Antibodies to intestinal microvillous membranes : ii. Inhibition of intrinsic factor-mediated attachment of vitamin b(12)to hamster brush borders.

Microvillous membranes isolated from the distal, but not proximal, half of hamster small bowel induced in rabbits the formation of antisera which inhibited intrinsic factor-mediated uptake of vitamin B(12) by hamster brush borders. The extent of inhibition was directly proportional to the concentration of antiserum, and an excess of IF-bound vitamin B(12) could overcome the inhibitory effect. The inhibitory factor was absorbed from antisera by brush borders isolated from the distal, but not proximal, half of the hamster intestine. Fractionation of antisera by gel filtration and DEAE-cellulose chromatography established that immunoglobulin G contained the inhibitory factor. Antisera capable of completely blocking uptake of IF-bound vitamin B(12) did not react with hamster IF or with the IF-vitamin B(12) complex, did not inhibit brush border disaccharidase activity and did not impair glucose transport by everted sacs of hamster intestine. These results demonstrate that an antibody to distal microvillous membranes competes with the IF-vitamin B(12) complex for a specific binding site or receptor located on the surface of distal hamster intestine.

Antibodies to intestinal microvillous membranes. I. Characterization and morphologic localization.

Antibodies (AbMVM) were produced in rabbits to microvillous membranes isolated from hamster small bowel. Incubation of frozen sections of hamster small bowel with fluorescein-labeled AbMVM showed specific reaction with brush borders, but not with other intestinal cellular components. Electron microscopy with ferritin-conjugated AbMVM localized the antigens more precisely to the surface mucopolysaccharide coat of the brush borders. AbMVM also reacted with the brush border of colon and of proximal renal tubules of hamsters but not with those of hamster stomach or gall bladder. It also reacted with the brush borders of some rat and human tissues, but not with those of rabbits. In addition, fluorescent-labeled AbMVM combined specifically with cell walls of some yeasts, but not of several bacteria. AbMVM also contained a weak precipitin to a component of hamster serum, which migrated like prealbumin in immunoelectrophoresis.

Structural features of the epithelio-mesenchymal interface of rat duodenal mucosa during development.

In fetal rats 5-7 days before birth, the duodenal epithelium is separated from mesenchymal cells by a well-defined basal lamina. By 3-4 days before birth, when small rudimentary villi are first seen, direct contact between epithelial and mesenchymal cells occurs by means of epithelial cell cytoplasmic processes which project through gaps in the basal lamina into the lamina propria. At contact sites, the epithelial and mesenchymal cell plasma membranes were less than 100 A apart but membrane fusion was not seen. In number and size these epithelial cell processes increase strikingly during the last 2 days of gestation, and they persist in large numbers until 7-10 days after birth. Thereafter, they decrease gradually in both number and size until 3-4 wk after birth, when the morphology of the epithelio-mesenchymal interface resembles that seen in adult rats, i.e., there are only rare epithelial cell processes which penetrate deeply into the lamina propria. The presence of a large number of epithelio-mesenchymal contact sites during the period of rapid growth and differentiation of duodenal mucosa may reflect epithelio-mesenchymal cell interactions which may facilitate the maturation of the duodenal mucosa.

Paneth and goblet cell renewal in mouse duodenal crypts.

Proliferation of Paneth and goblet cells of mouse duodenal crypts was studied by high resolution light microscope radioautography. In one group of mice, blood levels of thymidine-(3)H were sustained for up to 12 hr by repeated injections of isotope to facilitate identification of proliferating cells. In these animals, many goblet cell nuclei incorporated thymidine-(3)H whereas only 1 of 6261 tabulated Paneth cells was labeled. Cells intermediate in structure between undifferentiated and goblet cells and between undifferentiated and Paneth cells were identified and their light and electron microscopic features are described. A significant number of these "intermediate" cells incorporated thymidine-(3)H into their nuclei. Another group of mice received a single injection of thymidine-(3)H. These animals were killed 4 hr to 29 days after isotope administration. Goblet cells and intermediate cells with labeled nuclei were identified 4 hr after thymidine-(3)H but could not be seen after 15 days. In contrast, Paneth cells with labeled nuclei were not observed until 24 hr after thymidine-(3)H but were still present at 29 days, long after labeled undifferentiated, goblet, and intermediate cells had disappeared. We conclude that differentiated Paneth cells in mouse duodenum do not normally proliferate, but, instead, arise by differentiation from undifferentiated crypt cells or from intermediate cells. Moreover, once formed, Paneth cells persist in crypts for a prolonged period. In contrast, intermediate cells and crypt goblet cells proliferate actively and are less stable cell populations than differentiated Paneth cells. The precise function of the intermediate cells is not known, but they may represent transition forms between undifferentiated cells and the more matrure secretory cells. Damage of crypt epithelial cells, thought to be due to radiation effects, was evident in both groups of mice.

Sulfhydryl compounds may mediate gastric cytoprotection.

Ethanol induces hemorrhagic gastric erosions and causes a dose-dependent decrease in the concentration of nonprotein sulfhydryl compounds in rat gastric mucosa. Sulfhydryl-containing drugs protect rats from ethanol-induced gastric erosions, whereas sulfhydryl blocking agents counteract the mucosal cytoprotective effect of prostaglandin F2 beta. These observations suggest that endogenous nonprotein sulfhydryls may mediate prostaglandin-induced gastric cytoprotection and that sulfhydryl drugs may have potential for preventing or treating hemorrhagic gastric erosions.

Intestinal M cells: a pathway for entry of reovirus into the host.

Thirty minutes after inoculation of reovirus type 1 into the intestinal lumen of the mouse, viruses were found adhering to the surface of intestinal M cells but not other epithelial cells. Within 1 hour, viruses were seen in the M cell cytoplasm and were associated with mononuclear cells in the intercellular space adjacent to the M cell. These findings suggest that M cells are the site where reovirus penetrates the intestinal epithelium.

Organ culture of mucosal biopsies of human small intestine.

In vitro experiments of small intestinal mucosal function and metabolism utilizing excised tissue have been limited to a few hours by rapid epithelial cell necrosis which occurs with current incubation methods. We describe a method for culturing human mucosal biopsies for up to 24 hr employing organ culture methodology and demonstrate its potential application to studies of mucosal function. Peroral biopsies were placed in organ culture plates and maintained with modified Trowell's medium in 95% O(2)-5% CO(2) at 37 degrees C for 6-24 hr. To study cell proliferation, 2 muc of thymidine-(3)H was added per ml of medium. To study fat absorption, biopsies were exposed to micellar solutions of linolenic acid, monoolein, and taurodeoxycholate in Krebs-Ringer buffer for 15 min after culture in vitro for 24 hr. After 24 hr of culture, villi were shorter and wider. Cells in the lamina were reduced in number. Light and electron microscopic morphology of epithelial cells compared favorably to those of control biopsies except in occasional areas of partial necrosis. Some absorptive cells were more cuboidal and contained more lysosomes; many appeared entirely normal. Most crypt cells appeared normal; some contained increased glycogen and lysosomes. Mitoses were present, and labeled cells were abundant in crypts of biopsies after 6 hr of incubation with thymidine-(3)H-containing medium. By 24 hr. labeled cells migrated to the base of the villi. When biopsies cultured in vitro were subsequently exposed to micellar lipid, numerous lipid droplets were identified in the cytoplasm of absorptive cells. Thus, after 24 hr in vitro under these culture conditions, many human small intestinal epithelial cells maintain near normal morphology, epithelial cell proliferation proceeds, and fat absorption occurs.

Intrinsic factor-mediated attachment of vitamin B12 to brush borders and microvillous membranes of hamster intestine.

Hamster intrinsic factor (IF) preparations markedly enhanced the uptake of (57)cobalt-labeled cyanocobalamin (B(12)-(57)Co) by brush borders and microvillous membranes isolated from villous absorptive cells obtained from the distal but not the proximal half of hamster intestine. A similar effect was observed with rat and rabbit IF preparations, but IF preparations obtained from man, dog, and hog were ineffective. After fractionation of hamster IF preparations by gel filtration or ion exchange chromatography, the extent to which each fraction enhanced B(12)-(57)Co uptake by brush borders correlated closely with the vitamin B(12) binding capacity of the fraction. IF-mediated attachment of B(12)-(57)Co to brush borders occurred rapidly, was not diminished by removal of glucose or oxygen from the incubation medium, and was not significantly altered when incubation temperatures were reduced from 37 degrees C to 7 degrees C. Marked reduction in uptake occurred, however, in the absence of divalent cations. IF enhanced B(12)-(57)Co uptake by brush borders isolated from the proximal half of the intestine when these proximal brush borders were preincubated with supernatant fluid obtained after centrifugation of homogenates of distal intestinal mucosa at 28,500 g. The factor in this supernate responsible for the effect on proximal brush borders was shown to be particulate in nature upon centrifugation at speeds of 54,500 g or greater. The resultant pellet contained ribosomes and membranous fragments. Prolonged incubation of brush borders with crude saline extracts of hamster gastric mucosa resulted in decreased uptake of B(12)-(57)Co and marked lysis of brush borders with concomitant release of tissue nitrogen. Neither lysis of brush borders nor decreased uptake of B(12)-(57)Co with prolonged incubation was observed when hamster IF was partially purified. Furthermore, uptake of B(12)-(57)Co by brush borders increased with increasing purity of the IF preparation used. These results demonstrate IF-mediated attachment of B(12)-(57)Co to brush borders and microvillous membranes of hamster intestinal cells and provide further support for the presence of a specific receptor for IF-bound vitamin B(12) at the microvillous surface of the intestinal cell. IF-mediated attachment to the intestinal cell surface appears to be facilitated by divalent cations and to result from adsorption rather than an energy-requiring enzymatic reaction. Crude sources of hamster IF contain a factor which causes lysis of brush borders in vitro and which may explain in part the inhibitory effects of IF excess previously observed in vitro.

Glycoprotein synthesis and secretion by mucosal biopsies of rabbit colon and human rectum.

Elucidation of mechanisms involved in the control of colonic production of mucus requires direct examination of glycoprotein synthesis and secretion by colonic mucosa. In the past, the limited viability of intestinal mucosa in vitro has hampered such investigations. When maintained in an organ culture system, mucosal biopsies of rabbit colon and human rectum remained viable for 24 h as documented by morphologic appearance and a steady rate of protein synthesis and secretion. These biopsies also incorporated (14)C-labeled glucosamine into tissue glycoproteins and secreted labeled glycoproteins at a steady rate for 24 h. Glucosamine was predominantly incorporated into macromolecules that were ultimately secreted, in contrast to leucine, which was predominantly incorporated into tissue macromolecules. When studied by autoradiography, cultured rabbit colonic biopsies synthesized and secreted glycoproteins in vitro at cellular sites and over a time-course similar to those previously described for the intestine of intact animals. Acetylcholine consistently stimulated secretion of labeled glycoproteins but did not alter glycoprotein synthesis. In contrast, cycloheximide inhibited glycoprotein synthesis but had no effect on the secretion of newly synthesized glycoproteins. Rectal biopsies from patients with active ulcerative colitis incorporated increased amounts of [(14)C]glucosamine into glycoproteins during organ culture and secreted labeled glycoproteins more rapidly into the incubation medium when compared to biopsies obtained from healthy volunteers These findings indicate that organ culture provides a useful means of directly examining the synthesis and secretion of glycoproteins by healthy and diseased colonic mucosa.

Epithelial basement membrane of mouse jejunum. Evidence for laminin turnover along the entire crypt-villus axis.

Little is known regarding turnover of the epithelial basement membrane in adult small intestine. Are components degraded and inserted along the length of the crypt-villus axis or selectively in the crypt region with subsequent migration of basement membrane from crypt to villus tip in concert with epithelium? We injected affinity-purified sheep anti-laminin IgG or sheep anti-laminin IgG complexed to horseradish peroxidase (HRP) into mice to label basement membrane laminin in vivo. Fluorescence microscopy revealed linear fluorescence along the length of the jejunal epithelial basement membrane 1 d after anti-laminin IgG injection. By 1 wk, small nonfluorescent domains were interposed between larger fluorescent domains. Over the next 5 wk the lengths of nonfluorescent domains increased progressively whereas those of fluorescent domains decreased. Additionally, electron microscopy revealed HRP reaction product along the length of the epithelial basement membrane after 1 d whereas unlabeled or lightly labeled domains that increased in length with time were observed interposed between heavily labeled domains by 2 and 4 wk along the entire crypt-villus axis. We conclude that laminin turnover occurs focally in the epithelial basement membrane of mouse jejunum along the crypt-villus axis over a period of weeks and that this basement membrane does not comigrate in concert with its overlying epithelium.

Esophageal, gastric, and intestinal candidiasis.

Gastrointestinal Candida infection is more prevalent than previously recognized. It is most often seen in patients with underlying impairment of the immune system but may also occur in apparently normal individuals. Esophageal involvement is most common, presenting with odynophagia, dysphagia, or bleeding. Gastric Candida infection may cause diffuse mucosal involvement or focal invasion of benign gastric ulcers. Intestinal candidiasis is uncommon and poorly characterized. The diagnosis is usually established by visualizing the characteristic yeast or mycelial forms in endoscopic brushings and biopsies. Oral nystatin is effective therapy in many patients, but other antifungal agents may be needed in extensive or persistent disease, especially in immunocompromised patients.

Oral clotrimazole in the treatment of esophageal candidiasis.

Several antifungal regimens had failed to relieve severe, recurrent esophageal candidiasis in a 75 year old woman without predisposing disease whose serum transiently inhibited the candidacidal capacity of her polymorphonuclear leukocytes. Treatment with oral nystatin suspension was unsuccessful, whereas intravenous amphotericin B and miconazole induced only transient responses. Oral 5-fluorocytosine induced severe nausea and vomiting, and was discontinued. Oral clotrimazole troches produced prompt and sustained eradication of the patient's candidal esophagitis.

Immunocytochemical localization of actin in epithelial cells of rat small intestine by light and electron microscopy.

We used post-embedding immunocytochemical techniques and affinity-purified anti-actin antibody to evaluate localization of actin in epithelial cells of small intestine by fluorescence and electron microscopy. Small intestine was fixed with 2% formaldehyde-0.1% glutaraldehyde and embedded in Lowicryl K4M. One-micron or thin sections were stained with antibody followed by rhodamine- or colloidal gold-labeled goat anti-rabbit IgG, respectively. Label was present overlying microvilli, the apical terminal web, and the cytoplasm directly adjacent to occluding and intermediate junctions. Label was associated with outer mitochondrial membranes of all cells and the supranuclear Golgi region of goblet cells. Lateral cytoplasmic interdigitations between mature cells and subplasmalemmal filaments next to intrusive cells were densely labeled. The cytoplasm adjacent to unplicated domains of lateral membrane was focally labeled. Label was prominent over organized filament bundles within the subplasmalemmal web at the base of mature cells, whereas there was focal labeling of the cytoplasm adjacent to the basal membrane of undifferentiated cells. Basolateral epithelial cell processes were labeled. Label was focally present overlying the cellular ground substance. Our results demonstrate that actin is distributed in a distinctive fashion within intestinal epithelial cells. This distribution suggests that in addition to its function as a structural protein, actin may participate in regulation of epithelial tight junction permeability, in motile processes including migration of cells from the crypt to the villus tip, in accommodation of intrusive intraepithelial cells and in adhesion of cells to one another and to their substratum.

Alterations in blood vessels during gastric injury and protection.

Recent investigations suggest that the mucosal vascular endothelium is not a passive bystander, and that alterations within the blood vessel wall actively participate in the pathogenesis of gastric mucosal injury. We review here our data on rapidly developing vascular injury as detected by monastral blue deposition and increased vascular permeability measured by Evan's blue extravasation in dose- and time-dependent experiments with ethanol, HCl, and NaOH in the rat. In addition, using in vivo microscopy and laser-Doppler velocimetry, we demonstrate circulatory stasis in the superficial capillaries within about 1 min after topical application of damaging agents, and a gradual decrease in blood flow that correlates with the extent of hemorrhagic erosions. Prostaglandins or sulfhydryl agents prevented the circulatory standstill and the development of hemorrhagic mucosal lesions. We conclude that microvascular damage, increased vascular permeability, and capillary stasis precede the development of hemorrhagic mucosal lesions.

Diagnosis and treatment of celiac sprue.

Celiac sprue, also termed celiac disease or gluten-sensitive enteropathy, is a chronic disease in which malabsorption of nutrients is caused by a characteristic, but nonspecific, lesion of the small-intestinal mucosa. The lesion is produced, through unclear mechanisms, by protein constituents of some cereal grains. Exclusion of wheat gluten and rye, barley, and oat prolamins from the diet results in a prompt improvement in absorption, along with reversion, toward normal, of the associated small-intestinal lesion. The spectrum of manifestations of celiac sprue is remarkably broad, but the severity of disease generally correlates with the length of small intestine that is damaged. When most or all of the small-intestinal mucosa is involved, symptoms are severe and malabsorption is generalized. In such patients, a diagnosis of celiac sprue is usually considered. When, on the other hand, the mucosal lesion is limited to the duodenum and proximal jejunum, overt gastrointestinal symptoms and steatorrhea may be absent. In those patients, clinical manifestations, if present at all, may reflect malabsorption of only one or two substances, notably iron and folate, that normally are absorbed somewhat selectively by the proximal intestine. Arriving at the correct diagnosis in such cases may be quite challenging.

Structure and function of intestinal M cells.

M cells are structurally distinctive, uniquely permeable epithelial cells found only overlying the domes of mucosal lymphoid follicles. Antigenic macromolecules and some viruses, bacteria, and protozoa enter their apical surface by endocytosis or phagocytosis. These substances traverse the M-cell cytoplasm by transcytosis, breaching the epithelial barrier, and then interact with the subepithelial immunocompetent cells to initiate mucosal and systemic immune responses. The M cell serves as a portal of entry for selected pathogens that cause disease locally in the wall of the intestine or, following dissemination, at distant sites. The mechanisms that regulate adherence to and penetration of M cells by macromolecules and microorganisms are not known, but selective binding of secretory IgA to the luminal surface may be important. Whether M cells simply serve a sieving function and always transport substances unchanged across the epithelial barrier or whether they also sort and process antigens they endocytose and present them to adjacent lymphoid cells requires further study.