Stimulation of natriuretic peptide receptor C attenuates accumulation of reactive oxygen species and nitric oxide synthesis in ammonia-treated astrocytes

Abstract

J. Neurochem. (2010) 115, 1068–1076.

Abstract

Oxidative and nitrosative stress contribute to ammonia-induced astrocytic dysfunction in hepatic encephalopathy. Treatment of cultured astrocytes with 5 mmol/L ammonium chloride (‘ammonia’) increased the production of reactive oxygen species (ROS), including the toxic NADPH oxidase reaction product, •O₂⁻. Atrial natriuretic peptide (ANP), natriuretic peptide C and a selective natriuretic peptide receptor (NPR)-C ligand, cANP_(4–23), each decreased the total ROS content both in control cells and cells treated with ammonia. However, attenuation of •O₂⁻ accumulation by ANP and cANP_(4–23), was observed in ammonia-treated cells only and the effect of cANP_(4–23) was decreased when the NADPH oxidase-regulatory protein G_iα-2 was blocked with a specific anti-G_iα-2 antibody. Although in contrast to ANP, cANP_(4–23) did not elevate the cGMP content in control astrocytes, it decreased cAMP content and reduced the expression of G_iα-2, the NADPH oxidase-regulatory protein. The results show the presence of functional NPR-C in astrocytes, activation of which (i) attenuates basal ROS production, and (ii) prevents excessive accumulation of the toxic ROS species, •O₂⁻ by ammonia. Ammonia, ANP and cANP_(4–23) added separately, each stimulated formation of NO_x (nitrates + nitrites) which was associated with up-regulation of the activity [cANP_(4–23)] or/and expression (ammonia) of the endothelial isoform of nitric oxide synthase. However, the ammonia-induced increase of NO_x was not augmented by co-addition of ANP, and was reduced to the control level by co-addition of cANP_(4–23), indicating that activation of NPR-C may also reduce nitrosative stress. Future hepatic encephalopathy therapy might include the use of cANP_(4–23) or other NPR-C agonists to control oxidative/nitrosative stress induced by ammonia.

Abbreviations used:

ANP

atrial natriuretic peptide

CNP

C-type natriuretic peptide

DHEt

dihydroethidium

eNOS

endothelial nitric oxide synthase

HE

hepatic encephalopathy

KB

Krebs buffer

NO

nitric oxide

NPR

natriuretic peptide receptors

NPs

natriuretic peptides

ONS

oxidative/nitrosative stress

RNS

reactive nitrogen species

ROS

reactive oxygen species

VSMC

vascular smooth muscle cell

Excessive accumulation of ammonia in the brain is a causative factor of hepatic encephalopathy (HE), a neuropsychiatric syndrome associated with acute or chronic liver failure (Felipo and Butterworth 2002). Brain edema related to astrocytic swelling is a major cause of death in patients with acute HE. According to the current theory, ammonia causes astrocytic swelling by triggering a vicious cycle of oxidative/nitrosative stress (ONS) (Schliess et al. 2006; Häussinger and Görg 2010) and intramitochondrial accumulation of glutamine (Albrecht and Norenberg 2006). In cultured astrocytes, ammonia induces formation of reactive oxygen and nitrogen species (ROS/RNS) (Murthy et al. 2001; Schliess et al. 2002), in a process which also involves nitric oxide (NO) formation by overactivation of N-methyl-d-aspartate receptor (Schliess et al. 2002; Zielińska et al. 2003). The major toxic ROS species formed in ammonia-treated cultured astrocytes or brain slices is the superoxide anion (•O₂⁻) stemming from activation of NADPH oxidase isoforms, the activation involving serine phosphorylation on the p47^phox regulatory subunit of NADPH oxidase (Reinehr et al. 2007).

Molecular consequences of increased ROS/RNS production include protein and nucleic acids oxidation (Görg et al. 2008), protein tyrosine nitration and S-nitrosylation (Schliess et al. 2004), and modifications of lipid peroxidation (Swapna et al. 2006; Singh et al. 2008), all of which contribute to the pathogenesis of HE. Therefore, attenuation of free radicals production may be a desired target of HE therapy. We speculated that natriuretic peptides (NPs) are the compounds with a potential to counteract ONS produced by ammonia in the CNS. The hypothesis was essentially based on the data showing that NPs attenuate excessive ROS production in the cardiovascular system and liver cells (Woodard and Rosado 2008; De Vito et al. 2010). Of note, all three NPs: atrial natriuretic peptide (ANP), brain natriuretic peptide and C-type natriuretic peptide (CNP) are also present in the brain (Herman et al. 1996; Potter et al. 2006). NPs exert their actions by binding to natriuretic peptide receptors (NPRs): ANP binds preferentially to NPR-A, and brain natriuretic peptide and CNP to NPR-B, respectively (Lucas et al. 2000). All NPs bind with equal affinity to NPR-C (Lucas et al. 2000). NPR-A and NPR-B are particulate, cell membrane-bound guanylyl cyclase receptors which mediate the signal by increasing cGMP. NPR-C, earlier ascribed a role of a ‘clearance receptor’ removing the peptides to the periphery does not possess guanylyl cyclase activity (Potter et al. 2006; Rose and Giles 2008) and appears to have a complex signaling mechanism of its own. The best known immediate consequence of NPR-C activation is inhibition of adenylyl cyclase (cAMP) via guanine nucleotide inhibitory proteins (G_i proteins) (Potter et al. 2006). However, NPR-C is also coupled to endothelial nitric oxide synthase (eNOS)-dependent signal transduction, in which NO activates a soluble form of guanylate cyclase and in turn leads to production of cGMP (Murthy et al. 1998).

In peripheral tissues, NPs were found to attenuate ROS production by interacting with all three receptor classes. Thus, in hepatocytes and Kupffer cells, ANP reduced oxidative stress by activation of NPR-A/B (Pella 1991; Kiemer and Vollmar 1998; Carini et al. 2003). In turn, enhanced oxidative stress in vascular smooth muscle cells (VSMC) from hypertensive rats, was reduced by activation of NPR-C with its specific agonist, cANP_(4–23) (Saha et al. 2008). In astrocytes, NPs have been reported to elicit physiologically meaningful downstream signals by interacting with NPR-A (Boran and Garcia 2007; Prado et al. 2010) or NPR-B (Zielińska et al. 2007), but neither the effects of NPs on ONS, nor functioning of NPR-C in general have been studied in these cells. Although the presence of NPR-C in astrocytes was identified on the basis of ligand displacement analysis (Sumners and Tang 1992), no detailed functional characteristics of the receptor was provided. Basing on the observation that cANP_(4–23) reduced accumulation of ROS in VSMC (Saha et al. 2008), we envisaged the possibility that NPs may attenuate ammonia-induced ROS formation and that this may involve interaction with NPR-C. Here, we report that (i) ANP and CNP reduces ROS formation both in ammonia-treated and -non-treated astrocytes, (ii) the ROS-reducing effect also occurred upon incubation with the NPR-C-interacting ANP analog, cANP_(4–23), and (iii) the superoxide anion was the major ROS species accumulating following ammonia treatment and its accumulation was reduced both by ANP and cANP_(4–23). The study documented the presence of a functional NPR-C receptor on astrocytes the stimulation of which counteracts oxidative stress in these cells. Moreover, cANP_(4–23) attenuated, by an as yet unknown mechanism, ammonia-induced accumulation of NO, and could in this way reduce nitrosative stress.

Materials and methods

Cell culture and treatments

Primary astrocyte cultures were prepared from cortices of newborn Wistar rats using the method described by Hertz et al. (1989), with some modifications. Cortex, dissected from brain of 1-day-old rat was passed through Nitex nylon netting (pore size – 80 μm) into Dulbecco’s modified Eagle’s medium containing 20% fetal bovine serum. Cells were grown at 37°C in a humified atmosphere of 90% air and 10% CO₂, in 6- and 24-well plates or in 60 mm dishes. Medium was changed 2 days after plating and subsequently twice a week gradually changing to 10% fetal bovine serum and supplemented with dibutyryl-cAMP from the second week. Experiments were performed on 3-week astrocytes non-treated (Control’) and treated with ammonium chloride (‘Ammonia’) as indicated in the legends. NPs were added 15 min before addition of ammonia.

Fluorimetric measurement of reactive oxygen species

The total content of the different ROS species was determined fluorimetrically by the method of Murthy et al. (2001) using a fluorescent probe, 5-(and-6)-carboxy-2′,7′-dichlorodihydrofluorescein diacetate (carboxy-H₂DCFDA) (Molecular Probes, Eugene, OR, USA; Invitrogen, Carlsbad, CA, USA), slightly modified. Cells were washed with standard Krebs buffer (KB) (118 mmol/L NaCl, 25 mmol/L NaHCO₃, 4.7 mmol/L KCl, 1.2 mmol/L KH₂PO₄, 2.5 mmol/L CaCl₂, 1.2 mmol/L MgSO₄, 10 mmol/L glucose, aerated with 95% O₂ and 5% CO₂ at pH 7.4) and incubated with 100 μmol/L carboxy-H₂DCFDA for 30 min at 37°C (stock solution: 10 mmol/L in dimethylsulfoxide). Cells were washed two times with ice-cold phosphate-buffered saline and disintegrated by scraping and sonication in 0.2% Triton-X100. Cell extracts were centrifuged for 5 min at 6.500 g and the 200 μL of supernatants were used to measure the fluorescence in a Fluoroscan Ascent-FL microplate fluorimeter. Fluorescence intensity was directly proportional to the intracellular ROS accumulation and was measured at an excitation wavelength of 485 nm and emission wavelength of 515 nm. Protein concentration was estimated by the Lowry method using Modified Lowry Protein Assay Reagent (Pierce, Rockford, IL, USA).

Detection of superoxide anion (•O₂⁻) by confocal laser scanning microscopy

Superoxide anion production was evaluated by measuring the fluorescence of dihydroethidium (DHEt) probe (Molecular Probes; Invitrogen). Before the experiments, cells were washed two times with KB and incubated for 1 h at 37°C in KB containing 50 μmol/L DHEt (stock solution: 1 mmol/L in dimethylsulfoxide). After loading, cells were washed three times and fresh KB was added. A confocal laser scanning microscope (Zeiss LSM 510) (ZEISS; Carl Zeiss GmBh, Jena, Germany) was used to obtain fluorescence imaging. A helium–neon laser (543 nm) was used for excitation of DHEt fluorescence (ex 518) and emission was observed at 605 nm wavelength. Three randomly chosen fields were scanned per each well. Following signal acquisition, analysis of cell fluorescence intensity (directly proportional to •O₂⁻ generation) was processed using the ImageJ free software (NIH, http://rsbweb.nih.gov/ij/).

RNA isolation and real-time PCR

Total RNA was isolated using TRI Reagent (Sigma, St Louis, MO, USA), then 1 μg was reverse-transcribed using the High Capacity cDNA Reverse Transcriptase Kit (Applied Biosystem, Life Technologies Corporation, Carlsbad, CA, USA). Real-time PCR was performed in 96-well plates with the ABI 7500 apparatus (Applied Biosystems) using the MGB Taqman probe assay. Primers and probes for eNOS and endogenous control β-actin were purchased from Applied Biosystems (Rn 02132634-s1 and Rn 00667869-m1, respectively). Each reaction contained 5 μL Taqman Universal PCR Mastermix in a total volume of 10 μL, and 1 μL cDNA was added to the reaction. The real-time PCR reactions were performed at 95°C for 10 min, followed by 40 cycles of 30 s at 95°C and 1 min at 60°C. The results of the analysis were calculated in relation to the β-actin product, and results were calculated according to, and expressed by an equation (2^−ΔΔCt) that gives the amount of target, normalized to an endogenous reference and relative to a calibrator. C_t is the threshold cycle for target amplification (Livak and Schmittgen 2001).

Protein isolation and western blot analysis

After 10 min of incubation at 4°C with Triton Lysis Buffer containing protease inhibitors, astrocytes were scraped off and centrifuged for 10 min at 12 000 g and 4°C. The supernatant was transferred to a new Eppendorf tube and used for further investigations as a cytosolic fraction. Total protein concentration was determined by the Lowry method using Modified Lowry Protein Assay Reagent (Pierce). Protein (40 μg) was boiled with sodium dodecyl sulfate–polyacrylamide gel sample buffer for 10 min, separated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis and then transferred onto nitrocellulose membrane. Immunodetection of proteins were made using SNAP i.d. Protein Detection System (Millipore Corporation, Bedford, MA, USA) according to SNAP i.d. system protocol. Blots were blocked with 0.25% non-fat dry milk in Tris-Buffered Saline Tween-20. Incubation with antibodies against G_iα-2 (1 : 250; Santa Cruz Biotechnology, Santa Cruz, CA, USA) was performed in Tris-Buffered Saline Tween-20 with 0.25% non-fat dry milk at 20°C for 0.5 h followed by 10 min of incubation with peroxidase-conjugated-anti-rabbit antibodies (1 : 2500; Sigma) for detection by SuperSignal West Pico Chemiluminescent Substrate (Pierce). The first antibody was stripped off with 0.1 M glycine, pH 2.9, and second incubation was performed with an antibody against β-actin (10 min incubation at 20°C, 1 : 3300; Sigma).

Determination of nitrates + nitrites

Total nitrites and nitrates concentrations (marker of NO production) were assessed using Nitrate/Nitrite Colorimetric Assay Kit (Cayman, Chemical Company, Ann Arbor, MI, USA), according to the manufacturer’s protocols. Protein determination was carried out by Bradford’s (1976) method.

Determination of NOS activity

Nitric oxide synthase activity was measured using NOS detect Assay Kit (Stratagene, La Jolla, CA, USA) strictly in accordance to the manufacturer’s protocol. The NOS activity was quantified by counting the radioactivity of the radiolabeled arginine conversion product citrulline.

Measurement of intracellular cAMP and cGMP levels

Determination of cAMP and cGMP was performed, respectively, with cAMP and cGMP Enzymeimmunoassay Biotrak (EIA) System (Amersham Biosciences, GE Healthcare, Uppsala, Sweden) according to the manufacturer’s protocol, modified as described below. After treatment, the cells were washed with phosphate-buffered saline and incubated for 3 min in the KB with 0.5 mmol/L of a non-specific phosphodiesterase inhibitor 3-isobutyl-1-methylxantine (Sigma) and then resuspended in 250 μL of the assay kit buffer containing 4 mmol/L EDTA and disrupted by sonication. Samples were boiled, centrifuged (14 000 g, 5 min) and cAMP or cGMP was measured in the supernatant. Protein determination in the pellets was performed according to Bradford’s (1976) procedure.

Statistical analysis

Graph Pad (Prism) software was used for statistical analysis of data: one-way analysis of variance (one-way anova) followed by Dunnett’s post hoc comparisons.

Results

ANP and cANP_(4-23) decrease the total ROS pool incontrol- and ammonia-treated astrocytes

Treatment with 5 mM ammonium chloride (‘Ammonia’) for 1 (Fig. 1a) or 24 h (Fig. 1b) increased accumulation of ROS in astrocytes to ∼24% above control, consistent with earlier observations in other laboratories (Murthy et al. 2001). Addition of ANP or CNP at 10⁻⁶ mol/L concentration attenuated ROS accumulation in non-treated (‘Control’) cells by ∼20% and ∼35%, respectively, and in cells treated with ammonia for 1 h by ∼30% and ∼35% (Fig. 1a). Treatment with ANP or CNP for 24 h reduced ROS generation in both control- and ammonia-treated cells, in a dose-dependent manner. In control cells, ANP at 10⁻¹⁰ and 10⁻⁶ mol/L reduced ROS to ∼89% and ∼56% of the initial value, respectively, whereas CNP at the same concentrations reduced ROS to 69% and to 50%, respectively. In cells treated for 24 h with ammonia, ANP at 10⁻¹⁰ and 10⁻⁶ mol/L reduced ROS by ∼28% and ∼61%, respectively, and CNP at the same concentrations decreased ROS by ∼48% and ∼57%, respectively (Fig. 1b).

Figure 1.  Effect of ANP and CNP on ROS generation in cultured primary rat astrocytes: not treated (Control) and treated with ammonium chloride (Ammonia) at 5 mmol/L for 1 (a) and 24 h (b). CNP or ANP were added to astrocytic culture medium 15 min before ammonia exposure. DCF fluorescence was determined as described in Materials and methods section. Results are mean ± SD (n ≥ 3). *p < 0.05, **p < 0.01 versus Control; ^†p < 0.05, ^††p < 0.01 versus Ammonia.

cANP_(4–23), a selective NPR-C agonist added at 10⁻⁶ mol/L concentration, decreased ROS accumulation by ∼25% in control cells, and by ∼33% in cells treated with ammonia for 1 h. Treatment with cANP_(4–23) for 24 h reduced ROS production by ∼18% in control cells and by ∼21% in ammonia-treated cells. Differences between the net effects of ANP and cANP_(4–23) were statistically insignificant both in the presence and absence of ammonia (p > 0.05) (Fig. 2).

Figure 2.  Effect of cANP_(4–23) on ROS generation in cultured primary rat astrocytes: not treated (Control) and treated with ammonium chloride (Ammonia) at 5 mmol/L for 1 and 24 h. cANP_(4–23), a selective NPR-C agonist, was added to astrocytic culture medium 15 min before ammonia exposure. DCF fluorescence was determined as described in Materials and methods section. Results are mean ± SD (n ≥ 3). **p < 0.01 versus Control; ^††p < 0.01 versus Ammonia.

ANP and cANP_(4-23) attenuate •O₂⁻ production in ammonia-treated but not in control astrocytes

A 24-h ammonia treatment increased DHEt fluorescence (a marker of •O₂⁻) in astrocytes from 75.3 ± 13.8 to 105.9 ± 8.2 [f.u.], well in agreement with previous reports (Reinehr et al. 2007; Jayakumar et al. 2009) (Fig. 3a, for illustration see Fig. 3b). Both ANP and cANP_(4–23) prevented the ammonia-induced increase of DHEt fluorescence (•O₂⁻ production) (Fig. 3a, for illustration see Fig. 3b).

Figure 3.  Effect of cANP_(4–23) and ANP pre-treatment on ammonia-induced superoxide (•O₂⁻) production in cultured primary rat astrocytes. cANP_(4–23) or ANP were added 15 min before ammonia exposure. DHEt fluorescence was obtained using a LSC microscopy and analyzed as described in Materials and methods section. (a) Data obtained from fluorescence intensity analysis. (b) Representative images from confocal laser microscope. Scale bar in all images – 100 μm. Data are mean ± SD (n = 3–5). **p < 0.01 versus Control; ^†p < 0.05 versus Ammonia.

Effects of cANP _4-23 on G_iα-2 protein expression and production of cAMP and cGMP: evidence for functionality of NPR-C in astrocytes

A 24-h incubation with cANP_(4–23) (10⁻⁶ mol/L) decreased G_iα-2 protein expression both in ammonia-treated and control cells, by ∼54% and ∼56%, respectively (Fig. 4a; protein separation shown in Fig. 4b). Addition of cANP_(4–23) (10⁻⁶ mol/L) for 24 h decreased cAMP by ∼44%. A 24-h exposure of astrocytes to ammonia increased cAMP by ∼15%. Treatment with cANP_(4–23) markedly decreased cAMP accumulation in control astrocytes, and in a much lesser degree in astrocytes treated with ammonia (Fig. 5a).

Figure 4.  Effect of cANP_(4–23) on G_iα protein expression in not-treated (Control) and treated with 5 mmol/L ammonium chloride (Ammonia) for 24 h cultured primary rat astrocytes. cANP_(4–23) was added 15 min before ammonia exposure. Analysis of protein expression was performed as described in Materials and methods section. Proteins were quantified densitometrically. (a) Data obtained from western blot analysis. (b) Representative image of G_iα protein immunostaining. Results are mean ± SD (n = 4). **p < 0.01 versus Control.

Figure 5.  Effects of ANP and/or cANP_(4–23) pre-treatment on cAMP (a) and cGMP (b) content in non-treated astrocytes (Control) and treated with 5 mmol/L ammonium chloride (Ammonia) for 24 h cultured primary rat astrocytes. Peptides were added 15 min before ammonia exposure. The basal value of cAMP and cGMP production in control astrocytes was 100.7 ± 18.3 and 9.9 ± 1.7 [fmol/mg of protein], respectively. Results are mean ± SD (n = 4). *p < 0.05, **p < 0.01 versus Control; ^†p < 0.05, ^††p < 0.01 versus Ammonia.

As expected, ANP, acting via NPR-A, increased about threefold the cGMP level in both ammonia-treated and control cells (Fig. 5b). By contrast, cANP_(4–23) failed to increase cGMP production, confirming its exclusive interaction with NPR-C.

Attenuation of •O₂⁻ production in ammonia-treated astrocytes by cANP_(4-23) does not occur in the presence of an antibody against the G_iα-2 protein.

As predicted from experiments of Fig. 3, incubation with cANP_(4–23) alone reduced •O₂⁻ accumulation evoked by ammonia (Fig. 6). Addition to the incubation of an antibody raised against the G_iα-2 protein brought the •O₂⁻ level back to that recorded in the presence of ammonia confirming the involvement of NADPH oxidase (Fig. 6).

Figure 6.  Effect of cANP_(4–23) on superoxide (˙O₂⁻) production in cultured primary rat astrocytes not-treated (Control) and treated with 5 mmol/L ammonium chloride (Ammonia) for 24 h in the presence or absence of antibody against _-2 protein. Antibody against G_iα-2 (1 μg/well of 24-well culture plate) was added 30 min before cANP_(4–23). DHEt fluorescence was obtained using a LSC microscopy and analyzed as described in Materials and methods section. Data are mean ± SD (n = 3). **p < 0.01 versus Control; ^†p < 0.05 versus Ammonia.

NO synthesis in control- and ammonia-treated astrocytes: role of eNOS and modulation by cANP_(4-23) and ANP

Treatment with ammonia for 1 or 24 h increased nitrites +nitrates (NO_x) (marker of NO) by ∼40% (Fig. 7a and b, respectively). A 1-h incubation with ANP (10⁻⁶ mol/L) of control astrocytes increased the accumulation of NO_x by ∼29%, whereas cANP_(4–23) did not affect this parameter. Upon 24 h treatment, both ANP and cANP_(4–23) enhanced generation of NO_x in control astrocytes by ∼47% and ∼52%, respectively. However, in ammonia-treated astrocytes, ANP did not significantly increase the NO_x concentration above the level generated by ammonia alone, whereas cANP_(4–23) actually reduced the stimulatory effect of ammonia, bringing the level down to that measured in cells not treated with ammonia (Fig. 7b). The increase of NO_x accumulation at 24 h incubation with ammonia was abolished when incubations were carried out with 100 μM l-N^G-nitroarginine, a concentration preferentially inhibiting the constitutive forms of NOS (eNOS + nNOS) (Adamczyk et al. 2010) (Fig. 8a), and ammonia increased eNOS expression (Fig. 8b). The results indicate that in ammonia-treated astrocytes, increase of NO_x production is related to increased eNOS expression. By contrast, cANP_(4–23) stimulates NO_x accumulation without increasing eNOS expression. Both ammonia and cANP_(4–23) increased the specific activity of NOS (Fig. S1).

Figure 7.  Effect of ANP and cANP_(4–23) pre-treatment on nitrates and nitrites (NO_x) concentration in media from primary astrocyte cultures not treated (Control) and treated for 1 (a) or 24 h (b) with 5 mmol/L NH₄Cl (Ammonia). Results are mean ± SD (n = 3). **p < 0.01 versus Control; ^†p < 0.05, ^††p < 0.01 versus Ammonia.

Figure 8.  Effect of pre-treatment with cANP_(4–23) (a, b) and 100 μM l-N^G-nitroarginine (l-NNA) (a) on NO_x concentration in the media (a) and eNOS mRNA expression (b) in primary astrocyte cultures in not-treated (Control) and treated for 24 h with 5 mmol/L NH₄Cl (Ammonia). Results are mean ± SD (n = 3). *p < 0.05, **p < 0.01 versus Control; ^†p < 0.05 versus Ammonia.

Discussion

Oxidative/nitrosative stress resulting from excessive intracellular accumulation of ROS and RNS is one of the key mechanisms by which ammonia accumulating in brain during HE impairs astrocytic metabolism and function, leading to cell swelling and cerebral edema (Blei 2008; Häussinger and Görg 2010). Earlier studies with cultured astrocytes have demonstrated that induction of ROS/RNS and its pathophysiological manifestations in astrocytes in culture and/or in the brain in situ, can be prevented or attenuated by pharmacological manipulation with antioxidants (Murthy et al. 2001; Jayakumar et al. 2006), inhibitors of NO synthesis (Schliess et al. 2002; Zielińska et al. 2003), or by inhibiting the synthesis or mitochondrial transport of glutamine, the mediator of the oxidative stress induced by ammonia at the mitochondrial level (Pichili et al. 2007; Rama Rao et al. 2010; and references therein). In cardiovascular tissue and liver cells, NPs have proven effective in countering oxidative stress caused by a variety of causes (De Vito et al. 2010; for other references see introductory paragraphs). In this study, we assessed the potential of NPs to modulate production of ROS/RNS in ammonia-treated astrocytes.

In the present study, both ANP and CNP were found to significantly reduce ROS accumulation. Most interestingly, ROS accumulation was also attenuated by an ANP analog, cANP_(4–23), which is a specific agonist of NPR-C. Experiments with peripheral and neuroendocrine tissues revealed that NPR-C is coupled to inhibitory G proteins (G_i), the downstream signal depending upon the protein involved; coupling to G_iα inhibits adenylyl cyclase, whereas interaction with G_iβ,_γ stimulates phospholipase C (reviewed by Rose and Giles 2008). The cANP_(4–23)-induced decrease of cAMP level demonstrated in the present study documents the presence of a functional NPR-C receptor in astrocytes capable of signaling downstream the G_iα–adenylyl cyclase route. However, cANP_(4–23) failed to counter the accumulation of cAMP evoked by ammonia treatment. This intriguing observation appears compatible with earlier evidence that ammonia increases cAMP synthesis in glial cells by multiple mechanisms which may not be affected by activation of G_iα (Faff et al. 1996). Down-regulation of G_iα protein expression in astrocytes by exposure to cANP_(4–23) resembled the effect described by Saha et al. (2008) in VSMC and is likewise to be considered as evidence for the presence of a functional receptor in these cells. In VSMC from SHR rats, NPR-C activation decreased NADPH oxidase activity and subsequent accumulation of •O₂⁻, in parallel with decreasing G_iα protein expression. In the present study, cANP_(4–23) likewise decreased excess •O₂⁻ production by ammonia, as did native ANP. The reduced ability of cANP_(4–23) to attenuate ammonia-induced accumulation of •O₂⁻ when G_iα protein was blocked with an antibody (Fig. 6) bespeaks direct involvement of the NADPH oxidase–G_iα protein pathway. However, modulation of NADPH oxidase activity by G_iα must have occurred independently of altered G_iα expression because: (i) reduction of G_iα expression in control rats by cANP_(4–23) was not associated with reduced accumulation of •O₂⁻, and (ii) in contrast to VSMC from SHR rats, no elevation of G_iα protein expression was observed in ammonia-treated astrocytes. Evidently, although the amount of G_iα was reduced by ammonia, it turned out to be still enough to effectively respond both to cANP_(4–23) and ANP. From the theoretical perspective, ammonia could increase •O₂⁻ generation in astrocytes by mechanisms other than the NADPH oxidase reaction, such as enzymatic oxidation of xanthine or non-enzymatic generation of redox-reactive compounds such as semi-ubiquinone (Dröge 2002).To our knowledge, no study examining the effects of ammonia on the other routes of •O₂⁻ generation has been reported as yet.

As ROS present in control cells are engaged in physiological control of cell metabolism and function (for a review see Dröge 2002), their reduction by cANP_(4–23) in astrocytes not treated with ammonia may be a regulatory phenomenon. The relatively non-toxic classes of ROS the decrease of which was recorded in control cells with the use of carboxy-H₂DCFDA probe may have included singlet oxygen, hydrogen peroxide, hydroxyl radical, and peroxyl radical (Dröge 2002).

As earlier reported by others (Schliess et al. 2002), ammonia increased NO synthesis in astrocytes. The resulting nitrosative stress is reflected by increased protein nitration (Schliess et al. 2002; Häussinger and Schliess 2005). Of particular, relevance to the pathomechanism of HE is nitration of glutamine synthetase (Schliess et al. 2002), which contributes to impairment of ammonia detoxification during prolonged hyperammonemia (Kosenko et al. 2003).

The involvement of the different NOS isoforms in the ammonia-induced increase of NO synthesis has been a matter of controversy in the literature. Schliess et al. (2002) found increased iNOS expression in ammonia-treated astrocytes. By contrast, studies in in vivo models of hyperammonemia or HE have invariably pointed to increased expression of eNOS mRNA (Hernandez et al. 2004; Sawara et al. 2009) and protein (Blei 2005) in different brain regions of symptomatic animals. Of note, eNOS mRNA expression was found normalized when HE symptoms retreated (Sawara et al. 2009). The present study tends to support the data obtained in the in vivo models; increased NO_x accumulation in astrocytes treated with ammonia for 24 h was: (i) associated with increased eNOS mRNA expression (Fig. 8b) and (ii) not recorded when the constitutive forms of NOS (eNOS + nNOS) were preferentially blocked by l-N^G-nitroarginine (Fig. 8a). The fact that ammonia increased the specific activity of NOS (Fig. S1) shows it also modulates the activity at the post-translational level. The specific activity of NOS was also increased by cANP_(4–23). The observation that NPR-C stimulation activated NOS activity without modifying eNOS expression is analogous to what has been observed in the cardiovascular tissue stimulated with ANP (Elesgaray et al. 2008; William et al. 2008; Costa et al. 2010). The exact mechanism by which activation of NPR-C in astrocytes regulates eNOS activity, but also the question if enzyme isoforms other than eNOS are involved in the stimulation remain to be investigated. Of note, basing on the mechanism underlying the protection of hepatocytes against oxidative stress (Pella 1991; Kiemer and Vollmar 1998; Carini et al. 2003) contribution of NPR-A to the effect observed with ANP cannot be excluded and deserves more detailed analysis.

Interestingly, the increase of NO_x induced by ammonia or cANP_(4–23) was not reflected by an increase of cGMP accumulation. With regard to ammonia, the lack of increase could be due to down-regulation of soluble guanylate cyclase in ammonia-treated astrocytes (Konopacka et al. 2006). One other possibility which may apply to both agents is an increased activity of a feedback pathway in which cGMP synthesized in the NO-cGMP pathway activates cGMP-degrading phosphodiesterases; this pathway was found altered by ammonia (Monfort et al. 2004). In the case of ANP, cGMP accumulation was a straightforward reflection of the activation of NPR-A, the particulate form of guanylate cyclase which is present in the astrocytic membrane (Sumners and Tang 1992; Konopacka et al. 2006).

A decrease of ammonia-dependent NO production by cANP_(4–23) indicates that selective activation of NPR-C may be effective in counteracting nitrosative stress evoked in astrocytes by ammonia. However, the mechanism underlying this effect is unknown. In macrophages, NPs destabilize iNOS mRNA (Kiemer and Vollmar 1998, 2001) and/or decrease the lipopolysaccharide-induced expression of CAT-2B, an arginine-transporting protein (Kiemer and Vollmar 2001). These mechanisms are unlikely to apply to ammonia-treated astrocytes because: (i) in contrast to the report of Schliess et al. (2002), the present study does not implicate iNOS in the ammonia-induced increase of NO_x, and (ii) the effects described in macrophages are cGMP-dependent, implicating NPR-A rather than NPR-C. Possibility (ii) is thus difficult to reconcile with the intriguing absence of the NO-reducing effect of ANP in ammonia-treated astrocytes, because ANP also interacts with NPR-A and generates cGMP. Clearly, the relative roles of the different NOS isoforms in the ammonia-induced generation of NO_x, and their involvement in the mechanism by which activation of NPR-C reduces the response require a more detailed investigation.

Activation of NPR-C also signals via a pathway involving phospholipase C, which streams down to protein kinase C δ and results in phosphorylation of mitogen-activated protein kinase p38 (Rose and Giles 2008). This pathway was not investigated in the present study and its presence in astrocytes needs to be confirmed. However, even if present, it is not very likely to be engaged in diminishing oxidative stress induced by ammonia or other ROS-generating agents in astrocytes; activation of MAPK actually mediates the deleterious effects of ROS in cultured astrocytes treated with ammonia (Jayakumar et al. 2006) or menadione (Zhu et al. 2009).

In conclusion, NPR-C receptor activation attenuates ammonia-induced formation of ROS and NO in astrocytes. As NO and ROS contribute jointly to ONS, the key mechanism underlying ammonia neurotoxicity, NPR-C-preferring analogs of ANP may in the future become a useful therapeutic modality in overt stages of HE at which blood and brain ammonia concentrations are significantly raised. To achieve this goal, future experimental studies should account for the ability of these analogs to counter ammonia neurotoxicity in vivo.

Acknowledgements

This study was supported by Ministry of Science and Education grants S005/P-N/2007/01 (JA) and NN 401 0550 33 (MZ).