April 1995 Intestinal Disorders A301

• RELATIONSHIP OF THE UNSTIRRED LAYER TO ABSORPTIVE EPITHELIUM INTHE RAT JEJUNUM. M,D-,.J.,e_XiR, C, Fine, J. Fume, D.G. Levitt. VAMC Research Service, Mpls, and Dept. of Physiology, University of Minnesota.

Despite the villous architecture of intestinal mucosa, the interaction between the uustirred layer and absorptive epithelium commonly is modelled as a uniform thickness of unstirred water separating bulk luminal contents from a flat membrane. The g0al of the present study was to more accurately def'me the true relationship between the unstirred layer and epithelium. Rat jejunum (15 cm) was perfusod via implanted cannulas, and maltose hydrolysis determined over a range of concentrations. These data were fitted to a mucosal model in which the unstirred layer extended up the intervillous space to varying distances over the villous tips. Based on histological sections, villi were modeled as rectangular prisms (400x200 pro) with a 40 larn wide intervillous space. The Km (2.3 mM) and Vmax (0.56 Ilmol/sec) of maltase were determined in vitro and maltase activity was assumed to diminish by 60% from villous tip to base (J.Histochem and Cytoehem, 15:713,1968). Observed rams of maltose hydrolysis did not fit the conventional model of an unstirred layer over a flat membrane. The best fit was to a model in which an uustirred layer began 20 I.tm above the villous tips and extended into the intervlllous space. At very low maltose concentrations, hydrolysis occurred at or very near the the villous tips with a mean unstirred layer of about 20 p.m. In contrast, at high maltose concentrations, only 10% of hydrolysis occurred at the villous tips with the remaining 90% being hydrolyzed along the sides of the villi. A sizeable fraction of the maltose diffused through as much as 150 pm of unstirred fluid layer prior to reaching hydrolytic sites. The model predicted that secretion induced by a hypertonic solution should reduce maltose hydrolysis relative to that of an isotonic solution while water absorption induced by a hypotonic solution should not enhance hydrolysis (since villous tip maltaso was saturated in the isotonic solution). This prediction was experimentally confirmed: hydrolysis rates for 3 mM maltose were 0.53+005, 0,55+.007, and 0.295:.004 pmol/sec for isotonic, hypotonic and hypertonic solutions, respectively. We conclude that absorption cannot be modelled as a flat membrane of fixed surface area separated from bulk luminal contents by a single thickness of unstirred fluid. Rather the model must take into account that both the thickness of the aqueous diffusion barrier and the absorptive surface area increase with increasing solute concentration and that both these parumeters vary with solute transport eftlciency. Since double-reeiprocal plots assume constant membrane area and unstirred layer thickness, kinetics estimated via this commonly employed technique may contain sizeable errors.

• THE INTESTINE-SPECIFIC HOMEOBOX GENE CDX2 ACTIVATES INTESTINAL GENE TRANSCRIPTION VIA BOTH PROMOTER AND ENHANCER ELEMENTS. T. Levy, E. R. Suh, S. Siang, P. G. Traber. Division of Gastroenterology, University of Pennsylvania, Philadelphia, PA

The intestine-specific gene sucrase-isomaltase (SI) contains an evolutionarily conserved promoter element which binds two molecules of an intestine-specific homeodomain protein, Cdx-2 (Suh, et al. MCB 1994;14:7340). The binding of these two molecules provides additive transcriptional activation of the SI promoter. We examined whether the Cdx-2 protein may have a larger role in intestinal gene regulation by acting on other intestinal gene promoters and/or by binding to and activating enhancer elements. Two intestinal genes were identified that had promoters with potential Cdx-2 binding sites, lactase-phlorizin hydrolase (LPH) and AdRab-B (ARB), an intestine-specific rabbit gene encoding a phospholipase. Using electrophoretic mobility shift assays and antibodies to Cdx-2, we found that Cdx-2 binds to the previously described promoter element in LPH, LPH-N1, as a monomer. Likewise, a monomer binding site was confirmed in the ARB promoter just 5' to the start of transcription. In order to determine if this binding site was functional, chimeric ARB promoter-luciferase constructs were transfected into Caco-2 cells (which express Cdx-2) and co-transfected into NIH 3T3 cells with an expression vector for Cdx-2. Constructs were made with both the wild-type ARB promoter and a mutant with point mutations in the Cdx-2 binding site which abolishes binding. The results of these studies showed that Cdx-2 activates the ARB promoter by binding to the promoter element. We also examined whether the SIF1 bind!ng site (which binds Cdx-2) was able to act as an enhancer for gene transcription. In transfection experiments, we determined that the SIF1 element was able to activate transcription of multiple promoter elements (SI, thymidine kinase, and a minimal promoter from alkaline phosphatase) in a position and orientation independent fashion. Therefore, Cdx-2 may activate multiple genes by binding to enhancer elements located at a distance from the start of transcription. Taken together, these results suggest that the Cdx-2 protein is not limited to regulation of the SI gene, but likely has a much wider function in the regulation of intestinal gene transcription.

• WHIPPLE'S DISEASE PRESENTING WITH SEVERE GASTROINTESTINAL HEMORRHAGE. T.J.Lewis, S.A. Halter, R.M.Peek, R.C.Davis. Division of Gastroenterology, Vanderbilt University School of Medicine and Department of Veteran Affairs Medical Center, Nashville, TN.

Whipple's disease is a rare systemic bacterial illness caused by the recently identified bacillus, Tropheryma whippelii. This disease has a wide spectrum of clinical presentations, as it involves not only the gut and its associated lymphatics, but also the central nervous system, heart, and synovium. Common gastrointestinal manifestations in Whipple's disease include diarrhea, abdominal pain, and weight loss. However, to our knowledge, there have been no previous reports of significant gastrointestinal hemorrhage requiring multiple red cell transfusions.

We report a case of a 67 year-old male patient who presented complaining ofhematochezia without other abdominal complaints. On review of systems he reported a one year history of arthralgias, fatigue, anemia, and a 16 lbs. weight loss. Physical exam was remarkable for maroon blood in the rectum. Notable admission labs included albumin of 2.2 mg/di and hematocrit of 21%. Upper endoscopy was normal; colonoscopy revealed blood throughout the colon, with its source presumed to be proximal to the cecum. His hematochezia persisted and required ten units of packed red fells. Angiography and Tc/99-tagged red cell scanning were both normal. The patient underwent exploratory celiotomy which revealed extensive mesanteric adenopathy; intraoperative endoscopy showed diffuse hemorrhagic enteritis involving the distal jejunum and ileum. On microscopic exam, numerous intensely staining PAS positive macrophages were present throughout the lamina propria of the small intestine and the mesenteric lymph nodes. Special stains did not reveal acid-fast bacilli or fungal organisms. To confirm the diagnosis,genomic DNA was extracted from paraffin-embedded intestinal biopsies and lymph nodes. Using 2~ whippelii-specific oligonuclantide primers and template DNA from intestinal and lymph node tissue, PCR amplified a distinct product of 284 bp (expected size), which was not present in the control tissue. The patient was treated with tetracycline and trimethoprim-sulfa. He has done well with progressive weight gain and no further bleeding.

Whipple's disease should be considered as an etiology of severe gastrointestinal hemorrhage when other of its more common clinical manifestations are present.

NADPH DIAPHORASE REACTIVITY IN THE EPITHELIAL CELLS OF STOMACH, JEJUNUM AND COLON FROM THE RAT, DOG AND HUMAN. X. Liu. S. M. Miller and J. H. Szurszewski. Department of Physiology and Biophysics, Mayo Clinic and Mayo Foundation, Rochester,

55905 Nitric oxide (NO) synthesized from L-arginine by nitric oxide synthase

(NOS) mediates many important biological processes including non- adrenergic non-cholinergic neurotransmission throughout the gut. NOS is generally recognized to be nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d). This study examined the presumed location of NOS in epithelial cells of stomach, jejunum and colon from the rat, dog and human by using the NADPH-d histochemical technique. All tissues from the rat and dog used in this study were from normal animals. Biopsy materials from human stomach and jejunum were from patients operated on for morbid obesity. The human colon was obtained from a patient operated on for colon cancer. In human tissues, NADPH-d reactivity was found in gastric gland parietal cells and absorptive cells of the jejunum. Very little staining was observed in colonic absorptive cells. In dog tissue, NADPH-d reactivity was found in absorptive cells of the jejunum and colon. In both species, the staining was very intense in jejunal absorptive cells with the brush border regions exhibiting the greatest intensity of reactivity. The mature absorptive cells stained more intensely for NADPH-d reactivity than immature absorptive cells. In sharp contrast, NADPH-d reactivity was absent from epithelial cells of the rat stomach, jejunum and colon. These data show the existence of species and regional differences in NADPH..d reactivity in gastrointestinal epithelial cells of the human, dog and rat. They suggest an important role for NO in regulating absorption in the human and dog jejunum. These results are the first report of the occurrence of NOS in epithelial cells of the human gastrointestinal tract. (Supported by NIH Grant DK17238.)