A 7 4 0 AGA A B S T R A C T S ' GASTROENTEROLOGY, VoI. I O 8 , No. 4

TISSUE PARTITIONING AND SECRETION INTO BILE OF • UNESTERIFIED AND CHYLOMICRON DERIVED ARACHIDONIC ACID. T. Me]in. Qi Chert, Ake N'dsson. Depts Int. Meal. & Cell Biol. Lurid University Hosp. Lurid. Sweden.

This study examines the role of chyle lipoproteins and biliary Ilpids in the supply of arachidonic acid (AA) to the g e s t r o ~ tract. 14 C-AA as flee fatty acid (FFA) and 3H-AAlabeiled chylomicrons were injected intravenously into rats. Radioactivity in tissue llpids after different time intervals was measured in normal rats and in rats with b'di=ry drainage: Recoveryof3H in liver lipids was higher than of 14C after lh (4.9 % 3H vs 3.4% 14C per gram), 4h (3.5 vs 2.0°4 per gram) and 24h (2.6 vs 1.8% per gram). In stomach and small intestip.e 3H / g tissue was 15-40 fold lower and 14C / g tissue 7-17 fold lower than in liver after 10-60 min. The initial retention in these tissues of 14(2 was thus 1.5-3 fold higher than of 3H. Radioactivity of the small intestine increased with time up to 24h in controls but not in rats with biliary drainage. Cumulative output of 3H and 14C in biliary lipids was 8% of injected dose in 24h; the time course being related to the level of each isotope in liver phosphatidylcholine, In stomach and colon 3H/g tissue was still 16-20 fold lower than in the liver after 24h. The gastrointestinal tract thus takes up littie AA from ehylomierons. AA derived from biliary phospholipids contributes to the AA pool of the small intestinal mucosa. The equilibration between AA pools of the liver and small intestine, : and of other organs is a slow process.

ROLE OF HEPATOCYTE GROWTH FACTOR ON RETARDED WOUND REPAIR INDUCED BY INDOMETHACIN IN A CULTURED GASTRIC MUCOSAL CELL MODEL. H.Mikami. S.Watanabe, M.Hirese, T Murai. K. Endo, H.Tsubouchi*, N.Sato I ~ a s t r o e n t e r o l o g y , Juntendo University, Tokyo, Japan and *Dept. of Medicine, Miyazaki Medical Collage, Miyazaki, Japan

It is well known that NSAIDs. such as indomethacin(IND), cause the gastric mucosal injury. However, the 'cellular mechanism of IND-induced gastric mucosal damage and its recovery is still unclear Recently, we established a new gastric epithelial restoration model for quantitative assessment of wound repair In this model, mucosal restoration occurred by the initial cell migration followed by proliferation. Using this model, we found a promoting action of hepatocyte growth factor(HGF) on gastric mucosal restoration m v i t r o Also HGF has been reported to induce prostaglandin production. In this study, we investigated effects of IND and HGF on gastric epithelial restoration process. METHOD Isolated rabbit gastric epithelial cells(92%mucous cell] were cultured m F-12 medium and formed complete monolayer cell sheet in 48 h. A wound with call-free area of constant size(2mm 2) was created by mechanical cell denudation using rotating silicon tip. The restoration process was monitored by measuring woundsize every 12 h. The proliferative cells were detected by BrdU staining.Effects of IND( 10 .4- 8x 104M) and HGF(10ng/ml) were assessed RESULT: The change of the size of cell-free area was presented in a table Data:mean, n=4, number:mm'-, vs control

Oh 12h 24h 36h 48b

control 1.98" ' 0.64 0.20 0 0

indomethacin 3xl 0"4M 1.97 0.80* 0.32* 0.03* 0

indomethacin 8x 10-4M i98 111" 0.81" 0.59* 0.45*

indomethacin 8xl 0"4M 2.00 ' 0.71 0.16 0 0 + HGF 10hg/ml

IND wound repatr in a manner, and HGI e promoted the restoration rate in cohtrols. Addition of HGF in the medium treated with IND prevented the IND-induced retardation of gastric mucosal restoration IND also inhibited the cell proliferation, and HGF by contrast promoted cell proliferation. CONCLUSION: IND directly retards the wound repair of rabbit gastric mucosal cells. HGF accelerates the wound repair and abolished IND-mduced gastric damage. Therefore. HGF has cytoprotective effects on gastric mucosal cells.

• BIOTIN DEFICIENCY RESULTS FROM LONG-TERM THERAPY WITH ANTICGNVULSANTS. D.M. Mock, M.E. Dyken. Dept. of Pediatrics, Univ. of Arkansas For Med. Sciences, Arkansas Children's Hospital, Section of Gastroenterology, Little Rock, AR, and Dept. of Neurology, Col. of Medicine, Univ. of Iowa, Iowa City, IA.

One study has reported decreased blood Concentrations of biotin in 75% of 274 adults on 10ng-term anticonvulsant therapy (Krause, et al. Ann. Neurol. 1982;12:485). A study from our laboratory found normal blood concentrations of total avidin-binding substances (TABS) and urinary excretion of TABS in ten children. However, 3'hydroxyisovaleric acid (3-HIA) excretion (the result of deficient activity of the blotin-dependent enzyme, 3-methylcrotonyl-CoA carboxylase) was increased in 6 of 11 children. Thus, even though TABS excretion were normal, :biotin status might be decreased at the tissue level. Interference with biotin intestinal absorption by antieonvulsants has been suggested as the mechanism leading to biotin deficiency.

In a larger study of adults with more rigorous exclusion of subjects supplemented with biotin, we sought to determine the incidence of reduced biotin status in adults treated chronically with anticonvulsants as judged by urinary excretion of true biotin and 3-HIA. Fifteen adults (eight women) receiving single or some combination of anticonvulsants (phenytoin, phenobarbital, carbamazepine, felbamate, valproic acid) and not supplemented with biotin, provided complete 24 h urine coll~cti0ns. Ten normal healthy adults (five women) not supplemented with biotin served as controls. Mean urinary 3-HIA excretion was significantly increased [295+_ 76 vs 112 +- 12 p.mol/24 h; X +- SEM; p < 0.005 by Mann-Whitney test, (MW)]. Urinary 3-HIA was frankly abnormal in 9 of 15 subjects. Despite 3-HIA excretion consistent with biotin deficiency, mean urinary excretion. of the biotin metabolite bisnorbiotin Was significantly increased (72 +- I7 vs 26 + 3 nmol/24 h; X + SEM; p < 0.003 by MW test), and biotin excretion was not down regulated as much as expected (30+- 4 vs 41 +- 6 nmol/24 h; p> 0.1 by MW test). These findings are consistent with the hypothesis that long-term anticonvulsant therapy impairs biotin status by inducing biotin catabolism, impairing renal reclamation of biotin, or both. In addition, urinary excretion of TABS may not reflect biotin 'deficiency at the tissue level because of increased urinary excretion of metabolites or biotin or both. Certainly these exciting new findings emphasize the need to reproduce this phenomenon in an animal model and to pursue a more careful, invasive assessment of the multiple effects of anticonvulsant therapy on biotin status.

• ZINC INTESTINAL ABSORPTION IN PATIENTS WITH CIRRHOSIS. ~ A PROSPECTIVE, CONTROLLED STUDY. A.C. Montemavor. R. Rosas, R. Mufioz, F. Isoard, L. Aspe, A. Estanes, M. Uribe, J.L. Pod. Depts. of Gastroenterology and Physiology of Nutrition, Instituto Nacional de Nutrici6n, Salvador.Zubirdn, Mexico City.

Decrease zinc serum leVelS in cirrhosis Can be explained by 10w dietary intake, abnormal intestinal absorption, increased metabolic requirements or high urine excretion: Previous studies had found a low zinc intestinal absorption only in patients with active alcohol intake or severe liver disease, in this study, we aimed to evaluate zinc intestinal absorption in cirrhotic patients with compensated liver disease, without active alcoholism, compared to age- and gender-matched control subjects. Methods: We included 50 healthy subjects and 50 cirrhotic patients (viral 51%, alcoholic 19%, PBC 9%, auto immune 9%, idiopathic 1-2%). without active " alcohol, minerals, laxatives, or diuretics intake, Patients having uncontrolled ascifes or encephalopathy, cancer, renal failure or diabetes were excluded. The oral zinc tolerance test was used to evaluate zinc intestinal absorption. Zinc serum levels Were measured before and after oral intake of 15 mg of elementary zinc using and atomic absorption spectrophotometer, In addition, total caroric and protein intakes were calculated, and D-xylosa, b-carotene, albumin, hemoglobin and lymphocyte counts measured by conventional methods. The area under the curve as well as the maximum zinc concentration after oral zinc administration were calculated. ' . , . Results: Basal serum zinc concentrations were significantly lower in cirrhotic patients (73.2 + 29.4 vs. 101.4 -+ 24.8 pg/dl) compared to healthy subjects (p < 0.05). Body mass index (25.8 + 2.6, 27 +- 3.8), gender (female 73 vs. 65%) and age (49 + 14 VS. 45 _ 13 years) were similar betweefi groups whereas total caloric (1765 _+ 414 vs. 2089 _+ 598 Kcal/d), ahd proteifi intake (68 +- 22 vs. 82 -+ 25 g/d) ware lower in cirrhotic than controls (p < 0.05). Serum zinc concentrations (Ixg/dl) at 60 rain. (99 _+ 32 vs 134 _+ 30), 120 min. (118 + 38 vs. 155_+ 34), 180 min. (116 + 32 vs. 150 +_ 29)and 240 min. (98 ± 27 vs. ~ 129 +-26) were significantly lower in cirrhosis (p < 0.05) than control subjects. A stepwise decline in serum zinc with worsening Child-Pugh class was observed before (A: 88+21; B 66+_ 18;C: 51 -+ 17 p.g/dl) and 120 min. (A, 136 ± 36; B: 100 _+ 31 ; C: 85 -+ 21 pg/dl), after zinc administration. D-xylose absorption was normal in all but on? cirrhotic patient and b-carotene levels were lower in cirrhosis (102 + 53 vs. 126 ~ 53 p.g/dl, p > 0.05) than controls. " ~ . . . . Conclusion: Zinc intestinal absorption, assessed by the oral tolerance test, is significantly lower in cirrhotic patients with compensated tiver disease and without active alcohol intake than control subjects . . . . . . . . .