Nitration of cardiac proteins is associated with abnormal cardiac chronotropic responses in rats with biliary cirrhosis


  • Ali R. Mani,

    1. The UCL Institute of Hepatology, Department of Medicine, Royal Free & University College Medical School, University College London, London, UK
    Search for more papers by this author
  • Silvia Ippolito,

    1. The UCL Institute of Hepatology, Department of Medicine, Royal Free & University College Medical School, University College London, London, UK
    Search for more papers by this author
  • Richard Ollosson,

    1. The UCL Institute of Hepatology, Department of Medicine, Royal Free & University College Medical School, University College London, London, UK
    Search for more papers by this author
  • Kevin P. Moore

    Corresponding author
    1. The UCL Institute of Hepatology, Department of Medicine, Royal Free & University College Medical School, University College London, London, UK
    • The UCL Institute of Hepatology, Department of Medicine, Royal Free & University College Medical School, University College London, Rowland Hill Street, London NW3 2PF, UK
    Search for more papers by this author
    • fax: (44) 207 433 2871

  • Potential conflict of interest: Nothing to report.


Acceleration of the heart rate in response to catecholamines is impaired in cirrhosis. In this study, we tested the hypothesis that increased formation of reactive nitrogen species in biliary cirrhosis causes nitration of cardiac proteins and leads to impaired chronotropic function. Bile duct–ligated (rats with cirrhosis) or sham-operated rats were injected daily with either saline, NG-L-nitro-arginine methyl ester (L-NAME), or N-acetylcysteine for 7 days from week 3 to week 4 after surgery. Cardiac chronotropic responsiveness to β-adrenergic stimulation was assessed in vitro using spontaneous beating isolated atria. Nitration of cardiac proteins was measured by mass spectrometry and located by immunogold electron microscopy. Marked impairment of chronotropic responses of isolated atria to isoproterenol was seen in rats with cirrhosis, which normalized after the administration of N-acetylcysteine or L-NAME. The levels of protein-bound nitrotyrosine in atrial tissue increased from 16 ± 1 to 23 ± 3 pg/μg tyrosine in rats with cirrhosis, and decreased to 15 ± 1 and 17 ± 1 pg/μg after treatment with L-NAME and N-acetylcysteine, respectively (P < .05). Immunogold electron microscopy demonstrated increased nitration of mitochondrial proteins in the atria of rats with cirrhosis. The plasma nitrite/nitrate levels were elevated in rats with biliary cirrhosis, and decreased after administration of L-NAME but were unchanged by N-acetylcysteine. In conclusion, abnormal cardiac chronotropic function in cirrhosis is associated with increased nitration of cardiac proteins. Two independent treatments (N-acetylcysteine and L-NAME) that decrease nitration of cardiac proteins led to normalization of cardiac responses. Nitration of critical proteins in cardiac tissue may lead to abnormal cardiac function. (HEPATOLOGY 2006;43:847–856.)

Cardiac responses are impaired in cirrhosis, with a significant diminution of both chronotropic and inotropic changes after physiological or pharmacological stimulation.1 Thus, maneuvers leading to sympathetic activation, such as tilting, physical exercise, and pharmacological stimulation, do not evoke an adequate acceleration of heart rate in patients with cirrhosis compared with healthy subjects.2–4 Current evidence suggests that this is attributable to inherent abnormalities of the cardiac tissue, a phenomenon that has been controversially termed cirrhotic cardiomyopathy.1

The contribution of nitric oxide (NO) to the pathophysiology of cardiac dysfunction in cirrhosis was first described by Van Obbergh et al. in 1996.5 Subsequent studies by Lee and colleagues6 have shown that serum cytokine levels are increased in rats with cirrhosis, and that the negative inotropic effect of interleukin-1β (IL-1β) could be reversed by pre-incubation of isolated papillary muscle with an NO synthase (NOS) inhibitor. Similarly, we have shown that increased NO synthesis in bile duct–ligated (BDL) rats causes bradycardia by causing impaired responsiveness of cardiac pacemaker cells to adrenergic stimuli.7, 8 Nitric oxide is involved in the modulation of cardiac function in the normal heart, as well as in various cardiac diseases via mechanisms involving NO that are both dependent on and independent of guanylyl cyclase.9, 10 For example, evidence suggests NO attenuates cardiac pacemaker activity, mediated by cyclic guanosine monophosphate (cGMP).11 However, NO may also modify cardiac function through nitration or S-nitrosation of cardiac proteins.10, 12 Tyrosine residues in proteins are targets for nitration by reactive nitrogen species such as peroxynitrite or nitryl chloride (formed by a myeloperoxidase-dependent pathway), and may lead to loss of function.12–14 One example of this is actin, which when nitrated loses its ability to contract effectively.14 Likewise, cell signaling via phosphorylation of tyrosine residues also may become impaired after nitration of critical regulatory proteins.12 Moreover, the concept that protein function may be controlled by S-nitrosation of a critical cysteine residue has emerged in biology as a potentially important mechanism or pathway to control cellular function.12, 15

Although cirrhosis is well recognized as being associated with increased NO synthesis, the role of nitration or nitrosation of cardiac proteins in cardiac function in cirrhosis has not previously been investigated. Here we have investigated the effects of two independent treatments (a low-molecular-weight thiol and an NO synthase inhibitor), both of which can decrease nitration/nitrosation of proteins, on cardiac chronotropic responses in a rat model of biliary cirrhosis.


NO, nitric oxide; IL-1β, interleukin-1β; NOS, NO synthase; BDL, bile duct–ligated; L-NAME, NG-L-nitro-arginine methyl ester; ECGk electrocardiogram; HRV, heart rate variability; HF, high frequency; VLF, very low frequency; nu, normalized units; 2H6-PHPA, deuterated para-hydroxyphenylacetic acid;2H5-NHPA, deutrated 3-nitro-para-hydroxy-phenylacetic acid

Materials and Methods

Reagents and Materials.

All materials were purchased from Sigma (Pool, UK), unless otherwise specified in the text.


Male Sprague-Dawley rats (body weight 250–280 g) were obtained from the Comparative Biology Unit at the Royal Free and University College Medical School (Royal Free Campus, London, UK). All animal procedures were in accordance with Home Office (UK) recommendations. Animals were given free access to normal rodent chow and water up until the time of bile duct ligation or subsequent study. A midline abdominal incision was made under general anesthesia induced by intraperitoneal injection of diazepam (5 mg/kg) after intra-muscular administration of Hypnorm (fentanyl/fluanisone; 0.4 mg/kg). For those undergoing bile duct ligation, the common bile duct was isolated, triply ligated, and cut between the ligatures. Sham procedure involved a similar operation, but without ligation or cutting of the bile duct. All subsequent studies were performed at day 28 post operation, when biliary cirrhosis had developed in the animals undergoing bile duct ligation. Histological examination was performed on formalin-fixed liver stained with hematoxylin-eosin to confirm liver cirrhosis. Three weeks after the operation, the animals (BDL and control) were divided into three experimental groups, which received 130 mg/kg/twice a day of N-acetylcysteine (NAC; infusible solution, Celltech, UK), 0.5 mg/kg/twice per day of NG-L-nitro-arginine methyl ester (L-NAME) or an equivalent volume of saline subcutaneously for 1 week.16, 17 In week 4, the rats were anesthetized with isoflurane (1.5 %) and a lead II electrocardiogram (ECG) was recorded (15 minutes) for spectral analysis of heart rate variability (Powerlab, ADInstrument, Australia). The heart of each animal was then removed and the atria were dissected out from isolated hearts in cold oxygenated physiological solution. Blood was also collected in nitrate-free ethylendiamine-tetraacetic acid (EDTA) containing tubes for measurement of plasma nitrite/nitrate concentrations.

Preparation of Isolated Atria.

Isolated rat atria were used as a model system in this investigation because it is amenable to direct measurements of cardiac chronotropic response without being encumbered by confounding physiological mechanisms operative in more complex systems. In brief, the left and right atria were isolated in cold oxygenated physiological solution and suspended under isometric tension of 1,000 mg force in a 25-mL organ bath glass chamber. The temperature of the bathing solution was 37.0 ± 0.1°C, and pH was 7.4. The composition of physiological solution in millimolars was as follows: NaCl, 112; KCl, 5; CaCl2, 1.8; MgCl2, 1; NaH2PO4, 0.5; KH2PO4, 0.5; NaHCO3, 25; glucose, 10; and EDTA, 0.004. The solution was oxygenated with a gas mixture of 95% O2 and 5% CO2. The right atrium, which contains the sinoatriual node, was used for recording the spontaneous atrial beating, and was stimulated by increasing concentrations of isoproterenol from 10−10 to 10−6 mol/L (chronotropic study).

Heart Rate Variability (HRV) Study.

A continuous ECG was recorded from anesthetized rats under 1.5 % isoflurane anesthesia and spectral components of HRV were calculated using Fast Fourier Transform analysis (HRV software, ADInstrument, Australia). It is well known that beat-to-beat variations of heart rate contain information about the activity of the autonomic effectors controlling cardiac output.18 The interest of spectral analysis has focused on the quantitative separation of physiologically relevant mechanisms from a complex output, the beat-to-beat variation in heart rate. Power spectrum analysis of HRV were calculated in three spectral components as following19: (1) High frequency (HF: 1–3 Hz in rats) component which is caused by an inhibition of the vagal tone during inspiration. (2) Low frequency (LF; 0.04–1 Hz in rats) component, which is characterized by an oscillatory pattern, and in rats is mostly representative of cardiac sympathetic activity. (3) Very low frequency (VLF) component which accounts for all other heart rate changes, which might be oscillatory (when the period of oscillation is long, such as circadian rhythms) or chaotic.18 The data of HRV analysis were expressed in normalized units (nu) in order to prevent the influence of alteration in total power of HRV on each component18 [LFnu=LF/(Total power − VLF) and HFnu= HF/(Total power − VLF)]. The LF/HF ratio was also calculated and used as a measure of cardiac sympathovagal balance.18

Tissue Nitrotyrosine Levels.

Protein-bound nitrotyrosine levels were measured in the cardiac atrium, ventricle, liver and brain obtained from controls and rats with biliary cirrhosis using an isotope-dilution mass spectrometric method developed in our laboratory.20, 21 The effect of L-NAME and N-acetylcysteine on the levels of free nitrotyrosine was also measured in ventricular tissue, but not in the atria, since insufficient tissue was obtained. In brief, tissue was homogenized in a mixture of ice-cold saline (2 mL) and chloroform/methanol (2:1) containing 10 ng13C9-nitrotyrosine, and the protein precipitate (middle layer) was isolated by centrifugation at 2000 × g for 30 minutes at 4°C. The supernatant was used for measuring free nitrotyrosine (in ventricles) and the protein precipitates were lyophilized under vacuum for assessment of protein-bound nitrotyrosine. 1–1.5 mg of lyophilized protein was hydrolyzed for 15 hours at 120°C in 1 mL 4 M sodium hydroxide following the addition of 10 ng13C9-nitrotyrosine and 10 μg of2H4-tyrosine as stable isotopic internal standards. These conditions prevent the artifactual nitration of tyrosine that occurs during acidic hydrolysis conditions.20 Following two steps of solid phase extraction, nitrotyrosine and tyrosine were quantitated by stable isotope dilution gas chromatography/negative ion chemical ionization mass spectrometry.20, 21 Results of protein bound nitrotyrosine are expressed as a ratio of nitrotyrosine to tyrosine (pg/μg).

Assessment of Nitration Reactions In Vivo Using Deuterated PHPA as a Probe.

To determine whether N-acetylcysteine inhibits the formation of reactive nitrogen species in vivo, or inhibits the nitration of circulating phenolic compounds, a novel method was employed based on infusion of deuterated para-hydroxyphenylacetic acid (2H6-PHPA) followed by measurement of the urinary excretion of it's nitrated product using mass spectrometry.22, 23 This experimental approach allows us to assess the formation of reactive nitrogen species dynamically, since the formation of nitrated deuterated para-hydroxyphenylacetic acid can only occur through direct attack of its phenolic ring by reactive nitrogen species in vivo. Therefore, in a separate set of experiments, BDL or sham-operated rats were anesthetized with isoflurane (1.5%) and the external jugular vein and the urinary bladder were cannulated with PE-10 and PE-50 tubes respectively.2H6-PHPA (250 nmol) was infused intravenously by bolus injection, and the level of urinary deutrated 3-nitro-para-hydroxy-phenylacetic acid (2H5-NHPA) was measured sequentially during a 4-hour urine collection (nitration of2H6-PHPA produces2H5-NHPA as a result of replacement of a single deuterium of the aromatic ring by the NO2 group) using mass spectrometry.22, 23 The basis for this method is illustrated in Scheme 1.

Scheme 1.

Cardiac Concentrations of S-Nitrosothiols.

A reductive chemiluminescense-based assay was used for measuring cardiac levels of S-nitrosothiols.21 In a separate set of experiments, the left ventricle was perfused with ice-cold perfusion buffer (phosphate saline buffer, pH 7.4 containing a thiol blocking agent N-ethylmaleimide and diethylentriamine-pentaacetic acid of 5 and 1 mmol/L respectively). After homogenization of ≈500 mg of the snapped-frozen tissue in ice-cold perfusion buffer, S-nitrosothiols were determined by a copper (II)/iodine/iodine-mediated cleavage of S-nitrosothiols to NO, which was then quantified by its gas phase chemiluminescent reaction with ozone in a NO analyzer (NOA, Sievers, Boulder, CO) by a method developed in our laboratory.21

Measurement of Cardiac F2-Isoprostanes and Arachidonic Acid.

Since many reactive nitrogen species also initiate lipid peroxidation reactions in vivo we also measured the concentration of esterified F2-isoprostanes in cardiac tissue .24 In short, ≈250 mg whole cardiac tissue was homogenized in a mixture of saline and chloroform/methanol solution containing butylated hydroxytoluene (5%) (to inhibit ex vivo lipid peroxidation), and centrifuged at 3,000g for 10 minutes. This process results in three phases, an upper aqueous phase, separated from the lower lipid containing phase by a ring of protein precipitate. The lower lipid layer was aspirated and, following the addition of 500 pg of2H4-iso-PGFα and 50ng of2H8-arachidonic acid (Cayman Co., Ann Arbor, MI) as internal standards, was dried down under nitrogen, and hydrolyzed in methanolic 15% potassium hydroxide solution (1 hour, 37°C). To extract the F2-isoprostanes, the pH was adjusted to 3.0, and the samples were extracted on a C18 solid-phase extraction cartridge (Elstree, Hertsfordshire, Waters, UK) as described,24 converted to the pentaflurobenzyl ester, purified by thin-layer chromatography and analyzed as the tri-methysilyl ether. Detection was performed by selected ion monitoring gas chromatography negative ion chemical ionization/mass spectrometry with monitoring of ions at m/z 569 and 573.24 To control for the hydrolysis step, the levels of free (i.e., hydrolyzed) arachidonic acid were also measured by mass spectrometry. In brief, arachidonic acid was extracted on a C18 solid-phase extraction cartridge (Waters), converted to the pentaflurobenzyl ester, purified by thin-layer chromatography and analyzed by stable isotope dilution negative ion chemical ionization mass spectrometry by monitoring ions at m/z 303 and 311. The levels of F2-isoprostanes were expressed as the ratio of F2-isoprostanes to arachidonic acid in cardiac tissue homogenates.

Plasma Nitrite/Nitrate Concentrations.

Plasma nitrite/nitrate was measured by a chemiluminescence-based assay. In brief, nitrate was converted to nitrite using nitrate reductase. Samples were then injected into a reaction chamber containing acetic acid and potassium iodide (50 mg/mL) at a ratio of 4:1. This reduces nitrite to NO, which is purged from the refluxing solution by nitrogen and reacts with ozone before analysis by chemiluminescence (NOA). Measurements were calibrated against standard curves of sodium nitrate.

Immunogold Electron Microscopy.

In a separate set of experiments freshly-isolated atria from controls and rats with cirrhosis were excised and immersed in isotonic fixative (4% paraformaldhyde, 0.5% glutaraldehyde, in 0.1 mol/L phosphate buffer, pH 7.4, with 0.1 mol/L sucrose) for electron microscopic immunocytochemistry as has previously described (n = 2).25 Tissues were then infiltrated and embedded in LR white resin. Thin sections (70–90 nm) were cut and mounted on coated nickel grids. The grids were then blocked (0.1% bovine serum antigen, 0.1 mol/L glycine in PBS) for 30 minutes and incubated for 2 hours with rabbit polyclonal antibody raised against nitrotyrosine (anti-nitrotyrosine, 1:500, a kind gift of Dr. Joseph Beckman, University of Alabama at Birmingham, Alabama). Following a series of washes, grids were incubated for 1 hour with 10 nm Immunogold-linked, EM grade, goat anti-rabbit IgG (1:50 dilution). Following another series of washes, grids were successively stained with uranyl acetate and Reynold's lead citrate before visualization with a transmission electron microscope. Electron micrographs (31,000 × original magnification) were scanned using a digital imaging system.

Statistical Analysis.

The results are presented as mean ± SEM. One-way analysis of variance was applied with Student's t test and P values less than .05 were considered statistically significant. In conditions where the parametric test was not permitted (owing to heterogeneity of variances) a non-parametric Kruskal-Wallis test followed by Dunn's multiple comparison test were used for statistical analysis.


In Vitro Study.

Initial studies confirmed that chronotropic responses to isoproterenol were impaired in isolated atria (Fig. 1). Thus, there was a 1.8-fold increase in the EC50 of isoproterenol necessary to evoke the same response as normal rat atria in the spontaneously beating right isolated from rats with cirrhosis (P < .02). This confirms the observation that there are impaired cardiac responses to adrenergic stimulation in rats with cirrhosis. The maximum response (Rmax) to isoproterenol was not statistically different between controls and animals with cirrhosis (477 ± 10 beats/minutes vs. 466 ± 3 beats/minutes in control and animals with cirrhosis, respectively). Administration of N-acetylcysteine or L-NAME for 7 days normalized the EC50 of isoproterenol in rats with cirrhosis (P < .05) as shown in Fig. 1.

Figure 1.

Chronotropic responsiveness to β-adrenergic stimulation. Concentration-dependent responsiveness of isolated atria to isoproterenol in control or rats with cirrhosis treated with saline, N-acetylcysteine (NAC; A) or L-NAME (B). Chronotropic studies were carried out on spontaneous beating isolated right atrium. Data are expressed as mean ± SEM. 6-8 rats were used in each group.

Heart Rate Variability.

HRV depends on the interaction of the cardiovascular autonomic regulatory system. To determine whether there is activation of the sympathetic nervous system or vagal activity in our model of cirrhosis, heart rate variability was analysed using spectral analysis. The results of HRV analysis are shown in Fig. 2 and Table 1. The heart rate of rats with cirrhosis treated with saline was significantly (P < .01) lower than that of control rats. Administration of N-acetylcysteine or the NO synthase inhibitor L-NAME (P < .05) both led to normalization of the heart rate. Spectral analysis of HRV allows one to decompose different components of HRV based on their frequency; the HF spectrum reflects the influence of respiration on cardiac cycle and is a marker of cardiac vagal activity. The LF component is mostly representative of cardiac sympathetic activity in rats. The LF/HF ratio is a marker of cardiac sympathovagal balance.18 As shown in Fig. 2, the LF/HF ratio was increased significantly in rats with cirrhosis compared with controls, consistent with increased activation of the sympathetic nervous system. The same pattern of changes was observed in the LFnu component of heart rate variability (Table 1). Treatment with N-acetylcysteine did not normalize any of the components of HRV in either controls or rats with cirrhosis, suggesting that N-acetylcysteine does not work by altering activation of either the sympathetic or vagal nervous systems. The cardiac sympathetic activity (LFnu) and sympathovagal balance (LF/HF ratio) increased significantly in control rats given L-NAME but not in those with cirrhosis. VLF component of heart rate variability showed a significant reduction following the development of cirrhosis (P < .02). Neither N-acetylcysteine nor L-NAME normalized this reduction of VLF in the experimental groups.

Figure 2.

Basal heart rate (beats/min) and cardiac sympathovagal balance in control or rats with cirrhosis given saline, N-acetylcysteine (NAC; A), or L-NAME (B). An ECG was recorded in anesthetized rats and cardiac sympathovagal balance was assessed by spectral analysis of heart rate variability (LF/HF ratio). Data are shown as Mean ± SEM. (A) P < .05 compared with controls. (B) P < .05 compared with saline-treated BDL rats. 8-16 rats were used in each group.

Table 1. Heart Rate Variability Parameters in Controls and Rats With Cirrhosis Given Saline, N-acetylcysteine (NAC) or L-NAME
  • NOTE. Data are shown as Mean ± SEM.

  • Abbreviations: Total, total; VLF, very low frequency (<0.04 Hz); LF, low frequency (0.04-1 Hz); HF, high frequency (1-3 Hz) power of heart rate variability.

  • *

    P < .05 compared controls. 8-16 rats were used in each group.

Total (ms2)12.3 ± 2.713.2 ± 2.911.6 ± 3.04.1 ± 0.3*4.3 ± 1.1*4.8 ± 0.9
VLF (ms2)10.6 ± 2.611.2 ± 2.510.7 ± 3.12.9 ± 0.6*3.5 ± 1.13.6 ± 1.1
LF nu(%)8.5 ± 1.29.1 ± 1.315.1 ± 1.9*14.7 ± 2.1*14.2 ± 2.5*16.7 ± 2.8*
HF nu(%)91.2 ± 1.390.1 ± 2.284.7 ± 2.185.2 ± 2.185.8 ± 2.584.3 ± 4.8

Nitrotyrosine Concentrations in Cardiac Tissue and Nitration of Deuterated PHPA.

As shown in Table 2, protein-bound nitrotyrosine levels increased significantly in atrial tissue, ventricle and liver but not in the brain of rats with biliary cirrhosis compared with controls. The administration of L-NAME or N-acetylcysteine decreased atrial nitrotyrosine concentrations from 23 ± 3 pg/μg of tyrosine to 15 ± 1 and 17 ± 1 pg/μg respectively (Fig. 3). The same pattern of changes was observed in the ventricles as shown in Fig. 4. L-NAME decreases nitration of cardiac proteins by decreasing synthesis of nitric oxide, and the secondary formation of reactive nitrogen species which cause nitration of cardiac proteins. However, the mechanism by which N-acetylcysteine decreases nitration of proteins is unknown. One potential mechanism could be scavenging of reactive nitrogen species in vivo. To determine whether N-acetylcysteine scavenged and reduced the nitration potential of reactive nitrogen species, rats were infused with 2H6-PHPA, and the formation of2H5-NHPA was measured by mass spectrometry. Biliary cirrhosis was associated with increased formation of2H5-NHPA following intravenous infusion of 2H6-PHPA (P < .01; Fig. 5) consistent with increased formation of reactive nitrogen species in rats with cirrhosis and nitration of the phenolic ring of PHPA. However, administration of N-acetylcysteine did not decrease the nitration of 2H6-PHPA in vivo (Fig. 5). An alternative potential mechanism of N-acetylcystene could involve increased clearance and proteolysis of nitrated proteins in vivo, since others have shown that nitration of proteins is associated with increased proteolysis of the “abnormal” protein.26 To test this, we hypothesized that increased proteolysis would increase levels of free nitrotyrosine in cardiac tissue. Therefore free nitrotyrosine levels were measured in cardiac ventricular tissue of normal rats as well those with cirrhosis ± treatment with N-acetylcysteine. We measured free nitrotyrosine in ventricular tissue, since the tissue concentration of free nitrotyrosine is very low (<2 pg/mg wet tissue), making it undetectable in small tissues such as rat atrium. However, free nitrotyrosine was easily measured in ventricular tissues and the levels were not significantly different in controls and rats with cirrhosis. However, the administration of N-acetylcysteine led to a marked increase in the concentration of free nitrotyrosine in the ventricular tissue in rats with cirrhosis (P < .05; Fig. 4). Thus, whereas these data suggest N-acetylcysteine leads to increased proteolysis of nitrated proteins, the mechanism remains unclear.

Table 2. Comparison of Protein-Bound Nitrotyrosine Levels Between Controls and Rats With Cirrhosis
TissueProtein-Bound Nitrotyrosine (pg/μg tyrosine)
CTRBDLP (t test)
  1. NOTE. Proteins were extracted from different tissues (liver, atrium, ventricle and brain) obtained from sham operated (control) or bile duct ligated (BDL) rats and analyzed by mass spectrometry. Levels are corrected for tyrosine. Data are Mean ± SEM.

  2. *P < .05, compared with control. 6-8 rats were used in each group. ns: non statistically significant.

Liver63 ± 10178 ± 40<0.02
Atrium16 ± 123 ± 3<0.05
Ventricle22 ± 957 ± 8<0.05
Brain9 ± 112 ± 3ns
Figure 3.

Protein-bound nitrotyrosine content of atrial tissue obtained from control and rats with cirrhosis given saline, N-acetylcysteine (NAC; A), or L-NAME (B). Data are shown as mean ± SEM. (A) P < .05 compared with control. (B) P < .05 compared with saline-treated BDL rats. Four to six rats were used in each group.

Figure 4.

Protein-bound (upper panel, A-B) and free (lower panel, C-D) nitrotyrosine content of ventricular tissue obtained from control and rats with cirrhosis treated with saline, N-acetylcysteine (NAC; A,C) or L-NAME (B,D). Data are shown as Mean ± SEM. (A) P < .05 compared with controls. (B) P < .05. (C) P < .01 compared with saline-treated BDL rats. 8-10 rats were used in each group. Protein-bound nitrotyrosine levels are expressed as a ratio to tyrosine (pg/μg).

Figure 5.

Assessment of nitration reactions in vivo using deuterated PHPA (2H6-PHPA) as probe. Injection of 2H6-PHPA into rats with biliary cirrhosis leads to increased nitration of PHPA and formation of2H5-NHPA, which is excreted in urine. Pre-treatment of animals with cirrhosis with N-acetylcysteine (NAC) did not decrease2H5-NHPA formation in vivo (A), and indicates that NAC does not work by scavenging reactive nitrogen species. However, L-NAME, which inhibits nitric oxide synthase, significantly reduced2H5-NHPA formation in rats with cirrhosis (B). Data are expressed as mean ± SEM. 6-8 rats were used in each group. *P < .01 compared with control groups. #P < .05 in comparison with BDL (L-NAME) and control groups.

Cardiac S-Nitrosothiols.

To determine whether there was an increase in S-nitrosation of cardiac proteins we measured the concentration of S-nitrosothiols in cardiac tissue from contols,rats with cirrhosis with treatment with L-NAME or N-acteylcysteine, and rats with cirrhosis but without treatment. There was no difference in cardiac S-nitrosothiols following induction of cirrhosis, and neither N-acetylcysteine nor L-NAME had any effect on the cardiac levels of S-nitrosothiols (Table 3).

Table 3. Cardiac S-nitrosothiol and F2-isoprostanes Concentrations in Control and Cirrhotic Rats Treated With Saline, N-acetylcysteine (NAC) or L-NAME.
  • NOTE. F2-isoprostanes are expressed as a ratio of 8-iso-PGF to arachidonic acid (AA). S-Nitrosothiols are expressed as pmol/g of wet tissue. Data are Mean ± SEM.

  • *

    P < .05, compared with control (saline) group. 6-8 rats were used in each group.

S-nitrosothiol (pmol/g)56 ± 950 ± 949 ± 1161 ± 753 ± 1055 ± 9
F2-isoprostanes (ng/μg AA)1.8 ± 0.21.6 ± 0.21.8 ± 0.12.9 ± 0.5*2.4 ± 0.4*2.2 ± 0.4

Cardiac F2-Isoprostanes.

The ratio of esterified F2-isoprostanes to arachidonic acid was significantly increased in cardiac tissue from rats with cirrhosis at 2.9 ± 0.5ng/μg AA (arachidonic acid) compared to controls (1.8 ± 0.2ng/μg AA, P < .05), consistent with oxidative stress and increased lipid peroxidation in cardiac tissue. Treatment with either N-acetylcysteine or L-NAME had no significant effect on the tissue levels of F2-isoprostanes (Table 3).

Plasma Nitrite/Nitrate Concentrations.

As shown in Fig. 6, there was a significant increase in plasma concentrations of nitrite/nitrate in rats with biliary cirrhosis (13.9 ± 1.7 to 21.7 ± 1.8 μmol/L, P < .01), which decreased following inhibition of NO synthase by L-NAME (13.0 ± 0.6 μmol/L, P < .01 compared with saline-treated rats with cirrhosis). However treatment with N-acetylcysteine did not change the plasma concentrations of nitrite/nitrate levels in rats with biliary cirrhosis (27.4 ± 2.0 μmol/L).

Figure 6.

Plasma nitric oxide end products (nitrate + nitrite) are increased in rats with biliary cirrhosis. Treatment with L-NAME decreases plasma concentrations of nitrite/nitrate, but this is unaffected by NAC or saline. Data are shown as mean ± SEM, and significance as (A) P < .05 compared with controls. (B) P < .05 compared with saline-treated BDL rats. Six to eight rats were used in each group.

Immunogold Electron Microscopy Studies.

Representative electron micrographs from the atria of controls and rats with cirrhosis are shown in Fig. 7. Myofibrillar and mitochondrial organelles are clearly visible. Black electron dense circles (5 nm gold particles) represent positively stained tissue for nitrotyrosine. In control images, relatively sparse immunostaining was observed, with roughly equal staining density between myofibrillar and mitochondrial compartments. However, an obvious increase in staining of myofibrillar and mitochondrial proteins in the atria of rats with biliary cirrhosis was noted (Fig. 7).

Figure 7.

Immunogold electron microscopy for nitrotyrosine in atria from control (A) and bile duct ligated (B) rats. Electron-dense circles indicate positive staining for nitrotyrosine.


This study confirms the observation that cardiac chronotropic function is abnormal in rats with biliary cirrhosis. More importantly, it demonstrates that this abnormality is associated with increased levels of protein-bound nitrotyrosine in cardiac tissue. The administration of either L-NAME or N-acetylcysteine, both of which decrease the tissue levels of nitrotyrosine, led to normalization of heart rate as well as cardiac responsiveness to adrenergic stimulation in rats with biliary cirrhosis. Although this animal model of cirrhotic cardiomyopathy is well established,6, 27 its clinical significance in humans continues to be an area of some debate. In patients with cirrhosis, although resting tachycardia is part of the feature of the hyperdynamic circulation, maneuvers leading to sympathetic activation (e.g., ice-cold skin stimulation, mental stress, tilting and physical exercise), do not evoke an adequate acceleration of heart rate compared with healthy subjects.2–4 Conversely, when the sympathetic responses to tilting or exercise are assessed by the plasma concentrations of norepinephrine, the sympathetic response is enhanced.2, 3 This finding suggests the defect in cardiac responsiveness is located at the receptor and/or post receptor level. This notion is supported by the findings of several studies assessing the effects of ß-adrenergic agents. For example, in patients with cirrhosis, the ED25 of isoproterenol is reported to be threefold higher than healthy controls.28 Rats with biliary cirrhosis are bradycardic,29 and this condition is owing to a significant hypo-responsiveness to the chronotropic effect of catecholamines.7, 8 In the present study, we showed that two independent treatments (N-acetylcysteine and L-NAME), both of which can normalize chronotropic responses to adrenergic stimulation in vitro, also led to normalization of heart rate in vivo (Fig. 2). Moreover spectral analysis of HRV showed increased cardiac sympathetic activity in rats with cirrhosis, which was unchanged following either L-NAME or N-acetylcysteine treatment in rats with cirrhosis. These data suggest impaired end-organ responsiveness to adrenergic stimulation may have a role in pathogenesis of cardiac chronotropic dysfunction in cirrhosis and the effects of L-NAME and N-acetylcysteine are independent of changes in autonomic nervous system activity.

The mechanism of tyrosine nitration in proteins under pathological conditions is an area of active investigation. It can occur through the formation of peroxynitrite or nitryl chloride, formed by the reaction of nitric oxide and hypochlorous acid, both of which can react with tyrosine in proteins to form nitrotyrosine.13 Biliary cirrhosis is associated with enhanced nitric oxide production6 as well as increased activity of myeloperoxidase, which catalyzes the synthesis of hypochlorous acid.30 Nitric oxide reacts rapidly with superoxide anion (O2) to form peroxynitrite which can nitrate tyrosine residues in proteins.31 There is now good evidence that liver disease is associated with oxidative stress and increased formation of reactive oxygen species.32, 33 Ljubuncic et al. reported an increased formation of lipid peroxidation products in the cardiac tissue of rats with biliary cirrhosis compared with controls.33 This finding may be partly owing to enhanced generation of superoxide by NADPH oxidase in response to angiotensin II,34 endothelin,35 bile acids,36 or can occur secondary to the formation of peroxynitrite.24 The observation that there is increased tyrosine nitration in cirrhosis in liver as well as cardiac tissue, but not in brain suggests that nitrative stress is a generalized phenomenon in cirrhosis, which can involve different organs including liver and cardiovascular system. In the liver, for instance, it has recently been shown that tyrosine nitration contributes to inactivation of hepatic glutamine synthetase which in turn leads to the development of hyperammonemia in rats with cirrhosis and sepsis.37

We have previously observed that administration of N-acetylcysteine normalizes the systemic hemodynamic abnormalities in rats with portal hypertension,16 and lipoic acid (another low molecular weight thiol antioxidant) had a similar effect in rats with chronic bile duct ligation.32 Therefore, we investigated whether N-acetylcysteine could improve cardiac function in rats with biliary cirrhosis, and made the critical observation that administration of N-acetylcysteine led to normalization of bradycardia in vivo, as well as improved cardiac responses to adrenergic stimulation in vitro (Figs. 1 and 2). These data are of interest, because they have therapeutic implications, even though the mechanisms involved remain poorly understood. We observed that both N-acetylcysteine and L-NAME can each markedly decrease the levels of tissue nitrotyrosine. Because nitration of tyrosine can alter the protein's function, these data suggest nitration of cardiac proteins may lead to abnormal cardiac responses in cirrhosis. Reactive nitrogen species can also react with cysteine residues in proteins to form S-nitrosated proteins.12 Therefore, we also measured the levels of S-nitrosothiols in cardiac tissue of rats with biliary cirrhosis; however there were no significant differences in cardiac levels of S-nitrosothiols between controls and rats with cirrhosis. These data are consistent with the fact that certain reactive nitrogen species such as peroxynitrite are poor S-nitrosating agents, but potent at causing nitration.38

Both L-NAME and N-acetylcysteine can modulate the cardiac function and decrease cardiac nitrotyrosine levels in biliary cirrhosis. However, the effects of N-acetylcysteine is not attributable to downregulation of NO synthesis because the plasma concentrations of nitrite + nitrate (NO end-products) were still significantly elevated in rats with cirrhosis treated with N-acetylcysteine whereas L-NAME significantly decreased nitrate/nitrite concentrations in both groups (Fig. 6). The decrease of protein-bound nitrotyrosine after N-acetylcysteine treatment may be related to the ability of N-acetylcysteine, to either inhibit the formation of reactive nitrogen species or to increase the proteolytic degradation of the nitrated proteins. Using 2H6-PHPA as a probe to study formation of reactive nitrogen species in vivo, we could initially show enhanced nitration reactions in vivo and a corresponding increased formation of its nitrated product2H5-NHPA in rats with cirrhosis. However, based on this approach, we were unable to show decreased reactive nitrogen species formation following N-acetylcysteine administration as the urinary levels of2H5-NHPA were un-changed in N-acetylcysteine treated animals (Fig. 5). Because these data suggest N-acetylcysteine does not scavenge reactive nitrogen species in our model, the effect of N-acetylcysteine on nitrotyrosine levels could be explained if it enhances the degradation of nitrated proteins. Souza et al. have shown that nitration of tyrosine residue(s) in proteins is sufficient to induce an accelerated degradation of the modified proteins to release free amino acids as well as free nitrotyrosine.26 This study suggested that if N-acetylcysteine increased proteolysis of nitrated proteins then free nitrotyrosine levels may be increased in animals treated with N-acetylcysteine. This finding was confirmed when the levels of free nitrotyrosine were measured in cardiac ventricular tissue and found to be increased (Fig. 4). This outcome suggests that N-acetylcysteine may upregulate proteolytic degradation of proteins containing nitrotyrosine leading to a decrease in protein-bound nitrotyrosine content and corresponding to an increase in free nitrotyrosine levels, and is consistent with reports that cellular redox state is a major modulator of proteosme activity both in vivo and in vitro.39, 40

The mechanism by which protein nitration leads to abnormal cardiac function in cirrhosis is not clear, however a very similar mechanism has been recently postulated in cardiac dysfunction associated with high aldosterone levels.41 Also, Weber et al.have recently shown that chronic aldosterone/salt treatment is associated with a time-dependent generation of nitrotyrosine in cardiomyocytes, which was attenuated by the administration of N-acetylcysteine.41 The concept that nitration of cardiac proteins leads to cardiac dysfunction has also been described in AIDS-related cardomyopathy,42 and diabetes-associated cardiac dysfunction.43 Therefore, the development of new therapies, which prevent or modify nitration of cardiac proteins in cirrhosis, may have important implications in other types of cardiac dysfunction associated with systemic inflammation such as sepsis. Whereas the observation that increased nitration of cardiac proteins in cirrhosis is of interest, the question arises which proteins are nitrated and are they functionally important? In a recent report, Kanski et al. employed a proteomic approach to identify the cardiac proteins that undergo age-dependent protein tyrosine nitration.44 Among the identified proteins are important mitochondrial enzymes responsible for ATP production and metabolism; desmin, which is involved in sliding of myocardial filaments and cardiac contraction, is also included.44 Although we have tried to identify which proteins are nitrated in the current study (data not shown), we have not been able to obtain reproducible data, and at present the identity of the nitrated proteins remains unknown. However, the immunogold studies also suggest mitochondrial proteins are the target for reactive nitrogen species in the atria of rats with cirrhosis in a similar way to that observed in AIDS or diabetes mellitus.

We conclude that cirrhosis leads to increased nitration of cardiac proteins, oxidation of cardiac lipids, and abnormal cardiac function. The observation that two independent treatments, which decrease cardiac nitrotyrosine levels, also lead to normalization of cardiac function, suggests nitration of cardiac proteins leads to abnormal cardiac function.


The authors wish to thank Mr Innes R Clatworthy (Electron Microscopy Unit, Royal Free Campus, UCL, UK) for the excellent preparation of the electron microscopic images.