Upregulation of T-bet and tight junction molecules by Bifidobactrium longum improves colonic inflammation of ulcerative colitis

To the Editor: Manipulation of the mucosal microbiota to reduce the inflammatory potential of colonizing bacteria is an attractive therapy for patients with ulcerative colitis (UC). Therefore, probiotics treatments have been focused on the point of improving its intestinal microbial balance. We read with a great interest the review by Isaacs and Herfarth entitled “Role of probiotic therapy in IBD.”1 In that review, the authors described the effect of 4 weeks of symbiotic treatment (Bifidobactrium longum with Synergy 1) on patients with active UC reported by Furrie et al.2,Bifidobactrium longum (hereinafter BB536) was isolated from feces of a healthy baby in 1969.3 Thereafter, many physiological effects of BB536 have been reported and it has been used commercially for various food applications in several countries. The size of this study with BB536 was small, but there was a significant reduction of proinflammatory cytokines in mucosal biopsies in patients treated with the active therapy as compared to placebo. This finding was very promising for treatment with UC, but the exact mechanism of BB536 on patients with UC was not elucidated. Here we report data on an open-label trial of BB536 for Japanese patients with UC and demonstrate its mechanism of reducing intestinal inflammation in patients with UC. Human study: From 2005 to 2007, 14 patients with UC (5 male, 9 female; mean age 43 ± 5 years) who were refractory to more than 2250 mg of 5-aminosalicylate (5-ASA) were enrolled in this open-label study. The median disease duration prior to BB536 was 10 ± 3.3 years. Six patients of these patients (42.8%) were afflicted with pancolitis and 3 (21.4%) had left-sided colitis; 5 (35.7%) had proctitis. Clinical activity was assessed with the Clinical Activity Index (CAI). The mean CAI score was 7 ± 1.5 before BB536 administration. Patients who scored higher than 5 points in CAI were regarded as active phase and were treated with 2–3 × 1011 freeze-dried viable BB536 for 24 weeks. The clinical response was assessed at 24 weeks after administration of BB536. Clinical remission was defined as a decrease in the CAI to 4 or less. CAI was reduced in 12 of 14 patients at 24 weeks after starting BB536 therapy. Of note, 10 (67%) of 14 patients achieved clinical remission (2 had pancolitis; 3 had left-sided, and 5 had proctitis). The mean CAI significantly decreased at 8, 12, and 24 weeks compared to that before administration of BB536 (Fig. 1). Changes in the CAI during BB536 treatment. Mean CAI significantly decreased at 8, 12, and 24 weeks after treatment with BB536 compared with that before treatment. CAI was compared by Wilcoxon signed-rank test. A P-value < 0.05 was considered statistically significant. Next we investigated the molecular mechanism of BB536 on cytokine production and expression of molecules related to mucosal barrier function. 1) Splenocytes of T cell receptor-α knockout (KO) mice resembling cytokine phenotype of UC (Th2 dominant) were stimulated with CD3/CD28 after pretreatment of heat-inactivated BB536. The production of IL-12, IFN-γ, IL-4, and IL-13 in supernatants was measured by enzyme-linked immunosorbent assay (ELISA). T-bet and GATA-3 mRNA expression was evaluated by reverse-transcriptase polymerase chain reaction (RT-PCR). 2) After inoculation of BB536 (1.5 × 1010CFU/mouse) to wildtype and Toll like receptor 2 KO mice, the gene expression of tight junction molecule (Claudin-1 and ZO-1) in colonic mucosa were analyzed by RT-PCR. Stimulation with BB536 induced the production of both IL-12 and IFN-γ from splenocytes and inhibited the production of IL-4 and IL-13. Stimulation with BB536 resulted in induction of T-bet mRNA expression in splenocytes, while it reduced GATA-3 mRNA expression. Moreover, gene expressions of claudin-1 and ZO-1 was upregulated in colonic tissue of wildtype mice after stimulation with BB536, while these upregulations were not observed in that of TLR-2 KO mice (Fig. 2). Real-time PCR data showed a significant upregulation of gene expression of tight junction molecules (claudin-1 and ZO-1) in colonic mucosa of wildtype mice treated with BB536 compared to nontreated mice, while there was no significant difference of gene expression in these molecules in TLR2 KO mice between treated and nontreated with BB536. The Mann–Whitney U-test was used to analyze changes between 2 groups. A P-value < 0.05 was considered statistically significant. Our pilot study strongly demonstrated that administration of BB536 was well tolerated and effective for inducing remission of patients with UC as Furrie et al2 reported previously. In addition, an ex vivo and in vivo study showed that stimulation with BB536 skewed the Th1-dominant cytokine profile of splenocytes and induced tight junction molecules of colonic mucosa. We have reported that alteration of the Th1/Th2 ratio of peripheral CD4 T cells could be a biomarker of interferon-α and leukocytapheresis therapy for patients with UC.4,5 Shifting the Th1-dominant cytokine profile reflected improvement of patients with UC. Takahashi et al6 reported that immunostimulatory oligodeoxynucleotide (ODN) from BB536 induced the production of IL-12p40 and IFNγ from splenocytes, which was consistent with the present data. Our data also demonstrated the upregulation of T-bet in splenocytes treated with BB536. Taken together, BB536 can shift toward the Th1-dominant cytokine profile via upregulation of T-bet, resulting in amelioration of UC. Barrier function in the intestine is also important for attenuating intestinal inflammation of patients with UC. Therefore, we evaluated the effect of BB536 on gene expression of tight junction molecules (claudin-1 and ZO-1) in colonic tissue because epithelial cells comprise the largest amount of surface area and are sealed at paracellular interfaces by tight junctions. Interestingly, gene expression of claudin-1 and ZO-1 in colonic tissue was significantly upregulated in wildtype mice inoculated with BB536 compared to those with no treatment. Of note, this upregulation was not observed in TLR-2 KO mice. These findings first demonstrate that BB536 induced tight junction molecules via TLR-2, leading to enhancement of intestinal barrier function. In conclusion, our data strongly demonstrated that both alteration of immune balance by upregulation of T-bet and enhancement of mucosal barrier function with upregulation of tight junction molecules could be possible mechanisms of reducing intestinal inflammation in patients with UC. For evaluating the effect of BB536 on patients with UC, further study with a larger number of patients with UC is ongoing.