|Home | About | Journals | Submit | Contact Us | Français|
Campylobacter jejuni is a leading bacterial cause of human diarrheal disease in both developed and developing nations. Colonic mucosal invasion and the resulting host inflammatory responses are thought to be the key contributing factors to the dysenteric form of this disease. Dendritic cells (DCs) play an important role in both the innate and adaptive immune responses to microbial infection. In this study, the interaction between human monocyte-derived dendritic cells and C. jejuni was studied. We found that C. jejuni was readily internalized by DCs over a 2-h period. However, after a prolonged infection period (24 or 48 h) with C. jejuni, only a few viable bacteria remained intracellularly. Minimal cytotoxicity of C. jejuni to dendritic cells was observed. C. jejuni induced the maturation of dendritic cells over 24 h, as indicated by up-regulation of cell surface marker proteins CD40, CD80, and CD86. In addition, Campylobacter-infected DCs triggered activation of NF-κB and significantly stimulated production of interleukin-1β (IL-1β), IL-6, IL-8, IL-10, IL-12, gamma interferon, and tumor necrosis factor alpha (TNF-α) compared to uninfected DCs. Active bacterial invasion of DCs was not necessary for the induction of these cytokines, as heat-killed C. jejuni stimulated similar levels of cytokine production as live bacteria. Purified lipooligosaccharide of C. jejuni appears to be the major stimulant for the increased production of cytokines by DCs. Taken together, these data indicate that during infection, Campylobacter triggers an innate inflammatory response through increased production of IL-1β, IL-6, IL-8, and TNF-α and initiates a Th1-polarized adaptive immune response as predicted from the high level of production of IL-12.
Campylobacter jejuni is a spiral, gram-negative, microaerophilic bacterium. This pathogen is a leading bacterial cause of human diarrheal disease throughout the world. In the United States, about 2 million cases of diarrhea per year are caused by C. jejuni. Illness caused by C. jejuni is characterized by fever, headache, abdominal cramping, diarrhea, and the presence of erythrocytes and leukocytes in the stool (4, 5). Two to 3 weeks postinfection, C. jejuni is sometimes associated with the extraintestinal manifestations of Guillain-Barré syndrome or reactive arthritis (43). The pathogenic mechanisms of C. jejuni are not yet well understood. In developing countries, C. jejuni triggers primarily a noninflammatory, watery diarrhea. However, in developed countries infection via bacterial invasion of the colonic mucosa is thought to lead to intestinal inflammation and microulceration of the colonic mucosa. As is the case with shigellosis, the acute host inflammatory responses are thought to contribute heavily to disease pathogenesis (reviewed in reference 22).
Dendritic cells (DCs) play important roles in both the innate and adaptive immune responses to microbial pathogens. They are the major antigen-presenting cells (APCs) and are widely distributed in tissues, including the intestinal mucosa (reviewed in references 26, 32, 53). Dendritic cells regulate the type of T-cell-mediated immune response to an offending agent and are also a source of proinflammatory cytokines. They are the only cells that are able to initiate proliferation in naïve T cells, thereby inducing the primary immune response and permitting the establishment of immunological memory.
Immature DCs can capture antigens by phagocytosis, macropinocytosis, and endocytosis (17, 37, 38). Exposure of DCs to inflammatory stimuli converts these cells from antigen-capturing immature DCs to antigen-presenting mature DCs. This process is accompanied by up-regulation of major histocompatibility complex (MHC) class I and II molecules, costimulatory receptors, and adhesion molecules (2, 7, 11). Once a DC has captured an antigen, its ability to capture additional antigens rapidly declines. The process of differentiation from an immature DC into a mature professional APC can be induced by whole bacteria, their components, or cytokines as well as other inflammatory stimuli and infectious agents. Upon activation, DCs migrate from the inflammatory site into the lymphoid tissues, such as the lymph nodes and spleen, down-regulate their phagocytic ability, and up-regulate their antigen-presenting capacity (2).
Numerous studies have shown that DCs play an important role in the host-pathogen interaction during infection with enteric pathogenic bacteria. For example, DCs that phagocytose Salmonella enterica serovar Typhimurium can process and present bacterial antigens and produce cytokines that are critical to the immune response (25, 56). Helicobacter pylori also induces maturation and cytokine release from human DCs (24). Shigella flexneri infection of DCs leads to up-regulation of interleukin-1β (IL-1β) and IL-18 and rapid DC death (8). In contrast, Yersinia enterocolitica is able to invade DCs and does not induce necrosis or apoptosis but impairs DC function (40).
During infection, Campylobacter can invade and traverse the epithelial barrier and comes into contact with numerous leukocytes (49, 50). In addition, DCs are capable of traversing the tight junctions of the intestinal mucosa, which enables them to interact directly with bacteria on the mucosal surface (35, 39). Although the infection of macrophages by C. jejuni has been studied previously (14, 19, 20, 42, 44, 51), Campylobacter's interaction with DCs has not been reported to date. Understanding the interaction between DCs and C. jejuni will provide insight into the role of DCs in stimulating Campylobacter-induced inflammatory pathology and in controlling the infection. The goal of this work was to study the interaction between human DCs and C. jejuni with respect to DC activation and induction of a number of cytokines known to be involved in both the innate and adaptive immune responses.
Campylobacter jejuni strain 81-176, obtained following a disease outbreak in Minnesota, has been shown to cause colitis in two human challenge studies (4, 23; D. Tribble et al., unpublished data). C. jejuni 81-176 and its mutant RY213 (a cheY+ diploid) that is noninvasive (55), as well as the genome-sequenced C. jejuni NCTC 11168 (33), were grown in Mueller-Hinton (M-H) biphasic broth or on M-H agar (Difco) at 37°C under a Campylobacter gas atmosphere of 10% carbon dioxide, 5% oxygen, and 85% nitrogen (16).
Mononuclear cells were obtained by apheresis of normal volunteer donors as performed by the Blood Services Section of the Department of Transfusion Medicine at the National Institutes of Health Warren G. Magnuson Clinical Center (Bethesda, MD). The mononuclear cells were further enriched for monocytes by centrifugal elutriation as performed by the Cell Processing Section of the Department of Transfusion Medicine. Similar to the work of Pickering et al. (34), the elutriated monocytes were then cultured (1 × 106/well) in six-well tissue culture plates with 3 ml/well of RPMI 1640 medium (Mediatech, Inc., Herndon, VA) containing 5% human AB serum (Nabi, Miami, FL), 800 U/ml granulocyte-macrophage colony-stimulating factor (GM-CSF; Peprotech, Rocky Hill, NJ), and 500 U/ml IL-4 (Peprotech). The nonadherent and loosely adherent cells were harvested after incubating the plates for 6 days at 37°C in 5% CO2. The cells were centrifuged for 10 min at 1,200 rpm in a Beckman GS-6KR centrifuge and then plated in 96- (50,000 cells/well) or 6-well plates (106 cells/well) containing fresh RPMI (Mediatech) with 5% human AB serum (Nabi) and 800 U/ml GM-CSF (Peprotech). This resulted in the generation of immature monocyte-derived dendritic cells that were positive for CD11c but negative for CD14 (data not shown).
C. jejuni was added to the cultured DCs at different multiplicities of infection (MOIs). Bacterium-DC interactions were initiated by centrifugation at 1,000 rpm in a Beckman GS-6KR centrifuge for 5 min, followed by incubation at 37°C in 5% CO2. For assessment of the number of intracellular bacteria, 2 × 105 infected DCs/well in 24-well plates were cultured for 2, 4, 24, and 48 h. After the 4-h time point, gentamicin (20 μg/ml) was added to inhibit the growth of extracellular bacteria (19). At the specified time points (2, 4, 24, and 48 h), cells were washed three times with RPMI 1640, followed by incubation of DCs for 2 h in fresh medium including 100 μg/ml gentamicin to kill any remaining extracellular bacteria. The infected monolayers were then washed three times to remove the gentamicin. The DCs were lysed with 0.1% Triton X-100 in phosphate-buffered saline (PBS) for 15 min. Following serial dilutions in PBS, the viable internalized bacteria were enumerated by plate count on M-H agar. For cytokine measurements, the infected culture supernatants in 96-well plates were harvested, centrifuged, and frozen at −80°C prior to analysis (12, 34).
In control studies, 100 μg/ml gentamicin was found to kill all extracellular bacteria within 2 h. The MIC of gentamicin at which 50% of the C. jejuni 81-176 organisms were inhibited was ~6.25 μg/ml for C. jejuni 81-176. At 20 μg/ml of gentamicin, we found that extracellular C. jejuni could not multiply in cell culture medium. We used 20 μg/ml gentamicin to block any extracellular growth of released C. jejuni over longer periods (i.e., 24 to 48 h), because higher antibiotic concentrations over long periods can leak into the host cells and cause death of some intracellular bacteria.
The CytoTox 96 assay (Promega, Madison, WI) was used to determine host cell cytotoxicity over time induced by bacterial preparations that were added to the DCs. This assay qualitatively measures supernatant lactate dehydrogenase (LDH), a stable cytosolic enzyme that is released upon cell lysis. The DC number was evaluated via the assay instructions and determined to be optimal at 50,000 cells/well. The ratios of effector cells to bacteria used to assess host cell cytotoxicity were 1:1, 1:10, 1:20, and 1:100. Supernatants were collected following centrifugation at 4, 24, and 48 h postinfection with C. jejuni 81-176. Maximum LDH release was determined by measuring the amount of LDH release from uninfected DCs that were treated with lysis buffer. The percentage of cytotoxicity was calculated according to the following formula (OD is an abbreviation for optical density): [(ODsample) − (ODmedium/ODmax LDH release) − ODadjusted medium] × 100.
Immature DCs were incubated in 96-well plates (5 × 104 DCs/well) for 4, 24, and 48 h with live C. jejuni 81-176, heat-killed (i.e., 70°C for 30 min) C. jejuni 81-176, lipooligosaccharide (LOS) of C. jejuni 81-176 (extracted by using the hot phenol-water technique ), or hot phenol-water-extracted lipopolysaccharide (LPS) of Escherichia coli (Sigma, St. Louis, MO). PBS (Invitrogen Corporation)-incubated DCs served as unstimulated controls. Cytokine assays were performed by using human cytokine 10-plex and IL-12p70 antibody bead kits (Biosource International, Camarillo, CA) and a Luminex 100 analyzer (Luminex Corporation, Austin, TX) for IL-1β, IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12, gamma interferon (IFN-γ), tumor necrosis factor alpha (TNF-α), and GM-CSF. Cytokine data analysis was performed using the MasterPlex QT quantitation software (MiraiBio, Alameda, CA).
Flow cytometric analysis was performed in order to examine the cell surface maturation markers CD40, CD80, and CD86 in Campylobacter-infected DCs. Dendritic cells were infected as described above. Maturation of DCs was determined at 24 and 48 h after infection with C. jejuni (MOI, 10) by checking the expression of the surface maturation markers by flow cytometry. Cell samples (5 × 105 cells) were incubated (20 min, 4°C) with 4% human AB serum (Nabi) in PBS (Invitrogen Corporation) to block nonspecific binding sites and Fc receptors. The cells were washed via centrifugation (1,200 rpm; 10 min) with fluorescence-activated cell sorter buffer containing 0.2% bovine serum albumin (Sigma) and 0.1% sodium azide (Sigma) in PBS (Invitrogen Corporation) and then incubated (45 min, 4°C) with anti-CD40 (5C3), anti-CD 80 (L307.4), or anti-CD86 (2331 [FUN-1]) fluorescein isothiocyanate-labeled antibody (1 μl; Becton Dickinson, San Jose, CA) or the isotype-matched (1 μl of immunoglobulin G1; X40) control antibody (Becton Dickinson). After washing again, the cells were resuspended in 0.5% paraformaldehyde (Sigma) in cold PBS (Invitrogen) and 0.1% sodium azide. The cell-associated immunofluorescence was measured with a FACSCalibur flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, CA) and analyzed using CellQuest software (B-D Biosciences).
The TransAM NF-κB kit p65 (Active Motif, Carlsbad, CA) was employed to detect the activation of NF-κB. At 4 h postinfection, the nuclear extracts were prepared as per the manufacturer's instructions. Protein concentrations were determined using a 2′-benzoyloxycinnamaldehyde protein assay reagent kit (Pierce, IL). Nuclear extracts (10 μg/ml) were added into each well of a 96-well plate and incubated with anti-p65, and NF-κB concentrations were determined spectrophotometrically per the kit instructions.
Results are presented as the mean ± standard error of the mean from three independently conducted assays. P values were calculated with Student's t test.
C. jejuni was added to DCs at MOIs of 1, 10, 20, and 100. After a 2-, 4-, 24-, or 48-h invasion period, followed by a 2-h gentamicin (100 μg/ml) treatment to kill all extracellular bacteria, the numbers of viable, intracellular Campylobacter were quantified by plate count after host cell lysis. At 2 and 4 h postinfection, the uptake was generally dose dependent, and higher doses of bacteria led to greater levels of bacterial internalization by DCs (Fig. (Fig.1).1). The average number of viable intracellular bacteria per DC was ~1.5 bacteria/DC on average at 2 or 4 h at an MOI of 10 and increased about fourfold at an MOI of 100. By 24 or 48 h postinfection, however, the numbers of viable intracellular bacteria were dramatically reduced. For example, at an MOI of 100, only ~1% of DCs contained one viable bacterium, with even less observed at 48 h. These data suggest that C. jejuni is internalized at early time points, but most intracellular bacteria are killed by 24 and 48 h postinfection. This assumption is predicated on little or no cytotoxicity to DCs during bacterial infection, which was studied next.
To assess the viability over time of DCs following infection with C. jejuni, LDH release assays were performed using the supernatants of both uninfected and infected cells. No significant differences in DC viability were detected at 4 and 24 h after infection at any bacterial MOI, in comparison to the uninfected controls (Fig. (Fig.2).2). At 48 h postinfection, all infected DCs showed a minimal increase in cytotoxicity over the uninfected control. However, the 9.5% dead cells in the uninfected control at 48 h was only doubled to ~20% death in the infected DCs at the highest MOI. These data suggest that limited host cell cytotoxicity was associated with both long infection times and higher MOIs. However, infection-induced DC death over 48 h was very limited even at the highest MOI.
To study the effects on DC maturation of Campylobacter infection, the expression of cell surface costimulatory molecules CD40, CD80, and CD86 was measured with flow cytometry. Compared to uninfected cultures, C. jejuni-infected DCs showed increased expression of all three surface molecules at 24 h postinfection, as indicated by increased fluorescence intensity of DCs (Fig. (Fig.3).3). The averaged mean fluorescence intensity measurements for CD40, CD80, and CD86 at 24 h in all three sets of control DCs were 12.52± 0.99, 7.01 ± 0.38, and 45.28 ± 10.83, respectively. Infection with C. jejuni 81-176 significantly increased these maturation marker averaged mean fluorescence intensity measurements to 21.66 ± 3.84 (P< 0.05), 25.99 ± 10.83 (P < 0.01), and 96.16 ± 7.35 (P < 0.005), respectively. As a positive control, purified E. coli LPS induced DC maturation similarly to that observed with live C. jejuni (data not shown). These data show that C. jejuni infection stimulates an increase in the expression of DC surface costimulatory molecules, which are indicators of DC maturation.
The kinetics and profile of cytokine secretion from human DCs during Campylobacter infection were analyzed. Cell culture supernatants were collected at 4, 24, and 48 h from uninfected DCs or after infection with strain 81-176 at different MOIs, and cytokine levels were measured in multiplex bead assays. Campylobacter infection of DCs induced enhanced production of IL-1β, IL-6, IL-8, IL-10, IL-12, IFN-γ, and TNF-α (Fig. (Fig.4)4) but not IL-2, IL-4, or IL-5 (data not shown).
At 4 h postinfection with C. jejuni at an MOI of 10, the titers of IL-6, IL-8, and TNF-α were increased ~29-fold (837.80 ± 83.00 pg/ml), 34-fold (1,073.70 ± 282.90 pg/ml), and 1,959-fold (46,596.30 ± 11,977.78 pg/ml), respectively, compared to the uninfected controls (Fig. 4B, C, and G). The production of IL-6 and IL-8 as well as IL-12 and IFN-γ (Fig. 4E and F) was significantly increased at 24 and 48 h postinfection in comparison to the 4-h time point. TNF-α was detected at high levels at all time points from infected DCs. C. jejuni 81-176 induced comparatively lower quantities of IL-1β or IFN-γ than the other cytokines; however, IL-1β and IFN-γ levels were significantly increased at 24 and 48 h postinfection relative to the 4-h time point (Fig. 4A and F). Finally, substantial increases in IL-10 were only observed at the 48-h time point (Fig. (Fig.4D).4D). These data show that increases in the Campylobacter MOI from 10 to 100 had little impact on overall cytokine production. Cytokine production increased over time, with peak amounts at 24 or 48 h.
Wild-type invasive C. jejuni 81-176, the noninvasive RY213 mutant strain of 81-176 (55), NCTC 11168, a minimally invasive C. jejuni strain (1), and heat-killed 81-176 were compared for their ability to trigger DC secretion of selected cytokines. C. jejuni 81-176 was heat treated at 70°C for 30 min and subsequently shown by plate count to be nonviable. The production levels of IL-1β, IL-6, IL-8, IFN-γ, and TNF-α (Fig. 5A, B, C, F, and G) at 24 h postinfection were similar for all strains. The mutant RY213 and NCTC 11168 showed slightly lower IL-10 and IL-12 induction levels compared to live 81-176 (Fig. 5D and E). Heat-killed 81-176 cells generally induced slightly lower cytokine levels than live 81-176 cells, but this difference was, in most cases, not significant. Using an MOI of 100, the heat-killed bacteria did not induce any higher cytokine levels than when at an MOI of 20 (data not shown). These data suggest that neither active bacterial invasion of DCs nor bacterial viability is necessary for C. jejuni to induce cytokine production, but live wild-type 81-176 cells generally result in slightly better induction than other stimulants.
LOS was prepared from C. jejuni 81-176 by the hot phenol-water extraction method and identified as LOS on silver-stained sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels (data not shown). LOS (100 ng/ml) was added to DCs for 24 h. An equal amount of purified LPS from E. coli, known to induce cytokine production in DCs, was used as a control. Both LOS and LPS induced cytokine levels similar to that seen with live C. jejuni (Fig. (Fig.6),6), with the exception of a lower IL-10 induction by C. jejuni LOS. These data suggest that LOS of C. jejuni is a key contributor to the induction of cytokine production by DCs.
The host transcription factor NF-κB promotes transcription of various immune response genes, including cytokines (10). The nuclear extracts exhibited low NF-κB activation in the uninfected control DCs. In contrast, NF-κB activation increased threefold (P < 0.05) in DCs infected with C. jejuni 81-176 for 4 h (Fig. (Fig.77).
DCs play important roles in both the innate and adaptive immune responses to microbial pathogens (26). The intestinal immune system DCs lie beneath the mucosal surface in Peyer's patches and form an early line of defense against invading pathogens. Intestinal DCs can internalize bacterial pathogens following their uptake through M cells or more directly through paracellular dendrite extensions into the intestinal lumen (35). Here, we analyzed the interaction between C. jejuni and human DCs to investigate DC activation, potential bacterial or DC toxicity, and the induction of cytokines associated with disease inflammation and immune responses to infection. Our data show that DCs internalize C. jejuni at early time points (i.e., 2 h postinfection), taking up one to six bacteria per DC on average at MOIs of 10 to 100. Ninety-nine percent of internalized C. jejuni were killed within 24 h, and there was virtually no cytotoxicity to DCs induced over 24 h, with only minimal increased DC death (~10%) after 48 h. Wassenaar et al. (51) found that activated human monocytes/macrophages also efficiently killed C. jejuni, causing a 5-log reduction in 24 h. Our results differ from those of Siegesmund et al. (42), who reported that C. jejuni F38011 induced apoptosis in 63% of THP-1 (human monocyte line) cells 48 h after inoculation.
In striking contrast, many studies report that Salmonella is internalized by and survives within both human and murine DCs (6, 9, 18, 29, 45), although Salmonella has been reported to rapidly kill infected DCs and macrophages (21, 25, 48). Shigella infection of human DCs results in rapid IpaB-dependent DC death (8), which probably dampens the adaptive immune response to this enteric pathogen. Similar to our findings with Campylobacter, Y. enterocolitica enters DCs and does not induce necrosis or apoptosis (40). However, Yersinia reverses the activation of DCs, reduces DCs' abilities to stimulate an adaptive response, and prolongs infection. For Campylobacter and some enteric pathogens, it is still largely unclear as to what primary factor(s) determines intracellular bacterial survival or triggers host cell necrosis/apoptosis.
In this study, C. jejuni stimulation of immature DCs resulted in their maturation to APCs, as evidenced by enhanced expression of cell surface costimulatory molecules CD40, CD80, and CD86. Similarly, other enteric pathogens, such as S. enterica serovar Typhimurium (30) or Helicobacter pylori (24), have been shown to up-regulate DC surface expression of MHC class II, CD40, CD80, and CD86 molecules. Coordinated up-regulation of the costimulatory molecules and translocation of MHC molecules to the cell surface are essential molecular events for subsequent antigen presentation and ultimate activation of both CD4+ and CD8+ T cells (36).
Monocytes/macrophages are considered early host responders to infection and are also an important source of proinflammatory cytokines. C. jejuni has recently been reported to induce IL-1, IL-6, IL-8, and TNF-α in human monocytes (14, 19). Dendritic cells also produce various cytokines and chemokines upon interaction with pathogenic bacteria (8, 24, 25, 56). Our data show that C. jejuni activates the transcriptional factor NF-κB and induces the production of significant amounts of IL-1β, IL-6, IL-8, IL-10, IL-12, IFN-γ, and TNF-α in DCs. These cytokines play crucial roles in the induction of inflammation and in adaptive immune responses (28, 31, 47).
Proinflammatory cytokines IL-1β, IL-6, IL-8, and TNF-α were induced rapidly in DCs infected with C. jejuni and were secreted and maintained at high levels over 48 h. The cytokine expression pattern was not markedly altered by increasing the MOI from 10 to 100. These findings add to the growing body of information which shows that C. jejuni triggers the innate immune system to produce proinflammatory cytokines, which likely serve to initiate and modulate local inflammation in the Peyer's patches, leading to disease symptoms. Previous studies of C. jejuni-infected human cell lines have shown that this pathogen also induces epithelial cell secretion of IL-8 and some chemokines (13, 15, 27), attracting phagocytic cells that may be important in both containing and in exacerbating the infection (54). Consistent with these results, our studies show that many proinflammatory cytokines, including IL-8, were stimulated in a significant and rapid fashion in DCs.
As DCs mature, they synthesize cytokines essential for the development of T-cell interactions (i.e., adaptive responses). Compared to other types of APCs, DCs are 1,000-fold more efficient in activating resting T cells (3). Interleukin-12 appears to be a key cytokine produced by APCs to stimulate a Th1-directed response (7, 46). C. jejuni induced very high levels of IL-12 at 24 and 48 h after infection. IL-10 is an important Th2-type cytokine that up-regulates humoral and down-regulates cell-mediated immune responses (31, 41). The high IL-12 production (i.e., 10,000-fold induction) and relatively low IL-10 induction (i.e., 100- to 500-fold induction) by C. jejuni favors a Th1 response.
The induction of cytokine production in DCs was not significantly different between live and heat-killed bacteria, except for IL-10, which was slightly lower with the heat-killed bacterial stimulus. Similar findings were previously reported in C. jejuni-infected monocytes (19). The noninvasive RY213 strain and C. jejuni NCTC 11168, which have low invasion efficiencies (1), induced cytokines IL-10, IL-12, and TNF-α at levels reduced only by about one-half versus strain 81-176. These data argue that invasion is not necessary for the induction of these cytokines. In fact, purified LOS from strain 81-176 induced high levels of the selected cytokines studied, with the exception of IL-10. These data indicate that the predominant induction of cytokines in DCs may be due to interaction with LOS. Some alternate pathway is likely needed for full IL-10 induction. The signal transduction pathways modulated by the interaction of C. jejuni with DCs are currently being investigated in our lab. Taken together, these data suggest that DCs are an important component of the host-pathogen interaction with Campylobacter. Further, DCs play a role in triggering inflammatory cytokines likely involved in disease pathogenesis and in initiating a Th1-directed adaptive immune response during C. jejuni infection.
We acknowledge T. T. Wai for help in preparing Campylobacter LOS and thank R. I. Walker, M. Akkoyunlu, and S. Dharwan for critical review of the manuscript.
Editor: J. T. Barbieri