RNA oxidation and zinc in hepatic encephalopathy and hyperammonemia
- First Online:
- Cite this article as:
- Schliess, F., Görg, B. & Häussinger, D. Metab Brain Dis (2009) 24: 119. doi:10.1007/s11011-008-9125-2
- 213 Downloads
Hepatic encephalopathy is a neuropsychiatric manifestation of acute and chronic liver failure. Ammonia plays a key role in the pathogenesis of hepatic encephalopathy by inducing astrocyte swelling and/or sensitizing astrocytes to swelling by a heterogeneous panel of precipitating factors and conditions. Whereas astrocyte swelling in acute liver failure contributes to a clinically overt brain edema, a low grade glial edema without clinically overt brain edema is observed in hepatic encephalopathy in liver cirrhosis. Astrocyte swelling produces reactive oxygen and nitrogen oxide species (ROS/RNOS), which again increase astrocyte swelling, thereby creating a self-amplifying signaling loop. Astroglial swelling and ROS/RNOS increase protein tyrosine nitration and may account for neurotoxic effects of ammonia and other precipitants of hepatic encephalopathy. Recently, RNA oxidation and an increase of free intracellular zinc ([Zn2+]i) were identified as further consequences of astrocyte swelling and ROS/RNOS production. An elevation of [Zn2+]i mediates mRNA expression of metallothionein and the peripheral benzodiazepine receptor (PBR) induced by hypoosmotic astrocyte swelling. Further, Zn2+ mediates RNA oxidation in ammonia-treated astrocytes. In the brain of hyperammonemic rats oxidized RNA localizes in part to perivascular astrocyte processes and to postsynaptic dendritic spines. RNA oxidation may impair postsynaptic protein synthesis, which is critically involved in learning and memory consolidation. RNA oxidation offers a novel explanation for multiple disturbances of neurotransmitter systems and gene expression and the cognitive deficits observed in hepatic encephalopathy.
KeywordsAstrocytes Cell volume Oxidative stress Ammonia Brain Peripheral benzodiazepine receptor Metallothionein
Hepatic encephalopathy defines a broad spectrum of neuropsychiatric symptoms associated with acute and chronic liver failure, which all are potentially reversible (Brusilow 1986; Hazell and Butterworth et al. 1999; Häussinger and Schliess 2008). At the neurophysiological level hepatic encephalopathy is characterized by a disturbed cortico-cortical and cortico-muscular coupling (Timmermann et al. 2002; Timmermann et al. 2003). At the cellular level hepatic encephalopathy is considered as a primary gliopathy (Norenberg 1996).
Ammonia is a key toxin in hepatic encephalopathy (Norenberg 1996). Whereas in fulminant hepatic failure hyperammonemia contributes to the development of a cerebral edema (Vaquero et al. 2003), patients with liver cirrhosis and hepatic encephalopathy usually show no clinical signs of an overt cerebral edema and increased intracranial pressure. In 1994, in vivo proton magnetic resonance (1H-MRS) studies on human brain provided first evidence for a contribution of ammonia to the development of a low grade cerebral edema in cirrhotic patients (Häussinger et al. 1994). The recordings indicated an increase of brain glutamine/glutamate associated with a decrease of myo-inositol already in cirrhotics with subclinical hepatic encephalopathy, which was interpreted to reflect a (partially) compensated glial edema (Häussinger et al. 1994). The hypothesis was introduced, that astrocyte swelling integrates at least some of the pathogenetic signals released by ammonia and other hepatic encephalopathy-precipitating factors (Häussinger et al. 1994; Häussinger 2006; Häussinger and Schliess 2008). Indeed, precipitants of hepatic encephalopathy were shown to induce astrocyte swelling and/or to increase ammonia-induced astrocyte swelling (Norenberg et al. 1991; Bender et al. 1992; Bender and Norenberg 1998).
Swollen astrocytes predispose to neuronal dysfunction due to impairment of their protective homeostatic functions (Norenberg 1994; Hansson and Rönnbäck 1995; Kimelberg 1995; Sykova et al. 1999; Bezzi et al. 2001). Since 1994, a multitude of animal and human studies strengthened the view that a low grade cerebral edema due to a perturbed osmolyte homeostasis is an important pathogenetic event in hepatic encephalopathy (Ross et al. 1994; Cordoba et al. 1996; Morgan 1998; Cordoba et al. 2001; Shah et al. 2003, 2008; Shawcross et al. 2004; Minguez et al. 2006; Miese et al. 2006; Kale et al. 2006).
Osmotic and oxidative stress in ammonia toxicity to the brain
Ammonia in the brain is detoxified by the glutamine synthetase-catalyzed reaction, which is almost exclusively performed by the astrocytes (Martinez et al. 1977). Ammonia infusion into rats leads to an increase of brain glutamine and brain water, which is associated with astrocyte swelling (Takahashi et al. 1991; Willard-Mack et al. 1996; Brusilow 2002; Tanigami et al. 2005). In ammonia-infused portocaval anastomized rats the increase of brain water significantly correlated with the increase of cerebral blood flow that contributes to the development of brain edema (Larsen et al. 2001). Administration of methionine sulfoximine (MSO), a potent inhibitor of the glutamine synthetase (Eisenberg et al. 2000), not only prevented cerebral glutamine accumulation but also the increase of brain water and intracranial pressure (Blei et al. 1994). These and other observations suggest that glutamine at least in part mediates ammonia toxicity to the brain. On the other hand, it is well known that cell volume changes per se produce signals that are involved in the regulation of cell function at multiple levels including metabolism and gene expression (Häussinger and Lang 1992; Häussinger 1996; Schliess and Häussinger 2005; Schliess et al. 2007). It is therefore well conceivable that astrocyte swelling rather than glutamine accounts at least for some pathogenetic ammonia effects to the brain. Extensive surveys regarding the impact of astrocyte swelling on astrocyte function have been published earlier (Häussinger et al. 2000; Häussinger and Schliess 2008; Schliess et al. 2006a, b).
There is substantial evidence from animal and cell culture studies suggesting that ROS/RNOS play a major role in cerebral ammonia toxicity and hepatic encephalopathy (Norenberg 2003; Norenberg et al. 2004; Häussinger and Schliess 2008). Ammonia treatment of animals and cultured astrocytes increases the cerebral production of superoxide, nitric oxide (NO) and NO as well as protein tyrosine nitration (Hermenegildo et al. 1996; Master et al. 1999; Kosenko et al. 1999; Hermenegildo et al. 2000; Larsen et al. 2001; Schliess et al. 2002; Vaquero et al. 2003; Kosenko et al. 2004; Schliess et al. 2006a, b). Similarly, diazepam and endotoxin, which precipitate human hepatic encephalopathy, increased cerebral oxidative stress and protein tyrosine nitration in animals in vivo and in cultured astrocytes (Musavi and Kakkar 1998; Possel et al. 2000; Görg et al. 2003, 2006). Immunohistochemical analysis of brain slices from ammonia-, diazepam-, and endotoxin-treated rats unraveled that nitrotyrosine immunoreactivity was particularly enriched in perivascular astrocytes, which are constituents of the blood brain barrier (Schliess et al. 2002; Görg et al. 2003, 2006; Vaquero et al. 2003). Of interest, ammonia synergistically increases the protein tyrosine nitration of individual proteins induced by diazepam (Görg et al. 2003) and TNF-α (Häussinger and Schliess 2008) in cultured astrocytes.
Ammonia increases RNA oxidation in cultured astrocytes and in brain in vivo
Ammonia, hypoosmotic swelling, diazepam and TNF-α increase RNA oxidation in cultured rat astrocytes
A recent study demonstrated that ammonia potently increases RNA oxidation in cultured rat astrocytes (Görg et al. 2008). North-Western blot analysis of RNA with an anti-8OH(d)G antibody indicated that RNA oxidized in response to ammonia includes the ribosomal 28S and 18S RNAs, whereas the RT-PCR analysis of RNA purified with an anti-8OH(d)G antibody identified the glutamate aspartate transporter (GLAST)-encoding mRNA to be oxidized (Görg et al. 2008). On the other hand, oxidation of mRNAs encoding the γ-aminobutyric acid type B receptor subunit 1 (GABAB-R1b) and β-actin is not increased in response to ammonia (Görg et al. 2008), indicating some selectivity of RNA oxidation by ammonia. In line with this ammonia displays no effect on overall integrity of RNA as addressed by capillary electrophoresis (Görg et al. 2008). A selective RNA oxidation was also found in post mortem brain tissue of Alzheimer disease (AD) patients (Shan et al. 2003). This study even suggested that specific functional classes of mRNAs become oxidized, whose encoded proteins participate in AD pathogenesis (Shan et al. 2003). The reasons for the selective appearance of oxidized RNA are currently unknown but may include differential half-lifes, turnover rates, spatial conformations and accessibility to repair mechanisms. The mRNA abundance and specific sequence motifs are obviously not related to the extent of individual mRNA oxidation (Shan et al. 2003).
Of interest ammonia-induced RNA oxidation was reversible upon ammonia removal from the culture medium (Görg et al. 2008). This corresponds to the clinical observation of reversibility of hepatic encephalopathy upon treatment of the precipitating conditions. The mechanisms underlying the reversibility of ammonia-induced RNA oxidation are a matter of speculation but may include a decrease of the oxidative challenge combined with degradation and/or repair of oxidized RNA (Nunomura et al. 2006).
Pharmacological analysis of ammonia-induced RNA oxidation indicated the involvement of NMDA receptors, Ca2+ and the NADPH oxidase (Görg et al. 2008) that were also involved in the ROS/RNOS production stimulated by ammonia (Schliess et al. 2002; Reinehr et al. 2007). In addition polyphenon 60, a mixture of antioxidant polyphenols extracted from green tea (Saffari and Sadrzadeh 2004), inhibited NH4Cl-induced RNA oxidation in cultured astrocytes (Görg et al. 2008).
As demonstrated by hypoosmotic astrocyte exposure, astrocyte swelling was sufficient to increase RNA oxidation in cultured astrocytes (Görg et al. 2008). This points to the possibility that astrocyte swelling induced by ammonia contributes to ammonia-induced RNA oxidation. Also diazepam and TNF-α potently increased RNA oxidation in cultured astrocytes (Görg et al. 2008). This suggests that the action of different hepatic encephalopathy-precipitating conditions may integrate not only at the level of astrocyte swelling and ROS/RNOS production (Häussinger et al. 1994; Schliess et al. 2002, 2004; Häussinger 2006; Schliess et al. 2006a, b; Reinehr et al. 2007), but also at the level of RNA oxidation.
Ammonia-induced RNA oxidation in rat brain in vivo
Intraperitoneal injection of NH4Ac into rats increased RNA oxidation in rat brain in vivo (Görg et al. 2008). Whereas blood ammonia peaked after 30 min and thereafter declined to basal level, cerebral RNA oxidation was pronounced 6 and 24 h and was basal 72 h following NH4Ac injection (Görg et al. 2008). This indicates a reversibility of cerebral RNA oxidation by ammonia in vivo.
The cellular localization of NH4Ac-induced RNA oxidation in rat brain was addressed by immunohistochemical analysis (Görg et al. 2008). RNA oxidation was most pronounced in the neuronal somata but not in the cell nuclei throughout the cortical layers (Görg et al. 2008). Astroglial RNA oxidation was largely confined to perivascular astrocytic processes (Görg et al. 2008). The predominance of RNA oxidation in neurons is in line with earlier studies, which show that proteasome inhibitors or hydrogen peroxide in co-cultures of neurons and astrocytes produce a stronger RNA oxidation in the neurons (Ding et al. 2004; Shan et al. 2007). This suggests that astrocytes, compared to neurons, are fairly resistant to oxidative stress. The findings, however, do not exclude the possibility that astroglial ROS/RNOS account for neuronal RNA oxidation.
Potential impact of RNA oxidation on the pathogenesis of hepatic encephalopathy
mRNAs packaged together with proteins required for mRNA splicing and translation are transported along dendrites and can be released to dendritic spines upon stimulation of synapses (Schuman et al. 2006). This allows local synthesis of proteins such as Ca2+/calmodulin kinases, glutamate receptor subunits, elongation factors and the activity-regulated cytoskeleton-associated protein that account for synaptic plasticity (Schuman et al. 2006). Late phase long-term potentiation (L-LTP), an electrophysiological substrate for learning and memory consolidation, depends on stimulation of local protein synthesis (Kandel 2001).
The localization of oxidized RNA associated with the splicing protein NOVA-2 within RNA transport granules beneath post synapses in the hyperammonemic brain (Görg et al. 2008) (Fig. 3) suggests that ammonia interferes with the protein synthesis-dependent L-LTP. Thus, RNA oxidation may play a role in the pathogenesis of cognitive disturbances in hepatic encephalopathy and provide a link between astrocyte swelling and oxidative stress on the one hand and cognitive dysfunction on the other. In line with this, learning ability is disturbed in rats fed with a hyperammonemic diet (Erceg et al. 2005) and rats with portocaval anastomosis (Sergeeva et al. 2005). LTP is impaired in mouse brain slices exposed to ammonia (Chepkova et al. 2006) or TNF-α (Pickering et al. 2005). However, to what extent the oxidation of locally translated mRNA species contributes to the L-LTP impairment under these conditions remains to be established.
Hypoosmotic astrocyte swelling and ammonia increase intracellular free zinc in cultured astrocytes
Redox sensitivity of the intracellular free Zn2+ levels
Zinc in the periodic table is placed as a IIb group element. It is a transition metal and displays a negative standard reduction potential (−0.762 V; Zn → Zn2+ + 2e−) and in aqueous solutions exists as the two-fold charged ion (Zn2+).
In patients with liver cirrhosis, especially those with hepatic encephalopathy, serum Zn2+ levels are decreased (Vallee et al. 1957; Loomba et al. 1995; Yang et al. 2004; Rahelic et al. 2006). Experimental Zn2+ deficiency can precipitate HE in cirrhotic patients (van der Rijt et al. 1991) and in some studies oral Zn2+ supplementation improved the outcome of psychometric tests (Reding et al. 1984; Marchesini et al. 1996), Further, Zn2+ supplementation can counteract hyperammonemia by a stimulation of hepatic urea synthesis (Marchesini et al. 1996) and glutamine synthesis in skeletal muscle (Yoshida et al. 2001). In a rat model of fulminant hepatic failure the decrease of cerebral Zn2+ levels was associated with glutamate neurotoxicity (Zeneroli 1985; Zeneroli and Baraldi 1990), modifications in γ-amino butyric acid and serotonin action (Chetri and Choudhuri 2003) and a delay of visually evoked potentials, suggesting Zn2+ to be involved in neurotoxicity in hepatic encephalopathy (Hu and Chen 1992). Also in view of the astroglial ROS/RNOS production in response to ammonia and other hepatic encephalopathy-relevant toxins a pathogenetic role of an increased [Zn2+]i in cerebral ammonia toxicity and hepatic encephalopathy should be taken into account.
Swelling-dependent increase of [Zn2+]i in astrocytes
As demonstrated in a recent study (Kruczek et al. 2008) hypoosmotic astrocyte swelling rapidly increases the levels of cytosolic, mitochondrial and nuclear free Zn2+ in cultured rat astrocytes. The increase of [Zn2+]i is reversible upon re-institution of normoosmotic conditions and depends on NMDA receptor activation and nNOS-catalyzed NO synthesis (Kruczek et al. 2008). Further, the hypoosmotic [Zn2+]i increase is sensitive to antioxidant epigallocatechin gallate (Kruczek et al. 2008). Of interest, hypoosmotic swelling also increases [Zn2+]i in the human astrocytoma cell line MOG-G-CCM, whereas in the murine fibroblast cell line L929 hypoosmolarity is without effect on [Zn2+]i (Kruczek et al. 2008). This indicates cell type specificity of the [Zn2+]i response to cell swelling.
The hypoosmotic increase of [Zn2+]i is accompanied by a Zn2+-dependent translocation of the transcription factors MTF-1 and Sp1 into the nucleus (Kruczek et al. 2008). MTF-1 by a specific zinc finger motif with low affinity to Zn2+ senses [Zn2+]i elevation, which leads to MTF-1 translocation into the nucleus and transcriptional activation of a panel of genes including those encoding metallothionein 1 and 2 (Laity and Andrews 2007). In line with this hypoosmotic astrocyte swelling in a Zn2+-dependent manner increased the expression of mRNA encoding metallothionein 1 and 2 (Kruczek et al. 2008). Metallothionein protects astrocytes from metal toxicity and oxidative stress (Aschner 1996). Of interest, metallothionein induction by Zn2+ prevents astrocyte swelling by methylmercury due to attenuation of the methylmercury-induced Na+ uptake (Aschner et al. 1998). Thus, a swelling-dependent increase of metallothionein by hepatic encephalopathy-precipitating factors and conditions may counteract the vicious circle beween astrocyte swelling and ROS/RNOS and thereby contribute to reversibility of hepatic encephalopathy.
Sp1 is involved in transcriptional regulation of the peripheral-type benzodiazepine receptor (PBR) (Giatzakis and Papadopoulos 2004). Upregulation of the PBR was suggested to play a pivotal role in the pathogenesis of hepatic encephalopathy (Butterworth 2000, 2003). As shown earlier, hypoosmotic astrocyte swelling increases the number of PBR ligand binding sites in cultured astrocytes (Itzhak et al. 1994; Itzhak and Norenberg 1994). As shown recently (Kruczek et al. 2008), hypoosmotic astrocyte swelling increased PBR mRNA expression in a Zn2+-dependent manner. Further, experiments with the Sp1 inhibitor mithramycin A and a siRNA-mediated knockdown of Sp1 expression demonstrated that the hypoosmotic upregulation of PBR mRNA depends on Sp1 activation (Kruczek et al. 2008).
Increased PBR expression in cultured astrocytes may correspond well to the increase of cerebral PBR ligand binding sites observed in hepatic encephalopathy-relevant animal models (Giguere et al. 1992) and autopsy samples from cirrhotic patients with hepatic encephalopathy. PBR upregulation may enhance the synthesis of neurosteroids with positive GABAergic activity (Jackson et al. 2001) and increase the production of ROS/RNOS and astrocyte swelling (Panickar et al. 2007).
Besides hypoosmolarity also ammonia, diazepam and TNF-α increase [Zn2+]i in cultured astrocytes (Kruczek et al. 2008). This suggests that the action of different hepatic encephalopathy-precipitating conditions may integrate not only at the level of astrocyte swelling and the production of ROS/RNOS (Häussinger et al. 1994; Schliess et al. 2002, 2004; Häussinger 2006; Schliess et al. 2002, 2004, 2006a, b, 2007; Reinehr et al. 2007), but also at the level of [Zn2+]i. It is currently unknown to which extent astrocyte swelling contributes to the [Zn2+]i increase by ammonia, diazepam and TNF-α.
An increase of [Zn2+]i is involved in RNA oxidation by ammonia
In order to address the involvement of Zn2+ in the ammonia-induced production of oxidative stress, the effect of N,N,N’,N’-Tetrakis(2-pyridylmethyl) ethylenediamine (TPEN), a specific Zn2+ chelator (Arslan et al. 1985; Kruczek et al. 2008) on RNA oxidation in cultured rat astrocytes was investigated. RNA oxidation was monitored by Slot Blot analysis with an 8-OH(d)G-specific antibody as reported recently (Görg et al. 2008).
It is suggested that an astroglial increase of [Zn2+]i plays a role in signaling towards increasing levels of ROS/RNOS and RNA oxidation.
Ammonia and different hepatic encephalopathy precipitants trigger astrocyte swelling and the production of ROS/RNOS, which create a self-amplifying signaling loop under the participation of NMDA receptors, NADPH oxidase, glutamine synthetase and the PBR (Häussinger 2006; Schliess et al. 2006a, b) (Fig. 1). Under normal conditions the activation of this loop is limited by a marked volume-regulatory and antioxidative capacity of the astrocytes. However, in liver disease this loop may exacerbate and its outcome includes protein tyrosine nitration and phosphorylation and changes of gene expression.
Recently an increase of [Zn2+]i and RNA oxidation were identified as new consequences of astrocyte swelling and astroglial ROS/RNOS production under hepatic encephalopathy-relevant conditions (Kruczek et al. 2008; Görg et al. 2008) and, as shown in this article Zn2+ plays a crucial role in RNA oxidation induced by ammonia in cultured astrocytes. This suggests a new role of zinc in signal tranduction in hepatic encephalopathy and cerebral ammonia toxicity. The increase of [Zn2+]i may further increase ROS/RNOS production and thereby amplify the loop between swelling and oxidative stress in astrocytes.
RNA oxidation represents a novel mechanism in the pathophysiology of hepatic encephalopathy, which could well explain the multiple disturbances of neurotransmitter systems and gene expression in hepatic encephalopathy, thereby providing a link between ammonia-induced oxidative stress and cognitive dysfunction through impairment of local protein synthesis. The profiling of oxidized mRNAs in the brain may be crucial for further understanding of the pathomechanisms underlying hepatic encephalopathy and ammonia toxicity to the brain.