468 Mutant p53 Inhibits Helicobacter-induced Premalignant Lesions Through Downregulation of Thl Proinflammatory Responses James G. Fox, BarbaraJ. Sheppard,Charles A. Dangler, Mark T. Whary, Melanie Ihrig, MA Institute of Technology, Cambridge, MA; Timothy C. Wang, Univ of MA Medical Ctr, Worcester, MA Backgroundand Aims. P53 is a tumor suppressorgenemutated in many human malignancies, including gastric cancer. It is unclear why humans with p53 mutations don't appear to be at increased risk of gastric adenocarcinoma.Furthermore, the precise relationship between p53 mutations and helicobacter infection has been poorly investigated. Methods. To study the role of germline p53 mutations in modulating the progression to gastric cancer, p53~- and wild type (WT) C57BLB mice were infected with H. fells, and the resulting gastric pathology and immune responsein these two groups of mice analyzed for up to 15 months. The gastric fundus and antrum were evaluatedindependently using a 0-4 scale to score inflammation, parietal and chief cell loss and mucus metaplasia. The latter 3 indices were combined to create one variable representing pre-neoplastic lesions, with the frequency of invasive foci also recorded. An analysis of variance was performed to determine the effects of p53+~ , infection status and post-inoculation time (pi) on inflammation, pro-neoplasticand invasive lesions, mRNA expression for ~,IFN, IL-1, IL-10 and IL-4 was quantified by PCR. Sere were also evaluatedfor H. fells antibody by ELISA. Results. Antral inflammation increased signifi- cantly with time. There was a significant, protective effect on the development of both inflammation and pro-neoplastic fundic lesions due to the p53 mutation (p>0.O5 in each case), tnvasive loci were observed in 9 of 11 of WT infected mice ranging from 13 to 15 months pi; intravascularcarcinomaalso was present in 2 of these mice. Noneof these lesions were observed in thirty, p53+/- mice, infected or not, at any time pi. In sera from WT mice, IgG2a, considered a pro-inflammatory Thl response, continued to rise throughout the 15 month study (p<O.004). In contrast, IgG2a levels of the p53+~- mice were 50 to 60% lower than the WI mice at each time point (p value range of <0.012 to 0.002) and did not progress in magnitude between 2-15 months of chronic H. fells infection (p = 0.167). mRNA levels for ',/IFN and IL-1 were significantly upregulated in WT mice infected with H. fells (p<O.05) but were at background ~eveisin p53*~ mice. IL-10 and IL-4 mRNA were undstectable. Conclusions. Our results support the hypothesis that germline p53 mutations result in a Thl downregulation due to T cell senescence, and thus may offer protection against the develop- ment of gastric cancer and other epithelial-derived,immune-mediatedneoplasms. 461 SUmulaUon of Reactive Oxygen Species within Gastric Epithelial Cells by H. pylor/ is Responsible for the Increase in p53 Protein Levels. Duane T. Smoot, Amel Ahmed, Come, R. Allen, HassanAshktorab, Howard Univ, Washington, DC H. pylori induces cell cycle arrest and apoptosis in gastric epithelial cells. These processes appearto involve p53, which is increased in gastric cells after exposureto H. pylori. H. pylori adherence has also been shown to generate reactive oxygen species (ROS) within gastric epithelial cells. This is associatedwith changes in the activity of oxygen scavengingenzymes within gastric epithelial cells and results in formation of 8-OH deoxyguaoosine DNA adducts, indicating DNA damagefrom ROS. The following experiments were conducted to determine if the generation of ROS in gastric cells by exposureto H. pyloriare responsiblefor the rise in p53 protein. Theseexperimentswere conducted using a gastric carcinoma cell line (AGS). Gastric cells were exposedto cag-pos and cag-neg H. pylori strains for up to 12 hours at a concentration of 200 bacteria/epithelial cell. Gastric cells were rinsed with phosphatebuffered saline (PBS) after 2 hours of exposure to H. pylori, to remove non-adhereot bacteria. The cells were then treated with a non-fluorescent probe DCFDAthat fluoresces after exposure to H202.When exposed to a cag-neg strain there was a 2-fold increase in ROS production within 15 rain that increasedto 3-fold by 30 min and rose to 4-fold by 2 hours then leveled off. However with exposure of gastric cells to a cag-pos strain, there was a 7-fold increase in ROS production within 15 min, that increasedto 8-fold by 30 rain and 9-fold by 2 hours. N-acetyl cysteine (NAG), an antioxidaot, was used to inhibit ROS production. AGS cells pre- treated with NAC for one hour were found to have a minimal increase in ROS generation at 15 min to 44% above that of control cells cultured in medium alone and this remained constant up through 3 hours. In separate experiments gastric cells exposed to H. pylorifor up to 12 hours were analyzed for pB3 protein levels by westarnblot and quaotitstedby scanning densitometry. Exposureto H. pylori increased p53 concentration 2.5-fold after 6 hours and remained elevated through 12 hours of exposure. However, pro-exposure to NAG reduced p53 levels below that of controls after 3 hours and remained reduced through 12 hours. Inhibition of ROS production by H. pyloriin gastric cells preventsthe increase in p53 protein levels associated with bacterial exposure.This supports our hypothesis that the rise in p53 after exposureto H. pylorAs due to DNA damagefrom ROS.These data further support prior studies indicating that H. pylori is an indirect DNA damaging agent. 462 Gene Targeting As A Model To Study The Role Of PepUde YY In Endocrine Differentiation In The GI Tract Susan Schonhoff, Christelle Ratineau,Subir K. Ray, Andrew B. Leiter, New England Medical Ctr, Boston, MA We have previously shown that expression of peptide YY (PRY), is an early event in the differentiation of endocrine cells of the mouse pancreasand colon. PYY is expressed prior to the appearance of other hormones. When each endocrine cell lineagefirst appears during fetal development, PYY is coexpressedin most cells. Later in development,the percentage of cells coexpressing PYY falls for most cell types suggesting that colonic and pancreatic endocrine cells arise from a PYY-expressingprogenitor cell. To understand whether PYY expression is an obligatory step in enteric and pancreatic endocrine cell differentiation, we inactivated the PYY gone by gone targeting in ES cells and generatedmice that transmitted the targetedallelein the germline.Mice homozygous for the targetedallelegrew and reproduced normally. In the targeted allele, most of exons, 2 and 3 which encode PYY, were replaced with a bacterial ~-galactosidase gene containing a nuclear localizationsignal fused in frame with the N terminus of PYY. Thus cells expressing the disruptedgenecould be readilyidentified by histochemical analysis of 13-galactosidase activity, even in the absenceof PYY expression. To ensurethat 1~ galactosidase expressionwas identicalto the endogenousPYY gone without any influence of other sequencesincluded in the targeting vector, the pgk-neomycin cassette was flanked with IoxP sites and removed by crossing the mice with a Cre recombinase expressing strain. To determine whether PYY is required for endocrine differentiation we analyzedendocrinecell populationsin adult PYY - / - mice. The total number of enteroendo- crine cells in the ileum and colon, measuredby cells staining for chromograninA, was similar both in PYY - / - and PYY+/+ mice. We also analyzedspecific endocrinecell populations in null animals by staining for both (.~,-galactosidase activity plus specific hormones. GLP-1 was detectedin the ileum and colon of null mice and co-localizedwith ~-galactosidase.In the pancreataof null mice somatostatinand glucagon expressing cells stainedfor I]-galactosidase activity at the periphery of the islet while insulin positive cells in the interior of the islet did not. Ioterestingly, pancreatic polypeptide (PP) cells were absent in the P Y Y - / - mice. We conclude that PYY expression is not essential for either cell fate determination or terminal differentiation of most enteric or pancreatic endocrine cell types. However, the absence of PP cells suggests that the presence of the PYY locus is required for PP expression. 463 Thyroid Hormone and Gut Development: The Role of the T R y 2 Receptor Variant. Jason N. Badrinarain, Shufen Meng, Beth Israel DeaconessMedical Ctr and Harvard Medical Sch, Boston, MA; Eric Sibley, Rixun Fang, Stanford University; Richard A. Hodin, Beth Israel DeaconessMedical Ctr and Harvard Medical Sch, Boston, MA TRa-2 is a variant thyroid hormone (T3) receptor (TR) that binds to thyroid response elements but does not bind T3, and is thought to inhibit T3 action. Interestingly, endogenousTRo~2 levels in the gut are high during the suckling phase,a time in which the tissue is unresponsive to exogenousthyroid hormone. Giventhe critical importance of 1"3 in regard to gut mucosal development and differentiation, we performed studies to examine the effects of TRa-2 on the T3 target genes, intestinal alkaline phosphatase (lAP) and lactase. METHODS: Caco-2 cells were maintained in hypothyroid media (10% FBS stripped of thyroid hormone by the charcoal/resin method) and transiently transfected with either rat lactase (3.0 kb) or various deletional constructs of the human lAP luciferase (Luc) reporter plasmids. Co-transfections were carried out using the bona-fide T3 receptor, TRy-l, -/+ TRy2. A CMV-~gal plasmid was used to control for transtection efficiency. Cells were then treated - / + 10 nM T3 for 24 hr. All results shown were statistically significant, p<O.01. RESULTS:In the absence of 13, IAP-Luc activity was markedly decreasedby TRy1, consistent with the phenomenonof ligand-independent repression (LIR). Co-transfection of TRy-2 had no significant effect upon the magnitude of LIR. With T3 treatment, IAP-Luc activity was induced (lO-fold) but this effect was dramaticallyinhibited by TRa-2 (78%). Basedupon data using multiple lAP reporter constructs, a major T3 responseelement appearsto reside between-199 and -162 from the transcriptional start site. The ability of TRot2 to inhibit 13 action was then tested in the context of the lactase gene, which is negatively regulated by T3. Lactase-Luc activity was repressed by 1"3 treatment in the presenceof TR,8-1 (61%). In contrast to its effects on the lAP gene,however,TRo,-2 did not inhibit T3-mediatedrepressionof the lactase reporter gone. CONCLUSIONS: TRy2 appearsto inhibit T3-mediatedactivation of the lAP gene. In contrast, TRe-2 has no effects in the absenceof T3, nor does it inhibit T3-mediated repression of the lactase gone. 464 Altered Expression Of Epimorphin In Intestinal Myofibroblssts Affects C~ypt-Villus Morphogenesis Christine Fritsch, ElzbietaA. Swistlicki, Washington Univ Sch of Medicine, St. Louis, MO; Michela Plateroti, Ecole Normale Superieurede Lyon, Lyon France; Michele Kedinger, INSERM U 381, Strasbourg France; Marc S. Levin, DeborahC. Rubin, Washington Univ Sob of Medicine, St, Louis, MO Background:The formation of the crypt-villus axis during gut ontogeny requires continued reciprocalinteractions between the endodermand mesenchyme. The mesenchymalmolecules that regulate this process remain largely unknown. Epimorphin/syntaxin2 (epimorphin) is a mesenchymal protein related to the syotaxin family of vesicle docking proteins, that have been postulatedto play a role in skin, lung and intestinal development.Methods and Results: To elucidate the role of epimorphin in gut ontogeny, the full length epimorphin cDNA was transfected, in senseand antisenseorientations, into an intestinal myofibroblast cell line, MIC 216. Antisense (epFAS) cell lines exhibited a marked inhibition of epimorphin synthesis, whereas sense(epi-S) transfected cell lines showed -4 fold increased epimorphin expression. Epi-AS myofibroblasts induced enhanced proliferation of co-cultured Caco2 ceils, whereas sense-transtected myofibrobiasts induced a rounded morphology in Caco2 ceils, with decreased cell proliferation.Caco2-myofibroblast co-cultureswerealso performedusingTrans- well membranes,to preventcell-cell contact. Despitethe lack of cell-cell interactions, morpho- logic differences among the different epimorphin-Caco2cell cocultures were preserved.To determine the effects of epimorphin on crypt-villus axis formation, rat gut endoderm was combined with epimorphin transfected myofibroblasts and implanted into the chick intracoe- Iomic cavity. The grafts in which epimorphin was overaxpressed revealedwell-formed crypt- villus units, whereas those in which epimorphin was inhibited developed into small cystic structures without crypts or villi. Tenascinexpressionwas markedly increasedin the epimor- phin-overexpressing grafts. Conclusions:Changes in epimorphinexpression in intestinal myofi- broblasts produce profound effects on cellular proliferation and morphology in co-cultured Caco2 cells, and affect crypt-villus morphogenesisin chick intracoelomic grafts. Epimorphin may exert its effects by regulating secretion of morphogenic factors. These results suggest that epimorphin plays an important role in epithelial-mesenchymal interactions during crypt- villus morphogenesis. A-87