AB02AGA ABSTRACTS 4261 ACTIVATION OF COAGULATION AND FIBRINOLYSIS RE- SULTING IN A HEMOSTATIC IMBALANCE IN THE TREAT- MENT OF ACTIVE ULCERATIVE COLITIS. Adriaan A. van Bodegraven, Marianne Schoorl, Ronald K. Linskens, Jan Pa Baak, Piet Cm Bartels, Hans Are Tuynman, Free Univ Hosp, Amster- dam, Netherlands; Med Ctr Alkmaar, Alkmaar, Netherlands. Background: Activation of hemostasis has been demonstrated to be in- volved in the pathophysiology of ulcerative colitis (UC). Whether this hemostatic imbalance is due to coagulation or fibrinolytic disturbancies is unclear. Methods: Thirty-three patients with active UC were studied by determination of the thrombin-dependent generation of fibrin degradation products (FbDP) as a marker of coagulation and of the plasmin-dependent generation of fibrinogen degradation products (FgDP) as a marker of fibrinolysis. The balance of hemostasis was expressed as the ratio between FgDP and FbDP. Markers of inflammation (ESR, CRP, fibrinogen (Fg) and coagulation (fragment 1+2 (F1+2), thrombin-antithrombin (TAT» were also determined. Severity of disease was assessed by a 18-points endoscopic score, a 4-points pathohistologic score, and a 22-points patient score at baseline and at the third or fourth month of treatment. Reference values were determined in 22 healthy controls. Comparisons were per- formed by ANOVA and Student's T-test. Results: In active disease, all parameters differed significantly from healthy controls. During treatment, UC activity diminished from active to quiescent disease (endoscopically from 12.3 to 5.3; histopathologically from 2.21 to 0.75; patient score from 11.5 to 3.6). ESR, CRP and Fg also decreased significantly in the course of Uc. The hemostatic markers Fl +2, TAT, FgDP and FbDP decreased but not at a level of significance. In quiescent disease, all parameters of coagulation, including FgDP, were elevated in comparison with controls, indicating that activation of coagulation and fibrinolysis was persistent in the course of UC. The ratio between FgDP and FbDP was 0.69 in active UC, and was decreased in comparison with quiescent UC (1.12, p=O.Oll) and with controls (1.36, p < 0.0001). The FgDPIFbDP ratio in quiescent UC did not significantly differ from the findings in healthy controls. These findings are compatible with hypercoagulation in active UC and restoration of the hemostatic balance during treatment. Conclusions: In active UC, hypercoagulation and hyperfibrinolysis were established. The FgDPIFbDP ratio was decreased, compatible with a hemostatic imbalance in favor of hypercoagulation. During treatment of UC this ratio increased and the hemostatic balance restored, notwithstanding persistent activation of co- agulation and fibrinolysis. 4262 THE EFFECT OF 5-AMINOSALICYLIC ACID ON INDUCIBLE NITRIC OXIDE SYNTHASE EXPRESSION AND ACTIVITY IN THE COLONIC EPITHELIAL CELL LINE HT-29. Sean A. Weaver, George Kolios, Karen L. Wright, Stephen G. Ward, Duncan Af Robertson, Univ of Bath, Bath, United Kingdom. Background: Inducible nitric oxide synthase (iNOS) is a pleiotropic en- zyme that has been shown to be up-regulated in the colonic epithelium in inflammatory bowel disease (IBD). It is believed to play a part in the pathogenesis of this condition although its precise role remains uncertain. 5-Aminosalicylic Acid (5-ASA) is a compound commonly used in the treatment of IBD and has been shown to have a multiplicity of possible actions to bring about its beneficial effect. There has been some debate about whether 5-ASA exerts any of its effects via action on the expression or activity of iNOS and this work aims to address this issue Aim: To ascertain whether incubation with 5-ASA alters the expression or func- tional activity of iNOS in the colonic epithelial cell line, HT-29. Methods: Confluent monolayers of HT-29 cells were cultured in the presence of Tumour Necrosis Factor (TNF)-a (lOOng!rnl), Interleukin (Il.j-I« (lOng! rnI) and Interferon (IFN)-r (300u/ml) in the presence or absence of 5-ASA (lOOp,M) added both simultaneous to the stimulation or one hour prior. RNA was isolated over a 24 hour time course and probed for iNOS using Northern analysis. Subsequently HT-29 cells were stimulated as described above in the presence of varying concentrations of 5-ASA (lO-lOOOp,M) or the specific iNOS inhibitor 1400W (l-IOOp,M) for 48 hours. The cell culture supernatants were then flourometrically analysed for nitrite as a marker of iNOS functional activity Results: The described stimulation of colonic epithelial HT-29 cells induced iNOS RNA over 24 hours, peaking at 8 hours. Neither pre-treatment nor simultaneous treatment with 5-ASA at lOOp,M altered this RNA expression. Cytokine stimulation also induced nitrite production which was not significantly affected by 5-ASA at 10 or l00p,M. High dose 5-ASA at looop,M caused non-specific inhibition of all nitrite production, constitutive and inducible. The specific iNOS inhibitor I400W could inhibit induced nitrite production in a dose dependent man- ner. Conclusions: At a dose of 100p,M 5-ASA there is no effect on the expression or functional activity of iNOS. This dose is relatively high when compared to estimates of therapeutic tissue concentrations (5-50p,M). Therefore the beneficial effects of 5-ASA preparations observed in IBD are unlikely to involve alteration of iNOS expression or activity. GASTROENTEROLOGY Vol. 118, No.4 4263 PEROXISOME PROLIFERATOR-ACTIVATED RECEPTOR-f AGONISTS INHIBIT INOS AND RANTES GENE EXPRESSION IN COLONIC EPITHELIAL CELLS. Darran G. Cronshaw, Sean A. Weaver, Stephen G. Ward, Karen L. Wright, Univ of Bath, Bath, United Kingdom. Background: Peroxisome Proliferator-Activated Receptors (PPARs) are a family of nuclear hormone receptors, of which PPAR-r has been shown to promote differentiation and exhibit some anti-inflammatory properties. Thiazolidinedione (TZD) drugs and the naturally occurring 15-deoxy- J 2 (PGJ 2) are PPAR-r agonists and have been shown to inhibit Inducible Nitric Oxide Synthase (iNOS) in a variety of systems outside the gastrointestinal tract. In an intestinal system the chemokine Interleukin (IL)-8 has been shown to be inhibited by PPAR-r activation but further investigation of PPAR-r mediated suppression of inflammatory gene expression in the intestine is required. Aim: To study the effect of PPAR-r activation on the expression of iNOS and also the chemokine RANTES (Regulated on Activation, Normal T cell Expressed and Se- creted) in colonic epithelial cells. Methods: Confluent monolayers of the colonic epithelial cell line HT-29 were stimulated with Tumour Necrosis Factor (TNF)-a (lOOng!ml) and Interferon (lFN)'r (300u/ml) to induce RANTES, or TNF-a, IFN-r and IL-Ia OOng/ml) to induce iNOS with subsequent RNA expression being assayed by Northern analysis. The effect of simultaneous addition of the TZD drug Ciglitazone (l , 3 and lOp,M) or PGJ 2 (lOp,M) on RANTES and iNOS RNA production was then assessed. In addition the effect of these cytokine stimulations on expression of PPAR-r protein was assayed by Western analysis to verify that the amount of receptor remained constant Results: PPAR-r protein was shown to be present in these HT-29 colonic epithelial cells and it was not altered by the cytokine stimulations used. PGJ 2 (at lOp,M) and Ciglitazone (at 3 and lOp,M) both inhibited RANTES RNA production although this was done most effectively by PGJ 2. Similarly both PGJ 2 and Ciglitazone (both at lOp,M) both inhibited iNOS RNA production with PGJ 2 again being the more effective. Conclusions: The PPAR-r agonists PGJ 2 and Ciglitazone inhibited iNOS and RANTES RNA production as measured by Northern analysis in these colonic epithelial cells. This supports the hypothesis that PPAR-r activation may play an important part in down-regulating intes- tinal inflammation and may thus provide a therapeutic target. Furthermore it may be that the natural ligand and Cyclooxygenase 2 product, PGJ 2 , mediates this role in vivo and its defective function may playa pathogenic part in Inflammatory Bowel Disease. 4264 RESILIENCY OF A MODEL HUMAN INTESTINAL EPITHELIUM TO FAS-MEDIATED APOPTOTIC INJURY, EVIDENCE FOR JUNCTIONAL RESTRUCTURING IN ITS REPAIR. Maria T. Abreu, Andrew A. Palladino, Elizabeth T. Arnold, Richard S. Kwon, James A. McRoberts, Cedars-Sinai Med Ctr, Los Angeles, CA; VA Greater Los Angeles Healthcare System, Los Angeles, CA. Background: Intestinal epithelial cell apoptosis occurs continually without apparent permeability defects and is increased in response to intestinal inflammation. The Fas death receptor has been implicated as an effector of crypt epithelial cell apoptosis in mD. Hypothesis: Immune-mediated ap- optosis during inflammation may result in barrier dysfunction of the epithelium. Methods: T84 cells were cultured on solid supports to achieve an electrically resistant monolayer. Apoptosis was detected morphologi- cally by DNA staining and biochemically with a caspase-specific antibody. Transepithelial resistance (TER) was measured using an EVOM and per- meability was measured by mannitol and dextran-Fl'TC flux. Immunoflu- orescent staining (IF) of Fas and junctional proteins was analyzed by confocal microscopy. Results: T84 cells were found to express a high level of Fas which was predominantly basolateral in polarized T84 monolayers. In T84 monolayers exposed to agonist Ab to Fas, basolateral but not apical cross-linking of Fas resulted in T84 cell apoptosis and a loss of 50% of the cells within 24h. Apoptosis was coincident with a decrease in TER to 50% of baseline values. The decrease in TER was associated with increased flux of mannitol but not relatively small macromolecules ::=:3kD. Preservation of barrier function was associated with dramatic rearrangement of tight junctions and desmosomal junctions in apoptotic monolayers. In compar- ison to the compact lacework of junctional proteins in intact monolayers, apoptotic monolayers have a loose network with wide spaces. Triple IF-staining of nuclei, tight junctions and E-cadherin in apoptotic monolay- ers revealed multiple apoptotic and intact nuclei contained within large tight junction outlines but preservation of individual cell-cell contact through E-cadherin interactions. 3-D reconstruction of the monolayers shows flattening of apoptotic monolayers to 40% of control monolayer height. Conclusion: Immune-mediated apoptosis of intestinal epithelial cells may contribute to the permeability defects associated with inflamma- tory conditions of the bowel but the intestinal epithelium is remarkably resilient in the face of apoptosis. The findings in this study may represent a general property of epithelia and the ability to allow for elimination of cells by apoptosis without loss of barrier function by initiating a cell