Arginine-vasopressin immunoreactive material in the gastrointestinal tract

Like many other neuropeptides, vasopressin is not confined to the hypothalamic neurohypophysial system. Furthermore, vasopressin was found to be a potent vasoconstrictor in the rat jejunum, reducing myenteric artery flow. These associations were the basis of this investigation on the presence of vasopressin in the gastrointestinal (GI) tract by both RIA and immunohistochemistry. Portions of the gastrointestinal tract and pancreatic islets of the rat were extracted with 0.1 N HCl for RIA measurements of AVP content. Similar portions from the male cat GI tract were used for immunohistochemistry studies. Acid extracts of the GI tract were found to contain immunoreactive AVP with the highest concentration (pg/mg protein) in the fundus portion of the stomach (15.0 +/- 1.6) and slightly lower values down along the antrum-pylorus portion (6.7 +/- 0.6), proximal jejunum (8.6 +/- 0.2), distal ileum (9.7 +/- 0.3) and colon (11.9 +/- 0.5). In the pancreatic islets the concentration was much higher (72.0 pg/mg protein). The extract inhibition curves showed parallelism with the appropriate standard preparation of AVP in the specific RIA. Immunohistochemical localization showed IR-AVP in the nerve fibers around the myenteric plexus of the second portion of the duodenum. It was also found in fibers starting from where the myenteric plexus goes through the layer of muscle fibers, penetrating the submucosa and duodenal mucosa, ending near the capillaries situated along the basal side of the villous epithelium cells. Similar IR-AVP activity was found in cells located in the mucosal epithelium of the duodenum, jejunum, ileum, colon and rectum.(ABSTRACT TRUNCATED AT 250 WORDS)

Identification of selective, high affinity [125I]-angiotensin and [125I]-bradykinin binding sites in rat intestinal epithelia

Specific [125I]-angiotensin II (AII) and [125I]-bradykinin (Bk) binding sites have been identified within epithelial membranes from rat jejunum and descending colon. These high affinity intestinal sites exhibited KD values of 0.64 +/- 0.16 nM for [125I]-AII and 0.69 +/- 0.13 nM for [125I]-Bk, which were similar to those for [125I]-AII (0.85 nM) and [125I]-Bk binding sites (1.03 nM) previously identified in renal cortex epithelia. Specific [125I]-AII binding capacity was only 19.77 +/- 2.74 fmol mg-1 in small intestine and 11.31 +/- 2.66 fmol mg-1 in descending colon epithelia while a larger population, 332.0 +/- 72.9 fmol-mg-1 of specific [125I]-Bk sites were identified in epithelial membranes from small intestine. Significant hydrolysis of both free [125I]-AII and [125I]-Bk was observed while membrane bound peptides remained relatively resistant to degradation. Whilst no corrections have been made to the observed values of KD and Bmax quoted above, one may assume that the calculated reductions in the free hormone concentration will result in a decrease of the KD value for both peptides. Loss of membrane bound peptide, particularly of [125I]-AII, may indicate that the calculated Bmax value is an underestimation. Despite the rapid degradation of unbound [125I]-AII and [125I]-Bk during incubations the kinetics of specific peptide binding were reversible and highly selective. The order of potency for specific [125I]-AII binding was [Sar1, Leu8]-AII greater than or equal to [Sar1, Thr8]-AII greater than or equal to AII greater than [Sar1, Ile8]-AII greater than or equal to [Des, Asp1, Ile8] AII greater than AIII. Specific [125I]-Bk binding was also highly selective, the order of potency being Phe8-Bk greater than or equal to Tyr8-Bk greater than or equal to Lys-Bk much greater than Des, Arg1-Bk. AII exhibited an IC50 of greater than 1mM for specific [125I]-Bk binding and likewise Phe8-Bk for specific [125I]-AII binding.

Localization of Met-enkephalin-Arg6-Gly7-Leu8-like immunoreactivity in the gastrointestinal tract of rat and pig

One of the opioid precursor molecules, pre-pro-enkephalin A, contains within it, in addition to Leu-enkephalin (Leu-Enk) and Met-enkephalin (Met-Enk), Met-enkephalin-Arg6-Gly7-Leu8 (Met-Enk-8), which is specific to this precursor. This study deals with the localization of Met-Enk-8-like immunoreactivity in the gastrointestinal tract of rat and pig. Immunoreactivity was identified in intramural nerve elements of rat and pig, and in gut endocrine cells of pig. Immunoreactive (IR) nerve fibers were seen mainly in the myenteric plexus of rat and in both the myenteric and submucosal plexuses of pig. Some IR fibers were dispersed throughout the lamina propria mucosae of rat. Porcine IR endocrine cells were dispersed in the epithelium from the pyloric antrum to the ileum, existing concomitantly with enterochromaffin (EC) cells. Specificity tests revealed that immunoreactivity to Met-Enk-8 antiserum was not influenced by preincubation of the antiserum with Leu-Enk and Met-Enk. This suggests the possibility that pre-pro-enkephalin A is contained in the gastroenteric nerves of rat and pig and in a population of porcine EC cells.

Changes in collagen content of the small intestinal wall after anastomosis

Collagen content of small intestinal anastomosis in the rat was investigated in relation to previously reported changes in suture-holding capacity. The decrease in suture-holding capacity during the first 3 days previously observed, the inflammatory phase, did not correlate with collagen content during this period. However, the increase in suture-holding capacity from the third to the seventh days, the phase of fibroplasia, correlated well with an increased collagen content in the intestinal wall in both the ileum and jejunum. Anastomotic strength and collagen content correlated well after the third day. Deposition of collagen, not only in the wound but also in the intestinal wall and within the loop of the suture during the phase of fibroplasia, is the likely explanation for the rapid increase of suture-holding capacity during this period.

The effects of severe zinc deficiency on intestinal amino acid losses in the rat

To determine whether intestinal amino acid losses might occur during zinc deficiency, labeled aminoisobutyric acid was given parenterally to zinc deficient rats and to appropriate zinc-sufficient controls. After 24 hours, the aminoisobutyric acid loss into the intestinal lumen was measured by in situ perfusion of isolated intestinal segments under conditions of either net water absorption or water secretion. Net amino acid losses were larger in the jejunum of the zinc deficient rats and losses were exacerbated during net water secretion in the jejunum and colon segments. The contribution of amino acid losses to fecal nitrogen, particularly during osmotic diarrhea, may be important in the growth retardation of zinc deficiency. Further, these alterations may indicate defective enterocyte transport functions during severe deficiency.

Immunocytochemical localization of 140 kD cell adhesion molecules in cultured chicken fibroblasts, and in chicken smooth muscle and intestinal epithelial tissues

A monoclonal antibody (JG22 MAb) that was previously raised to a chick embryo myogenic cell preparation had been shown to produce rounding and other morphological changes in myogenic cells in culture, and, in some cases, their detachment from the substratum. In other studies it was shown that the epitope recognized by JG22 was associated with a set of 140 kD cell surface glycoproteins. It is shown that this antigen occurs in a wide variety of cell types; in cultured fibroblasts, it is distributed equally between the dorsal and ventral cell surfaces shortly after plating, but appears to become concentrated on the ventral surface as cell spreading proceeds; by immunoelectron microscopic labeling experiments, it is absent from the focal adhesion contact sites formed by fibroblasts with their substrata and with one another, but is present in clusters at the edge of focal adhesions, and within the close contact sites and extracellular matrix contact sites; in smooth muscle cells, it is absent from the membrane-associated dense plaques, but is located in clusters at adjacent membrane sites; in intestinal epithelium, it is present in clusters at the basolateral membranes, but not at the microvilli or within junctional complexes of the brush border of the cell layers. These and other results are consistent with the suggestion that the antigen recognized by JG22 MAb is important cell adhesion molecules, and performs a characteristic function in a variety of cell-cell contacts and cell adhesions.

Selenium absorption by canine jejunum

Deficiency of the trace element selenium causes disease in domestic animals and may also be implicated in the pathogenesis of some human illness. In this study, the triple-lumen perfusion method was used to measure the rate of absorption of trace quantities of selenium (50 micrograms/liter in a physiological electrolyte solution) from the jejunum when given as D,L-selenomethione, D,L-selenocystine, or sodium selenite to healthy dogs in vivo. Selenium absorption from the test segment (expressed as percent administered dose per centimeter +/- SEM) was 1.97 +/- 0.04 from D,L-selenomethionine, 1.15 +/- 0.06 from D,L-selenocystine, and 0.51 +/- 0.07 from sodium selenite (P less than 0.01, N = 5). In separate studies in four anesthetized dogs, the jejunum was perfused with L-[75Se] selenomethionine while concentrations of 75Se were measured in the portal venous blood; these studies established that [75Se]selenomethionine disappearing from the gut lumen corresponded quantitatively to 75Se appearing in the portal venous effluent (74 +/- 6%) and incorporated into intestinal tissue (24 +/- 5%). These results are consistent with the hypothesis that the absorption of amino acid-bound selenium is accelerated by the specific amino acid active transport mechanisms in the gut mucosa. Sodium selenite is absorbed more slowly, possibly by simple diffusion through the intestinal mucosa, than the amino acid-bound selenium compounds.

Short-chain fatty acids in the small-bowel bacterial overgrowth syndrome

The short-chain fatty acids (SCFAs) have been measured by gas chromatography in fasting jejunal secretions, saliva, and feces from 8 patients with the small-bowel bacterial overgrowth syndrome (BO) and 9 control patients; in jejunal secretions and saliva from 6 healthy subjects; and in feces from 20 healthy subjects. The concentrations of SCFAs (median (range), mumol/l) in jejunal secretions of BO patients were as follows: total, 990 (210-12,370); acetic acid, 650 (170-6770); propionic acid, 110 (16-3070); isobutyric acid, 26 (1-310); n-butyric acid, 90 (12-1340); isovaleric acid, 35 (2-680); n-valeric acid, 7 (3-200). In BO patients the total concentration of SCFAs in jejunal secretions was approximately four times higher than in control patients (p less than 0.01) and in healthy subjects (p less than 0.025). The relative distribution of the acids resembled the distribution found in feces more than that of saliva or the normal jejunal secretions. These findings indicate that patients with BO have a colon-like flora in the small intestine and that the main part of the SCFAs in the jejunal secretions of these patients is produced by the altered microbial flora in the jejunum. Combined with other tests, analyses of intestinal SCFAs may prove to be valuable in the diagnosis of small-bowel bacterial overgrowth.

Distribution of neurotensin in the canine gastrointestinal tract

The distribution of immunoreactive neurotensin in the canine gastrointestinal tract from the esophagus to the rectum, as well as in the pancreas, was determined by a specific neurotensin radioimmunoassay. Immunoreactive neurotensin was found throughout the gastrointestinal tract and in the pancreas. The highest concentrations of immunoreactive neurotensin were found in the mucosal extracts of the jejunum (422 +/- 68 ng/gm) and ileum (3025 +/- 289 ng/gm). Small but substantial amounts of immunoreactive neurotensin were found in the esophagus, fundus (includes fundus and corpus), antrum, duodenum, colon, and pancreas. The concentrations of neurotensin in the mucosal extracts of the jejunum and ileum increased in a graded fashion from the proximal jejunum to the distal ileum. The neurotensin concentration in extracts of the seromuscular layers of jejunum (73 +/- 14 ng/gm) and ileum (187 +/- 38 ng/gm) were statistically higher in comparison with other gut loci.

Location and distribution of a receptor for the 987P pilus of Escherichia coli in small intestines

Frozen sections of rabbit or pig small intestines were stained with fluorescein-labeled antibody specific for the 987P receptor isolated from adult rabbit small intestines. The 987P receptor was present along the entire villous surface and in goblet cells in adult rabbits, but only in goblet cells in infant rabbits. In adult rabbits, the receptor was distributed equally in the jejunum and the ileum. Material antigenically similar to the rabbit 987P receptor was demonstrated in goblet cells in neonatal piglet ileum.

Distribution and molecular forms of peptides containing somatostatin immunodeterminants in extracts from the entire gastrointestinal tract of man and pig

Specimens from human porcine mucosal and muscular tissue from the entire gastrointestinal tract were extracted in acid ethanol, subjected to chromatography and analysed for somatostatin-like immunoreactivity by region-specific radioimmunoassays. The concentration of somatostatin-like immunoreactivity from man and pig ranged from 1.13 +/- 0.37 to 101.15 +/- 33.93 pmol/g wet weight, and from 7.64 to 159.48 +/- 23.79 pmol/g wet weight, respectively. In both species the highest concentrations were found in the jejunum. The immunoreactivity in intestinal mucosal extracts was distributed among four major peaks, two of which were identified by HPLC as somatostatin 1-28 and somatostatin 1-14, respectively. A peak of approx. 10 kDa was resolved by ion exchange plus HPLC into three components, two containing at least part of the somatostatin 1-14 sequence as well as (part of) the somatostatin 1-28(1-14) sequence (but differing in charge), the third containing only the 1-28(1-14) sequence. These peptides probably represent uncleaved and partially cleaved prosomatostatin. The fourth component to be identified by gel filtration was slightly larger than somatostatin 1-14. Extracts from the antrum, the pancreas and from muscular tissues contained almost exclusively somatostatin 1-14, and very little somatostatin 1-28, indicating that the somatostatin precursor is processed differently at these sites. Furthermore, extracts of porcine gastric antrum, analysed for somatostatin 1-28(1-14) immunoreactivity, showed two immunoreactive forms in the mucosa and three major forms in the muscular layers.

Coexistence of peptide YY and glicentin immunoreactivity in endocrine cells of the gut

Endocrine cells containing peptide YY (PYY) were numerous in the rectum, colon and ileum and few in the duodenum and jejunum of rat, pig and man. No immunoreactive cells could be detected in the pancreas and stomach. Coexistence of PYY and glicentin was revealed by sequential staining of the same section and by staining consecutive semi-thin sections. Since the PYY sequence is not contained in the glucagon/glicentin precursor molecule the results suggest that the PYY cell in the gut expresses two different genes coding for regulatory peptides of two different families.

Cellular retinol-binding protein. Quantitation and distribution

The distribution of cellular retinol-binding protein was localized in three tissues involved in the transport of vitamin A using the peroxidase-antiperoxidase technique. In addition, a sensitive radioimmunoassay was developed to quantitate cellular retinol-binding protein in various tissues. In liver, the protein was found in the hepatocytes and in the perisinusoidal fat-storing cells. The columnar epithelial cells of the jejunum and the cells of the proximal tubuli also exhibit high concentrations of cellular retinol-binding protein. Estimations of the content of the protein in various tissues from normal and retinol-deficient rats showed that the retinol status did not affect the tissue levels or the subcellular distribution. The appearance of high amounts of this protein in cells involved in the dietary uptake, storage, mobilization, and resorption of vitamin A suggest that one function of the cellular retinol-binding protein is to act as a vehicle in the intracellular transport of retinol through these organs.

A protein inhibitor of phosphofructokinase from the mucosa of rat small intestine. A mechanism for the regulation of glycolysis that is independent of glucose

A protein inhibitor of phosphofructokinase was detected by the chromatography on Sephadex G-100 of crude mucosal extracts prepared from the small intestine of starved rats. Addition of the protein inhibitor to extracts from fed rats increased the K0.5 of phosphofructokinase for fructose 6-phosphate to that for starved rats. The protein inhibitor provides a mechanism for the hormonal regulation of glycolysis independently of glucose supply.

Oophorectomy leads to a selective decrease in 1,25-dihydroxycholecalciferol receptors in rat jejunal villous cells

The distribution of 1,25-dihydroxycholecalciferol [1,25-(OH)2D3] receptors in isolated rat jejunal cells was investigated in normal and oophorectomized female rats and also in vitamin D-deficient animals. Oophorectomy caused a selective reduction in 1,25-(OH)2D3 receptor numbers in villous cells, the reduction being similar (37%) in both normal and vitamin D-deficient animals. The data suggest that this effect of oophorectomy reflects a direct effect of oestrogens in the intestinal synthesis of 1,25-(OH)2D3 receptors, independent of the renal synthesis of 1,25-(OH)2D3. It is proposed that this phenomenon may, in part, explain the reduction in calcium absorption which occurs after oophorectomy or the menopause.

Glucagon-37 (oxyntomodulin) and glucagon-29 (pancreatic glucagon) in human bowel: analysis by HPLC and radioreceptorassay

A method for assaying specifically the biologically active peptides Glucagon-37 (G-37/Oxyntomodulin/bioactive Enteroglucagon) and Glucagon-29 (G-29/pancreatic Glucagon) has been developed by use of high performance liquid chromatography (HPLC) of crude tissue extracts followed by radioreceptorassay in liver membranes. The peaks observed with this method in samples from human bowel have also been analysed in two other assays: stimulation of cyclic AMP accumulation in gastric glands and radioimmunoassay. Owing to the different patterns of activity of porcine G-37 and G-29 in these assays, the comparison of the data obtained allows to discriminate between the two peptides. The same behaviour in both HPLC and the three assays of the human peaks on one hand and the porcine peptides on the other strongly suggests that human intestine contains a very similar or the same molecules as that isolated from the porcine tissues. Whatever the portion of small intestine, G-37 represented ca 90% of G-37 + G-29. A decreasing concentration gradient of both G-37 and G-29 was also observed from ileum to descending colon.

Recovery of intestinal membrane binding sites for K88 E. coli from pig mucosal organ cultures

Putative receptors for K88 + E. coli from piglet intestinal epithelium were released into the organ culture medium and were demonstrated by direct binding with K88 + E. coli through the utilization of an in vitro binding procedure or by immunoprecipitation with K88 antigen. Incorporation of 14C-glucosamine by newborn to day old and 3-week to 6-week old piglet jejunal and ileal mucosa, in organ culture, occurred throughout the 24 hr culture period. Uptake in both age groups and both areas of the intestine was similar with a somewhat greater incorporation by the older age group. Secretion of 14C-glucosamine-labeled components into the culture medium was demonstrated by gel filtration of the concentrated medium. Some large molecular weight components eluted in the void volume in excess of 2 X 10(6) daltons. A second peak of activity was spread from approximately 690K to 25K daltons. All eluted fractions demonstrated binding to K88 + E. coli. Antibodies to purified brush borders from susceptible pigs produced prominent precipitation bands following double diffusion with concentrated organ culture media which confirmed that the organ culture media contained labeled proteins of brush border origin. Immunoprecipitation of the intestinal mucosal organ culture media with K88 + pili and pilus antisera, followed by electrophoresis with SDS and reduced conditions, demonstrated a subunit of approximately 35K daltons.

Abundance of VIP in duodenal mucosa of coeliacs and duodenal ulcer patients

VIP levels were determined in gastroduodenal mucosal biopsies of 8 duodenal ulcer patients, of 5 coeliac sprue patients, and of 8 volunteers without upper gastrointestinal disease. In duodenal ulcer patients, mucosal VIP concentrations were significantly elevated in the proximal duodenum (e.g., in the duodenal bulb 225 +/- 48 versus 95 +/- 17 pmol/g in controls), while in coeliac sprue VIP levels tended to be increased in the whole duodenum and upper jejunum (e.g., descending duodenum 409 +/- 161 versus 81 +/- 16, p less than 0.05). In both disease entities, the rise in mucosal VIP may be a reaction of the peptidergic nervous system to chronic mucosal irritation and a reason for enhanced fluid and electrolyte secretion in the affected areas.

Chemical and histochemical studies of normal and diseased human gastrointestinal tract. II. A comparison between histologically normal small intestine and Crohn's disease of the small intestine

Comparative chemical and histochemical studies were performed on formalin-fixed, surgical specimens of human small intestine from cases of Crohn's disease and normal controls. The sialic acids of the crude glycoproteins isolated from normal ileum were significantly less neuraminidase-susceptible and more C4 substituted (P less than 0.01) than those of the glycoproteins isolated either from normal upper small intestine (duodenum and jejunum) or from cases of Crohn's disease of the ileum. Fractionation yielded two major sialic acid-containing fractions, eluting from DEAE-cellulose with 0.2 M or 0.3 M sodium chloride. Both fractions contained fucose, galactose, glucosamine and galactosamine in addition to sialic acids both with and without O-acyl substituents at position C4 and/or in the side-chain (side-chain O-acylated sialic acids were also detected by histochemical procedures). The fractions differed significantly from one another with respect to the neuraminidase susceptibility of their sialic acids (P less than 0.01), the percentage of C4 (P less than 0.01) and side-chain substituted sialic acids (P less than 0.05), and the molar fucose-sialic acid ratio (P less than 0.05). The O-acyl substitution patterns of the sialic acids of both the 0.2 M and 0.3 M fractions of the upper small intestine glycoproteins differed significantly from those of the corresponding fractions from normal ileum, while the sialic acids of the 0.2 M fractions from Crohn's disease of the ileum differed significantly from normal with respect to neuraminidase susceptibility (P less than 0.01) and percentage C4 substitution (P less than 0.01); the 0.3 M fractions differed only in the percentage of sialic acids substituted at C4. The differences between the sialic acids from the normal and Crohn's disease specimens were shown to be independent of either the anatomical origin of the specimen or the histopathological sub-group of the Crohn's disease specimens; no significant differences were noted between the sub-groups but all the sub-groups differed from normal.

Neuronal cholecystokinin, gastrin-releasing peptide, neurotensin, and beta-endorphin in the intestine of the guinea pig. Distribution and possible motor functions

The guinea-pig intestine was found to harbor nerve fibers containing immunoreactive cholecystokinin (CCK), gastrin-releasing peptide (GRP), neurotensin or beta-endorphin. Such fibers occurred in the myenteric and submucous ganglia and in the smooth muscle. GRP- and CCK-fibers, in addition, were found in the mucosa. Following colchicine treatment, neuronal perikarya in the myenteric ganglia displayed CCK-, GRP-, or beta-endorphin immunoreactivity. CCK-immunoreactive perikarya were located also in the submucous ganglia. Neurotensin-immunoreactive cell bodies could not be detected. The presence of immunoreactive neuronal perikarya in intramural ganglia indicates that CCK-, GRP- and beta-endorphin-containing fibers are intrinsic to the gut wall. GRP, neurotensin, and beta-endorphin were identified in extracts of smooth muscle by immuno-chemical and chromatographic analysis. CCK-8, GRP and neurotensin contracted the isolated taenia coli. Tetrodotoxin reduced the response to CCK-8 but not that to GRP and neurotensin, suggesting that the two latter peptides act directly on smooth muscle receptors. The effect of CCK-8 is partly mediated by cholinergic nerves, since not only tetrodotoxin but also atropine greatly reduced the CCK-8-induced contractile response. The substance P (SP) antagonist, (D-Pro2, D-Trp7,9)-SP1-11 had no effect on the CCK-8-induced contraction of the taenia. CCK-8 enhanced the SP-mediated (atropine-resistant) contractile response to electrical stimulation but not that mediated by acetylcholine. beta-Endorphin had no effect on the tension of the muscle but reduced the response to electrical stimulation (cholinergic as well as SP-mediated) through a naloxone-sensitive mechanism. While CCK-8 and beta-endorphin seem to play neuromodulatory roles in the taenia coli, the significance of GRP and neurotensin remains enigmatic.

Plasma zinc and jejunal mucosal zinc in gastric carcinoma

The plasma zinc in both gastric carcinoma and in other malignancy requiring surgical treatment was significantly higher than in patients with non-malignant disease but there was no difference between the gastric carcinoma and other malignancy levels. The jejunal mucosal zinc levels showed similar results with both gastric carcinoma and other malignancy requiring surgery being higher than non malignant disease but no difference between them. It was not possible to correlate plasma and jejunal mucosal zinc levels overall. This increased level of body zinc does not appear to be specific to gastric carcinoma and may be associated with non specific mucin changes.

Substance P-containing nerves in the human small intestine. Distribution, ultrastructure, and characterization of the immunoreactive peptide

Light and electron microscopic immunocytochemical techniques were used to examine the distribution and ultrastructure of substance P-immunoreactive nerves in human jejunum and distal ileum. The organization of human enteric substance P-containing nerves closely resembled that in other species. Dense arrays of varicose immunofluorescent fibers occurred in myenteric and submucous ganglia (which contained immunoreactive nerve cell bodies) and in the mucosa. There were fibers in both muscle layers, in the muscularis mucosae, and around blood vessels. Fibers in the myenteric plexus contributed to both ascending and descending pathways. Substance P-immunoreactive axon profiles contained small round and large round vesicles and were apposed to nerve cell bodies, and nonimmunoreactive and immunoreactive axon profiles. Synapselike contacts were occasionally observed on nerve cell bodies and processes. The substance P-like material was characterized by high pressure liquid chromatography and radioimmunoassay and found to be indistinguishable from the authentic undecapeptide. These results suggest that enteric nerves containing substance P may play similar roles in humans as in other species.

Immunolocalization of the 33 kD protein in the microvilli of rabbit small-intestinal epithelial cells

Rabbit small-intestinal microvilli isolated by a Ca2+ precipitation method contain a 33 kD protein, which has not been observed in microvilli isolated in the presence of Ca2+-chelators. The intracellular localization of this protein in rabbit intestinal epithelial cells was studied by immunofluorescence and immunoperoxidase microscopy, and was compared with that of aminopeptidase M, a well-known microvillus membrane-bound enzyme. The results obtained show that the 33 kD protein is located in the inside of the microvillus, but not in the terminal web of the epithelial cell. The protein may also be located on the basolateral surface of the cell.