Helicobacter pylon is a microaerophilic bacterium found in the stomach of asymptomatic humans well as patients with acid peptic disease and gastric adenocarcinoma. We have developed an in situ adherence assay to examine the cel lineage-specific nature of binding of this organism and to characterize the nature of cell surface receptors that recognize its adhesin. Fluorescein isothiocyanatelabeled H. pylon strains were bound to surface mucous cells present in the pit region of human and rat gastric units but not to mucous neck, parietal, or chief cell lineages present in the glandular domains of these units. Binding was abolished by proteinase K treatment of tissue sections and by pretreatment of the bacteria with bovine submaxillary gland mucin, a rich source of fucosylated and sialylated carbohydrates. Several lines of evidence suggest that binding to surface mucous cells is not dependent upon terminal nonsubstituted a2,3-and a2,6-linked sialic acids in the adhesin receptor: (s) binding was not inhibited by incubating H. pylon strains with sialylated glycoconjugates such as fetuin and free sialyRactose; (it) immunohiochemical stainings using the sialic acid-specific Sambucus nigra and Maackia amurensis lectins and the cholera toxin B subunit did not detect any sialylated glycoconjugates in these epithelial cells; and (ii) binding was not sensitive to metaperiodate under conditions that selectively cleaved carbons 8 and 9 of terminal nonmodifled sialic acids. A role for fucosylated epitopes in the glycoprotein(s) that mediate binding ofH. pylon to surface mucous cells was suggested by the facts that this lineage coexpresses the adhesin receptor and major fucosylated histo-blood group antigens, that monoclonal antibodies specific for histo-blood group antigens H, B, and Leb block binding, and that the lectin Ukx europaeus type 1 aggutinin, which is specific for a-L-fucose, also bound to the same ceDls that bound the bacteria. Furthermore, human colostrum secretory IgA inhibited adhesion in a metaperiodate-and a-L-fucosidasesensitive but neuraminidase-independent fashion. The in situ adherence assay should be useful in further characterizing the H. pyloni adhesin and its receptor and for identifying therapeutically useful compounds that inhibit strain-specific and cell lineage-specific binding of this human pathogen.

. Min is a fully penetrant dominant mutation that leads to the development of multiple intestinal adenomas throughout the duodenal-to-colonic axis.Minl+ C57BL6/J mice have an average life-span of 120 d. Mufti-label immunocytochemical studies of these lesions demonstrate patches of differentiated enterocytes, and scattered enteroendocrine, goblet and Paneth cells . Expression of endogenous marker genes within these differentiated cells can be directly correlated with the position occupied by the adenoma along the duodenal-to-colonic axis and mirrors the regional differentiation of the normal gut epithelium . The presence of multiple lineages in adenomas together with their retention of spatial information suggests that tumorigenesis in Minl+ mice may be initiated in a multipotent stem cell normally located at the base of intestinal crypts . To study the time-dependent properties of these tumors, genetic conditions were T HE mouse intestinal epithelium undergoes perpetual renewal of its four principal terminally differentiated cell types: the polarized absorptive enterocyte, the mucus-producing goblet cell, a complex population of enteroendocrine cells, and the defensin/lysozyme containing Paneth cell which is thought to function as part ofthe biological barrier to bacterial translocation across the gut (reviewed in Gordon, 1989) . Renewal and differentiation are rapid, extraordinarily well organized in several spatial dimensions, and dependent upon multipotent stem cells that are functionally anchored at the base of intestinal crypts (Cheng and Leblond, 1974). In the adult small intestine, 6-10 crypts, each composed of -250 cells, surround each villus. ['H]Thymidine labeling studies indicate that 150 cells, located in the middle portion of each crypt, pass through the cell cycle every 12 h resulting in the generation of 300 new cells/ crypt/d (reviewed in Potten and Loefer, 1990). These cells undergo a bipolar migration . -12 cells emerge from each crypt per hour and are translocated in coherent vertical bands up the adjacent villus (Cheng and Leblond, 1974; Potten and Loefer, 1990;Schmidt et al., 1985b). Differentiation ofenterocytes, goblet and enteroendocrine cells occurs during this upward migration which is completed in 3 d when the sooner. These studies indicate that the Minl+ mouse is a powerful model system for analyzing the mechanisms that establish and maintain a balance between proliferation and differentiation in the continuously renewing gut epithelium and for an assessment of the multi-step hypothesis of intestinal neoplasia . cells are exfoliated into the gut lumen at the apical extrusion zone of the villus . Each villus contains a steady state level of -3,500 surface epithelial cells and sheds -1,400 cells/

We have used histochemical methods to survey the cellular patterns of binding of a panel of 45 lectins with well-defined carbohydrate specificities to sections prepared from various regions of the gastric-to-colonic axis of fetal, neonatal, and adult FVB/N mouse gut. The results suggest that lectins can be used as remarkably sensitive tools to describe the differentiation programs of gastric and intestinal epithelial cell lineages as a function of their position along the cephalocaudal axis of the gut and as a function of developmental stage. Studies of intestinal isografts and transgenic mice that express Simian virus-40 T antigen in enterocytes suggest that many of these cell lineage-specific and spatial patterns of glycoconjugate production can be established and maintained in the absence of exposure to luminal contents and in the presence of specific proliferative abnormalities. This lectin panel should be useful for operationally defining subpopulations of the principal gut epithelial cell lineages in normal strains of mice, for describing variations in gut epithelial cell differentiation programs in mutant and transgenic mice, and for recovering specific epithelial cell lineages or subpopulations.

Abstract. The mouse intestinal epithelium represents a unique mammalian system for examining the relationship between cell division, commitment, and differentiation. Proliferation and differentiation are rapid, perpetual, and spatially well-organized processes that occur along the crypt-to-villus axis and involve clearly defined cell lineages derived from a common multipotent stem cell located near the base of each crypt. Nucleotides -1178 to +28 of the rat intestinal fatty acid binding protein gene were used to establish three pedigrees of transgenic mice that expressed SV-40 large T antigen (TAg) in epithelial cells situated in the uppermost portion of small intestinal crypts and in already committed, differentiating enterocytes as they exited these crypts and migrated up the villus. T antigen production was associated with increases in crypt cell proliferation but had no apparent effect on commitment to differentiate along enterocytic, enteroendocrine, or Paneth cell lineages. Singleand multilabel-immunocytochemical studies plus RNA blot hybridization analyses suggested that the differentiation programs of these lineages were similar in transgenic mice and their normal litterrnates. This included enterocytes which, based on the pattern of [3H]thymidine and 5-bromo-2'-deoxyuridine labeling and proliferating nuclear antigen expression, had reentered the cell cycle during their migration up the villus. The state of cellular differentiation and/or TAg production appeared to affect the nature of the cell cycle; analysis of the ratio of S-phase to M-phase cells (collected by metaphase arrest with vincristine) and of the intensities of labeling of nuclei by [3H]thymidine indicated that the duration of S phase was longer in differentiating, villus-associated enterocytes than in the less well-differentiated crypt epithelial cell population and that there may be a block at the G2/M boundary. Sustained increases in crypt and villus epithelial cell proliferation over a 9-mo period were not associated with the development of gut neoplasms-suggesting that tumorigenesis in the intestine may require that the initiated cell have many of the properties of the gut stem call including functional anchorage.T HE manner in which epithelial cell renewal occurs in the mouse intestine provides a unique opportunity to address questions about the relationships between cell division and differentiation programs in a mammalian system. The opportunity arises because of three features: (a) Address reprint requests to Jeffrey Gordon, Department of Molecular Biology and Pharmacology, Box 8103, Washington University, 660 So. Euclid Ave., St. Louis, MO 63110.Gfinter H. Schmidt's present address is A.B.C. Cambridge, Ltd., Cambridge, England. proliferation and differentiation in the gut are spatially well organized along the crypt-to-villus axis; (b) the processes occur rapidly (completed in •3 d) and perpetually; and (c) cell lineage relationships are well known and involve derivation from a single, active multipotent stem cell.Each adult sma...

Abstract. Transgenes consisting of segments of the rat liver fatty acid-binding protein (L-FABP) gene's 5' nontranscribed domain linked to the human growth hormone (hGH) gene (minus its regulatory elements) have provided useful tools for analyzing the mechanisms that regulate cellular and spatial differentiation of the continuously renewing gut epithelium. We have removed the jejunum from normal and transgenic fetal mice before or coincident with, cytodifferentiation of its epithelium. These segments were implanted into the subcutaneous tissues of young adult CBY/B6 nude mouse hosts to determine whether the bipolar, migration-dependent differentiation pathways of gut epithelial cells can be established and maintained in the absence of its normal luminal environment. Immunocytochemical analysis of isografts harvested 4-6 wk after implantation revealed that activation of the intact endogenous mouse L-FABP gene (fabpl) in differentiating enterocytes is perfectly recapitulated as these cells are translocated along the crypt-to-villus axis. Similarly, Paneth and goblet cells appear to appropriately differentiate as they migrate to the crypt base and villus tip, respectively. The enteroendocrine cell subpopulations present in intact 4-6-wk-old jejunum are represented in these isografts. Their precise spatial distribution along the crypt-to-villus axis mimics that seen in the intact gut. A number of complex interrelationships between enteroendocrine subpopulations are also recapitulated. In both "intact" and isografted jejunum, nucleotides -596 to +21 of the rat L-FABP gene were sufficient to direct efficient expression of the hGH reporter to enterocytes although precocious expression of the transgene occurred in cells located in the upper crypt, before their translocation to the villus base. Inappropriate expression of hGH occurred in a high percentage (>80%) of secretin, gastrin, cholecystokinin, and gastric inhibitory peptide producing enteroendocrine cells present in the intact jejunum of 4-6-wkold L-FABP -596 to +2VhGH transgenics. Addition of nucleotides -597 to -4,000 reduced the percentage of cells co-expressing this reporter four-to eightfold in several of the subpopulations. Jejunal isografts from each transgenic pedigree studied contained a lower percentage of hGH positive enteroendocrine cells than in the comparably aged intact jejunum. Together our analyses suggest that (a) the complex program of enteroendocrine cell differentiation can be expressed in the absence of extracellular luminal contents; (b) L-FABP/hGH transgenes appear to be useful reporters of a subtle heterogeneity in the regulatory environments of enteroendocrine cell subpopulations and a sensitive marker of the fact that luminal contents can regulate gene expression within and between subpopulations; (c) in contrast to enteroendocrine cells, transgene expression in enterocytes does not appear to be modulated by luminal factors; and (d) transgenic intestinal isografts represent a general and powerful tool not only for assessing the influence...