H₂O₂ degradation by NO, namely to inhibition of catalase by NO itself (predominant in hepatocytes) and/or to inhibition of glutathione peroxidase by NO_x (prevailing in endothelial cells).