Helicobacter pylori Impairs Murine Dendritic Cell Responses to Infection

Background Helicobacter pylori, a human pathogen associated with chronic gastritis, peptic ulcer and gastric malignancies, is generally viewed as an extracellular microorganism. Here, we show that H. pylori replicates in murine bone marrow derived-dendritic cells (BMDCs) within autophagosomes. Methodology/Principal Findings A 10-fold increase of CFU is found between 2 h and 6 h p.i. in H. pylori-infected BMDCs. Autophagy is induced around the bacterium and participates at late time points of infection for the clearance of intracellular H. pylori. As a consequence of infection, LC3, LAMP1 and MHC class II molecules are retained within the H. pylori-containing vacuoles and export of MHC class II molecules to cell surface is blocked. However, formalin-fixed H. pylori still maintain this inhibitory activity in BMDC derived from wild type mice, but not in from either TLR4 or TLR2-deficient mice, suggesting the involvement of H. pylori-LPS in this process. TNF-alpha, IL-6 and IL-10 expression was also modulated upon infection showing a TLR2-specific dependent IL-10 secretion. No IL-12 was detected favoring the hypothesis of a down modulation of DC functions during H. pylori infection. Furthermore, antigen-specific T cells proliferation was also impaired upon infection. Conclusions/Significance H. pylori can infect and replicate in BMDCs and thereby affects DC-mediated immune responses. The implication of this new finding is discussed for the biological life cycle of H. pylori in the host.


Introduction
Helicobacter pylori is a Gram-negative, spiral-shaped, microaerophilic bacterium which colonizes the gastric mucosa. Infection occurs worldwide and is correlated with socioeconomic condition [1]. The prevalence among middle-aged adults is over 80 percent in many developing countries, as compared with 20 to 50 percent in industrialized countries. Overt diseases, however, occur in only 10-20% of infected individuals. It plays a causative role in chronic gastritis, peptic ulcer disease and is strongly associated with the development of gastric adenocarcinoma and mucosa-associated lymphoid tissue lymphoma [2]. The histopathological hallmarks of H. pylori-induced disease are a massive inflammatory cell infiltration of the lamina propria and erosion of the gastric epithelium. The gastric mucosa is well protected against bacterial infections. However, H. pylori is able to adapt and reside in the mucus, attach to epithelial cells, evade immune responses, and achieve persistent colonization in the stomach. It is not clear why the immune system fails to clear H. pylori infection. Furthermore, the mechanisms controlling the induction and maintenance of the H. pylori-induced chronic inflammation are only partly understood [3]. H. pylori is generally viewed as an extracellular pathogen colonizing the luminal side of the gastric epithelium and inducing an immune-inflammatory response on the stroma side of the mucosa. Paradoxically, H. pylori-specific gastric mucosal T cells generally present a Th 1 phenotype. It has been shown that Th1 cytokines promote gastritis, whereas Th2 cytokines are protective against gastric inflammation [4]. DCs are potent antigenpresenting cells that initiate T-cell responses. An abnormal pattern of DC maturation could promote chronic inflammatory processes [5]. Thus, the dysfunction of DCs may contribute to the local inflammation by producing inflammatory mediators and inappropriate activation of T cells.
Virulent H. pylori is capable of escaping phagocyte killing by delaying phagosome formation [6,7]. However, we have reported the Taiwan clinical isolates such as HP238 and HP250, not only survive but also replicates in human macrophage cell line THP-1 cells [8]. In this study, we further extend this research to show that H. pylori can multiply in BMDCs, and impair the function of BMDC.

Replication of H. pylori in murine BMDCs
H. pylori can multiply in macrophage autophagosomes [8]. To extend to DC, murine BMDC were infected with HP238 at a MOI of 10 for 1 h. Gentamycin was used to kill extracellular H. pylori for another 1 h. After gentamicine treatment, the cells were washed with PBS for 3 times and then cultured for additional 6, 12 and 24 h. At various time points post-infection (p.i.), BMDC were lysed, and viable H. pylori were re-cultured and plate-counted on a CDC plate for colony forming units (CFU) determination. As shown in Fig. 1A, CFU counts at 2 h p.i. corresponded to the amount of internalized live H. pylori. At 6 h p.i., a 10-fold increment was observed indicating that internalized H. pylori underwent replication. The number of live H. pylori gradually decreased at 12 h till 24 h p.i., suggesting that BMDCs gradually cleared the bacteria. The extracellular H. pylori was killed by the gentamicin, no viable H. pylori can be cultured from supernatant at 6, 12 and 24 h p.i. (data not shown). The role of VacA or CagA was also evaluated. The cagA and vacA mutants were also able to replicate in BMDC, but were cleared at a faster rate than wild type. No viable H. pylori could be recovered at 24 h p.i. (Fig. 1A).
H. pylori-associated autophagy in infected murine BMDCs Replication of H. pylori in murine BMDCs was examined under transmission electron microscopy. Intracellular H. pylori were found surrounded by a double layer membrane at 2, and 6 h p.i. (Fig. 1B-b,c), a characteristic of autophagosomes. Dividing H. pylori could also be observed in double-layer large membrane vacuoles ( Fig. 1B-f). At 12 h p.i., several autophagosomes fused with lysosomes to form a onion-like structure of multiple layer vesicles (Fig. 1B-e) and bacteria degradation were observed at 24 h p.i. (Fig. 1B-d). The autophagic marker LC3-II co-localized with the H. pylori-containing vacuole at 2 h p.i. (Fig. 2A). Increased numbers and densities of LC3 punctate staining were associated with replicating H. pylori at 6 h p.i., but gradually decreased at 24 h p.i. On Western blot analysis, the signaling molecules such as LC3 II conversion, BNIP3 formation, but not in Beclin1, were also found to increase after infection compared to uninfected cells (Fig. 2B). The p62 was induced at 2 h p.i., and degrade gradually till 24 h, p.i., indicating that an autophagic flux was induced post infection. This suggests that H. pylori infection induces autophagy and autophagosomes seem to be the site in which H. pylori transiently replicate before degradation by autophagolysosome.
Membrane trafficking of H. pylori and impairment of MHC class II molecule surface expression A delayed phagosome maturation associated to the formation of megasomes was reported to delay the clearance of H. pylori in macrophages [7]. Therefore, H. pylori trafficking in murine BMDCs was studied by confocal microscopy. H. pylori was detected in EEA1 early endosome-positive vacuoles at 2 h p.i.. Then, H. pylori wild type, cagA and vacA mutants were found in a LAMP-1-positive compartment at 24 h p.i. (Fig. 3B). At 48 h p.i., loss in the intensity of anti-H. pylori antibody staining was observed indicating that H. pylori had started to be degraded (Fig. 3A).
DCs are the most potent antigen presenting cells. Activated DCs express activation molecules, such as CD80 and CD86 and MHC class II molecules on the cell surface to present antigenic peptides to activate T cells. In H. pylori-infected murine BMDCs, MHC class II molecules were not transported to the cell surface and accumulated in the cytoplasm (Fig. 4A). In fact, MHC molecules were found to localize with the H. pylori-containing vacuoles (Fig. 4B). Under these conditions, H. pylori infection also impaired the IFN-c -induced up-regulation of MHC class II expression. MHC class II molecule transport to cell surface was also defective in formalin-fixed H. pylori infected BMDCs. Since the formalinfixed H. pylori is not capable of replicating, this suggests that H. pylori replication is not a prerequisite for the block of class II molecule transport to cell membrane (Fig. 4C). We used flow cytometry to stain the surface expression of MHC class I, II molecules and CD80 and CD86 in H. pylori-infected BMDC. The IFN-c was used as a positive control to enhance the surface expression of MHC class I and II molecules, CD80 and CD86. MHC class I expression was enhanced after H. pylori infection and the addition of IFN-c further increased MHC class I expression, indicating that H. pylori indeed activated MHC class I expression in BMDC. In contrast, MHC class II, CD80 and CD86 surface expression remained at the same level as untreated control immature cells (Fig. 4D). H. pylori infection also down-regulated the IFN-c enhanced MHC class II, CD80 and CD86 surface expression. Interestingly, formalin-fixed H. pylori infection induced the same activities as untreated bacteria. Therefore, transport of MHC class II molecules from the cytoplasm to the plasma membrane seems to experience interference from the presence of the H. pylori, either viable or formalin-fixed.
H. pylori infection stimulates murine BMDCs to secrete TNF-a, IL-6 and IL-10, but not IL-12 The cytokine secretion pattern after H. pylori infection was further analyzed at 24 h p.i.. Salmonella enteric serovar Thyphimurium infection was used as a positive control. TNF-a and IL-6 secretion in supernatants from H. pylori or formalin-fixed H. pylori infected cells were found to be similar (Fig. 5A). Interestingly, IL-10 secretion required BMDCs stimulated by live H. pylori only. No IL-12 secretion was observed from the H. pylori-infected BMDCs ( Fig. 5A-d). The vacA and cagA mutants were also tested for the production of IL-12. These mutants also failed to produce IL-12 ( Fig. 5A-e). These results suggest that H. pylori infection impairs the activation and maturation of DCs, probably by down regulating the secretion of pro-iflammatory cytokines such as IL-12 and, in the case of live bacteria by promoting the secretion of IL-10.
Negative role of TLR4 or TLR2 on the BMDC maturation post H. pylori-infection H. pylori LPS uses both TLR4 and TLR2 for cell activation [9]. Hence, we compared H. pylori infection in BMDCs derived from TLR4-deficient C3H/HeJ mice and in C3H/HeN control mice. Although the presence of H. pylori in BMDC from TLR4-deficient C3H/HeJ mice is slightly lower than that from C3H/HeN mice ( Figure S1A), H. pylori can replicate in TLR4-deficient BMDC. However, the MHC class II surface expression in infected TLR4deficient BMDCs is not inhibited (Fig. 6B) and no co-localization of MHC class II molecules with H. pylori-containing vesicles was found (Fig. 6A). For cytokine production, TNF-a and IL-6 secretion were lower in infected TLR4-deficient BMDC than in wild type BMDC. However, IL-10 secretion was higher in TLR4deficient BMDC than in wild type BMDC (Fig. 5B). No IL-12 production was detected for both BMDCs (data not shown).
We also compared DC activities in wild type versus TLR2 knockout mice. H. pylori can replicate in BMDCs derived from either TLR2 wild type or TLR2 -/-( Figure S1B). In TLR2 -/-BMDC, MHC class II surface expression inhibition was transient, found only at 24 h p.i. Then, an enhanced surface expression of MHC class II molecules was observed at 48 h p.i. (Fig. 6C). The cytokine expression was also modulated in TLR2 -/-. TNF-a and IL-6 were lower in TLR2 -/-BMDC than in wild type BMDC. However, in contrast to what is observed in TLR4-deficient mice, IL-10 was not detected in the supernatants of TLR2 -/infected BMDC (Fig. 5C).
Finally, we assayed antigen presentation in H. pylori-infected BMDC to H. pylori-specific T cells. Lymph node T cells derived from H. pylori-immunized mice were stimulated with H. pyloriinfected BMDC in a ratio of 1:5 and the BrdU incorporation was determined. Comparing with the H. pylori-infected TLR2 -/-BMDC that has an increased MHC class II molecule surface expression post-infection, H. pylori-induced T cell proliferation by wild type BMDC was significantly lower than that of TLR2 -/-BMDC (Fig. 6D). Whether H. pylori infected BMDC, killed H. pylori-treated BMDC, or IFN-c treated H. pylori-infected BMDC, the same pattern was displayed, indicating that H. pylori-treated BMDC showed a defect in antigen presentation.

Discussion
We have demonstrated here that H. pylori is capable of multiplying in BMDCs after invasion. A 10-fold increase of CFU between 2 h and 6 h p.i. is found in H. pylori-infected BMDCs. The doubling time of the H. pylori in BMDC cells was calculated to be 1.5-2 h. Autophagy is induced by H. pylori and the double-membrane autophagosomes is associated H. pylori after its invasion. Then, the H. pylori-containing autophagosomes fuse with lysosomes to lead to degradation of the bacteria. The finding of this transient replication of H. pylori in BMDCs suggests that H. pylori can be considered as a facultative intracellular microorgan-ism under certain condition. Moreover, its multiplication in BMDCs would impair the functions of the dendritic cells.
H. pylori is generally viewed as a non-invasive pathogen localized only in the lumen of the stomach and attached to gastric epithelial cells. It has adapted several mechanisms to reside and persist in the mucus, avoid phagocytosis and evade immune responses. The phagocytosis process is retarded and the membrane trafficking of phagosome maturation is disrupted to form a megasome that contains the ingested bacteria [6,7]. Many reports have shown that H. pylori is invasive, and can be considered a facultative intracellular organism [10,11]. However, no direct demonstration of its replication by plaque counting in infected cells has so far been provided. We have found that H. pylori could multiply not only in macrophages [8], and in epithelial cells (data not shown), but also in BMDCs as shown in this study. Two Taiwanese clinical isolates of HP238 (from gastric adenocarcinoma specimen) and HP250 (from mucosa-associated lymphoid-tissue lymphoma speciment) can multiply in macrophage cell line THP-1 [8] whereas standard strains of ATCC43504 and J99 can multiply in AGS or MKN45 epithelial cell line (unpublished data). For BMDCs, only HP238 is found to replicate in this study. This new finding has several implications for the biological life cycle of H. pylori in the host, especially considering the role of DC as controllers of the immune responses. Wunder et al. reported that H. pylori is auxotrophic for cholesterol, and can extract cholesterol from epithelial membrane. The glycosylation of cholesterol would inhibit the phagocytosis of H. pylori and promote its immune evasion [12]. Cholesterol has been found to play an essential role in the establishment of the chronic infection of intracellular microorganisms, in that the phagosomal lipid metabolism of the host macrophage was modulated by the virulent factors of the bacteria to favor the survival of the invading bacteria [13][14][15]. H. pylori can also infect the AGS cells via a sipper-like mechanism, involving intimate contact with AGS cells microvilli and surface membrane pseudopod structure. The internalized H. pylori are found in LAMP-1-containing vacuoles [16]. Cytoskeletal rearrangement with tyrosine phosphorylation was observed in the promixity of intracellular H. pylori. We found that H. pylori infection induced autophagosome formation in either phagocytic THP-1 cells, BMDC, or non-phagocytic AGS epithelial cells. H. pylori resides and replicates within autophagosomes, probably reflecting its dependence on the cell membrane probably because of its cholesterol requirement for growth. The CagA can be translocated to plasma membrane and form rafts with the invading H. pylori [17]. In our study, Cag A or VacA mutants can still replicate in BMDC, but were cleared in a fast rate probably due to the interference in the autophagosome formation by Cag A or VacA. This needs further investigation. Rittig et al reported the H. pylori-induced homotypic phagosome fusion is independent of the VacA or Cag A in human monocytes [18]. Since we used murine BMDC, whether H. pylori can replicate in human dendritic cells need to be checked in the future.
In view of the observation that H. pylori infection induces autophagy, there remains the question as to what role autophagy plays in the defense against H. pylori. Autophagy has been found to be a component of the innate cellular immune responses against not only intracellular but also against extracellular microorganisms [19,20]. After phagocytosis of the microorganism, the autophagosomes will be formed to degrade the ingested bacterium by the lysosomal killing mechanism. The intracellular bacteria such as Legionella pneumophila, Listeria monocytogenes, Mycobacterium, and Shigella have developed different mechanisms to evade the autophagic cellular surveillance in macrophage. The double membranebound compartments that bear autophagic markers were induced, but its maturation into autophagolysosomes will be arrested or delayed. Allen et al. have reported the megasome formation induced by H. pylori via homotypic phagosome fusion to delay its clearance in macrophage [6,7]. However, for the H. pylori-induced autophagosome, the autophagosome not only provides the site for replication, but also the place for the destruction of the replicating bacterium. After phagocytosis into BMDC, H. pylori co-localized with EEA1 early endosomal marker and then with the LAMP1 +   HeN) was compared to the TLR 4 deficient (C3H/HeJ) mice. In C, the B6 wild type was compared to the TLR2 knock out mice. (D) Antigen-specific T cell late endosomal/lysosomal marker (Fig. 3). Multiple vesicle vacuoles formed at 12-24 h p.i. that subsequently resulted in the destruction of the bacteria due to fusion with lysosomes (Fig. 1B). IFN-c treatment enhanced the clearance of the H. pylori also supporting the notion that autophagy helps the clearance of internalized H. pylori in the endosome/autophagosome after fusion with lysosomes ( Figure S2).
TLRs are involved in the innate immune response to H. pylori infection [9,[21][22][23]. H. pylori has adapted its LPS structure for colonization, adhesion and escaping from immune responses. The phase-variable expression of these Lewis antigens seems to affect the induced immune response by modulation of cell activation through DC-SIGN. Le X antigen negative expressed H. pylori escape binding to dendritic cells and induce a strong Th1 response. In contrast with Le X antigen negative expression, H. pylori variants that express Le X /Le Y can bind to DC-SIGN on dendritic cells and enhance the production of IL-10 which promote Th2 response [24,25]. Cell activation by H. pylori LPS has been shown to be both TLR2-and TLR4-dependent, although less stimulatory than E. coli LPS, and sometime antagonistic for TLR4 [26][27][28]. For H. pylori LPS, TLR2 is primarily activated [21,28], and induces the production of IL-10, also skewing the immune response to Th2 arm [9,25,29,30]. In our study, the secretion of TNFa and IL-6 was inhibited in TLR4and TLR2-deficient BMDCs, indicating that both TLR2 and TLR4 participate in the response to H. pylori LPS. Intriguingly, IL-10 production was enhanced in TLR4-deficient BMDCs, but completely abolished in TLR2 -/-BMDCs. TLR4 was reported to negatively regulate the fusion between phagosomes and lysosomes in macrophage for the degradation of the engulfed antigen [31]. IL-10 has been found to down-regulate the MHC class II export to plasma membrane, probably via inhibition of cathepsin S [32,33]. How the H. pylori components interfere with the MHC class II molecules transport to cell surface is intriguing. VacA is initially characterized as its ability to induce the formation of intracellular vacuoles in cell lines, and has been found to decrease the pH value of acidic vesicle. Molinari M., et al., demostrated VacA interferes the Ii-dependent pathway of antigen presentation in B cells [34]. During the phagosome maturation, the VacA-containing H. pylori or VacA recruit and retain the coronin 1 to disrupt the phagosome maturation and inhibit the antigen presentation [34,35]. The urease of H. pylori can also bind to CD74 (the class II-associated invariant chain that regulate the intracellular transport and functions of class II) to affect the cell motility of dendritic cells [36], or cause megasome formation [37]. In addition, several molecules such as RP105, ST2, SIGIRR, or Galectin-3 have been reported to regulate the TLR4 or TLR2 signaling negatively [38][39][40][41][42][43]. In our study, surface MHC class II molecules, or even after IFN-c-treatment, were down-regulated by H. pylori infection in BMDCs. The formalin-fixed and heat-killed H. pylori can still remain the inhibition of the MHC class II expression. The MHC class II molecule seems to be trapped in H. pylori containing vesicles. The TLR-2 and 4 signaling also play a negative role on the class II surface expression. It must be a multiple factors involvement. The Ii, urease, IL-10, TLR2/4 signaling will all contribute to the inhibition of class II export and these are under current investigation.
DC can phagocytose and process bacteria for antigenic peptide presentation by MHC class II molecules to specific T cells. DCs are activated and secrete IL-12, which favors the differentiation of splenocytes into Th1 [4,27,[44][45][46]. Inhibition of IL-12 secretion by H. pylori was also reported [47,48]. In this study, IL-10, but not IL-12, was detected after H. pylori infection. In addition, MHC class II molecule export to the cell surface is blocked, which results in an impaired antigen presentation. This will favor the induction of a non-Th1 response. A complexity of interaction between H. pylori and host immune defenses is responsible for the persistence of H. pylori infection [3]. For H. pylori, Th1 has been supposed to be protective as well as immunopathogenic, which is paradoxical. CD8 + T cells are also activated and are associated with severe gastritis in H. pylori-infected mice [49]. The replication of H. pylori within the cells can explain the endogenous antigen presentation and activation of CD8 + T differentiation post infection. In conclusion, the finding that H. pylori can multiply in dendritic cells will add a new element to the complexity of the immune responses to the H. pylori infection, and shed light on further understanding of immunopathology in H. pylori-induced disease.

Bacterial strains and culture
The H. pylori clinical isolate (HP238) was obtained from the Department of Pathology, National Cheng Kung University Hospital. The mutant of cagA and vacA were used as previously described [8]. H. pylori were grown on CDC anaerobic 5% sheep blood agar plates (BBL, Becton-Dickinson, USA) under microaerophilic conditions (5% O 2 , 10% CO 2 , 85% N 2 ) and 85% humidity in a Nuaire incubator (Plymonth, Minnesota, USA) at 37uC. Fresh plates were started from glycerol stocks and subculture every 48 hours. In bacterial replication assays, the strain used in this study were HP238, and its vacA or cagA isogenic mutants [8] at a multiplicity of infection of 1:10. Bacteria were centrifuged onto BMDCs at 400 g for 6 min at 4uC and incubated at 37uC with 5% CO2 atmosphere one hour. After washed with phosphate-buffer saline (PBS) for 3 times, the cells were incubated with 5% FBS, 100 mg/ml gentamicin RPMI1640 media at 37uC for another one hour. After gentamicine treatment, the cells were washed with PBS for 3 times and then cultured with dendritic cell culture media for 6, 12 and 24 h. After co-incubation of BMDC with H. pylori for 6, 12 or 24 h, BMDC were lysed with 0.5 ml of 0.1% saponine (in DPBS) and then plated on CDC plates with serial dilution to determine the viable bacteria. Colonies were grown and counted after 7 days culture. The time point of 2 h is considered as the entry of H. pylori into the cells. The increase of CFU at 6 h is interpreted as result from multiplication of H. pylori within the cells. There was no viable bacterial cultured from supernatant post 6, 12 and 24 h post infection (p.i.). Bacteria were treated with 4% formalin for 10 min and then washed 3 times with PBS. H. pylori was heat-killed at 95uC for 20 min.

Bone marrow-derived dendritic cells culture
Bone marrow cells were collected from the bones of 6-10-week old C57BL/6 (wild and TLR2 knockout) mice, C3H/HeN or C3H/HeJ female mice. The 1.66106/well cells were cultured in 6-well plate with RPMI1640 media containing 10% fetal bovine proliferation. BMDCs derived from B6 or TLR2 -/mice were infected with H. pylori, heat-killed H. pylori, or IFN-c plus H. pylori for 24 h. Then BMDCs were mixed with H. pylori-primed lymph node cells at a 1:5 ratio. The T cell proliferation was determined after 5 days by BrdU cell proliferation ELISA kit. The absorbance at 450 nm was depicted. Wild type (&) and TLR2 -/-( ) BMDCs were compared for their antigen presentation to H. pylori-specific T cells. H. pylori-stimulated wild type BMDCs have lower antigen presentation than that of TLR2 -/-. * p,0.05 via student t-test. doi:10.1371/journal.pone.0010844.g006 serum, 50 mM 2-mercaptoethanol and 20 ng/ml mouse recombinant GM-CSF at 37uC for 5 days to differentiate into dendritic cells. The media were replaced with fresh media on 3rd day. The mice were maintained in the pathogen-free facility of the Animal Laboratory of National Cheng Kung University (Tainan, Taiwan) that were raised and cared for according to the guidelines set up by the National Science Council, ROC. The use of mouse experiments were approved by the institutional animal care and use committee of NCKU. Immunofluorescence confocal microscopy BMDC were incubated with either H. pylori, or formalin-fixed H. pylori for 1 h, then washed and treated with 100 mg/ml gentamicin. For the IFN-c treated groups, recombinant mouse IFN-c (100 U/ml) was added after the gentamicin treatment step, and were then fixed with 3% paraformaldehyde, pH 7.4, at room temperature for 10 min, and washed with phosphate buffer saline 3 times. The primary antibodies used were: rabbit anti-H. pylori antibody (ABR, USA), mouse anti-H. pylori (Abcam, UK), rabbit anti-LC3 II antibody (Abgent, USA); rat anti-lysosomal-associated membrane glycoprotein 1 (LAMP-1) antibody 1D4B (eBioscience, USA); goat anti-human EEA-1 antibody (Santa Cruz, USA), and rabbit anti-MHC class II antibody. The secondary antibodies used were: FITC-conjugated goat anti-rabbit, TexRed-conjugated donkey anti-rat and cyanin-5-conjugated anti-goat antibodies. Samples were observed on a Zeiss LSM 510 laser scanning confocal microscope for image acquisition.

Transmission electron microscopic examination
For ultrastructural analysis, H. pylori infected cells at different time points were fixed with 4% glutaraldehyde and post-fixed in 1% OsO 4 . The cells were observed under transmission electron microscopy (Hitachi 7000, Japan).