There is compelling evidence that dysregulation of. Mice Deficient in Th1- and Th2-Type Cytokines Develop Distinct Forms of Hapten-Induced Colitis

Transcription

1 GASTROENTEROLOGY 2000;119: Mice Deficient in Th1- and Th2-Type Cytokines Develop Distinct Forms of Hapten-Induced Colitis TAEKO DOHI,*, KOHTARO FUJIHASHI, HIROSHI KIYONO,, CHARLES O. ELSON, and JERRY R. McGHEE* Immunobiology Vaccine Center, Departments of *Microbiology, Oral Biology, and Medicine, University of Alabama at Birmingham, Birmingham, Alabama; Department of Mucosal Immunology, Research Institute for Microbial Diseases, Osaka University, Osaka, Japan; and Department of Gastroenterology, Research Institute, International Medical Center of Japan, Tokyo, Japan Background & Aims: Most experimental models for inflammatory bowel disease in mice are associated with production of interferon (IFN)- and other proinflammatory cytokines. We hypothesized that T-helper 2 (Th2)- type cells could also contribute to the colitis and cause inflammation different than that mediated by Th1-type cells. Methods: Trinitrobenzene sulfonic acid (TNBS)-induced colitis in C57BL/6 background mice genetically deficient in interleukin (IL)-12 p40 (IL-12 / ), IFN- (IFN- / ), or IL-4 (IL-4 / ) was examined in comparison with control mice (C57BL/6 / ). Results: C57BL/6 /, IFN- /, and IL-12 / mice developed patterns of colitis characterized by distortion of crypts, loss of goblet cells, and mononuclear cell infiltration with fibrosis of the mucosal layer. IL-4 / mice had greater mortality than other groups because of penetrating ulcers; however, survivors developed milder lesions that were limited to focal acute ulceration. Colonic CD4 T cells from normal, IFN- /, or IL-12 / mice produced both IL-4 and IL-5. Conclusions: In TNBS colitis, Th1-like cytokine responses induce fatal, acute, transmural, and focal types of lesions, whereas Th2-like cytokine responses play a significant role in the diffuse atrophic changes in crypts and the mucosal layer that occur in the late stages of this disease. There is compelling evidence that dysregulation of the mucosal immune system is a major contributing factor to the pathogenesis of inflammatory bowel disease (IBD). 1,2 Several mouse models for IBD have shown that alterations in cytokine levels result from a failure of regulation by T-helper (Th) cells leading to chronic inflammation in the intestine. Initially, excessive production of interferon (IFN)- and exaggerated Th1-type responses were key immunologic events in the induction of inflammation in most experimental models. For example, increased IFN- synthesis characterized spontaneous colitis in interleukin (IL)-10 deficient mice 3 and in those with selective loss of a G protein, 4 as well as colitis induced by adoptive transfer of T cells to severe combined immunodeficient (SCID) mice. 5,6 On the other hand, studies in other models such as T-cell receptor (TCR)- gene knockout mice suggested a significant role for IL-4 in the spontaneous development of intestinal inflammation. 7,8 This point was directly confirmed by recent studies using an in vivo anti IL-4 monoclonal antibody (mab) treatment system and a double mutation model with IL-4 / mice. 9,10 Hapten-induced colitis with colonic administration of oxazolone was also induced by an IL-4 mediated pathway. 11 Thus, chronic intestinal inflammation may occur in immunologically altered conditions in which either Th1- or Th2-type responses are dominant. Colitis induced by the hapten 2,4,6-trinitrobenzene sulfonic acid (TNBS) was first shown to be a Th1-type of inflammation in SJL/J mice However, we have recently found that TNBS colitis can be induced in the absence of IFN- in BALB/c mice. 15 This result led us to determine the pathophysiologic differences between Th1- and Th2-mediated colonic inflammation. To define the role of Th2-type cells and derived cytokines, we assessed TNBS colitis in IL-12 gene disrupted (IL- 12 / ) mice to eliminate the possibility of a Th1-driven immune response and directly compared this with colitis in IL-4 / mice, in which Th2-type responses are absent. Because BALB/c mice are known to favor Th2-type immune responses, the C57BL/6 mouse strain was chosen for this study to insure that Th2-type responses associated with IBD were not inherent to the mouse strain itself. We now report that both IL-12 / and Abbreviations used in this paper: Ab, antibody; IBD, inflammatory bowel disease; IFN- /, interferon gamma knockouts; IL-4 /, interleukin 4 knockouts; IL-12 /, interleukin 12 p40 knockouts; mab, monoclonal antibody; LPL, lamina propria lymphocyte; TCR, T-cell receptor; TGF-, transforming growth factor ; TNBS, 2,4,6-trinitrobenzene sulfonic acid; TNF-, tumor necrosis factor ; TNP, trinitrophenyl residue by the American Gastroenterological Association /00/$10.00 doi: /gast

2 September 2000 TNBS COLITIS EXHIBITS BOTH Th1 AND Th2 PHASES 725 IFN- / C57BL/6 mice develop significant TNBS colitis, whereas IL-4 / mice of the same C57BL/6 background show a different pattern of disease. Our comparative study of normal, IL-12 /, IFN- /, and IL-4 / mice is the first to show that a single antigen can induce different types of inflammation depending on the dominance of Th1- or Th2-type responses. Materials and Methods Mice Normal, IFN- gene disrupted (IFN- / ), IL-4 gene disrupted (IL-4 / ), and IL-12 p40 gene disrupted (IL- 12 / ) mice, all of C57BL/6 background, and IL-12 p40 gene disrupted (IL-12 / ) and IL-4 / BALB/c mice were purchased from the Jackson Laboratory (Bar Harbor, ME). The mice were kept in microisolator cages in animal facilities at the Immunobiology Vaccine Center University of Alabama at Birmingham. Induction of Colitis Mice were given a solution of TNBS (Research Organics, Cleveland, OH) dissolved in a mixture of phosphatebuffered saline (PBS, ph 7.2) and then mixed with an equal volume of ethanol for a final concentration of 2% TNBS in 50% ethanol. On days 0 and 7, the TNBS enema was administered to mice anesthetized with ketamine and xylazine via a glass microsyringe equipped with a gastric intubation needle, and tissues and cells were assessed 3 days later (day 10). In C57BL/6 mice, 40 g TNBS per gram of their weight was administered intrarectally unless otherwise indicated. Because BALB/c mice were more sensitive to the TNBS enema than C57BL/6 mice, 36 g/g weight was given, as optimized in a previous study. 15 Treatment With mabs In some experiments, mice were treated with mabs to IFN-, IL-12, or IL-4. Separate groups of mice were given 1 mg of rat anti IFN- (XMG1.2), or a mixture of anti IL-12 p40 (clone 18.2) and IL-12 p35 (clone 15.1), or anti IL-4 (11B11) intraperitoneally on days 0 and 7. Control groups received intraperitoneal doses of normal rat immunoglobulin (Ig) G ( Jackson ImmunoResearch Laboratories Inc., West Grove, PA). Histologic Analysis The colon was removed from its mesentery to the pelvic brim by blunt dissection, opened longitudinally, and fixed in 5% glacial acetic acid in ethanol (vol/vol). After embedding in paraffin, 4- m serial sections were prepared and stained with hematoxylin. Preparation of Colonic CD4 T Cells Colonic lamina propria lymphocytes (LPLs) were prepared as described previously, 16 with some modifications. The isolated LPLs were further purified and separated from epithelial cells by centrifugation through a Percoll gradient. LPLs were then stained with fluorescein isothiocyanate labeled anti- CD4 mab (clone GK1.5; UAB Core Facility, Birmingham, AL) and subjected to flow cytometry with a FACStar Plus to obtain purified CD4 T cells ( 99% CD4 T cells). Cytokine-Specific Reverse-Transcription Polymerase Chain Reaction Total RNA fractions were prepared from freshly isolated CD4 LPLs by the acid guanidium thiocyanate phenol chloroform extraction method. 17 Cytokine-specific reverse-transcription polymerase chain reaction (RT-PCR) was performed, as described in detail previously, 18 with minor modifications. Cytokine Assay For stimulation of T cells through the TCR-CD3 complex, culture plates were precoated with 10 g/ml of anti-cd3 mab (clone 145-2C11; PharMingen, San Diego, CA). CD4 T cells isolated from LPLs were added to plates at a concentration of cells/ml in RPMI 1640 supplemented with 10% fetal calf serum, sodium pyruvate, L-glutamine, HEPES, 50 mmol/l 2-mercaptoethanol, 100 U/mL penicillin, 100 g/ml streptomycin, 40 g/ml gentamicin, and 1 g/ml amphotericin B (complete medium) at 37 C in an atmosphere of 5% CO 2 in air. After 48 hours, culture supernatants were subjected to cytokine-specific enzyme-linked immunosorbent assay (ELISA), as described previously. 16,19,20 Serum IFN- levels were also determined by this assay. Transforming growth factor (TGF- ) was detected by luminometry using streptaquorin (Sealite Sciences, Inc., Bogard, GA) with a Microtiter Plate Luminometer ML3000 (Dynatech Laboratories, Chantilly, VA). Detection of Anti Trinitrophenyl Residue Abs Serum and fecal extracts were collected on day 10, and anti trinitrophenyl residue (TNP) Abs were determined by an endpoint titer ELISA. Briefly, microtiter plates were coated with TNP-coupled ovalbumin (0.1 mg/ml), blocked by 1% bovine serum albumin, and incubated with appropriately diluted serum or fecal extract. For detection, peroxidase-labeled secondary Abs against mouse Ig isotypes (Southern Biotechnology Associates Inc., Birmingham, AL) were used. IgG subclasses were determined by use of biotinylated anti-mouse IgG subclass Abs (PharMingen) followed by peroxidase-labeled anti-biotin antibody (Ab) (Vector Laboratories, Inc., Burlingame, CA). Tetramethylbenzidine was used as a substrate for peroxidase. Endpoint titers were expressed as the reciprocal log 2 of the last dilution, which gave an optical density of 0.1 greater than the serum obtained before TNBS enema of individual mice.

3 726 DOHI ET AL. GASTROENTEROLOGY Vol. 119, No. 3 Table 1. Survival Rates of Control and Cytokine-Deficient C57BL/6 Mice After Induction of TNBS Colitis Mice Statistics Dose ( g/g wt) No. of survivors/ total no. of mice (%) C57BL/6 / 25 5/5 (100) C57BL/6 / 40 22/25 (88) C57BL/6 / 50 2/5 (40) IL-12 / 40 17/20 (85) IFN- / 40 19/20 (95) IL-4 / 25 14/15 (93) IL-4 / 40 12/23 (52) a a Significantly lower than other mouse groups given 40 g/g wt. Statistical significance was determined by the Mann Whitney U test using the Statview-J 4.11 statistical program (Abacus Concepts, Berkeley, CA) for Macintosh computers. P values of 0.05 were considered significant. Results and Discussion Sensitivity of Cytokine-Deficient Mice to TNBS Colitis For induction of TNBS colitis in cytokine gene disrupted and control C57BL/6 (C57BL/6 / ) mice, we initially optimized the dose of hapten required to induce colitis. Both mortality and colitis in C57BL/6 / mice were dose dependent, and 40 g/g body wt induced colitis with minimum death (Table 1). This dose was subsequently used in all experiments. Both IL-12 / and C57BL/6 / mice showed similar degrees of survival after TNBS enema treatment, but IFN- / mice tended to be more resistant, although no significant statistical differences were noted. On the other hand, IL-4 / mice showed greater sensitivity to TNBS; only 52% of mice survived TNBS treatment, and the majority of deaths occurred within 4 days after the first TNBS enema (Table 1). Interestingly, all IL-4 / mice that survived the initial dose also tolerated the second TNBS challenge. There were no differences in survival rates between male and female mice in any of the groups tested. Histologic Features of TNBS Colitis When TNBS-treated C57BL/6 / mice were assessed for colitis, we noted a distortion of crypts, loss of goblet cells, infiltration of mononuclear cells, and focal ulcers (Figure 1A). These diffuse, atrophic changes in crypts were accompanied by fibrosis and were similar to those in IFN- / or IL-12 / mice (Figure 1B and C). Histologically, no significant differences were observed in the 3 mouse groups. On gross examination, we observed shortening of colon length, turbid mucosa, and loss of normal longitudinal folds in the lower half of the colon (Figure 2). In contrast, histologic findings in IL- 4 / mice surviving this treatment were different from those of C57BL/6 /, IFN- /, or IL-12 / mice. The lesions tended to be milder and were limited to focal ulcers (Figure 1D). Crypt distortions were found only in areas adjacent to acute ulcers and were not diffuse. Goblet cells were mostly preserved, even in the areas surrounding deep, penetrating ulcers. The thickness of the mucosal layer in IL-4 / mice was increased with reactive hyperplasia of epithelial cells. The atrophic changes seen in IFN- /, IL-12 /, and C57BL/6 / mice were rarely observed in IL-4 / mice. Because death of mice occurred mainly in the early stages of TNBS colitis, we also assessed colitis on day 2 after the first dose of hapten, noting focal ulcers in C57BL/6 / mice. Most ulcers reached the muscular layer, and the mucosal layer bordering the ulceration showed diffuse cell infiltration and distortion of crypts resembling those observed on day 10 (Figure 1E). In contrast, approximately 50% of IL-4 / mice analyzed on day 2 developed large necrotizing ulcer rings (Figure 2), which most likely caused their subsequent death. However, the epithelium adjacent to ulcers was hypertrophic and still maintained goblet cells (Figure 1F). Distortion of crypts was found only at the edge of the ulcer in IL-4 / mice. Because we previously reported hypertrophy of colonic patches in TNBS colitis of IFN- / BALB/c mice, 15 it was important to examine whether these patches developed in C57BL/6 mice. Before TNBS enema, colonic patches were not visible even under examination with a dissecting microscope. After induction of colitis, we observed enlarged colonic patches in C57BL/6 /, IFN- /, and IL-12 / mice (Figures 2 and 3); however, the numbers of colonic patches in C57BL/6 mice were smaller than those in BALB/c mice. No hypertrophy of patches was seen in IL-4 / mice. The extent of colitis on day 10 was assessed by a standard scoring system (Figure 4). Crypt distortion and loss of goblet cells were observed in C57BL/6 / and IFN- / mice, were most diffuse in IL-12 / mice, and were limited in IL-4 / mice. Chronic inflammation was also less severe in IL-4 / mice, and these mice showed minimal histopathologic changes in comparison with other groups, despite the initial high mortality in the early stage of disease. Although distinct types of colitis were observed in IL-4 / and IL-12 / mice, it cannot be simply assumed that normal mice without either IL-4 or IL-12 develop a similar disease. In these gene knockout mice, it is possible that compensation of IL-4 or IL-12 deficiency through other pathways of the cytokine

4 September 2000 TNBS COLITIS EXHIBITS BOTH Th1 AND Th2 PHASES 727 Figure 1. Histologic analysis of TNBS colitis. (A D) Colon samples obtained on day 10; (E and F ) samples obtained on day 2. Results for (A and E ) control C57BL/6 / mice, (B) IL-12 / mice, (C) IFN- / mice, and (D and F ) IL-4 / mice. (Original magnification: A and F,20 ; B E,10.) network results in distinct phenotypes. To address this possibility, normal and cytokine-deficient mice were treated with anticytokine mabs. Treatment of C57BL/ 6 / mice with anti IL-12 mabs resulted in less frequent acute ulcers, although changes in epithelia and chronic inflammation were observed to the same degree as in control, nontreated C57BL/6 / mice. Total blockade of Th1-type regulatory responses by treatment of IL-12 / mice with anti IFN- did not ameliorate chronic inflammation. Although these mice showed dis- Figure 2. Typical macroscopic features of TNBS colitis induced in IL-12 / (middle, killed on day 10) and IL-4 / (bottom, killed on day 2) mice. Normal colon taken from C57BL/6 / mice is seen on top. Dissected colon was opened and fixed with 5% acetic acid in ethanol. IL-12 / mice showed diffuse inflammation with a small ulcer; IL-4 / mice developed a large penetrating ulcer. Arrow indicates an enlarged colonic patch.

5 728 DOHI ET AL. GASTROENTEROLOGY Vol. 119, No. 3 Figure 3. Hypertrophy of colonic patches in TNBS colitis. (A) C57BL/ 6 / mice; (B) IFN- / mice; (C) IL-12 / mice. tortion of crypts and loss of goblet cells, they were less prone to acute ulceration. The significant involvement of IL-4 in the development of colitis was shown by the amelioration of crypt changes and chronic inflammation in C57BL/6 /, IFN- /, and IL-12 / mice treated with anti IL-4 mabs. When IL-4 / mice were treated with anti IFN- mab, all mice survived and were less subject to acute ulcers than were untreated mice at both day 1 (data not shown) and day 10. We also confirmed that the TNBS colitis with blockade of IL-12 or IL-4 induced a similar type of colitis in BALB/c mice as those seen in mice on a C57BL/6 background (Figure 4). T-Cell and Cytokine Analysis To investigate production of cytokines by colonic T cells, we purified colonic lamina propria CD4 T cells from C57BL/6 /, IL-12 /, and IL-4 / mice on day 0 (before TNBS enema) as well as days 1, 3, 5, and 10 and analyzed the expression of cytokine-specific messenger RNA (mrna) (Figure 5A). In comparison with cytokine profiles of naive mice, mrna for both IFN- and tumor necrosis cactor (TNF)- was up-regulated on day 1 in C57BL/6 / mice. On days 3 and 5, no up-regulation of IFN- or TNF- was noted (data not shown). On day 10, mrna for IL-4 and IL-13 was also expressed in C57BL/6 / mice. In IL-12 / mice, there was no upregulation of IFN-, TNF-, or IL-4 on day 1. Messenger RNA for TGF- was expressed in similar levels throughout the course of the disease in these 2 groups of mice. In IL-4 / mice, mrna for IFN- and TNF- appeared on day 1, and at day 10 the cytokine pattern resembled that seen at day 0. The expression of mrna for IL-10, IL-13, and TGF- in IL-4 / mice was variable in 3 independent experiments. These data suggest that cell responses with a Th1-like cytokine profile were initially seen in C57BL/6 / and IL-4 / mice, and Th2-like responses occurred in the late phase of the disease in C57BL/6 / and IL-12 / mice. These early and late cytokine profiles were related to the histologic features (Figure 5B). The low scores of IL-12 / mice and high scores of IL-4 / mice for acute ulcers on day 1 were noted, which paralleled the expression of IFN- and TNF-. To further investigate the characteristics of CD4 T cells on days 1 and 10, purified colonic lamina propria T cells were triggered with anti-cd3 mab, and cytokine production was assessed in the culture supernatants (Table 2). IFN- was not detected on day 1 in IL-12 / mice. On day 10, CD4 T-cell responses of C57BL/6 / mice were higher than on day 1, especially for production of IL-4 and IL-5. This pattern of cytokine secretion was similar to that noted in IL-12 / mice. In IFN- / mice, IL-4 and IL-5 levels were approximately 5 6-fold higher than those in IL-12 / mice. Because IL-12 is produced by antigen-presenting cells but not by CD4 T cells, it was not surprising that purified T cells from IL-12 / mice showed almost normal cytokine responses after stimulation with anti-cd3 mab. However, we cannot assume that this marked difference in levels of IL-4 production between IL-12 / and IFN- / mice also occurs in vivo. CD4 T cells from IL-4 / mice produced levels of IFN- comparable to those produced by T cells from C57BL/6 / mice; however, IL-5 production in IL-4 / mice was less than observed in the other mouse groups. It is unlikely that TGF- or IL-10 producing T cells play major roles in ameliorating the disease in IL-4 / mice, because TGF- and IL-10 levels after CD3 triggering were either minimal or below the limits of sensitivity (15 pg/ml) of the assay used. Serum IFN- Levels in Early and Late Phases of Colitis Because production of IFN- by CD4 T cells on day 1 in C57BL/6 / and IL-4 / mice was associated with focal ulcers, we hypothesized that fatal lesions in early stages of the disease may have resulted from up-

6 September 2000 TNBS COLITIS EXHIBITS BOTH Th1 AND Th2 PHASES 729 Figure 4. Histologic scores of TNBS colitis. Grading was done according to criteria described previously. 15 Each value is shown as the mean with an SEM. The scores for control groups given rat IgG were the same as the scores for untreated mice in each mouse group. Mouse groups indicated with asterisks had lower scores than control groups (white bars) with statistical significance. regulation of IFN- production. To address this possibility, we measured serum IFN- levels in treated mice. When the serum samples on days 1, 2, 5, 7, and 10 from C57BL/6 / mice were examined in a preliminary study, IFN- levels on day 2 were the highest and the least variable among individual mice. Thus, we chose the time points of day 2 and day 10 (when all mice were killed) in addition to day 0 (before TNBS enema). Serum IFN- levels increased sharply on day 2 in C57BL/6 / mice but had decreased by day 10 (Figure 6). Interestingly, IL-12 / mice did not show changes in serum IFN- levels during the course of the disease. In IL-4 / mice, we arbitrarily divided the data obtained on day 2 into 2 separate groups: mice that died before day 7 and mice that survived both treatments. The serum IFN- levels of the former group were significantly higher than those of the latter group. IFN- levels measured before TNBS enema treatment did not vary among the groups. Serum concentrations of IL-4 and IL-5 were also measured; however, serum levels of these 2 cytokines were below the sensitivity of the assay (5 pg/ml for IL-4 and 0.1 U/mL for IL-5) at all time points and in all groups of mice. Ab Responses After TNBS Enema Because Th2-type responses provide effective help for B cells, we measured TNP-specific Ab production (Figure 7). Serum samples were collected from mice at day 10. Total TNP-IgG and IgM Ab titers were not different among these groups of mice. On the other hand, IL-4 / mice had lower IgG1 Ab titers than C57BL/ 6 /, IFN- /, or IL-12 / mice. In contrast, IFN- / mice had a significantly higher IgG1 Ab titer than other mouse groups. Anti-TNP IgG2a Abs were not detected in 3 of 5 or 2 of 5 IL-12 / or IFN- / mice, respectively. There was no significant difference in

7 730 DOHI ET AL. GASTROENTEROLOGY Vol. 119, No. 3 Figure 5. Cytokine profiles and histologic scores in early and late phases of TNBS colitis. (A) Expression of mrna for cytokines in freshly isolated colonic lamina propria CD4 T cells was assessed by RT-PCR. Representative results from 3 independent experiments. Expression of mrna for IL-10, IL-13, and TGF- in IL-4 / mice was not reproduced in each experiment. (B) Histologic scores for days 1 and 10. Scores were determined on days 1 (_) and 10 ( ). Each value is shown as a mean with an SEM. Asterisks indicate statistically significant difference between day 1 and day 10 in each group of mice. Score of IL-12 / mice on day 1 ( # ) was significantly lower than those of C57BL/6 / mice or IL-4 / mice on day 1. Grading was done as in Figure 4. IgG2b Ab titers among the 4 mouse groups. IgG3 Ab titers were low in all groups of mice. Because IL-4 / mice are mostly free from colitis on day 10, the high levels of IgG1 Abs may play a role in the formation of lesions seen in IL-12 / and IFN- / mice. Anti-TNP IgA Ab titers in fecal extracts and serum were either low or were not detected in C57BL/6 / mice as well as in the other cytokine-deficient mouse groups. Our current results clearly showed that pathologically different types of TNBS colitis were induced in situations in which Th1- or Th2-like immune responses were dominant. An increased mortality for penetrating ulcers was seen in IL-4 / compared with IFN- / mice, and amelioration of acute inflammation with anti IL-12 or anti IFN- mabs indicated that Th1-type immune responses were responsible for acute focal ulcers. This finding was also supported by the up-regulation of IFN- and TNF- in lamina propria CD4 T cells on day 1 in C57BL/6 / and IL-4 / mice. IL-4 / mice, in which the colitis proved to be lethal, and C57BL/6 mice had sharp increases in serum IFN- levels on day 2. Because acute ulcers were less severe in IL-4 / mice treated with anti IFN- mab, IFN- likely plays a significant role in the formation of acute fatal lesions. The histologic features of this type of inflammation included penetrating focal ulcers and reactive hypertrophy of the mucosal layer. Epithelial cells were minimally damaged, and goblet cells were preserved. Of interest, IL-12 / mice also developed acute ulcer type lesions on day 10 in addition to diffuse crypt changes with concomitant expression of mrna for IFN- (Figures 4 and 5). We postulate that these changes were caused by compensatory up-regulation of IFN- in IL-12 / mice. Indeed, anti IFN- treatment of IL-12 / mice significantly reduced the scores of acute lesions (Figure 4). On the other hand, the failure of anti IL-12 and anti IFN- mabs to prevent these lesions indicates that the damage to the crypts was not caused by Th1-type immune responses. Surviving IL-4 / mice and C57BL/ 6 / mice treated with anti IL-4 mab developed milder lesions. The cytokine analysis of colonic CD4 T cells and histologic analysis of all groups of mice on day 10

8 September 2000 TNBS COLITIS EXHIBITS BOTH Th1 AND Th2 PHASES 731 Table 2. Cytokine Production by Anti-CD3 mab-stimulated Colonic Lamina Propria CD4 T Cells Th1-type cytokines Th2-type cytokines Mouse groups C57BL/6 / (Background) Day showed that production of IL-4 and IL-5 accompanied crypt damage and chronic inflammatory changes. These results support active roles for IL-4 in diffuse atrophic types of damage in crypts in the later stages of TNBS colitis. Such damage includes the distortion of crypts, loss of goblet cells, and infiltration of mononuclear cells accompanied by diffuse fibrotic lesions that are limited to the mucosal layer. Because the lesions do not extend beyond the layer of muscularis mucosae, fibrosis does not cause stricture, as seen in Crohn s disease, but leads to a loss of crypts often seen in ulcerative colitis. We speculate that stimulation of Th2-like immune responses may interfere with the normal tissue repair processes. Histologic findings of Th2-like inflammation were characterized by regeneration of fibrotic tissue at the site of erosion, which may prevent regeneration of normal configuration of crypts. Continuous stimulation of Th2-like T cells by luminal antigens (and by a second dose of TNBS) may maintain crypt inflammation, cell infiltration, and fibrosis in the mucosal layer. In fact, repeated TNBS enemas weekly for 4 weeks failed to induce focal ulcers but did perpetuate the diffuse changes in crypts and the mucosal layer. Recently, the efficacy of anti IL-12 mab in treating TNBS colitis by regulating apo- Anti- CD3 IFN- (ng/ml) IL-2 (U/mL) IL-4 (pg/ml) IL-5 (U/mL) IL-6 (ng/ml) IL-10 (ng/ml) TGF- (pg/ml) IL-12 / IFN- / IL-4 / NOTE. Colonic lamina propria CD4 T cells prepared on days 0 (before TNBS enema), 1, and 10 were stimulated with anti-cd3 mab for 48 hours, and culture supernatants were assessed for Th1 and Th2 cytokines. Values are shown as the average 1 SEM of 3 independently prepared cell fractions. On day 10, significant increases of all cytokines (except IL-6 and TGF- ) were noted after anti-cd3 mab treatment in C57BL/6 / and IL-12 / mice. The increases in IL-10 levels in C57BL/6 /, IL-12 /, and IFN- / mice were not statistically significant. No increases in TGF- levels were observed. Figure 6. Serum levels of IFN- in TNBS colitis. Serum samples were obtained on days 0, 2, and 10., C57BL/6 / mice;, IL-12 / mice; E, IL-4 / mice that died after day 2; F, IL-4 / mice that survived until day 10. Each point is shown as the mean of 6 17 mice with an SEM. Serum levels of IFN- on day 2 in C57BL/6 / mice were significantly higher than at day 0 or day 10. IFN- levels at day 2 in IL-4 / mice that subsequently died were significantly higher than in their day 0 samples, and also higher than serum IFN- levels from IL-4 / mice that survived both TNBS enemas and were killed on day 10.

9 732 DOHI ET AL. GASTROENTEROLOGY Vol. 119, No. 3 Figure 7. Characterization of TNP-specific Ab titers in serum after TNBS enema. Serum samples were obtained on day 10 after TNBS enema. Each data point shows the Ab titer of individual mice. Open circles show samples in which positive reactions were not detected at the 2 8 dilution of serum. Anti-TNP IgA Ab in serum or secretory IgA in fecal extracts was low or not detected. ptosis was reported 22. This research group has demonstrated TNBS colitis using SJL/J mice with drastic weight loss and severe lesions in the entire layer of the colonic wall, concomitant with Th1-type cytokine responses. 12,14 We assume that the difference from our results using anti IL-12 mab derived from distinct types of colitis due to the different mouse strains and the dose of TNBS. They apparently induced colitis with dominant Th1-type cytokine responses. Nonetheless, in terms of preventing ulcer-type colitis induced with Th1- like cytokine responses by anti IL-12 mab, our results do not conflict with this report. From clinical studies, it is generally recognized that Th1-type immune responses may have a more significant role in Crohn s disease, whereas Th2-type responses may be more involved in ulcerative colitis Initially, TNBS colitis in mice was introduced as a Th1-type inflammation that was mediated by IFN- driven by IL-12 producing antigen-presenting cells using SJL/J mice. 12,14 Mice in this study showed a severe form of colitis that damaged all layers of the colon. On the other hand, studies of TNBS colitis in BALB/c mice showed that Th2-type CD4 T cells played a major role in IBD; in this model, pathologic findings were mainly associated with the atrophic changes of crypts. 15 Furthermore, a recent report showed that the use of oxazolone as a hapten for induction of colitis in SJL/J mice promoted the production of Th2-type cytokines by colonic T cells, resulting in lesions characterized by leukocyte infiltration limited to the superficial layer of the mucosa. 11 Together, these reports suggest that intestinal inflammation associated with Th1-type responses more closely resembles Crohn s disease, whereas Th2-type responses induce lesions more closely resembling ulcerative colitis. This analysis also raises the question of whether the differential immune responses are caused by genetic factors or the nature of antigens encountered in the colon. We have now shown that both types of colitis, colitis resembling Crohn s disease and that resembling ulcerative colitis, are inducible in a single strain of mice using an identical hapten. It indicates that the phenotype of intestinal inflammation in an individual can be modified with dysregulation of T-helper cells with excess Th1 or Th2 responses. However, the initial factors that determine the direction of Th1 vs. Th2 type must be investigated in more detail. In summary, our comparative study of experimental colitis in Th1- and Th2-deficient mice shows that Th1- type responses are involved in acute focal lesions and Th2-type responses play a more important role in diffuse atrophic changes of crypts. These results provide significant new evidence for the differential pathogenesis of Crohn s disease and ulcerative colitis and may also contribute to future immunologic interventions in IBD. References 1. Braegger CP, MacDonald TT. Immune mechanisms in chronic inflammatory bowel disease. Ann Allergy 1994;72: Brandtzaeg P, Haraldsen G, Rugtveit J. Immunopathology of human inflammatory bowel disease. Springer Semin Immunopathol 1997;18: Kühn R, Löhler D, Rennick D, Rajewsky K, Müller W. Interleukin- 10 deficient mice develop chronic enterocolitis. Cell 1993;75: Rudolph U, Finegold MJ, Rich SS, Harriman GR, Srinivasan Y, Brabet P, Boulay G, Bradley A, Birnbaumer L. Ulcerative colitis and adenocarcinoma of the colon in G i2 -deficient mice. Nat Genet 1995;10: Morrissey PJ, Charrier K, Braddy S, Liggitt D, Watson JD. CD4 T cells that express high levels of CD45RB induce wasting disease when transferred into congenic severe combined immunodefi-

10 September 2000 TNBS COLITIS EXHIBITS BOTH Th1 AND Th2 PHASES 733 cient mice. Disease development is prevented by cotransfer of purified CD4 T cells. J Exp Med 1993;178: Powrie F, Leach MW, Mauze S, Menon S, Caddle LB, Coffman RL. Inhibition of Th1 responses prevents inflammatory bowel disease in SCID mice reconstituted with CD45RB hi CD4 T cells. Immunity 1994;1: Mizoguchi A, Mizoguchi E, Chiba C, Spiekermann GM, Tonegawa S, Nagler-Anderson C, Bhan AK. Cytokine imbalance and autoantibody production in T cell receptor mutant mice with inflammatory bowel disease. J Exp Med 1996;183: Takahashi I, Kiyono H, Hamada S. CD4 T-cell population mediates development of inflammatory bowel disease in T-cell receptor chain deficient mice. Gastroenterology 1997;112: Iijima H, Takahashi I, Kishi D, Kim J-K, Kawano S, Hori M, Kiyono H. Alteration of interleukin 4 production results in the inhibition of T helper type 2 cell dominated inflammatory bowel disease in T cell receptor chain deficient mice. J Exp Med 1999;190: Mizoguchi A, Mizoguchi E, Bhan AK. The critical role of interleukin 4 but not interferon gamma in the pathogenesis of colitis in T-cell receptor mutant mice. Gastroenterology 1999;116: Boirivant M, Fuss IJ, Chu A, Strober W. Oxazolone colitis: a murine model of T helper cell type 2 colitis treatable with antibodies to interleukin 4. J Exp Med 1998;188: Neurath MF, Fuss I, Kelsall BL, Stüber E, Strober W. Antibodies to interleukin-12 abrogate established experimental colitis in mice. J Exp Med 1995;182: Elson CO, Beagley KW, Sharmanov AT, Fujihashi K, Kiyono H, Tennyson GS, Cong Y, Black CA, Ridwan BW, McGhee JR. Hapteninduced model of murine IBD: mucosal immune response and protection by tolerance. J Immunol 1996;157: Strober W, Kelsall B, Fuss I, Marth T, Ludviksson B, Ehrhardt R, Neurath M. Reciprocal IFN- and TGF- responses regulate the occurrence of mucosal inflammation. Immunol Today 1997;18: Dohi T, Fujihashi K, Rennert PD, Iwatani K, Kiyono H, McGhee JR. Hapten-induced colitis is associated with colonic patch hypertrophy and T helper cell 2-type responses. J Exp Med 1999;189: Taguchi T, McGhee JR, Coffman RL, Beagley KW, Eldridge JH, Takatsu K, Kiyono H. Analysis of Th1 and Th2 cells in murine gut-associated tissues. Frequencies of CD4 and CD8 T cells that secrete IFN- and IL-5. J Immunol 1990;145: Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 1987;162: Hiroi T, Fujihashi K, McGhee JR, Kiyono H. Polarized Th2 cytokine expression by both mucosal and T cells. Eur J Immunol 1995;25: Fujihashi K, Yamamoto M, McGhee JR, Beagley KW, Kiyono H. Function of TCR intestinal intraepithelial lymphocytes: Th1- and Th2-type cytokine production by CD4 CD8 and CD4 CD8 T cells for helper activity. Int Immunol 1993;5: VanCott JL, Staats HF, Pascual DW, Roberts M, Chatfield SN, Yamamoto M, Coste M, Carter PB, Kiyono H, McGhee JR. Regulation of mucosal and systemic antibody responses by T helper cell subsets, macrophages, and derived cytokines following oral immunization with live recombinant Salmonella. J Immunol 1996; 156: Fuss IJ, Marth T, Neurath MF, Pearlstein GR, Jain A, Strober W. Anti interleukin 12 treatment regulates apoptosis of Th1 T cells in experimental colitis in mice. Gastroenterology 1999;117: Breese E, Braegger CP, Corrigan CJ, Walker-Smith JA, MacDonald TT. Interleukin-2 and interferon- secreting T cells in normal and disease human intestinal mucosa. Immunology 1993;78: Niessner M, Vold BA. Altered Th1/Th2 cytokine profiles in the intestinal mucosa of patients with inflammatory bowel disease as assessed by quantitative reversed transcribed polymerase chain reaction (RT-PCR). Clin Exp Immunol 1995;101: Fuss IJ, Neurath M, Boirivant M, Klein JS, de la Motte C, Strong SA, Fiocchi C, Strober W. Disparate CD4 lamina propria (LP) lymphokine secretion profiles in inflammatory bowel disease. J Immunol 1996;157: Monteleone G, Biancone L, Marasco R, Morrone G, Marasco O, Luzza R, Pallone F. Interleukin 12 is expressed and actively released by Crohn s disease intestinal lamina propria mononuclear cells. Gastroenterology 1997;112: Received April 6, Accepted April 12, Address requests for reprints to: Taeko Dohi, M.D., Department of Gastroenterology, Research Institute, International Medical Center of Japan, Toyama, Shinjuku-ku, Tokyo , Japan. dohi@ri.imcj.go.jp; fax: (81) Supported by U.S. Public Health Service National Institutes of Health grants DK 44240, AI 18958, DE 09837, DE 12242, AI 43197, and AI 35932; and by a grant and contracts from Ministry of Education, Science, Sports and Cultures, the Ministry of Health and Welfare, and Organization for Pharmaceutical Safety and Research of Japan. The authors thank Dr. Giorgio Trinchieri for providing hybridomas for anti IL-12 monoclonal antibodies; Dr. Ed Leiter for discussions of this work and manuscript; Drs. Prosper N. Boyaka and John L. VanCott for lending their expertise with IL-12 / and IFN- / mice; Drs. Casey Weaver and Audrey J. Lazenby for important discussions on histopathology of lesions; Dr. Raymond J. Jackson for help with the TGF- assay; Annette M. Pitts for preparation of the tissues for histology; and Dr. Kim McGhee for editorial advice.