| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
An incorrect version of this article was published in Am. J. Trop. Med. Hyg., Vol. 68, No. 1, January 2003. The correct version of the article appears here. The Journal staff regrets this error.
Campylobacter jejuni causes a spectrum of disease in humans. Infection is initiated in the gastrointestinal tract, but can become extraintestinal in severe cases, particularly in immunocompromised hosts.7,8 In clinical reports describing primary infections with C. jejuni in developed countries, infection with mucosal disease predominates with symptoms of diarrhea, abdominal pain, and blood in the stool.9–11 However, infrequently, infection with systemic spread, infection without disease with short-term bacterial persistence, and infection with resistance and no bacterial persistence occur.9–11 In developing countries, C. jejuni enteritis usually occurs in infants; adults rarely have symptomatic disease.12 In these cases, the disease spectrum includes severe inflammatory illness, mild secretory diarrhea, or an asymptomatic carrier state. In addition, the duration of disease is shorter in children in hyperendemic areas compared with those with little previous exposure.
Campylobacter species possess significant genotypic and phenotypic diversity.13 However, C. jejuni strain variation does not account for all of these differences; the same strains producing watery diarrhea in children in developing countries have been isolated from visitors with acute inflammatory disease.14 In another study, there was no difference in serotypes isolated from symptomatic and asymptomatic children.15 The main explanation for the different disease expression in developing versus developed countries has been the early acquisition of immunity by children exposed to hyperendemic and/or polymicrobial infection.16–18 There is documentation of concurrent infections precipitating C. jejuni enteritis.19 These observations support the notion that this organism is capable of producing a spectrum of disease scenarios depending on the immune status of the host, virulence determinants of the bacterium, and complexity of concurrent infections.
In previous studies, inoculation of weaned swine with low levels of Trichuris suis (whipworm) initiated an interaction with resident bacterial flora to induce mucohemorrhagic enteritis in the colon.20 These pigs exhibited diarrhea, mucosal edema, inflammatory cell infiltration, and bacterial accumulation at the site of worm attachment in the proximal colon. In pigs infected with T. suis, lymphoglandular complexes (LGCs) in the submucosa of the distal colon were enlarged and contained numerous extra cellular and intracellular bacteria, eosinophils, lymphocytes, and macrophages. Control pigs infected with T. suis and treated with antibiotics had lesions localized to the site of worm attachment and histologically normal LGCs with no invasive bacteria in the distal colon. Another control group not inoculated with T. suis, but exposed to the same bacterial flora had no disease and no pathology. There was a T. suis dose-dependent increase in size of the follicular germinal centers and in the size of the lymphoid follicles within the LGCs. Multiple bacterial species were isolated from the LGC follicles; however, C. jejuni was the most common isolate. These results suggested that T. suis facilitated C. jejuni invasion of deeper tissues of the colon.
One billion people are infected with intestinal nematodes worldwide, and several million of these individuals are symptomatic.21,22 Ascaris lumbricoides, hookworms, and Trichuris trichiura are the most commonly diagnosed gastrointestinal nematodes associated with intestinal disease.22 Polyparasitism with hookworm, Ascaris, and Trichuris is common.23 It is possible that polymicrobial infections may explain some of the differences in clinical symptoms observed in young children with Campylobacter infections in developing countries. Based on the results in pigs, the following study was designed to examine whether a combined inoculum of C. jejuni and T. suis could synergize to enhance invasion of the colon by C. jejuni in an immunologically naive, germ-free pig and, if so, to explore the location of the invasion site. The results suggest that co-infection with a helminth that inhabits the proximal colon causes more severe consequences in immunologically naive pigs and could serve as a model to explore host resistance to C. jejuni, especially germinal center development in lymphoid tissues of the distal colon.
Preparation of germ-free incubators. Germ-free incubators were prepared in an animal containment facility. This facility had limited access with shower-in and shower-out facilities. Air entering the separate containment rooms was filtered and waste leaving the rooms was autoclaved or heated before disposal. Gnotobiotic swine incubators were sequentially disinfected and sterilized with Nolvasan (Fort Dodge Animal Health, Fort Dodge, IA), PolyClean (Madeira Inc., Spring, TX), and peracetic acid. All canned food and equipment required were passed through a barrier port and sterilized with peracetic acid spray.
Surgical procedure. Surgery was conducted in the germ-free incubators. Piglets were taken from sows by caesarean section at approximately 112 days gestation under standard surgical procedure in the containment facility surgery room. Sows were given atropine (0.045 mg/kg) intramuscularly, sedated with Acepromazine® maleate (Boehringer Ingelheim Vetmedica, Inc., St. Joseph, MO) (0.5 mg/kg intramuscularly), and an epidural anesthesia with 2% lidocaine was administered. Each sow was placed in lateral recumbency, and a standard surgical approach was performed through a paramedian incision. Pigs were delivered into sterile drapes and passed through a sterile sleeve into the gnotobiotic incubators by random assignment. When all piglets were delivered, the sow was given an overdose of sodium pentobarbital. Piglets were housed in germ-free incubators in individual cages in groups of 4–6 for social support and fed a sterilized complete milk diet (Esbilac; Borden Co., Hamshire, IL) for the duration of the experiment.
Experimental design. At three days of age, gnotobiotic pigs were infected with the following pathogens according to group; Group 1 (five pigs): control group (no infection), Group 2 (four pigs): infected with only with T. suis, Group 3 (five pigs): infected only with C. jejuni, and Group 4 (six pigs): infected with T. suis and C. jejuni. Pigs were examined three times per day for adverse signs. Abnormal findings, including descriptions and degree of diarrhea, were recorded. Fecal swabs were taken for bacteriologic isolation and identification on days 6 and 27 after infection. Pigs were maintained until severe clinical signs were observed in some, at which time, they were killed. At 27 days after infection, pigs were removed from the incubators and humanely killed with an intravenous overdose of sodium pentobarbital. Complete post mortem examinations were done, and tissues were recovered for various analyses.
Experimental infections. Trichuris suis. Embryonated eggs were prepared from adult whipworms isolated from colonic mucosa of experimentally infected pigs as described previously.24 Briefly, the worms were washed in sterile saline, then in sterile Hanks’ balanced salt solution (sHBSS), incubated in 5x concentrated antibiotics (500 U/ml of penicillin, 500 µg/ml of streptomycin, 1.25 µg/ml of amphotericin B, and 350 µg/ml of chloramphenicol) in RPMI 1640 medium for 16–24 hours, incubated in 1x antibiotic cocktail without chloramphenicol for 16–24 hours, and washed three times (two hours per wash) in sHBSS to remove residual antibiotics. Finally, worms were incubated for 10 days in RPMI 1640 medium containing 1% glucose (4 worms/ml) at 37°C in a humidified atmosphere of 5%CO2 for collection of eggs. Eggs were allowed to embryonate, examined microscopically, and embryonated eggs were picked with a pipette into individual doses. Piglets in groups 2 and 4 were inoculated by oral gavage with 3,000 embryonated T. suis eggs.
Campylobacter jejuni. Pigs in groups 3 and 4 were inoculated orally with a low dose of C. jejuni (1 x 106 colony-forming units [cfu]). The C. jejuni strain was second passage ATCC strain 33292 (American Type Culture Collection, Manassas, VA) isolated from a human with enteritis and characterized by testing with various media and conditions to optimize growth and ensure virulence. To test for virulence, three-day-old colostrum-deprived piglets were inoculated orally with 5 x 109 cfu of C. jejuni strains 33291, 33292, and 33560. Piglets inoculated with strain 33292 developed diarrheal disease within two days post-inoculation and C. jejuni were re-isolated from feces, amplified to generate low-passage bacterial stocks, and maintained in sheep blood at -80°C. Doses for infections were prepared from the second passage frozen stock. Bacteria were streaked onto Brucella agar supplemented with 5% sheep blood and incubated for 48 hours at 37°C in an atmosphere of 5% CO2. A single colony was picked into Brucella broth, which was grown at 37°C for 20 hours to early log phase. Confirmation of early log phase growth was based on growth curve determinations for strain 33292, optical density readings at 560 nm (OD560), and dark field microscopy in which the majority of organisms showed a spiral form and darting motility. Doses were made by resuspending 0.5 mL of the 0.1 OD560 inoculum in 49.5 mL of milkwarmed to 37°C. To confirm the desired inoculum of approximately 1 x 106 cfu per pig, limiting dilution analyses was done on the prepared doses of C. jejuni.
Necropsy procedure. The entire gastrointestinal tract was removed, divided into small intestines, cecum, and large intestines, opened longitudinally, and walls were examined for pathologic lesions. A staging system developed previously was used to score and rank the severity of the colonic lesions.20 Twenty-five LGCs were measured from the colon of each pig using a micrometer to determine the overall expansion of the lymphoid follicles. After lesions were evaluated, tissue samples were taken, the remaining colon was scraped with a glass slide, and the scraped material was incubated with sterile saline (0.85%) with 10 mM EDTA for 30 minutes at 37°C. After incubation, colonic scrapings were allowed to sediment for 30 minutes in graduated cylinders, the supernatant was aspirated, and the sediment was examined and enumerated for larval T. suis.
Histopathology. For each pig, full-thickness tissue samples were taken from the cecum, proximal colon, and distal colon for histopathologic evaluation. All samples were fixed in 10% formalin, embedded in paraffin, sectioned at 5 µm, stained with hematoxylin and eosin, and sections were observed and photographed (Optiphot-2 microscope with an H-III automatic camera; Nikon, Corporation, Tokyo, Japan) A scoring system was developed to evaluate histopathologic changes in representative villus crypt units of the proximal and distal colon from each pig. Each feature was given a rank from 0 to 4 (e.g., 0 = few cells or no pathologic change, 1 = slight cellular infiltrate, mild pathologic change, 2–3 = a moderate number of cells or moderate pathologic change, and 4 = many cells or maximal pathologic change). Specific features that were evaluated were as follows. The lumen was evaluated for excess mucus and inflammatory exudate. The epithelium was evaluated for surface integrity, number of intraepithelial lymphocytes, goblet cell hypertrophy, goblet cell depletion, crypt hyperplasia, crypt atrophy, adenomatous change in the crypt, and crypt inflammation. The lamina propria and submucosa were evaluated for increases in inflammatory or immune cells and their distribution, and the submucosa was evaluated for fibrosis. The thickness of the lamina propria, submucosa, and muscularis from the proximal and distal colon was measured in five fields with an ocular micrometer and the average thickness for these layers was recorded for each pig.
Immunohistochemistry. Twenty-five LGCs, full-thickness sections of the proximal and distal colon, and colonic Peyer’s patches from the cecum immediately distal to the ileal aperture were sampled in their entirety for immunohistochemical analysis. Tissue snips were overlayed with OCT compound (Sakura Finetek, Inc., Torrance, CA), snap frozen in liquid nitrogen, and maintained at -80°C until use. Tissues were sectioned at 4–6 µm with a cryostat (Reichert-Jung Cryocut Bensheim, Germany). Sections were adhered to either poly-L-lysine (Sigma, St. Louis, MO)–coated glass slides or charged glass slides (Fisher Scientific, King of Prussia, PA) and fixed in 4°C acetone for 5 minutes.
Tissue sections were ringed with PAP marker (Research Products International Corp., Mount Prospect, IL), rehydrated with TTBS (50 mM Tris, 0.1% Tween 20, 150 mM NaCl) for 10 min, and cross-reactive sites were blocked with B-TTBS (TTBS plus 1% bovine serum albumin) for 30 min. The primary monoclonal antibody (anti-C. jejuni outer membrane neat hybridoma supernatants or 1:500 purified ascites; Biogenesis, Sandown, NH) was diluted 1:10 with B-TTBS. Other sections from the LGCs were prepared and stained with the monoclonal antibody 5C9, which is directed against the IgM-µb of swine (kindly provided by Dr. Joan Lunney, U.S. Department of Agriculture, Beltsville, MD). Negative controls were treated with B-TTBS only instead of primary antibody, and a mixture of irrelevant anti-rat monoclonal antibodies. Sections were incubated with primary antibody at 22°C for three hours, washed with TTBS, and incubated for one hour with biotin-labeled F(ab’)2 fragments of sheep anti-mouse antibody (Sigma) diluted 1:100 in B-TTBS (previously cross-absorbed against human serum). Sections were then incubated for one hour with streptavidin–alkaline phosphatase (Sigma, St. Louis, MO) diluted 1:200 in TBS (50 mM Tris, 150 mM NaCl), followed by three washes (five minutes per wash) with TTBS. For colorimetric detection, 5 mg of naphthol AS-MX phosphate (Sigma) was dissolved in 250 mL of N,N-dimethylformamide and added to 10 mg of Fast-Red TR salt (Sigma) in TTBS. This solution was applied to tissue sections for 10–15 minutes, or until the background stain increased. Sections were counterstained with Gill’s I hematoxylin for 15 seconds and rinsed for two minutes in distilled water. The slides were dried and coverslips applied using aqueous mounting medium (Biomeda Corp., Foster City, CA).
Electron microscopy. Another set of samples from the same tissues was placed into Timm’s fixative for electron microscopy to correlate ultrastructural findings to individual cell data and to identify Campylobacter bacteria in situ.25 All samples were fixed for transmission electron microscopy (TEM) by immersing the pellet in 20 volumes of osmium tetroxide for one hour with gentle agitation. The samples were washed three times (10 minutes per wash) in Sorenson’s buffer, followed by three five-minute washes in sterile, double-distilled water. Samples were then incubated in 0.5% uranyl acetate overnight, washed in sterile, double-distilled water, and dehydrated in a graded ethanol series. Following dehydration, the samples were infiltrated by passing through increasing concentrations of Spurr’s medium (3:1, 1:1, 1:3, 100% ethanol to Spurr’s medium).25 Specimens were embedded in Spurr’s medium, cut in 60–90-nm sections on a Sorvall microtome (Kendro Laboratory Products, Asheville, NC), mounted on HX 200 copper grids, examined using a 100 CX electron microscope (JEOL, Inc., Peabody, MA), and photographed.
Statistical analysis. The size and number of the LGCs, the number of T. suis, and the number of episodes of a particular clinical sign from the four treatment groups were compared using the Kruskal-Wallis rank sum test. The pathology scores were expressed as the percentage of pigs with a particular stage of pathology over the total number of pigs evaluated.
Trichuris suis. Pigs in groups 2 and 4 were given 3,000 T. suis eggs in 5 mL of diluent. The actual percentage of the T. suis dose recovered was 0.1–1% (mean ± SD = 17 ± 13) in Group 2 (T. suis only) and 0.3–1% (mean ± SD = 21 ± 11) in Group 4 (T. suis and C. jejuni). Although the doses were low, all T. suis-infected pigs had T. suis and uninfected groups had no T. suis.
Clinical signs.
Clinical signs experienced by all pigs are summarized in Table 1
. Diarrhea was the most prominent adverse sign. Uninfected pigs had 11 episodes of diarrhea. Pigs given only T. suis had nine episodes of diarrhea. Pigs given both T. suis and C. jejuni had 68 episodes of severe diarrhea with blood and inflammatory cells. Pigs given only C. jejuni had transient fever for 24 hours after infection. They had 11 episodes of diarrhea over the observation period, which did not correlate temporally with fever. One pig each in groups 1, 3, and 4 died at day one of age before the experimental infections were performed due to blood loss when the umbilical clamps came off. Campylobacter jejuni was isolated at 6 and 27 days post-infection from feces of all pigs in groups 3 and 4, but not from any pig in groups 1 and 2.
|
and Figure 1D). Four of five pigs in this group had mucosal hemorrhage over more than 50% of the colon surface. The colon was inflamed, thickened, edematous, the surface epithelium was denuded in some areas, and many LGCs were filled with mucopurulent debris. Uninfected pigs and pigs with only T. suis had little gross pathology (Table 2 and Figure 1A and C). Two pigs given only C. jejuni had slight hyperemia located mainly in the Peyer’s patches of the small intestines and in small areas of the proximal colon surrounding and including the ileocecal Peyer’s patches (Table 2 and Figure 1B). In pigs infected with both T. suis and C. jejuni, there was a statistically significant increase in the size and in the number of LGCs in the distal colon (P 
0.05) (Table 3).
|
|
|
and 3). Uninfected control pigs had no lesions and served as the control for evaluating increases in thickness of the layers of the proximal colon (Figure 2, Group 1 and Figure 3A). Their mean ± SD values were mucosal = 295 ± 5 µm, submucosal = 57 ± 16 µm, and muscularis = 110 ± 23 µm. Pigs given only T. suis had crypt hyperplasia and increased thickness of all layers of the proximal colon (mucosal = 346 ± 24 µm, submucosal = 73 ± 21 µm, and muscularis = 142 ± 2 µm) with goblet cell hypertrophy (Figure 2, Group 2 and Figure 3C). Pigs given only C. jejuni had mild lesions and no crypt hyperplasia (mucosal = 298 ± 4 µm, submucosal = 72 ± 2 µm, and muscularis = 197 ± 50 µm) or goblet cell hypertrophy (Figure 2, Group 3 and Figure 3B). Pigs with dual infections had the greatest crypt hyperplasia with increased thickness of all layers of the colon (mucosal = 324 ± 9 µm, submucosal = 96 ± 2 µm, and muscularis = 160 ± 39 µm) and goblet cell hypertrophy (Figure 2, Group 4 and Figure 3D).
|
|
and 3). More marked infiltration of cells was seen in the pigs with dual infections in both the proximal and distal colon (Figure 3D). These pigs had large numbers of lymphocytes, macrophages, neutrophils, plasma cells, and eosinophils in the lamina propria and the submucosa (Figure 2, Group 4). Only one of five pigs given only C. jejuni had all of these cell types (Figure 2, Group 3 and Figure 3B). The predominant reactive cell types in pigs given only T. suis were lymphocytes, macrophages and eosinophils, which were in higher numbers than in the pigs infected only with C. jejuni (Figure 2, Group 2 and Figure 3C).
Immunohistochemistry.
IgM staining of the LGCs showed that germinal centers were present in all pigs given either T. suis, C. jejuni, or both. All pigs given C. jejuni had germinal centers in the LGCs, but the overall follicle size was significantly increased only in pigs infected with both agents (Table 3 and Figure 4D).
|
). Pigs with only C. jejuni had anti-C. jejuni staining in the LGC-entrapped crypts and within the follicle-associated epithelium of the LGCs (four of four pigs infected with C. jejuni) (Figure 5C). Pigs infected with both T. suis and C. jejuni had anti-C. jejuni staining in the LGC-entrapped crypts, within the follicle-associated epithelium of the LGCs, and within individual cells of the lymphoid follicle (five of five pigs infected with C. jejuni and T. suis) (Figure 5D). In the dual infected group, there was overall more anti-C. jejuni staining because the follicles were significantly larger than in the group given only C. jejuni. In the distal colon, none of the LGCs that were sectioned had T. suis adults or larvae within the nodule. Histology confirmed the presence of LGCs in all uninoculated, untreated, control pigs and in those pigs infected only with T. suis, but they were small, had no apparent germinal centers, and no anti-C. jejuni staining (none of two pigs infected with T. suis and none of four control pigs) (Figure 5A and B).
|
and 7). In the mucosal samples of pigs infected only with C. jejuni, there were lumenal bacteria, epithelial cell-associated bacteria, goblet cell-associated bacteria, and crypt-associated bacteria, but no bacteria within epithelial cells, the lamina propria, or beneath the muscularis mucosae (Figure 6). Dually infected pigs had bacteria in the lumen, associated with epithelial cells, associated with goblet cells, associated with the crypts, and within some epithelial cells (Figure 7). One pig in this group had C. jejuni in moderate numbers in the lamina propria and beneath the muscularis mucosae.
|
|
). In LGCs, bacteria were found in entrapped crypt mucus and superficially associated with goblet cells of the follicle-associated epithelium (Figure 6a–d). Pigs with dual infections had LGCs invaded by C. jejuni. Bacteria were found in epithelial cells in the follicle-associated epithelium, as well as in crypt mucus and goblet cells (Figure 7a–d). In these pigs, some C. jejuni were within vacuoles while others had no limiting phagocytic membrane that could be detected around bacteria within the epithelial cells.
In susceptible humans, C. jejuni may colonize either the small27,28 or large intestines, but most often results in colitis or typhlitis.29 Invasion of cells is suggested by the presence of abdominal pain, fever, diarrhea, frank blood in stools, and inflammatory cells in stools.10,29 Histologic sections of colonic biopsies show that patients with acute colitis have crypt abscesses, depletion of goblet cells, and inflammatory infiltrates of the lamina propria composed of polymorphonuclear leukocytes, lymphocytes, and plasma cells.29,30 Monocytes may play a role in the translocation of the organism to the blood stream, but the role of phagocytosis as a defense mechanism against C. jejuni seems to be secondary with the humoral response having the primary role against infection.8
There was site specificity of the pathologic lesions observed in dual infected pigs 27 days after infection. In the proximal colon where the worms reside, there was hemorrhage, crypt hyperplasia, thickening of submucosa and muscularis layers, edema, epithelial cell destruction, goblet cell hypertrophy and increased numbers of lymphocytes, macrophages, neutrophils, and eosinophils in the lamina propria and in the submucosa. In the distal colon, LGCs were significantly enlarged due to crypt abscesses and enlarged germinal centers. Pigs infected with T. suis only had mild crypt hyperplasia and goblet cell hypertrophy in the proximal colon and normal LGCs, suggesting that bacteria such as C. jejuni are required to produce the degree of pathology usually observed in field cases. Only two of four pigs given only C. jejuni had pathology with mild hyperemia in Peyer’s patches of the small intestines and colon. Possibly lymphoid tissues serve as a reservoir in swine for C. jejuni to maintain the colonized state.
Campylobacter jejuni were isolated and observed in the colon of all pigs receiving the bacterium. However, the presence of T. suis greatly influenced the location of C. jejuni. Pigs with dual infections had C. jejuni in crypt abscesses in the mucosa and the LGCs, associated with the surface of goblet cells, within epithelial cells especially in the follicle-associated epithelium of the LGCs, and within the lamina propria in inflamed areas of the proximal colon. Pigs infected only with C. jejuni had C. jejuni in colonic sections, especially in colonic crypts, where they were found mainly in the lumen or associated with mucus near goblet cells, but not within epithelial cells.
Babakhani and others found similar pathology when they orally infected newborn, colostrum-deprived pigs with a different invasive strain of C. jejuni.31 On necropsy of pigs from day 1 to day 6 after infection, they observed edema, hyperemia and mucus, as we saw in our dually infected pigs at day 27. In their study, histopathologic examination and transmission electron microscopy demonstrated damage to surface epithelial cells and the presence of intracellular bacteria in the large intestines. In our studies, it is unlikely that the group infected only with C. jejuni experienced pathology as severe as observed early in infection in their study because clinical signs were absent, except for transient fever. These varying results appear to be largely due to the different C. jejuni doses of these two virulent strains that were used. However, differences in C. jejuni strain virulence, the absence of other enteric flora, or genetic differences in the swine hosts may also play a role. In our laboratory, preliminary infections of germ-free pigs with a high dose of C. jejuni strain 33292 produced acute disease similar to that seen by Babakhani and others.31 Also, in several studies, dual infected pigs with naturally acquired C. jejuni showed the most significant effects 21–27 days after infection with T. suis. These data suggest that the worm has an effect and bacterial strain differences alone do not account for the differences in timing of the expression of pathology in these studies. Despite the differences in the experimental systems, the cell invasion pattern and electron microscopic and histopathologic lesions observed in our study and that of Babakhani and others31 were similar. Recent data shows that C. jejuni has an adhesin that binds to a cellular receptor.32 Ketley has hypothesized two models to explain C. jejuni epithelial cell invasion.11 He suggests motility and binding to epithelial cells via an adhesin leads to invasion by either 1) endocytic uptake via coated pits and association of the vacuole with microtubules, or 2) interaction with molecules associated with caveolae leads to transduction of a signal resulting in endocytosis involving actin filaments. Our data does not conclusively support either mechanism and suggests that a time course infection in young pigs with labeled C. jejuni with and without T. suis would be useful.
The LGCs appear to be important sites of C. jejuni colonization/invasion and immune induction sites for C. jejuni. This is likely due to their characteristic anatomy. They are submucosal secondary lymphoid structures in the distal colon of swine, humans, cattle, dogs, and other animals.33 They have structural features similar to Peyer’s patches with propria nodules that lie beneath the muscularis mucosae and mucosal glands that pass through the muscularis mucosae and lie among the nodule. A specialized follicle-associated epithelium (FAE) is present in these entrapped glands. The FAE directly overlays the follicle and forms the lining of entrapped, radially ramifying crypts within the follicle. It is similar to the FAE of jejunal and ileocecal Peyer’s patches, being composed of goblet cells, enteroendocrine cells, columnar and cuboidal enterocytes, and M-type cells associated with intra-epithelial lymphocytes.34 In this study, C. jejuni invaded the entrapped crypts of the LGCs, and all pigs exposed to C. jejuni had enlarged germinal centers in the LGCs. These findings suggest that C. jejuni is processed by M cells in the LGC FAE even in the absence of T. suis. In dually infected pigs, the dramatic increase in the size of the LGCs, expansion of the germinal centers, and the presence of C. jejuni within epithelial cells of the FAE suggest that T. suis enhances this response. Epithelial cells appear to be damaged in C. jejuni infections in most people with primary infections;10 however, the mechanisms controlling this are not known. In these pigs, electron microscopy showed that C. jejuni does invade epithelial cells in the FAE of the LGC. Also, immunohistochemistry showed anti-C. jejuni staining of cells with reticular morphology within the LGC follicles of dual infected pigs. It may be that macrophages within the LGC propria nodule, and especially in the pocket of M cells underlying the LGC FAE, are important in translocation of the bacterium from the lumen. Campylobacter jejuni has been found in M cells in Peyer’s patches of rabbits.35 This structure may provide a route for dissemination to deeper tissues and possibly the blood stream. Further studies with labeled C. jejuni are needed to identify the cells with C. jejuni inside the LGC follicle. Pigs given only C. jejuni had smaller LGCs with smaller germinal centers. The majority of C. jejuni-specific immunohistochemical staining of these LGCs was limited to the FAE of the entrapped crypt, but electron microscopy showed no bacteria within these cells. Therefore, we believe that anti-C. jejuni staining of epithelial cells in pigs infected only with C. jejuni is due to binding of antibody to C. jejuni breakdown products within vacuoles. In previous experiments, the propria nodules of LGCs increased in size with increasing numbers of T. suis inocula.20 Therefore, it was surprising that in this experiment pigs given only T. suis had no significant increase in the size of germinal centers in the LGCs. It appeared that the increase in size of the germinal centers was in response to the presence of C. jejuni. The nature of this epithelium, its interface with the lymphoid follicle, and the results of these experiments supports the role of the LGC in antigen sampling.
The mechanism by which T. suis synergize with C. jejuni is unknown but is likely to involve direct mechanical damage as well as cytokine-mediated damage. It is known that in the proximal colon T. suis first stage larvae invade basal crypt cells where they develop in a syncytium of epithelial cells.36 In this study, it appears that the majority of lesions are due to bacteria, but the parasite is required to elicit disease and pathology. It is likely that the lesions seen in dual-infected pigs resulted from the compound assault of colonic crypts by T. suis and C. jejuni. Earlier studies supported a synergistic relationship of T. suis and the microbial flora in producing the pathogenesis of mucohemorrhagic colitis in swine.37 Trichuris suis excretory-secretory products (ESPs) affect epithelial cell integrity,38 and ESPs cause increases in numbers of C. jejuni within in vitro cultured cells if cells are treated before addition of C. jejuni (L. D. Cunningham, Mansfield L. S., unpublished data). We speculate that T. suis larvae in the proximal colon may expose extracellular matrix necessary as cellular targets for adhesins of the bacterium such as fibronectin.32
Cytokine dysregulation may be an important contributor to this phenomenon, but functions differently than in inbred mice with similar dual infections. Trichuris causes marked inflammation and strong TH-2-associated immune responses in the colon of young experimentally infected, conventionally reared pigs,20,39 which apparently leads to susceptibility to C. jejuni. This model has proved useful for converting a C. jejuni colonized state to a disease state to examine the role of immune modulation in the intestine as a contributor to the pathogenesis of Campylobacter infections. Additionally, hyper-Th-1 polarized interleukin (IL)-10/4 knockout (KO) mice infected with T. muris developed severe ulcerating lesions, which were reversed by antibiotic treatment.40 However, it will be important to examine the difference between these systems and that of concommitant Helicobacter felis and Heligomoisoides polygyrus in C57BL/6 mice in which the helminth down-regulates immune-mediated pathology in the stomach due to the bacterium.41 Likewise, Toxoplasma gondii and Schistosoma mansoni synergize to promote hepatocyte dysfunction in C57BL/6 mice, but gut pathology is decreased in dual infected mice.42 Differences in worm species, worm location, worm-feeding mechanisms, and the stereotypical immune response patterning of these inbred mice compared with outbred pigs and specific IL-10/4 KO mice could explain the differences observed.
Received May 18, 2002. Accepted for publication September 11, 2002.
Acknowledgments: We thank Dr. Robert Holland for providing training with anesthesia and surgical techniques for the caesarian sections of the donor sows, Dr. Jon Patterson for reviewing the histopathologic and electron microscopic images, Sally Burns for technical assistance with electron microscopy, and Dr. Joan Lunney for providing monoclonal antibodies. We also thank Awdry Jones for expert technical assistance in providing T. suis eggs for infection of pigs.
Financial support: This work was supported by National Institutes of Health grant 61-0954, U.S. Department of Agriculture Animal Health Formula Funds grant 6881, National Food Safety and Toxicology Center grants 71-2954 and 71-9487, Michigan State University (East Lansing, MI), and Agricultural Research Service grant CHRIS 1265-32000-049.
Authors’ addresses: Linda S. Mansfield, David T. Gauthier, Sheila R. Abner, Kathryn M. Jones, and Stacey R. Wilder, B43 Food Safety Toxicology Center, Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, MI 48824, Telephone: 517-432-6309, Fax: 517-432-2310, E-mail: Mansfie4{at}cvm.msu.edu. Joseph F. Urban, Nutrient Requirements and Function Laboratory, Beltsville Human Nutrition Research Center, Agricultural Research Service, United States Department of Agriculture, Beltsville, MD 20705.
and early death in C57BL/6 mice. J Immunol 163: 2089–2097.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
