In its response to tissue damage and inflammation induced by a variety of xenobiotics, endotoxins and disease states (such as viral hepatitis), post-ischemic and regenerative injury, the liver produces a large quantity of nitric oxide (NO). Nearly all cell types in liver tissue, including hepatocytes, Kupffer cells, stellate cells and endothelial cells, have the capacity for generating NO. It has been documented that NO is capable of promoting apoptosis in a number of different cell types, generally classified as cGMP-dependent or cGMP-independent 1234. The potential of chemical NO donor sodium nitroprusside (SNP) to induce apoptosis directly from NO liberation has been established in vitro 5. The fact that NO is capable of triggering apoptosis is consistent with its ability to induce DNA damage, the inhibition of DNA synthesis and cell cycle arrest 67. The reaction product formed between NO^• and superoxide [i.e., peroxynitrite (ONOO^-)] plays a critical role in the induction of inflammatory reaction and apoptosis, but is also associated with tumor promotion and/or progression. Potentially toxic levels of peroxynitrite can be achieved whenever NO^• and O₂^.- production is stimulated, due to the fact that a 100-fold increase in the rate of peroxynitrite formation occurs for every 10-fold increase in NO^• and O₂^.- concentration 8.

Apoptosis, frequently termed "programmed cell death", is the form of cell death that occurs in normal liver in the course of its development and organogenesis, and in adult liver during the renewal of hepatocytes. In addition, apoptosis can be triggered by several hepatotropic viruses and toxic drugs, as well as in various liver diseases and experimental liver conditions such as hepatic allograft rejection. Degradation of the nuclear DNA, a common phenomenon observed in many organisms throughout the evolutionary scale, is one of the best-characterized biochemical features of apoptotic cell death. It has been established that the cell undergoes epigenetic reprogramming in the D₁ phase of programmed cell death, the result of which is the activation of double-stranded DNA fragmentation in the F phase during which the nuclear morphology dramatically changes 9. The cleavage of DNA may have a protective function in that it reduces the likelihood for genes in a potentially active site to be transferred from dying cells to the nuclei of viable neighboring cells. It is possible that various endonucleases exert a DNA degrading activity, as well as that many proteins can receive DNA degrading properties upon change of pH conditions 11. The most important aspect of apoptosis is the universal property of some proteins to exert a dual function: the protection against proteolysis and the maintenance of the structure and function of normal cells. Being free from the inhibitory complex, however, these proteins may also contribute to protein or chromatin cleavage during apoptosis 1213. Changes in DNA degradation may lead to the pathogenesis in various disorders, such as liver cancer 1415.

On the basis of the pharmacological data supporting the critical role of NO and peroxynitrite in apoptosis, current research studies have evaluated the activity of alkaline and acid DNase during the administration of SNP or of peroxynitrite, as well as the changes in numerous susceptible parameters of nitrosative stress, including SH group oxidation, carbonyl group formation, lipid peroxidation and DNA modification. An assay of enzyme activity was performed using liver tissue after in vivo administration and in vitro treatment of isolated rat hepatocytes or purified commercial enzymes either with SNP or authentic peroxynitrite.

There are few data concerning the in vivo susceptibility of liver tissue to NO donor SNP and authentic peroxynitrite. A considerable attention has been paid to the establishment of in vivo tolerability and to the markers of apoptotic effects. Both NO and peroxynitrite can directly react with aromatic and sulfhydryl nucleophiles and nitrate aromatic residues. Sulfhydryl groups oxidation was documented in the plasma and, almost equally, in liver tissue. Peroxynitrite administration led to a more pronounced decrease in the concentration of other plasma free radical scavengers such as uric acid, and was followed by an increase in plasma nitrate concentration (Tables 1 and 2). According to the data suggesting that peroxynitrite decomposes rapidly to OH^• and NO₂ ^•-like species at physiological pH, it was assumed that the carbonyl groups and the lipid peroxidation product [i.e., malondyaldehyde (MDA)] may play a significant role in liver cell toxicity. Neither plasma nor liver protein carbonyls showed any significant increase. This may be a likely consequence of the significant increase in aromatic amino acids nitration, presumably tyrosine the spectral contribution of which was substracted from the samples treated with 2,4-dinitrophenylhidrazine. Plasma and liver MDA concentrations were not significantly changed, either. The obtained results do not support the data suggesting that oxygen radicals, probably generated during cellular SNP metabolism, may mediate cell toxicity and apoptosis, but do confirm previous in vitro observations 16. Plasma alanine aminotranferase (ALT) activity, used as the standard liver functional test, decreased in the peroxynitrite-treated group (Table 1). The activity of liver alkaline DNase increased and that of acid DNase decreased after in vivo treatment with either SNP or peroxynitrite (Table 2).

The ultraviolet spectra of DNA were obtained by spectrophotometric scanning between 230–500 nm on a scan detecting system. According to the data by Yermilov et al. 17, the appearance of a peak between 375 and 405 nm (depending on pH) corresponds to 8-nitroguanine. In the scan analysis (Figure 1), the peak was between 390 and 410 nm with a maximum absorbance at 405 nm in alkaline conditions (DNA extract was adjusted to pH 10). This peak may correspond to the formation of nitro-derivatives, most probably of 8-nitroguanine. The nitroguanine peak at 405 nm was particularly apparent in peroxynitrite-treated samples.

In vivo administration of SNP or peroxynitrite tended to increase the rate of DNA fragmentation, but it was not statistically significant. The rate was estimated according to the percentage of DNA resisting centrifugation at 27 000 g (Table 2).

After in vitro exposure of isolated hepatocytes to SNP or peroxynitrite, the activity of both alkaline and acid DNase decreased in a dose-dependent fashion (Figure 2). During in vitro incubation of purified enzymes DNase I and DNase II with SNP or peroxynitrite, a dose-dependent decrease of enzyme activity was also documented (Figure 3).

NO^•, a free radical gaseous molecule is one of the simplest compounds found to be continuously produced in humans and animals. It can be derived from L-arginine through the enzyme nitric oxide synthase (NOS) and by different NO donors, including SNP. NO has been shown to play an unprecedented range of roles in biological systems, acting as a universal intracellular and transcellular signaling molecule and the regulator of vascular tone, cell proliferation and apoptosis 181920. Peroxynitrite is a strong, relatively long-lived oxidant with a half-life of approximately 0.5–1 s under physiological conditions. Our study confirmed that in both plasma and liver tissue peroxynitrite causes a rapid oxidation of sulfhydryl groups and thioethers, as well as the nitration and hydroxylation of aromatic compounds (Tables 1 and 2). A chronic exposure of hepatocytes to reactive nitrogen species exhibits a cytotoxic and cytostatic activity leading to functional and morphological alterations 821. Cell death after exposure to different NO-donors such as SNP has been to date established through the expression of tumor suppressor gene p53 and pro-apoptotic genes such as bax, cyclin-dependent kinase inhibitor p21, the inhibited expression of anti-apoptotic protein bcl-2, the inhibited NF-κB binding activity, ERK and p-38-dependent cytochrome c release, and caspase-3 activation 222324. In contrast, the anti-apoptotic effects of NO may be mediated through the mechanisms such as blockade of the recruitment of pro-caspase-9 to the Apaf-1 apoptosome, stimulation of c-GMP-dependent protein kinase, control of mitochondrial permeability transition, induction of the heat shock protein HSP 70, and interaction with the ceramide pathway 2526. The prolonged damage of p53 gene by peroxynitrite has been associated with tumor formation. Recent results by Vincent and Maiese 3 indicate that NO donor SNP (at 300 μmol concentration) is capable of inducing strong apoptotic effects via DNA fragmentation and induction of Mg²⁺-dependent endonuclease activity in the culture of neuronal cells. In our in vivo study (Table 2), the activity of alkaline DNase increased within 24 h after exposure to SNP (achieving approximately a similar blood concentration of about 250 μmol) or to authentic peroxynitrite. Several molecules involved in nuclear DNA fragmentation have been detected and characterized based on their ionic sensitivity. Besides the presence of constitutive Ca²⁺/Mg²⁺-dependent endonucleases, a great deal of endonuclease activity within a 7.2–8.0 pH range most probably represents the inducible form of DNase. The molecular weights of the constitutive (NO-independent) and inducible (NO-dependent) endonuclease are similar, as well as their optimum pH range (7.5–8.0). A likely conclusion is that Mg²⁺-dependent endonuclease seems to be a result of de novo synthesized or the pre-existing Ca²⁺/Mg²⁺-dependent endonuclease activation. Up to now, several Mg²⁺- or Ca²⁺/Mg²⁺-dependent alkaline DNases (DNase I) with an optimum activity within the range of 7.5–9.5 have been purified. Some of them, including specific caspase3-activated DNase (CAD), are active upon release of the specific inhibitor ICAD 2728. DNase gamma has been documented as a critical component of apoptotic machinery, in that it cleaves the chromosomal DNA into nucleosomal units, thus leading to DNA ladder formation 29. The alkaline DNase, active only during apoptosis, has been documented to be inherent to cyclophilins (A, B and C) as well, irrespective of their protein folding (peptidylprolyl cis-trans-isomerase) activity. All of them have the ability to degrade the supercoiled, single stranded and double stranded DNA 3031. Besides alkaline DNases, the cation-independent endonuclease with an optimum activity at pH 5, known as acid or DNase II, was identified. One leucocyte elastase inhibitor (LEI) can also exert an acid DNase activity after post-translational modification through the proteolytic cleavage 32. The specific involvement of DNase II in physiological nuclear degradation during apoptosis could not be excluded upon decrease of intracellular pH values below 7 with a proton ionophore. Three potential N-nitrosylation sites are important for DNase II regulation 3233. Since our experimental data indicated a decrease in acid DNase activity 24 h after exposure to SNP or peroxynitrite (Table 2), the inhibition of DNase II may be explained by the nitrosylation of its susceptible sites. Indeed, when isolated hepatocytes were exposed to SNP or peroxynitrite for 1 h, a dose-dependent inhibition of DNase II was also documented (Fig. 2). The same result was obtained after exposure of purified enzyme to SNP (in the presence of the reducing agent cysteine 1 mmol) or peroxynitrite (Fig. 3).

The formation of 8-nitroguanine, 8-oxo-deoxyguanine and oxazolone and the oxidative modification of 2'-deoxyribose into TBA-responsive compounds are the most prominent nucleotide modifications after reactive nitrogen species attack 3435. A highly potential mutagenic product 8-nitroguanine can be depurinated yielding apurinic sites capable of inducing GC→TA transversions, GC→CG transversions and deletions 1736. The appearance of the nitroguanine peak during the scan analysis of purified DNA at 405 nm was documented in our study (Fig. 1). The rate of DNA fragmentation tended to be increased, but the difference was not significant (Table 2).

In vivo administrated SNP and peroxynitrite increase the activity of alkaline DNase. They also induced DNA modifications, such as nitroguanine formation. The obtained DNase activation seems to be associated with the programmed destruction of the genome and cell death. Given the above results and observations, the elimination of damaged hepatic cells appears to be a likely factor in prevention against mutation and carcinogenesis.

GK carried out the in vivo and in vitro experiments, culture experiments and wrote the paper. RP and GN carried out DNA spectral analysis. TC performed measurement of SH groups and lipid peroxides. IS performed the measurement of nitrates and nitrites. TJ assisted during in vivo and in vitro experiments and did the graphical presentation. DS performed statistical analysis and assisted during in vivo experiments. DP and RK assisted during in vitro experiments and participated in the design of the study. All the authors read and approved the final manuscript.