Ammonia impairs neutrophil phagocytic function in liver disease

Authors

  • Debbie L. Shawcross,

    1. Liver Failure Group, The University College London (UCL) Institute of Hepatology, Division of Medicine, University College London, London, UK
    Search for more papers by this author
  • Gavin A. K. Wright,

    1. Liver Failure Group, The University College London (UCL) Institute of Hepatology, Division of Medicine, University College London, London, UK
    Search for more papers by this author
  • Vanessa Stadlbauer,

    1. Liver Failure Group, The University College London (UCL) Institute of Hepatology, Division of Medicine, University College London, London, UK
    Search for more papers by this author
  • Stephen J. Hodges,

    1. Liver Failure Group, The University College London (UCL) Institute of Hepatology, Division of Medicine, University College London, London, UK
    Search for more papers by this author
  • Nathan A. Davies,

    1. Liver Failure Group, The University College London (UCL) Institute of Hepatology, Division of Medicine, University College London, London, UK
    Search for more papers by this author
  • Caroline Wheeler-Jones,

    1. Department of Veterinary Basic Sciences, Royal Veterinary College, University of London, London, UK
    Search for more papers by this author
  • Andrew A. Pitsillides,

    1. Department of Veterinary Basic Sciences, Royal Veterinary College, University of London, London, UK
    Search for more papers by this author
  • Rajiv Jalan

    Corresponding author
    1. Liver Failure Group, The University College London (UCL) Institute of Hepatology, Division of Medicine, University College London, London, UK
    • Liver Failure Group, The UCL Institute of Hepatology, Division of Medicine, University College London, 69-75 Chenies Mews, London, WC1E 6HX, UK
    Search for more papers by this author
    • fax: 02073800405


  • Potential conflict of interest: Nothing to report.

Abstract

Hyperammonemia is a feature of liver failure, which is associated with increased risk of infection. The aims of the present study were to determine in vitro, in rats fed an ammoniagenic diet and in patients with cirrhosis, whether induction of hyperammonemia results in neutrophil dysfunction. As hyperammonemia produces cell swelling, we explored the role of the osmoregulating, p38 mitogen-activated protein kinase (p38MAPK) pathway in mediating this neutrophil dysfunction. Neutrophils were isolated from blood of healthy volunteers and incubated with either 75 μM ammonia or phosphate-buffered saline. Both groups were studied under hyponatremic conditions and/or with the addition of p38MAPK modulators. Neutrophil phagocytosis was measured in naive rats and rats fed an ammoniagenic diet and in patients with stable cirrhosis given placebo (n = 8) or an amino acid solution inducing hyperammonemia (n = 8). Cell volume and phagocytosis was analyzed by fluorescent-activated cell sorting using fluorescein isothiocyanate–labeled E. coli. p38MAPK phosphorylation was measured by western blotting. In healthy neutrophils incubated with ammonia and in rats fed an ammoniagenic diet, neutrophils showed evidence of swelling, impaired phagocytosis, and increased spontaneous oxidative burst compared to controls. Phagocytosis was significantly impaired in patients with induced hyperammonemia compared to placebo. The effects of hyperammonemia and hyponatremia were synergistic. The p38MAPK intracellular signaling pathways were activated in healthy neutrophils exposed to ammonia in association with increased burst activity. Neutrophil phagocytic dysfunction was abrogated by the addition of a p38MAPK agonist. Conclusion: Ammonia produces neutrophil swelling and impairs neutrophil phagocytosis. The p38MAPK intracellular signaling pathway has been shown to be important in mediating the ammonia-induced neutrophil dysfunction. (HEPATOLOGY 2008.)

Patients with cirrhosis are particularly prone to infection,1, 2 which complicates the course of about 40% of hospitalized patients and is a major cause of death.3, 4 The increased risk of infection is secondary to impairment of several host defense mechanisms including neutrophil function.5, 6 Our group has recently shown that neutrophil dysfunction with high spontaneous oxidative burst and reduced phagocytic capacity was present in patients with cirrhosis and alcoholic hepatitis, and was associated with a significantly greater risk of infection, organ failure, and mortality.7

Hyperammonemia is common in liver disease and its severity relates to the degree of liver dysfunction.8 Recent studies suggest that in addition to ammonia, infection is a common precipitant of the syndrome of hepatic encephalopathy.9–11 Studies have shown that ammonia may reduce neutrophil chemotaxis by causing a decrease in the affinity of the neutrophil chemotactic peptide receptor for its ligand.12 Furthermore, ammonia inhibits neutrophil phagocytosis, degranulation, and oxygen metabolism in human periodontal infection13 and decreases stimulated oxidative burst in normal neutrophils exposed to 10 mM ammonium chloride without inducing a pH change.14 This leads us to ask if it is the hyperammonemia associated with liver disease itself that alters neutrophil function, thereby increasing the susceptibility of the patient with cirrhosis to infection.

Additionally, hyponatremia is an independent predictor of mortality and may predispose to infection.15, 16 Both hyperammonemia and hyponatremia have the ability to produce swelling of susceptible cells by altering the osmotic balance.17–19 It is known that p38 mitogen-activated protein kinase (p38MAPK) acts as an osmosensor which is activated in response to increased cell hydration or swelling.20, 21 Therefore, we chose to explore the effect of ammonia on neutrophil functions such as phagocytosis and oxidative burst and whether activation of the p38MAPK plays a role in this process by studying normal neutrophils in vitro incubated with different concentrations of ammonia and under hyponatremic conditions, as well as in a hyperammonemic rat model. Finally, in a subgroup of an ancillary study in which patients were primarily recruited to determine the neuropsychological effects of induction of hyperammonemia by administration of an oral bolus of a specifically prepared solution that mimics the amino acid composition of the hemoglobin molecule,22 16 patients were randomized in a double-blind fashion into two groups; one group received amino acid solution and the other received placebo. Neutrophil phagocytosis and changes in blood ammonia were measured at 0 and 4 hours following administration of the amino acid solution or placebo.

Abbreviations

FACS, fluorescence-activated cell sorting; FITC, fluorescein isothiocyanate; HD, high protein/ammoniagenic; IgG, immunoglobulin G; MAPK, mitogen-activated protein kinase; MPO, myeloperoxidase; PBS, phosphate-buffered saline.

Materials and Methods

As endotoxin was convincingly demonstrated to induce neutrophil dysfunction in our previous study,7 in all experiments strict precautions were taken to avoid endotoxin contamination by working aseptically and using endotoxin-free equipment.

Effect of Ammonia on Neutrophil Phagocytosis and Oxidative Burst

Ammonia-Fed Rats.

All animal experiments were conducted according to Home Office guidelines under the UK Animals in Scientific Procedures Act 1986. Male Sprague-Dawley rats (body weight 230-280 g) were housed in the unit and given free access to water, with a light/dark cycle of 12 hours, at a temperature of 19-23°C and humidity of approximately 50%. Rats (n = 4) were administered a high protein/ammoniagenic (HD) diet or standard liquid rodent feed (naive) (n = 4) ad libitum for 5 days prior to termination. The HD diet consisted of a liquid rodent feed (Bioserve, Frenchtown, NJ) and a tailor-made mixture mimicking the amino acid composition of the hemoglobin molecule (4 g/kg/day Nutricia; Numico, Cuijk, The Netherlands) as described,22 mixed with commercially available gelatin to prevent sedimentation. This regimen produces chronic hyperammonemia to levels that are similar to those observed in bile duct–ligated animals.23 Control rats were administered standard chow. Rats were killed by exsanguination under anesthesia (Hypnorm 200 μL/kg intramuscularly), 20 minutes after administration of diazepam (1 mg/kg intraperitoneally). Blood was withdrawn from the descending aorta and 300 μL of whole blood was withdrawn into lithium-heparin pyrogen-free tubes for analysis of phagocytosis and oxidative burst. The remainder of the blood was immediately put into ice-cold tubes containing heparin or ethylene diamine tetraacetic acid that were centrifuged at 4°C, and the plasma was collected and stored at −80°C until assayed. Samples (200 μL) were analyzed for plasma biochemistry using a Cobas Integra 400 multianalyzer with the appropriate kits (Roche Diagnostics, West Sussex, UK).

Effect of Induced Hyperammonemia on Neutrophil Phagocytosis in Patients with Cirrhosis.

The study was undertaken with full approval of the local research ethics committee and with written informed consent from each patient in accordance with the Declaration of Helsinki (1989) of the World Medical Association. The safety of administering an amino acid solution to patients with cirrhosis has been extensively studied11, 22 without any complication such as development of overt encephalopathy.

Patients.

Patient characteristics are summarized in Table 1. A total of 16 patients were randomized by the closed envelope method in a double-blind fashion into two groups; one group received the amino acid solution (n = 8) and the other received placebo (n = 8). Both the investigator and patient were blinded to the content of the solution that was administered. All patients had clinical and histological evidence of cirrhosis. Patients were excluded if they had clinical evidence of overt hepatic encephalopathy, diabetes, cardiovascular disease, renal dysfunction (serum creatinine >150 μmol/L), serum sodium <130 mmol/L, serum potassium <3.2 mmol/L or >5 mmol/L, concomitant neurological disease, recent gastrointestinal bleeding (within the previous 4 weeks), malignancy, or pregnancy. Patients had to be abstinent from alcohol and benzodiazepines for at least 1 month prior to the study. All patients were metabolically stable, had no evidence of clinically overt infection, and were studied following an overnight fast.

Table 1. Patient Characteristics
 Amino Acid Solution (n = 8)Placebo (n = 8)
  1. Data are presented as mean ± standard error.

  2. Abbreviations: PSC, primary sclerosing cholangitis; M, males.

Age (years)55.3 ± 7.550.7 ± 4.4
Sex5M4M
Etiology  
 Alcohol43
 Hepatitis C32
 Cryptogenic12
 PSC1
Child-Turcotte-Pugh grade  
 A23
 B54
 C11
Previous encephalopathyNoneNone
Body mass index29 ± 1.827 ± 2.9
Ascites55
Portal hypertension and varices78
Venous ammonia (μmol/L)74 ± 669 ± 11
Oral Amino Acid Solution.

Hyperammonemia was induced by administration of an oral bolus of 75 g of a specifically prepared solution (Nutricia; Numico, Cuijk, The Netherlands; product number 24143) that mimics the amino acid composition of the hemoglobin molecule.22 The solution was freshly made in 200 mL of water and xanthum gum was added to prevent sedimentation. The placebo solution was comprised of water and xanthum gum.

Measurement of Ammonia and Neutrophil Phagocytosis.

Venous blood was taken immediately prior to and 4 hours following administration of the amino acid or placebo solution. Phagocytosis was quantified using the Phagotest (Orpegen Pharma, Heidelberg, Germany) as described below. Plasma was obtained by centrifugation, deproteinized with trichloroacetic acid, and stored at −80°C for spectrophotometric determination of ammonia (CobasMiraS; Hoffmann-La Roche, Basel, Switzerland) at a later date.

In Vitro Studies.

Venous blood was collected aseptically into pyrogen-free tubes (BD Vacutainer Lithium-Heparin, 60 U per tube; Becton Dickinson, Plymouth, UK) from 12 fasted healthy nonsmoking volunteers and was placed immediately in ice for 10 minutes to precool the neutrophils to 0°C. All in vitro studies were performed in pyrogen-free conditions and utilized nonpyrogenic tubes (Corning, Inc., Corning, NY). A total of 2 mL of blood was gently mixed on a blood rotator for 1 minute, then incubated at 37°C in a water bath with 0-500 μM NH4Cl for 90 minutes. Each concentration was tested in triplicate. After 30 and 60 minutes of incubation, the samples were gently remixed for 1 minute. The effect of incubation with ammonia for 90 minutes with up to 500 μM NH4Cl on extracellular pH, chloride concentration, and partial pressure of carbon dioxide (pCO2) was measured using a Radiometer ABL 700 Series blood gas analyzer.

Quantitation of Neutrophil Phagocytosis and Oxidative Burst.

Neutrophil function was examined in neutrophils isolated from whole blood and incubation with ammonia was performed in whole blood to more closely resemble physiological conditions in patients with liver dysfunction.

The Phagotest.

This in vitro test allows the quantitative determination of leukocyte phagocytosis in heparinized blood. This measures the overall percentage of monocytes and granulocytes showing phagocytosis. The Phagotest was used to measure phagocytosis by using fluorescein isothiocyanate (FITC)-labeled opsonized E. coli bacteria as described.7 In brief, 100 μL of whole blood was mixed with 20 μL of FITC-labeled E. coli bacteria (2 × 107) at 37°C for 20 minutes. Fluorescence of bacteria at the cell surface was quenched using trypan blue, red cells were lysed, and neutrophils were washed with sterile phosphate-buffered saline (PBS) prior to analysis. To identify neutrophils, cells were stained with anti-CD16-PE immunoglobulin G (IgG)1 (IOTest; Beckman Coulter) or anti-CD11b (BD Pharmingen) and analyzed on a fluorescence-activated cell sorting (FACS) analyzer (FACScan; Becton Dickinson, San Jose, CA). Phagocytosis is expressed as the percentage of cells undergoing phagocytosis (double-positive cells) and the number of bacteria engulfed per individual granulocyte (geometric mean of fluorescence intensity). Samples were analyzed in triplicate. The intraassay and interassay coefficient of variation was 1.6% and 10.1%, respectively.

The Burstest.

The Burstest (Orpegen Pharma, Heidelberg, Germany) was used to measure the percentage of phagocytic cells that produce reactive oxidants with or without E. coli as described.7 In brief, 100 μL of heparinized whole blood were incubated for 20 minutes with 20 μL of opsonized E. coli (2 × 107), phorbol 12-myristate 13-acetate (8.1 μM) or sterile PBS at 37°C. Formation of oxidative reactants was monitored by oxidation of dihydrorhodamine (DHR 123) to rhodamine, which gives green fluorescence. To identify neutrophils, cells were stained with anti-CD16-PE IgG1 and analyzed by FACS. Neutrophils were gated on forward-scatter and side-scatter characteristics and, subsequently, the percentage of CD16-positive cells producing reactive oxygen metabolites (green fluorescence) was calculated. Samples were analyzed in triplicate. The intraassay coefficient of variation was 4.7% and 2.4% for spontaneous and stimulated burst, respectively. The interassay coefficient of variation was 5.4% and 4.2% for spontaneous and stimulated burst, respectively.

Myeloperoxidase Assay.

Myeloperoxidase (MPO) activity, a further measure of oxidative burst, was performed using a classical guaiacol peroxidation assay.24, 25 It is affected by the presence of even a small number of eosinophils since eosinophils have an even higher eosinophil peroxidase activity and content than neutrophils, therefore this assay utilizes 3-amino-1,2,4-triazole, which inhibits eosinophil peroxidase. Blood was incubated with 0 μM or 75 μM ammonium chloride (NH4Cl) for 90 minutes and the granulocytes were isolated using Polymorphoprep (Axis-Shield, Oslo, Norway). The MPO buffer was prepared by adding 0.02% weight/volume cetyltrimethylammonium bromide and 13 mM guaiacol to 200 mL of PBS, pH 7.0. MPO buffer (3 mL) was then added to a cuvette, along with 50,000 neutrophils and 6 μL of 2 mM 3-amino-1,2,4-triazole. The reaction was then started by adding 4 μL of 1 μM hydrogen peroxide while the reaction mixture was maintained at 37°C by a heated mount; the increase in optical density was followed kinetically at 470 nm using a spectrophotometer.

Effect of Hyponatremia on Neutrophil Phagocytosis

The effect of incubation in hyponatremic blood on neutrophil function alone or coincubation with 75 μM NH4Cl for 90 minutes in the presence of hyponatremia was studied in three healthy individuals by buffering whole blood with sterile PBS containing 124 mM instead of 138 mM sodium chloride.

Neutrophil Viability and Volume Changes Following Incubation with Ammonia

Neutrophil Isolation.

Neutrophils were aseptically isolated from whole blood by a one-step gradient centrifugation. In brief, whole blood (4 mL) was layered over 5 mL of Polymorphoprep in a pyrogen-free tube and spun for 30 minutes at 400g at room temperature. Neutrophils were harvested from the second interface and washed with sterile PBS. Cells were counted in a Thoma hemocytometer and resuspended in PBS at a density of 1 × 106 cells in 400 μL.

Cell Viability.

Cell viability was tested by the trypan blue exclusion test. Ammonia-induced apoptosis or necrosis was assessed by staining with propidium iodide and annexin V-FITC (BD Pharmingen) as described elsewhere.26, 27 In brief, isolated neutrophils from whole blood were resuspended in 100 μL of sterile binding buffer (10 mM 4-2-hydroxyethyl-1-piperazine ethanesulfonic acid/NAOH, pH 7.4, 140 mM NaCl, and 2.5 mM CaCl2). A total of 5 μL of Annexin V-FITC and 10 μL of propidium iodide were added and the solution was gently mixed. The tubes were then incubated for 15 minutes at room temperature (20-25°C) in the dark, then analyzed by FACS.

Changes in Cell Volume.

The changes in neutrophil volume were investigated by flow cytometry using FACS and Cellquest software, using a method previously reported for hepatocyte cell volume regulation.28 Neutrophil volume was expressed as arbitrary units after normalization of 100,000 control neutrophils. The intraassay coefficient of variation was 4.7% and the interassay coefficient of variation was 5.4%.

Effect of Ammonia on Neutrophil p38MAPK Phosphorylation

Whole blood was incubated for 90 minutes with 0 μM or 75 μM NH4Cl. The selective p38MAPK agonist (1 μM isoproterenol; Sigma, Poole, UK) or antagonist (10 μM SB203580; Merck Biochemicals, Nottingham, UK) were used as positive and negative controls, respectively. Each incubation was performed in triplicate and reproduced in three healthy individuals. Neutrophils were isolated by a one-step gradient centrifugation as described above. Protein separation and transfer were performed using a NuPAGE precast gel system (Invitrogen, Ltd., UK). Rabbit polyclonal IgGs for detection of phosphorylated p38MAPK and total p38MAPK (Santa Cruz Biotechnology, Santa Cruz, CA) with a secondary goat polyclonal antibody to rabbit IgG, horseradish peroxidase conjugated (Hycult Biotechnology, the Netherlands) were all used at a dilution of 1:1,000. Protein bands were visualized by using ECL Advance western blotting detection reagents (Amersham) and Hyperfilm (GE Healthcare, UK).

Alterations in Neutrophil Phagocytosis, Oxidative Burst, and Cell Volume Following Incubation with Modulators of p38MAPK

The effects of a selective p38MAPK agonist (1 μM isoproterenol) or antagonist (10 μM SB203580) were examined in three healthy individuals. A total of 2 mL of heparinized whole blood was incubated at 37°C in a water bath with 10 μM SB203580 dissolved in sterile dimethyl sulfoxide or 1 μM isoproterenol, and 75 μM NH4Cl for 90 minutes. Control samples were incubated with an equivalent volume of sterile dimethyl sulfoxide. Flow cytometry, Phagotest, and Burstest were then performed.

Statistics

The data are expressed as the mean and standard error of the mean. Groups were compared using the Student t test and Mann-Whitney U test as appropriate. For comparison of more than two groups, one-way analysis of variance with Tukey's (equal variance) or Dunnett C (nonequal variance) post hoc analysis was used. A P < 0.05 was considered as statistically significant. GraphPad Prism 4.0 (GraphPad Software Inc., San Diego, CA) software was used.

Results

Effect of Ammonia on Neutrophil Phagocytosis and Oxidative Burst in Ammonia-Fed Rats and Patients with Cirrhosis and Induced Hyperammonemia

Ammonia-Fed Rats.

Table 2 outlines the biochemical indices, plasma ammonia, and percentage of neutrophils undergoing phagocytosis and stimulated and spontaneous oxidative burst of the study rats. Plasma ammonia concentration in HD rats was significantly higher compared to naive control rats (P < 0.05). Phagocytosis was significantly impaired, with only 55% of neutrophils undergoing phagocytosis in the HD rats, compared to 83% in naive rats (P < 0.05). There was no difference in the mean number of bacteria ingested by the neutrophils in either group. Spontaneous oxidative burst activity was significantly increased, with 31% of neutrophils undergoing burst in the HD rats, compared to 8% in naive rats (P < 0.001). No difference in burst activity was seen between the stimulated neutrophils in the HD and naive rats.

Table 2. Plasma Biochemistry and Neutrophil Function of Control and Ammonia-Fed (HD) Rats
 Control Rats (naive)Ammonia-Fed (HD) RatsP Value
  • Data are presented as mean ± standard error. P < 0.05 was considered statistically significant.

  • *

    P < 0.05;

  • **

    P < 0.001.

  • Abbreviations: ns, not significant; AST, aspartate aminotransferase; ALT, alanine aminotransferase.

AST (IU/L)258.8 ± 38.9241.5 ± 26.1ns
ALT (IU/L)56.8 ± 1.641.8 ± 7.2ns
Total bilirubin (μmol/L)11.5 ± 0.616.2 ± 1ns
Sodium (mmol/L)130.9 ± 0.6133.5 ± 0.5ns
Creatinine (μmol/L)30.5 ± 0.942.1 ± 0.6ns
Plasma ammonia (μmol/L)31.2 ± 6.2200.1 ± 16.70.03*
Neutrophils undergoing phagocytosis (%)83 ± 3.355 ± 70.03*
Neutrophils undergoing stimulated oxidative burst (%)73 ± 1.174 ± 3.7ns
Neutrophils undergoing spontaneous oxidative burst (%)8 ± 1.931 ± 5.3<0.001**

Induced Hyperammonemia in Patients with Cirrhosis.

Patients in both groups were well-matched (Table 1) and did not describe any significant side effects from the amino acid solution.

Ammonia.

Ammonia concentrations at baseline were similar in the two groups and increased in the patients administered the amino acid solution from a baseline mean of 74 ± 6 μmol/L to 111 ± 9 μmol/L at 4 hours (P < 0.001). No significant changes were observed in the patients administered the placebo solution (baseline: 69 ± 11 μmol/L to 73 ± 7 μmol/L; P = 0.4).

Phagocytosis.

Phagocytosis at 4 hours was significantly impaired in patients administered the amino acid solution compared to those given placebo, from 76 ± 2.7% to 62 ± 2.5% (P = 0.0002) (Fig. 1). There was no difference in the mean number of bacteria ingested by the neutrophils in either group and no difference in phagocytosis at baseline between the groups (76 ± 2.7% versus 76 ± 3.6%).

Figure 1.

A graphical representation to illustrate the reduced phagocytosis at 4 hours in each of the eight patients administered the amino acid solution compared to the eight given the placebo solution. A paired t test showed this to be significant at P = 0.0002.

Effect of Ammonia on Neutrophil Phagocytosis and Oxidative Burst in Healthy Individuals

Neutrophils Exposed to Ammonia.

Phagocytosis.

Phagocytosis was significantly impaired by a mean of 11.9 ± 2.7% following incubation with 75 μM NH4Cl (P < 0.0001). The critical ammonia concentration that had a significant effect on phagocytosis was 75 μM (Fig. 2A -D). In the representative experiment shown in Fig. 2, 78% (Fig. 2B) and 60% (Fig. 2C) of neutrophils are undergoing phagocytosis (neutrophils falling in the right upper quadrant) compared to 95% in the control panel (Fig. 2A), while 17% (Fig. 2B) and 36% (Fig. 2C) are falling in the left upper quadrant compared to 0.7% (Fig. 2A) following incubation with 75 and 500 μM NH4Cl, respectively. The geometric mean of fluorescence intensity (mean number of bacteria phagocytosed per neutrophil) was not affected by incubation with rising ammonia concentrations. Incubation with up to 500 μM NH4Cl did not alter the expression of CD11b during phagocytosis. This effect was seen in the entire granulocyte population but, as >95% of the granulocytes were neutrophils and the remaining <5% were basophils and eosinophils, only the effects on the gated neutrophil population are reported in this work.

Figure 2.

(A-C) Typical flow cytometry plots to demonstrate the neutrophils gated (n) undergoing phagocytosis following incubation for 90 minutes with 0, 75, and 500 μM ammonium chloride (NH4Cl). (D) The mean percentage of neutrophils undergoing phagocytosis, following incubation with 0-150 μM ammonium chloride (NH4Cl) in 12 healthy individuals.

Oxidative Burst.

Stimulated oxidative burst following the addition of opsonized E. coli approached 95%-98% in all samples in both the presence (Fig. 3A) and absence of ammonia (Fig. 3B), which was similar to that produced by adding the positive control phorbol 12-myristate 13-acetate with and without ammonia (data not shown). Spontaneous oxidative burst in the absence of bacteria was increased following incubation with 75 μM NH4Cl by 10.7 ± 1.8% (P < 0.0005) compared to controls (Fig. 3C,D). MPO activity, a further measure of oxidative burst, increased significantly following incubation with 75 μM NH4Cl (P < 0.001).

Figure 3.

(A,B) Typical plots obtained from neutrophils challenged with FITC-labeled E. coli (stimulated). In (A), 95% of control neutrophils (no ammonia) have undergone oxidative burst and have been gated in the right upper quadrant; in (B), this is not changed by the presence of 75 μM NH4Cl (98%). (C,D) Typical plots obtained from FITC-labeled neutrophils with no bacteria added (spontaneous burst). In (C), 5% have undergone burst in the resting state, whereas (D) in the presence of 75 μM NH4Cl, this has increased to 12% of the neutrophils.

Cell Viability.

Incubation of whole blood with ammonia did not affect viability of neutrophils (1.6%, 1.8%, and 1.8% nonviable for 0, 75, and 500 μM NH4Cl, respectively). Up to 5 hours of incubation with 75 μM NH4Cl resulted in <2% apoptotic neutrophils (annexin V-FITC–positive, propidium iodide–negative) and <2% necrotic neutrophils (annexin V-FITC–positive and propidium iodide–positive). When incubated for >5 hours, >20% necrotic cells were observed. Furthermore, short-term incubation with ammonia for up to 120 minutes at concentrations of NH4Cl up to 500 μM did not induce apoptosis or cell death.

Cell Volume.

Cell volume increased by 2 ± 0.4% with ammonia compared to controls (P < 0.05) (Fig. 4).

Figure 4.

Graphical representation of the percentage change in neutrophil volume compared to control neutrophils following incubation for 90 minutes with hyponatremia (sodium 124 mM) alone, ammonia (75 μM NH4Cl) alone, coincubation with hyponatremia and ammonia, the selective p38MAPK antagonist SB203580 (10 μM) alone, coincubation of p38MAPK antagonist with hyponatremia and/or ammonia, the selective p38MAPK agonist isoproterenol (1 μM) alone, and coincubation of p38MAPK agonist with hyponatremia and/or ammonia as measured by flow cytometry using FACS.

Neutrophils Exposed to Hyponatremia.

Phagocytosis and Oxidative Burst.

Phagocytosis (in three healthy controls) was impaired in a hyponatremic environment (124 mmol/L) (Fig. 5). Hyponatremia did not have an effect on stimulated or spontaneous oxidative burst.

Figure 5.

Diagrammatic representation to show the mean percentage of neutrophils undergoing phagocytosis in the presence or absence of 75 μM NH4Cl, 124 mM hyponatremia, p38MAPK agonist (1 μM isoproterenol) and p38MAPK antagonist (10 μM SB203580) in three healthy individuals.

Cell volume.

Cell volume increased by 3.4 ± 0.6% in a hyponatremic environment compared to controls (P < 0.05) (Fig. 4).

Neutrophils Exposed to Ammonia and Hyponatremia.

Phagocytosis.

The effect of exposing neutrophils to ammonia in a hyponatremic environment was synergistic. Phagocytosis (in three healthy controls) was impaired by 24.4 ± 5% (P < 0.05), compared to 10.1 ± 1.6% in a hyponatremic environment (124 mmol/L) alone and 10.1 ± 1.8% with ammonia (75 μM NH4Cl) alone (Fig. 5).

Cell Volume.

Impaired phagocytosis was associated with an increase in cell volume to 5.7 ± 1.2% with ammonia in a hyponatremic environment compared to controls (P < 0.05). Cell volume increased to 2 ± 0.4% with ammonia alone and 3.4 ± 0.6% in a hyponatremic environment alone (Fig. 4).

Effect of p38MAPK Modulators on Neutrophil Function.

p38MAPK phosphorylation.

Neutrophils isolated from whole blood which had been incubated with 75 μM NH4Cl for 0 and 90 minutes, demonstrated increased expression and activation of phosphorylated (active) p38MAPK. Expectedly, incubation of whole blood with the selective p38MAPK agonist isoproterenol for 90 minutes increased phosphorylation of p38MAPK. No change was observed following incubation with the p38MAPK antagonist, SB203580 (Fig. 6).

Figure 6.

Typical western blot from a healthy individual showing increased expression of phosphorylated p38MAPK following incubation of whole blood with 75 μM NH4Cl compared to control incubated with sterile PBS for 90 minutes. Incubation of whole blood with the selective p38MAPK agonist (1 μM isoproterenol) and antagonist (10 μM SB203580) for 90 minutes were used as positive and negative controls, respectively. α-Tubulin was used as the house keeping protein.

Phagocytosis.

The percentage of neutrophils in whole blood of healthy controls undergoing phagocytosis was 91% when the blood was incubated in the absence of NH4Cl; following incubation with 75 μM NH4Cl, this percentage reduced to 63 ± 3.8% (P < 0.05). Coincubation of whole blood with 75 μM NH4Cl and 10 μM p38MAPK antagonist SB203580 had no significant effect on neutrophil phagocytosis (64 ± 6.2%) above that observed with ammonia alone (63 ± 3.8%). The selective p38MAPK agonist isoproterenol prevented the ammonia-induced impairment of phagocytosis seen with 75 μM NH4Cl (63 ± 3.8%) with 96 ± 0.7% of neutrophils phagocytosing, similar to the levels seen in controls (91 ± 1.2%) (Fig. 7A).

Figure 7.

(A) Bar graph representing the effect of the selective p38MAPK agonist isoproterenol (1 μM) and the selective p38MAPK antagonist SB203580 (10 μM) on the mean percentage of neutrophils undergoing phagocytosis, with and without 75 μM NH4Cl, in three healthy individuals. (B) Bar graph representing the effect of the selective p38MAPK agonist isoproterenol (1 μM) and the selective p38MAPK antagonist SB203580 (10 μM) on the mean percentage of neutrophils undergoing spontaneous oxidative burst, with and without 75 μM NH4Cl, in three healthy individuals.

Oxidative Burst.

Incubation of whole blood with 75 μM NH4Cl, with and without the selective p38MAPK antagonist SB203580 and the selective p38MAPK agonist 1 μM isoproterenol, did not alter the number of neutrophils undergoing oxidative burst with E. coli. However, the p38MAPK antagonist SB203580 prevented the ammonia-induced increase in spontaneous oxidative burst reducing the levels back to those in controls (5%) (Fig. 7B). The percentage of neutrophils undergoing spontaneous oxidative burst was not altered by coincubation with the selective p38MAPK agonist.

Cell volume.

The increase in cell volume by ammonia and hyponatremia were abrogated by the selective p38-MAPK agonist isoproterenol and exacerbated by the addition of SB203580, a selective p38MAPK antagonist (Fig. 4. Thus, the p38MAPK agonist isoproterenol led to a decrease in neutrophil volume compared to controls and prevented ammonia-induced neutrophil swelling. The p38MAPK antagonist SB203580, however, increased neutrophil volume by 2.7 ± 0.43% compared to controls (P < 0.05), which was exacerbated by the addition of 75 μM NH4Cl to 5.2 ± 0.92% (P < 0.05), impairing normal neutrophil osmoregulation (Fig. 4.

Discussion

The main finding of our study was the observation that hyperammonemia impaired neutrophil function by decreasing phagocytosis and increasing spontaneous oxidative burst. In rats fed an ammoniagenic diet, impaired phagocytosis and increased spontaneous oxidative burst of neutrophils was demonstrated compared to controls, reproducing the in vitro observations. In patients with cirrhosis, phagocytosis was significantly impaired within 4 hours of receiving an ammonia load. This is the first direct evidence that pathophysiological concentrations of ammonia impair neutrophil function in the absence of other physiological and biological factors that could contribute to neutrophil dysfunction in liver disease.

The observed reduction in phagocytosis occurred in the absence of an extracellular pH change, and without any evidence of increased apoptosis or necrosis. Our studies support previous observations which indicate that ammonia may decrease neutrophil function through alterations in its ability for ligand binding, thereby affecting chemotaxis. Additionally, it has been hypothesized that the chronicity of H. pylori infection may relate to the inability of the mucosal granulocytes to phagocytose the bacteria because of its ability to generate ammonia.29 The recent studies by the Felipo group30 have suggested that hyperammonemia may produce long lasting dysfunction of the lymphocytes through upregulation of protein kinase, dependent on the cyclic adenosine monophosphate pathway.

We hypothesized that the detrimental effects of ammonia on neutrophil function may be due to its effects on cell volume regulation. In keeping with the hypothesis, we observed consistent and significant evidence of cell swelling induced by pathophysiological concentrations of ammonia. Several studies have defined the mechanisms involved in phagocytosis and show clearly that an increase in rounding of cells that is introduced by swelling would reduce its phagocytic function31 (reviewed in Ref.32). Furthermore, a similar reduction in phagocytic capacity following induction of hyponatremia, which is a well-known stimulus for cell swelling, argues for the role of cell swelling being an important mechanism involved in phagocytic dysfunction. Importantly, the additive effect of hyponatremia and hyperammonemia in causing a more pronounced cell swelling and further phagocytic dysfunction may have important clinical implications and may help to explain the poor outcomes of patients with liver failure (who are known to be hyperammonemic) who become hyponatremic.15, 16

To explore the potential mechanism of how ammonia-induced cell swelling may impair neutrophil function, we studied the p38MAPK pathway, which is a well-preserved intracellular signaling mechanism activated by numerous stimuli. p38MAPK is ubiquitous throughout the human body and is critical in cell volume regulation.21 Hypoosmotic and hyperosmotic stress20, 33 have been shown to activate intracellular p38MAPK in many cell types and hyponatremia has been shown to induce cell swelling in association with activation of p38MAPK.34 However, the specific relationship between ammonia, cell swelling, and p38MAPK has only been previously described in hepatocytes35 and astrocytes.36 Therefore, we aimed to explore the role of the p38MAPK pathway in the context of ammonia-induced neutrophil dysfunction. We first hypothesized that ammonia, which leads to cell swelling, activated the p38-MAPK pathway. We were able to show evidence of activation of the p38MAPK by demonstrating increased amount of phosphorylated p38MAPK in ammonia-treated neutrophils. To investigate the functional importance of this finding, the effect of p38MAPK modulators on neutrophil function was tested. Unexpectedly, the impaired phagocytic capacity induced by ammonia, hyponatremia, or in combination was not restored by the p38MAPK antagonist SB203580 but the p38MAPK agonist isoproterenol was able to restore phagocytic capacity. Importantly, the p38MAPK agonist also prevented both ammonia-induced and hyponatremia-induced cell swelling, supporting the hypothesis that cell swelling is the central mechanism associated with the reduced neutrophil phagocytic function. This observation suggests that activation of the p38MAPK pathway is not the cause but rather the consequence of ammonia-induced neutrophil swelling. Our results can therefore be interpreted as showing that ammonia leads to cell swelling, which leads to phosphorylation of p38MAPK. These observations are also supported by the findings of Jayakumar et al.,36 who explored the effects of ammonia on the p38MAPK pathway in primary cultured astrocytes exposed to a supraphysiological dose of ammonia (5 mM NH4Cl). They also observed that ammonia-induced cell swelling was associated with significant increase in the phosphorylation of p38MAPK, which could be ameliorated by coculture with MAPK inhibitors.36

The increase in spontaneous oxidative burst by similar concentrations of ammonia may indicate a paradox, as an increase in burst is usually taken as an indication of a neutrophil being in a state of increased function. However, as we have shown recently in patients with alcoholic hepatitis, increased spontaneous burst is associated with increased risk of infection and consequent mortality.7 Expectedly, the ammonia-induced increased spontaneous burst was associated with increased MPO activity, which has the potential to increase oxidative stress often observed in patients with liver failure. The observation that the p38MAPK antagonist SB203580 prevented the ammonia-induced increase in spontaneous oxidative burst suggests that the increased spontaneous burst may well be a consequence of ammonia-induced phosphorylation of p38MAPK.

Taken together, the results of our studies leads us to postulate that ammonia induces neutrophil swelling through an unknown mechanism; this swelling activates the p38MAPK signaling pathway, which is key in cellular osmoregulation, thereby attempting to reduce neutrophil swelling. The consequence of this protective mechanism is the activation of oxidative burst, which can be prevented by inhibition of the p38MAPK signaling pathway. A selective agonist of p38MAPK, however, abrogates the ammonia-induced neutrophil swelling and impairment in phagocytosis. This hypothesis is summarized in Fig. 8.

Figure 8.

Diagram illustrating the hypothesis that ammonia induces neutrophil swelling and activates the osmoregulator p38MAPK, which abrogates the ammonia-induced swelling and impairment of phagocytosis. This is at the expense of inducing spontaneous oxidative burst in unstimulated neutrophils, which can be inhibited by the p38MAPK inhibitor SB203580.

In summary, hyperammonemia and hyponatremia act synergistically to significantly impair neutrophil phagocytosis possibly by affecting neutrophil cell volume. The p38MAPK intracellular signaling pathways are activated by cell swelling, resulting in increased spontaneous oxidative burst and release of MPO. Our results suggest that prompt treatment of hyperammonemia and hyponatremia may not only reduce oxidative injury but also improve neutrophil function in patients with cirrhosis.

Ancillary