April 1998

• G4114

SELECTIVE INHIBITION OF TNF-(x-MEDIATED COX-2 BUT NOT CONSTITUTIVE COX-1 GENE EXPRESSION IN HT-29 CELLS USING A SPECIFIC NF-t:B INHIBITOR. C. Jobin, O. Morteau, DS. Han, R.B Sartor. Center for GI Biol. & Dis., Univ. of N. Carolina, Chapel Hill.

Introduction: Prostaglandins influence gastrointestinal fluid secretion and have cytoprotective on the gastrointestinal mucosa. Cyclooxygenase (COX) is the key regulatory enzyme of the prostaglandin/eicosanoid pathway. The two COX isoforms include the constitutively expressed COX-1 and the inducible COX-2. The COX-2 gene promoter contains a ~:B binding site for the transcription factor NF-t:B but its role in COX-2 expression in intestinal epithelial cells (IEC) is unclear. NF-~:B activity is controlled by the cytoplasmic inhibitor protein IkB. We recently reported inhibition of proinflammatory gene expression in IEC by an NF-~:B super-repressor delivered by an adenoviral vector Ad51t:B (Jobin et al, J. Immunol 1998). Aims: i) Investigate the role of NF-t:B in COX-1 and COX-2 gene expression after TNF-a stimulation of HT-29 cells and 2) determine if inducible and/or constitutive prostaglandin production could be inhibited by the NF-rB super repressor. Methods: Uninfected or Ad51KB or Ad5LacZ- (13-galactosidase; control virus) infected HT-29 cells were stimulated with TNFa (30ng/ml) or media alone for 0 to 24 h. COX-1 and COX-2 gene expression were analyzed by RT-PCR and Western Blot and secreted PGE 2 was quantified by radioimmunoassay. N F - K B activity was evaluated by p65 immunofluorescence and by electrophoretic mobility shift assay. The transfection efficiency was monitored indirectly by X-Gal staining and directly by 1KB immunofluorescence using an antibody specific for the boB mutant. Results: The NF-t:B super-repressor was efficiently delivered into IEC. COX-1 mRNA was constitutively expressed in HT-29 ceils with no apparent up-regulation following TNF-ct stimulation. COX-2 mRNA expression was undetectable in unstimulated cells but was strongly up-regulated by TNFa stimulation at 3 and 6h. COX-2 mRNA up-regulation was totally abolished in Ad51KB-infected HT-29 cells but not in Ad5LacZ- infected control cells. Ad51rB did not affect COX-I mRNA constitutive expression. TNFa increased PGE 2 production in both uninfected (<10 fold) and Ad5LacZ-infected HT-29 cells (<20 fold) but not in Ad51~B-infected cells (<2fold) in agreement with the COX-2 mRNA findings. NF-t:B activity was inhibited in Ad51KB-infected HT-29 cells but not in Ad5LacZ-infected ceils. Conclusion: NF-KB activation is critical in mediating COX-2 gene expression in HT-29 cells. Blocking NF-t:B activity prevents the induction of COX-2 expression and of PGE 2 production by TNF-a, but has no effect on COX-1 mRNA. Selectively inhibiting COX-2 expression with the NF-KB super- repressor may be useful in understanding the role of prostanglandins in intestinal inflammation and provides greater specificity than pharmacologic inhibitors.

• G4115

TNF~ SIGNALING BLOCKADE BY A DOMINANT-NEGATIVE TRAF-2 DELIVERED BY ADENOVIRAL VECTOR INHIBITS NF-~:B AND JNK ACTIVATION AND IL-8 EXPRF_SSION IN HUMAN INTESTINAL EPITHELIAL CELLS. C. Jobin, J-L. Klein, C. Brandham, K. Streetz, D.A. Brenner, R.B. Sartor. Center for GI Biol. & Dis., Univ. Of N. Carolina, Chapel Hill.

Introduction: Cytokine signaling involves the participation of many adaptor proteins including the docking proteins TRAF-2 and TRAF-6, which signal through TNF~ and IL-113 respectively. Following TNFcx cell stimulation, TRAF-2 associates with the NF-KB-indncing kinase (NIK) which triggers a cascade of events leading to the phosphorylation and degradation of 1KBcx, which releases NF-~:B. Aims: We characterized the effect of a dominant negative TRAF-2 protein (Ad5dn TRAF-2) on the TNF~-signaling cascade and investigated whether inhibiting this pathway could efficiently block TNF-(x induced 1L-8 gene expression in intestinal epithelial cells (IEC). Methods: The dn TRAF-2 gene was delivered by either adenoviral infection or lipofectamine. The effect of dn TRAF-2 on IL-8-1uciferase gene expression was studied in NIH 3T3 cells after IL-113 (10ng/ml) or TNF-~ (10ng/ml) stimulation. HT-29 and IEC-6 cells were infected with Ad5dn TRAF2 or Ad5LacZ (13-galactosidase; control virus) for 4-12h using various multiplicities of infection (moi). IL-8 gene expression was analyzed by RT-PCR, Northern Blot and ELISA. JNK activity was assayed by in vitro phosphorylation of a GST-c-Jun fusion protein. 1KB steady-state level was analyzed by Western blot and NF-~:B activity was evaluated by immunofluorescence and electrophoretic mobility shift assay. Results: A high efficacy and level of Ad5dn TRAF-2 gene transfer was obtained in IEC using an moi of 50. dn TRAF-2 transfection specifically inhibited TNF-a but not IL-lfl induced IL-8-LUC reporter gene expression in NIH 3T3 ceils. In addition TNF-ct but not IL-113 induced l~:Ba degradation and NF-~:B activation was prevented in Ad5dn TRAF-2-infected NIH 3T3 and IEC-6 cells. Induction of IL-8 expression by TNF-a was strongly inhibited in Ad5dn TRAF-2-transfected HT-29 but not in control Ad5LacZ-infected cells. Surprisingly, IL-113-mediated IL-8 gene expression was also inhibited i n HT-29 cells as measured by Northern blot and ELISA. In addition, TNF-a induced JNK activation was also inhibited in Ad5dn TRAF-2-infected HT-29 cells as demontrated by GST-cJun phosphorylation. Condusion: The Ad5dn TRAF-2 blocks the TNF-a signaling pathway in IEC by inhibiting both

Tmmunology, Microbiology, and Inflammatory Disorders A1005

NF-tcB and JNK activation. TRAF-2 protein may also be involved in a novel IL-lfl signaling pathway in HT-29 cells. Manipulation of cytokine signaling pathways represents a new approach for inhibition of proinflammatory gene expression in IEC.

64116

SYMPTOMATIC SACROILIITIS IN CROHN'S DISEASE. HW. Jones, P Gordon, J Hibbert, T Gibson, JD Sanderson. Guy's Hospital, London, UK.

The published incidence of sacroiliitis in Crohn's disease (CD) varies from 11-30% depending on mode of imaging. Males predominate and many patients are asymptomatic. The aim of this study was to examine the incidence of sacroiliitis in patients with CD with symptoms of backpain. 110 consecutive patients (57 female, 53 male) with CD attending an inflammatory bowel disease clinic completed a questionnaire regarding musculoskeletal symptoms. Those reporting backpain or stiffness were offered full clinical evaluation including x-ray and CT scan of sacroiliac joints. Of the 110 patients, 54 (49%) reported backpain. 27/54 (50%) had sacroiliitis confirmed on CT scan. There were no cases of x-ray sacroiliitis with normal CT. In contrast, x-rays were considered normal or equivocal in 14 (52%) of cases abnormal on CT. Sacroiliitis was more common in males (9F:18M, X2=3.97;p<0.05). An associated peripheral arthropathy was present in 10 cases (37%). CT defined sacroiliitis correlated closely with clinical features of axial early morning stiffness of greater than 30 minutes (r=0.502;p<0.001) and sacroiliac tenderness (r=0.458; p<0.001). There was no correlation between ESR, CRP or site of disease and sacroiliitis. Amongst the 27 cases of CD without sacroiliitis, lumbar spondylosis and apophyseal osteoarthrosis were the most frequent causes of back pain. Back pain was common in patients with CD. Sacroiliitis was more common than previously observed and occurred more frequently in men. CT scanning is considerably more sensitive than x-ray in detecting sacroiliitis in CD.

G4117

HYPO-RF~PONSlVE TO CD2-MEDIATED ACTIVATION BY ULCERATIVE COLITIS (UC), BUT NOT CROHN'S DISEASE (CD), MUCOSAL T CELLS. N.E. Joseph, C. Fiocchi, A.D. Levine. Departments of Medicine and Surgery, Case Western Reserve University School of Medicine, Cleveland, OH.

UC- and CD-derived lamina propria T cells (LPT) are activated during chronic inflammation, yet the intrinsic reactivity of these cells is markedly different. While CD LPT proliferate well after IL-2 stimulation, we have reported previously that UC LPT exhibit a decreased response to IL-2 (GE 112:A1021 1997). This study was designed to investigate the mechanism(s) responsible for this difference in proliferation by examining the contribution of the CD2 signaling pathway and IEC co-stimulation to the activation of UC vs. CD LPT. Methods LPT, isolated by negative selection using magnetic sorting from UC, CD and control mucosa, were cultured in the absence and presence of stimulatory anti-CD2 antibodies with and without allogeneic intestinal epithelial cells (IEC). T cell proliferation was evaluated by [3HI thymidine incorporation on day 5. Results When intestinal epithelial cells alone were used to induce proliferation, UC and CD LPT were less responsive than control LPT. When stimulation with anti-CD2 alone was examined, the response of UC LPT was significantly less than that of LPT from control or CD (p < 0.03 and p < 0.015, respectively). UC, CD and control LPT all showed an enhanced response to co-stimulation with IEC and anti-CD2, but the overall response of UC LPT to co-stimulation was 5-fold less than normal or CD LPT (p < 0.01). Conclusion When compared to CD and control LPT, UC LPT are hyporesponsive to activation through the CD2 receptor, regardless of IEC co-stimulation. These results suggest that altered LPT signaling is a defect associated with the pathogenesis of UC and not a non-specific consequence of gut inflammation.

anti- CD2 Stimulation anti. CD2 + IEC

i 5"I 1.3--41100 "i "P < 0.03 ~ i t,~ CO StlmulatlOll • • .,,_o.o, 0,

o ~ NL CD UC NL CD UC