Sa1289

Non Celiac Gluten Sensitivity: Lack of Response to In Vitro Gliadin Challengeand Basophils Activation AssayCristina Bucci, Fabiana Zingone, Ilaria Russo, Ivonne Morra, Raffaella Tortora, NorbertoPogna, Giulia Scalia, Paola Iovino, Carolina Ciacci

In recent years it has become more frequent for physicians to come across types of glutensensitivity other than celiac disease (CD). Regardless of the absence of any serological CDmarkers or intestinal damage, gluten sensitive (GS) patients often report intestinal and extra-intestinal symptoms shortly after the ingestion of gluten and the disappearance of suchsymptoms on a gluten free diet (GFD). Regrettably, at present, there is no evidence of anymucosal or serological modifications in GS patients, or that gluten is really the triggerfactor in gluten sensitivity. We aimed at evaluating the impact of gliadin challenge on twoexperimental settings, i.e. in vitro gliadin challenge of duodenal mucosa and peripheralblood basophils in GS patients compared to CD patients and healthy controls. We consecu-tively enrolled patients referred to tertiary centers for food intolerance and celiac disease.Participants underwent a complete clinical screening in order to rule out CD (serum a-tTG,EMA, total IgA and HLA status, where appropriate), an interview for evaluation of diet andsymptoms, a skin prick test to exclude wheat allergy, upper endoscopy for duodenal biopsiesused for both histological assessment and for the in vitro evaluation of the gliadin-inducedmucosal expression of early and late inflammatory markers: PY99, epithelial HLA-DR, ICAM-1, crypt HLA-DR, CD3, CD25 and CD69. In addition, a basophils activation assay, in whichpatients' basophils were challenged with a gliadin extract, was performed. Basophils wereanalyzed for the expression of hematopoietic membrane antigens CD203c, CD63 and CD45by flow cytometry looking at differences in antigen expression. 119 subjects were screenedfor CD and GS and, according to the results, four groups were obtained: GS patients (no.16),CD on gluten free diet (no.34), CD on gluten containing diet (no.35) and healthy controls(no.34). As for the in vitro gliadin challenge, all CD patients, regardless of their dieteticstatus, showed a clear mucosal activation (increased immunofluorescence intensity both forearly and delayed inflammation markers) when stimulated with gliadin, while only 3 controlsand 3 GS patients showed an weak response for some, not all, inflammation markers. Also,there were no significant differences in the skin prick test results or in the level of CD63and CD203c basophils expression, used as a marker of cell activation, among the groups.The present study shows that in GS patients, gliadin challenge test did not show anysignificant modifications in the expression of those mucosal inflammation markers foundincreased in CD patients. Therefore, this tool should not be used as a diagnostic instrumentin case of GS. Moreover, the lack of differences in basophils activation and skin prick testseems to exclude this form of cellular sensitivity in our GS series also.

Sa1290

Utilizing High Throughput Discovery Approach to Identify Optimal CandidateDeamidated Gliadin Peptides for the Identification of Celiac Disease.Joseph A. Murray, Melissa Snyder, Carol T. Van Dyke, Tricia L. Brantner, VasanthJayaraman, Kang Bei, Tianhao Wang, Hari K. Krishnamurthy, John J. Rajasekaran

Celiac disease (CeD) is an immune response to gluten especially to deamidated gliadin. Itis detected by antibodies to tissue transglutaminase (tTg) or deamidated gliadin(DG). Theperformance of DGP antibodies is quite variable and limits their use compared to tTgantibodies. The current DG tests have been based on epitopes recognized by t cells or byuse of random libraries. The sequences developed commercially are trade secrets. Goals:Identify sequences of the deamidated gliadin family (alpha, beta, gamma, omega gliadin)that are most predictive of celiac disease using a high-density in situ synthesized peptidemicroarrays in combinatorial analysis. Methods: Sera from untreated CeD (n=50) werecompared to controls (n=50), both on a gluten-containing diet. A 2-stage process wasutilized. In the 1st stage high-density microarrays with all native and systemically deamidatedgliadin 12 mer overlapping sequences. Each microarray consisting of over 55,000 uniquepeptides with duplicates were in-situ synthesized onto an area of 2.35mm2. The mostinformative 3-mer subsequences were paired and in-situ synthesized into the 2nd phase ofarrays into 6mers in forward and reverse combinations. Using high-density analysis, specific3-mer subsequences paired into novel 6-mers, which provided the highest percentage ofpositives, were identified. A matrix of subsequences and the highest percentage of sequencesshowing IgG and IgA antibody response among all positive samples were then filtered outusing a novel algorithm. A matrix of diagnostic utility was constructed by combining thesequences with the highest occurrence. Novel sequences could then be synthesized withlearning from the first library, and this set is continuously evolving as more samples arerun. Results: With this, it is possible to demonstrate that combining subsequences, eachwith 3 amino acids, in a specific order shows a high accuracy for CeD. (Table) Interestingly,in reversing the subsequences substantially impairs the recognition. This suggests thatantibody recognition of epitopes of DG in CeD is highly sequence specific and only a fewsequences are recognized across the entire spectrum of patients with CeD. Similar sequencesare recognized by IgA and IgG but thus far the IgA isotype is much more sensitive thanIgG (90% vs 68%) Conclusion: A systematic approach to the determination of antibodyrecognition of deamidated gliadin peptides in celiac disease allows for greater precision oftesting. It also identifies potentially immunodominant epitopes in a way independent of tcells epitopes. This mass scale analysis of targeted DG recognition in CeD, is expected tobe further refined as more samples are run. This method is entirely scalable and allows forthe precise interrogation of humoral immunity in subjects with celiac disease and otherautoimmune diseases.Matrix Combination of Gliadin Subsequences

S-253 AGA Abstracts

% accuracy of 6mer combination of 3 mer subsequences

Sa1291

Screening for Celiac Disease in Women With Preterm ChildbirthRaffaella Nenna, Rosalia Ferro, Laura Petrarca, Paola Favata, Federica Lucantoni, MaurizioMennini, Valentina Fiorenza, Matteo Florio, Pisani Valentina, Patrizia Colarizi, ClaudioTiberti, Margherita Bonamico

Celiac disease (CD) is a chronic, immuno-mediated entheropathy caused by ingestion ofgluten in genetically predisposed individuals. It may present with a variety of symptoms(such as diarrhea, weight loss, iron deficient anemia, etc…) or completely silent. Theassociation between CD and pregnancy outcome has been investigated, with contrastingresults. Our aim was to screen for CD women with preterm childbirth or low birth weight.In our prospective study, we performed a CD screening in a group of women with pretermchildbirth (defined as ,37 weeks of gestational age) or with low birth weight ( , 2500 gr),using RIA IgA anti-transglutaminase antibodies (tTGAb). Positive patients were referred toour Operative Unit on Celiac disease, where they were submitted to a further serologicalevaluation. Confirmed positive patients were submitted to upper gastrointestinal endoscopywith multiple duodenal biopsies, and if diagnosed to be celiac, started the gluten-free diet(GFD). A total of 75 women (age range: 19-49 years) participated in the study. Four (5.3%)women resulted tTGAb positive (age range: 20-31 years, gestational age range: 31weeks+6days- 35weeks+ 6 days, birth weight range: 1650-2720 gr). The four women were allprimiparae but one, who already had a preterm child. All the patients were confirmed atthe second serum sample. One of these women performed upper endoscopy with duodenalbiopsies, showing subtotal villous atrophy (3b sec. Marsh modified by Oberhuber) andstarted the GFD. The other three women are still under evaluation, waiting to perform theupper endoscopy. In conclusion, our study demonstrates that CD prevalence is higher inwomen with preterm childbirth or low birth weight rather than in the Italian population.Screening for CD could be advisable as part of the diagnostic flow-chart in these patients.In fact, a timely diagnosis and the prompt GFD, in otherwise asymptomatic patients, couldimprove the pregnancy outcome.

Sa1292

Lamina Propria Lymphocytes Are Positive for IL-15 in (Refractory) CeliacDiseaseSascha Gross, Petula Nijeboer, Chris J. Mulder, Gerd Bouma, Boudewina M. vonBlomberg, Hetty Bontkes

A small fraction of celiac disease (CD) patients does not recover even after strict adherenceto the gluten-free diet (GFD). The diagnosis of refractory celiac disease (RCD) in thesepatients is only possible by exclusion of other enteropathies. In RCD type I (RCDI) patientsthe IEL have a normal phenotype (surface CD3 and CD8 positive), while in RCDII patientsa large proportion of the IEL consists of aberrant cells (surface CD3, CD4 and CD8 negativebut positive for cytoplasmic CD3). Inappropriate activation of intraepithelial lymphocytes(IEL) by enterocyte derived IL-15 is thought to contribute to villous atrophy. Furthermore,IL-15 is thought to be an important growth and survival factor for aberrant IEL found inRCDII patients, and anti-IL-15 is therefore currently being explored as a therapy for RCD.We hypothesised that IL-15 levels may be a useful marker for the diagnosis of RCD. Here,we investigated IL-15 levels in enterocytes of small bowel biopsies of active CD (ACD),GFD, RCD-I and RCD-II patients. The biopsies were separated into epithelial layer containingenterocytes and IEL and lamina propria containing other immune cells including T-lympho-cytes and dendritic cells. As IL-15 is mostly receptor-bound and hardly secreted in a solubleform, we measured IL-15 in cell lysates from freshly isolated cell fractions or by FACS onintact specific cell types. When IL-15 was measured in lysates of the complete epitheliallayer, including IEL, levels were the lowest in RCD-I. Interestingly, after removal of lympho-cytes from the epithelial fraction by CD45-MACS, IL-15 levels were found to be considerablylower. In the lamina propria lysates, the levels of IL-15 were comparable to those of the

AG

AA

bst

ract

s

AG

AA

bst

ract

scomplete epithelial layer. Furthermore, lymphocytes in the lamina propria showed IL-15surface staining by FACS analysis. Our results so far suggest that in CD/RCD patients besidesenterocytes also lymphocytes bear a considerable amount of IL-15. However, IL-15 measuredby ELISA or FACS does not discriminate RCD patients from either ACD or GFD well enoughto serve as a diagnostic marker.

Sa1293

GPR30, a Novel Estrogen Receptor, Enhances Cholesterol Cholelithogenesisby Inhibiting Cholesterol 7α-Hydroxylase (CYP7A1) and the Classic Pathwayof Bile Acid SynthesisOrnella de Bari, Helen H. Wang, Changnyol Paik, Gabriella Garruti, Min Liu, PieroPortincasa, David Q. Wang

We found that a novel estrogen receptor, the G protein-coupled receptor 30 (GPR30), ismapped on mouse chromosome 5 and co-localized with Lith18, a new gallstone gene. Ourmolecular and genetic data support the candidacy of Gpr30 for Lith18. Moreover, thelithogenic effect of GPR30 is independent from that of the classic estrogen receptor α.However, identifying the lithogenic mechanism of GPR30 remains a significant challengebecause it is unclear how estrogen, through GPR30, increases susceptibility to gallstoneformation. Methods: We studied the lithogenic actions of GPR30 on cholesterol gallstoneformation in gonadectomized GPR30 (+/+) mice treated with the GPR30-selective agonistG-1 (200 ng/day) or the GPR30-specific antagonist G15 (900 ng/day), and fed a lithogenicdiet for 8 weeks. To elucidate the metabolic abnormalities underlying the major source ofthe excess cholesterol molecules leading to cholesterol-supersaturated bile as induced byGPR30, we examined whether GPR30 regulates CYP7A1 and the classic pathway of bileacid synthesis through the epidermal growth factor receptor (EGFR) pathway. Results:Inhibition of GPR30 by G15 induced cholesterol-supersaturated bile by relative excesscholesterol in relation to low amounts of bile acids, and promoted gallstone formation. Incontrast, GPR30 activation by G-1 reduced gallstone prevalence. This therapeutic effect wasmediated by GPR30-dependent increases in biliary bile acid concentrations, which restoredcholesterol solubility in bile. After GPR30 activation by G-1, there was a significant dose-dependent increase in mRNA and protein levels of EGFR and liver receptor homolog-1(LRH-1), coupled with elevated CYP7A1 mRNA and protein levels in mice. In contrast,suppressing GPR30 by G15 led to a significant decrease in mRNA and protein levels of

S-254AGA Abstracts

EGFR and LRH-1, coupled with reduced mRNA and protein levels of CYP7A1 in a dose-dependent manner. In vitro studies in cultured mouse hepatocytes showed that when EGFRwas inhibited by AG1478, a highly potent EGFR kinase inhibitor, mRNA and protein levelsof CYP7A1 were unchanged even though GPR30 was activated by its agonist G-1 or repressedby its antagonist G15. Our results indicated that GPR30 regulates Cyp7a1 expression throughthe EGFR cascade. Conclusions: In the lithogenic state, reduced hepatic synthesis of bileacids from the classic pathway, due to GPR30 repression of CYP7A1, significantly inhibitsthe conversion of cholesterol to bile acids. Thus, the balance of biliary lipids for keepingcholesterol solubilized becomes perturbed by a significant increase in the ratio of cholesterolto bile acids in bile, contributing to the formation of cholesterol-supersaturated bile. Thisstudy may provide an efficacious novel strategy for the prevention of gallstones by stimulatingthe GPR30 activity with a liver-specific, GPR30-selective agonist.

Sa1295

Serum MicroRNA: Novel Prognostic Biomarkers in Primary SclerosingCholangitis Patients Treated With High-Dose Ursodeoxycholic AcidYu Li, Nihar Shah, James E. Nelson, Keith D. Lindor, Jayant Talwalkar, Kris V. Kowdley

Purpose: Identification of serummicroRNAs that suggest the progression of primary sclerosingcholangitis (PSC) and predict serious adverse events associated with high-dose ursodeoxy-cholic acid treatment. Background: Primary sclerosing cholangitis (PSC) is a chronic inflam-matory liver disease that may progress to biliary cirrhosis, hepatic failure and liver and bileduct cancer. Investigated as a potential treatment of PSC, high-dose ursodeoxycholic acid(UDCA) was shown to be associated with increased risk of liver-related death or livertransplantation despite improvement in clinical serologic tests among patients. MicroRNAsare increasingly being used as biomarkers for monitoring disease progression, thus wehypothesized that serum microRNA may be useful to predict PSC progression and prognosisof UDCA treatment. To test this hypothesis, we assessed the expression profiles of 376microRNAs in 105 serum samples from a multicenter, double-blind controlled, high-doseUDCA trial study [Lindor et al Hepatology 2009]. Results: A significant increase in numberof cellular microRNAs in serum was observed after Year 3 post-enrollment in both treatmentand placebo groups. Notably, a group of cancer-related microRNAs, such as let-7, miR-126and miR-26a, were significantly up-regulated (p,0.05) in Year 4 and 5 in both groups.The over-expression of these microRNAs may underscore a dysregulated tumorigenesis eventduring the progression of PSC. Interestingly, the treatment- and the placebo- groups canbe segregated by the expression pattern of a different set of microRNAs including miR-125band miR-99a (Figure 1). Bioinformatic analysis on the function of this microRNA set inUDCA-treated patients suggested altered cell proliferation involving the p53 and PPAR αpathways. Furthermore, some UDCA treatment-associated microRNAs demonstrated differ-ent expression patterns between patients who progressed to liver-related death comparedto survivors. This differential expression pattern may provide clues for understanding theworse prognosis of patients under high-dose UDCA treatment. Conclusion: We identifiedchanges in serum microRNA expression patterns over time in PSC patients treated withhigh-dose UDCA, as well as differences among patients with liver-related death comparedto survivors. These data provide mechanistic insights into the effects of high-dose UDCAtreatment in PSC.

Figure 1: Serum microRNA expression patterns segregate patients with high-dose UDCAtreatment or placebo.

Sa1296

Symptomatic Gallstone Disease in the Young: Shifting Trends Towards EarlierAge of Onset in Bronx and New York StateHarish Patel, Manoj Bhandari, Jasbir S. Makker, Bhavna Balar, Sridhar Chilimuri

Background: Gallstone is a common problem affecting those between 20 to 40 years of age,but they do not become symptomatic until late. In our institution we noticed an increasein proportion of teenagers undergoing endoscopic retrograde cholangiopancreatography(ERCP), which can be due to earlier manifestation of the gallstone disease. Recent data frompediatric population has also shown rising prevalence of gallstone disease. This rise can bedue to the increasing prevalence of risk factors like physical inactivity, obesity, diabetes,pregnancy and oral contraceptive use. Methods: A retrospective review of 15-year data (1996- 2010) obtained from SPARCS (Statewide Planning and Research Cooperative System)database of New York State Department of Health was done. Patients with ICD-9 code 574for cholelithiasis and its complication in any of the initial three discharge diagnosis were