Lower Villus Cells Research Articles

170 GENE EXPRESSION OF INSULIN-LIKE GROWTH FACTOR-l IN THE SMALL INTESTINE OF SUCKLING AND ADULT RATS.

Insulin-like growth factor-I (GF-I) is a peptide that induces cell division and differentiation in a wide range of tissues and cell types. IGF receptors are present in almost all organs including the small intestine. A major site for the synthesis of circulating lGF-l is the liver, but lGF-l is also produced in a number of other tissues. Gene expression of IGF-I in fetal and adult rat small intestine was first detected by Lund (1986). However, intestinal levels of lGF-l mRNA during the suckling period, when IGF-I peptide is delivered via milk, are not known. In order to measure IGF-I mRNA levels in intestinal samples of suckling and adult rat, we established a very sensitive quantitative reverse transcription (RT) competitive-PCR assay. A known amount of an IGF-I competitor cDNA was added prior to PCR reaction and served as a quantitative standard. Interestingly, the expression of IGF-I mRNA in jejunum of sucklings was 3-fold lower (p 0.01) than in adult rats. We speculate that this difference may relate to the enteral intake of milk-borne IGF-I. Furthermore, exact cellular localization of IGF-I mRNA transcripts in the small intestine is not defined. To clarify this we used a non-radioactive in situ hybridization technique (digoxigenin labeling). In adult male rats an intense positive signal was evident in the upper crypt enterocytes and goblet cells, and sometimes in the lower villus cells. Probably because of the lesser sensitivity and in agreement with the previous data indicating a low amount of IGF-I mRNA in suckling rats in comparison with adults, no IGF-I transcripts were detected in suckling rats. We have thus demonstrated IGF-I mRNA presence in the intestinal mucosa of suckling rats and characterized its localization in the small intestinal mucosa of adult rats.

Patterns of gene expression along the crypt-villus axis in mouse jejunal epithelium.

There is considerable interest in gene expression along the crypt-villus axis of the small intestinal epithelium, particularly in the identification of genes expressed in intestinal crypts. In an attempt to identify crypt-expressed genes, single-stranded cDNA made from normal mouse jejunal epithelium was used in subtractive hybridization against single-stranded cDNA from epithelium from which crypt cells were depleted by 2,000 rads of gamma irradiation. Partial DNA sequence and in situ hybridization of 72 resulting clones were determined. The sequence of 45 clones matched previously published genes. Gene expression patterns fell into three categories: expression throughout the crypt-villus axis, expression restricted to the villus, and expression restricted to the crypt. Clones in the first two categories could be further divided into three subgroups: those with uniform expression, those with an increasing gradient of expression, and those with a decreasing gradient of expression along the crypt-villus axis. Twenty two clones showed a stronger expression in crypt and lower villus cells, four of these were differentially localized to the crypt. Two of the crypt localized clones were uniformly expressed throughout the crypt, expression of one was stronger in the lower crypt, and expression of the remaining clone was enhanced Paneth cells. We report the full-length cDNA sequence of the Paneth-cell-enhanced clone. The screen isolated crypt-expressed genes that may prove useful tools in the study of crypt biology. In a companion report, we characterize one of the crypt clones.

Patterns of gene expression along the crypt‐villus axis in mouse jejunal epithelium

Background There is considerable interest in gene expression along the crypt-villus axis of the small intestinal epithelium, particularly in the identification of genes expressed in intestinal crypts. Methods In an attempt to identify crypt-expressed genes, single-stranded cDNA made from normal mouse jejunal epithelium was used in subtractive hybridization against single-stranded cDNA from epithelium from which crypt cells were depleted by 2,000 rads of gamma irradiation. Partial DNA sequence and in situ hybridization of 72 resulting clones were determined. Results The sequence of 45 clones matched previously published genes. Gene expression patterns fell into three categories: expression throughout the crypt-villus axis, expression restricted to the villus, and expression restricted to the crypt. Clones in the first two categories could be further divided into three subgroups: those with uniform expression, those with an increasing gradient of expression, and those with a decreasing gradient of expression along the crypt-villus axis. Twenty two clones showed a stronger expression in crypt and lower villus cells, four of these were differentially localized to the crypt. Two of the crypt localized clones were uniformly expressed throughout the crypt, expression of one was stronger in the lower crypt, and expression of the remaining clone was enhanced Paneth cells. We report the full-length cDNA sequence of the Paneth-cell-enhanced clone. Conclusions The screen isolated crypt-expressed genes that may prove useful tools in the study of crypt biology. In a companion report, we characterize one of the crypt clones. © 1996 Wiley-Liss, Inc.

The de novo and salvage pathways for the synthesis of pyrimidine residues of RNA predominate in different locations within the mouse duodenal epithelium.

RNA synthesis was examined by radioautography in mouse duodenal epithelium using 3H-uridine as a tracer of the salvage pathway and 3H-orotic acid as a tracer of the de novo pathway. The incorporation of the two precursors was estimated by counting silver grains in light-microscopic and electron-microscopic radioautographs at successive levels of crypt and villus. With both precursors, silver grains were found over all epithelial nuclei, but in numbers varying by location. Thus, after 3H-uridine injection, the number of grains was high over nucleolus and nucleoplasm in the base of the crypt, declined gradually in the middle and top of the crypt, and was low along the villus. After 3H-orotic acid, the number of grains was fairly low throughout, but peaked over the nucleoplasm in lower villus cells. The 3H-uridine reaction over nucleolus and nucleoplasm in crypt cells was interpreted as synthesis by the salvage pathway of ribosomal RNA and heterogeneous RNA, respectively, whereas the 3H-orotic acid reaction over the nucleoplasm of some villus cells indicated that these cells synthesized heterogeneous RNA by the de novo pathway.

Radioautographic localization of epidermal growth factor receptors in human fetal gut

The present investigation was undertaken to study the localization and accessibility of epidermal growth factor binding sites in the human fetal gut (15–19 weeks of gestation) using light-microscopic and quantitative autoradiography. Exposure of colonic explants to 5 nmol/L 125I-labeled epidermal growth factor for 60 minutes at 22 °C revealed extensive accumulation at binding sites in undifferentiated cells of the crypts and at the base of the villus, as well as in the inner circular layer of the muscularis externa bordering the submucosa. Some labeling was also present in the mesenchymal and vascular elements of the lamina propria. Labeling was virtually absent on the brush border at all levels of the epithelium. Quantitative analysis revealed a distinct gradient in grain density along the various compartments of the crypt-villus axis. Epithelial cells in the deep portions of the crypts showed the highest grain density (9.2 grains/μm2) with values gradually decreasing to 6.5 in upper crypt and 3.9 in lower villus cells. The upper third of the villus showed very little labeling (0.4 grains/μm2). Cellular distribution of silver grains in lower villous cells revealed a polarization of labeling in the basolateral infranuclear region. Experiments performed at 4 °C and at various incubation times showed similar results. Using isolated loops of intact colon and jejunum, segments in which labeled epidermal growth factor was only accessible on the serosal side showed extensive labeling and distribution similar to that found in explanted tissue. On the other hand, labeled epidermal growth factor could not access these same receptor sites when infused into the lumen, either at 22 °C or 4 °C. These results show that in the human fetal gut (a) the greatest concentration of epidermal growth factor binding sites is found in regions of high proliferative activity and (b) binding sites are absent from the brush border, suggesting that, under normal circumstances, systemic but not luminal epidermal growth factor has free access to its specific receptor.

Diet‐Induced Alterations of Lipids During Cell Differentiation in the Small Intestine of Growing Rats

SummaryThe lipid components of columnar cells harvested from rat small intestine were analyzed at each step of cell maturation. The effect of dietary lipids on the evolution of lipids in differentiating cells was studied using two diets representative either of a control or of an essential polyunsaturated fatty acid deficient lipid supply. Two groups of weanling rats were fed a semisynthetic diet composed of corn oil (control diet) or hydrogenated coconut oil (deficient diet) for 11 weeks. Intestinal cells were extracted according to their position along the villus column. Linoleic and arachidonic acids constitutive of cell phospholipids increased as cells migrated from the lower to the mid‐part of the crypt‐villus axis only with the control diet. This gradient disappeared after a 3‐week feeding period with hydrogenated coconut oil diet so that columnar cells of essential fatty acid deficient rats exhibited the same overall fatty acid composition all along the crypt‐villus axis. Essential fatty acid deficiency resulted in an increase of both the triene to tetraene and arachidonic acid to linoleic acid weight ratios regardless of the maturational step of cells. As compared to the control diet, the essential fatty acid deficient diet induced a decrease of both the cholesterol and free fatty acid to phos‐pholipid molar ratios only in lower villus cells. We conclude that (a) progressive compositional modifications of lipids occur while ascending the crypt‐villus axis and (b) essential fatty acid deficiency induces substantial alterations of the lipid evolution normally linked to cell differentiation.

Diet-Induced Alterations of Lipids During Cell Differentiation in the Small Intestine of Growing Rats

The lipid components of columnar cells harvested from rat small intestine were analyzed at each step of cell maturation. The effect of dietary lipids on the evolution of lipids in differentiating cells was studied using two diets representative either of a control or of an essential polyunsaturated fatty acid deficient lipid supply. Two groups of weanling rats were fed a semisynthetic diet composed of corn oil (control diet) or hydrogenated coconut oil (deficient diet) for 11 weeks. Intestinal cells were extracted according to their position along the villus column. Linoleic and arachidonic acids constitutive of cell phospholipids increased as cells migrated from the lower to the mid-part of the crypt-villus axis only with the control diet. This gradient disappeared after a 3-week feeding period with hydrogenated coconut oil diet so that columnar cells of essential fatty acid deficient rats exhibited the same overall fatty acid composition all along the crypt-villus axis. Essential fatty acid deficiency resulted in an increase of both the triene to tetraene and arachidonic acid to linoleic acid weight ratios regardless of the maturational step of cells. As compared to the control diet, the essential fatty acid deficient diet induced a decrease of both the cholesterol and free fatty acid to phospholipid molar ratios only in lower villus cells. We conclude that (a) progressive compositional modifications of lipids occur while ascending the crypt-villus axis and (b) essential fatty acid deficiency induces substantial alterations of the lipid evolution normally linked to cell differentiation.

Expression of Developmentally Regulated Crypt Cell Antigens in Human and Rat Intestinal Tumors<xref ref-type="fn" rid="FN1">2</xref>

The expression of 5 antigens specific for adult intestinal crypt cells and defined by monoclonal antibodies prepared to surface membrane components of the human colon tumor cell line CaCo-2 was studied during fetal and postnatal development of Sprague-Dawley rat small and large intestines. In the small intestine, all epithelial cells were stained during fetal life; antigen distribution became restricted to the crypt and lower villus cells in suckling animals and to the crypt cells after weaning. In the colon, these antigens could be detected only during a short period of development, comprising the last 3-4 days of fetal life and the first 8-10 days after birth. Expression of the antigens defined by this group of antibodies was investigated in rat intestinal tumors induced by 1,2-dimethylhydrazine (CAS: 540-73-8) and in normal and diseased human colons. In rats, these antigens were detected in all poorly and moderately differentiated adenocarcinomas of the small and large intestines. In most specimens, antigen distribution was not uniform; intensely stained areas were surrounded by completely negative tumor regions. Antigen expression was less intense in well-differentiated tumors, and about half of the tumors were negative for antigen expression. A similar pattern of expression of these antigens was observed in all human colonic adenocarcinomas examined; all samples of normal colon, benign polyps, and inflammatory bowel diseases examined were negative. These results suggest that this group of monoclonal antibodies recognizes oncofetal rat antigens expressed in chemically induced rat intestinal tumors and human colonic adenocarcinomas.

Fetal characteristics of small intestinal crypt cells.

Nine monoclonal antibodies were prepared against luminal membranes purified from rat intestinal cells at day 19 of gestation, and seven of them were found to define antigens common to adult crypt cells and fetal or embryonic intestinal epithelial cells. The FBB 2/29 antigen was first detected over the entire intestinal epithelial population at days 14-15 of gestation, a period of development characterized by formation of a stratified intestinal epithelium and differentiation of the surrounding mesenchyme. This antigen, identified as a set of high molecular mass proteins, became restricted to the crypt and lower villus cells after birth and was exclusively expressed by the crypt cells in adult intestine. It also was found to be expressed by the epithelial cells of the distal tubuli in the kidney of adult rats and by cultured human tumor colonic cells. The FBB 1/54/1, FBB 3/46, and FBB 3/78/9 antibodies stained only the epithelial cells present at the base of the villi in fetal intestine, starting at days 20-21 of gestation (about 1-2 days before birth), and stained the crypt and lower villus cells in newborn and adult intestine; these antigens may be regarded as specific markers for the developing crypt cells in fetal intestine shortly before birth. The FBB 1/20 and FBB 4/2 antigens were first detected on the fetal intestinal cells at day 18 of gestation; they were located over the entire epithelium in newborn rats and became restricted to the crypts after weaning. The FBB 2/28 antigen was expressed by the entire intestinal epithelium at all stages of development, starting from days 18-19 of gestation in the fetus. Two antibodies, FBB 3/4 and FBB 3/24, were found to be specific for lactase. These results have demonstrated the expression of cell- and tissue-specific components in rat intestine during early embryonic development and revealed a marked similarity in surface membrane antigens between fetal intestinal epithelial cells and adult crypt cells.

Glycoconjugate distribution and mobility on apical membranes of absorptive cells of suckling rat ileum in vivo.

The luminal membrane of ileal absorptive cells in suckling rats includes two domains: microvillar membranes and deep invaginations between microvilli. We examined the fates of foreign macromolecules that bind to anionic or saccharide sites on these domains after infusion into ligated loops in vivo. Cationized ferritin (CF) and ferritin-RCAI (beta-galactosyl) binding sites were distributed over the entire apical membrane. Ligands bound to apical invaginations were rapidly endocytosed, but ligands on microvilli were not. After CF binding, anionic sites on microvilli were mobile in the plane of the membrane and formed CF clusters at the tip and base of each microvillus. RCAI binding sites did not cluster. Wheat germ agglutinin (WGA, sialic acid) labeling was restricted to microvillus tips of mature cells but was dispersed over the microvillar surfaces of lower villus cells. Ferritin conjugates of Concanavalia ensiformis (Con A), Ulex europaeus agglutinin (UEA), and Dolichos biflorus agglutinin (DBA) did not bind to cell surfaces in vivo. Aldehyde fixation dramatically altered lectin binding patterns, resulting in unmasking and labeling of Con A, WGA, and DBA binding sites that were unavailable in vivo.

Specific binding of high density lipoprotein (HDL3) is not related to sterol synthesis in rat intestinal mucosa.

There is good evidence that high density lipoprotein (HDL) is involved in the flux of cholesterol into the cells of some organs and out of the cells of other tissues. Because we have previously found that HDL is bound specifically by mucosal cells of the small intestine, we have examined the possibility that this was associated with regulation of cholesterol flux. We have, therefore, compared the specific binding of 125I-labeled HDL3 with cholesterol synthesis in mucosal cells obtained from rats that had been treated to alter intestinal cholesterol metabolism. The rate of sterol synthesis measured in tissue slices, by the incorporation of [3H]water into sterols, was altered up to fivefold by treatment with cholestyramine (to induce bile salt loss), by surformer treatment (to reduce absorption of cholesterol), and by biliary diversion. Yet the capacity of mucosal cells to bind, internalize, and degrade 125I-labeled HDL3 was unchanged. Cholesterol feeding influenced neither the interaction of 125I-labeled HDL3 with cells nor the rate of sterol synthesis. Furthermore, the interactions of 125I-labeled HDL3 with mucosal cells isolated from the proximal and distal halves of the intestine or between the upper and lower villus cells were similar, despite differences in sterol synthesis. These data suggest that, in rat intestine, the specific binding of HDL is not related to sterol synthesis.(ABSTRACT TRUNCATED AT 250 WORDS)

Characterization of isolated epithelial cells from rat small intestine.

Intestinal mucosal cells from the rat have been isolated by a new technique involving intravascular perfusion of an intestinal segment with collagenase. Detached cells were flushed from the intestinal lumen with a second perfusion circuit containing an oxygenated buffered solution with 1% bovine serum. Sequential collection of cells at intervals during the period of perfusion revealed that villus-tip cells are recovered first (after 15 min of collagenase perfusion), followed by midvillus (after 25 min) and lower villus cells (after 35 min). The isolated cells were judged intact and viable by the criteria of trypan blue dye exclusion, ultrastructural appearance, and metabolic activity. They were characterized as villus-tip, midvillus, and lower villus-crypt cells by their alkaline phosphatase and sucrase activity, glycoprotein formation, and [3H]thymidine incorporation. Microsomal monooxygenase activity was four to five times greater in villus-tip than in lower villus cells, whereas heme oxygenase exhibited a reverse gradient. The isolated cells synthesized heme and bilirubin under cell culture conditions.