Final Progress Report
Proposal No. IBD-0048
Principal Investigator: Emiko Mizoguchi, M.D., Ph.D.
Applicant Organization: Massachusetts General Hospital (Boston, U.S.A.)
Project Title: Role of tumor necrosis factor receptors of colonic epithelial cells in inflammatory bowel disease
Period of Award: March 1, 2003 - May 31, 2005
A. Aims of this project
The grant proposal of our research project consisted of two aims:
Aim-I: To examine the biological effect of TNFR1 and TNFR2 in experimental models of colitis and human inflammatory bowel disease (IBD) at different stages and conditions.
Aim-II: To define the functional role of a potential inflammatory mediator, Chitinase 3-like-1, in the development of intestinal inflammation.
B. Accomplishments towards meeting those aims
Aim I: By utilizing a DSS-induced colitis model in TNFR-deficient rearrangement activating gene-1 (RAG-1) knockout (KO) mice, we found that TNFR1-signaling cascade in colonic lamina propria (LP) myeloid lineage cells might ameliorate colitis by enhancing innate immune responses through activation of MAPK signaling pathways in CEC.
Aim II: During the first BMRP funding period, we identified that chitinase-3-like-1 (CHI3L1/YKL-40/HCgp39) is specifically upregulated in inflamed colonic mucosa by DNA microarray screening. Expression of CHI3L1 protein was clearly detectable in LP and CEC in several murine colitis models and ulcerative colitis (UC) and Crohn's disease (CD) patients, but absent in normal controls. Functionally, CHI3L1 plays a pathogenic role in intestinal inflammation by enhancing the bacterial adhesion and the invasion of intracellular bacteria such as Salmonella typhimurium in CEC. Inhibition of CHI3L1 activity would be a novel therapeutic approach for intestinal inflammation.
C. A list of significant results
I-1. Absence of TNFR1 significantly influences the severity of DSS-induced colitis. Transplant of the RAG-1 KO mice derived BM cells rescued the lethal damage from RAG x TNFR1 DKO mice. It has been demonstrated that TNFR2 on lamina proprial T cells plays a critical role for the pathogenesis of CD and promotes experimental colitis. To further examine the role of TNFRs expressed on non-lymphoid cells, we generated RAG-1 deficient-TNFR1 DKO and -TNFR2 DKO mice. After the administration of 4% DSS for five days, more than 85% of RAG x TNFR1 DKO mice, which are able to constitutively express TNFR1 on CEC and LP cells, survived even after seven days cessation of DSS (DSS day 12). In contrast, 90% of RAG x TNFR1 DKO mice, which can express TNFR2 on CEC and LP cells after DSS administration, died before day 12. For further characterizing the TNFRs expressing cell population, which may be involved in the regulation of colitis, BM chimera mice were generated by transplanting RAG-1 KO mice derived BM cells into the irradiated RAG x TNFR1 DKO mice. After transplantation of the RAG-1 KO-derived BM cells that are able to express TNFR1 on DSS days two and four, and 90% of the recipient RAG x TNFR1, DKO mice were rescued from the lethal damage of colitis. From the result, the TNFR1 expressed on LP cells, which are RAG-1 KO mice derived myeloid lineage cells including macrophages and dendritic cells but not T cells and B cells, is a critical regulatory factor for recovering from DSS-induced lethal colitis.
I-2. Prolonged activation of p42/p44 MAPK of CEC during the course of DSS colitis after BM transplanted to RAG x TNFR1 DKO mice. TNFR-mediated signaling cascades have been demonstrated to activate NF-κB and MAPK pathways. To examine whether TNFR1 and TNFR2 cooperatively affect the MAPK pathway activation in CEC and LP cells, we analyzed the phosphorylation of p42/p44 MAPK in RAG-1 KO, RAG x TNFR1 DKO and BM transplanted RAG x TNFR1 DKO mice. After the BM transplant, MAPK p42/p44 activation was significantly upregulated in CEC, but not in LP cells, on day six of DSS treatment. There was no significant difference in the phosphorylation of MAPKp38 or c-Jun NH2-terminal kinase (JNK) among the three groups. This result raises the possibility that some factors that are generated by colonic LP cells (myeloid lineage cells originating from BM cells of RAG-1 KO mice) might contribute to the specific activation of a MAPK p42/p44 pathway in the CEC during the early phase of recovering from acute colitis.
I-3. Both TNFR1 and TNFR2 were highly upregulated in CEC of IBD patients. To determine the expression of TNFR1 and TNFR2 in human IBD, samples from patients with colonic disease were subjected to immunohistochemical analysis with anti-TNFR1- or TNFR2-monoclanal antibody. TNFR2 expression was found in about 87% of patients with CD (n=19). In addition, TNFR1 expression was also strongly detected in about 80% of CD patients (n=20). Both TNFR1 and TNFR2 were mostly presented in crypt epithelial cells and some scattered positive cells in LP (including intra-epithelial cells and presumably myofibroblasts). In UC patients (n=17), both TNFR1 and TNFR2 were mainly detected in CEC, but less or absent in LP cells. The staining seemed to be more intense than in CD patients. There were less TNFR2 positive cells, but some scattered TNFR1 positive cells in the colonic LP of UC patients. In normal colon (n=12), TNFR2 expression was completely absent, but TNFR1 was expressed both in CEC and LP cells. In infectious colitis (n=4) and diverticulitis (n=7), both TNFR1 and TNFR2 expression were moderately upregulated in CEC and LP cells. Phospho-MAPKp42/p44 was significantly activated in parallel to the severity of inflammation in UC and CD, but less significantly upregulated in normal individuals or diverticular disease patients. All the patients with CD with fistula (n=5) or history of perforations (n=3) significantly upregulated expression of TNFR1 and TNFR2 as well as MAPKp42/p44 activation in CEC.
II-1. Expression of Chitinase 3-like-1 (CHI3L1) molecule upregulated in colonic mucosa with inflammation. The CHI3L1 upregulation is positively associated with the severity of colitis. DNA microarray and quantitative polymerase chain reaction (Q-PCR) analyses using a DSS colitis model showed that CHI3L1 gene expression was significantly (30-50 fold) increased in the colonic LP cells in DSS day eight compared with DSS day four. Immunohistochemical analysis also confirmed that CHI3L1 expression is significantly upregulated in the colonic LP cells on DSS day eight. In contrast to LP cells, CHI3L1 expression was detectable in colonic surface epithelium, but not crypts on DSS days two and three. We also confirmed that CHI3L1 expression was upregulated in animal models of colitis, including TCRα KO, B cell deficient TCRa KO and IL-10 KO mice models. Furthermore, CHI3L1 mRNA level was increased in active UC and involved region in CD compared to inactive UC, uninvolved region of CD and normal individuals. Immunohistochemical analysis confirmed that CHI3L1 expression in the colon is undetectable or very low in the control group. In contrast, the CHI3L1 expression becomes clearly detectable in the colon of both active UC and involved region of CD patients. Interestingly, CHI3L1 expression was more restricted to crypt and surface epithelium of colon.
II-2. CHI3L1 is constitutively expressed in some human colonic epithelial cells and upregulated the expression by proinflammatory cytokine stimulation. In vitro studies using several human cell lines revealed that CHI3L1 expression is detectable in THP-1 (macrophage cell line) as well as colonic epithelial cell lines including CMT93, Caco-2, SW480, and differentiated T84 cells. In contrast, the expression was not detected in HT29 or COLO205 colonic epithelial cell lines or Jurkat T cell line. As the CHI3L1 expression was induced only under inflammatory but not normal-conditions, we next analyzed the effect of proinflammatory (IL-1β and TNFα), Th1- (IFNγ), and Th2- (IL-4 and IL-13) cytokines for CHI3L1 expression in SW480 and differentiated T84 cells. Proinflammatory cytokines including TNFα and IL-1β at the concentration of 10 ng/ml significantly enhanced the CHI3L1 expression in both SW480 and T 84 cell lines.
II-3. Overexpression of CHI3L1 significantly upregulated the adhesion and internalization of Salmonella typhimurium in colonic epithelial cells. We have examined a biological effect of CHI3L1 in CMT93 and Caco-2 cells (endogenously express CHI3L1) and HT29 cells (lack the endogenous CHI3L1 expression. These cells were transiently transfected with pCDNA4/CHI3L1 or mock (pCDNA4) plasmids and subsequently infected with Salmonella typhimurium. Surprisingly and unexpectedly, transfection with pCDNA4/CHI3L1 compared to pCDNA4 mock transfection significantly upregulated the S. typhimurium adhesion (6.24 fold) and internalization (6.61 fold) after 1h infection in CMT93 cells. HT29 and Caco-2 cells also show increased rate of adhesion and internalization of S. typhimurium. To test the specificity of CHI3L1 expression in bacterial adhesion or invasion ability, CHI3L1 transfected cells were treated with 100 μg/ml anti-CHI3L1 Ab. Preincubation with anti-CHI3L1 Ab reduces the S. typhimurium adhesion and invasion level to 43.4% and 19.7% of initial binding in pCDNA4/CHI3L1 transfected CMT93 cells. These findings suggest that the invasiveness of S. typhimurium is directly or indirectly regulated by CHI3L1.
II-4. Inhibition of CHI3L1 contributes to the suppression of DSS-induced colitis by reducing the internalization of luminal bacteria to colonic mucosa. To examine the functional role of CHI3L1 in vivo, anti-CHI3L1 Ab or control rabbit IgG was intraperitoneally administrated into a DSS-colitis model. Since the CHI3L1 protein induced in acute phase (days 2 and 3) on CEC, and in recovery phase (day 8) on LP cells, we administrated the Ab or control rabbit IgG at the two different time course: days –1, 0, 1, and 2 or days 5, 6, 7, and 8. In both time courses of treatment, anti-CHI3L1 Ab administration enhanced the recovery from acute colitis as indicated by the improvement of body weight loss and clinical score compared to the control group. Interestingly, more effective suppression of colitis was observed in the group (days -1 to 2) in which Ab administration was started one day before starting DSS compared to another group (day 5 to 8) in which Ab administration was started on day 5. In addition, anti-CHI3L1 Ab-administrated group exhibited significantly less epithelial damage and inflammatory cell infiltration on day 12 after the induction of DSS colitis. The anti-CHI3L1 Ab administrated (days -1 to 2) group showed less number of both gram positive and gram negative bacteria as well as acid fast positive intracellular bacteria in the distal part of colon compared to the control group. Furthermore, anti-CHI3L1 Ab administrated (days -1 to 2) mice showed significantly lower bacterial loads in spleen, mesenteric lymph nodes and liver compared to control group on DSS day 6. Taken together, these in vivo studies suggest that the inhibition of CHI3L1 contributes to the suppression of DSS-induced colitis, presumably by reducing the internalization of luminal bacteria to colonic mucosa.
D. Lay summary
Intestinal inflammation including Crohn’s disease and ulcerative colitis was caused by abnormal interaction among host epithelial cells, luminal components (bacteria and food), and the subjacent immune cells just beneath the epithelial layer. The colonic epithelial cells play a central role to maintain the host homeostasis by expressing several important suppressive molecules during the inflammatory conditions. However, some molecules expressed on colonic epithelial cells (CEC) actively exacerbate the colitis. The Principal Investigator (PI) identified that receptors (binding partners) against tumor necrosis factor (TNFR) play a key role for CEC/ lamina proprial interaction. Especially, TNFR type-I is involved in the regulation of colitis by enhancing innate immune responses through activation of cell signaling pathways in CEC. In addition, the PI identified a unique molecule, chitinase-3-like-1 (CHI3L1), which is the enzyme for breaking down chitin (polymer of sugar). Chitin exists on nematode, insects, fungi, crustaceans (shrimp and crab), and house dust mites, but is completely absent in mammals. Therefore, exposures or infections with chitin- or chitin-like structure containing antigens (pathogens) can induce strongly exaggerated immune responses. Under the intestinal inflammatory conditions, CHI3L1 production dramatically increased in CEC. The CHI3L1 binds with luminal bacteria and help their adhesion and invasion into the CEC. The PI produced a specific antibody that binds to the CHI3L1 molecule. This antibody effectively suppressed the excess luminal bacterial adhesion and invasion to CEC and enhanced the recovery from the acute intestinal inflammation in mice. We believe that the CHI3L1 plays an important pathogenic effect during the development of intestinal inflammation and the CHI3L1 Ab administration may effectively suppress the development of colitis.