Article Text

This article has a correction. Please see:

Download PDFPDF

Bacterial DNA activates cell mediated immune response and nitric oxide overproduction in peritoneal macrophages from patients with cirrhosis and ascites
  1. R Francés1,
  2. C Muñoz1,
  3. P Zapater2,
  4. F Uceda3,
  5. I Gascón3,
  6. S Pascual3,
  7. M Pérez-Mateo3,
  8. J Such3
  1. 1Department of Immunology, Hospital General Universitario, Pintor Baeza s/n, Alicante, Spain
  2. 2Department of Clinical Pharmacology, Hospital General Universitario, Pintor Baeza s/n, Alicante, Spain
  3. 3Liver Unit, Hospital General Universitario, Pinto Baeza s/n, Alicante, Spain
  1. Correspondence to:
    Dr J Such
    Liver Unit, Hospital General Universitario, Pintor Baeza s/n, 03010 Alicante, Spain;


Background and aims: Translocation of intestinal bacteria to ascitic fluid is probably the first step in the development of episodes of spontaneous bacterial peritonitis in patients with cirrhosis. We have recently reported the detection of bacterial DNA in blood and ascitic fluid from patients with advanced cirrhosis, what we consider as molecular evidence of bacterial translocation. Several studies have shown the immunogenic role of bacterial DNA in vitro, and we hypothesised that the presence of bacterial DNA could activate the type I immune response in peritoneal macrophages from these patients, leading to greater cytokine synthesis (interleukin (IL)-2 and IL-12, tumour necrosis factor α, and interferon γ) and effector molecules such as nitric oxide.

Methods: Peritoneal macrophages obtained from patients with cirrhosis and culture negative non-neutrocytic ascitic fluid were collected and characterised by flow cytometry. Inducible nitric oxide synthase, nitric oxide levels, and cytokine production were measured by immunoenzymometric assays in basal and harvested conditions according to the presence/absence of bacterial DNA.

Results: The ability of peritoneal macrophages to synthesise nitric oxide and levels of all cytokines were significantly increased in patients with bacterial DNA. There was a positive correlation between inducible nitric oxide synthase and nitric oxide levels.

Conclusions: The presence of bacterial DNA in patients with decompensated cirrhosis is associated with marked activation of peritoneal macrophages, as evidenced by nitric oxide synthesising ability, together with enhanced cytokine production.

  • ascites
  • bacterial translocation
  • cirrhosis
  • cytokines
  • nitric oxide
  • SBP, spontaneous bacterial peritonitis
  • AF, ascitic fluid
  • BT, bacterial translocation
  • bactDNA, bacterial DNA
  • NK, natural killer
  • IFN-γ, interferon γ
  • NO, nitric oxide
  • iNOS, inducible NO synthase
  • PMN, polymorphonuclear leucocytes
  • LPS, lipopolysaccharide
  • MoAbs, monoclonal antibodies
  • TNF-α, tumour necrosis factor α
  • IL, interleukin

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Bacterial infections are frequently observed complications arising in patients with cirrhosis, and among them, spontaneous bacterial peritonitis (SBP) is probably the most relevant.1,2 SBP probably occurs as a consequence of repeated access of bacteria from the intestinal lumen through a haematogenous route reaching the mesenteric lymph nodes3,4 and, eventually, gaining access to ascitic fluid (AF) in a process that has been termed bacterial translocation (BT).5 Supporting this hypothesis, in a previous work we detected the presence of bacterial DNA (bactDNA) fragments simultaneously in blood and AF in as many as 32% of patients with advanced cirrhosis and sterile non-neutrocytic AF,6 a fact that we interpret as molecular evidence of BT.

BactDNA includes a repeated series of unmethylated CpG motifs that bind to the toll-like receptor 9, thus becoming a potent activator of cells of the innate immune system, namely macrophages, dendritic cells, and natural killer cells,1,7 and leading to increased synthesis of proinflammatory cytokines and effector molecules. Similarly, CpG motifs stimulate natural killer (NK) cells to synthesise interferon γ (IFN-γ),8 which in turn enhances the toxicity of other agents such as lipopolysaccharide (LPS)9 and prime nitric oxide (NO) synthesis through expression of the inducible form of NO synthase (iNOS)10 that has been reported to play an important role in the macrophage mediated response to infectious agents.11,12

As we detected the presence of bactDNA in a subgroup of patients with decompensated cirrhosis, we hypothesised that peritoneal macrophages obtained from these patients should be primed for higher NO and cytokine release. Thus the aims of this study were to evaluate the NO synthesising ability, together with the cytokine cascade probably involved in the process, in peritoneal macrophages from our series of patients, according to the presence of bactDNA.



Since April 2000, all patients with cirrhosis and culture negative non-neutrocytic AF have been consecutively included in a prospective study in our Liver Unit to detect and identify the presence of bactDNA in blood and/or AF, according to the methodology previously described.6 A series of 10 consecutively admitted patients with or without bactDNA in blood and AF were included in the study protocol.

Cirrhosis was diagnosed by histology or by clinical, laboratory, and/or ultrasonographic findings. Exclusion criteria were the presence of culture positive blood or AF, neutrocytic AF (⩾250 polymorphonuclear leucocytes (PMN)/μl), signs or symptoms of systemic inflammatory response syndrome according to previously published criteria,13 upper gastrointestinal bleeding or intake of antibiotics in the preceding two weeks, including selective intestinal decontamination with norfloxacin, hepatocellular carcinoma and/or portal thrombosis, alcoholic hepatitis, and refusal to participate in the study. The ethics committee of Hospital General Universitario approved the study protocol and all patients gave informed consent for inclusion in the study.

Blood was obtained for routine haematological, biochemical, and coagulation studies. Simultaneously, large volume paracentesis was performed in all patients on admission under aseptic conditions, following the usual procedures,13 and samples for routine biochemical study and PMN count were obtained. Total protein, albumin, leucocyte count, and PMN count were performed on all AF specimens. Both blood and AF were inoculated at the bedside in aerobic and anaerobic blood culture bottles, 10 ml each.14 Samples of blood and AF were inoculated in rubber sealed pyrogen free tubes (Endo Tube ET; Chromogenix AB, Vienna, Austria).

Peritoneal macrophages culture

A volume of 800–1500 ml of AF, collected after therapeutic paracentesis, was centrifuged at 1500 g for 10 minutes to obtain the cellular pellet. After washing with phosphate buffer saline (PBS) three times, more than 80% of the macrophage population was counted from a total number of 10–16 million cells, in all cases using Testsimplets prestained slides (Roche Diagnostics Ltd, Sussex, UK). Cell viability was also evaluated by trypan blue (Sigma, Madrid, Spain) and resulted in more than 95% viability. Cells (4×106) were separated for macrophage characterisation by flow cytometry; 1×106 cells were prepared to obtain the cell lysates for iNOS quantification according to the manufacturer’s instructions and 5×106 macrophages were resuspended in 5 ml of phenol red free RPMI 1640 medium (Gibco BRL, Life Technologies, Paisley, UK) supplemented with 10% human serum AB (BioWhittaker, Walkersville, Maryland, USA), 100 IU/ml penicillin/streptomycin, and 2.5 μg/ml amphotericin B (Gibco BRL). Macrophages (2×106) were incubated as controls, and 2×106 macrophages were cultured adding 0.2 μg of LPS from E coli serotype 0111:B4 (Sigma) in a 24 well plate for 24 hours at 37°C in a humidified atmosphere with 5% CO2. Supernatants were obtained by centrifugation at 2500 g for five minutes and stored at −80°C until assay. The maximum period of time between being frozen and thawing was three months and in the current study none of the specimens was thawed and refrozen prior to analysis.

Immunostaining of cell surface and characterisation of the monocyte/macrophage population

At the indicated time of culture, cells were harvested and washed once in ice cold PBS, suspended in 500 μl of PBS, and distributed (50 μl per tube) to polystyrene round bottom tubes (Becton Dickinson, San Diego, California, USA) for immunolabelling. Surface determinant immunostaining was performed with unfixed cells using fluorochrome conjugated monoclonal antibodies (MoAbs) against the surface determinants. The combinations CD45 PerCP/CD3 FITC, CD45 PerCP/CD33 PE, CD45 PerCP/CD14 FITC, and CD14 FITC/CD33 PE were added for better discrimination between monocytes and lymphocytes in AF mononuclear cells during flow cytometry analysis. Cells were incubated in the dark for 15 minutes, washed in cold PBS, fixed, and permeabilised with IntraPrep permeabilisation reagent (Immunotech Beckman Coulter, Marseille, France). Then, cells were washed in PBS and stained (30 minutes in the dark) for intracellular cytokines using FITC conjugated MoAbs against tumour necrosis factor α (TNF-α) and interleukin (IL)-6, and PE conjugated MoAbs against IL-8 and IL-1-β. FITC and PE conjugated isotype controls were used in parallel. After cells were washed in PBS they were suspended with paraformaldehyde 1% in PBS for flow cytometry analysis. Samples were analysed in a FACS sort flow cytometer (Becton Dickinson, Immunocytometry Systems, Palo Alto, California, USA) using Cellquest (version 3.2) software. Typically, list mode data for 20 000 events for CD45+ cells in a “live gate” mode were acquired.

Measurement of NO production and human iNOS levels

The sum of the NO metabolites nitrite (NO2) and nitrate (NO3) is widely used as an index of NO generation and expressed as NOx levels per ml, which corresponds to 106 cells in this study. NOx levels in basal AF, and in supernatants from cultured macrophages, were calculated by measuring conversion of NO3 to NO2 by the enzyme nitrate reductase using an ELISA assay (R&D Systems, Minneapolis, USA) based on the Griess reaction that absorbs visible light at 540 nm. All samples were tested in duplicate and values were corrected by running samples with culture media without macrophages to assess background NOx levels. Quantitative determination of iNOS concentrations was performed through the Quantikine Human iNOS Immunoassay (R&D Systems) in cell lysates, obtained according to the manufacturer’s instructions and expressed as U/ml.

Measurement of IFN-γ, TNF-α, IL-2, and IL-12

Immunoenzymometric assays for quantitative measurement of human IFN-γ, TNF-α, IL-12, and IL-2 in AF samples were performed by handling Biosource IFN-γ EASIA kit (Biosource Europe SA, Nivelles, Belgium), human TNF-α HS Quantikine, and IL-12 and IL-2 Quantikine (R&D Systems) according to the manufacturer’s instructions. All samples were tested in duplicate and read at 450 nm and 490 nm in a ThermoMax microplate reader (Molecular Devices, Sunnyvale, California, USA).

Statistical analysis of experimental data

All observations are reported as mean (SD). Statistical differences in basal characteristics between groups were analysed using the χ2 test for categorical data applying Yates’ correction when required, the Mann-Whitney U test for quantitative data without a normal distribution according to the Komolgorov-Smirnov test, and the Student’s t test for variables with a normal distribution. The ANOVA test with the post hoc Tukey test was used for multiple comparisons. Bivariate correlations were evaluated using the Pearson test. All p values were two tailed. A p value <0.05 indicated statistical significance. Analyses were performed with the SPSS statistical package (SPSS Inc. version 11.0, Chicago, Illinois, USA).


Characteristics of patients and laboratory data

All patients fulfilling the inclusion and exclusion criteria were considered for entry into the study. Ten consecutively admitted patients in whom we did not detect or detected the presence of bactDNA comprised groups I and II, respectively. Clinical and analytical characteristics of the patients in both groups at admission are shown in table 1. No significant differences were observed in any of the parameters between the two groups of patients.

Table 1

Basic clinical and analytical characteristics of the cohort distributed according to the absence (group I) or presence (group II) of bacterial DNA in serum and ascitic fluid

BactDNA was always detected simultaneously in blood and AF. Similarity in sequences between blood and AF was more than 99.5% in all cases, and identifications obtained from the NCBI Database by the advanced BLAST search15 included E coli (n = 7), Klebsiella (n = 2), and S aureus (n = 1).

Upper gastrointestinal endoscopy was performed at the index admission or in the preceding three months in eight patients in group I and in seven patients in group II. No differences were observed in the presence or size of oesophageal varices or in portal hypertensive gastropathy between the two groups of patients.

Patients were followed up during admission from inclusion in this study. Two patients from group II died during admission; one from terminal liver failure and one from acute pancreatitis. One patient in group I developed culture negative spontaneous bacterial peritonitis and one patient in group II developed a vertebral abscess. Infections developed seven and 12 days, respectively, after the initial paracentesis.

Gating of monocyte population and production of intracellular cytokines

Monocyte populations were defined on the basis of cell granularity and fluorochrome conjugated MoAbs CD14-FITC and CD33-PE (fig 1). No differences between groups were found in an ANOVA test, with the stimuli and MoAbs as factors in the analysis. The average monocyte population activated was 76.1 (SD 14.7)%.

Figure 1

Gating of the monocyte population. Monocyte populations were defined on the basis of cell granularity (SSC-H) and the fluorochrome conjugated monoclonal antibodies anti-CD33-PE (FL2-H) and anti-CD14-FITC (FLH-1). Regions R1 and R2 correspond to monocytes and polymorphonuclear cells, respectively.

AF NOx and cytokine levels

Basal AF NOx levels in patients from group II were higher than those observed in patients from group I although values did not reach statistical significance, probably due to the reduced number of cases (table 2). NOx levels measured in the supernatant of non-stimulated macrophages from patients in group II were significantly higher than those observed in patients in group I. Furthermore, addition of cellular stimulants to the culture media induced a marked increase in NOx levels in both groups of patients but values only reached significance in patients in group I (table 2).

Table 2

NOx, TNF-α, IFN-γ, and IL-12 levels in ascitic fluid in patients in groups I and II

Basal values of all four cytokines were significantly higher in patients in group II than in patients in group I. Values obtained in the supernatant of non-stimulated cultures were significantly higher in patients in group II than in patients in group I, with the exception of IFN-γ. However, we did not observe significant differences for any of the cytokines in the supernatant of stimulated cultures between groups. Addition of LPS to cultures induced a significant increase in all cytokines in comparison with values observed without stimulation in both groups of patients (table 2).

Although the limited number of bacteria isolated (seven E coli, two K pneumoniae, one S aureus) precludes any statistical analysis, we did not observe any difference in the immune response according to the type of bacteria detected in patients in group II.

Human iNOS concentrations in cell lysates

Relative mass values for naturally occurring iNOS under all three experimental conditions are detailed in table 2. Statistically significant differences were observed between values obtained at baseline and in non-stimulated cultures between both groups of patients. Addition of LPS to the culture medium induced a significant increase in iNOS concentration compared with samples obtained without stimulation in both groups of patients (table 2).

A direct and significant correlation was found between basal iNOS and NOx levels in the overall group of patients (r2 = 0.72, p = 0.03) (fig 2) and also between basal values of iNOS and all cytokines studied, with the exception of TNF-α (IFN-γ r2 = 0.77, p = 0.04; TNF-α r2 = 0.65, p = 0.09; IL-12 r2 = 0.79, p = 0.001; IL-2 r2 = 0.84, p = 0.001).

Figure 2

Correlation between basal inducible NO synthase (iNOS) and NOx (sum of nitric oxide metabolites, nitrite and nitrate) levels in peritoneal macrophages obtained from ascitic fluid in patients with decompensated cirrhosis.


In this study, we have shown that peritoneal macrophages, obtained from patients with cirrhosis and AF, and the presence of bactDNA are primed to synthesise significantly higher amounts of NO than macrophages obtained from patients without bactDNA, and this is associated with marked activation of the cytokines implicated in the type 1 immune response.

Bacterial infections are common complications in patients with advanced cirrhosis, and SBP is the most frequent and clinically relevant.1 The classical pathogenic theory of SBP suggests that bacteria of intestinal origin move across the intestinal wall,5 reaching mesenteric lymph nodes and other organs. Bacteria can then obtain access to AF, and a SBP episode may eventually develop if the local bactericidal mechanisms are insufficient to mount an adequate response.16,17

We have recently described the presence of bactDNA in patients with cirrhosis and culture negative non-neutrocytic AF, a fact that we interpret as molecular evidence of BT.6 It is likely that bacteria reaching AF may activate the cellular component of the innate immune system, modifying its inflammatory response and NO synthesising ability, a situation similar to what has been described in in vitro models.9,18 Thus the aim of this study was to assess the macrophage immune response according to the presence or absence of bactDNA in patients with cirrhosis.

BactDNA is characterised by the presence of short repeated sequences of unmethylated CpG dinucleotides. Different experimental studies have shown the immunomodulatory role of these fragments, capable of inducing a similar immune response to that produced by a complete microorganism in vitro,19–21 thus becoming potent activators of cells of the innate immune system (namely, NK, macrophages, and dendritic cells) by binding toll-like receptor 9.7

BactDNA mediated macrophage activation leads to TNF-α, IL-6, and IL-12 synthesis.20,22 These last two cytokines activate B and T cells23,24 and modulate IFN-γ production by contributing to NK cell activation in vivo and in vitro.8,25,26 As IFN-γ receptors are also expressed on the macrophage surface, the presence of increased levels of IFN-γ further enhances TNF-α production and upregulates the transcription of the iNOS gene27–32 leading to an increase in macrophage NO synthesising capability.

Peritoneal macrophages from cirrhotic patients with ascites and a previous episode of SBP have been shown to be activated, with NO overproduction,33 a situation similar to that described here, as suggested by the fact that NOx levels were higher in patients with molecular evidence of BT (group II) compared with patients in group I (table 2). Also, the existence of a direct and significant correlation between iNOS and NOx levels strongly suggests that NOx detected in the supernatant of cultures of peritoneal macrophages is iNOS derived. One of the best described functions of iNOS is its role in the macrophage mediated response to infectious agents.11,12 Furthermore, iNOS produces physiological concentrations of NO in the nanomolar range.34 Micromolar concentrations observed in our study are consistent with these data, considering overactivation of macrophages in response to bacterial DNA.

As described previously, the increase in NO synthesis is not an isolated phenomenon but a consequence of a complex sequence of events in which several cytokines play a significant role. In fact, NO has been shown to downregulate cytokine release from macrophages.35 Supporting this, the reported differences in the ability of macrophages to synthesise NO is associated with a marked inflammatory scenario. We observed significant basal differences in all four cytokines evaluated (table 2) between both groups of patients, which indicates two clearly differentiated populations according to the absence (group I) or presence (group II) of bactDNA.

We observed a positive correlation between IL-2, IL-12, and IFN-γ and iNOS values, and the cytokine dependent induction of iNOS has been reported previously.36 Elevated production of IFN-γ by CpG DNA activated NK cells is mediated by IL-1237 and, in fact, we observed a significant correlation between basal levels of these two cytokines, which clearly suggests activation of a type I immune response in these patients. Supporting this assumption, TNF-α, which is directly synthesised by macrophages and also enhanced by IFN-γ, was significantly higher in patients in group II compared with group I.

In summary, peritoneal macrophages from patients with cirrhosis and ascites and the presence of bactDNA are primed for higher cytokine and NO synthesis. Our data suggest identification of a new subset of patients, with a marked pre-activation of the cellular component of the immune system, probably in response to the immunogenicity of CpG motifs present in bactDNA.


Parts of this work were supported by an unrestricted grant from Schering Plough, and grants GV00-070-12 from Conselleria de Cultura, Generalitat Valenciana, PI 021291 from Fondo de Investigaciones Sanitarias, and Instituto de Salud Carlos III (C03/02), Spain.


Linked Articles

  • Commentary
    B A Runyon
    BMJ Publishing Group Ltd and British Society of Gastroenterology