Analysis of the Contribution of Salmonella Pathogenicity Islands 1 and 2 to Enteric Disease Progression Using a Novel Bovine Ileal Loop Mode
http://www.100md.com
感染与免疫杂志 2005年第11期
Michael Smith Laboratories
Department of Microbiology and Immunology
University of British Columbia, Vancouver, British Columbia V6T 1Z4, Vaccine and Infectious Disease Organization, University of Saskatchewan, Saskatoon, Saskatchewan S7N 5B4
Department of Veterinary Pathology, Western College of Veterinary Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
ABSTRACT
We have developed a novel ileal loop model for use in calves to analyze the contribution of Salmonella enterica serovar Typhimurium type III secretion systems to disease processes in vivo. Our model involves constructing ileal loops with end-to-end anastamoses to restore the patency of the small intestine, thereby allowing experimental animals to convalesce following surgery for the desired number of days. This model overcomes the time constraint imposed by ligated ileal loop models that have precluded investigation of Salmonella virulence factors during later stages of the infection process. Here, we have used this model to examine the enteric disease process at 24 h and 5 days following infection with wild-type Salmonella and mutants lacking the virulence-associated Salmonella pathogenicity island 1 (SPI-1) or SPI-2 type III secretion systems. We show that SPI-2 mutants are dramatically attenuated at 5 days following infection and report a new phenotype for SPI-1 mutants, which induce intestinal pathology in calves similar to wild-type Salmonella in the 5-day ileal loop model. Both of these temporal phenotypes for SPI-1 and SPI-2 mutants were corroborated in a second animal model of enteric disease using streptomycin-pretreated mice. These data delineate novel phenotypes for SPI-1 and SPI-2 mutants in the intestinal phase of bovine and murine salmonellosis and provide working models to further investigate the effector contribution to these pathologies.
INTRODUCTION
Salmonella enterica are facultative intracellular enteric bacteria that cause a range of clinically important infections in humans and commercial livestock. Within the S. enterica species, there are numerous serovars, discriminated by their lipopolysaccharide and flagellar and capsular antigens. The natural progression of Salmonella disease is strongly influenced by both the S. enterica serovar and the host. For example, S. enterica serovar Typhimurium is a pathogen with a wide host range, infecting humans, cattle, mice, and chickens. In humans and cattle, infection with S. enterica serovar Typhimurium manifests as enterocolitis that rarely spreads to systemic organs. In susceptible mice, however, serovar Typhimurium efficiently crosses the gut epithelium and colonizes the spleen and the liver (22, 27). Infection of chickens is mainly a colonization model that rarely manifests in disease.
Two major virulence determinants involved in Salmonella pathogenesis are encoded in large chromosomal pathogenicity islands called Salmonella pathogenicity island 1 (SPI-1) (9) and SPI-2 (20, 24). Both SPI-1 and SPI-2 encode separate type III secretion systems that introduce virulence proteins (called effectors) into the host environment either by translocation directly into host cells or possibly by secretion into the vicinity of host cells (10, 14, 18). Effector proteins translocated by the SPI-1 type III secretion system influence early host cell cytoskeletal and membrane rearrangements involved in bacterial uptake into target cells (8) whereas SPI-2 is generally thought to play a role during intracellular infection by allowing the formation of Salmonella replicative vacuoles and evading host cell defenses (30).
Animal models are available to study both intestinal and systemic phases of salmonellosis caused by S. enterica serovar Typhimurium. Cattle have been used to study the enteric disease caused by this organism, whereas infected mice develop a systemic infection that shares features with human typhoid. S. enterica serovar Dublin is also commonly used in bovine infection models because this particular serotype is able to initiate both intestinal and systemic phases of infection in cows.
An important theme to emerge from these models is the seemingly dichotomous role played by the SPI-1 and SPI-2 type III secretion systems during the intestinal and systemic phases of salmonellosis. Whereas the SPI-1-encoded type III secretion system plays an essential role in colonization of the bovine intestine and in bovine enteropathogenesis (12, 19, 29, 31), this virulence trait has been reported to play little to no role in systemic infection (11). Conversely, the SPI-2-encoded type III secretion system is more strongly associated with systemic virulence and associated pathology (4, 17, 20, 23) than intestinal disease. The role of the SPI-2-encoded type III secretion system in enteric disease has been less studied and appears more contentious. Earlier work demonstrated that SPI-2 mutants maintain intestinal virulence during oral infection of cows (26) and in ileal loops in rabbits (7). However, one study using bovine ileal loops infected with S. enterica serovar Dublin (2) indicated that mutations in SPI-2 reduced the intestinal secretory response compared to wild-type serovar Dublin. Another study using oral infections of calves with a bovine isolate of serovar Typhimurium (26) indicated that SPI-2 mutants produced intestinal lesions of reduced severity in calves but nevertheless caused mortality and acute diarrhea. One of the confounding factors for the study of enteric disease in cattle caused by S. enterica serovar Dublin is the highly invasive nature of the infection that results in a disseminating bacteremia (33). In contrast, natural and experimental infection of calves with S. enterica serovar Typhimurium results in an enteric disease localized to the gut without systemic involvement, which shares similar pathological and clinical features to human enterocolitis caused by this same organism (33).
Recently, a mouse model of Salmonella-induced colitis has been developed that relies on the pretreatment of animals with streptomycin (1). This model has been used to demonstrate that Salmonella-induced colitis at day 2 after infection requires the SPI-1 effectors SipA, SopE and SopE2 (15), plus flagella and chemotaxis (25). Using this model, we (5) and others (16) have extended these findings by showing that the SPI-2 type III secretion system contributes to the intestinal inflammatory phenotype in the cecum and proximal colon between day 2 and day 5 following oral infection of mice with S. enterica serovar Typhimurium. Although these studies have identified a role for SPI-2 in the ceca of susceptible mice after 2 days of infection, as yet these data have not been extended to the bovine ileal loop infection model at the same time points.
The ligated ileal loop model in cattle is a proven tool for studying enteropathogenesis of Salmonella serotypes (28, 33). However, due to obstruction of the small bowel with multiple ligatures, one recognized limitation of this model in its present usage is the inability to monitor pathological changes in response to Salmonella at times longer than 12 h postinfection (34). This has precluded the investigation of the relative contribution of SPI-1 and SPI-2 to the progression of intestinal disease in bovine loops and the comparison to other intestinal inflammatory models of serovar Typhimurium infection. To address this limitation, we have developed a novel calf ileal loop model that restores the patency of the small intestine following surgery. This feature overcomes the time constraints of traditional loop experiments, permitting the examination of virulence determinants at later stages of enteric disease in a highly controlled environmental context. Here, we establish the relative contributions of S. enterica serovar Typhimurium SPI-1 and SPI-2 type III secretion to early and late intestinal inflammation. Our results identify novel phenotypes for S. enterica serovar Typhimurium mutants with deficiencies in SPI-1 and SPI-2 type III secretion systems, which we have demonstrated in a murine model of infectious colitis and confirmed in the extended bovine ileal loop model. The working model presented here supports the view that during the progression of enteric disease, the presence of SPI-2 overcomes the previously reported requirement of SPI-1 for intestinal inflammatory disease.
MATERIALS AND METHODS
Bacterial strains and culture conditions. Salmonella enterica serovar Typhimurium strain SL1344 (32) was used as the wild-type strain throughout this study and mutants lacking SPI-1-mediated type III secretion (SB103; invA::Kan) (11) or SPI-2-mediated type III secretion (ssaR [cows]; ssaR bearing pAT113-GFP [mice]) (3) were isogenic derivatives of this strain. An S. enterica serovar Typhimurium double mutant defective for both SPI-1 and SPI-2-mediated type III secretion was constructed by generalized P22 transduction of a marked invA::Kan mutation to a strain with an unmarked, in frame deletion of ssaR (6). For infection of experimental animals, bacterial cultures were grown for 18 h at 37°C with shaking in Luria-Bertani (LB) medium and then diluted to the appropriate concentration in phosphate-buffered saline (PBS).
Mouse infection experiments. Male C57BL/6 mice (6 to 8 weeks old) were purchased from Jackson Laboratories (Maine, USA). Mice were housed in sterilized, filter-top cages under specific pathogen-free conditions at the University of British Columbia Animal Facility. The protocols used here were in direct accordance with animal care guidelines as outlined by the University of British Columbia Animal Care Committee and the Canadian Council on the Use of Laboratory Animals. The streptomycin pretreatment model of murine infectious colitis was used as described previously (1, 5). Mice were deprived of food and water for four hours and then given 20 mg of streptomycin by oral gavage. Two hours after antibiotic treatment, food and water were provided for twenty hours. Four hours prior to infection, food and water were removed once again and then mice were infected orally with 3 x 106 or 3 x 108 bacteria in a 0.1 ml volume. Control mice were given 0.1 ml of sterile LB medium. Water and food were provided ad libitum following infection. Mice were euthanized with CO2 at 2 days and 5 days after infection and tissues were harvested for bacterial enumeration and histopathology.
Bacterial enumeration from mouse tissue. Mouse colon, cecum, spleen, and liver were collected into 1.5 ml of cold sterile PBS and homogenized using a tissue homogenizer (Polytron MR-21; Kinematic). Serial dilutions of the homogenized organs were spread on LB agar plates containing 100 μg/ml streptomycin and incubated overnight at 37°C.
Murine histopathology. Colons, ceca and ilea of experimental animals were fixed in 3% formalin for 18 h followed by 18 h in 70% ethanol and then embedded in paraffin, sectioned at 5 μm and stained with hematoxylin and eosin. For pathological scoring, six fields per sample were examined and scored as described previously (5). This scoring system was as follows: for lumen, sum of empty (score = 0), necrotic epithelial cells (scant = 1, moderate = 2, dense = 3), and polymorphonuclear cells (PMNs) (scant = 2, moderate = 3, dense = 4); for surface epithelium, sum of no pathological change (score = 0), regenerative change (mild = 1, moderate = 2, severe = 3), desquamation (patchy = 1, diffuse = 2), and PMNs in epithelium (score = 1), ulceration (score = 1); for mucosa, sum of no pathological change (score = 0), crypt abscesses (rare [<15%] = 1, moderate [15% to 50%] = 2, abundant [>50%] = 3), presence of mucinous plugs (= 1), and presence of granulation tissue (= 1); for submucosa, sum of no pathological change (= 0), mononuclear cell infiltrate (1 small aggregate = 0, more than one aggregate = 1, large aggregates plus increased single cells = 2), PMN infiltrate (none = 0, single = 1, aggregates = 2), and edema (mild = 0, moderate = 1, severe = 2).
Calf intestinal loop surgeries. All animal experiments were conducted in accordance with the Guide to the Care and Use of Experimental Animals, provided by the Canadian Council on Animal Care. One month old, male Holstein calves were housed in single isolation cubicles at the Vaccine and Infectious Disease Organization (VIDO) animal facility. All animals were clinically healthy prior to surgery and rectal swabs from all calves were tested for Salmonella prior to experimentation and were found to be negative in all cases. The calves were fasted for 24 h prior to surgery and then premedicated with butorphanol (0.2 mg/kg of body weight) and diazepam (0.1 mg/kg). Anesthesia was induced with 6 to 8 ml of 5% thiopental sodium prior to placement of an endotracheal tube and maintenance with isoflurane. The intestinal ileal ‘loop’ model developed in sheep (13) was adapted for use in one-month old male calves. A laparotomy was performed and the small intestine was exteriorized until six consecutive Peyer's patches were identified. The exposed small intestine was frequently moistened with sterile PBS prewarmed to 37°C. An intestinal segment containing two Peyer's patches separated by interspaces without Peyer's patches was demarcated by intestinal clamps approximately 30 cm proximal and 30 cm distal to the first and last Peyer's patch and then transected. This intestinal segment was flushed twice with 100 ml of warm sterile saline to clean the gut of its contents. The patency of the intestine was restored with end-to-end anastomoses by aligning the mesenteric and antimesenteric borders of the transected intestine and closing with interrupted and continuous sutures. Silk ligatures were tied approximately 8 cm proximal and distal to each Peyer's patch to create 16- to 18-cm isolated segments containing a Peyer's patch, separated by interspaces of various length that lacked Peyer's patches. Three anastamoses were created in each animal, generating three intestinal segments each containing two loops that provided two independent sites for duplicate samples. For each animal, the intestinal segments received (i) sterile saline, (ii) wild-type Salmonella, and (iii) a Salmonella mutant under investigation. The loops were infected with 3 ml of sterile PBS or 3 ml of a bacterial suspension diluted in PBS to contain either 1x 103 or 1 x 106 CFU. Intestinal segments were marked by silk sutures and the succession and size of each internal loop and interspace was recorded. The loops were then replaced into the abdominal cavity and the surgical incision in the abdominal wall was sutured. This surgery was completed in a total of six animals with two calves euthanized at 24 h postsurgery and four calves euthanized at 5 days postsurgery.
Calf intestinal tissue collection, specimen handling and histopathology. Immediately after euthanasia, intestinal loops were exposed from the abdominal cavity and the fluid volume was collected. Samples of lumenal fluid were taken for bacteriologic isolation and intestinal tissues from loops were taken for histopathology. For histopathology of intestines, sections of ileum were fixed in 10% neutral-buffered formalin, embedded in paraffin, sectioned at a 5-μm thickness, and stained with hematoxylin and eosin. A standard histology scoring system similar to previous studies using bovine ileal loops was used. A histological score from 0 to 4 was assigned based on the tissue reactions to the various doses of mutant and wild-type Salmonella as follows: 0, no visible lesions; 1+, mild inflammatory cell infiltration, submucosal edema with villus atrophy; 2+, moderate inflammatory cell infiltration, submucosal edema with villus atrophy; 3+, moderate inflammatory cell infiltration, submucosal edema, necrosis, vascular thrombosis with villus atrophy; 4+, severe inflammatory cell infiltration, submucosal edema, necrosis, vascular thrombosis with villus atrophy.
Statistical analysis. Total pathological scores were compared using Dunn's Multiple Comparison and Kruskal-Wallis nonparametric tests. Bacterial load was compared using analysis of variance with Tukey's multiple-comparisons posttest. All analyses were performed using Graphpad Prism, version 3.0.
RESULTS
Ceca of streptomycin-treated mice infected with SPI-1 and SPI-2 mutants show attenuated pathology at 48 h postinfection. A model of Salmonella-induced colitis using mice pretreated with streptomycin has been developed (1). Using this model we (5) and others (16) reported that SPI-2 contributes to the induction of intestinal inflammation in mice. In the latter study, a SPI-1 apparatus mutant was compared to a SPI-2 translocon mutant, however a direct comparison in mice of SPI-1 and SPI-2 apparatus mutants devoid of all type III secretion and translocation activity has not been previously described later than 2 days postinfection. In order to directly compare the relative contribution of SPI-1 and SPI-2 to the full course of murine intestinal inflammation, streptomycin-treated mice were infected with wild-type S. enterica serovar Typhimurium or mutants lacking functional SPI-1, SPI-2, or both SPI-1 and SPI-2 type III secretion systems. Intestinal inflammation was evaluated using previously described histopathology scoring methods (5) and bacterial loads were enumerated in the intestine and spleen. By day 2, significant differences in the severity of histopathology elicited by wild-type, SPI-1, and SPI-2 mutant bacteria were evident (Fig. 1A). Wild-type Salmonella elicited severe and diffuse inflammatory changes in the ceca of infected mice. In contrast, attenuation of inflammatory pathology was apparent in SPI-2 infected ceca, including decreased lumenal inflammatory cell infiltrate and epithelial debris, mild or patchy surface epithelial necrosis and regenerative change, little inflammatory cell recruitment into the mucosa, and diminished submucosal edema (Fig. 2). Inflammatory pathology was virtually absent at 2 days in tissues infected with Salmonella lacking SPI-1 or both SPI-1/SPI-2 type III secretion systems (Fig. 1A). Intestinal colonization by wild-type Salmonella and each of the mutant strains was similar at 2 days and translocation to the spleen was observed at similar levels in all strains under investigation (Fig. 1C and 1D).
Murine ceca infected with either SPI-2 or SPI-1/SPI-2 double mutants display little pathology at 5 days after infection, while ceca infected with SPI-1 mutants show severe pathological changes. Our previous work using streptomycin-pretreated mice demonstrated a role for SPI-2 in an inflammatory phenotype at 5 days after infection, however we did not assess the contribution of SPI-1 to this disease process. In order to compare the relative contributions of SPI-1 and SPI-2 virulence mechanisms during the course of intestinal inflammatory pathogenesis in mice, we infected streptomycin-treated mice orally with wild-type, SPI-1, SPI-2, or SPI-1/SPI-2 mutant S. enterica serovar Typhimurium and assessed intestinal disease at day 5. In contrast to the lack of intestinal inflammation elicited by SPI-1 mutants at 2 days after infection, significant histopathological changes were apparent in the ceca of mice infected with either wild-type or SPI-1 mutant Salmonella at day 5 after infection (Fig. 1B). SPI-1 mutants of S. enterica serovar Typhimurium that retained a functional SPI-2 type III secretion system elicited severe and diffuse inflammatory changes in the ceca of infected mice. In contrast, in mouse tissues infected with mutants lacking a functional SPI-2 type III secretion system, or both SPI-1 and SPI-2 type III secretion systems, we observed a near total attenuation of inflammatory pathology including little or no lumenal inflammatory cell infiltrate and epithelial debris, normal surface epithelial appearance and architecture, little inflammatory cell recruitment into the mucosa and no submucosal edema (Fig. 2). The presence of SPI-1 was not sufficient to induce an inflammatory response at this time in the absence of SPI-2. These data indicate that although SPI-1 is important for the induction of inflammatory responses in the gut of Salmonella infected mice early in infection, SPI-2 appears to play the dominant role in intestinal inflammatory pathogenesis in mice and is required for intestinal disease irrespective of the presence of SPI-1 encoded virulence mechanisms.
Calf intestinal loops infected with SPI-1 and SPI-2 mutants show reduced secretory response and similar pathological changes at 24 h postinfection. Using a 12-h calf ileal loop model with S. enterica serovar Dublin, Bispham and colleagues (2) reported that an sseD mutant (a component of the SPI-2 translocation machinery that probably introduces a pore in host cell plasma membranes) induced a weaker secretory and inflammatory response than the parental wild-type strain. Unlike serovar Dublin, which progresses to a systemic infection in calves, serovar Typhimurium produces an enteric infection in the small intestine that rarely progresses beyond the cow gut (33). Because the maximum infection time achievable with the existing bovine ileal loop model is 12 h, we developed an alternative model that overcomes this time constraint, and examined intestinal pathology at 24 h and 5 days following infection with wild-type S. enterica serovar Typhimurium, or SPI-1 and SPI-2 mutants. Ileal loops infected with wild-type Salmonella, the SPI-1 apparatus mutant (invA) or the SPI-2 apparatus mutant (ssaR) were examined for secretory response into the loop and for pathological changes at 24 h after infection. At 24 h postinoculation, intestinal fluid accumulation was higher in loops infected with wild-type Salmonella (Fig. 3A). This secretory response was consistent with other studies that measured fluid accumulation in ileal loops at 12 h postinfection (2). Loops infected with either SPI-1 or SPI-2 mutants showed a reduction in the secretory response at 24 h compared to the response in loops infected with wild-type Salmonella (Fig. 3A). At 24 h after infection, there was little difference in the degree of colonization between wild-type Salmonella and the SPI-1 or SPI-2 mutants. However, by 5 days after infection, a role for SPI-2, but not SPI-1, in intestinal colonization was apparent (Fig. 3B). These colonization data at day 2 and day 5 after infection parallel those seen in the colon of streptomycin-pretreated mice. Mucosal necrosis and edema with acute inflammation of the intestinal loops due to inoculation with wild-type Salmonella or the ssaR or invA mutants was similar at 24 h after infection (Fig. 4A, Fig. 5). As another measure of pathological change in the epithelium, we measured the villus height in several tissue sections of intestinal epithelium from loops infected with wild-type Salmonella and each of the SPI-1 and SPI-2 mutants. Villus height correlated with the pathological score and demonstrated more severe villus atrophy in tissues with greater pathological changes (Fig. 4C). At 24 h after infection, there was little correlation between the degree of pathological change and the initial infectious dose of Salmonella.
Calf ileal loops infected with SPI-2 mutants display little pathology at 5 days after infection. Previous calf infection models employing ligated ileal segments to examine pathology during Salmonella infection have been limited to terminally anesthetized animals at 8 h (34) or 12 h after infection (2, 21). To date, no studies have examined pathological changes in Salmonella-infected ileal loops after these time points. In order to examine whether the natural progression of intestinal disease in calves following infection with various Salmonella mutants follows the same course as that observed in mice, we infected bovine ileal loops for 5 days and examined intestinal pathology. After 5 days of infection, tissue damage and inflammation in the intestinal loops infected with wild-type Salmonella was severe and demonstrated tissue necrosis, submucosal edema and fluid accumulation into the intestinal lumen (Fig. 4B and 5). In contrast, intestinal tissue infected with SPI-2 mutant Salmonella had mild to moderate inflammation with mild villus atrophy, and by day 5 following infection, began to show signs of tissue regeneration and restoration (Fig. 5). At 5 days after infection, pathological scores generally correlated with the initial infectious dose of Salmonella, where loops infected with the lower infectious dose (1 x 103 CFU) had lower pathological scores than loops infected with the higher infectious dose (1 x 106 CFU). No pathological features of disease were observed in control loops inoculated with saline (Fig. 4B) indicating that the surgical procedure did not account for the observed changes.
SPI-1 mutants produce severe intestinal pathology in calf ileal loops at 5 days postinfection. Previous studies employing the 8 to 12 h infection model of calf ileal loops have established a role for SPI-1 in the early stages of intestinal pathology (26, 34). In order to examine whether the pathology at later times after infection was also SPI-1-dependent, we infected calf ileal loops for 5 days prior to examination of intestinal pathology. As observed previously, tissue damage and inflammation in the intestinal loops infected with wild-type Salmonella was severe, demonstrating tissue necrosis, submucosal edema and fluid accumulation in the intestinal lumen (Fig. 4B). Similarly, intestinal tissues infected with SPI-1 mutant Salmonella displayed severe pathological lesions characterized by severe inflammatory cell infiltration, submucosal edema, necrosis, and vascular thrombosis with villus atrophy at 5 days after infection (Fig. 4B and 5). Loops inoculated with saline or with SPI-2 mutants (described above) did not show these pathological changes indicating that neither the surgical procedure nor the presence of Salmonella per se in the intestinal loops contributed to this associated pathology. Loops infected with SPI-1 mutants also showed a greater degree of villus atrophy compared to loops infected with SPI-2 mutants (Fig. 4D). These data indicate that the later progression of bovine enteric disease induced by S. enterica serovar Typhimurium is dominated by the SPI-2, and not the SPI-1, type III secretion system.
DISCUSSION
As the role for SPI-2 in Salmonella enteropathogenesis emerges, it is of considerable importance to explore how this virulence system imparts its function in multiple models of Salmonella enterocolitis. The bovine ligated ileal loop model is a proven tool for studying enteropathogenesis of Salmonella serotypes (28, 33). Combined with the murine model of colitis, these models provide significant investigative power to study the role of SPI-1 and SPI-2 in intestinal salmonellosis. While bovine loop infection has previously been limited to studying the host-pathogen interactions between Salmonella and host tissues at early times (<12 h) after infection (34) due to the obstruction of the small bowel with multiple ligatures, we have developed a calf ileal loop model that can be used to explore infections of longer duration by restoring the patency of the small intestine following surgery. This model incorporates the advantages of ligated ileal loops such as (i) delivery of a synchronized and metered dose of infectious organism to a known site, (ii) the ability to create multiple experimental sites in one animal for positive and negative controls plus an experimental variable under study, (iii) a reduction in the number of experimental animals required since oral infections can accommodate only one strain in each animal and are often more variable due to heterogeneity of the outbred animal population and variability in the infectious dose reaching the intestine, and (iv) the ability to monitor pathological changes in intestinal segments containing a fully intact blood supply and lymphatic system. Importantly, our model overcomes the time constraints imposed by previous ligated ileal loop methods and is fully compatible with infection experiments of long duration in which it is desirable to monitor pathological changes within the advantageous context described above.
By using the murine model of Salmonella-induced colitis and by extending the duration of bovine ileal loop infections we have assessed the contribution of two major virulence factors, the SPI-1 and SPI-2 type III secretion systems, to intestinal disease progression. The 5-day ileal loop model confirmed a critical role for SPI-2 in the pathogenesis of Salmonella enterocolitis and revealed a previously unrecognized pathological phenotype for SPI-1 mutants in the intestine of cows. In this model SPI-1 mutants retain the ability to induce severe inflammation and intestinal pathology similar to that induced by infection with wild-type Salmonella. This pathology is dependent on a functional SPI-2 apparatus since SPI-2 mutants produced only a mild inflammatory response at day 5 after infection of ileal loops.
It has been reported previously using the mouse streptomycin-pretreatment model that the SPI-1 effectors, SipA, SopE and SopE2, are required for Salmonella-induced intestinal pathology at day 2 after oral infection (15). Indeed, in our studies using the similar model, we observed that Salmonella mutants lacking a functional SPI-1 type III secretion system are significantly attenuated for the induction of colitis after 2 days of infection. However in the former study, time points longer than 2 days were not examined for intestinal pathology. Here, we have extended these results by showing that during later stages of the disease process, mutants lacking a SPI-1 type III secretion system but retaining a functional SPI-2 secretion system produce severe intestinal pathology, similar to that induced by wild-type Salmonella, at 5 days after oral infection. We confirmed the requirement of SPI-2 for this pathology by using single and double mutants defective for both SPI-1 and SPI-2-mediated type III secretion, which had a similar phenotype in the 5-day murine model more resembling that of a single SPI-2 mutant. It has been reported previously using the mouse model of Salmonella colitis that SPI-1 mutants produced a delayed colitis at day 4 after oral infection, but that this disease was still significantly attenuated in magnitude compared to colitis induced by wild-type Salmonella (16). Our results using this mouse model showed that SPI-2 mediates significant intestinal inflammation in the absence of SPI-1 at 5 days after oral infection. While SPI-2 mutants were significantly attenuated for induction of cecal pathology at day 2 and day 5 postinfection, SPI-1 mutants were not significantly attenuated in their ability to induce intestinal pathology at day 5 after infection. One difference between our study and that reported previously using streptomycin-pretreated mice (16) is the class of SPI-2 type III mutant used. While we used a type III secretion system apparatus mutant (ssaR), incapable of secreting SPI-2 effectors, Hapfelmeier and colleagues used a translocon mutant (sseD), incapable of translocating protein into the host cell during intracellular infection but still potentially able to secrete effectors out of the bacteria. The observation that SPI-2 is actively expressed in the lumen of the gut (N. Brown and B. Finlay, unpublished observations) supports the involvement of SPI-2-mediated type III secretion in the extracellular phenotype we observed. One other difference includes the sex of mice infected, as our studies exclusively used males, which appear less susceptible to inflammation than females (B. Coburn and B. Finlay, unpublished data). Together, these data suggest that the pathological basis of enteric Salmonella disease at later stages of the infection process is more dependent on the SPI-2 type III secretion system and its associated effectors rather than the contribution of the SPI-1 type III secretion system. This notion was supported in the bovine ileal loop model where SPI-2 mediated significant intestinal inflammation in the absence of SPI-1 at 5 days after infection.
One of the important findings of this work was the concordance between the mouse model of infectious colitis and the 5-day ileal loop model in calves for discriminating the phenotypes associated with major Salmonella virulence factors. A similar requirement for SPI-2 in the progression of intestinal inflammation was found in both animal models despite noteworthy differences between murine and bovine infections. Such differences include (i) the presence of co-occurring systemic infection in mice and the absence of this systemic component in cows, (ii) the induction of inflammation in the small bowel in bovine infection and in the large bowel in murine infection, and (iii) the necessity to treat mice with antibiotics in order to induce intestinal disease, without any such requirement in cows. The dependence on SPI-2 for the progression of intestinal disease in an infection model largely confined to the intestine (bovine) is noteworthy, as the superimposed systemic disease in streptomycin treated mice can confound the interpretation of intestinal phenotypes. A similar progression of S. enterica serovar Dublin infection of cattle, in which the highly invasive bacteria disseminate to systemic sites of infection, can also encumber the interpretation of enteric phenotypes of certain bacterial mutants under study. Although the microenvironment and physiology of a ligated intestinal loop and a patent gut likely differ, the use of ligated loops affords desirable experimental conditions and endpoints for the study of S. Typhimurium pathogenesis. Likewise, the early colonization by Salmonella of the murine intestine following streptomycin treatment requires careful interpretation due to the absence of normal microbiota. These caveats notwithstanding, we suggest that the streptomycin-treated mouse model and the extended duration bovine ileal loop model presented here can provide complementary approaches to the study of Salmonella enteric disease. These new models provide a framework to further investigate the SPI-2 effectors that contribute to inflammatory pathology and persistence during the intestinal phase of salmonellosis in cows and mice.
ADDENDUM IN PROOF
A recently published report (L. A. Knodler, A. Bestor, C. Ma, I. Hansen-Wester, M. Hensel, B. A. Vallance, and O. Steele-Mortimer, Infect. Immun. 73:7027-7031, 2005) has demonstrated that the presence of a fluorescent protein-bearing plasmid can significantly attenuate the virulence of Salmonella enterica serovar Typhimurium in cultured cells and susceptible mice. The SPI-2 mutant strain used in the mouse experiments in our study bears a green fluorescent protein-expressing plasmid. This strain was used for the mouse experiments only; plasmid-free strains were used in bovine ligated intestinal loop experiments. The strains are otherwise identical. Since our original submission, we have directly compared these strains and found that the presence of this plasmid increases the attenuation of the SPI-2 mutant strain in the murine enterocolitis model of infection at 5 days but not at 2 days. At the later timepoint, the attenuation of intestinal virulence and colonic bacterial load in the SPI-2 mutant strain not bearing the plasmid is less severe than that in the strain (bearing the plasmid) used in our study. However, the attenuation of the plasmid-free strain is still significant.
ACKNOWLEDGMENTS
We thank the Finlay laboratory for helpful discussions and insightful comments. We also gratefully acknowledge the animal care staff at VIDO for animal husbandry assistance and Gordon Crockford for technical assistance.
Funding for this work was provided by grants to B.B.F. from the Canadian Institutes of Health Research (CIHR) and the Howard Hughes Medical Institute (HHMI) and to A.A.P. from CIHR, the Canadian Bacterial Disease Network and the Natural Sciences and Engineering Research Council of Canada (NSERC). B.K.C. is the recipient of postdoctoral fellowships from the CIHR and Michael Smith Foundation for Health Research (MSFHR), and B.A.C. is the recipient of postgraduate scholarships from the CIHR and MSFHR. B.B.F. is a CIHR Distinguished Investigator, an HHMI International Research Scholar, and the University of British Columbia Peter Wall Distinguished Professor. A.A.P. holds an NSERC Senior Industrial Research Chair position.
B.K.C. and B.A.C. contributed equally to this work.
REFERENCES
1. Barthel, M., S. Hapfelmeier, L. Quintanilla-Martinez, M. Kremer, M. Rohde, M. Hogardt, K. Pfeffer, H. Russmann, and W. D. Hardt. 2003. Pretreatment of mice with streptomycin provides a Salmonella enterica serovar Typhimurium colitis model that allows analysis of both pathogen and host. Infect. Immun. 71:2839-2858.
2. Bispham, J., B. N. Tripathi, P. R. Watson, and T. S. Wallis. 2001. Salmonella pathogenicity island 2 influences both systemic salmonellosis and Salmonella-induced enteritis in calves. Infect. Immun. 69:367-377.
3. Brumell, J. H., C. M. Rosenberger, G. T. Gotto, S. L. Marcus, and B. B. Finlay. 2001. SifA permits survival and replication of Salmonella typhimurium in murine macrophages. Cell. Microbiol. 3:75-84.
4. Cirillo, D. M., R. H. Valdivia, D. M. Monack, and S. Falkow. 1998. Macrophage-dependent induction of the Salmonella pathogenicity island 2 type III secretion system and its role in intracellular survival. Mol. Microbiol. 30:175-188.
5. Coburn, B., Y. Li, D. Owen, B. A. Vallance, and B. B. Finlay. 2005. Salmonella enterica serovar Typhimurium pathogenicity island 2 is necessary for complete virulence in a mouse model of infectious enterocolitis. Infect. Immun. 73:3219-3227.
6. Coombes, B. K., M. E. Wickham, N. F. Brown, S. Lemire, L. Bossi, W. W. Hsiao, F. S. Brinkman, and B. B. Finlay. 2005. Genetic and molecular analysis of GogB, a phage-encoded type III-secreted substrate in Salmonella enterica serovar typhimurium with autonomous expression from its associated phage. J. Mol. Biol. 348:817-830.
7. Everest, P., J. Ketley, S. Hardy, G. Douce, S. Khan, J. Shea, D. Holden, D. Maskell, and G. Dougan. 1999. Evaluation of Salmonella typhimurium mutants in a model of experimental gastroenteritis. Infect. Immun. 67:2815-2821.
8. Galan, J. E. 1999. Interaction of Salmonella with host cells through the centisome 63 type III secretion system. Curr. Opin. Microbiol. 2:46-50.
9. Galan, J. E. 1996. Molecular genetic bases of Salmonella entry into host cells. Mol. Microbiol. 20:263-271.
10. Galan, J. E. 2001. Salmonella interactions with host cells: type III secretion at work. Annu. Rev. Cell Dev. Biol. 17:53-86.
11. Galan, J. E., and R. Curtiss III. 1989. Cloning and molecular characterisation of genes whose products allow Salmonella typhimurium to penetrate tissue culture cells. Proc. Natl. Acad. Sci. USA 86:6383-6387.
12. Galyov, E. E., M. W. Wood, R. Rosqvist, P. B. Mullan, P. R. Watson, S. Hedges, and T. S. Wallis. 1997. A secreted effector protein of Salmonella dublin is translocated into eukaryotic cells and mediates inflammation and fluid secretion in infected ileal mucosa. Mol. Microbiol. 25:903-912.
13. Gerdts, V., R. R. Uwiera, G. K. Mutwiri, D. J. Wilson, T. Bowersock, A. Kidane, L. A. Babiuk, and P. J. Griebel. 2001. Multiple intestinal ‘loops’ provide an in vivo model to analyse multiple mucosal immune responses. J. Immunol. Methods 256:19-33.
14. Ghosh, P. 2004. Process of protein transport by the type III secretion system. Microbiol. Mol. Biol. Rev. 68:771-795.
15. Hapfelmeier, S., K. Ehrbar, B. Stecher, M. Barthel, M. Kremer, and W. D. Hardt. 2004. Role of the Salmonella pathogenicity island 1 effector proteins SipA, SopB, SopE, and SopE2 in Salmonella enterica subspecies 1 serovar Typhimurium colitis in streptomycin-pretreated mice. Infect. Immun. 72:795-809.
16. Hapfelmeier, S., B. Stecher, M. Barthel, M. Kremer, A. J. Muller, M. Heikenwalder, T. Stallmach, M. Hensel, K. Pfeffer, S. Akira, and W. D. Hardt. 2005. The Salmonella pathogenicity island (SPI)-2 and SPI-1 type III secretion systems allow Salmonella serovar typhimurium to trigger colitis via MyD88-dependent and MyD88-independent mechanisms. J. Immunol. 174:1675-1685.
17. Hensel, M., J. E. Shea, S. R. Waterman, R. Mundy, T. Nikolaus, G. Banks, A. Vazquez-Torres, C. Gleeson, F. C. Fang, and D. W. Holden. 1998. Genes encoding putative effector proteins of the type III secretion system of Salmonella pathogenicity island 2 are required for bacterial virulence and proliferation in macrophages. Mol. Microbiol. 30:163-174.
18. Hueck, C. J. 1998. Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol. Mol. Biol. Rev. 62:379-433.
19. Jones, M. A., M. W. Wood, P. B. Mullan, P. R. Watson, T. S. Wallis, and E. E. Galyov. 1998. Secreted effector proteins of Salmonella dublin act in concert to induce enteritis. Infect. Immun. 66:5799-5804.
20. Ochman, H., F. C. Soncini, F. Solomon, and E. A. Groisman. 1996. Identification of a pathogenicity island required for Salmonella survival in host cells. Proc. Natl. Acad. Sci. USA 93:7800-7804.
21. Reis, B. P., S. Zhang, R. M. Tsolis, A. J. Baumler, L. G. Adams, and R. L. Santos. 2003. The attenuated sopB mutant of Salmonella enterica serovar Typhimurium has the same tissue distribution and host chemokine response as the wild type in bovine Peyer's patches. Vet. Microbiol. 97:269-277.
22. Santos, R. L., S. Zhang, R. M. Tsolis, R. A. Kingsley, L. G. Adams, and A. J. Baumler. 2001. Animal models of Salmonella infections: enteritis versus typhoid fever. Microbes Infect. 3:1335-1344.
23. Shea, J. E., C. R. Beuzon, C. Gleeson, R. Mundy, and D. W. Holden. 1999. Influence of the Salmonella typhimurium pathogenicity island 2 type III secretion system on bacterial growth in the mouse. Infect. Immun. 67:213-219.
24. Shea, J. E., M. Hensel, C. Gleeson, and D. W. Holden. 1996. Identification of a virulence locus encoding a second type III secretion system in Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 93:2593-2597.
25. Stecher, B., S. Hapfelmeier, C. Muller, M. Kremer, T. Stallmach, and W. D. Hardt. 2004. Flagella and chemotaxis are required for efficient induction of Salmonella enterica serovar Typhimurium colitis in streptomycin-pretreated mice. Infect. Immun. 72:4138-4150.
26. Tsolis, R. M., L. G. Adams, T. A. Ficht, and A. J. Baumler. 1999. Contribution of Salmonella typhimurium virulence factors to diarrheal disease in calves. Infect. Immun. 67:4879-4885.
27. Tsolis, R. M., R. A. Kingsley, S. M. Townsend, T. A. Ficht, L. G. Adams, and A. J. Baumler. 1999. Of mice, calves, and men. Comparison of the mouse typhoid model with other Salmonella infections. Adv. Exp. Med. Biol. 473:261-274.
28. Wallis, T. S., and E. E. Galyov. 2000. Molecular basis of Salmonella-induced enteritis. Mol. Microbiol. 36:997-1005.
29. Wallis, T. S., M. Wood, P. Watson, S. Paulin, M. Jones, and E. Galyov. 1999. Sips, Sops, and SPIs but not stn influence Salmonella enteropathogenesis. Adv. Exp. Med. Biol. 473:275-280.
30. Waterman, S. R., and D. W. Holden. 2003. Functions and effectors of the Salmonella pathogenicity island 2 type III secretion system. Cell. Microbiol. 5:501-511.
31. Wood, M. W., M. A. Jones, P. R. Watson, S. Hedges, T. S. Wallis, and E. E. Galyov. 1998. Identification of a pathogenicity island required for Salmonella enteropathogenicity. Mol. Microbiol. 29:883-891.
32. Wray, C., and W. J. Sojka. 1978. Experimental Salmonella typhimurium infection in calves. Res. Vet. Sci. 25:139-143.
33. Zhang, S., R. A. Kingsley, R. L. Santos, H. Andrews-Polymenis, M. Raffatellu, J. Figueiredo, J. Nunes, R. M. Tsolis, L. G. Adams, and A. J. Baumler. 2003. Molecular pathogenesis of Salmonella enterica serotype typhimurium-induced diarrhea. Infect. Immun. 71:1-12.
34. Zhang, S., R. L. Santos, R. M. Tsolis, S. Stender, W. D. Hardt, A. J. Baumler, and L. G. Adams. 2002. The Salmonella enterica serotype typhimurium effector proteins SipA, SopA, SopB, SopD, and SopE2 act in concert to induce diarrhea in calves. Infect. Immun. 70:3843-3855.(Brian K. Coombes, Bryan A)
Department of Microbiology and Immunology
University of British Columbia, Vancouver, British Columbia V6T 1Z4, Vaccine and Infectious Disease Organization, University of Saskatchewan, Saskatoon, Saskatchewan S7N 5B4
Department of Veterinary Pathology, Western College of Veterinary Medicine, University of Saskatchewan, Saskatoon, Saskatchewan, Canada
ABSTRACT
We have developed a novel ileal loop model for use in calves to analyze the contribution of Salmonella enterica serovar Typhimurium type III secretion systems to disease processes in vivo. Our model involves constructing ileal loops with end-to-end anastamoses to restore the patency of the small intestine, thereby allowing experimental animals to convalesce following surgery for the desired number of days. This model overcomes the time constraint imposed by ligated ileal loop models that have precluded investigation of Salmonella virulence factors during later stages of the infection process. Here, we have used this model to examine the enteric disease process at 24 h and 5 days following infection with wild-type Salmonella and mutants lacking the virulence-associated Salmonella pathogenicity island 1 (SPI-1) or SPI-2 type III secretion systems. We show that SPI-2 mutants are dramatically attenuated at 5 days following infection and report a new phenotype for SPI-1 mutants, which induce intestinal pathology in calves similar to wild-type Salmonella in the 5-day ileal loop model. Both of these temporal phenotypes for SPI-1 and SPI-2 mutants were corroborated in a second animal model of enteric disease using streptomycin-pretreated mice. These data delineate novel phenotypes for SPI-1 and SPI-2 mutants in the intestinal phase of bovine and murine salmonellosis and provide working models to further investigate the effector contribution to these pathologies.
INTRODUCTION
Salmonella enterica are facultative intracellular enteric bacteria that cause a range of clinically important infections in humans and commercial livestock. Within the S. enterica species, there are numerous serovars, discriminated by their lipopolysaccharide and flagellar and capsular antigens. The natural progression of Salmonella disease is strongly influenced by both the S. enterica serovar and the host. For example, S. enterica serovar Typhimurium is a pathogen with a wide host range, infecting humans, cattle, mice, and chickens. In humans and cattle, infection with S. enterica serovar Typhimurium manifests as enterocolitis that rarely spreads to systemic organs. In susceptible mice, however, serovar Typhimurium efficiently crosses the gut epithelium and colonizes the spleen and the liver (22, 27). Infection of chickens is mainly a colonization model that rarely manifests in disease.
Two major virulence determinants involved in Salmonella pathogenesis are encoded in large chromosomal pathogenicity islands called Salmonella pathogenicity island 1 (SPI-1) (9) and SPI-2 (20, 24). Both SPI-1 and SPI-2 encode separate type III secretion systems that introduce virulence proteins (called effectors) into the host environment either by translocation directly into host cells or possibly by secretion into the vicinity of host cells (10, 14, 18). Effector proteins translocated by the SPI-1 type III secretion system influence early host cell cytoskeletal and membrane rearrangements involved in bacterial uptake into target cells (8) whereas SPI-2 is generally thought to play a role during intracellular infection by allowing the formation of Salmonella replicative vacuoles and evading host cell defenses (30).
Animal models are available to study both intestinal and systemic phases of salmonellosis caused by S. enterica serovar Typhimurium. Cattle have been used to study the enteric disease caused by this organism, whereas infected mice develop a systemic infection that shares features with human typhoid. S. enterica serovar Dublin is also commonly used in bovine infection models because this particular serotype is able to initiate both intestinal and systemic phases of infection in cows.
An important theme to emerge from these models is the seemingly dichotomous role played by the SPI-1 and SPI-2 type III secretion systems during the intestinal and systemic phases of salmonellosis. Whereas the SPI-1-encoded type III secretion system plays an essential role in colonization of the bovine intestine and in bovine enteropathogenesis (12, 19, 29, 31), this virulence trait has been reported to play little to no role in systemic infection (11). Conversely, the SPI-2-encoded type III secretion system is more strongly associated with systemic virulence and associated pathology (4, 17, 20, 23) than intestinal disease. The role of the SPI-2-encoded type III secretion system in enteric disease has been less studied and appears more contentious. Earlier work demonstrated that SPI-2 mutants maintain intestinal virulence during oral infection of cows (26) and in ileal loops in rabbits (7). However, one study using bovine ileal loops infected with S. enterica serovar Dublin (2) indicated that mutations in SPI-2 reduced the intestinal secretory response compared to wild-type serovar Dublin. Another study using oral infections of calves with a bovine isolate of serovar Typhimurium (26) indicated that SPI-2 mutants produced intestinal lesions of reduced severity in calves but nevertheless caused mortality and acute diarrhea. One of the confounding factors for the study of enteric disease in cattle caused by S. enterica serovar Dublin is the highly invasive nature of the infection that results in a disseminating bacteremia (33). In contrast, natural and experimental infection of calves with S. enterica serovar Typhimurium results in an enteric disease localized to the gut without systemic involvement, which shares similar pathological and clinical features to human enterocolitis caused by this same organism (33).
Recently, a mouse model of Salmonella-induced colitis has been developed that relies on the pretreatment of animals with streptomycin (1). This model has been used to demonstrate that Salmonella-induced colitis at day 2 after infection requires the SPI-1 effectors SipA, SopE and SopE2 (15), plus flagella and chemotaxis (25). Using this model, we (5) and others (16) have extended these findings by showing that the SPI-2 type III secretion system contributes to the intestinal inflammatory phenotype in the cecum and proximal colon between day 2 and day 5 following oral infection of mice with S. enterica serovar Typhimurium. Although these studies have identified a role for SPI-2 in the ceca of susceptible mice after 2 days of infection, as yet these data have not been extended to the bovine ileal loop infection model at the same time points.
The ligated ileal loop model in cattle is a proven tool for studying enteropathogenesis of Salmonella serotypes (28, 33). However, due to obstruction of the small bowel with multiple ligatures, one recognized limitation of this model in its present usage is the inability to monitor pathological changes in response to Salmonella at times longer than 12 h postinfection (34). This has precluded the investigation of the relative contribution of SPI-1 and SPI-2 to the progression of intestinal disease in bovine loops and the comparison to other intestinal inflammatory models of serovar Typhimurium infection. To address this limitation, we have developed a novel calf ileal loop model that restores the patency of the small intestine following surgery. This feature overcomes the time constraints of traditional loop experiments, permitting the examination of virulence determinants at later stages of enteric disease in a highly controlled environmental context. Here, we establish the relative contributions of S. enterica serovar Typhimurium SPI-1 and SPI-2 type III secretion to early and late intestinal inflammation. Our results identify novel phenotypes for S. enterica serovar Typhimurium mutants with deficiencies in SPI-1 and SPI-2 type III secretion systems, which we have demonstrated in a murine model of infectious colitis and confirmed in the extended bovine ileal loop model. The working model presented here supports the view that during the progression of enteric disease, the presence of SPI-2 overcomes the previously reported requirement of SPI-1 for intestinal inflammatory disease.
MATERIALS AND METHODS
Bacterial strains and culture conditions. Salmonella enterica serovar Typhimurium strain SL1344 (32) was used as the wild-type strain throughout this study and mutants lacking SPI-1-mediated type III secretion (SB103; invA::Kan) (11) or SPI-2-mediated type III secretion (ssaR [cows]; ssaR bearing pAT113-GFP [mice]) (3) were isogenic derivatives of this strain. An S. enterica serovar Typhimurium double mutant defective for both SPI-1 and SPI-2-mediated type III secretion was constructed by generalized P22 transduction of a marked invA::Kan mutation to a strain with an unmarked, in frame deletion of ssaR (6). For infection of experimental animals, bacterial cultures were grown for 18 h at 37°C with shaking in Luria-Bertani (LB) medium and then diluted to the appropriate concentration in phosphate-buffered saline (PBS).
Mouse infection experiments. Male C57BL/6 mice (6 to 8 weeks old) were purchased from Jackson Laboratories (Maine, USA). Mice were housed in sterilized, filter-top cages under specific pathogen-free conditions at the University of British Columbia Animal Facility. The protocols used here were in direct accordance with animal care guidelines as outlined by the University of British Columbia Animal Care Committee and the Canadian Council on the Use of Laboratory Animals. The streptomycin pretreatment model of murine infectious colitis was used as described previously (1, 5). Mice were deprived of food and water for four hours and then given 20 mg of streptomycin by oral gavage. Two hours after antibiotic treatment, food and water were provided for twenty hours. Four hours prior to infection, food and water were removed once again and then mice were infected orally with 3 x 106 or 3 x 108 bacteria in a 0.1 ml volume. Control mice were given 0.1 ml of sterile LB medium. Water and food were provided ad libitum following infection. Mice were euthanized with CO2 at 2 days and 5 days after infection and tissues were harvested for bacterial enumeration and histopathology.
Bacterial enumeration from mouse tissue. Mouse colon, cecum, spleen, and liver were collected into 1.5 ml of cold sterile PBS and homogenized using a tissue homogenizer (Polytron MR-21; Kinematic). Serial dilutions of the homogenized organs were spread on LB agar plates containing 100 μg/ml streptomycin and incubated overnight at 37°C.
Murine histopathology. Colons, ceca and ilea of experimental animals were fixed in 3% formalin for 18 h followed by 18 h in 70% ethanol and then embedded in paraffin, sectioned at 5 μm and stained with hematoxylin and eosin. For pathological scoring, six fields per sample were examined and scored as described previously (5). This scoring system was as follows: for lumen, sum of empty (score = 0), necrotic epithelial cells (scant = 1, moderate = 2, dense = 3), and polymorphonuclear cells (PMNs) (scant = 2, moderate = 3, dense = 4); for surface epithelium, sum of no pathological change (score = 0), regenerative change (mild = 1, moderate = 2, severe = 3), desquamation (patchy = 1, diffuse = 2), and PMNs in epithelium (score = 1), ulceration (score = 1); for mucosa, sum of no pathological change (score = 0), crypt abscesses (rare [<15%] = 1, moderate [15% to 50%] = 2, abundant [>50%] = 3), presence of mucinous plugs (= 1), and presence of granulation tissue (= 1); for submucosa, sum of no pathological change (= 0), mononuclear cell infiltrate (1 small aggregate = 0, more than one aggregate = 1, large aggregates plus increased single cells = 2), PMN infiltrate (none = 0, single = 1, aggregates = 2), and edema (mild = 0, moderate = 1, severe = 2).
Calf intestinal loop surgeries. All animal experiments were conducted in accordance with the Guide to the Care and Use of Experimental Animals, provided by the Canadian Council on Animal Care. One month old, male Holstein calves were housed in single isolation cubicles at the Vaccine and Infectious Disease Organization (VIDO) animal facility. All animals were clinically healthy prior to surgery and rectal swabs from all calves were tested for Salmonella prior to experimentation and were found to be negative in all cases. The calves were fasted for 24 h prior to surgery and then premedicated with butorphanol (0.2 mg/kg of body weight) and diazepam (0.1 mg/kg). Anesthesia was induced with 6 to 8 ml of 5% thiopental sodium prior to placement of an endotracheal tube and maintenance with isoflurane. The intestinal ileal ‘loop’ model developed in sheep (13) was adapted for use in one-month old male calves. A laparotomy was performed and the small intestine was exteriorized until six consecutive Peyer's patches were identified. The exposed small intestine was frequently moistened with sterile PBS prewarmed to 37°C. An intestinal segment containing two Peyer's patches separated by interspaces without Peyer's patches was demarcated by intestinal clamps approximately 30 cm proximal and 30 cm distal to the first and last Peyer's patch and then transected. This intestinal segment was flushed twice with 100 ml of warm sterile saline to clean the gut of its contents. The patency of the intestine was restored with end-to-end anastomoses by aligning the mesenteric and antimesenteric borders of the transected intestine and closing with interrupted and continuous sutures. Silk ligatures were tied approximately 8 cm proximal and distal to each Peyer's patch to create 16- to 18-cm isolated segments containing a Peyer's patch, separated by interspaces of various length that lacked Peyer's patches. Three anastamoses were created in each animal, generating three intestinal segments each containing two loops that provided two independent sites for duplicate samples. For each animal, the intestinal segments received (i) sterile saline, (ii) wild-type Salmonella, and (iii) a Salmonella mutant under investigation. The loops were infected with 3 ml of sterile PBS or 3 ml of a bacterial suspension diluted in PBS to contain either 1x 103 or 1 x 106 CFU. Intestinal segments were marked by silk sutures and the succession and size of each internal loop and interspace was recorded. The loops were then replaced into the abdominal cavity and the surgical incision in the abdominal wall was sutured. This surgery was completed in a total of six animals with two calves euthanized at 24 h postsurgery and four calves euthanized at 5 days postsurgery.
Calf intestinal tissue collection, specimen handling and histopathology. Immediately after euthanasia, intestinal loops were exposed from the abdominal cavity and the fluid volume was collected. Samples of lumenal fluid were taken for bacteriologic isolation and intestinal tissues from loops were taken for histopathology. For histopathology of intestines, sections of ileum were fixed in 10% neutral-buffered formalin, embedded in paraffin, sectioned at a 5-μm thickness, and stained with hematoxylin and eosin. A standard histology scoring system similar to previous studies using bovine ileal loops was used. A histological score from 0 to 4 was assigned based on the tissue reactions to the various doses of mutant and wild-type Salmonella as follows: 0, no visible lesions; 1+, mild inflammatory cell infiltration, submucosal edema with villus atrophy; 2+, moderate inflammatory cell infiltration, submucosal edema with villus atrophy; 3+, moderate inflammatory cell infiltration, submucosal edema, necrosis, vascular thrombosis with villus atrophy; 4+, severe inflammatory cell infiltration, submucosal edema, necrosis, vascular thrombosis with villus atrophy.
Statistical analysis. Total pathological scores were compared using Dunn's Multiple Comparison and Kruskal-Wallis nonparametric tests. Bacterial load was compared using analysis of variance with Tukey's multiple-comparisons posttest. All analyses were performed using Graphpad Prism, version 3.0.
RESULTS
Ceca of streptomycin-treated mice infected with SPI-1 and SPI-2 mutants show attenuated pathology at 48 h postinfection. A model of Salmonella-induced colitis using mice pretreated with streptomycin has been developed (1). Using this model we (5) and others (16) reported that SPI-2 contributes to the induction of intestinal inflammation in mice. In the latter study, a SPI-1 apparatus mutant was compared to a SPI-2 translocon mutant, however a direct comparison in mice of SPI-1 and SPI-2 apparatus mutants devoid of all type III secretion and translocation activity has not been previously described later than 2 days postinfection. In order to directly compare the relative contribution of SPI-1 and SPI-2 to the full course of murine intestinal inflammation, streptomycin-treated mice were infected with wild-type S. enterica serovar Typhimurium or mutants lacking functional SPI-1, SPI-2, or both SPI-1 and SPI-2 type III secretion systems. Intestinal inflammation was evaluated using previously described histopathology scoring methods (5) and bacterial loads were enumerated in the intestine and spleen. By day 2, significant differences in the severity of histopathology elicited by wild-type, SPI-1, and SPI-2 mutant bacteria were evident (Fig. 1A). Wild-type Salmonella elicited severe and diffuse inflammatory changes in the ceca of infected mice. In contrast, attenuation of inflammatory pathology was apparent in SPI-2 infected ceca, including decreased lumenal inflammatory cell infiltrate and epithelial debris, mild or patchy surface epithelial necrosis and regenerative change, little inflammatory cell recruitment into the mucosa, and diminished submucosal edema (Fig. 2). Inflammatory pathology was virtually absent at 2 days in tissues infected with Salmonella lacking SPI-1 or both SPI-1/SPI-2 type III secretion systems (Fig. 1A). Intestinal colonization by wild-type Salmonella and each of the mutant strains was similar at 2 days and translocation to the spleen was observed at similar levels in all strains under investigation (Fig. 1C and 1D).
Murine ceca infected with either SPI-2 or SPI-1/SPI-2 double mutants display little pathology at 5 days after infection, while ceca infected with SPI-1 mutants show severe pathological changes. Our previous work using streptomycin-pretreated mice demonstrated a role for SPI-2 in an inflammatory phenotype at 5 days after infection, however we did not assess the contribution of SPI-1 to this disease process. In order to compare the relative contributions of SPI-1 and SPI-2 virulence mechanisms during the course of intestinal inflammatory pathogenesis in mice, we infected streptomycin-treated mice orally with wild-type, SPI-1, SPI-2, or SPI-1/SPI-2 mutant S. enterica serovar Typhimurium and assessed intestinal disease at day 5. In contrast to the lack of intestinal inflammation elicited by SPI-1 mutants at 2 days after infection, significant histopathological changes were apparent in the ceca of mice infected with either wild-type or SPI-1 mutant Salmonella at day 5 after infection (Fig. 1B). SPI-1 mutants of S. enterica serovar Typhimurium that retained a functional SPI-2 type III secretion system elicited severe and diffuse inflammatory changes in the ceca of infected mice. In contrast, in mouse tissues infected with mutants lacking a functional SPI-2 type III secretion system, or both SPI-1 and SPI-2 type III secretion systems, we observed a near total attenuation of inflammatory pathology including little or no lumenal inflammatory cell infiltrate and epithelial debris, normal surface epithelial appearance and architecture, little inflammatory cell recruitment into the mucosa and no submucosal edema (Fig. 2). The presence of SPI-1 was not sufficient to induce an inflammatory response at this time in the absence of SPI-2. These data indicate that although SPI-1 is important for the induction of inflammatory responses in the gut of Salmonella infected mice early in infection, SPI-2 appears to play the dominant role in intestinal inflammatory pathogenesis in mice and is required for intestinal disease irrespective of the presence of SPI-1 encoded virulence mechanisms.
Calf intestinal loops infected with SPI-1 and SPI-2 mutants show reduced secretory response and similar pathological changes at 24 h postinfection. Using a 12-h calf ileal loop model with S. enterica serovar Dublin, Bispham and colleagues (2) reported that an sseD mutant (a component of the SPI-2 translocation machinery that probably introduces a pore in host cell plasma membranes) induced a weaker secretory and inflammatory response than the parental wild-type strain. Unlike serovar Dublin, which progresses to a systemic infection in calves, serovar Typhimurium produces an enteric infection in the small intestine that rarely progresses beyond the cow gut (33). Because the maximum infection time achievable with the existing bovine ileal loop model is 12 h, we developed an alternative model that overcomes this time constraint, and examined intestinal pathology at 24 h and 5 days following infection with wild-type S. enterica serovar Typhimurium, or SPI-1 and SPI-2 mutants. Ileal loops infected with wild-type Salmonella, the SPI-1 apparatus mutant (invA) or the SPI-2 apparatus mutant (ssaR) were examined for secretory response into the loop and for pathological changes at 24 h after infection. At 24 h postinoculation, intestinal fluid accumulation was higher in loops infected with wild-type Salmonella (Fig. 3A). This secretory response was consistent with other studies that measured fluid accumulation in ileal loops at 12 h postinfection (2). Loops infected with either SPI-1 or SPI-2 mutants showed a reduction in the secretory response at 24 h compared to the response in loops infected with wild-type Salmonella (Fig. 3A). At 24 h after infection, there was little difference in the degree of colonization between wild-type Salmonella and the SPI-1 or SPI-2 mutants. However, by 5 days after infection, a role for SPI-2, but not SPI-1, in intestinal colonization was apparent (Fig. 3B). These colonization data at day 2 and day 5 after infection parallel those seen in the colon of streptomycin-pretreated mice. Mucosal necrosis and edema with acute inflammation of the intestinal loops due to inoculation with wild-type Salmonella or the ssaR or invA mutants was similar at 24 h after infection (Fig. 4A, Fig. 5). As another measure of pathological change in the epithelium, we measured the villus height in several tissue sections of intestinal epithelium from loops infected with wild-type Salmonella and each of the SPI-1 and SPI-2 mutants. Villus height correlated with the pathological score and demonstrated more severe villus atrophy in tissues with greater pathological changes (Fig. 4C). At 24 h after infection, there was little correlation between the degree of pathological change and the initial infectious dose of Salmonella.
Calf ileal loops infected with SPI-2 mutants display little pathology at 5 days after infection. Previous calf infection models employing ligated ileal segments to examine pathology during Salmonella infection have been limited to terminally anesthetized animals at 8 h (34) or 12 h after infection (2, 21). To date, no studies have examined pathological changes in Salmonella-infected ileal loops after these time points. In order to examine whether the natural progression of intestinal disease in calves following infection with various Salmonella mutants follows the same course as that observed in mice, we infected bovine ileal loops for 5 days and examined intestinal pathology. After 5 days of infection, tissue damage and inflammation in the intestinal loops infected with wild-type Salmonella was severe and demonstrated tissue necrosis, submucosal edema and fluid accumulation into the intestinal lumen (Fig. 4B and 5). In contrast, intestinal tissue infected with SPI-2 mutant Salmonella had mild to moderate inflammation with mild villus atrophy, and by day 5 following infection, began to show signs of tissue regeneration and restoration (Fig. 5). At 5 days after infection, pathological scores generally correlated with the initial infectious dose of Salmonella, where loops infected with the lower infectious dose (1 x 103 CFU) had lower pathological scores than loops infected with the higher infectious dose (1 x 106 CFU). No pathological features of disease were observed in control loops inoculated with saline (Fig. 4B) indicating that the surgical procedure did not account for the observed changes.
SPI-1 mutants produce severe intestinal pathology in calf ileal loops at 5 days postinfection. Previous studies employing the 8 to 12 h infection model of calf ileal loops have established a role for SPI-1 in the early stages of intestinal pathology (26, 34). In order to examine whether the pathology at later times after infection was also SPI-1-dependent, we infected calf ileal loops for 5 days prior to examination of intestinal pathology. As observed previously, tissue damage and inflammation in the intestinal loops infected with wild-type Salmonella was severe, demonstrating tissue necrosis, submucosal edema and fluid accumulation in the intestinal lumen (Fig. 4B). Similarly, intestinal tissues infected with SPI-1 mutant Salmonella displayed severe pathological lesions characterized by severe inflammatory cell infiltration, submucosal edema, necrosis, and vascular thrombosis with villus atrophy at 5 days after infection (Fig. 4B and 5). Loops inoculated with saline or with SPI-2 mutants (described above) did not show these pathological changes indicating that neither the surgical procedure nor the presence of Salmonella per se in the intestinal loops contributed to this associated pathology. Loops infected with SPI-1 mutants also showed a greater degree of villus atrophy compared to loops infected with SPI-2 mutants (Fig. 4D). These data indicate that the later progression of bovine enteric disease induced by S. enterica serovar Typhimurium is dominated by the SPI-2, and not the SPI-1, type III secretion system.
DISCUSSION
As the role for SPI-2 in Salmonella enteropathogenesis emerges, it is of considerable importance to explore how this virulence system imparts its function in multiple models of Salmonella enterocolitis. The bovine ligated ileal loop model is a proven tool for studying enteropathogenesis of Salmonella serotypes (28, 33). Combined with the murine model of colitis, these models provide significant investigative power to study the role of SPI-1 and SPI-2 in intestinal salmonellosis. While bovine loop infection has previously been limited to studying the host-pathogen interactions between Salmonella and host tissues at early times (<12 h) after infection (34) due to the obstruction of the small bowel with multiple ligatures, we have developed a calf ileal loop model that can be used to explore infections of longer duration by restoring the patency of the small intestine following surgery. This model incorporates the advantages of ligated ileal loops such as (i) delivery of a synchronized and metered dose of infectious organism to a known site, (ii) the ability to create multiple experimental sites in one animal for positive and negative controls plus an experimental variable under study, (iii) a reduction in the number of experimental animals required since oral infections can accommodate only one strain in each animal and are often more variable due to heterogeneity of the outbred animal population and variability in the infectious dose reaching the intestine, and (iv) the ability to monitor pathological changes in intestinal segments containing a fully intact blood supply and lymphatic system. Importantly, our model overcomes the time constraints imposed by previous ligated ileal loop methods and is fully compatible with infection experiments of long duration in which it is desirable to monitor pathological changes within the advantageous context described above.
By using the murine model of Salmonella-induced colitis and by extending the duration of bovine ileal loop infections we have assessed the contribution of two major virulence factors, the SPI-1 and SPI-2 type III secretion systems, to intestinal disease progression. The 5-day ileal loop model confirmed a critical role for SPI-2 in the pathogenesis of Salmonella enterocolitis and revealed a previously unrecognized pathological phenotype for SPI-1 mutants in the intestine of cows. In this model SPI-1 mutants retain the ability to induce severe inflammation and intestinal pathology similar to that induced by infection with wild-type Salmonella. This pathology is dependent on a functional SPI-2 apparatus since SPI-2 mutants produced only a mild inflammatory response at day 5 after infection of ileal loops.
It has been reported previously using the mouse streptomycin-pretreatment model that the SPI-1 effectors, SipA, SopE and SopE2, are required for Salmonella-induced intestinal pathology at day 2 after oral infection (15). Indeed, in our studies using the similar model, we observed that Salmonella mutants lacking a functional SPI-1 type III secretion system are significantly attenuated for the induction of colitis after 2 days of infection. However in the former study, time points longer than 2 days were not examined for intestinal pathology. Here, we have extended these results by showing that during later stages of the disease process, mutants lacking a SPI-1 type III secretion system but retaining a functional SPI-2 secretion system produce severe intestinal pathology, similar to that induced by wild-type Salmonella, at 5 days after oral infection. We confirmed the requirement of SPI-2 for this pathology by using single and double mutants defective for both SPI-1 and SPI-2-mediated type III secretion, which had a similar phenotype in the 5-day murine model more resembling that of a single SPI-2 mutant. It has been reported previously using the mouse model of Salmonella colitis that SPI-1 mutants produced a delayed colitis at day 4 after oral infection, but that this disease was still significantly attenuated in magnitude compared to colitis induced by wild-type Salmonella (16). Our results using this mouse model showed that SPI-2 mediates significant intestinal inflammation in the absence of SPI-1 at 5 days after oral infection. While SPI-2 mutants were significantly attenuated for induction of cecal pathology at day 2 and day 5 postinfection, SPI-1 mutants were not significantly attenuated in their ability to induce intestinal pathology at day 5 after infection. One difference between our study and that reported previously using streptomycin-pretreated mice (16) is the class of SPI-2 type III mutant used. While we used a type III secretion system apparatus mutant (ssaR), incapable of secreting SPI-2 effectors, Hapfelmeier and colleagues used a translocon mutant (sseD), incapable of translocating protein into the host cell during intracellular infection but still potentially able to secrete effectors out of the bacteria. The observation that SPI-2 is actively expressed in the lumen of the gut (N. Brown and B. Finlay, unpublished observations) supports the involvement of SPI-2-mediated type III secretion in the extracellular phenotype we observed. One other difference includes the sex of mice infected, as our studies exclusively used males, which appear less susceptible to inflammation than females (B. Coburn and B. Finlay, unpublished data). Together, these data suggest that the pathological basis of enteric Salmonella disease at later stages of the infection process is more dependent on the SPI-2 type III secretion system and its associated effectors rather than the contribution of the SPI-1 type III secretion system. This notion was supported in the bovine ileal loop model where SPI-2 mediated significant intestinal inflammation in the absence of SPI-1 at 5 days after infection.
One of the important findings of this work was the concordance between the mouse model of infectious colitis and the 5-day ileal loop model in calves for discriminating the phenotypes associated with major Salmonella virulence factors. A similar requirement for SPI-2 in the progression of intestinal inflammation was found in both animal models despite noteworthy differences between murine and bovine infections. Such differences include (i) the presence of co-occurring systemic infection in mice and the absence of this systemic component in cows, (ii) the induction of inflammation in the small bowel in bovine infection and in the large bowel in murine infection, and (iii) the necessity to treat mice with antibiotics in order to induce intestinal disease, without any such requirement in cows. The dependence on SPI-2 for the progression of intestinal disease in an infection model largely confined to the intestine (bovine) is noteworthy, as the superimposed systemic disease in streptomycin treated mice can confound the interpretation of intestinal phenotypes. A similar progression of S. enterica serovar Dublin infection of cattle, in which the highly invasive bacteria disseminate to systemic sites of infection, can also encumber the interpretation of enteric phenotypes of certain bacterial mutants under study. Although the microenvironment and physiology of a ligated intestinal loop and a patent gut likely differ, the use of ligated loops affords desirable experimental conditions and endpoints for the study of S. Typhimurium pathogenesis. Likewise, the early colonization by Salmonella of the murine intestine following streptomycin treatment requires careful interpretation due to the absence of normal microbiota. These caveats notwithstanding, we suggest that the streptomycin-treated mouse model and the extended duration bovine ileal loop model presented here can provide complementary approaches to the study of Salmonella enteric disease. These new models provide a framework to further investigate the SPI-2 effectors that contribute to inflammatory pathology and persistence during the intestinal phase of salmonellosis in cows and mice.
ADDENDUM IN PROOF
A recently published report (L. A. Knodler, A. Bestor, C. Ma, I. Hansen-Wester, M. Hensel, B. A. Vallance, and O. Steele-Mortimer, Infect. Immun. 73:7027-7031, 2005) has demonstrated that the presence of a fluorescent protein-bearing plasmid can significantly attenuate the virulence of Salmonella enterica serovar Typhimurium in cultured cells and susceptible mice. The SPI-2 mutant strain used in the mouse experiments in our study bears a green fluorescent protein-expressing plasmid. This strain was used for the mouse experiments only; plasmid-free strains were used in bovine ligated intestinal loop experiments. The strains are otherwise identical. Since our original submission, we have directly compared these strains and found that the presence of this plasmid increases the attenuation of the SPI-2 mutant strain in the murine enterocolitis model of infection at 5 days but not at 2 days. At the later timepoint, the attenuation of intestinal virulence and colonic bacterial load in the SPI-2 mutant strain not bearing the plasmid is less severe than that in the strain (bearing the plasmid) used in our study. However, the attenuation of the plasmid-free strain is still significant.
ACKNOWLEDGMENTS
We thank the Finlay laboratory for helpful discussions and insightful comments. We also gratefully acknowledge the animal care staff at VIDO for animal husbandry assistance and Gordon Crockford for technical assistance.
Funding for this work was provided by grants to B.B.F. from the Canadian Institutes of Health Research (CIHR) and the Howard Hughes Medical Institute (HHMI) and to A.A.P. from CIHR, the Canadian Bacterial Disease Network and the Natural Sciences and Engineering Research Council of Canada (NSERC). B.K.C. is the recipient of postdoctoral fellowships from the CIHR and Michael Smith Foundation for Health Research (MSFHR), and B.A.C. is the recipient of postgraduate scholarships from the CIHR and MSFHR. B.B.F. is a CIHR Distinguished Investigator, an HHMI International Research Scholar, and the University of British Columbia Peter Wall Distinguished Professor. A.A.P. holds an NSERC Senior Industrial Research Chair position.
B.K.C. and B.A.C. contributed equally to this work.
REFERENCES
1. Barthel, M., S. Hapfelmeier, L. Quintanilla-Martinez, M. Kremer, M. Rohde, M. Hogardt, K. Pfeffer, H. Russmann, and W. D. Hardt. 2003. Pretreatment of mice with streptomycin provides a Salmonella enterica serovar Typhimurium colitis model that allows analysis of both pathogen and host. Infect. Immun. 71:2839-2858.
2. Bispham, J., B. N. Tripathi, P. R. Watson, and T. S. Wallis. 2001. Salmonella pathogenicity island 2 influences both systemic salmonellosis and Salmonella-induced enteritis in calves. Infect. Immun. 69:367-377.
3. Brumell, J. H., C. M. Rosenberger, G. T. Gotto, S. L. Marcus, and B. B. Finlay. 2001. SifA permits survival and replication of Salmonella typhimurium in murine macrophages. Cell. Microbiol. 3:75-84.
4. Cirillo, D. M., R. H. Valdivia, D. M. Monack, and S. Falkow. 1998. Macrophage-dependent induction of the Salmonella pathogenicity island 2 type III secretion system and its role in intracellular survival. Mol. Microbiol. 30:175-188.
5. Coburn, B., Y. Li, D. Owen, B. A. Vallance, and B. B. Finlay. 2005. Salmonella enterica serovar Typhimurium pathogenicity island 2 is necessary for complete virulence in a mouse model of infectious enterocolitis. Infect. Immun. 73:3219-3227.
6. Coombes, B. K., M. E. Wickham, N. F. Brown, S. Lemire, L. Bossi, W. W. Hsiao, F. S. Brinkman, and B. B. Finlay. 2005. Genetic and molecular analysis of GogB, a phage-encoded type III-secreted substrate in Salmonella enterica serovar typhimurium with autonomous expression from its associated phage. J. Mol. Biol. 348:817-830.
7. Everest, P., J. Ketley, S. Hardy, G. Douce, S. Khan, J. Shea, D. Holden, D. Maskell, and G. Dougan. 1999. Evaluation of Salmonella typhimurium mutants in a model of experimental gastroenteritis. Infect. Immun. 67:2815-2821.
8. Galan, J. E. 1999. Interaction of Salmonella with host cells through the centisome 63 type III secretion system. Curr. Opin. Microbiol. 2:46-50.
9. Galan, J. E. 1996. Molecular genetic bases of Salmonella entry into host cells. Mol. Microbiol. 20:263-271.
10. Galan, J. E. 2001. Salmonella interactions with host cells: type III secretion at work. Annu. Rev. Cell Dev. Biol. 17:53-86.
11. Galan, J. E., and R. Curtiss III. 1989. Cloning and molecular characterisation of genes whose products allow Salmonella typhimurium to penetrate tissue culture cells. Proc. Natl. Acad. Sci. USA 86:6383-6387.
12. Galyov, E. E., M. W. Wood, R. Rosqvist, P. B. Mullan, P. R. Watson, S. Hedges, and T. S. Wallis. 1997. A secreted effector protein of Salmonella dublin is translocated into eukaryotic cells and mediates inflammation and fluid secretion in infected ileal mucosa. Mol. Microbiol. 25:903-912.
13. Gerdts, V., R. R. Uwiera, G. K. Mutwiri, D. J. Wilson, T. Bowersock, A. Kidane, L. A. Babiuk, and P. J. Griebel. 2001. Multiple intestinal ‘loops’ provide an in vivo model to analyse multiple mucosal immune responses. J. Immunol. Methods 256:19-33.
14. Ghosh, P. 2004. Process of protein transport by the type III secretion system. Microbiol. Mol. Biol. Rev. 68:771-795.
15. Hapfelmeier, S., K. Ehrbar, B. Stecher, M. Barthel, M. Kremer, and W. D. Hardt. 2004. Role of the Salmonella pathogenicity island 1 effector proteins SipA, SopB, SopE, and SopE2 in Salmonella enterica subspecies 1 serovar Typhimurium colitis in streptomycin-pretreated mice. Infect. Immun. 72:795-809.
16. Hapfelmeier, S., B. Stecher, M. Barthel, M. Kremer, A. J. Muller, M. Heikenwalder, T. Stallmach, M. Hensel, K. Pfeffer, S. Akira, and W. D. Hardt. 2005. The Salmonella pathogenicity island (SPI)-2 and SPI-1 type III secretion systems allow Salmonella serovar typhimurium to trigger colitis via MyD88-dependent and MyD88-independent mechanisms. J. Immunol. 174:1675-1685.
17. Hensel, M., J. E. Shea, S. R. Waterman, R. Mundy, T. Nikolaus, G. Banks, A. Vazquez-Torres, C. Gleeson, F. C. Fang, and D. W. Holden. 1998. Genes encoding putative effector proteins of the type III secretion system of Salmonella pathogenicity island 2 are required for bacterial virulence and proliferation in macrophages. Mol. Microbiol. 30:163-174.
18. Hueck, C. J. 1998. Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol. Mol. Biol. Rev. 62:379-433.
19. Jones, M. A., M. W. Wood, P. B. Mullan, P. R. Watson, T. S. Wallis, and E. E. Galyov. 1998. Secreted effector proteins of Salmonella dublin act in concert to induce enteritis. Infect. Immun. 66:5799-5804.
20. Ochman, H., F. C. Soncini, F. Solomon, and E. A. Groisman. 1996. Identification of a pathogenicity island required for Salmonella survival in host cells. Proc. Natl. Acad. Sci. USA 93:7800-7804.
21. Reis, B. P., S. Zhang, R. M. Tsolis, A. J. Baumler, L. G. Adams, and R. L. Santos. 2003. The attenuated sopB mutant of Salmonella enterica serovar Typhimurium has the same tissue distribution and host chemokine response as the wild type in bovine Peyer's patches. Vet. Microbiol. 97:269-277.
22. Santos, R. L., S. Zhang, R. M. Tsolis, R. A. Kingsley, L. G. Adams, and A. J. Baumler. 2001. Animal models of Salmonella infections: enteritis versus typhoid fever. Microbes Infect. 3:1335-1344.
23. Shea, J. E., C. R. Beuzon, C. Gleeson, R. Mundy, and D. W. Holden. 1999. Influence of the Salmonella typhimurium pathogenicity island 2 type III secretion system on bacterial growth in the mouse. Infect. Immun. 67:213-219.
24. Shea, J. E., M. Hensel, C. Gleeson, and D. W. Holden. 1996. Identification of a virulence locus encoding a second type III secretion system in Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 93:2593-2597.
25. Stecher, B., S. Hapfelmeier, C. Muller, M. Kremer, T. Stallmach, and W. D. Hardt. 2004. Flagella and chemotaxis are required for efficient induction of Salmonella enterica serovar Typhimurium colitis in streptomycin-pretreated mice. Infect. Immun. 72:4138-4150.
26. Tsolis, R. M., L. G. Adams, T. A. Ficht, and A. J. Baumler. 1999. Contribution of Salmonella typhimurium virulence factors to diarrheal disease in calves. Infect. Immun. 67:4879-4885.
27. Tsolis, R. M., R. A. Kingsley, S. M. Townsend, T. A. Ficht, L. G. Adams, and A. J. Baumler. 1999. Of mice, calves, and men. Comparison of the mouse typhoid model with other Salmonella infections. Adv. Exp. Med. Biol. 473:261-274.
28. Wallis, T. S., and E. E. Galyov. 2000. Molecular basis of Salmonella-induced enteritis. Mol. Microbiol. 36:997-1005.
29. Wallis, T. S., M. Wood, P. Watson, S. Paulin, M. Jones, and E. Galyov. 1999. Sips, Sops, and SPIs but not stn influence Salmonella enteropathogenesis. Adv. Exp. Med. Biol. 473:275-280.
30. Waterman, S. R., and D. W. Holden. 2003. Functions and effectors of the Salmonella pathogenicity island 2 type III secretion system. Cell. Microbiol. 5:501-511.
31. Wood, M. W., M. A. Jones, P. R. Watson, S. Hedges, T. S. Wallis, and E. E. Galyov. 1998. Identification of a pathogenicity island required for Salmonella enteropathogenicity. Mol. Microbiol. 29:883-891.
32. Wray, C., and W. J. Sojka. 1978. Experimental Salmonella typhimurium infection in calves. Res. Vet. Sci. 25:139-143.
33. Zhang, S., R. A. Kingsley, R. L. Santos, H. Andrews-Polymenis, M. Raffatellu, J. Figueiredo, J. Nunes, R. M. Tsolis, L. G. Adams, and A. J. Baumler. 2003. Molecular pathogenesis of Salmonella enterica serotype typhimurium-induced diarrhea. Infect. Immun. 71:1-12.
34. Zhang, S., R. L. Santos, R. M. Tsolis, S. Stender, W. D. Hardt, A. J. Baumler, and L. G. Adams. 2002. The Salmonella enterica serotype typhimurium effector proteins SipA, SopA, SopB, SopD, and SopE2 act in concert to induce diarrhea in calves. Infect. Immun. 70:3843-3855.(Brian K. Coombes, Bryan A)