The role of specific microbial Ags in the induction of experimental inflammatory bowel disease is poorly understood. Oral infection of susceptible C57BL/6 mice with Toxoplasma gondii results in a lethal ileitis within 7–9 days postinfection. An immunodominant Ag of T. gondii (surface Ag 1 (SAG1)) that induces a robust B and T cell-specific response has been identified and a SAG1-deficient parasite (Δsag1) engineered. We investigated the ability of Δsag1 parasite to induce a lethal intestinal inflammatory response in susceptible mice. C57BL/6 mice orally infected with Δsag1 parasites failed to develop ileitis. In vitro, the mutant parasites replicate in both enterocytes and dendritic cells. In vivo, infection with the mutant parasites was associated with a decrease in the chemokine and cytokine production within several compartments of the gut-associated cell population. RAG-deficient (RAG1−/−) mice are resistant to the development of the ileitis after T. gondii infection. Adoptive transfer of Ag-specific CD4+ effector T lymphocytes isolated from C57BL/6-infected mice into RAG−/− mice conferred susceptibility to the development of the intestinal disease. In contrast, CD4+ effector T lymphocytes from mice infected with the mutant Δsag1 strain failed to transfer the pathology. In addition, resistant mice (BALB/c) that fail to develop ileitis following oral infection with T. gondii were rendered susceptible following intranasal presensitization with the SAG1 protein. This process was associated with a shift toward a Th1 response. These findings demonstrate that a single Ag (SAG1) of T. gondii can elicit a lethal inflammatory process in this experimental model of pathogen-driven ileitis.
The intestinal epithelium provides both a physiologic and immunologic barrier to a range of microorganisms and foreign substances. When an imbalance does occur in the regulation of this response, lethal inflammatory disease may develop (1). Complex interactions between invading pathogen, commensal bacteria, and the host immune response are involved in the development of inflammatory bowel disease (IBD)3 in humans (2). Individuals with Crohn’s disease mount T cell proliferative responses to autologous microbial Ags, suggesting that chronic inflammation of the gastrointestinal mucosa may be sustained by an Ag. Germfree conditions in murine models generally do not favor the development of IBD (3, 4). Similarly, Crohn’s disease and ulcerative colitis in humans respond to antibiotics (5, 6). A number of bacteria including Helicobacter hepaticus (7, 8) Citrobacter rodentium (9), mycobacteria, Campylobacter, and Clostridium (10) can elicit intestinal inflammation. In addition, viruses such as herpes simplex (11) and Epstein-Barr, the primary measles virus (12), can cause IBD-like disease under specific circumstances (13). The role of microbial Ags in influencing the local inflammatory response and the contribution of these Ags to the development of lesions have yet to be well understood.
Oral infection with Toxoplasma gondii in susceptible C57BL/6, but not in the resistant BALB/c mice, leads to a Th1-type acute and lethal ileitis (14, 15). In the absence of genetic or chemical manipulation, this pathogen-driven experimental model develops remarkable similarities to human ileitis (14). Following oral infection, C57BL/6 mice exhibit discontinuous areas of transmural intestinal inflammation, throughout the distal ileum. Histological examination demonstrates mononuclear and polymorphonuclear cell (PMN) infiltrates in the lamina propria (LP), submucosa, and muscle layers. Inflamed small intestinal mucosa exhibits partial villous blunting and hemorrhages. The intestinal pathology induced by T. gondii requires the induction of proinflammatory cytokines such as IFN-γ, TNF-α, and inducible NO synthase (iNOS) as well as the regulatory effects of TGF-β (16). Neutralization of either IFN-γ or CD4+ T cells during oral infection prevents necrosis of the ilea and acute mortality (14, 17).
T. gondii is an obligate intracellular protozoan that infects a wide variety of vertebrate hosts, including humans. Following oral infection of tissue cyst bradyzoites, the parasite transforms through an as of yet unknown mechanism into a rapidly replicating tachyzoite that can then actively invade target cells, beginning the lytic cycle of the parasite. The surface of T. gondii comprises a family of developmentally regulated GPI-linked proteins (surface Ag (SAG)-related sequences), of which SAG1 is the prototypic member (18). SAG1 protein is exclusively expressed on the tachyzoite (19, 20). The biological role for this superfamily of surface proteins remains mostly enigmatic, although there is evidence for a role in parasite attachment. SAG1 induces the dominant Ab response during infection (21, 22) and a strong, systemic Th1-like T cell response characterized by high titer IFN-γ production by CD4 and CD8 T lymphocytes (23).
A SAG1 null mutant was engineered by homologous recombination and used to infect C57BL/6 mice. This mutant was shown in vitro to adhere and to replicate in fibroblasts at the same or even at a better rate than the control parental strain (M. Grigg and J. Boothroyd, manuscript in preparation). In vivo, we demonstrate that this Ag-deficient parasite is not able to induce ileitis following intralumenal infection. Although this mutant can replicate in both the host and in vitro cell culture, infection is associated with a decrease in both innate and adaptive inflammatory immune responses. Host infection with the mutant strain affects the capacity of enterocytes to secrete chemokines, of dendritic cells (DCs) to produce IL-12, and of LP CD4+ T cells to elicit a proinflammatory Th1-like cytokine response. Adoptive transfer of T lymphocytes sensitized with the wild-type parasite, but not with the mutant, confers susceptibility to the infection of RAG−/− mice, normally resistant to the development of the ileitis. SAG1 sensitization of resistant BALB/c mice, followed by Ag re-exposure, results in the development of acute ileitis.
Materials and Methods
Parasites and mouse infection
Tachyzoites from the wild-type (RHΔ) (24), the Δsag1 (lacking the sag1 gene), the sag1-complemented M34 (generously provided by J. Boothroyd and M. Grigg, Stanford, CA), and the SAG3 knockout (KO) strains (generously provided by S. Tomavo, Lille, France) were used for intraintestinal infection. The 76K strain cysts were maintained in the brain of chronically infected mice and were used for the BALB/c mouse infection. All of these Toxoplasma strains are isogenic.
Five- to 7-wk-old female C57BL/6, BALB/c, and RAG-deficient mice (RAG−/−) on a C57BL/6 background (The Jackson Laboratory, Bar Harbor, ME), housed under approved conditions of the animal research facility at Dartmouth Medical School, were used in this study. Infection of C57BL/6 was established by surgery injection of 10,000 tachyzoites directly into the intestines of anesthetized mice. For presensitization procedures, BALB/c mice received intranasally two equivalent doses 28 days apart of 10 μg of purified SAG1, 10 μg of SAG1 plus 0.5 μg of cholera toxin (CT; List Biological Laboratories, Campbell, CA), or 0.5 μg of CT alone. The sensitization efficiency was verified, as described (25). One week after the last boost, BALB/c mice were challenged by gavage with 100 cysts from the 76K strain. RAG−/− mice were challenged with 35 cysts from the 76K strain.
mICcl2 cells (26), derived from C57BL/6 mice, were grown in complete DMEM/Ham’s F12 (1/1, v/v; Invitrogen Life Technologies, Grand Island, NY) and were plated on collagen I-coated plate (2 × 106 cells/well for RNA study and 1 × 105 cells/well for [3H]uracil incorporation).
For cell culture, DCs were purified from the spleen of Fms-like tyrosine kinase 3 ligand-treated mice (27) by a negative selection with anti-CD3-coated magnetic beads, followed by a positive selection with anti-CD11c-coated magnetic beads using MidiMACS columns (Miltenyi Biotec, Auburn, CA). Purity of magnetically sorted CD11c cell population was >95%, as attested by FACScan analysis (BD Biosciences, San Jose, CA). DCs were plated at 1 × 105 cells/well for [3H]uracil incorporation in complete medium.
T. gondii intracellular multiplication
Ten-day differentiated mICcl2 cells (1 × 105 cells/well) and fresh splenic CD11c+ cells (1 × 105 cells/well) were infected with either wild-type (RHΔ) or mutant Δsag1 (at a ratio 1:2 parasite-cell). Two hours later, unbound parasites were removed, and new medium with 0.5 μCi of [3H]uracil (Amersham, Piscataway, NJ) was added. Twenty-four hours later, the radioactivity was counted with a scintillation counter (Beckman Coulter, Fullerton, CA).
Quantitation of parasite DNA by real-time PCR
DNA from tissues was extracted using the Qiamp tissue kit (Qiagen, Chatsworth, CA); amplification of parasite DNA was performed using specific primers for the Toxoplasma B1 gene (5′-GGAACTGCATCCGTTCATGAG-3′ and 5′-TCTTTAAAGCTTCGTGGTC-3′). The mouse β-actin housekeeping gene (5′-AGAGGGAAATCGTGCGTGAC and 5′-CAATAGTGATGACCTGGCCGT) was also amplified at the same time to normalize the quality and quantity of DNA between samples. A plasmid containing the same primer template sequences was used as a standard to calculate the number of parasites per μg of DNA. Real-time PCR was conducted with the SYBR Green PCR Core Kit (Applied Biosystems, Warrington, U.K.) on an iCycler iQ instrument (Bio-Rad, Hercules, CA). Amplification conditions were 95°C for 8 min, followed by 40 cycles of 94°C for 15 s, 63°C for 45 s, and 72°C for 15 s.
Purification of LP CD4+ T lymphocytes and intestinal CD11c+ cells
LP CD4+ T cells and CD11c+ were purified from at least 12 mice, as described (17). Peyer’s patches were removed, and sliced intestines were placed in RPMI 1640 with 25 mM EDTA for 45 min. Intestinal fragments were then incubated for 2 h at 37°C in RPMI 1640 containing 125 UI/ml collagenase VIII (Sigma-Aldrich, St. Louis, MO). After centrifugation on a Ficoll layer, the CD4+ LP T cells were purified by positive selection using anti-CD4 beads and CD11c+ cells using anti-CD11c (L3T4 microBeads; Miltenyi Biotec). The purity of CD4+ LP cell population was >90%, as attested by FACScan analysis (clone L3T4; BD Biosciences). CD11c+ cells were also isolated from the mesenteric lymph nodes following the same purification procedures as for the LP.
CD4+CD45RBhighCD25− T cell transfer
Mononuclear cells isolated either from the LP or the mesenteric lymph nodes were stained with anti-CD25 PE, anti-CD45RB FITC, and anti-CD4 CyChrome (BD Pharmingen, San Diego, CA) Abs. CD4+CD45RBhighCD25− T cells (named T effector cells) were then sorted by FACSorter (BD Biosciences). The purified cells (1 × 105) were adoptively transferred by i.v. route into immunocompromised RAG1−/− mice. The purity of the sorted cells was >97%. One day after the transfer, recipient mice were infected orally with the parasites.
SAG1 Ag purification
The SAG1 protein was purified by immunoaffinity (28), using the monoclonal IgG anti-SAG1 (mAb 1E5) (29). A sonicated Ag was prepared from RH tachyzoites, as described elsewhere, and referred to as crude extract (30). SAG1 was purified using an anti-SAG1 mAb affinity column, as previously described (28). The purity was determined by electrophoresis, followed by a transfer to a nitrocellulose paper probed with the mAb 1E5. The quantification of SAG1 was performed with a bicinchoninic acid protein assay reagent kit (Pierce, Rockford, IL).
To detect total TGF-β, 100 mg of infected tissues was homogenized and assayed by ELISA (BioSource International, Camarillo, CA). IL-12p70 production by the CD11c+ cells isolated from the mesenteric lymph nodes and from the LP (4 × 105 cells/well; DC-parasites ratio 1:1) was assayed by ELISA.
RNase protection assay (RPA)
Total mRNA either from freshly isolated LP mononuclear cells (LPMC) (infected or uninfected) or from the mICcl2 cells (infected or uninfected) was extracted using the TRIzol reagent (Invitrogen Life Technologies). Cytokine and chemokine expression was detected using the RiboQuant multiProbe RPA system kit (BD Pharmingen). For quantification, band densities were analyzed by NIH Image 1.61/ppc software using a Macintosh computer. Results were expressed as the percentage of the band intensity relative to the intensity of the housekeeping (GADPH and L32) RNA.
Histological assessment of intestinal inflammation
Intestines were embedded in paraffin and stained with H&E for histologic examination. Only specimens exhibiting longitudinally oriented sections through the crypts were measured. Histological inflammatory score ranging from 0 to 4, as previously described (31), was applied in a blinded fashion to estimate intestinal inflammation: 0, no inflammation; 1, slight infiltrating cells in LP with focal acute infiltration; 2, mild infiltrating cells in the LP with increased blood flow and edema; 3, diffuse and massive infiltrating cells leading to disturbed mucosal architecture; 4, crypt abscess and necrosis of the intestinal villi.
For ELISA experiments, results were expressed as the mean of cytokine concentration ± SD. Statistical differences between groups were analyzed using Student’s t test. Value of p < 0.05 was considered as significant.
Δsag1 parasites fail to induce lethal necrosis of the ilea
C57BL/6 mice are susceptible to infection with T. gondii, and acute mortality occurs within 7 days (15) as a result of IFN-γ-mediated necrosis and hemorrhage of the ileum (14). To evaluate the ability of the SAG1 Ag to induce intestinal inflammation, C57BL/6 were infected with a lethal dose of either parental RHΔ or Δsag1 parasites. Disease progression was assessed daily. Both the time to death and survival were significantly reduced after infection with the wild-type compared with the Δsag1 mutant (Fig. 1,A). Morphologic and histologic examination of intestines at days 5, 7, and 9 postinoculation demonstrated an absence of inflammation in those mice infected with Δsag1 parasites. By day 7, severe necrosis of the ilea, predominantly within the villi, a total absence of columnar epithelial cells, and massive accumulation of polymorphonuclear inflammatory cells were detected in the intestinal epithelium of mice infected with RHΔ (Fig. 1,B). In striking contrast at both days 7 (Fig. 1 D) and 9 (data not shown) postchallenge, the intestinal epithelium of mice infected with Δsag1 parasites was consistent with the saline controls and without evidence of significant inflammation.
Genetically manipulated strains that express a functional sag1 gene and the corresponding protein on their surface were added as additional controls and tested for their ability to induce the intestinal pathology. These controls included the M34 strain that has been engineered by reinserting the sag1 gene in the Δsag1 strain and a deficient parasite for SAG3, SAG3KO strain (M. E. Grigg, manuscript in preparation) (32). Intestinal infection with 10,000 parasites from either the M34 or the SAG3KO strains induced an intestinal inflammation 7 days later (Fig. 1, E and F).
Taken together, these results emphasized the important role of SAG1 in intestinal inflammation following T. gondii infection in C57BL/6.
Increased parasite burden associated with Δsag1 parasite infection
One possible explanation for Δsag1 weak intestinal pathogenicity might be that the Δsag1 mutant is less capable of colonizing and replicating in its host. For the wild-type RHΔ, parasite burden peaked 5–7 days postinfection before declining to undetectable levels in the liver, intestine, and spleen of surviving mice. In contrast, parasite burden in Δsag1-infected mice was comparable to RHΔ infections over the first 5 days, but then rose exponentially over the next 4 days, peaking by day 9 postinfection. Between days 7 and 9, parasite load in the intestine and spleen was typically 2-fold greater in mice infected with the Ag-deficient parasite than RHΔ-infected mice (Fig. 2 A).
Both enterocytes and DCs are important as first lines of defense against the invading parasite. An vitro assay was developed to evaluate whether the replication of the Δsag1 mutant parasite was consistent with that of the wild-type strain in these different cell compartments.
The mICcl2 cell line that is derived from the enterocytes of C57BL/6 mice as well as splenic DCs was infected either with the wild-type (RHΔ) or mutant Δsag1 parasites, and 24 h later, parasite replication was measured via the [3H]uracil incorporation test. The rate of parasite replication in the enterocyte line was consistent among the different parasite strains tested (Fig. 2,B). Similar rates of replication were observed in the splenic derived DCs (Fig. 2 C). These data suggest that the reduced pathogenicity of the Δsag1 parasites is not related to a decrease in replication in these two important cell populations or in the whole host.
Reduced production of inflammatory innate immune response after infection with Δsag1 parasites
Oral infection with the wild-type strain enhances the production of chemokines in the small intestine (17). Chemokines are essential for the attraction of inflammatory cells, including macrophages, DCs, PMNs, and T and B cells. All of these cells have been shown to play a role in the clearance of parasites from the infected host. Compared with mice infected with the wild-type parasites, mice infected with the Ag-deficient parasites display decreased levels of chemokine mRNA in their intestinal tissue, in particular for MIP-2 and MCP-1 (Fig. 3, A and B). Enterocytes that are the first barrier of defense against pathogen invasion are important for the production of these molecules (17). Enterocytes from the mICcl2 cell line were examined for their ability to produce a number of chemokines in response to parasite infection. mRNA was isolated at 6 h after infection with either the mutant Δsag1 or the wild-type (RHΔ) parasites. The mRNA expression for MIP-2 (Fig. 3,C) and MCP-1 (CCL-2) (Fig. 3,D) was significantly higher in the mICcl2 cells infected with the wild-type parasites in comparison with Δsag1 parasites. The DCs might also be a crucial cell population for intestinal chemokine production. Ex vivo CD11c+ cells isolated from the mesenteric lymph nodes display a lower ability to produce chemokine MIP-1α (CCL3) and RANTES (CCL5) (Fig. 3, E and F). Similar results were obtained with CD11c+ isolated from the LP (data not shown). These data indicate that SAG-1 protein can induce the expression of several chemokines that may be important in the initiation of the inflammatory response in the intestine. In response to pathogen exposure, enterocytes and DCs can produce additional molecules such as NO and/or IL-12p70 that would directly participate in the innate inflammatory process (33). In the intestinal tissue, taken as a whole, the production of iNOS (Fig. 4,A) was significantly decreased in mice infected with the Δsag1 parasites as compared with the wild-type RHΔ parasites. A substantial difference in iNOS production was observed when LPMC were compared. LPMC isolated from mice infected with RHΔ parasites produced a substantially greater quantity of mRNA for iNOS, compared with those mice infected with the Δsag1 mutant parasites that failed to express iNOS mRNA (Fig. 4,B). In addition, we observed that following infection with the Δsag1 parasites, enterocytes from the mICcl2 cell line (p < 0.05) (Fig. 4,C) synthesized less iNOS mRNA. Ex vivo CD11c+ cells isolated from the mesenteric lymph nodes have a lower ability to produce Th1 cytokines such as IL-12p40 (Fig. 4 D). Similar results were obtained with CD11c+ isolated from the LP (data not shown). These observations suggest that SAG1 protein may be important in both trafficking and activation of professional and nonprofessional APCs. This observation also suggests that the SAG1 protein is implicated in iNOS activation.
Reduced production of inflammatory cytokines and adaptive immune response after infection with Δsag1 parasites
IFN-γ acts synergistically with TNF-α to activate the inflammatory response that limits parasite proliferation (23). If uncontrolled, overexpression of these cytokines leads to the overproduction of NO. Previous experiments have indicated maximal IFN-γ production at day 7 postinfection in susceptible mice that develop acute ileitis. Cytokine quantitation by RT-PCR shows that intestinal IFN-γ and TNF-α mRNA in mice infected with the RHΔ parasites were present at higher levels at day 7 postinfection than those of mice infected with the Δsag1 parasites (Fig. 5, A and B). A similar decrease in proinflammatory cytokine mRNA in response to the Ag-deficient mutant was observed in both mesenteric lymph nodes and spleen.
In this model of experimental ileitis, the deleterious effect of this response has been reported to be mediated by TNF-α- and IFN-γ-producing CD4+ T cells from the LP (CD4+ LP T cells) (17). The absence of inflammatory immune response following intraintestinal inoculation of the Δsag1 parasites is due perhaps to the lack of CD4+ LP T cell activation secondary to the diminished production of chemokines, IL-12, and iNOS. CD4+ LP T cells were more abundant in the mice infected with the wild-type parasites (6 × 105± 3 × 104 cells/mouse) compared with the Ag-deficient strain (3 × 105± 6 × 104 cells/mouse) and uninfected mice (1 × 105± 5 × 104 cells/mouse). The failure of CD4+ LP T cell activation after infection with the Δsag1 parasites was further investigated by mRNA analysis for cytokine production. CD4+ LP T cells from the mice infected with the Δsag1 parasites produced significantly less (p < 0.05) IFN-γ (Fig. 5,C) and TNF-α (Fig. 5 D) than the controls isolated from mice infected with the RHΔ parasites. This observation suggests that the SAG1 protein is implicated in immune cell activation.
The defect in activation of the strong immune response after infection with the Δsag1 parasites might be related to the overstimulation of regulatory cytokines. There was no difference in IL-4 mRNA expression in the intestinal tissue regardless of the infecting strain (Fig. 5,E). IL-10 could not be detected in any of the groups at day 7 after infection (data not shown). The role of TGF-β in the control of the hyperinflammatory response has been previously reported (16). Intraepithelial lymphocytes (IELs) have been identified as a major source of TGF-β production in our model. There was no significant difference in the quantity of TGF-β secreted by IELs purified from mice infected with either RHΔ or Δsag1 parasites (Fig. 5 F). These data suggest that although IL-4 and TGF-β are increased in response to parasite infection, they do not appear to be responsible for the differences noted in the inflammatory response to the Δsag 1 parasites.
Transfer of Ag-specific CD4+CD45RBhigh T cells from the LP of mice infected with the Δsag1 parasites into resistant RAG1−/− mice failed to confer the intestinal pathology
CD4+ T cell population characterized by high expression of the CD45RB Ag contains cells that mediate both protective and pathogenic Th1 responses, and the reciprocal CD45RBlow population can suppress both of these responses (34). Functionally specialized regulatory T cells exist as part of the normal immune repertoire, preventing the development of pathogenic responses to both self and intestinal Ags (35). Based on these observations, the CD4+ T cell population was isolated from the LP of infected mice, and by cell sorting, potential T regulatory cells (CD25+) were discarded, whereas the likely most efficient effector T cells (CD45RBhigh) were sorted. To further investigate the role of immune components, and more specifically of SAG1-specific T cells, in the development of the intestinal inflammation after T. gondii oral infection, adoptive transfer of CD4+CD45RBhighCD25− T cells (effector T cells) into RAG1−/− mice was performed. RAG1−/− mice orally infected with T. gondii do not develop the ileitis. However, 13 days after the transfer of T. gondii-specific effector T cells isolated from mice infected with the wild-type ΔRH strain, RAG1−/− mice exhibited intestinal inflammation following infection (Fig. 6,A). RAG1−/− mice adoptively transferred with specific effector T cells isolated from Δsag1-infected mice (Fig. 6 B) failed to develop such a pathology. Control uninfected RAG1−/− mice never developed the acute ileitis at least within the short 13-day window observation, regardless of the source of effector cells transferred (ΔRH- or Δsag1-primed T cells). Similarly, transfer of unprimed T cells isolated from naive LP was unable to induce the acute ileitis even upon parasite re-exposure in RAG1−/− mice. Together, these observations suggest that T. gondii Ag-specific T effector cells are able to transfer the inflammatory disease upon restimulation, and that SAG1-specific T cells are most likely involved in this process.
Presensitization with the SAG1 protein increases susceptibility of resistant BALB/c mice
Intranasal SAG1 immunization triggers a strong Th1-like immune response (25). To assess the ability of SAG1 to induce inflammation in resistant mice, inbred BALB/c mice were sensitized via intranasal route with SAG1 protein and CT and challenged orally with tissue cysts. BALB/c mice were chosen for the sensitization studies because this strain is resistant to the development of acute ileitis following oral parasite infection. The ability of CT to act as a mucosal adjuvant has been reported (25). Based on our early experience using the intranasal instillation of the SAG1 protein (25), CT was added to the SAG1 protein (SAG1 + CT), and mice were immunized, as outlined in Materials and Methods. Mice were challenged with parasites following the completion of the vaccination with SAG1 + CT. It was observed at day 7 postinfection that mice sensitized with SAG1 + CT developed histologic evidence of an acute inflammatory ileitis. Histological score and examination revealed extensive disruption of the villi from SAG1 + CT-presensitized mice. In comparison, neither of the control groups, naive mice or mice vaccinated with CT alone, developed intestinal pathology (Fig. 7). Moreover, 3 of 10 SAG1 + CT-vaccinated mice died from lethal ileitis following parasite challenge. These data suggest that resistant inbred mice could be rendered susceptible to acute ileitis when sensitized with the single parasite Ag SAG1 in the presence of an appropriate mucosal adjuvant such as CT.
Increased Th1 cytokine expression in SAG1 + CT-presensitized BALB/c mice
The cytokine profile in the intestine was examined after challenge for IFN-γ mRNA expression in those mice presensitized with SAG1 + CT (Fig. 8,A). Expression was up-regulated for both IFN-γ and TNF-α (data not shown) in mice that had been presensitized with SAG1 + CT compared with control mice. NO, essential for the development of experimental ileitis, as well as other proinflammatory cytokines such as IL-1α were increased in the group immunized with SAG1 + CT (Fig. 8,B). No difference was observed in cytokine production in the control groups either only challenged or treated with CT alone. The role of IL-10 and TGF-β in the regulation of the inflammatory response was also evaluated. There was no difference in either IL-10 mRNA expression (Fig. 8,C) or TGF-β protein in the gut whatever the experimental groups (Fig. 8 D). These data demonstrate that sensitization with SAG1 + CT is associated with the induction of an intestinal Th1-like immune response and turns the resistant phenotype of BALB/c mice into susceptible.
We demonstrate that acute inflammatory ileitis in mice can be elicited by exposure to a surface Ag of T. gondii that is recognized by both acute and convalescent sera from humans infected with this parasite. Although the Δsag1 parasites can infect and replicate both in vitro and in vivo with equal efficiency to the wild-type parental strain, these Ag-deficient parasites fail to elicit an inflammatory response in the infected mouse small intestine. Consistent with this lack of histologic evidence of inflammation is an insufficient up-regulation of the appropriate chemokines and cytokines that initiate and mediate this inflammatory response. The anti-inflammatory response, including TGF-β secretion, IL-10, and Il-4 production, does not appear to be implicated. In addition, T. gondii-specific effector CD4+ T (CD25−CD45RBhigh) lymphocytes, when stimulated with the wild-type parasite, but not with the Δsag1 mutant parasite, can transfer the inflammatory disease into normally resistant RAG1−/− mice following infection. This indicates that SAG1-specific T lymphocytes might be involved in this process.
Microbial Ags may provide a local environmental trigger that initiates and perpetuates intestinal inflammatory response (36). The most compelling evidence for involvement of microflora in the pathogenesis of mucosal inflammation models has been examined in several experimental murine models of IBD (37). Previous observations have shown that oral inoculation of T. gondii can induce acute ileitis in C57BL/6 mice (14, 33). This T. gondii-driven process of acute ileitis in a genetically unmanipulated strain of mouse provides a reasonable experimental model to evaluate the importance of specific microbial Ags. Histological analyses of the intestinal tissue from these mice demonstrate many morphologic changes consistent with human ileitis. We have observed that in contrast to the wild-type parasites, the Ag-deficient parasites induce substantially less inflammation in the gut of the parasite-infected host.
The failure to initiate an inflammatory response is not linked to a decreased ability of the mutant parasite to either invade or to replicate in host cells. In vivo observations demonstrate that the mutant replicates at a rate consistent with the wild-type strain. In vitro, no difference was observed between the replicative rate of the mutant and the wild-type parasites in host cells. We observed that mice infected with the mutant parasite eventually died from parasitemia without evidence of ileitis. The recruitment and activation of leukocytes at sites of intestinal inflammation and injury are the hallmark of intestinal inflammation. Chemokines are central to the recruitment and activation of leukocytes and have overlapping functions during the process of inflammation. Both MIP-2 (IL-8) and MCP-1 (CCL2) play an important role in the immunopathogenesis of intestinal inflammatory disease (38, 39). In vitro, we show that the SAG1KO parasites were less efficient in stimulating the expression of MIP-2 and MCP-1 (CCL-2) in a C57BL/6-derived enterocyte line (mICcl2 cells) when compared with wild-type parasites. We speculate that this reduction in chemokine expression would lead to decreased recruitment and activation of inflammatory cells such as PMNs, macrophages, DCs, and B and T lymphocytes with a decrease in the severity of the inflammatory response. Increased iNOS expression has been associated with the inflammatory disorder observed after infection with the parental strain (33). In vitro, enterocytes produce less iNOS when infected with the Δsag1 parasites compared with wild-type parasites. This further emphasizes the diminished capacity of the mutant strain to initiate an inflammatory response.
Intestinal inflammation and tissue damage are characterized by an exaggerated immune response mediated by both TNF-α and IFN-γ (40) and heightened sensitivity of intestinal epithelial cells to TNF-α (41, 42, 43). IFN-γ is the cytokine mediator associated with intestinal inflammation following oral infection with T. gondii. Recent studies have identified that at least one source of IFN-γ is LP CD4+ T lymphocytes (17). Host exposure to SAG1 protein affects the LP CD4+ T cell response. We observed a significant reduction in the expression of both TNF-α and IFN-γ by the CD4+ LP T lymphocytes isolated from mice infected with the Δsag1 parasites compared with mice infected with RHΔ. In vivo and in vitro, chemokine expression was significantly reduced following infection with the Δsag1 parasites, and this observation is consistent with the low attraction and activation of the LP CD4+ T lymphocytes. A decrease in the production of the proinflammatory cytokine, IL-12, was seen in those DCs infected with the SAG1-deficient parasite. The diminished IL-12 production may be responsible for the reduced capacity of CD4+ T lymphocytes to secrete TNF-α and IFN-γ. Together, these observations imply that the SAG1 Ag is an important mediator of the increased levels of IFN-γ, TNF-α, and iNOS following oral parasite infection. This Ag appears to be directly linked to the development of the lethal ileitis following oral infection in the C57BL/6 mice. We have generated data using another strain of genetically engineered parasites that is deficient for SAG3, another parasite surface Ag (32), and have found that these KO parasites are capable of inducing lethal ileitis similar to the parental strains. Although there may be other parasitic Ags that are yet to be identified as capable of eliciting this lethal inflammatory process, our data support a unique immunogenic role for the SAG1 protein. Previous investigations from our laboratory and others have clearly demonstrated the ability of this particular parasite Ag, SAG1, to induce a population of IFN-γ-producing Ag-specific CD4+ T cells (44). Furthermore, it has been shown that SAG1-specific CD4+ T cells are involved in long-term immunity in those individuals with chronic infection (45). To further investigate the specific role of CD4 T cells in the development of this acute inflammatory response, we used RAG−/− mice in which the innate immune response against the parasite is intact, but that lack all the T and B cell adaptive immune responses. Although DCs from the RAG−/− mice are capable of responding to the T. gondii infection by producing high amounts of IL-12 and IFN-γ (data not shown), and despite a normal parasite replication, these mice do not develop the acute inflammatory intestinal disease after the infection with T. gondii. Transfer of T. gondii-specific effector T cells with the phenotype CD4 CD25−CD45RBhigh stimulated with the parental strain into infected RAG−/− mice resulted in the development of a lethal inflammatory response day 13 postinfection by T. gondii, whereas the same CD4+ cell subpopulation cells stimulated with the Δsag1-deficient parasites did not. This emphasizes the important role of Ag-specific CD4+ T cells in the pathogenesis of this infection, and reinforces the idea that this is the unregulated immune response rather than the parasite itself that leads to the lethal damages observed following infection with T. gondii (46, 47).
Susceptible mice can be rendered resistant to ileitis by alteration or deletion of specific regulatory immune products. For example, C57BL/6 mice deficient in IL-12, CD40, or CD40L fail to develop gut inflammation in response to parasite infection (48). Conversely, we have also observed that resistant mice (49) are made susceptible by either blocking TGF-β (16) or deleting the gene for IL-10 (50), respectively. We demonstrate that intranasal sensitization of resistant BALB/c mice, followed by Ag re-exposure can render a resistant animal susceptible to the development of ileitis. This change in phenotype is associated with an increase in the production of Th1-type cytokines rather than a reduced anti-inflammatory response from an increased TGF-β, IL-10, or IL-4 production. It has been demonstrated that in these conditions of SAG1 protein stimulation, Ag-specific CD4+ T cell-producing IFN-γ could be generated (25). We argue that presensitization with SAG1 generates an inflammatory Ag-specific response.
Oral infection with T. gondii in B6 mice provokes a robust inflammatory response that results in the development of lethal ileitis. In this study, the identification of a single microbial Ag that can elicit this response provides a unique and potentially useful model to study inflammation in the small intestine and its regulation.
We gratefully acknowledge Joe Schwartzman for his assistance in histology slide analysis, and members of the Kasper lab for many helpful discussions. We also thank Dr. F. Velge-Roussel for providing us with purified SAG1 protein.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Grants AI19613, AI30000, and TW011003.
Abbreviations used in this paper: IBD, inflammatory bowel disease; CT, cholera toxin; DC, dendritic cell; IEL, intraepithelial lymphocyte; iNOS, inducible NO synthase; KO, knockout; LP, lamina propria; LPMC, LP mononuclear cell; PMN, polymorphonuclear cell; RPA, RNase protection assay; SAG, surface Ag.