Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Inflammatory bowel diseases (IBDs) are common immune-mediated disorders of the gastrointestinal tract. An imbalance between Th1 proinflammatory and Th2 anti-inflammatory subtypes of immune responses plays a role in the pathogenesis of these disorders (Podolsky, 1991; Mizoguchi et al., 1996; Adorini and Francesco, 1997). In both experimental colitis, and in patients with IBD, the disease is a Th1-mediated immune disorder, resulting in a life-long inflammatory response against the colon. Secretion of proinflammatory cytokines such as IFN has been described and contributes to an increased mucosal permeability in these settings (Strober et al., 1997). Anti-inflammatory cytokines such as IL10 down-regulate the proinflammatory effects of Th1-mediated cytokines, thereby alleviating the disease (Neurath et al., 1996; Madsen et al., 1997; Van Deventer et al., 1997).

Oral tolerance is a recognized procedure for induction of antigen-specific peripheral immune hyporesponsiveness (Weiner, 1997). Enteral exposure to high antigen doses induces tolerance by clonal inactivation of antigen-specific T cells. Feeding low antigen doses leads to induction of regulatory cells secreting immunosuppressive cytokines that suppress the generation of antigen-specific effector cells (Miller et al., 1992). Oral tolerance was previously shown effective in preventing immune-mediated disorders in animal models, such as experimental allergic encephalomyelitis, collagen-induced arthritis, diabetes, and chronic graft versus host disease (Zang et al., 1990; Trentham et al., 1993; Von Herrath et al., 1996; Ilan et al., 2000a). Oral tolerance toward adenoviral antigens effectively prevents the antiviral immune response (Ilan et al., 1997c) and seems superior to other modes of immune tolerance induction (Ilan et al., 1996, 1997a,b; Takahashi et al., 1996). Recently, it has been shown by us and others that oral tolerance can be used to prevent experimental colitis in a model system using mice treated with 2,4,6-trinitrobenzenesulfonic acid (TNBS) (Trop et al., 1999; Ilan et al., 2000b). This material induces an autoimmune response resembling Crohn's disease in humans (Neurath et al., 1996). Initiation of oral tolerance via feeding of colitis-extracted proteins, before colitis induction, ameliorates the disease. However, clinical application of oral tolerization for patients with IBD would require tolerance induction in the presence of preexisting immunity and established colitis. It would also necessitate the use of surrogate antigens rather than colitis-extracted proteins. Oral administration of antigens has been shown to alleviate antigen-specific immune response and secondary immune response in the presence of preexisting immunity (Ilan et al., 1998). Both in animals and humans it has had a beneficial effect in alleviating preexisting autoimmune disorders such as scleroderma, rheumatoid arthritis, and diabetes (Weiner et al., 1993; Strobel and Mowat, 1998; Trentham, 1998; Gotsman et al., 2000).

In the present study we sought to evaluate whether induction of oral tolerance toward colitis- or normal colon-extracted proteins (CEPs and NCEPs, respectively) could be used to alleviate preexisting host anticolon immune response and ameliorate experimental colitis. The results show that induction of oral tolerance in the presence of established disease, and the use of surrogate antigens, had a beneficial effect in alleviating experimental colitis.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals

BALB/c male mice, 2 to 4 months old, were obtained from Harlan Laboratories (Indianapolis, IN) and maintained in the Animal Core of the Hadassah-Hebrew University Medical School, Jerusalem, Israel. Mice were maintained on standard laboratory chow and kept in 12-h light/dark cycles. All animal experiments were carried out according to the guidelines of the Hebrew-University-Hadassah Institutional Committee for Care and Use of Laboratory Animals and with the committee's approval.

Induction of Colitis

TNBS-colitis was induced by rectal instillation of TNBS (Sigma Diagnostics, St. Louis, MO), 1 mg/mouse, dissolved in 100 µl of 50% ethanol as described (Trop et al., 1999).

Preparation and Administration of the Oral Antigens

The colon was removed from normal or from TNBS-induced colitis colons of mice, 3 days following colitis induction; cut into small strips; and mechanically homogenized. Following filtration through a 40-µm nylon cell strainer, the intact cells were spun down and removed. Proteins were quantified using a protein assay kit (Bio-Rad, Richmond, CA). NCEPs and CEPs were orally administered into mice from all experimental groups, using a feeding atraumatic needle, 50 µg/mouse, every other day for 10 days (a total of five doses) as described below. This dose was found to be in the low-dose oral tolerance range.

Experimental Groups

Six groups of mice, consisting of 10 animals each, were studied (Table 1). All mice were challenged with rectal TNBS on day 1 of the study. Mice in groups A, B, and E were fed every other day for 10 days starting on the day of colitis induction. For evaluation of the effect of tolerance induction on established experimental colitis, mice in groups C, D, and F were fed with identical doses starting on day 7 following induction of colitis. Mice in groups A and C were fed with CEP. For evaluation of the effects of surrogate antigens on tolerance induction, mice in groups B and D were fed with NCEP. Mice in control groups E and F were fed with identical doses of bovine serum albumin (BSA). Mice in all groups were sacrificed 21 days following colitis induction.

Evaluation of the Effect of Preexisting Immunity and Surrogate Antigens on Oral Tolerance Induction in the Experimental Colitis Model. The role of presensitization toward target antigens and surrogate antigens on tolerance induction was evaluated by monitoring the following parameters for colitis.

Clinical assessment of colitis. Diarrhea was followed daily throughout the study. Body weights were measured before colitis induction, and at the end of study in all experimental groups.

Macroscopic score of colitis. Colitis assessment was performed 21 days following colitis induction, using four standard macroscopic parameters, namely, degree of colonic ulcerations, intestinal and peritoneal adhesions, wall thickness, and degree of mucosal edema (Ilan et al., 2000b). Each parameter was graded on a scale from 0 (completely normal) to 4 (most severe). Coded specimens were scored by two experienced examiners.

Grading of histological lesions. For histological evaluation of inflammation, distal colonic tissue (last 10 cm) was removed and fixed in 10% formaldehyde. Five paraffin sections from each mouse were then stained with H&E by using standard techniques. The degree of inflammation on microscopic cross sections of the colon was graded semiquantitatively from 0 to 4 (Ilan et al., 2000b): grade 0, normal with no signs of inflammation; grade 1, very low level of leukocyte infiltration; grade 2, low level of leukocyte infiltration; grade 3, high level of infiltration with high vascular density and bowel wall thickening; and grade 4, transmural infiltrates with loss of goblet cells, high vascular density, wall thickening, and disruption of normal bowel architecture. Coded specimens were scored by two experienced examiners.

Evaluation of the Mechanism of Immune Tolerance Induction. Serum IL4 and IFN levels. Serum IL4 and IFN levels were measured by a "sandwich" enzyme-linked immunosorbent assay, using Genzyme Diagnostics kits (Genzyme Diagnostics, Boston, MA) according to manufacturer's instructions. Serum levels were measured in all mice from all experimental and control groups 20 days following colitis induction.

Statistical Analysis

Statistical differences were calculated by the Student's t test. Mice were compared with their respective control group: Mice in experimental groups A and B were compared with control group E, and mice in experimental groups C and D were compared with control group F.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Evaluation of Effect of Preexisting Immunity and Surrogate Antigens on Oral Tolerance Induction in Experimental Colitis Model

Clinical Assessment of Colitis. A marked decrease in diarrhea was observed in tolerized mice from group A fed with mouse CEP. A similar beneficial effect was observed in mice treated with mouse-derived NCEP in group B, which were fed starting on day 1. A partial beneficial effect was observed in mice in which tolerance was induced in the presence of established colitis (groups C and D). In contrast, mice in control groups E and F fed with BSA suffered severe diarrhea.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Effect of tolerance induction toward colitis and normal colon-extracted proteins on body weights in mice with experimental colitis. Follow-up of mouse body weight disclosed a statistically significant increase in body weight in tolerized mice in group A compared with control mice in group E (p < 0.005). A similar beneficial effect was observed in mice from group B treated with mouse-derived NCEP (p < 0.005). A favorable effect was also observed in mice fed with CEP and NCEP starting 7 days following induction of colitis (p < 0.005 compared with control mice). In contrast, no significant increase in body weight was noted in mice treated with BSA.

Macroscopic Grading of Colitis. Induction of oral tolerance by feeding of mouse-extracted colitis-derived proteins (group A), or with mouse normal colon extracted proteins (group B) markedly alleviated the macroscopic grading of colitis. The scores for tested macroscopic parameters of colitis were degree of colonic ulceration, intestinal and peritoneal adhesions, wall thickness, and degree of mucosal edema. Total macroscopic scores measured 0.4 and 0.7 in group A and B mice, respectively, compared with 3.3 and 3.1 in nontreated controls in groups E and F, respectively (p < 0.005, Fig. 2). A favorable effect was also observed in mice fed with CEP starting 7 days following colitis induction (group C). Their total macroscopic score measured 1.2. A partial beneficial effect was noted in mice fed with NCEP starting 7 days following colitis induction, and their scores were 1.9 (p < 0.005 compared with control mice in group F, Fig. 2).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Effect of tolerization on macroscopic grading of colitis. Parameters tested were graded on a scale from 0 (completely normal) to 4 (most severe), including degree of colonic ulceration, intestinal and peritoneal adhesions, wall thickness, and degree of mucosal edema. All tested parameters were significantly down-regulated in tolerized mice fed with mouse-extracted colitis-derived proteins (group A), or with mouse normal colon-extracted proteins (group B), compared with nontreated controls in groups E and F (p < 0.005). A favorable effect was also observed in mice fed with CEP starting 7 days following colitis induction (group C, p < 0.005 compared with control mice). A partial beneficial effect was noted in mice fed with NCEP starting 7 days following colitis induction.

Grading of Histological Lesions. Histological evaluation of bowel tissue showed a marked reduction in inflammatory response and mucosal ulcerations in tolerized mice in groups A, B, C, and D, compared with nontolerized control mice in groups E and F. In mice in groups A and B, almost normal sections or only minimal lymphocytic infiltration was detected (Fig. 3, A and B). In contrast, severe inflammatory reaction (grade 3-4) was observed in bowel specimens taken from nontolerized control mice (Fig. 3, E and F). A partial effect was observed in mice fed with CEP or NCEP starting 7 days following colitis induction in groups C and D (Fig. 3, C and D). The results of standard pathological tests in tolerized mice from groups A, B, C, and D produced scores of 0.4, 1.2, 1.4, and 1.95, respectively. Control nontreated mice in groups E and F exhibited severe microscopic colitis, with pathological scores of 3.1 and 3.25, respectively (p < 0.005 compared with control mice, Fig. 4).

View larger version (182K): [in this window] [in a new window]   Fig. 3.   Effect of tolerization on histological evaluation of bowel mucosa in experimental colitis: a marked reduction in inflammatory response and mucosal ulcerations was noted in tolerized mice in groups A, B, C, and D, compared with nontolerized control mice in groups E and F. In mice in groups A and B, almost normal sections, or only minimal lymphocytic infiltration, was detected (A and B). In contrast, severe inflammatory reaction (grade 3-4) was observed in bowel specimens taken from nontolerized control mice (Fig. 3, E and F). A partial effect was observed in mice fed with CEP or NCEP starting 7 days following colitis induction in groups C and D (C and D).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effect of tolerization on the microscopic grading of experimental colitis. The results of standard pathological tests in tolerized mice from groups A, B, C, and D produced significantly lower scores compared with control nontreated mice in groups E and F (p < 0.005).

Mechanism of Immune Tolerance Induction by Administration of Surrogate Antigens in the Experimental Colitis Model

Serum IL4 and IFN Levels. Tolerized mice manifested a shift from a Th1 to Th2 immune response cytokine secretion. The feeding of mouse-derived CEP induced an increase in IL4 levels and a decrease in IFN serum levels 20 days following colitis induction, to 22.4 ± 5 and 1.2 ± 0.7 pg/ml, respectively (group A, p < 0.005 compared with control group E, Fig. 5). Similarly, the feeding of mouse-derived NCEP induced a Th1 to Th2 immune shift with IL4 and IFN serum levels of 18.4 ± 1.9 and 2.4 ± 0.9, respectively (group B, p < 0.005). Primed mice that were tolerized in the presence of preexisting disease fed with CEP or NCEP exhibited a similar cytokine paradigm, and their IL4 levels increased to 16.4 ± 2.9 and 10.3 ± 1.1, whereas their IFN serum levels measured 6.3 ± 0.9 and 7.3 ± 1.1 (groups C and D, respectively). In contrast, mice from nontolerized control groups E and F, exhibited high IFN and low IL4 serum levels (18.4 ± 2.7 and 1.9 ± 0.3 and 22.1 ± 1 and 2.1 ± 0.5, respectively, p < 0.005 compared with control mice in group F, Fig. 5).

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Effect of oral tolerization on serum IL4 () and IFN () serum levels. Tolerized mice manifested a shift from a Th1 to Th2 immune response cytokine secretion. The feeding of mouse-derived CEP induced an increase in IL4 levels and a decrease in IFN serum levels 20 days following colitis induction (group A). Similarly, the feeding of mouse-derived NCEP induced a Th1 to Th2 immune shift with IL4 and IFN serum levels (group B, p < 0.005). Primed mice that were tolerized in the presence of preexisting disease fed with CEP or NCEP exhibited a similar cytokine paradigm, with increased IL4 and decreased IFN serum levels (groups C and D, respectively, p < 0.005 compared with control mice in group F). In contrast, mice from nontolerized control groups E and F exhibited high IFN and low IL4 serum levels.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Oral administration of low doses of colitis-extracted proteins to mice with experimental colitis was shown to have induced immune tolerance in the presence of established colitis. Tolerance induction down-regulated the inflammatory response and alleviated the disease. Moreover, the feeding of normal colon-extracted proteins had a significant beneficial effect.

Oral tolerance is an effective means of inducing antigen-specific immunological hyporesponsiveness (Weiner, 1997; Trop et al., 1999). Oral tolerance has been shown effective in preventing various immune-mediated disorders. Both in experimental animals and in humans, tolerance induction is associated with a shift from a Th1 proinflammatory to a Th2 anti-inflammatory state, with secretion of immunosuppressive cytokines (Ilan et al., 2000b). The preventive role of oral tolerance via administration of low-dose colitis-extracted proteins on development of experimental colitis was recently shown by others and by us (Ilan et al., 2000b). In a model system that uses mice injected with TNBS, a Th1-mediated immune disorder is induced, leading to an experimental colitis similar to the human disease. By feeding a mixture of colonic proteins exposed to TNBS or with colitis-extracted proteins, the host colonic inflammatory response was inhibited, thereby preventing the development of clinical, macroscopic, and microscopic manifestations of colitis (Neurath et al., 1996; Ilan et al., 2000b). Tolerance induction in this model replaced a Th1-proinflammatory with a Th2-anti-inflammatory type of immune response (Ilan et al., 2000b). Adoptive transfer of tolerance by transplantation of splenocytes from tolerized donors to sublethally irradiated recipients further supports the presence of suppressor cells in this setting. However, clinical application of this mode of therapy would require tolerance induction in the presence of preexisting colitis and the use of surrogate antigens extracted from a nondiseased bowel.

The results of the present study show for the first time that it is possible to induce tolerance toward colitis-extracted proteins in animals with established colitis. Although tolerance induction in immunologically primed animals with preexisting disease was less pronounced than when preventive oral tolerization was used, it alleviated the disease and was clinically significant compared with nontolerized control groups. Immunologically primed mice with established colitis fed with colitis-extracted proteins exhibited a significant increase in body weight and an improvement in all macroscopic and microscopic parameters of colitis (experimental group C).

In several animal models, such as experimental encephalitis, uveitis, and arthritis, oral tolerance was shown to suppress an established immune response (Miller et al., 1992; Lieshman et al., 1998). Feeding ovalbumin 7 days following immunization produces suppression of delayed type hypersensitivity, via suppression of T cells (Lieshman et al., 1998). Oral tolerance down-regulates the secondary immune response in the presence of preexisting antiadenovirus immunity in animals (Ilan et al., 1998). Enteral administration of adenoviral structural proteins into preimmunized rats down-regulated both a humoral and cellular antiviral immune response. Oral administration of low-dose hepatitis B virus envelope proteins effectively inhibits antihepatitis B antibody production in mice with secondary antiviral immunity (Gotsman et al., 2000). Several clinical trials using oral tolerance in humans with autoimmune diseases in whom the immune response is well established, including multiple sclerosis, rheumatoid arthritis, scleroderma, and diabetes, have been published, which show promising results (Weiner et al., 1993; Trentham, 1998). In contrast, several studies have shown that oral tolerance induction in the setting of a previously primed animal is less effective in down-regulating immune-mediated inflammatory response. In allergic contact hypersensitivity models in guinea pigs, oral administration of nickel to naive animals induces immune hyporesponsiveness and prevents subsequent induction of contact dermatitis. In contrast, oral tolerance to sensitized animals did not prevent subsequent contact dermatitis (Van Hoogstraten et al., 1994). Antigenic variability, the administered dose, and different mechanisms in various disorders may be responsible for this variance. The results of the present study show that in the experimental colitis model, tolerance induction in primed animals, although somewhat less pronounced, was clinically significant (experimental groups A and C versus E and F).

Clinical application of oral tolerance in patients with IBD would require the use of surrogate antigens. A similar approach has been used in trials in patients with multiple sclerosis, diabetes, and rheumatoid arthritis (Weiner et al., 1993; Weiner, 1997; Trentham, 1998). A bystander effect is known to play a role in oral tolerance induction (Miller et al., 1991; Karpus et al., 1995; Lundin et al., 1996; Von Herrath, 1997). It involves regulatory cells secreting nonantigen-specific cytokines that suppress inflammation in the microenvironment where the fed antigen is localized. Mucosal Th2/Th3 cells, secreting transforming growth factor 1 generated by intermittent feeding of a low-dose antigen may play a role in bystander tolerance (Von Herrath, 1997). Several studies showed only partial or no effect of bystander oral tolerance (Miller et al., 1993; Balans et al., 1996; Carvalho et al., 1997). In the present study we have shown that surrogate antigens derived from mouse normal colonic wall induced a beneficial effect, although less pronounced than the one achieved by the feeding of mouse-derived colitis-extracted proteins. This effect was more prominent when proteins were administered preventively than when they were given to preexisting colitis (experimental groups B and D). These results suggest that surrogate antigens, related to the disease-target epitopes, may have a similar immunomodulating effect. They imply that closely related proteins are being presented and processed by gut-associated lymphoid tissue in a similar way. Therefore, both administration of an antigenically similar epitope, or of an epitope distinct from the disease-target antigen, but found in the target organ, can down-regulate peripheral immune activation. When suppressor T cells activated by antigen-presenting cell presentation of these proteins encounter similar epitopes in the colon, they secrete anti-inflammatory cytokines (Miller et al., 1991; Karpus et al., 1995). These cytokines are not antigen-specific. We believe that tolerance is induced because both disease-target antigen and tolerizing epitopes are located in proximity to each other. Geographical proximity and/or antigenic similarity to the disease-target antigen may be held responsible for these effects.

The pathogenesis of IBD involves exposure of specific epitopes on diseased bowel mucosa through a toxic, infectious, or immune-mediated effect (Hibi et al., 1983; Das et al., 1990; Podolsky, 1991; Dasgupta et al., 1994; Neurath et al., 1995). These cryptic antigens incite an active autoimmune inflammatory response (Takahashi and Das, 1985; Z'graggen et al., 1997). Both in humans with IBD and in animals with TNBS-induced experimental colitis, the disease is a Th1-type immune-mediated disorder (Mizoguchi et al., 1996; Neurath et al., 1996). Stimulated cells in the inflamed mucosa produce increased amounts of IFN and IL2 and reduced amounts of IL4, thereby attracting inflammatory cells and disrupting mucosal integrity. In contrast, anti-inflammatory cytokines such as IL10 down-regulate the proinflammatory effects of Th1 cytokines and may alleviate the disease (Madsen et al., 1997). The results of the present study showed that induction of oral tolerance in mice with experimental colitis led to a reverse of the cytokine secretion paradigm with increased secretion of IL4 and decreased secretion of IFN. In contrast in nontolerized animals, a proinflammatory secretion pattern was observed. Successful treatment of IBD, similar to that of other immune-mediated disorders requires overcoming the immune response toward disease-associated target antigens. This involves generalized immunosuppression, which can bring many undesirable side effects. In contrast, induction of specific tolerance toward colitis or normal colon-extracted antigens could potentially allow for long-term alleviation of the disease, while leaving the general immunological defense of the recipient intact.

In conclusion, in the present study we have demonstrated that oral administration of colitis-extracted proteins or of surrogate antigens induced immune tolerance and down-regulated the anticolon inflammatory immune response, thus alleviating established colitis. This method opens up the possibility of understanding immune target antigens involved in the pathogenesis of IBD and has a potential use in clinical practice.