OUP user menu

Helicobacter pylori lipopolysaccharide in the IL-2 milieu activates lymphocytes from dyspeptic children

Wieslawa Rudnicka, Agnieszka Jarosinska, Leokadia Bak-Romaniszyn, Anthony Moran, Izabela Planeta-Malecka, Torkel Wadstrom, Magdalena Chmiela
DOI: http://dx.doi.org/10.1016/S0928-8244(03)00023-3 141-145 First published online: 1 May 2003

Abstract

In this study, we assessed the proliferative response of peripheral blood mononuclear leukocytes (PBML) from 33 children/young adolescents with chronic dyspepsia, to H. pylori LPS in the presence and absence of IL-2 as a T cell growth factor. A rapid urease test (RUT) and a presence of Helicobacter-like organisms (HLO) in the biopsy specimens allowed us to distinguish RUT/HLO-positive (17/33) and -negative (16/33) patients. H. pylori LPS alone induced a proliferation of PBML from 4 out of 33 dyspeptic patients. IL-2 increased the prevalence of the response to LPS to 59% and 74% of RUT/HLO-positive and -negative patients, respectively. PBML from RUT/HLO-positive patients responded significantly less intensively to H. pylori LPS in the presence of IL-2, to IL-2 alone and to H. pylori LPS+IL-2. However, there was no difference in PHA-driven proliferation of PBML from the patients of those two groups. A negative correlation between the responsiveness to H. pylori LPS of PBML and occurrence of type B inflammation in gastric mucosa was demonstrated. The results suggest a contribution of H. pylori LPS to an outcome of H. pylori infection. It is speculated that H. pylori LPS by an activation of immunocompetent cells may reduce gastric inflammation, decrease bacterial load and prolong H. pylori infection.

Keywords
  • Helicobacter pylori
  • LPS
  • IL-2
  • Cell proliferation
  • Inflammation

1 Introduction

Lipopolysaccharides (LPS) endotoxins of the outer leaflet of Gram-negative bacteria have a variety of biological effects on numerous types of mammalian cells. B lymphocytes are activated by LPSs to differentiate into antibody secreting plasma cells in the absence of specific antigens [1]. The macrophages activated by LPSs produce a variety of cytokines (TNF, IL-1, IL-6), procoagulant activators, eicosanoids, reactive oxygen and nitrogen intermediates. LPS may also activate T cells. However, the majority of the effects of LPSs on T cells seem to be derived indirectly from the interaction between T cells and LPS-stimulated macrophages and from the mediators released from the LPS-activated accessory cells and cytokine-activated T cells [2]. A role of an interaction between macrophage CD14 antigen, a cellular binding site for LPSs, with IL-2 produced by T cells, in the activation of human monocytes has been reported [3]. Like the cell envelopes of other Gram-negative bacteria, that of Helicobacter pylori contains LPS. However, H. pylori LPS shows significantly lower mitogenicity, pyrogenicity, toxicity [4] and activity in mediation of macrophage activation [5] than LPSs of enterobacterial bacteria do. On the basis of structural study, the lower immunological activity of H. pylori LPS has been explained by underphosphorylation and underacylation of its lipid A moiety compared with those from enterobacterial LPS [6]. Considering the chronic nature of H. pylori infections it was suggested that H. pylori LPS evolved its present structure as a consequence of adaptation to chronic infection of the gastric mucosa. Perhaps the evolution has led to co-expression of common complex carbohydrates, such as Lewis X and Lewis Y determinants, which are present on the LPSs of most H. pylori strains and in human cell surface glycoconjugates of blood cells and gastric mucosa [7,8]. Based on this molecular mimicry Appelmelk et al. [9] speculate on the role of autoimmune mechanisms in H. pylori-associated type B gastritis. Recently, it has been shown that an intravenous administration of a purified LPS from H. pylori induced in rats an increased vascular permeability in cardiac, renal, hepatic, pulmonary and intestinal tissues, which could be prevented by the selectively inducible nitric oxide synthase inhibitor [10,11]. In contrast, H. pylori LPS-induced gastric mucosal inflammatory response in rats was enhanced by aspirin [12]. During natural H. pylori infections in humans, antibodies (mainly of IgG class) to polysaccharide chains of LPSs of those bacteria are produced [13,14]. However, the antigenicity of the polysaccharide varied, depending on the strain. It has been found that smooth H. pylori strains isolated from the tumors of patients with gastric cancer showed significantly lower antigenicity than smooth strains derived from patients with chronic gastritis and gastric and duodenal ulcers [13]. If anti-H. pylori LPS IgG and IgA antibodies are produced in H. pylori-infected subjects one may speculate that an activation of T lymphocytes or at least of T helper cells by bacterial LPS is very likely. In this study, we examined the proliferative response to H. pylori LPS of peripheral blood mononuclear leukocytes (PBML) from children and young adolescents with chronic dyspepsia, seropositive for anti-H. pylori IgG, in cultures with or without IL-2 as a growth factor of T cells. All patients were examined by endoscopy and a type B gastritis in gastric mucosa was evaluated. Gastric mucosal biopsy specimens were used for detecting bacterial urease in a rapid urease test (RUT) and Helicobacter-like organisms (HLO) in histological sections. The results suggest that H. pylori LPS may influence an outcome of H. pylori infection by activation of T cells. Probably, activated lymphocytes may, at least in some patients, facilitate a control of bacterial growth and diminish inflammation in gastric mucosa by releasing cytokines, possibly of Th2 type.

2 Materials and methods

2.1 Patients

A total of 33 children and young adolescents (aged 7–17 years) with chronic dyspepsia were studied. Endoscopic examination was used to determine a type B inflammation in the gastric mucosa. Four biopsy specimens were collected and used for the RUT and HLO in sections stained with hematoxylin and eosin. The blood was collected for cell cultures and serology. The sera of all patients were monitored for IgG antibodies to a glycine acid extract (GE) of H. pylori G27 and to a recombinant fusion protein carrying an immunodominant portion of H. pylori CagA expressed in E. coli (rCagA). Both H. pylori G27 (CagA and VacA positive) bacteria and rCagA were obtained by courtesy of A. Covacci, Institute for Immunobiology Research, Siena, Italy. The levels of anti-GE and anti-CagA IgG were determined in ELISA as previously described [15]. Ethical approval was obtained for this study.

2.2 Cell cultures

Heparinized venous blood was collected before endoscopic examination. Peripheral blood mononuclear leukocytes were separated by Lymphoprep (Nycomed, Pharma, AJ, Norway) gradient centrifugation as recommended by the manufacturer. The RPMI-1640 medium containing 10% heat-inactivated fetal calf serum, 200 mM l-glutamine and 0.05 mg ml−1 gentamicin was used for cell cultures. The PBML were stimulated for 3 days (37°C, 5% CO2) with 0.01 mg ml−1H. pylori LPS of Lewis XY type [6] alone, IL-2 (10 U ml−1) alone, LPS+IL-2, PHA (0.01 mg ml−1) or in the culture medium with no stimuli. PBML proliferation was estimated on the basis of 3H-thymidine incorporation in DNA and measured in Wallac MicroBeta scintillation counter after harvesting cells on fiber filters (Wallac, Turku, Finland).

2.3 Statistics

Data were analyzed by Statistica 5.0 PL program. The results are expressed as mean cpm/culture±standard deviation. Differences between the intensity of proliferation were determined by an unpaired t test. The Pearson χ2 test corrected for small sample sizes was used to compare the prevalence of anti-CagA IgG, type B inflammation and non/low or high H. pylori LPS responders in patient groups. Probability values less than 0.05 were considered as significant.

3 Results

3.1 The proliferation of PBML from RUT/HLO-positive and -negative dyspeptic patients

All 33 children and young adolescents with chronic gastroduodenal disease produced IgG to antigens present in the glycine extract (GE) of H. pylori. The biopsy specimens obtained from 17 out of 33 patients were positive in rapid urease test and in histological examination for Helicobacter-like organisms. Sixteen patients were RUT/HLO-negative. H. pylori LPS alone induced the proliferation of PBML in 4 out of 33 patients (two RUT/HLO-positive and two RUT/HLO-negative). However, in the presence of IL-2, PBML from 10 out of 17 RUT/HLO-positive and from 12 out of 16 RUT/HLO-negative patients responded intensively to H. pylori LPS. These patients were designated as the high LPS responders (cpm/culture>1600). In the cultures containing IL-2, the cells from 7 and 4 RUT/HLO-positive and -negative patients, respectively, showed no or a very weak proliferative response to H. pylori LPS. These patients were designated as the low LPS responders (cpm/culture<1600).

Data in Table 1 demonstrated that the proliferative responses of PBML from RUT/HLO-positive patients to H. pylori LPS, IL-2 alone and to H. pylori LPS+IL-2, but not to PHA were significantly decreased as compared with the responses of PBML from RUT/HLO-negative patients. In the cultures with no stimuli, a spontaneous proliferation of PBML from RUT/HLO-positive patients was also significantly less intensive than the spontaneous proliferation of PBML from RUT/HLO-negative patients. In the sera from RUT/HLO-positive compared with RUT/HLO-negative patients, significantly higher levels of anti-EG IgG were detected (Table 1). Moreover, anti-CagA IgG were produced more frequently by RUT/HLO-positive (77%) than -negative (31%) patients. There were no differences in the levels of anti-EG or anti-CagA IgG between non/low and high LPS responders in the RUT/HLO-positive or -negative patients (Table 2). But in RUT/HLO-negative patients, IgG to CagA were detected in 5 out of 12 of high and not one out of 4 non/low LPS responders. The endoscopic examination revealed the type B inflammation in all RUT/HLO-positive and in 44% RUT/HLO-negative patients (Table 1).

View this table:
Table 1

The proliferation of PBML from RUT/HLO-positive and -negative dyspeptic patients

Dyspeptic patientsLPS (IL-2)IL-2 aloneLPS+IL-2PHANo stimuliAnti-GE IgG log titerAnti-CagA IgG prevalence log titerType B inflammation prevalence
cpm/culture
RUT/HLO-positive, n=171480±8323105±14975490±201489202±19029956±7704.0±0.513 (77%) 3.1±0.417 (100%)
RUT/HLO-negative, n=172461±17844397±14068382±2912100277±202591522±6333.1±0.55 (31%) 3.1±0.37 (44%)
P0.0480.020.0200.110.020.0005χ2=5.1 P=0.0240χ2=10.4 P=0.0012
  • The response to LPS in the presence of IL-2 [cpm (LPS+IL-2)-cpm (IL-2 alone)].

View this table:
Table 2

Anti-GE and CagA IgG in the sera from non/low and high responders to H. pylori LPS

Dyspeptic patientsRUT/HLO-positive IgG toRUT/HLO-negative IgG to
GECagAGECagA
Non/low LPS responders4.1±0.66/7 (86%) 3.1±0.33.0±0.60/4
High LPS responders4.0±0.57/10 (70%) 3.3±0.43.1±0.55/12 (41%) 3.1±0.3

3.2 The proliferation of PBML from dyspeptic patients with and without type B inflammation in gastric mucosa

In the presence of IL-2, PBML from dyspeptic patients without type B inflammation in gastric mucosa (all RUT/HLO-negative) showed enhanced proliferative response to H. pylori LPS and LPS+IL-2 but not to IL-2 alone or PHA (Table 3). In comparison, identically stimulated PBML from the patients with type B inflammation in gastric mucosa (RUT/HLO-positive and -negative) responded significantly less intensively to H. pylori LPS or LPS+IL-2, although their responses to IL-2 alone and PHA were at the same range. The levels of IgG to EG but not to CagA in the sera from the patients with type B inflammation in gastric mucosa were higher than in the sera from the patients without gastric inflammation. In the group of RUT/HLO-negative patients, significantly decreased prevalence of gastric mucosa inflammation in the LPS high responders compared with non/low responders was demonstrated (Table 4).

View this table:
Table 3

The proliferation of PBML from the patients with and without type B inflammation in gastric mucosa

Gastric mucosaLPS (IL-2)IL-2 aloneLPS+IL-2PHANo stimuliAnti-GE IgG log titerAnti-CagA IgG prevalence log titer
cpm/culture
Type B inflammation, n=171520±9773588±16656189±245794061±295991116±7463.7±0.714 (60%) 3.1±0.4
No inflammation, n=93119±18744113±13078766±313595934±199311534±7233.2±0.64 (44%) 3.1±0.3
P0.0030.4020.0180.8100.1500.07χ2=0.1 P=0.74
  • The response to LPS in the presence of IL-2 [cpm (LPS+IL-2)-cpm (IL-2 alone)].

View this table:
Table 4

Type B inflammation in gastric mucosa of non/low and high responders to H. pylori LPS in the presence of IL-2

Dyspeptic PatientsNon/low LPS respondersHigh LPS respondersχ2
Type B inflammation
RUT/HLO-positive, n=177/710/10
RUT/HLO-negative, n=164/43/124.15 P=0.040
Total, n=3311/1113/224.30 P=0.038

4 Discussion

This study showed that in the presence of IL-2, which is a T cell growth factor, H. pylori LPS may be an effective stimulus for human PBML, although H. pylori LPS alone, unlike LPSs from other Gram-negative bacteria [2], showed a very weak if any capacity to stimulate human PBML. The proliferative response to H. pylori LPS, in the medium with IL-2, was determined for PBML from children/young adolescents with chronic dyspepsia, seropositive for anti-H. pylori IgG. These patients could be divided in two groups: one RUT/HLO-positive with probably a dense H. pylori colonization, and another RUT/HLO-negative with less dense bacterial colonization which did not allow the detection of H. pylori bacteria and their urease in mucosa. In the presence of IL-2, 59% RUT/HLO-positive and 74% RUT/HLO-negative patients responded to H. pylori LPS. However, the intensity of this response was significantly increased in the cultures of PBML from RUT/HLO-negative as compared with -positive patients. Similarly, spontaneous (in culture medium alone) as well as IL-2 driven proliferation of PBML from RUT/HLO-positive patients were decreased as compared with the response of PBML from RUT/HLO-negative patients. The decreased proliferation of PBML from RUT/HLO-positive patients could be caused by anti-proliferative [1618] and/or promoting T cell apoptosis [19,20] activities of H. pylori bacteria. It could result in a lymphocyte dysfunction, decreased clearance of infection and dense bacterial colonization. A dysfunction of T lymphocytes from RUT/HLO-positive patients was confirmed by their reduced responsiveness to IL-2, which depends on an activating state of the cells. Our results do not allow us to know if proliferative response of PBML from dyspeptic seropositive patients to H. pylori LPS resulted from the endotoxic or antigenic quality of this molecule. However, this response was independent of the level of IgG to glycine extract of H. pylori and production of IgG to CagA (Tables 1 and 2). Also, we could see no difference between the proliferative response of PBML from RUT/HLO-positive and -negative patients to PHA, which is a very strong mitogen of T cells.

Infection with H. pylori causes active chronic type B gastritis, which is characterized by a dense mucosal infiltration by neutrophils, monocytes/macrophages, lymphocytes and some plasma cells [21,22]. The inflammation plays an important role in the pathogenesis of H. pylori infections and it is a risk factor for the development of duodenal and gastric ulcers and adenocarcinoma. H. pylori-induced gastric inflammation is thought to be mainly regulated by cytokines [23,24]. It has been demonstrated that H. pylori infections leading to chronic gastric diseases are associated with an overproduction of Th1 cytokines, IFN-γ and IL-2 as well as IL-12 which promotes the development of Th1 lymphocytes. This is why, in this study, we used IL-2 to co-stimulate PBML in the cultures with H. pylori LPS. We found that in the presence of IL-2, PBML from dyspeptic children without gastric inflammation responded significantly more intensively to H. pylori LPS than PBML from children with type B inflammation in gastric mucosa (P=0.003) (Table 3). Moreover, in the group of RUT/HLO-negative patients, the gastric inflammation was observed in 4 out of 4 dyspeptic children whose PBML showed a very weak if any response to H. pylori LPS and in only 3 out of 12 patients whose PBML responded intensively to this molecule (χ2=4.15, P=0.040; Table 4). The negative correlation between the type B inflammation and responsiveness of PBML to H. pylori LPS suggests that this molecule, by activation of immunocompetent cells, may reduce H. pylori-induced gastric inflammation, decrease bacterial load and prolong the infection. It is possible that H. pylori LPS may favor a shift from a Th1 to a Th2 cell pattern of cytokine secretion, which seems to be more beneficial to the host [23,24]. Recently it has been reported [25] that H. pylori LPS, like E. coli LPS, induced in human monocytes a large production of neutrophil-recruiting CXC chemokines (IL-8 and growth-related oncogene alpha). However, H. pylori LPS was less potent in inducing production of proinflammatory cytokines by human monocytes and was also a weak inducer of the CC chemokine RANTES. In this study, H. pylori LPS and IL-2 were used for the stimulation of unseparated fraction of blood mononuclear cells containing mainly T lymphocytes but also B lymphocytes, some monocytes and few dendritic cells. Thus, we cannot exclude a possibility that H. pylori driven proliferation of PBML resulted from the response of IL-2-activated T cells to the mediators released from H. pylori LPS-stimulated accessory cells, monocytes or activated B cells. In conclusion, our results confirm a contribution of H. pylori LPS to an outcome of H. pylori infection.

Acknowledgements

This work was supported by a grant of the Polish Research Committee (KBN, 4P05 B 059 14) and the University of Lodz.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
View Abstract