A814 AGA ABSTRACTS GASTROENTEROLOGY, Vol. 108, No. 4

• CYTOKINE EXPRESSION IN GENETICALLY SUSCEPTIBLE ITY s AND CONGENIC RESISTANT ITY R STRAINS OF MICE FOLLOWING SALMONRIIA DUBLIN INFECTION. L. ~ a n n , J. Fierer, M.F. Kagnoff. Dept. of Medicine, University of California, San Diego, La Jolla, CA

The course and outcome of microbial infections can be determined by the array of eytokines produced by the host. For example, IFN-% TNFot and IL-1 are required for a successful host response to Sa/monella infection in mice. Genes at the Ity locus on chromosome 1 are expressed in macrophages, and are important for determining the outcome of mufine Salmonella infection. In this study; we asked whether Ity alleles govern the outcome of infection with Salmonella dublin by affecting the expression of host eytokines. Methods: Salmonella- susceptible BALB/e (lty s) and Salmonella-re~stant congenic BALB/e flty R) mice were infected with $. dublin i.p. Cytokine mRNA levels were determined by quantitative RT-PCR. Cytokine secretion was assessed by ELISA. Results: BALB/c (Ity s) mice died within 5 days after infection with 102 to 104 S. dublin, whereas BALB/c (Ity R) mice survived infection. Similar increases in mRNA levels for TNFa, IL-la, IL-6, and IL-12p40 were seen in the spleens and livers of both mouse strains after infection. In addition, mRNA levels for IFN-~ and IL-10, cytokines known to eross- regulate each other in vitro, were equally increased in both strains following infection. Five days after infection, spleen cell eultur~ from both strains secreted increased levels of IFN-% ~ 1 0 and IL-6 relative to uninfected controls, and infected mice of both strains had increased serum IFN-? levels, mRNA levels for IL-4, IL-5, TGFgl and g-actin in spleen and liver were decreased, or not altered, in infected mice. Thus, susceptible and resistant strains of mice produce a similar array of cytokines in response to S. dublin infection. Condusinns: BALB/c (Ity s) mice produce an array of cytokines that are required for a successful host response to Salmonella, but expression of these cytokines alone does not determine host survival. The array of eytokines produced in response to S. dublin infection is independent of alleles at the Ity locus. Supl~rtod by NIH grant DK35108 and a CCFA research fellowship.

• INDUCTION OF CYCIX~XYGENASE IN EP1TI~JJAL CELLS IN RESPONSE TO SALMON~I.rA DUBLIN INVASION. L. ~ k m ~ n , M.F. Kagnoff, W.F. Stenson. Depts. of Medicine, Univ. of California, San Diego, CA, and Washington University, St, Louis, MO

To initiate systemic infection, invasive bacteria enter, and pass through, intestinal epithelial cells. One response of these cells to bacterial invasion is increased secretion of chemoattractant cytokines, including IL-8 and MCP-1, which may be important for attracting inflammatory cells to the site of infection. Prostaglandins (PG) are increased during intestinal inflammation and can modulate several inflammatory events. PGs are formed from arachidonic acid through two isoforms of cyclonxygenase (COX), the constitutively expressed COX-I and the cytokine-inducible COX-H, Here we have investigated regulation of COX levels and PG production in epithelial cells in response to bacterial invasion. Methods: Monolayers of human I407 epithelial cells were infected for up to 7 h with the invasive bacteria Salmonella dublin and Yersinla enterocoh'~ca, a noninvasive EschericMa coli, or stimulated with IL-Ifl or TNFa. PGE2 in the supernatsnts was measured by ~LI.qA. COX-I and COX-H mRNA levels were determined by ribonuclease protection assay. Results: Infection of 1407 cells with S. dublin increased PGE2 levels in the supernatants > 30-fold over controls (0.2 ng/ml) after 7 h incubation. IL-IB and TNFa stimulation increased PGE2 levels 3-fold and 2.5-fold, respectively. In contrast, infection with Y. enterocolitica or E. coli did not alter PGE2 production. COX-I mRNA was detectable in conlrol cells, and levels did not increase after co-culture with bacteria, or after eytoldne stimulation. COX-H mRNA was not detectable in control cells, but was expressed in 5. dublin-infected cells, and after H.rl8 or TNFa stimulation. COX-H mRNA levels were proportional to PGE2 levels in the supernatants, with the highest levels being noted 7 h after S. dublin infection. Coneluslens: Bacterial invasion of epithelial cells with & dublin induces COX-H mRNA, and increases PGE2 production. This response may contribute to the pathophysiol0gie events seen after Salmonella infection, since, for example, PGE2 is known to increase CI" secretion by intestinal epithelial cells. Supported by NIH grant DIC35108 and a CCFA research fellowship.

• BUDESONIDE CONTROLLED ILEAL RELEASE (CIR) CAPSULES AFFECT PLASMA CORTISOL LESS THAN PREDNISOLONE

S Edshiicker, P Larsson, M Nilsson, JE Wirdn. Astra Draco AB, Land, Sweden.

Budesnnide CIR-capsules (BUD), 9 mg/day, are as effective as prednisolone (PRED), 40 mg/day, in inducing remission in patients with Crohn's disease affecting the terminal ileum and/or the ascending colon (Farguson et al. AGA 1995). The primary aim of this study was to compare the effect of BUD and PRED on the cortisol plasma concentration vs time curve (AUC0_24h).

Methods: The study was double-blind, randomised and six-period cross-over. Twentyfour healthy volunteers participated. Each treatment period consisted of 5 days with 2 weeks washout between periods. BUD was given once daily in the morning (o.m.) or twice daily (b.iA.) morning and evening in the following doses: 3 mg o.m., 9 mg o.m., 15 mg o.m. and 4.5 mg b.i.d.. PRED, 20 nag, was given o.m. and placebo was given b.i.d.. Blood samples were taken at 3 hour intervals after the last morning dose and plasma cortisol concentrations were measured by HPLC. ANOVA was used for pairwise comparisons.

Results: Cortisol AUC0.uh (% of placebo, mean values [CV%])

BUD PRED

3 mg o.m. 9 mg o.m. 15 mg o.m. 4.5 mg b.i.d. 20 mg o. m.

72.7 [22.2] t~ 44.2 [46.4] tt) 26.5 [81.3] ttt) 45.8 [76.1] 19.7 [59.3]

t) p=0.01 vs placebo t t ) not star. different from 4.5 rag BID t t t ) different from PRED (p---0.01)

The effect on plasma cortisol after o.m. dosing of BUD increased with dose, and 20 mg PRED could be estimated (by linear extrapolation) to correspond to 29 nag BUD (95 % confidence interval 16 to 89 nag).

Conclusion: Budesonide CIR capsules, in clinical doses and excessive (15 rag/day), affected plasma cortisol less than a moderate dose of PRED (20 rag/day).

CD45-POSITIVE CELLS IN FECES: A MARKER FOR DISEASE ACTIVITY IN INFLAMMATORY BOWEL DISEASE J.Egan, K. Salata* J. Cremins*, J.W. Kikendail, P. Nair t, A. Lohani*, J. Hershey*, S. Sharmi t, K. Davis t, R.K.H. Wong. Gastroenterology Service, Division of Clinical Investigation*, Walter Reed Army Medical Center, Washington, DC, Lipid Nutrition Laboratory r, USDA. Beltsville, MD, Fitzsimons Army Medical Center, Denver, CO ~ BACKGROUND: The cellular component of feces can be isolated utilizing a Percoll gradient. These isolated cells, in turn, can be characterized using fluorescein labeled antibodies. CD45 is expressed on the surface of all white blood cells (wbcs). Fluorescein-conjugated antibodies directed against this surface marker can be used to detect wbcs in a fecal specimen. In Inflammatory Bowel Disease (IBD) the mucosal integrity is disrupted, and the number of cells in feces is increased. Our hypothesis was that inflammatory cells in feces would increase in patients with active IBD and this increase would reflect the level of disease activity. METHODS: Fresh human stool was dispersed in buffered saline and fractionated over PercoII/BSA gradients. The harvested cellular component was mixed with fluorescein isothiocyanate (FITC) labeled antibody directed against CD45. The percentage of CD45-positive cells was determined using flow cytometry. RESULTS:

Subjects number CD45 (%)

Normal controls 9 <5

Ulcerative colitis 1 12

IBD (in remission) 4 <5 The percentage of CD45-positive cells is routinely <5 in normal

controls. Similar results are found in those with quiescent IBD. In the single UC patient with active disease 12% of the cells were CD45-positive. CONCLUSIONS: 1) This method can be used to further characterize the cellular component of feces in patients with IBD. 2) The results could aid in assessing the inflammatory activity of the IBD and give insight into pathophysiologic mechanisms.