Abstract
Helicobacter pylori infection leads to chronic gastric inflammation. The current study determined the response of human APCs, NK cells, and T cells toward the bacteria in vitro. Human monocyte-derived dendritic cells (DC) were incubated with bacteria for 48 h. Intact H. pylori at a multitude of infection 5 stimulated the expression of MHC class II (4- to 7-fold), CD80, and CD86 B7 molecules (10- to 12-fold) and the CD83 costimulatory molecule (>30-fold) as well as IL-12 secretion (>50-fold) in DCs, and thereby, strongly induced their maturation and activation. CD56+/CD4− NK cells, as well as CD4+/CD45RA+ naive T cells, were isolated and incubated with DCs pulsed with intact bacteria or different cellular fractions. Coculture of H. pylori-pulsed DCs with NK cells strongly potentiated the secretion of TNF-α and IFN-γ. Coculture of naive T cells with H. pylori-pulsed DCs significantly enhanced TNF-α, IFN-γ, and IL-2 secretion as well as T-bet mRNA levels, while GATA-3 mRNA was lowered. However, the effect appeared attenuated compared with coculture with Escherichia coli. A greater stimulation was seen with naive T cells and DCs pulsed with H. pylori membrane preparations. Intact H. pylori potently induced the maturation and activation of human monocyte-derived DC and thereby promote NK and Th1 effector responses. The strong activation of NK cells may be important for the innate immune response. Th1-polarized T cells were induced especially by incubation with membrane preparations of H. pylori, suggesting that membrane proteins may account for the specific adaptive immune response.
Helicobacter pylori is a Gram-negative bacterium which colonizes the gastric mucosa and elicits a complex immune response, thereby initiating innate as well as adaptive immune responses (1). The dense infiltration of the mucosa with cells of the immune system suggests that a complex interplay between APCs and other immune cells may be important for the development of gastric pathologies; nevertheless, the chronic persistence suggests that the bacteria might interfere with crucial players of the immune system that would usually help to cure the infection. IFN-γ-producing Th1-polarized T cells and activated NK cells have been suggested to play an important role for the development of severe pathologies (2) or the elimination of the bacteria (3). Both cell types secrete IFN-γ; in vivo neutralization of IFN-γ results in a decrease of gastric inflammation in Helicobacter felis-infected, as well as immunized mice (4, 5). IFN-γ and Th1-deficient mice cannot be immunized against the infection, indicating that IFN-γ secretion is critical to eliminate the bacteria (6, 7). These observations underscore the importance of NK as well as Th1 cells for gastric inflammation and bacterial elimination. However, details about activation of these IFN-γ-producing cells and the related cellular mechanisms herein have remained obscure. Thus, a more detailed knowledge about the initiation of the host response is necessary to understand the mechanisms of how H. pylori induces a chronic activation of these cells in vitro.
The initial immune response toward bacteria is typically induced by APCs. Dendritic cells (DCs)3 as professional APCs discriminate between different bacteria, present bacterial Ags, and thereby induce innate and adaptive immune responses (8, 9). DCs do not act individually, because their function depends on environmental and tissue factors as well as on the microorganisms themselves (10). Recent studies have reported that DCs can integrate stimuli derived from microbial pathogens and other cells present at, or recruited to, the site of infection (10). These interactions can determine the success or failure of the response induced against pathogens. Indeed, follicular DCs have been detected in the H. pylori-inflamed gastric mucosa (11). H. pylori infection also up-regulated MIP-3-α gene expression in gastric epithelial cells and induced an influx of myeloid DCs in the lamina propria of the gastric mucosa in mice (12). However, the exact nature and function of these DCs has remained unclear because the cells are difficult to isolate and not abundant in the gastric mucosa.
Monocyte-derived DC (Mo-DCs) provide a suitable model in vitro for the determination of the antigenic potential of intact H. pylori or specific bacterial proteins. Maturation and Ag presentation in DCs, as monitored by FACS analysis of B7 molecules such as CD80 and CD86 and the costimulatory molecule CD83, is accompanied by cytokine secretion and can further lead to the stimulation of NK as well as T cells (9, 13, 14, 15). Activation and differentiation of the latter cells depends on 1) expression of peptide-MHC class II complexes on the surface of DCs that initiate a signal transduction cascade, 2) expression of B7 and costimulatory molecules that amplify the signaling process and stabilize the immunological synapse, and 3) the amount of the secreted cytokines such as IL-12, IFN-γ, and TNF-α (7, 16, 17). Although IL-12 can be regarded as a cytokine-inducing NK as well as Th1 activation, IL-10 secretion from DCs favors a Th2 polarization (9, 13, 14, 15). Thus, determination of the cytokine secretion from DCs in response to the bacteria might help to describe the nature of the subsequent response, but may not be unequivocal.
Activation of T cells and NK cells by H. pylori has not been investigated in vitro using preparations of these cells in combination with Mo-DCs. However, activation of Th1 cells during the adaptive immune response toward H. pylori infection has been identified in vivo and ex vivo using gastric biopsies or PBMC, respectively (18). H. pylori-specific T cells have been isolated from the mucosa of infected patients (19). Thus, incubation of Th1 cells with Ag-pulsed DC in vitro is of special interest to determine the actual components that are responsible for this immune response in detail.
IFN-γ-producing NK cells represent a special entity during the current immune reaction toward H. pylori and are key players of the innate immune system. Interaction between NK cells and DCs in bacterial infections results in rapid activation of NK cells (20, 21). Similar to Th1-polarized T cells, NK cells are activated by the release of IL-12 from DCs and secrete IFN-γ themselves (22, 23). Activation of NK cells with H. pylori in vitro has not been reported so far.
Therefore, we determined activation of these different cell types during the interaction with H. pylori. We found that Mo-DCs and NK cells become activated during bacterial contact in vitro and may thus be crucial players for the process of H. pylori recognition and elimination (18, 24, 25, 26). Th1-polarized T cells, key players of the adaptive immune response, are more sensitive to stimulation with Mo-DC pulsed with H. pylori membrane proteins, indicating an important role of H. pylori membrane proteins for induction of the specific immune response.
Materials and Methods
Bacteria
H. pylori wild-type strain, G27 (G27Wt), is a laboratory strain which carries an intact cag pathogenicity island and produces the vacuolating cytotoxin A. The bacteria were grown on Wilkins-Chalgren blood agar plates under microaerobic conditions (10% CO2, 5% O2, 85% N2; 37°C) for 3–4 days. Escherichia coli (strain XL1-blue) were grown on Luria-broth agar plates at 37°C overnight. Cells were harvested, washed once with PBS, pH 7.5, and subsequently added to the Mo-DCs at multitude of infection (MOI) 5 (5 bacteria per DC). MOI was determined by growing H. pylori to an OD600 = 1.0. Subsequently, serial dilutions of the cultures were plated on Wilkins-Chalgren agar. Colonies were counted after 4–5 days of culture and the number of viable cells per milliliter was calculated. We determined ∼2 × 108 H. pylori cells per milliliter at OD600 = 1.0.
Preparation of bacterial fractions
H. pylori G27Wt was grown in brain heart infusion broth supplemented with 10% horse serum under microaerobic conditions (10% CO2, 5% O2, 85% N2; 37°C) to an OD600 = 1.5. Cells were harvested by centrifugation, washed one time with PBS, and resuspended in 25 mM Tris/HCl, pH 7.5; 1 mM DTT; 10 mM MgCl2; 1 mM PMSF (reducing the volume ∼250-fold). Bacteria were lysed by several freeze and thaw cycles (liquid nitrogen, water bath at 37°C with a following sonication step). Cells underwent centrifugation at 10,000 × g for 20 min at 4°C to remove nondisrupted cells and debris. The supernatant was diluted five times in 25 mM Tris/HCl, pH 7.5; 1 mM DTT; 10 mM MgCl2; 1 mM PMSF. Centrifugation at 100,000 × g resulted in the membrane fraction (pellet) and the cytosolic fraction (supernatant). The membranes were washed one time in 1 mM Tris/HCl, pH 7.5; 5 mM EDTA, pH 8.0; 1 mM PMSF. The protein content of the fractions was determined by the method according to Bradford with BSA as standard. All fractions were aliquoted and stored at −20°C.
Flow cytometry
Flow cytometry analysis was performed using the FACSCalibur and the CellQuest software package (BD Biosciences, Heidelberg, Germany). Cells (105) were incubated in 100 μl of FACS buffer (PBS, 1% (w/v) BSA), and 10 μl of each of the appropriate FITC and PE-labeled Ab for 30 min on ice in the dark. After incubation, the cells were washed with FACS buffer and resuspended in 500 μl of FACS buffer before automated analysis with the FACSCalibur. The purity of monocytes, immature Mo-DCs, mature Mo-DCs, naive T and NK cells was determined. Cells were stained with the following anti-human Abs: monocytes with anti-CD45 and anti-CD14; immature Mo-DCs with anti-CD1a, anti-CD14, anti-HLA-DR, anti-CD80, anti-CD83, and anti-CD86; mature Mo-DCs with anti-HLA-DR, anti-CD80, anti-CD83, anti-CD86, and anti-CD1a; naive T cells with anti-CD45RA and anti-CD4; and NK cells with anti-CD56 and anti-CD4. All Abs used for FACS analysis were purchased from BD Biosciences.
Cell isolation
PBMC were isolated from heparinized venous blood of healthy adult volunteers not infected with H. pylori by Ficoll-Hypaque (Pharmacia Biotech, Uppsala, Sweden) density gradient centrifugation as recommended by the manufacturer. DCs, NK cells, and T cells for one experimental set up were isolated from the same donor to exclude HLA interactions.
Generation and infection of Mo-DCs
Monocytes were isolated from PBMC by MACS with the Monocyte Isolation Kit II (Miltenyi Biotec, Bergisch-Gladbach, Germany) and their purity was determined by FACS by staining the cells with anti-CD14 and anti-CD45. Mo-DCs were generated by culturing monocytes in X-VIVO 15 medium (Cambrex, Vervier, Belgium), 1% heat-inactivated (hi) FCS, 1% l-glutamine, 1% streptomycin, 1% penicillin, 1000 U/ml human rIL-4 (Strathmann Biotech, Hannover, Germany), and 1600 U/ml human rGM-CSF (Leucomax; AESCA, Traiskirchen, Austria) for 6 days in 6-well plates at 5 × 106 cells per well. At day 7, cells were harvested and analyzed by FACS for CD1a expression. DCs derived from the culture of monocytes in the presence of GM-CSF and IL-4, but unexposed to the stimulus, were considered “differentiated”. DCs induced to up- or down-regulate surface markers were termed “mature”, and DCs that secrete cytokines were named “activated” DC.
A total of 5 × 105-differentiated Mo-DCs were incubated with H. pylori at MOI 5 and with E. coli at MOI 5 as positive control in 24-well plates in 1 ml of medium (X-VIVO 15, 1% hi FCS, 1% streptomycin, 1% penicillin, and 1% l-glutamine) for 48 h. The DCs were analyzed for CD80, CD83, CD86, CD1a, and MHC class II expression by flow cytometry to determine their maturation. The secretion of IL-12 and IL-10 in the supernatant was determined by ELISA. Differentiated DCs were incubated with cytochalasin (Ct)A or CtD (which are known to block internalization of particles) at 5 μg/ml for 1.5 h at 37°C before the pulse with H. pylori. The medium was then removed by centrifugation and fresh medium was added for further incubation with H. pylori for 48 h. After 48 h, IL-12 and IL-8 secretion in the supernatant was determined by ELISA.
NK cell isolation and stimulation
NK cells were isolated from PBMC with the NK Cell Isolation Kit II (Miltenyi Biotec) by negative selection and the purity was determined by FACS analysis using FITC-conjugated anti-CD56 and PE-conjugated anti-CD4. CD56+/CD4− cells were used as NK cells, added to the H. pylori-pulsed DCs (2–4 NK cells per DC) and incubated in 24-well plates in 1.0 ml of medium (X-VIVO 15, 1% hi FCS, 1% streptomycin, 1% penicillin, and 1% l-glutamine). After 24 h of coculture, secretion of the cytokines IFN-γ and TNF-α in the supernatant was determined by ELISA. As a positive control, NK cells were added to E. coli-pulsed DCs (2–4 NK cells per DC) and incubated under the same conditions as with H. pylori-pulsed DCs. Cytotoxicity of NK cells was determined with the nonradioactive cytotoxicity assay CytoTox 96 (Promega, Mannheim, Germany) using K-562 cells (ATCC no. CCL-243; American Type Culture Collection, Manassas, VA) cultured in RPMI 1640/10% FCS as target cells. The assay is a colorimetric alternative to 51Cr release assays and quantitatively measures the release of the stabile cytoplasmic enzyme lactate dehydrogenase in a coupled enzymatic assay which results in the conversion of a tetrazolium salt into a red formazan product. The assay was performed according to the manufacturers protocol with 2.5 × 103 K-562 cells per well as target (determined by an assay to assess the optimized target cell number) with the indicated E:T ratios.
Naive T cell isolation and stimulation
Naive T cells were isolated in a two step procedure as follows: first, CD3+ T cells were isolated from PBMC by negative selection with the Pan T Cell Isolation Kit II (Miltenyi Biotec); second, naive T cells were obtained from the CD3+ T cell fraction with CD45RA MicroBeads (Miltenyi Biotec) by positive selection. The purity of the naive T cells was evaluated by flow cytometry by staining the cells with anti-human CD4-FITC and anti-human CD45RA-PE. CD4+/CD45RA+ T cells were used as naive T cells. Naive T cells were added to H. pylori-pulsed DCs and to E. coli-pulsed DCs (10 T cells per DC) and incubated for 72 h in 1.5 ml of medium (X-VIVO-15, 1% hi FCS). Subsequently, the secretion of IL-4, TNF-α, IL-2, and IFN-γ was determined by ELISA in the culture supernatant.
Quantification of cytokines by immunoassay
Cytokine secretion from the different experiments was determined in the culture supernatant by commercially available specific ELISA according to standard procedures. In the assays the lower limits of detection were 7.8 pg/ml for IL-10, IL-12 (p40), IFN-γ, IL-4, and TNF-α (Biosource, Nivelles, Belgium) and 6.5 pg/ml for IL-10 and 3.9 pg/ml for IL-2 (Biocarta, Hamburg, Germany).
Quantitative TaqMan real-time PCR analysis
Quantitative real-time PCR was performed as published previously (27, 28). Total RNA from DC/T cell cocultures was extracted by the phenol chloroform method (29). The obtained RNA underwent an additional purification step by lithium chloride precipitation. Therefore, RNA was eluted in 100 μl of 4 M LiCl solution and incubated for 3 h at 4°C. After subsequent centrifugation at 10,000 × g for 30 min, the supernatants were discarded; the pellet was washed with 70% ethanol and resuspended in 50 μl of RNase-free water. RNA was reverse transcribed using Multiscribe reverse transcriptase (PerkinElmer, Weiterstadt, Germany) according to the manufacturer’s instructions. TaqMan primers (MWG Biotec, Ebersberg, Germany) and probes (PerkinElmer) were designed to span exon junctions or to lie in different exons to prevent amplification of genomic DNA. Primer and probe sequences were: T-bet forward primer, 5′-TCA GCA CCA GAC AGA GAT GAT CA-3′; T-bet reverse primer, 5′-GCC ACA GTA AAT GAC AGG AAT GG-3′; T-bet probe, 5′-CCA AGC AGG GAC GGC GGA TGT-3′. GATA-3 forward primer, 5′-TCT ATC ACA AAA TGA ACG GAC AGA A-3′; GATA-3 reverse primer, 5′-GCT CTC CTG GCT GCA GAC A-3′; GATA-3 probe, 5′-CGG CCC CTC ATT AAG CCC AAG C-3′. GAPDH forward primer, 5′-ACG GAT TTG GTC GTA TTG GGC-3′; GAPDH reverse primer, 5′-TTG ACG GTG CCA TGG AAT TTG-3′; GAPDH probe, 5′-CCT GGT CAC CAG GGC TGC TTT TAA-3′. PCR was performed for 40 cycles with the following conditions: 95°C for 15 s; 60°C for 20 s; and 72°C for 30 s. To determine absolute mRNA copy numbers, standard curves were generated using plasmid dilution series containing the corresponding target sequence. T-bet and GATA-3 copy numbers were normalized to GAPDH copies.
Statistical analysis
Data were analyzed by one-way ANOVA followed by Newmann-Keuls testing for dependent variables. Values of p ≤ 0.05 were considered to be significant.
Results
Generation of differentiated monocyte-derived DCs
Monocytes were isolated from PBMC by MACS. The purity of the cell population was determined by FACS analysis of CD14 and CD45 expression on the cell surface. More than 90% of the isolated cells were CD14+ and CD45+ (not shown). Culture of these monocytes in the presence of IL-4 and GM-CSF for 6 days resulted in the generation of differentiated Mo-DCs and was monitored by FACS analysis of the expression of MHC class II, CD1a, CD14, costimulatory molecule CD83 and B7-1 (CD80) and B7-2 (CD86) molecules on the cells (Fig. 1⇓). CD1a expression is restricted to differentiated Mo-DCs and gating the cells on CD1a+ and HLA-DR+ revealed 91.64% double-positive cells (Fig. 1⇓A). The expression of CD14 and CD80 is restricted to monocytes and mature Mo-DCs, respectively. As shown in Fig. 1⇓B, when cells were gated on CD14+ (upper left (UL), 1.54%), CD80+ (lower right (LR), 2.2%), and CD14/CD80 double-positive cells (upper right (UR), 0.22%). Similarly, there were 0.59% CD4+ cells (LR), representing T lymphocytes, 0.11% CD86+ cells (UL) and 0.19% CD4/CD86 double-positive cells (UR) in the preparation (Fig. 1⇓C). CD80 (B7-1) and CD86 (B7-2) are two major markers for mature DCs in addition to the MHC class II and are important to establish the immunological synapse with T cells. CD83 costimulatory molecule was expressed during the maturation process of DCs as shown in Fig. 1⇓D by gating the cells on CD4+ (UL, 1.06%), CD83+ (LR, 5.8%), and CD4/CD83 double-positive cells (UR, 0.14%). To rule out a possible contribution of contaminating CD4+ T cells in the DC preparation, CD4+/CD45RA+ naive T cells were isolated by magnetic cell separation and incubated with H. pylori. Basal release of IL-12 or IL-10 from CD4+/CD45RA+ cells ranged from 1.98 to 4.2 pg/ml, but no increment in IL-12 or IL-10 secretion was observed (n = 3 experiments, data not shown).
Flow cytometry of the Ag profile of differentiated Mo-DCs. Monocytes were generated from PBMC by MACS; >90% of the isolated cells were CD14+ and CD45+ (not shown). Mo-DCs were generated in vitro by culturing purified CD14+ monocytes in the presence of GM-CSF and IL-4 for 6 days. Differentiated DCs were characterized by expression of HLA-DR and high surface expression of CD1a (A), minimal expression of CD80 (B) and CD86 (C), and almost no expression of CD83 costimulatory molecules (D).
H. pylori induced the maturation and activation of Mo-DCs
Differentiated Mo-DCs were cultured with H. pylori and E. coli as positive control at MOI 5 for 48 h and the maturation process was subsequently determined by FACS analysis. The immunophenotypic characteristics of mature DCs can be shown by the high expression of MHC class II, CD80, CD83, and CD86 on the cell surface as shown in Fig. 2⇓A. H. pylori (▪) stimulated the expression of CD80 and CD86 10- to 12-fold, CD83 expression increased >30-fold, and MHC class II expression 4- to 7-fold in comparison to basal conditions (medium alone, □). In accordance, there was no longer expression of CD1a, the major marker of differentiated DCs compared with 57% expression under basal conditions. To determine the optimal time point of DC activation, secretion of IL-12 was determined after 24, 48, and 72 h by ELISA. Forty-eight hours were chosen for further analysis of H. pylori pulsed-DCs because this time point showed the highest IL-12 secretion that decreased again at 72 h (Fig. 2⇓B).
Intact H. pylori induced the maturation and activation of Mo-DCs. Differentiated Mo-DCs were incubated with H. pylori at MOI 5 or, as negative control, in medium alone. A, After 48 h, Mo-DCs were stained with Abs and analyzed by FACS. Compared with controls (□), H. pylori (▪) induced an increase of the expression of the costimulatory molecule CD83, B7 molecules (CD80 and CD86) and MHC class II indicating maturation of the cells. In contrast, expression of CD1a, a marker of differentiated DCs, was decreased. Data are shown as mean of n = 5 ± SEM. B, To determine the optimal time point of cytokine secretion, supernatants were obtained after 24, 48, and 72 h and the IL-12 secretion was determined by ELISA. For additional experiments, 48-h incubation was chosen because it gave the highest secretion of IL-12. Data are shown as mean of n = 3 ± SEM. C, IL-10 and IL-12 secretion of the cell preparations was determined by ELISA after a 48-h incubation. H. pylori induced a >50-fold increase in IL-12 (p40) secretion (▪), while IL-10 secretion (□) was only marginally stimulated. Incubation with E. coli was used as a control and yielded similar results. Data are shown as mean of n = 5 ± SEM. The differentiated DCs were preincubated for 1.5 h with 5 μg/ml CtA or CtD before incubation with H. pylori for another 48 h. D, IL-12 secretion was completely blocked by CtA and CtD respectively. E, CtA and CtD have no effect on IL-8 secretion. Data are shown as mean of n = 3 ± SEM.
As shown in Fig. 2⇑C, the activation of DCs was accompanied with a >50-fold induction of IL-12 secretion (▪) compared with DCs cultured in medium alone. The absolute IL-10 secretion (□) under these conditions was low compared with IL-12 and only a slight increase of IL-10 secretion from H. pylori-pulsed DCs was detected. MHC-class II up-regulation and IL-12 secretion after incubation with H. pylori was also tested in the presence of polymyxin B. Polymyxin B is known to deorganize the cell wall of Gram-negative bacteria and, therefore, can be used to prevent LPS activity potentially present in the preparation. In the current experiments, addition of 1 μg/ml polymyxin B sulfate (Applichem, Darmstadt, Germany) did not significantly alter the H. pylori-induced response (n = 3, data not shown), ruling out the possibility that H. pylori LPS is responsible for the effects observed in DCs.
Fig. 2⇑, D and E, shows the effects of CtA and CtD added over 1.5 h to the DCs before incubation with H. pylori. CtA and CtD both blocked IL-12 production significantly (Fig. 2⇑D) but did not have inhibitory effects on IL-8 secretion (Fig. 2⇑E). CtA and CtD are known to inhibit actin polymerization, thus preventing uptake of bacteria.
H. pylori promoted the activation of NK cells
NK cells were isolated from PBMC using MACS and the purity of NK cells was determined by staining of these cells with FITC-conjugated anti-CD56 and PE-conjugated anti-CD4. As shown in Fig. 3⇓A the purity of CD56+/CD4− cells was ≥82%. Isolated NK cells were cocultured with H. pylori-pulsed DCs for 24 h as described before. The incubation resulted in activation of NK cells that was indirectly determined by the measurement of cytokines known to be produced by active NK cells, such as TNF-α and IFN-γ, in the coculture supernatant. As shown in Fig. 3⇓, B and C, the secretion of TNF-α and IFN-γ was significantly induced, indicating a strong activation and stimulation of NK cells. Addition of H. pylori to NK cells alone lead to a small increment in cytokine secretion. However, incubation of NK cells with H. pylori-pulsed DC significantly enhanced IFN-γ and TNF-α secretion. The secretion of TNF-α was increased >30-fold in the coculture of NK cells with H. pylori-pulsed DCs in comparison to the coculture with nonpulsed DCs (Fig. 3⇓B). The secretion of IFN-γ was increased 8-fold under the same conditions (Fig. 3⇓C).
H. pylori-pulsed DCs promoted NK cell activation. Isolation of NK cells from PBMC was performed by MACS. A, The purity of NK cells was determined by staining the cells with anti-CD56-FITC and anti-CD4-PE and was shown to be >82%. NK cells were cocultured with DCs pulsed with intact H. pylori, intact E. coli or vehicle for 24 h. Activation of NK cells was determined by measurement of TNF-α secretion (B) or IFN-γ secretion (C) in the supernatant, yielding a significant stimulation only in preparations with H. pylori-pulsed DCs. Data are shown as mean of n = 5 ± SEM.
Induction of cytotoxicity in NK cells in response to incubation with H. pylori was also investigated. As shown in Table I⇓, addition of IL-2, as positive control, strongly induced cytotoxicity of NK cells. Incubation of NK cells with H. pylori for 48 h did not stimulate cytotoxicity and there was also no cytotoxicity when NK cells were cocultured with H. pylori-pulsed DCs.
Cytotoxicity of NK cells after incubation with H. pyloria
H. pylori induced the differentiation of naive T lymphocytes into Th1 cells
Naive T cells were isolated from PBMC by magnetic cell separation and the purity of the cells was evaluated by FACS analysis by staining the cells with FITC-conjugated anti-CD45RA and PE-conjugated anti-CD4. More than 80% of the isolated cells were CD4+/CD45RA+ (Fig. 4⇓A). Naive T cells were added to the H. pylori-pulsed DCs at a ratio of 10 T cells per DC. Addition of H. pylori to T cells alone lead to a small increment in cytokine secretion. However, incubation of T cells with H. pylori-pulsed DCs significantly enhanced IFN-γ and TNF-α secretion. After 72 h of coculture, the secretion of the cytokines TNF-α (Fig. 4⇓B) and IFN-γ (Fig. 4⇓C) was determined in the supernatant by ELISA. Secretion of TNF-α and IFN-γ was stimulated compared with the coculture of nonpulsed DCs with naive T cells. IFN-γ secretion was even more stimulated after incubation with membrane preparations from H. pylori, while cytosolic preparations had no effect (Fig. 4⇓D).
Activation of naive T cells with intact H. pylori, membrane and cytosolic preparations. Naive T cells were isolated from PBMC by MACS. A, The purity of the naive T cell preparation was evaluated by FACS analysis using FITC-conjugated anti-CD45RA and PE-conjugated anti-CD4. More than 80% of the isolated cells were CD4+ and CD45RA+. B and C, Subsequently, naive T cells were added to DCs pulsed with intact H. pylori, intact E. coli or vehicle for 72 h at a ratio of 10 T cells per DC. Activation of T cells was determined by the measurement of TNF-α secretion (B) and IFN-γ secretion (C) in the supernatant, showing a significant stimulation of TNF-α or IFN-γ with H. pylori compared with stimulation with vehicle only. D, Naive T cells were added to DCs pulsed with vehicle, intact H. pylori, membrane preparations (MP) or cytosolic preparations (Cyt) from H. pylori. Membrane preparations induced a significant increase in IFN-γ secretion compared with intact bacteria (p < 0.05) or incubations with vehicle (p < 0.001). Data are shown as mean of n = 3 ± SEM.
Fig. 5⇓ presents the effect of various incubation time periods (24, 48, and 72 h) on cytokine production after incubation of H. pylori pulsed-DC with naive T cells. The 72-h time point was determined to yield a maximal stimulation. A strong activation was determined by measuring increased secretion of IFN-γ, TNF-α, and IL-2 and CD69 expression T cell surface following incubation of T cells with H. pylori pulsed DC (Fig. 5⇓).
Time course of activation T cell. Naive T cells were isolated from PBMC by MACS and were cocultured with Mo-DCs pulsed with H. pylori or vehicle for 24, 48, and 72 h at a ratio of 10 T cells per DC. Activation of T cells was determined by the measurement IFN-γ secretion. A, TNF-α secretion (B) and IL-2 production (C) in the supernatant, showing a significant stimulation of all cytokines in the coculture with H. pylori-pulsed DCs compared with controls (vehicle only) with maximum secretion at 72 h. After 72 h, T cells were stained with anti-CD69 and analyzed with FACS. Compared with controls (□), H. pylori (▪) induced an increase of the expression of CD69 on T cell surface indicating activation of these cells (D).
TaqMan PCR analysis of T-bet and GATA-3 expression in T cells
mRNA expression of transcription factors regulating Th1 and Th2 cell differentiation, T-bet, and GATA-3, respectively, was determined quantitatively by TaqMan PCR. As shown in Fig. 6⇓A, expression of T-bet was up-regulated in T cells cocultured with H. pylori-pulsed DCs, compared with T cells which were cocultured with nonpulsed DCs. In contrast, mRNA levels of GATA-3 were lower in T cells cocultured with H. pylori-pulsed DCs than in the nonpulsed DC/T cell coculture. Thus, the stimulation of naive T cells with H. pylori-pulsed DCs resulted in differentiation of naive T cells into a Th1 subpopulation. T-bet expression (Fig. 6⇓B, right panel) was more pronounced in T cells stimulated with membrane preparations of H. pylori while there was no effect with cytosolic preparations. The GATA-3 expression (Fig. 6⇓B, left panel) was not affected by using membrane or cytosolic preparations compared with intact H. pylori.
TaqMan PCR analysis of quantitative mRNA expression for transcription factors T-bet and GATA-3. Mo-DC were pulsed with intact H. pylori and E. coli as control (A) or cellular fractions (membrane preparations and cytosolic preparations) of H. pylori (B) and subsequently cocultured with naive CD4+/CD45+ T cells. mRNA expression of T-bet (marker for Th1 differentiation) and GATA-3 (Th2 marker) was determined by TaqMan PCR. Data are shown as mean mRNA levels of n = 3 ± SEM.
Discussion
Limited information is available on the direct effect of intact H. pylori on DCs, and in particular of their subsequent effect on NK and T cells, which represent major players of the human immune system. The latter information is rather important because it may shed light on cellular interactions occurring during innate as well as adaptive immune responses against H. pylori in the course of infection. Our work was focused on three key cell types that might play a role during the response toward H. pylori: DCs, NK, and T cells.
First, our present data clearly demonstrate exposure toward the pathogen H. pylori rapidly induced maturation and activation of human Mo-DCs. Cytokine secretion was accompanied by increased expression of functionally important MHC class II surface molecules, as well as B7 (CD80, CD86) and CD83 costimulatory receptors (21). Expression of MHC class II, B7 and CD83 molecules has specifically been reported during contact of Mo-DCs with live E. coli or Mycobacterium bovis bacillus Calmette-Guérin (21, 20). These studies also revealed that B7-1 (CD80) and B7-2 (CD86) molecules seem to be of special importance for the subsequent immune response because increased expression amplifies the signaling process at the immunological synapse established between dendritic cells and other cells of the immune system (21). Cellular synapses occurring between the two cell types are likely to be crucial by greatly increasing the local concentration of DC-released soluble mediators (20).
The current data reveal a possible pathway how H. pylori is recognized and processed in the gastric mucosa. Indeed, increased numbers of activated DCs and activated T cells have been detected in the lamina propria of H. pylori-infected individuals (30) (12). It should be noted that the activated DCs were localized in the bottom portion of the lamina propria. Thus, the activated DCs might migrate to the regional lymph nodes where further interactions with immune cells can occur (12).
In accordance with the model of an immunological synapse between DCs and other immune cells, H. pylori infection led to a remarkable up-regulation of IL-12 secretion in DC preparations, while IL-10 secretion was very low. Inhibitory effects of CtA and D on IL-12 secretion suggest that internalization of the bacteria or bacterial proteins is important for activation of DCs. These cytokines are important for the subsequent activation of innate as well as adaptive responses. Recent studies reported a 5- to 10-fold activation of human Mo-DCs during H. pylori infection in vitro (31) which is in accordance to our current work, although the response observed here is significantly higher (>50-fold). Our data further suggest that H. pylori LPS does not account for the activation of Mo-DCs, because polymyxin B was ineffective in preventing this response. In line with our results, H. pylori LPS has been shown to have a 1000-fold lower biological activity than enterobacterial LPS (32).
Regarding a second key player of the immune system, we found that incubation of H. pylori-pulsed Mo-DCs with NK cells lead to the activation of cytokine secretion from NK cells. IL-12 secretion from stimulated DCs is known to promote NK cell activation (33). Increased levels of IL-12, IFN-γ, and TNF-α have indeed been detected in the H. pylori-infected mucosa in humans (27, 34, 35, 36, 37, 38, 39), which is in accordance to our in vitro findings. DCs that have encountered H. pylori in the gastric epithelium may thereby initiate important effector arms of the innate immunity by secreting IL-12. Activation of NK cells in vitro was determined by measurement of TNF-α and IFN-γ. The high amounts of TNF-α and IFN-γ secreted by activated NK cells suggests that NK cell-derived cytokines play a major role in the homeostasis of the immune response during H. pylori infection. In this context, TNF-α and IFN-γ have been shown to promote activation of macrophages and differentiation of T cells. Overall, our studies report first and important information that NK cells are highly activated during H. pylori infection.
A further question related to our present data was whether bacterial infection could lead to a rapid induction of NK cell cytotoxicity. NK cells have been shown to have cytotoxic effects against tumor cells (40). It has already been documented that contact of lymphocytes with H. pylori augments NK cell activity and induces production of IFN-γ in vivo, suggesting the activation of cytotoxic NK cells; H. pylori culture supernatants contain a factor which can activate resting large granular lymphocytes to exhibit cytolytic activity (41). In our current experiments, we did not observe the activation of cytotoxic NK cells after exposure to H. pylori-pulsed DCs. This observation is in line with previous publications suggesting that two different subsets of NK cells exist: The more cytotoxic subset of NK cells expresses higher levels of Ig-like NK receptors, while the other subset has the capacity to produce abundant cytokines following activation of monocytes (42). Furthermore, other works report that the cytotoxic activity of NK cells does not seem to exert any direct effect against bacteria at the surface of the epithelium and may therefore be of less importance (43). Therefore, we propose a mechanism in which cytokine secretion from activated NK cells play a major role in the homeostasis of the immune response during bacterial infection, while cytotoxic effects against intact H. pylori seem to play a less important role during this process.
The third and probably most important key player of the immune system investigated here are T cells. As expected, coculture of naive T cells with H. pylori-pulsed DCs enhances expression of the transcription factor T-bet which directs Th1 lineage commitment (44, 45). T-bet is a Th1-specific T box transcription factor that controls the expression of the hallmark Th1 cytokine, IFN-γ. T-bet expression correlates with IFN-γ expression in Th1 (44). Accordingly, IFN-γ secretion in the coculture was up-regulated, whereas no IL-4 protein was detected. These data suggest a Th1 differentiation. Interestingly, this effect was mainly observed when membrane preparations of H. pylori were used. Gastric T cells from H. pylori-infected animals and humans produce predominantly IFN-γ, but not IL-4, typical for a Th1-polarized T cell response in vivo (2, 4, 5, 6), which is in accordance to our results. Similar to the activation of NK cells, previous studies have revealed that IL-12 secretion from DCs seems to play a key role for polarization of naive T cells into Th1 cells mediating adaptive immune responses (16, 46).
A preparation of CD4+/CD45RA+ naive T cells (>80%) was used to investigate the influence of H. pylori on the activation and differentiation of CD4+ naive T cells because previous experiments have determined that the differentiation into Th1 cells is of special importance for therapeutic vaccination against the bacteria in mice (5). Increased density of CD4+ and also CD8+ cells have been detected in the gastric mucosa of H. pylori-inflamed tissue; however, the importance of CD8+ cells remains speculative during the infection (47).
IFN-γ secretion in the H. pylori-pulsed DC/T cell coculture was lower than in the E. coli-pulsed DC/T cell coculture. Moreover, compared with the incubation of DCs with NK cells, the total amount of IFN-γ secreted from T cells was much lower. Previous studies reported that the IFN-γ promoter is hypermethylated in naive T cells and may thus be not effectively stimulated by intact bacteria (48, 49). It also has to be mentioned that T cells react predominantly to known Ags, and the total number of T cells carrying suitable T cell receptors may be only a small percentage of the total number of T cells used, and thus, the amount of IFN-γ released via activation of T cell receptors is lower.
Most interestingly, the Th1 effector responses seemed to be more pronounced when membrane preparations of H. pylori were used. It appears that fractionation of the bacteria into subcellular fractions might facilitate phagocytosis of the antigenic proteins. In contrast, fractionation might yield a higher concentration of proteins because the other, nonimmunogenic fractions, have been removed. These results suggest that the true nature of the Th1 response appears to be driven by membrane proteins exposed on the surface and underline the importance of these proteins for induction of a Th1 response. It may be speculated that these proteins are captured by Mo-DCs and some of them become exposed on MHC class II complexes, leading to a Th1 differentiation; this process appears facilitated when using membrane preparations in the current in vitro setup. This speculation would be in accordance with previous observations made with Gram-negative bacteria using Mo-DCs and T cells (50). A further possibility could be that membrane proteins also react with TLRs, thus stimulating intracellular signaling cascades and initiating cytokine secretion. Indeed, H. pylori-induced NF-κB activation was decreased in MKN45 gastric epithelial cells by transfection of dominant-negative versions of TLR2 and TLR5, but not TLR4 (51). These studies suggest that gastric epithelial cells recognize and respond to H. pylori infection at least in part via TLR2 and TLR5. Nevertheless, epithelial cells may respond differentially to incubation with H. pylori than dendritic cells and may not be the key players in initiating the immune response.
In summary, mature DCs, activated NK cells, as well as the Th1 subsets, seem to be crucial players for the process of H. pylori recognition and presentation. Our current work has triggered an intensive search for bacterial epitopes inducing a particular Th1-effector response that may be of great importance for a successful elimination of the bacteria (2, 18, 24, 25, 26).
Acknowledgments
We thank Dr. Martin Mempel (Department of Dermatology, Technical University of Munich, Munich, Germany) for helpful instructions and discussions.
Footnotes
-
↵1 This work was sponsored by Else Kröner-Fresenius Stiftung Germany, Graduiertenkolleg 333, and Deutsche Forschungsgesellschaft. C.P. is a recipient of the Heisenberg programe (DFG 411-7/1).
-
↵2 Address correspondence and reprint requests to Dr. Christian Prinz, Department of Medicine II, Technical University Ismaningerstrasse, 22 D-81675 Munich, Germany. E-mail address: christian.prinz{at}lrz.tum.de
-
↵3 Abbreviations used in this paper: DC, dendritic cell; Mo-DCs, monocyte-derived DC; MHC-II, major histocompatibility complex class II; MOI, multitude of infection; UR, upper right; UL, upper left; LR, lower right; hi, heat-inactivated, Ct, cytochalasin.
- Received January 12, 2004.
- Accepted May 13, 2004.
- Copyright © 2004 by The American Association of Immunologists
References
In this issue
