The Effects of Nitric Oxide Synthase Inhibitors in Acetaminophen-Induced Hepatic Injury in Mice

: (1) Aim of the Study: In this study, we aimed to evaluate the vascular damage and the effects of nitric oxide synthase (NOS) enzyme inhibitors in hepatic damage caused by high doses of acetaminophen (APAP). (2) Material and methods: Fifty-three Swiss albino male mice were used for this study. Hepatic and thoracic aorta toxicity caused by 2 or 6 h exposures to APAP (300 mg/kg intraperitoneally (i.p.)) were evaluated. The general NOS inhibitor N G -nitro-L-arginine methyl ester (L-NAME: 25 mg/kg and 50 mg/kg, i.p.) and the neuronal NOS inhibitor 7-nitroindazole (7-NI: 15 mg/kg, i.p.) were administered one hour before APAP exposure. (3) Results: Significant morphological deteriorations were observed after 6 h of APAP exposure in histopathological examinations of hepatic sections. Pre-treatment with L-NAME (at 50 mg/kg) or 7-NI before a 6 h APAP exposure significantly decreased hepatic toxicity ( p < 0.05). Significant increases in ALT levels in 6 h of APAP exposure were decreased by both L-NAME (with the 25 mg/kg but not at 50 mg/kg) and 7-NI pre-treatments. No significant change was observed in the measured nitrate/nitrite levels and total antioxidant status in either serum or liver homogenates. No significant deteriorations were observed during either hematoxylin-eosin or immunohistochemical staining in thoracic aorta sections. In the thoracic artery sections, no statistical difference was found in acetylcholine-mediated relaxation, which may indicate endothelial dysfunction. (4) Conclusions: This study demonstrated that APAP-induced hepatic toxicity, especially neuronal NOS inhibitors, may decrease hepatic toxicity. It was also shown that APAP-induced hepatic toxicity was not accompanied by vascular dysfunction. non-parametric Kruskal–Wallis test and Dunn’s comparison analysis used to compare semiquantitative pathological scores. One-way ANOVA for biochemical analyses and Dunnett’s comparative analysis test for multiple group comparisons were used. For repeated measures in aortic responses, two-way ANOVA and Dunnett’s multiple comparison tests were used. p < 0.05 was chosen as the threshold for statistical significance.


. Introduction
Acetaminophen (APAP) is a commonly prescribed antipyretic and pain-relieving therapeutic. It is known that the side e ects of APAP at therapeutic doses (below g daily dose) are low. However, acute high doses and long-term chronic use can cause damage to the liver and lead to fatal outcomes if not treated e ectively [ ].
APAP is known to be metabolized to the toxic N-acetyl-β-benzoquinone imine (NAPQI), which has high reactivity with cytochrome P enzymes in the liver [ , ]. In hepatic cells, NAPQI is e ectively detoxi ed by glutathione [ ]. Due to glutathione depletion, at toxic doses of APAP, NAPQI cause necrotic damage in the liver by covalent attachment of centrilobular to intracellular proteins in the liver [ ]. In experimental studies carried out on animals, during NAPQI-induced centrilobular liver injury attributed to high-dose APAP exposure, the release of reactive oxygen/nitrogen radicals due to glutathione depletion was observed [ ].
The increased reactive oxygen/nitrogen radicals, nitrotyrosine compounds, and in ammatory cytokines cause mitochondrial dysfunction [ -]. It has been reported that nitrotyrosine compounds in the liver may be associated with nitric oxide synthesized from sinusoids and the hepatic microcirculation [ ]. It has been shown in studies that the activity of inducible nitric oxide synthase (iNOS), which increases nitric oxide levels during APAP toxicity, leads to nitrotyrosine accumulation [ , -]. However, pharmacological inhibition of iNOS did not decrease APAP-induced liver toxicity in mice [ ]. Studies in iNOS null mice have shown a decrease in protein nitration; however, there was also an increase in lipid peroxidation leading to APAP-induced toxicity [ ]. A newly developed iNOS inhibitor has been shown to have a reducing e ect on APAP-induced hepatic toxicity [ ]. Thus, it was thought that iNOS might have a role in protein nitration, but the mechanism of its toxicity is unclear [ ]. Delayed hepatic toxicity and the development of fewer nitrotyrosine compounds were shown in wild-type mice with neuronal NOS enzymes compared with untreated wild-type mice, re ecting the role of initial toxicity in APAP-induced hepatic toxicity in mice [ ]. Thus, neuronal NOS might be a pathological source of nitric oxide; however, there also might be another independent pathological pathway related to APAP-induced hepatic toxicity [ ].
Several authors have proposed that endothelial cell damage and microvascular dysfunction might have a role in the toxicity of APAP [ , ]. Randle and colleagues showed increased catecholamine levels in APAP-induced hepatic toxicity, and decreased toxicity was observed with administration of alpha receptor antagonists, calling attention to the hepatic microcirculation [ ]. We aimed to evaluate whether severe macrovascular dysfunction accompanies microvascular dysfunction in APAP-induced hepatic toxicity in mice. Additionally, in this study, we attempted to show the e ects of both the non-speci c NOS inhibitor NG-nitro-L-arginine methyl ester (L-NAME) and the neuronal NOS inhibitor -nitroindazole ( -NI), which may have a role in the hepatic circulation during hepatic injury caused by APAP.

. . Animals
The protocol for this study was evaluated by Hacettepe University Faculty of Medicine Experimental Animals Ethics Committee (approval number: --), and this study was carried out in accordance with the "Regulation on the Welfare and Protection of Animals Used for Experimental and Other Scienti c Purposes". In this study, adult, male Swiss albino mice (average weight: . ± . g) from the Hacettepe University Medical Faculty Experimental Animal Unit were used. Before the study protocol, mice used in experiments were housed on a h light/dark cycle and with no food or drink restriction. , ( ),

. . Study Protocol
The mice to be used during the experiment were fasted overnight. Subsequently, mice were given intraperitoneal (i.p.) APAP ( mg/kg, dissolved in warmed saline solution). Measurements were made and h after exposure to APAP. For the control group, the same amount of saline was given. Animals were divided into groups: h of APAP exposure, h of APAP exposure, and a saline control. In order to show the role of the NOS pathway, h prior to administration of APAP, the NOS inhibitor L-NAME ( mg/kg and mg/kg, i.p.) or the neuronal NOS inhibitor -NI ( mg/kg, i.p.) were given. The mice to be examined were anesthetized with urethane ( . g/kg, i.p.) before the experiment.
The hepatic tissues of all animals were photographed before blood samples were taken. The blood was obtained intracardially and then centrifuged ( min at g). Obtained samples were stored at − • C. Portions of hepatic tissue were prepared fresh and stored at − • C until examination. Another piece of hepatic tissue and a piece of the thoracic aorta were stored in % formaldehyde solution for histopathological examination. Another part of the thoracic aorta was isolated from the surrounding tissue and kept in physiological saline solution (PSS), which was aerated with a % oxygen and % carbon dioxide mixture at • C for functional measurements. PSS contained (in mmol/L) NaCl: ; KCl: . ; NaHCO : ; MgSO : . ; KH PO : . ; CaCl : . ; glucose . and EDTA: . ; pH . .

. . Pathological Evaluation
The liver and aorta tissues xed in % formaldehyde solution were embedded in para n after tissue processing. The para n blocks were cut in -µm-thick and the sections were stained with haematoxylin-eosin stain. Prepared slides were examined under a light microscope (Olympus, BX , Tokyo, Japan).
Liver damage was assessed in terms of the following parameters: . Cytoplasmic vacuolization in hepatocytes; . Increase in eosinophilia in hepatocytes; . Hepatocellular necrosis; . Haemorrhage; . Periportal area in ammation.
Thoracic aorta tissue was evaluated by haematoxylin-eosin staining. CD , a transmembrane glycoprotein synthesized by endothelial cells, has been used to demonstrate damage in vascular endothelial cells. Endothelial cells stained immunohistochemically with a CD -speci c antibody were also examined microscopically. The concentrations of hepatic tissue nitrate/nitrite and TAS are expressed as µmol/g protein and mmol/g protein, respectively. All measurements were made under the instruction manual of the kit.

. . Functional Measurements
A wire myograph system ( M Danish Myo-Technology, Aarhus, Denmark) was used, and isometric arterial responses were recorded using a data acquisition system (Powerlab/ SP, AD Instruments, CO, USA). The isolated thoracic aortic rings were cut into lengths of~ mm rings. The portion was isolated from the surrounding tissues and was prepared for examination in the myography system. Two steel wires µm in diameter were passed through the lumen of the artery. With the help of the steel wire, the ring was mounted to a mL myograph chamber maintained at • C and continuously aerated with a % oxygen and % carbon dioxide mixture. The prepared arterial rings were rested for min and then stretched to the calculated normalized diameter step by step using the normalization module program (Danish Myo-Technology, Aarhus, Denmark) as described in the literature [ ]. By the end of normalization, the measurements were recorded in mN/mm. After normalization and resting, the arteries were challenged with high K + by replacing the PSS with mM K + containing PSS, and the reference contraction curves were recorded. After the contraction reached a plateau, the mM K + containing PSS was washed out. Then, phenylephrine ( nM-µM) concentration-isometric contraction responses were recorded. The submaximal contraction responses ( -% of the mM K + contraction reference value) were recorded. After the submaximal response was obtained, concentration-relaxation response curves of acetylcholine ( nM-µM) added cumulatively to the organ bath were recorded to show endothelium-dependent relaxation responses.

. . Reagents and Statistical Analysis
All drugs used in the experiments were obtained from Sigma-Aldrich (St. Louis, MO, USA). The GraphPad Prism (v , La Jolla, CA, USA) program was used to prepare the graph images and statistics. A non-parametric Kruskal-Wallis test and Dunn's comparison analysis were used to compare semiquantitative pathological scores. One-way ANOVA for biochemical analyses and Dunnett's comparative analysis test for multiple group comparisons were used. For repeated measures in aortic responses, two-way ANOVA and Dunnett's multiple comparison tests were used. p < . was chosen as the threshold for statistical signi cance.

. . Pathological Evaluation
Structural damage was observed macroscopically in livers exposed to APAP for h and h, while no visible damage was observed in the saline-treated control group ( Figure A-C). Macroscopic improvement of hepatic tissue was observed following administration of mg/kg and mg/kg L-NAME and -NI ( mg/kg), h before APAP (Figure D-F).
When liver tissue sections were examined by haematoxylin-eosin staining, histopathologic examination revealed eosinophilia, vacuolisation, hepatocyte necrosis, and di use parenchymal haemorrhage in the portal area and near the central vein ( Figure A-C).
Compared with the control group exposed to APAP for h, when the severity of damage was scored on a -point semiquantitative scale, there was a statistically signi cant impairment in hepatocellular necrosis, haemorrhage, and periportal area in ammation in cytoplasmic vacuolization in hepatocytes (Table ). Administration of mg/kg L-NAME resulted in a partial improvement similar to that in the control group. , ( ), Signi cant improvements were obtained with a mg/kg dose of L-NAME, compared with h of APAP injury. A similar improvement to the control was obtained with pathological assessment results following -NI pre-treatment, which was found to be signi cantly di erent from h of APAP exposure (Table ).  Mouse thoracic aortas and endothelial layers were observed as slightly corrupted but not signi cantly damaged when examined pathologically with haematoxylin-eosin staining. Moreover, there was no signi cant damage in the endothelial layer with immunohistochemical staining using the endothelial marker CD . Mild non-signi cant damage of the endothelial layer seen in the examined sections is shown in Figure . , ( ),

. . Biochemical Measurements
As a result of APAP exposure, the measured ALT activity in serum was increased statistically signi cantly after h of exposure (Figure ). -NI and L-NAME ( mg/kg) given before APAP resulted in a decrease in measured ALT activity (Figure ). No signi cant changes were observed in the further biochemical parameters measured in both serum and hepatic tissues (Figure ).

. . Functional Evaluation of Thoracic Aorta
Thoracic aortic rings analysed in the myograph system showed no di erence in smooth muscle contraction forces (mN/mm) against mM KCl stimulation ( Figure A). Of the arterial rings (in all arterial rings) were excluded from the study due to the irresponsiveness to phenylephrine concentrations. In the h APAP-induced group, the concentration-contraction curves induced by phenylephrine ( nM-µM) were statistically signi cantly di erent ( Figure B), but there were not signi cant di erences in the relaxation curves obtained by acetylcholine ( nM-µM) in all groups analysed ( Figure C).

. Discussion
In this study, hepatic toxicity was observed following systemic administration of high-dose APAP to mice, and it was also shown that pre-administration of the general NOS inhibitor L-NAME and the neuronal NOS inhibitor -NI partially reduced hepatic damage and decreased hepatic damage histopathologically, respectively. However, as a result of pharmacological functional and pathologic evaluations, no signi cant alterations were observed in thoracic aorta.
In this study, hepatic toxicity, which started at h of APAP exposure and became evident at h, was shown in mice macroscopically and with pathological examination using routine haematoxylin-eosin staining. The reported toxicities were thought to be consistent with the literature [ , ]. In the literature, APAP was reported to have a hepatic toxicity beginning and progressing at and h, respectively, after administration of mg/kg gavage or i.p. [ , ]. Additionally, in our study, mg/kg dose of acetaminophen, which causes a moderate level of hepatotoxicity in previous studies, was chosen [ ]. Statistically signi cant deteriorations were observed due to APAP exposure for h in sections of hepatic tissues by using a semiquantitative pathological method. Additionally, h of APAP exposure caused a signi cant increase in measured ALT levels in mouse serum. The toxicity was attributed to the depletion in glutathione, which led to the uncompromised increase in reactive oxygen/nitrogen species [ ]. The complex role of radical nitrogen species in APAP-induced hepatic toxicity was extensively studied, and decreased toxicity to hepatocytes was shown by inhibition of nitric oxide synthase [ ]. Additionally, in our study, pre-treatment with L-NAME, the general nitric oxide synthase inhibitor, partially reversed the toxicity caused by APAP in the liver. These improvements in histopathological ndings were found to be more prominent in the mg/kg L-NAME than in the mg L-NAME pre-treatment. These results suggest that L-NAME is susceptible to APAP-induced damage reduction. The e ect seen at a dose of mg/kg may not have been observed due to shifting the majority toxicity pathway towards superoxide-derived oxidants, as speculated in the literature [ ], or due to the signi cant hemodynamic disturbances that may be caused by L-NAME administered at a dose of mg/kg. In the mg/kg L-NAME and -NI pre-treatment groups, mouse serum ALT levels were measured similar to the control group. The partial improvement e ect shown for L-NAME has not been shown in the literature [ , ]. In a previous study, NOS inhibitors (N-monomethyl-L-arginine (L-NMMA) ( mg/kg), L-NAME ( mg/kg), L-N-( -iminoethyl)lysine (L-NIL) ( mg/kg), and aminoguanidine ( mg/kg)) were not shown to reduce the hepatotoxicity of high doses of , ( ), APAP in mice [ ]. The augmentation of hepatotoxicity by reducing NOS inhibitors appeared to shift the toxicity pathway toward reactive oxygen species [ , ]. However, in our study, L-NAME was administered at high doses of mg/kg and mg/kg, unlike in the literature. It has been thought that, at higher doses, L-NAME might reduce in all other NOS enzymes, which may lead to a complex e ect on hepatic toxicity. In addition, the NO releasing and nitrovasodilator properties of L-NAME have been shown in the literature, as well as the exacerbation of these properties during in ammation [ ], therefore it is understood that more studies are needed to elucidate the complex e ects of L-NAME. In a cell culture study, decreased mitochondrial damage with reduction in the initial phase of cellular damage was evaluated by adding the general NOS inhibitor L-NMMA and the neuronal NOS inhibitor -nitroindazole in the re-incubation media, but not with the inducible NOS inhibitor L-NIL (N -( -iminoethyl)-L-lysine)) [ ]. Therefore, consistent with our results, neuronal NOS might have a prominent role in APAP-induced hepatotoxicity. However, APAP-induced hepatotoxicity was present in NOS -null mice [ ]. Therefore, neuronal NOS is important, but is not the only pathologic pathway in APAP-induced hepatotoxicity. It has also been shown that a newly developed inducible NOS inhibitor had a protective e ect against hepatic toxicity caused by APAP [ ].
An increase in nitrite/nitrate levels in mouse serums of APAP treatment, which was shown in previous studies [ ], could not be shown in our study. We did not observe signi cant changes between treatment groups. It was thought that if the exposure duration of APAP was increased to or h, signi cant changes might be observed. This issue might also be related to di erences in the animal strains used in our study and the fact that our method lacks the sensitivity to detect di erences.
APAP-induced hepatic damage was thought to be related to endothelial cell injury, microvascular dysfunction, and hepatocyte necrosis [ ].
In an APAP toxicity model in rats, aortic endothelial-and non-endothelial-induced relaxation responses impaired along with vessel morphology [ ]. Therefore, in this study, the in ammation developed by APAP toxicity in mice may also a ect the aortic tissue. Here, we tried to show widespread vascular and endothelial impairment. However, in a histopathological examination of vascular tissue with routine haematoxylin-eosin staining, no signi cant impairment was observed. Besides showing the super cial endothelial cell structure, we assessed CD glycoprotein by immunohistochemical staining, but no signi cant endothelial damage was shown. It was therefore concluded that the h APAP exposure did not cause widespread vascular pathology in a histopathological manner. Although histopathologic damage was not demonstrated in the thoracic aorta, parts of the arterial rings were evaluated in the myograph system to assess the possible damage at the functional receptor level. In our myograph experiments, the phenylephrine-induced contractions were signi cantly di erent in the h APAP exposure group. In mice, it was shown that plasma catecholamine levels were increased h after APAP administration, and increased catecholamine levels were associated with APAP-induced hepatotoxicity [ ]. Therefore, the hyperactivity of the artery segments in the h APAP group might be related to the early increase in catecholamine concentrations or in ammatory mediators. The increased contractility seen at h of APAP exposure but not at h of APAP exposure may also speculate a down-regulation of post-receptor mechanisms. However, as mentioned before in the results section, few arterial samples were irresponsive to cumulative concentrations of phenylephrine. This irresponsiveness seen in some arteries may be due to the trauma that the tissue was exposed to during preparation for myograph experiments.

. Conclusions
It was concluded that in APAP-induced hepatic toxicity in mice, a measurable vascular pathology could not be established. In our study, the inability to evaluate the microcirculation in the liver was seen to be a limitation of our study and remains to be established in future studies. In conclusion, this study showed that nNOS in particular may have a clear role in APAP-induced hepatic injury; however, macrovascular functional