Apoptosis is characterized by cell shrinkage, nuclear condensation, DNA fragmentation, and apoptotic body formation. These features separate apoptosis from other forms of cell death, such as necrosis (Kerr et al., 1974). It is well known that shrunken hepatocytes are more susceptible to stress-induced cell death, which may explain the clinical observation that transplantation oflivers from hypernatremic donors shows high rates of primary graft dysfunction. Apoptotic volume decrease (AVD) is an early and ubiquitous event in apoptotic cell death (Bortner and Cidlowski, 2001, 2002; Lang et al., 2000a; Yu et al., 2001). Ion fluxes and cellular mechanisms of volume regulation participate in the regulation ofAVD as well as in apoptosis induction. For example, in Jurkat cells, CD95 activation inhibits Na+/H+-antiport (Lang et al., 2000a), the Na+/K+-ATPase (Nobel et al., 2000), activates an outwardly rectifying chloride channel (Szabo et al., 1998), and stimulates efflux of taurine (Lang etal., 1998), resulting in AVD. Furthermore, inhibition of AVD by inhibition of volume-sensitive K+ or Cl~ channels (Maeno et al., 2000) or by elevation of extracellular K+ (Gomez-Angelats et al., 2000; Thompson )

counteracts apoptosis, whereas inhibition of Na+/H+ exchange accelerates apoptotic cell death (Lang et al., 2000a).

Apparently, in hepatocytes, hyperosmolarity-induced CD95 activation and CD95 activation induced by the CD95 ligand (CD95L) or proapopto-tic bile salts share almost the same signaling events, which finally lead to activation of the apoptotic machinery. Thus the question arises at which point in the respective intracellular signaling cascades induced by these distinct stimuli does signal transduction converge? One early signaling event shared by all these stimuli was shown to be an almost instantaneously generated oxidative stress response (Reinehr et al., 2003a,b).

In rat hepatocytes, oxidative and osmotic stresses are apparently closely interlinked. On the one hand, hyperosmotic cell shrinkage generates oxidative stress (Reinehr et al., 2003a,b) and on the other, oxidative stress induces cell shrinkage (Becker et al., 2007a; Hallbrucker et al., 1994;

Saha et al., 1993). Apparently the balance between intracellular metabolic H2O2 generation and its removal by detoxication systems such as catalase and glutathione peroxidase may be one determinant for hepatocellular K+ balance and, accordingly, cell volume. Hyperosmolarity, as well as other proapoptotic stimuli, such as CD95L or proapoptotic bile salts, was reported to trigger an almost instantaneous oxidative stress response in rat hepato-cytes (Reinehr et al., 2003a,b). Like tumor necrosis factor-a (Stadler et al., 1992), proapoptotic bile acids were shown to induce oxygen radical formation by mitochondria (Krahenbuhl et al., 1994; Sokol et al., 1995, 1998); however, this may represent a downstream consequence, not the cause of CD95 activation. Similar considerations may apply for the oxygen radical formation due to an endoplasmic reticulum stress response, which is induced by hydrophobic bile acids (Crowley et al., 2000; Sokol et al., 1998). In rat hepatocytes, NADPH oxidase isoforms, which are found in many tissues and exhibit sequence homology to the classical phagocyte NADPH oxidase gp91phox (Nox 2) (Lambeth, 2002), have been identified as the source of the instantaneous reactive oxygen species generation upon proapoptotic stimulation, that is, hyperosmolarity, CD95L, or proapoptotic bile salts (Reinehr et al., 2005a,b, 2006). NADPH oxidase isoforms, which are homologues of the gp91phox, are called Nox and Duox (Edens et al., 2001; Lambeth, 2002; Lambeth et al., 2000) and participate in a variety of signal transduction cascades (Lambeth et al., 2000). NADPH oxidases are activated by a self-assembly, regulated by proteins such as p47phox, p67phox, and Rac (Bokoch et al., 2002; Nauseef et al., 2004; Pani et al., 2001; Vignais, 2002). Thereby, p47phox is thought to be critical for normal NADPH oxidase function because p47phox acts as an adapter protein, which facilitates stimulus-induced binding ofp67phox to the enzyme complex (Nauseef et al., 2004; Vignais, 2002). Evidence has been given that the small G protein Rac acts as an upstream signaling event of hyperosmotically induced caspase 3 activation in NIH 3T3 fibroblasts (Friis et al., 2005).

Within the last decade, the contribution of AVD and ROS formation to apoptotic signal transduction has been under intense investigation. Hyper-osmolarity mimics AVD and induces ROS formation, thereby stimulating CD95-mediated apoptosis in hepatocytes; other death receptors were also shown to be activated upon hyperosmotic exposure. For example, hyperosmotic exposure can induce JNK activation and TNF receptor clustering (Rosette and Karin, 1996). This, however, is cell type specific and may depend on the efficacy ofvolume-regulatory increase mechanisms (Bortner and Cidlowski, 1996; Reinehr et al., 2003b). As outlined in more detail later, hyperosmotic cell shrinkage in rat hepatocytes triggers ROS formation, which leads to CD95 activation, oligomerization, and subsequent membrane trafficking, thereby sensitizing hepatocytes toward CD95L-induced apoptosis (Reinehr et al., 2002, 2003b).

2. Hyperosmotic Activation of the CD95 System in Hepatocytes

Whereas CD95 in normosmotically exposed rat hepatocytes is localized inside the cells and exhibits no detectable membrane localization, this death receptor is rapidly targeted to the plasma membrane in response to proapoptotic stimuli, such as hyperosmotic exposure, CD95L, or bile acids (Eberle et al., 2005, 2007; Reinehr and Haussinger, 2004; Reinehr et al., 2002, 2003a,b, 2004a,b,c). An easy way to visualize CD95 translocation to the plasma membrane is to detect CD95 by immunocytochemistry under permeabilizing conditions, which allow for intracellular and membrane CD95 staining, compared to nonpermeabilizing conditions, which allow for plasma membrane CD95 staining only (Reinehr et al., 2002). Other ways to detect CD95 membrane trafficking are Western blot analysis of membrane and cytosolic fractions using ultracentrifugation (Reinehr et al., 2003b) or transfection of a yellow fluorescent protein (YFP)-coupled CD95 receptor and subsequent analysis of living cells by fluorescence microscopy (Reinehr et al., 2004c) (Figs. 8.1 and 8.2).

Hyperosmotic CD95 activation and subsequent translocation of the CD95 to the plasma membrane are complex processes. Hyperosmotic exposure leads to an almost instantaneous increase of ceramide levels measured by lipid extraction and subsequent high-performance thin-layer chromatography, which was sensitive to inhibition of the acidic sphingo-myelinase (ASM) using inhibitors such as AY9944 or desipramine or by ASM protein knockdown using ASM-specific antisense oligonucleotides (Reinehr et al., 2006).

Hyperosmotic hepatocyte shrinkage increases the cytosolic Cl- concentration (Graf and Haussinger, 1996; Haussinger etal., 1990; Lang etal., 1993) not only because ofthe almost immediate osmotic water efflux in response to hyperosmolarity, which has a concentrative effect on cytosolic chloride concentration, but also because of the ionic mechanisms of volume regulatory increase. The latter involves a hyperosmotic activation of Na+/H+ exchange and Na+/K+-ATPase together with HCO3~/Cl~ exchange, resulting in net accumulation of Na+, K+, and Cl~ in rat hepatocytes (Graf and Haussinger, 1996; Haussinger et al., 1990; Lang et al., 1993). Evidence has shown that hyperosmotic ASM activation is probably regulated by changes in the intracellular chloride concentration ([Cl-];). In hepatocytes, ASM localizes at the cellular membrane or in the lysosomes (Goni and Alonso, 2002; Ohanian and Ohanian, 2001), but also in an early endosomal compartment with an apparent pHves of about 6.0 (Reinehr et al., 2006). In order to detect pH changes in this presumably endosomal compartment, hepatocytes were allowed to endocytose FITC-coupled dextran molecules

Was this article helpful?

0 0

Post a comment