Helicobacter pylori Induces Activation of Human Peripheral γδ+ T Lymphocytes

Helicobacter pylori is a Gram-negative bacterium that causes gastric and duodenal diseases in humans. Despite a robust antibody and cellular immune response, H. pylori infection persists chronically. To understand if and how H. pylori could modulate T cell activation, in the present study we investigated in vitro the interaction between H. pylori and human T lymphocytes freshly isolated from peripheral blood of H. pylori-negative donors. A direct interaction of live, but not killed bacteria with purified CD3+ T lymphocytes was observed by microscopy and confirmed by flow cytometry. Live H. pylori activated CD3+ T lymphocytes and predominantly γδ+ T cells bearing the TCR chain Vδ2. Upon interaction with H. pylori, these cells up-regulated the activation molecule CD69 and produced cytokines (such as TNFα, IFNγ) and chemokines (such as MIP-1β, RANTES) in a non-antigen-specific manner. This activation required viable H. pylori and was not exhibited by other Gram-negative bacteria. The cytotoxin-associated antigen-A (CagA), was at least partially responsible of this activation. Our results suggest that H. pylori can directly interact with T cells and modulate the response of γδ+ T cells, thereby favouring an inflammatory environment which can contribute to the chronic persistence of the bacteria and eventually to the gastric pathology.


Introduction
Helicobacter pylori (H. pylori) is a spiral shaped Gram-negative bacterium that causes gastric and duodenal disorders. The H. pylori infection is typically acquired in early childhood via person-toperson spread, via oral-oral or fecal-oral transmission. The majority of infected individuals remain asymptomatic, and only a 5-15% develops serious complications. Chronic infection with Helicobacter pylori is the major known risk factor for duodenal and gastric ulcer diseases and cancer [1,2], which are frequently associated with the expression of CagA antigen [3,4,5].
Helicobacter pylori infection induces a strong local immune response with infiltration of the mucosa by neutrophils, macrophages and lymphocytes. Many studies reported that the T cell response to H. pylori is prevalently of Th1 type with infiltration of IFN-c producing T cells in the site of infection [6]. In addition, unconventional T cell populations may also intervene at the mucosal level in response to H. pylori stimuli and modulate the outcome of the infection, leading to local inflammation, chronic persistence of lesions and eventually cancer [1]. Some studies have described the involvement of cd+ T cells in Helicobacter pylori gastritis [7,8,9]. In particular, one study has reported the infiltration of cd+ T cells in H. pylori infected biopsies that were significantly higher in grade III gastritis while strongly decreased after eradication therapy [10]. Moreover cd+ T cells appear to have both pro-inflammatory and regulatory functions: they can act as a bridge between innate and adaptive immunity early in the response and down-modulate inflammation once the infection is cleared [7].
In the present study we investigated the interaction of H. pylori with human T cell populations, including cd+ T cells and how this interaction modulated their state of activation and ability to produce cytokines.

H. pylori directly interacts with T lymphocytes
To investigate whether H. pylori physically interacted with human T cells, T lymphocytes were purified from peripheral blood of H. pylori negative donors and co-cultured with viable G27 H. pylori strain. After 4 h of culture cell clustering was visible microscopically suggesting a direct interaction between T lymphocytes and the live bacteria ( Figure 1B). In contrast, formaldehyde fixed H. pylori were unable to exert the same effect ( Figure 1C). Lymphocyte activation was also evident by cytofluorimetric analysis because of an increase of cellular complexity (side scatter) of T cells cultured with bacteria, as compared to unstimulated control (data not shown).
To ascertain whether the T cell clustering was due to a direct interaction of the bacteria with purified T lymphocytes, cocultures were also examined by confocal microscopy, using GFPtransfected H. pylori. As shown in Figure 2A, green fluorescent bacteria were tightly surrounding most of the purified T cells. In addition, we assessed by flow cytometry the percentage of purified CD3+ cells that co-localized with GFP fluorescent bacteria. As compared to the control (Figure 2A), about 80% of CD3+cells had fluorescent bacteria bound to them ( Figure 2B). Interestingly, this interaction was nearly absent if the bacteria were treated with formaldehyde ( Figure 2C). Taken altogether these data strongly suggest that the bacteria directly interacted with T cells and that this interaction required viable H. pylori.

H. pylori activation of purified T lymphocytes in short term co-cultures
To investigate if the observed interaction also modulated the function of T lymphocytes, purified CD3+ cells were co-cultured with viable H. pylori, using fixed bacteria as a control, to assess the up-regulation of CD69, known to be an early activation marker antigen of lymphocytes. Figure 3 shows that CD69 was significantly up-regulated by CD3+ cells co-cultured with live bacteria, but not with killed bacteria. These data show that H. pylori driven T lymphocytes activation occurred in the absence of APCs, and suggest that this effect was independent of their antigen specificity. Moreover, we also found that the T cell responsiveness was not increased when we used PBMCs from H. pylori positive subjects (supplementary materials Figure S1). This suggests that the activation mechanism is not antigen-specific, and it does not depend on previous infections with H. pylori.
The ability to induce up-regulation of CD69 on CD3+ cells was also evaluated using a mutant of H. pylori G27 unable to synthesize CagA (DCagA). It is well known that CagA is translocated into gastric epithelial cells causing changes in cell structure, function and morphology [11]. The CD69 up-regulation was partially reduced when cells were co-cultured with the bacteria lacking CagA, as compared to wild type bacteria (average of 42% of reduction). In conclusion, bacterial viability, rather than integrity is required for CD3+ lymphocytes activation, with CagA being partially involved in this process.

H. pylori induced cytokine production by T cells in the absence of APCs
Subsequently we evaluated whether this activation of CD3+ cells after co-culture with H. pylori was accompanied by production of cytokines/chemokines in the supernatants. Indeed, H. pylori induced the production of cytokines such as TNFa, IFNc and chemokines such as MIP1-b, Rantes by CD3+cells. Very low levels of IL-2 were detected; moreover IL-6 was undetactable, indicating that our system was highly purified from APCs. Note that IL-6 was detectable at high levels when unfractionated PBMCs were stimulated with live H. pylori (for IL-6: medium = 33626 pg/ml versus live H. pylori = 13386 421 pg/ml). This effect was measurable already after 4 hours (Table 1), and increased during overnight stimulation. Production of cytokines and chemokines was confirmed by intracellular staining, after stimulation of purified CD3+ cells for 16 h with viable H. pylori. In addition, CD3+ T lymphocytes did not    Purified CD3+ cells from peripheral blood of H. pylori negative donors produce a wide range of cytokines and chemokines after 4 h of culture with H. pylori. Culture supernatants was collected and analyzed by bioplex assay. Data represent the mean and the range of cytokines and chemokines produced by T cells. Statistical significances of the differences between the cytokines production of CD3+/H. pylori co-culture compared to the control group (medium) were assessed using paired t test and reported in the third column. The differences in cytokines production in presence of H. pylori wt and DCagA were also calculated and compared, as reported in the last column of produce cytokines when co-cultured with killed bacteria (data not shown).
To evaluate whether this effect was specific of H. pylori or it was shared by other Gram-negative bacteria, some experiments were carried out with Escherichia coli. Unlike H. pylori, E.coli was unable to induce cytokines production by CD3+ T cells (Figure 4), nor CD69 up-regulation (not shown) suggesting that this stimulatory effect was peculiar of H. pylori, and not shared with other Gramnegative bacteria.
H. pylori induced up-regulation of CD69 and cytokines production by cd+ T cells in the absence of APCs We then asked which CD3+ T cells populations were preferentially activated by H. pylori. We observed that the majority of CD3+ producing cytokines were CD3+CD4-CD8-double negative. Therefore we asked whether, among these cells, cd+ T cells were preferentially activated. We found that 30-60% of CD3+CD4-CD8-T cells were TCR cd+ and 10% of total cd+ T cells produced TNFa and IFNc after co-culture with the bacteria ( Figure 5). These cd+ T cells also produced the chemokines MIP1-b and Rantes and about 90% of them up-regulated the activation marker CD69 (data not shown). It is worth of note that the activation of cd+ T cells by H. pylori clearly was independent on professional APCs, indicating that the mechanism was not antigen specific.
In order to better characterize the phenotype of cd+ T cells principally involved in the response to H. pylori, we evaluated the contribution of the Vd2+ T cell subset. This subset of cd+ T cells has been frequently reported to play a role in immunity against bacterial, parasitic pathogens and tumors [12]. We found that the majority of cytokine secreting CD3+T cells after co-culture with H. pylori were Vd2+ ( Figure 6). In addition DCagA was still able to induce T lymphocytes activation, although at a level lower than H. pylori wild type. Finally, no cytokine production was observed after stimulation of cd T cells with purified bacterial protein CagA (not shown), indicating that active processes mediated by H. pylori were required to achieve the highest activation of T cells and suggesting that CagA had to be actively translocated into the cells to exert its effect.

Discussion
In this study we show that H. pylori is able to stimulate peripheral blood T lymphocytes from H. pylori negative donors without the need for APCs. The fact that this effect happens on H. pylori negative subjects supports the idea that the stimulation is not antigen specific. This activation requires direct contact between viable bacteria and T cells, suggesting that the binding to the T cell membrane is necessary with a metabolically active bacterium. In particular, we found that live H. pylori is a potent activator of peripheral blood cd+ T cells. Indeed after only 4 hours of co-culture these cells showed significant up-regulation of the activation molecule CD69 and release of a wide range of cytokines (such as TNFa, IFNc) and chemokines (such as MIP-1b, RANTES). Moreover, the majority of cd+ T cells producing cytokines expressed the Vd2 TCR chain. Vd2+ subset is reported to be involved in the response against a wide range of pathogens [8,13] although there is still much to understand about the fine antigen specificity of these cells, especially in the context of H. pylori recognition.
Our findings that H. pylori is able to induce activation of T cells are consistent with a previous report [14] describing murine CD4+ T cell clones that were activated by the bacteria in the absence of APC. However in our study cells activated after contact with bacteria were essentially CD3+CD4-CD8-cd+ Tcells. This could be explained by the fact that we stimulated freshly isolated human peripheral blood instead of differentiated Th1 and Th2 murine clones. As suggested by the authors, terminally differentiated Th1 and Th2 murine cells could express receptors that allow activation by H. pylori [14]. Remain to clarify if also in human the CD4+ T cells after maturation could become more responsive to the H. pylori activation, although available evidence with human CD4+ T cell clones does not appear to support the murine data [6].
Since H. pylori resides at the apical side of the epithelial layer of the gastric mucosa, one can raise the question on how viable bacteria (and not just bacterial antigens) can interact with lymphoid cells. The interaction between H. pylori and T cells can occur after a damage of the epithelium cell layer during infection. The tissue injury can be mediated by the release of many bacterial products, such as the vacuolating toxin VacA, that is able to alter tight junctions, increasing permeability [15] or CagA, which is actively injected into the epithelial cells where it is responsible for the alteration of epithelial cell morphology [16] or via factors intervening at the basolateral side of the epithelium, such as reactive oxygen intermediates induced by bacterial products secreted or released after autolysis [17]. The damage of the gastric barrier may lead to infiltration of H. pylori in the sub-mucosa, generating an inflammatory status with infiltration of mononuclear cells [18,19]. The activation of cd+ T cells that we observed after in vitro stimulation with live H. pylori suggests that in vivo these cells may recognize and interact directly with the bacteria infiltrating the lamina propria. It has been reported that Vd2+ cells which are also found in the intestinal epithelium, might contribute to the epithelial homeostasis and might play an important role during the early stage of the immune response against pathogens [12]. Furthermore, some studies have reported an accumulation of cd+ T cells in the gastric mucosa of H. pylori infected subjects that seem to correlate with the severity of gastritis and infiltration of inflammatory cytokines [10].
In the present study we demonstrate, for the first time, that the activation of cd+ T cells requires viable H. pylori and in particular we report that Vd2+ cell population is affected by this stimulatory effect.
A wide variety of molecules have been described being able to activate Vd2+ T cells [12,20,21]. Among these, there are small compounds (like alkylamines and phosphoantigens) derived from stress-associated surface molecules as well as small bacterial metabolites or microbial compounds produced during infection and cellular stress. The activation of cd T cells by bacterial metabolites has been demonstrated in several studies using bacterial extract and supernatants containing purified antigens [20,22]. However, in our experimental conditions we found that H. pylori viability is necessary to induce the activation and requires whole bacterium/cell contact. Therefore this leads us to speculate that phosphoantigens or alkylamines are not the major components responsible for this process. Moreover recent reports have shown a functional expression of TLR 2, 3, 5 and 6 in freshly isolated blood Vd2+ cells [23,24]. This may suggest that PAMPs, expressed on Gram-negative bacteria surface, could be the putative molecules that activate cd+ T cells in our in vitro system. However, the fact that live E.coli or killed H. pylori are unable to exert this stimulatory effect tends to rule out an involvement of TLRs agonists in this activation process. Nevertheless, these observations do not exclude the possibility of a partial involvement of TLR agonists as co-receptors.
According to that recent studies reported that TCR cross-linking is required for TLR-mediated costimulatory effects on human cd T cells activation and expansion [25,26].
Our data support the notion that other H. pylori specific components must intervene in the cd+ T cells activation, which are produced by metabolically active bacteria. CagA represents an ideal candidate, since it is actively inoculated by the bacterium into the epithelial cells via its type IV secretion system [16,27,28]. We have shown that the activation of cd+ T cells by live bacteria is reduced when cells are co-cultured with bacteria lacking CagA. In the present study the activation of cd+ T cells require live CagA positive bacteria but it was not observed after stimulation with the purified CagA protein alone, suggesting the need for metabolically active bacteria able to mediate translocation of CagA into the lymphocytes. In agreement with the idea that the protein CagA may be internalized by the lymphocytes, as it happens with epithelial cells, we observed by flow cytometry a change in the cellular complexity (Side Scatter) after bacterium-T cell interaction that was nearly absent when T cells were cultured with the strain lacking CagA. The importance of bacterium viability is supported by the fact that treatments leading to kill H. pylori, such as formaldehyde fixation, neutralize its stimulatory abilities. On the contrary, the stimulatory activity is not altered after irradiation that makes the bacterium unable to reproduce itself while retaining its vital function. CagA, however, cannot be the only responsible for cd+ T cells activation, since DCagA strains retain their stimulatory ability. Several other factors, such as H. pylori type IV organelle, which are not well characterized, could contribute to the H. pylori stimulatory ability [29,30]. Thus, it is likely that contact-dependent secretion of one or more factors transported by the type IV apparatus could actively promote cd+ T cells activation. Further experiments will be necessary to better clarify all the molecular mechanisms that lie behind H. pylori/cd T cells interaction. Note that experiments of stimulation with VacAknockout H. pylori mutant strain showed that the stimulatory ability of the bacteria was retained even in the absence of VacA (supplementary materials Figure S2).
Our results on the in vitro H. pylori activation of cd+ T cells after short time interaction leads us to speculate that the bacteria could take advantage form this activation, by creating an environment that prevents the complete clearance of the pathogen. In fact, despite the strong immunological response, the pathogen is rarely eliminated and, in the absence of treatment, infection can persists for life. Previous studies have reported that chronic phase of inflammation is also characterized by the concomitant presence of regulatory T cells at the infection site can contribute to H. pylori persistence by suppressing antibacterial responses [31]. According to this, we observed that the direct contact of H. pylori with T lymphocytes activates T cells and in particular cd+ T cells to produce CCR5 agonists (such as MIP1-b) that could participate in the recruitment of Tregs at the site of infection, where they may start to exert their suppressive functions. Moreover it has been hypothesized that the Tregs cell homing in sites of infection may be driven by the chemokine receptor CCR5 that is described to be preferentially expressed by Tregs compared with normal CD4+ T cells [32]. In this frame we could speculate that the bacterium activates cd+ T cells to produce chemokines that may positively control the recruitment of regulatory T cells to the sites of gastric lesions.
Overall our findings showed that H. pylori can actively modulate the function of CD3+cells and particularly cd+ T cells. This activation may turn out to play a role in the maintenance of the local inflammation and eventually to the gastric and duodenal diseases.

Cell isolation and culture
Buffy coats were obtained from blood of donors serologically negative for H. pylori. PBMC were separated by Ficoll density centrifugation (Ge Healthcare, Little Chalfont, United Kingdom). The PBMC layer was recovered, washed and then resuspended in RPMI 1640 complete medium, (RPMI 1640 supplemented with L-glutamine and 25 mM HEPES, containing 5% Human Serum (Celbio, Milan, Italy) or 2% of FBS (HyClone South, Logan, UT).
For the co-culture experiments with H. pylori, PBMCs were seeded in 96-well flat-bottom plates at a density of 2610 6 cells/ well. In all tests performed, T cells viability was assessed over time using flow cytometric analysis and was comparable to that of unstimulated cells. CD3+ T cells were isolated from PBMCs by magnetic cell separation using the Dynabeads Cell Isolation kits (Invitrogen, San Diego, CA). The purity of cell preparations was confirmed by flow cytometry and was found to be greater than 98%. Some experiments were performed with sorted CD3+, CD3+CD4+, CD3+CD8+, CD3+CD4-CD8-cells using FACSAria cell sorter (BD, Becton, Dickinson Franklin Lakes, NJ). The yield of purification, confirmed by FACS, was greater than 99%. In addition, in some experiments we also used CD14 marker (BD, Becton, Dickinson Franklin Lakes, NJ) to ascertain that our cell preparation did not contain CD14+ cells following magnetic separation and/or cell sorting.
Bacteria were cultured microaerobically, using Campygen gas generating system (Oxoid, Cambridge, UK) for 12 hours at 37uC, on solid media consisting of Tryptic (trypticase) Soy Agar (TSA) plates containing 5% FBS and H. pylori selective Agar (DENT), supplemented with 200 mg/ml of kanamycin for the growth of kanamycin-resistant mutants. Bacteria harvested from the plates, were used to inoculate liquid cultures starting from an A 535 of 0.2 in 15 ml of BHI supplemented with 5%FBS and DENT (+/2 Kanamycin) and grown at 37uC in microaerophilic conditions with vigorous shaking (180 rpm) to an A 535 of 0.5. Bacteria were finally resuspended with BHI and 10 ml of bacteria suspension were added to T cells suspension in order to obtain a multiplicity of infection (MOI) of 100. After 14 h at 37uC under 5% CO 2 humidified atmosphere, cultures were observed with an optical microscope (Leica, Wetzlar, Germany) with 10x lents. For the formaldehyde fixation, bacteria from liquid culture were incubated in 2% formaldehyde in PBS for 30 min at room temperature and washed four times in PBS. For peptidase digestion, bacteria were incubated for 2.5 hours with proteinase K (Sigma-Aldrich, Taufkirchen, Germany) at final concentration of 2.5 mg/ml. Irradiated H. pylori were obtained treating 10 9 bacteria/ml of PBS at 6000 rad and then washed and resuspended at the working concentration.
Interaction between GFP-bacteria and T cells was imaged by Zeiss Observer LSM 710 confocal microscope. DAPI staining was used to determine the number of nuclei and to assess gross T cells morphology. Laser lines at 405 nm, 488 nm were used for excitation of DAPI and GFP respectively.

Analysis of activation markers and cytokines production by Flow cytometry (FACS)
T cell response was assessed by stimulating purified T cell populations with viable bacteria for 18 hours. Brefeldin (BFA, BD Biosciences, Franklin Lakes, NJ) was added after 2 hours in order to block proteins secretion. PBMC cultures in medium alone were included as negative control.
Samples were acquired with a FACS LSRII (BD Bioscience) and analyzed using Flowjo analysis software.

Statistical Analysis
Statistical analysis was done by the paired Student's paired T test with a two-tailed distribution.

Supporting Information
Figure S1 CD3+ cells from peripheral blood of H. pylori-positive (n = 3) donors produce cytokines and chemokines comparable to H. pylori-negative subjects (n = 4) (A). Culture supernatants was collected after 4 h of co-culture with H. pylori and analyzed by bioplex assay. Data represent the means and the range of cytokines and chemokines produced by T cells. No increase in cytokine and chemokine production was observed with PBMCs from H. pyloripositive subjects compared to the H. pylori negative. Note: n.d = not detectable. B. The percentage of CD69 up-regulation induced by viable H. pylori on CD3+ T cells is comparable in H. pyloripositive and H. pylori-negative subjects. Purified CD3+ cells were co-cultured with H. pylori (MOI 100). After 18 h cells were stained with anti CD3-PB and anti-CD69-APC. Numbers represent the percentage of CD3+CD69+ cells. The average was calculated from three independent experiments. (TIFF) Figure S2 H. pylori DVacA activate CD3+ T cells in a nonantigen-specific fashion after 16 hours of co-culture by inducing IFN-c production. No differences have been found between G27 wild type and H. pylori VacA knockout mutant, suggesting that VacA is not involved in this activation mechanism. On the contrary, in the presence of the mutant DCagA a reduction of IFN-c production was observed. Data are representative of two independent experiments with similar results. The numbers in each panel represent the percentage of IFN-c-producing CD3+ cells. (TIFF)