A 7 9 2 AGA ABSTRACTS GASTROENTEROLOGY, Vol. 108, No. 4

• IN VlVO DETERMINATION OF THE INTRACOLONIC RELEASE OF IMMUNOGLOBULINS IN CHRONIC ULCERATIVE COLITIS. F. Casellas, M. Papo, F. Guarner, M. Antolfn, R. Jardf, J.R. Armengol, J-R. Malagelada. Digestive System Research Unit and Department of Biochemistry, Hospital General Vail d'Hebron, Barcelona, Spain.

Studies in mucosal biopsies from patients with active inflammatory bowel disease suggest that mucosal IgA/IgG secretory ratio is shifted in favour of IgG. Our objective was to investigate whether immunoglobulin release into the colonic lumen is increased in patients with chronic ulcerative colitis (CUC) and whether the magnitude of the release is related to disease activity. We performed colonic perfusion studies in 25 patients with untreated CUC, and in 7 patients with irritable bowel syndrome (IBS} who served as disease controls. In 7 CUC patients the disease was inactive and in 18 was mild to moderately active according to clinical and colonoscopic Criteria. Colonic perfusion was performed via a double lumen tube placed into the descending/sigmoid colon. An isotonic solution was continuously infused 50 cm from the anal verge at 5 ml/min, and recovered 30 cm distally by siphonage. After 30 rain washout and 30 rain equilibration periods, 10 min effluent collections were obtained for 40 min. Aspirates were analyzed for IgG and IgA by nephelometry and for total proteins by a photometric method.Results: In the 7 IBS controls, total protein release was of 1.8-+0.9 mg/min, IgG release was undetectable in all and IgA was only detectable in 2. In active CUC total protein release was increased to 4.4-+0.7 rag/rain (p<0.05) and both IgG and IgA were detectable in 17/18 patients (IgG: 383-+76 pglmin, IgA: 281 -+68 pg/min) with an IgG/IgA ratio of 1.4. In inactive CUC there was a trend towards lower total Protein release (2.4-+0.7 mg/min, p = 0.07 vs active CUC). Moreover, in the 5/7 inactive CUC patients in who IgG and IgA were detectable, rates were also lower (IGG:129.14-_65 pg/min, IGA:102.1-+42 pg/min,p<O.05 vs active CUC). However, colonic IgG release rates in these inactive C U C patients were significantly higher than in IBS controls (p<O.05). The IgG/IgA ratio remained unmodified in inactive CUC (ratio 1.2, p = ns vs active CUC). Conclusion: Intracolonic IgG release is undetectable in IBS controls. In contrast, CUC patients show sustained release of IgG into the colonic lumen that worsens with disease activity.

ORAL IgA TREATMENT OF CHRONIC NON-SPECIFIC DIARRHOEA IN INFANTS AND CHILDREN Thomas H. Casswall (1), Lennart HammarstrSm (2), Anna Bogstedt (2), Carl Erik Nord (3), B~la Veress (4), Ulrika Brockstedt (4) and Karl Anders Dahlstr6m (1). Dpts of Pediatrics (1), Clinical Immunology (2), Oral Microbiology (3) and Pathology (4), Huddinge University Hospital, Karolinska Institute, Huddlnge, Sweden.

Background Chronic non-specific diarrhoea of infancy (CNSDI) is a common problem in children. We have earlier shown that children with CNSDI had bacterial overgrowth of the small intestine by bacteria from the upper respiratory tract and that the diarrhoea was succesfully treated with Trimetoprim-Sulphamethoxazole .The systemic immune system matures slowly in children which may also include the local mucosal immune system. Therefore we wanted to study the clinical effect of orally given IgA on the diarrhoea and on the possible pathophysiological mechanisms of CNSDI.

Method Between june 1992 and september 1993, 7 children with CNSDI participated in this prospective study. The chi!dren were investigated for malabsorption with routine blood tests including gliadin antibodies and serum immunogiobulins, oral absorption tests, stool analyses and cultures and in some patients breath hydrogen lactulose test for bacterial overgrowth of the small intestine. Small bowel biopsies were taken before and after treatment and were analysed with light- and electronmicroscopy, immunohistechemistry and cultured for aerobic and anaerobic bacteria. The children were treated with IgAbutin® (60% IgA) in a dose of 1000 mg per day divided into two doses during 3 weeks. Fecal IgA was analyzed to estimate the amount of oral IgA passing the GI tract.

Results. Seven children (5 boys and 2 girls) with a median age of 26 months with median of 3.5 loose stools per day and a duration of diarrhoea for a median of 32 weeks were included in the study. The children did not show any signs of malabsorption. Gliadie antibodies and biopsy specimens were normal. After treatment the recorded stools per 24 hours in the children decreased from 3.5 to 1.5, p=< 0.01. Fecal IgA concentration increased in the majority of children. The electronmicroscopic examinations of the biopsies, in 4 patients, reveald supranuc]ear vacuoles, containing fine granular material, dilatation of endoplasmatic reticulum, hyperplasia and dilatation of Golgi apparatus and some microvillous abnormalities. The significance of these findings is unclear. Cultures from small bowel biopsies showed a normal pattern and no bacterial overgrowth, futher Supported by normal breath hydrogen lactulose test in the investigated children.

Di~¢~.~ion. In contrast to our previous study bacterial overgrowth was not found to be the cause of CNSDI. Orally given IgA survived the passage through the GI tract. Clinical improvement was seen, but the mechanism for this is unknown. Our data suggests that oral treatment with IgA may be used for treatment of chronic diarrhoea of infancy.

• PURIFIED SACCHAROMYCES BOULARDII PROTEASE INHIBITS C, DIFFICILE TOXIN A EFFECTS IN RAT ILEUM. I. Casta.qliuolo, J. T. LaMont, C. Baker, D. D. Guzman, S.T. Nikulasson, C. Pothoulakis. Section of Gastroenterology and Department of Pathology, Boston University School of Medicine, Boston MA.

Saccharomyces bou/ardii (Sb), a nonpathogenic yeast, is effective in treating some patients with C. diffici/e diarrhea and colitis. We have previously shown that Sb inhibits C. diffici/e toxin A mediated secretion in rat ileum and that this effect may be mediated by a secreted protease. The aim of this study was to purify and characterize this protease and study its inhibitory action on toxin AI Methods: Sb protease was purified to homogeneity by gel filtration on G-50 Sephadex and OctyI-Sepharose and tested for: a) inhibition of toxin A secretion, epithelial permeability and morpholog.~ in rat

ileal loops in vivo,(n=7-23 loops per group) b) inhibition of-H-toxin A binding to purified rat ileal brush border (BB) in vitro and c) cleavage of the toxin A molecule in vitro. Results: A 54 kDa serine protease was purified by gel chromatography from Sb broth culture. A purification of more than 300-fold with a yield of 0.4% was achieved.

3 Sb protease reduced H-toxin A binding to rat ileal BBin vitro by 42% (p<0.01). Pretreatment of toxin A with purified protease partially digested the toxin A molecule as shown by autoradio~raphy on

. O , SDS-PAGE and by a TCA precipitation technique of H-toxin A. Pretreatment of toxin A with~urified protease in vitro reduced fluid

secretion (by 46%, p<0.01) H-mannitol permeability (by 74.!%, p<0.01 ) and morphologic damage in rat ileal loops in vivo, but had no effect on toxin A-induced cell rounding in cultured fibroblasts. Summary and Conclusions: Sb serine protease inhibits the intestinal effects of C. diffici/e toxin A by cleavage of the toxin A molecule and its intestinal receptor: Our results may be relevant to the mechanism by which Sb exerts its protective effects in C. diffici/e infection in humans. (This research was partially funded by Biocodex Laboratories, Montrouge, France).

• RECOMBINANT HUMAN INTERLEUKIN-11 (rhlL-11) INHIBITS CLOSTRIDIUM DIFFICILE TOXIN A-MEDIATED ENTEROTOXICITY IN RAT ILEUM. I. Casta,qliuolo, J.T. LaMont, C. Baker, S.T. Nikulasson, J.C. Keith, Jr, and Pothoulakis C. Section of Gastroenterology and Department of Pathology, Boston University School of Medicine, Boston MA, and Genetics Institute, Inc., Cambridge MA.

Intedeukin-11 a multifunctional cytokine which stimulates hematopoetic stem cells, was previously shown to decrease acetic acid- induced colonic injury in rats Toxin A from C. difficile, binds to a brush border toxin receptor, causing secretion of fluid, increased intestinal permeability and an intense inflammatory infiltrate in rat ileum. The goal of this study was tO examine whether pretreatment with rhlL-11 would alter the intestinal effects of toxin A. Methods: Adult rats were injected (S.C.} with either rhlL-11 (150 ug/kg) or vehicle 20 min before administration of toxin A. Rat ileal loops (5 cm) were then injected with toxin A (5 ug) or buffer (0.4 ml) and 4 hours later. enterotoxicity was assessed by fluid secretion (mg/cm) and blood-to- lumen excretion of 3H-mannitol (dpm/Ioop).

Results: Secretion Permeability Buffer (n=8) 109+6 1,450+270 Toxin A (n=12) 368+14- 45,200+4,500"" rhlL-11 + Toxin A (n=5) 211+15 .... 5,460+2030"** n=number of rats tested. *p<O.05 and ** p<0.01 vs Buffer, ++ p<0.01 vs

Toxin A Histologic evaluation indicated that rhlL-11 attenuated epithelial damage caused by toxin A. but had no effect on neutrophil infiltration or congestion and edema of the mucosa. Preincubation of rhlL-11 (50 ug/ml) with purified toxin A (5 ug/ml) in vitro did not alter toxin A-induced cell rounding in cultured lung fibroblasts or 3H-toxin A specific binding to rat ileal brush borders Summary and Conclusions: rh-lL11 significantly inhibits C. difficile toxin A-mediated secretion (by 45.6%) and permeability (by 88%} and reduces mucosal damage in v/vo, but has no effect on fibroblast rounding or toxin A binding to its membrane receptor in vitro. These results suggest that IL- l 1 inhibits toxin A secretion and intestinal permeability by a post-receptor mechanism