Microbiota Modulates Host Gene Expression via MicroRNAs

p<0.01, n=4). Since Na+/K+-ATPase provides the favorable Na+ gradient for Na+-dependent cotransport processes, activity of this enzyme was determined by measuring the Pi level in cellular homogenates and it was found to be unaffected (control vs treated: 19.1±1.8 vs 18.0±1.5 nmol/mg protein per min, n=4). Kinetic studies revealed that TNF-α inhibited Na+-Met cotransport by decreasing the number of cotransporter for Met (Vmax: normal vs TNF-α = 2.47 ± 0.16 vs 0.89 ± 0.05 nmol/mg protein per 30 sec, p<0.01, n=3). Preliminary data showed that TNF-α decreased SNAT2 specific mRNA abundance and protein expression in TNF-α treated cells. Conclusion: We conclude from these results that Na+-dependent Met cotransport is present in the IEC-6 cells and is significantly inhibited by TNF-α treatment. This inhibition is not due to an alteration in Na+/K+-ATPase activity rather due to a decrease in the number of Met cotransporter.

Mo1848

Background: Increasing evidence implicates gut flora in colitis and colonic cancer. We are studying the role of enterotoxigenic Bacteroides fragilis (ETBF) in colon tumorigenesis via a multiple intestinal neoplasia (Min) mouse model. The only known ETBF virulence factor, B. fragilis toxin (BFT), is a zinc-dependent metalloprotease. ETBF produce 1 of 3 related BFT isoforms. The reference wild-type (WT) laboratory strains VPI-13784 (VPI), 8654432-2 (86), and Korea 570 (570) produce BFT1, BFT2 and BFT3 respectively. Due to strainspecific differences we observed in degree of colitis induced in WT mice, we hypothesized that ETBF virulence may be BFT isoform dependent. Aims: 1. To compare the virulence of ETBF strains 2. To assess contribution of bacterial host genetic background to BFT virulence Methods Specific pathogen-free C57Bl/6 Min mice were gavaged with 1-5 X 108 CFU of bacteria and colons harvested after 14 or 60 days. B. fragilis were cultured from mouse feces and colon tissue homogenates to assess gut colonization and mucosal adhesion respectively. BFT-induced inflammation was assayed via mucosal IL-17A expression using qRTPCR. Results Min mice infected with ETBF VPI, 86, and 570 for 2 months developed significantly more colon adenomas than sham controls (Table). Colonization, mucosal adherence and fecal bft expression was comparable between strains. To further evaluate the role of BFT in ETBF virulence, we constructed isogenic bft deletions in VPI and 86. Compared to WT parental ETBF strains, mouse infection with bft mutants failed to increase polyposis above controls. Mutant colonization and adherence was comparable to WT. These data indicate that BFT is necessary for tumor induction and suggest that 86, through BFT2, is more virulent than other ETBF. Consistent with this hypothesis, ETBF 86 induced significantly higher levels of cecal IL-17A expression than 570 and VPI (Table). To further evaluate the contribution of BFT2 to tumorigenesis and the potential impact of bacterial host strain background, non-toxigenic B. fragilis (NTBF) NCTC 9343 (9343) and gut symbiont B. thetaiotamicron (Btheta) were transfected with a plasmid expressing bft2 from 86 to generate strains 9343-bft2 and Btheta-bft2. Whereas polyp numbers in Min mice infected with NTBF 9343 were similar to sham controls, infection with 9343-bft2 and Btheta-bft2 lead to marked colitis and polyposis comparable to 86 infected mice. Conclusions: 1. There are strainspecific variations in ETBF virulence likely mediated by the BFT isoform expressed by the ETBF strain 2. BFT expression in avirulent Bacteroides spp. is sufficient for virulence transformation suggesting that host background has minimal effect on BFT action Additional studies to understand the structural characteristics of BFT and impact of isoform on colon tumorigenesis are warranted. Table: Relative virulence of WT ETBF Strains

Mo1846 A Study of the Prevalence of Genetic Polymorphisms in Bile Acid Diarrhea Patients Sanjeev S. Pattni, Shivani Pathmasrirengan, Peter H. Dixon, Catherine Williamson, Julian R. Walters Background and Aims: Bile acid diarrhea (BAD) is a common cause of chronic diarrhea and results from high concentrations of bile acids in the colon. It has recently been suggested that impaired production of Fibroblast Growth Factor 19 (FGF19) in the ileum can result in excessive hepatic bile acid synthesis through reduced negative feedback regulation, and this, rather than malabsorption, is the cause of the diarrhea. The reason for the reduced ileal FGF19 production remains unclear but it has been established that ileal FGF19 synthesis is markedly stimulated by bile acids binding to the farnesoid X receptor (FXR). In the liver, FGF19 acts through the FGFR4 receptor and its co-receptor Klotho β. Organic Solute Transporter α (OSTα) is the ileal basolateral bile acid transporter. The aim of this study was to determine the prevalence of 6 common single nucleotide polymorphisms (SNPs) in 4 genes involved in bile acid metabolism and FGF19 action, comparing the prevalence in patients with primary BAD, a similar number of controls with diarrhea, and a larger numbers of healthy controls. Methods: Two common SNPs (rs948992 and rs1789170) in the FGF19 gene were analysed as well as 2 SNPs (M173T, rs61755050 and -1g>t, rs56163822) in the FXR gene. Mutations in FXR are implicated in patients with obstetric cholestasis and have been shown to reduce function and translation efficiency of FXR. A SNP in Klotho β (rs17618244 ) has been associated with colonic transit in IBS-D patients. A further SNP in the OSTα gene (rs939885) was also studied. Genomic DNA was extracted from 92 primary BAD patients diagnosed on the basis of an abnormal (<15%) SeHCAT test and 75 diarrhea controls. The frequencies of some of these SNPs in our local healthy population has been previously reported. Genotyping of the SNPs was performed using Taqman allelic discrimination assays. Results: Polymorphisms in all the genes studied were detectable in controls and in cases. No significant differences in the genotypes and allele frequencies were found in the Caucasian cohorts in the control and primary BAD groups (p > 0.1). Further analysis of patients of different ethnic origin again did not show significant differences in the allele frequencies. Conclusion: Major differences in the frequency of these genetic polymorphisms in 4 candidate genes involved with FGF19 and bile acid metabolism are unlikely to be associated with primary BAD. However, the SNPs identified may contribute small effects, with a combination of SNPs in several genes producing a cumulative effect. Larger numbers of cases are required to define specific interactions or minor effects. At present the cause of the reduced FGF19 action in BAD remains uncertain. Mo1847

Mo1849

Polyethylene Glycol (PEG) Diminishes Pathological Effects of Citrobacter rodentium Infection by Blocking Bacterial Attachment to the Colonic Epithelia Wentao Qi, Suhasini Joshi, Christopher R. Weber, Ramesh K. Wali, Hemant K. Roy, Suzana D. Savkovic

Microbiota Modulates Host Gene Expression via MicroRNAs Guillaume Dalmasso, Hang Thi Thu Nguyen, Yutao Yan, Hamed Laroui, Moiz A. Charania, Saravanan Ayyadurai, Shanthi V. Sitaraman, Didier Merlin

Introduction: Enteropathogenic bacteria (EPB) remain one of the most common human health threats, despite the use of existing vaccines and antibiotics. EPB infections are extensive in underdeveloped countries (e.g. recent outbreak of Cholera in Haiti), but also occur in developed countries due to food processing and vast global travel. Furthermore, an emerging concern is increasing resistance to antibiotics, and thus, alternative approaches to prevent EPB infection are needed. EPB can utilize a number of mechanisms to infect the host cells, however establishing a close physical contact and interaction to the host cell surface receptors is first step. We have shown that polyethylene glycol (PEG), a polymer of ethylene oxide, downregulates intestinal surfaces receptors. Thus, we hypothesize that PEG may be a low cost and safe means to protect intestinal epithelia from EPB attachment. Aim: To assess the effectiveness of PEG (MW 3,500) in the protection of intestinal epithelia from EPB attachment. Methods: Mouse colonic CMT-93 cells and C57BL/6J mice were infected with murine EPB Citrobacter rodentium (CR). In Vitro, cells were pretreated with PEG (5%, 3 h) and CR infected (5 h) with or without PEG. In Vivo, the following groups were established: control, CR infected (2 wks), PEG (5% daily, 3 wks), PEG (1 wk)+concurrent CR infection (2 wks), and PEG pretreatment (1 wk) followed by CR infection (2 wks). Bacterial attachment and shedding were assessed by growth on MacConkey plates and immunofluorescence. Inflammation was semi-quantitatively scored, and signaling pathways were assessed by immunoblot. Results: Pretreatment or concurrent treatment with PEG inhibited CR attachment to host cells In Vitro, without lowering bacterial growth, suggesting that PEG has preventive properties. Similarly, In Vivo, the number of adherent bacteria to colonic mucosa and the weight of CR infected colons was significantly decreased by concurrent or pretreatment with PEG. These findings correlated with reduced CR-induced inflammation in colons with PEG treatment. CR-induced pathways in colonic epithelia, such as EGFR expression and phosphorylation, were also significantly attenuated by PEG. Furthermore, during infection the presence of PEG lowered bacterial shedding into the stool (by 99%), and this is known to be critical to the spread of bacteria to other hosts. Conclusion: PEG is an effective inhibitor of CR attachment and CR-associated disease. We speculate that PEG may be an efficient strategy to prevent the spread of EPB infections during outbreaks, and may also serve as prophylaxis when traveling to underdeveloped regions.

Background/Aims: Microbiota is known to modulate host gene expression, yet the underlying molecular mechanisms remain elusive. MicroRNAs (miRNAs) are importantly implicated in many cellular functions by post-transcriptionally regulating gene expression via binding to the 3′-untranslated regions (3′-UTRs) of the target mRNAs. However, a role for miRNAs in microbiota-host interactions remains unknown. Here we investigated if miRNAs are involved in microbiota-mediated regulation of host gene expression. Methods: Germ-free mice were colonized with the microbiota from pathogen-free mice for 4 days. Colonic and ileal total RNAs isolated from germ-free and colonized mice were applied for comparative profiling of miRNA and gene expression using miRNA arrays and DNA microarray, respectively. Expression levels of dysregulated miRNAs and their target genes were verified by quantitative RT-PCR. Regulation of target genes by dysregulated miRNAs was verified In Vitro using the murine macrophage RAW 264.7 cell line. Cells were transfected with the miRNA precursors, and target gene expression at mRNA and protein levels were assessed by quantitative RTPCR and Western blot. The 3'-UTR of target mRNA was cloned into the pMIR-REPORT Luciferase vector or the pEGFP-C1 vector, and these constructs were transfected into the cells with the miRNA precursors. Levels of luciferase activity and GFP expression were assessed by luciferase assay and fluorescent microscopy, respectively. Results: MiRNA arrays revealed 1 and 8 miRNAs that were differently expressed in the ileum and the colon, respectively, of colonized mice relative to germ-free mice. A computational approach was employed to predict genes that were potentially targeted by the dysregulated miRNAs during colonization. Overlapping the miRNA potential targets with the microbiota-induced dysregulated genes detected by a DNA microarray revealed several host genes that were regulated by miRNAs in response to colonization. Among them, Abcc3 was identified as a highly potential miRNA target during colonization. Mmu-miR-665, which was dysregulated during colonization, down-regulated Abcc3 expression by directly targeting the Abcc3 3′UTR in RAW 264.7 cells. Conclusion: Our study demonstrates that microbiota modulates microRNA expression, which in turn regulates host gene expression. Thus, our study is the first to explore the implication of miRNAs in host response to microbiota.

S-663

AGA Abstracts

AGA Abstracts

Virulence of Enterotoxigenic Bacteroides Fragilis (ETBF) Strains is B. Fragilis Toxin (BFT) Isoform Dependent Shaoguang Wu, Kenolisa Onwueme, XinQun Wu, Guillermo Ortega, Augusto FrancoMora, Franck Housseau, Drew M. Pardoll, Cynthia L. Sears