Volume 125, Issue 6, Pages 1762-1773 (December 2003) Inhibition of indoleamine 2,3-dioxygenase augments trinitrobenzene sulfonic acid colitis in mice  Gregory J Gurtner, Rodney D Newberry, Suzanne R Schloemann, Keely G McDonald, William F Stenson  Gastroenterology  Volume 125, Issue 6, Pages 1762-1773 (December 2003) DOI: 10.1053/j.gastro.2003.08.031

Figure 1 Induction of IDO mRNA and protein in TNBS colitis. Colitis is induced in the distal colon of SJL/J mice by the administration of TNBS (0.5 mg) in 35% EtOH per rectum. (A ) Real-time PCR for IDO mRNA in the colons of mice 4 days after receiving TNBS in EtOH, EtOH alone, and untreated controls. Each group contained 5 mice, and each sample represents an individual mouse colon. There is an 8-fold induction of IDO mRNA in the colons of TNBS-treated mice vs. controls (P < 0.006). IDO mRNA levels in SJL/J colons after receiving a 35% EtOH enema are not significantly different from untreated controls. There was evidence of colitis only in the TNBS—treated group. (B) Western blot analysis for IDO and β-actin protein expression in colon lysates from untreated mice, mice treated with a 35% EtOH enema, and mice receiving enemas with TNBS in 35% EtOH. Each lane represents an individual mouse colon. A dark band at 42 kilodaltons representing IDO is present only in mice with active TNBS colitis. Control and EtOH-treated animals occasionally demonstrate a faint band representing baseline colonic IDO expression. The Western blot shown is representative of 1 of 3 separate experiments with consistent results. Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)

Figure 2 IDO and quinolinic acid expression in the colonic lamina propria. High-power immunofluorescence of mouse colon frozen sections demonstrates IDO in the cytoplasm of mononuclear cells in the LP surrounding the colonic crypts (red). (A ) Ethanol enema-treated colon at day 4. (B) In TNBS-treated colons at day 4, there is both an increase in cytoplasmic IDO-staining intensity and an increased number of IDO-staining LPMNCs. Nuclear counterstain (blue) is demonstrated in these cells as well as in the crypt epithelium. Quinolinic acid immunohistochemistry of formalin-fixed tissues reveals staining in LPMNCs with a similar morphology to the IDO-staining LPMNCs in B. (C ) Ethanol enema—treated colon day 4. (D) Ethanol enema—treated colon at day 4 of a mouse treated with 1-mT to inhibit IDO. (E ) TNBS-treated colon at day 4. (F ) TNBS-treated colon at day 4 of a mouse treated with 1-mT to inhibit IDO. There is decreased QA staining in the mouse colons treated with 1-mT (D and F ) consistent with diminished IDO activity. Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)

Figure 3 The effect of IDO inhibition on the course of TNBS colitis. (A ) Survival curve of mice treated with TNBS enemas and subcutaneous pellets containing either placebo or 1-mT. (B) Low-power H&E- stained section of the distal portion of a normal untreated SJL/J colon. (C ) Colon section from a mouse 4 days after treatment with a 35% EtOH enema and a subcutaneous pellet containing placebo. (D) Colon section from a mouse 4 days after treatment with a 35% EtOH enema and a subcutaneous pellet containing 1-mT. (E ) Colon section from a mouse 4 days after treatment with a TNBS enema and a subcutaneous pellet-containing placebo. (F ) Colon section from a mouse 4 days after treatment with a TNBS enema and a subcutaneous pellet containing 1-mT. (G) Colon section from a mouse 6 days after treatment with TNBS + placebo. (H ) Colon section from a mouse 6 days after treatment with TNBS + 1-mT. After day 3, mice receiving TNBS and 1-mT have significantly more mucosal injury compared with those receiving TNBS and placebo as evidenced by an increased transmural inflammatory infiltrate, a complete loss of mucosal architecture, and the development of a progressive loss of vascularity resulting in necrosis. Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)

Figure 4 IDO inhibition leads to toxic colonic dilation in the setting of TNBS colitis. Representative colons from untreated mice and from mice receiving a TNBS enema plus subcutaneous pellets containing either placebo or 1-mT are shown. Animals were killed 4 days after the administration of TNBS. Five out of 7 (71%) mice receiving TNBS and 1-mT developed visible colonic dilation with stool retention by day 6 as compared with zero out of 9 (0%) among those receiving TNBS and placebo (P < 0.005). Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)

Figure 5 Proinflammatory cytokines expressed within the colons of mice treated with TNBS and placebo or 1-mT real-time PCR were used to quantify the mRNA expression of various cytokines in colon lysates from mice treated with TNBS plus either placebo or 1-mT. The mice were killed 4 days after TNBS administration. The data are expressed as fold increase over untreated control animals. The error bars represent standard deviations. There is significantly higher expression of IL-12, IFN-γ, IL-2, in the TNBS- and 1-mT—treated mice vs. TNBS- and placebo-treated mice. IL-1β expression is not significantly different when comparing the 2 TNBS-treated groups. There is no significant change in TGF-β expression when comparing control and TNBS and placebo or TNBS and 1-mT treated animals. Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)

Figure 6 IDO expression in colonic LPMNCs. LPMNCs were isolated, magnetically fractionated, and cultured in the presence and absence of rIFN-γ as described in the Materials and Methods section. Western blots were then performed to assess IDO protein expression, and the cellular composition of each fraction was assessed by flow cytometry using anti-CD11c (dendritic cells) and F4/80 (macrophage) antibodies. Weak basal IDO expression was seen in unfractionated colonic LPMNCs, which increases following stimulation with rIFN-γ. The B220− /MHCII+ fraction enriched for professional APCs expresses moderate amounts of IDO at baseline, which increases after incubation with IFN-γ. There is minimal or undetectable baseline IDO protein expression in the B220−/MHCII− fraction, which is moderately induced by incubation with IFN-γ. IDO protein is not detectable in the B220+ population but can be induced to a minimal extent by rIFN-γ. The results shown are representative of 3 independent experiments with consistent results. Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)

Figure 7 Decreased IDO expression in STAT-1−/− LPMNCs and increased inflammation in STAT-1−/− mice in response to TNBS. Colonic LPMNCS from C57BL/6 and STAT-1−/− mice were isolated and cultured for 24 hours in the presence of rIFN-γ and LPS and examined for IDO expression by Western blotting. (A ) Colonic LPMNCs from C57BL/6 mice demonstrated basal IDO expression, which could be slightly increased by incubation with LPS and significantly induced by rIFN-γ. (B) Colonic LPMNCs from STAT-1−/− mice had reduced basal IDO expression, and IDO expression could not be appreciably augmented by culturing with LPS and/or rIFN-γ. (C ) Low-power view of a colon section from a C57BL/6 mouse 4 days after treatment with a 35% EtOH enema. (D) Colon section from a STAT-1−/− mouse 4 days after treatment with a 35% EtOH enema. (E ) Colon section from a C57BL/6 mouse 4 days after treatment with 2 mg TNBS. (F ) Colon section from a STAT-1−/− mouse 4 days after treatment with 2 mg TNBS. Note the extensive mucosal ulceration and crypt destruction in the STAT-1−/− mouse as compared with the focal ulceration in the WT mouse. The histologic scores were 4.6 ± 0.5 and 3.6 ± 0.5 for STAT-1−/− and control mice, respectively (N = 5 per group,∗P = 0.02). Gastroenterology 2003 125, 1762-1773DOI: (10.1053/j.gastro.2003.08.031)