Assembly of the intrinsic factor domains and oligomerization of the protein in the presence of cobalamin.

Human intrinsic factor (IF) was purified from the recombinant plant Arabidopsis thaliana by affinity chromatography. Cobalamin (Cbl) saturated protein was separated by gel filtration into peaks I and II, which contained according to SDS electrophoresis the 50 kDa full-length protein IF(50) and a mixture of two fragments, respectively. Two components of peak II were identified as the 30 kDa N-terminal peptide IF(30) and the 20 kDa C-terminal glycopeptide IF(20). Measurements of M(w) under the nondenaturing conditions were conducted by static light scattering. They revealed 100 kDa IF dimers in peak I, whereas 50 kDa cleaved monomers were found in peak II. The protein devoid of Cbl dissociated to the elementary units incapable of association in the absence of Cbl. The individual proteolytic fragments bound Cbl at high concentration of the ligand; however, neither IF(30).Cbl nor IF(20).Cbl oligomerized. A mixture of two fragments IF(30) + IF(20) and Cbl produced a firm complex, IF(30+20).Cbl, which could not associate to dimers. In contrast to IF(30+20).Cbl, the saturated full-length monomers IF(50).Cbl dimerized with K(d) approximately 1 microM. We suggest a two-domain organization of the full-length protein, where two distant units, IF(30) and IF(20), can be assembled only by Cbl. They are connected by a protease-sensitive link, whose native structure is likely to be important for dimerization. However, linkage between two domains is not compulsory for Cbl binding. Advantages of the two-domain structure of IF are discussed.

Human intrinsic factor expressed in the plant Arabidopsis thaliana.

Intrinsic factor (IF) is the gastric protein that promotes the intestinal uptake of vitamin B12. Gastric IF from animal sources is used in diagnostic tests and in vitamin pills. However, administration of animal IF to humans becomes disadvantageous because of possible pathogenic transmission and contamination by other B12 binders. We tested the use of recombinant plants for large-scale production of pathogen-free human recombinant IF. Human IF was successfully expressed in the recombinant plant Arabidopsis thaliana. Extract from fresh plants possessed high B12-binding capacity corresponding to 70 mg IF per 1 kg wet weight. The dried plants still retained 60% of the IF activity. The purified IF preparation consisted of a 50-kDa glycosylated protein with the N-terminal sequence of mature IF. Approximately one-third of the protein was cleaved at the internal site em leader PSNP downward arrow GPGP. The key properties of the preparation obtained were identical to those of native IF: the binding curves of vitamin B12 to recombinant IF and gastric IF were the same, as were those for a B12 analogue cobinamide, which binds to IF with low affinity. The absorbance spectra of the vitamin bound to recombinant IF and gastric IF were alike, as was the interaction of recombinant and native IF with the specific receptor cubilin. The data presented show that recombinant plants have a great potential as a large-scale source of human IF for analytical and therapeutic purposes.

Effect of gastrointestinal proteases on purified human intrinsic factor-vitamin B12 (IF-B12) complex.

Intrinsic factor (IF) from human gastric juice was purified and complexed with vitamin B12 (IF-B12 complex) on Sepharose-vitamin B12 affinity matrix. By labeling studies, using [(57)Co] vitamin B12 and (125)I, the specific B12 binding activity of IF was found to be 23 microg B12/mg protein, and the molecular size by gel filtration 60 kDa. Proteolysis of the IF-B12 complex by sequential treatment with pepsin, trypsin, alpha-chymotrypsin and carboxypeptidase A, followed by chromatography of proteolysed complex and IF-B12 showed higher mobility of proteolysed fraction. Gel filtration, however, showed same molecular size for both proteolysed and the IF-B12 complex. On SDS-PAGE, purified IF-B12 appeared as a single band of 60 kDa. The proteolysed complex had higher mobility on SDS-PAGE and did not bind to zirconium phosphate gel. Immunodiffusion with rabbit antisera had positive reaction with IF-B12, but there was no reaction with the proteolysed sample.

Purification by cobalamin-Sepharose affinity chromatography and intrinsic factor-binding activity of an extramembrane proteolytic product from pig ileal mucosa.

We have purified a cobalamin-binding protein obtained by papain digestion of pig intestine by cobalamin-AH-Sepharose affinity chromatography, with a purification factor of 17,300, a yield of 63% and a cobalamin-binding activity of 11,260 pmol/mg of protein. The protein contained 3.8% carbohydrate and was O- and N-glycosylated. Its molecular mass was 69 kDa on SDS/PAGE and its isoelectric point was 5.1. It had a binding activity for both [57Co]cobalamin and [57Co]cobalamin-intrinsic factor in native PAGE autoradiography and it inhibited the binding of intrinsic factor to the intact intestinal receptor with an IC50 of 49.31 nmol/l in a radioisotope assay. In conclusion, the purified protein shared a binding activity for both cobalamin and intrinsic factor-cobalamin complexes and could correspond to the extracellular domain of the ileal intrinsic factor receptor.

Gastric intrinsic factor and its receptor.

Cbl metabolism has been the subject of many studies since the existence of Cbl was suspected in the first decades of the twentieth century. These studies have confirmed the high complexity of the assimilation of Cbl by the organism. During absorption, Cbl is bound to two glycoproteins, Hc and IF, in a consecutive manner. Over the last few years, it has been demonstrated that Cbl bound to Hc in the stomach is only transferred to IF after the action of pancreatic trypsin. It is also possible that Hc-bound biliary Cbl is transferred to IF in this way and that the Cbl in the Cbl-IF complex is absorbed in the terminal ileum, thus constituting an enterohepatic cycle. Knowledge concerning the sites of synthesis and secretion of IF is becoming more detailed due to the use of immunocytochemistry and in situ hybridization techniques. It is now certain that in man, IF is not only localized in gastric parietal cells, but also in other foregut-derived cells. This observation may explain the multiple physiological stimuli involved in mediating IF secretion. Determination of the molecular structure of purified Cbl binders can be added to the significant progress made due to the application of molecular biology techniques to the field of isolation and structural characterization of cDNA encoding Cbl binders, and particularly IF. Studies of IF, Hc and TC in different species and those into the properties of acceptor fragments have allowed the distinction between the Cbl binding site on IF and the IF-Cbl binding site on the IFCR. The absence of experimental models cause difficulties in studying transcytosis of Cbl through the enterocyte. There are also problems in determining the structure of IFCR as it is difficult to obtain a large quantity of a molecule which denatures very quickly. Studies into IFCR expression in polarized cancerous cells of intestinal or renal origin, including the effects of different pharmacological agents, along with the results of immunochemical investigations are beginning to clarify the pathway involved in the transport of Cbl through the enterocyte.

Intrinsic factor covalently bound to Sepharose as affinity medium for the purification of a soluble intrinsic factor receptor from human urine.

We have identified a soluble receptor for intrinsic factor (IF) in human urine. The purification of this protein by affinity chromatography required a preliminary purification of IF from hog pyloric mucosal extract. This was achieved by thermolabile cobalamin-ethanol-aminohexane Sepharose affinity chromatography with a 133-fold purification, a yield of 45% and a specific binding activity of 15720 pmol/mg protein. The purified Cbl-IF complex was coupled to epoxy-Sepharose with a yield of 23.8% and a specific activity of 1.2 nmol per mol of gel. The soluble IF receptor was purified form 200 ml of urine concentrate of pregnant women. Desorption was performed at pH 5.0 and in the presence of 5 mM EDTA. The soluble IF receptor was purified 17,200-fold with a yield of 52% and a IF binding capacity of 3260 pmol per mg of protein. A single protein with a Mr of 70,000 was found in silver-stained SDS-PAGE.

Secretion of intrinsic factor from cultured rat gastric chief cells.

Intrinsic factor (IF) is a vitamin B12 binding protein that is secreted from the gastric mucosa. We tested secretagogues which stimulate IF secretion in rat gastric perfusion and found that carbachol and cholecystokinin octapeptide (CCK-8) stimulated secretion, but histamine and tetragastrin did not. To confirm these results, we examined IF secretion from isolated rat chief cells. For this purpose, we established an enzyme immunoassay (EIA) using an avidin-biotin peroxidase complex to measure small amounts of IF. To prepare an anti-rat IF, IF was isolated from the stomach, and was injected into a rabbit for immunization. Rat gastric chief cells were isolated from the gastric mucosa with Dispase and a Percoll gradient centrifugation, and were cultured. We examined the effects of chemicals by adding them to culture dishes of chief cells in a CO2 incubator. Released IF in culture medium was determined by EIA. Carbachol, CCK-8 and secretin stimulated IF secretion from cultured chief cells, while histamine and tetragastrin did not; Forskolin and A23187 also stimulated the secretion. We concluded that carbachol and CCK-8 stimulated IF secretion via an increase of intracellular Ca2+ concentration and that secretin did so via a cAMP accumulation.

The intrinsic factor (IF)-cobalamin receptor binding site is located in the amino-terminal portion of IF.

Intrinsic factor has two binding sites, one each for cobalamin and for the ileal receptor recognizing the intrinsic factor-cobalamin complex. To obtain initial functional mapping of these domains, cDNAs encoding intact rat and human intrinsic factor or fragments thereof were expressed transiently in COS-1 cells or in an in vitro transcription/translation system. Deletion of as little as 12% of the amino acids from the carboxyl terminus resulted in loss of cobalamin binding activity. On the other hand, the receptor binding region of intrinsic factor appears localized to a restricted region in the amino-terminal portion of the protein. Only those transcription/translation fragments of rat or human intrinsic factor tested that contained amino acid residues 25 to 62 (out of 399) showed calcium-dependent binding to isolated kidney brush borders, the shortest sequence corresponding with 20 consecutive amino acids. In contrast, a 232-amino acid carboxyl-terminal fragment of rat intrinsic factor and 243- and 338-amino acid carboxyl-terminal fragments of human intrinsic factor showed no receptor binding activity.

In vitro expression and secretion of functional mammalian intrinsic factor using recombinant baculovirus.

Intrinsic factor was produced at levels of 1-2 mg per 1 (0.25 micrograms per 10(6) cells) by growth of recombinant baculovirus-infected Sf9 cells in spinner culture. The recombinant IF showed a binding affinity for cobalamin (2.6.10(-10) M) and for the intrinsic factor-cobalamin receptor (3.5.10(-10) M) nearly identical with native IF. Purification of the recombinant intrinsic factor could be accomplished by affinity chromatography, but final purification by gel chromatography (FPLC) was necessary to separate intrinsic factor from a 62 kDa protein secreted from uninfected Sf9 cells. This protein binds selectively to the cobalamin-Sepharose column, but demonstrates no cobalamin binding activity after elution. Microgram quantities of radiolabelled protein could be produced for metabolic and autoradiographic studies. The stability of intrinsic factor to pancreatic proteinases was nearly identical with human gastric intrinsic factor, both native and recombinant as produced in mammalian cells. Glycosylation of the intrinsic factor was demonstrated by lectin binding to the recombinant protein separated on SDS-PAGE, and by a shift in apparent molecular mass from 47 kDa to 43 kDa following treatment of Sf9 cells with tunicamycin. Most of the recombinant IF was produced by Sf9 cells in the first 48 h post infection.

Leupeptin and ammonium chloride inhibit intrinsic factor mediated transcytosis of [57Co]cobalamin across polarized renal epithelial cells.

The [125I] intrinsic factor (IF) mediated transcytosis of [57Co]Cyanocobalamin (Cbl) by polarized opossum kidney cells was inhibited (greater than 80%) by preincubation of the cells with lysosomotropic agents leupeptin or ammonium chloride. Inhibition of Cbl transcytosis resulted in the intracellular accumulation of both [125I]IF (48 kDa) and [57Co]Cbl. Intracellular degradation of [125I]IF occurred during normal cellular transcytosis of [57Co]Cbl and in one h following internalization the major intracellular degradation products of IF were two polypeptides of Mr 29 kDa and 19 kDa. The size of the major degradation product of IF in the basolateral media was 10 kDa. Based on these results, we suggest that IF is internalized by the renal epithelial cells and is degraded by leupeptin-sensitive acid proteases during Cbl transcytosis.

Human gastric intrinsic factor: characterization of cDNA and genomic clones and localization to human chromosome 11.

A human gastric intrinsic factor (IF) cDNA clone was isolated using a rat cDNA clone as a probe. Comparison of the predicted amino acid sequence revealed 80% identity of human IF with rat IF. These cDNA clones were used to isolate and map two overlapping clones encoding the human IF gene. The first exon of the cloned region (exon 2) contains 30 bp of the 5' untranslated region, the signal peptide, and the first 8 amino acids of the mature protein. Exons 3-10 encode the remainder of the coding and 3' noncoding regions. Southern analysis of genomic DNA indicated the presence of a single human IF gene and also revealed the presence of strong hybridizing sequences in genomic DNA from monkey, rat, mouse, cow, and human, suggesting that the IF gene is well conserved. The IF gene was localized to human chromosome 11 by concurrent cytogenetic and cDNA probe analysis of a panel of human X mouse somatic cell hybrids. Southern analysis of genomic DNA from patients with congenital pernicious anemia (lacking intrinsic factor) revealed normal restriction fragment patterns, suggesting that a sizable gene deletion was not responsible for the deficiency.

Expression of cobalamin transport proteins and cobalamin transcytosis by colon adenocarcinoma cells.

Human colon adenocarcinoma (Caco-2) cells express both intrinsic factor-cobalamin receptor and transcobalamin II (TC II). The expression of these activities began to rise by day 6 and reached peak levels between 10 and 15 days in culture. The postconfluent Caco-2 cell membranes bound approximately 30-35 fmol of intrinsic factor (IF) [57Co]Cbl/mg protein. The size of the mature receptor expressed in the apical brush border had a relative molecular mass of 230 kDa. The intracellular form of TC II had a Mr of 43, 5 higher than the secreted form of TC II. TC II was secreted unidirectionally via the basolateral direction when Caco-2 cells were grown on culture inserts. When grown on culture inserts, the Caco-2 cells were polarized (electrical resistance greater than 200 omega/cm2) and transcytosed [57Co]Cbl bound to IF from apical-to-basal but not from basal-to-apical direction. Under these conditions, [57Co]Cbl complexed to haptocorrin was not transported. These cells also transcytosed free [57Co]Cbl, although less efficiently. The [57Co]Cbl transcytosed using either IF[57Co]Cbl or free [57Co]Cbl as ligands was bound exclusively to TC II. Intracellular [57Co]Cbl decreased during transcytosis with a slow (t1/2 = 4 h) transfer of [57Co]Cbl from IF to TC II. These results show that the transport of Cbl in Caco-2 cells is very similar to the human enterocyte system.

Monoclonal antibodies to human intrinsic factor.

Mice were immunized with human intrinsic factor, and their lymph node cells were fused with a myeloma cell line by standard hybridoma techniques. Eleven of the resulting 227 hybridomas secreted immunoglobulin G capable of binding to intrinsic factor-cobalamin complex. Cloning by limiting dilution gave 6 clones secreting anti-intrinsic factor antibodies that bound human intrinsic factor-cobalamin complex with affinities of 13-116 nM; 3 antibodies also bound rabbit intrinsic factor-cobalamin complex. Five antibodies inhibited to some degree the binding of cobalamin by intrinsic factor, and 2 also prevented attachment of intrinsic factor-cobalamin complex to guinea pig ileal receptors. Anti-rabbit intrinsic factor antibodies specifically precipitated a peptide of molecular weight 53,000, corresponding to the molecular weight of rabbit intrinsic factor from homogenates of rabbit gastric mucosal explants biosynthetically labeled with [35S]methionine and from culture medium in which the explants were incubated. Indirect fluorescence immunocytochemistry with the antibodies in human and rabbit gastric mucosal sections showed intense selective staining of parietal cells. These results (a) document species differences between human and rabbit intrinsic factors not previously demonstrable with polyclonal anti-intrinsic factor sera; (b) confirm earlier evidence that cobalamin binding and receptor functions occur at separate sites in intrinsic factor; and (c) provide a useful approach to studying structure-function relations of the intrinsic function molecule.

Physicochemical characterization and biological activity of intrinsic factor in cystic fibrosis.

Absorption of crystalline labeled cobalamin is strongly decreased in cases of cystic fibrosis. In order to determine if this is due to an alteration or a lack of activation of intrinsic factor by proteases, the physicochemical properties and biological activity of intrinsic factor have been studied. Intrinsic factor was purified 800-fold from stimulated gastric juice of cystic fibrosis patients with a yield of 64.2%. Cystic fibrosis intrinsic factor had an estimated Mr of 57,000 in SDS-polyacrylamide gel electrophoresis. Its carbohydrate content resembled that of normal human intrinsic factor, except that the ratio fucose/sialic acid was higher (6.1 and 1.6, respectively) and that the content in N-acetylgalactosamine was decreased. The same alterations in carbohydrate composition were observed for Hc purified from cystic fibrosis saliva. Purified intrinsic factor from cystic fibrosis gastric juice was biologically active in vitro in the presence of ileal solubilized receptor as well as in vivo (Schilling test). The fate of iodinated cystic fibrosis intrinsic factor in guinea pig ileum studied by high-resolution radioautography was similar to that of normal intrinsic factor. In conclusion, despite modifications of the carbohydrate content of the molecule, the biological activity of intrinsic factor is not altered in cases of cystic fibrosis. The malassimilation of crystalline cobalamin observed in cystic fibrosis is due to a mechanism independent from intrinsic factor secretion.

Purification of intrinsic factor receptor from pig ileum using as affinity medium human intrinsic factor covalently bound to Sepharose.

Intrinsic factor receptor was purified from hog ileum using human intrinsic factor covalently bound to Sepharose. A yield of 49.6% and a specific activity of about 2500 pmol/mg protein were achieved. The purified receptor was very unstable: 24 h of storage or addition of sodium phosphate precipitated it. The association constant of the receptor for the cyano[57Co]cobalamin-intrinsic factor complex was estimated to be 2.1 nM-1. In native polyacrylamide gel electrophoresis it resolved in two 256 and 320 kDa bands; beta-mercaptoethanol treatment cleared it into four bands corresponding to molecular masses of 107, 81.8, 63.5 and 53.2 kDa. An additional 39.3 kDa band was considered to be an artefact due to the presence of Triton X-114. Isoelectric focusing polyacrylamide gel electrophoresis resolved the receptor into two isoproteins isoelectric at pH 4.7 and 5.1. A similar result was obtained in column electrofocusing with the 125I-iodinated receptor. The 125I-labelled receptor did not crossreact with rabbit anti-human intrinsic factor antiserum. The electrophoretic properties of the receptor purified with intrinsic factor covalently bound to Sepharose were compared to those of the receptor purified by the use of the classical cobalamin-affinity medium. It was concluded that a disassembled receptor was produced using the classical method.

Identification and characterization of a pancreatic intrinsic factor in the dog.

An intrinsic factor has been identified in the canine pancreas, and output and properties of this protein have been compared with those of gastric intrinsic factor in the dog. Mean concentrations of intrinsic factor and peak outputs per minute were approximately 5- to 10-fold higher in pure pancreatic juice after stimulation with secretin and cholecystokinin, respectively, than in pentagastrin-stimulated gastric juice. Purified gastric and pancreatic intrinsic factors had an identical molecular mass of 65 kDa, estimated by gel filtration on Sephacryl S-200, while sodium dodecyl sulfate-polyacrylamide gel electrophoresis demonstrated single bands corresponding to 53 kDa. Immunoblots showed that rabbit polyclonal antiserum to canine gastric intrinsic factor cross-reacted with canine pancreatic intrinsic factor. Gastric and pancreatic intrinsic factor-cyano[57Co]cobalamin complexes exhibited comparable association constants for ileal receptors in canine brush-border vesicles, while there was minimal binding to jejunal vesicles. These findings demonstrate that the canine pancreas is an important source of an intrinsic factor that closely resembles gastric intrinsic factor in the dog.

Lectin binding to the porcine and human ileal receptor of intrinsic factor-cobalamin.

The purified porcine receptor for the intrinsic factor-cobalamin complex bound to concanavalin A, lentil lectin and wheat germ lectin covalently coupled to Sepharose and was eluted with the corresponding soluble sugars. In contrast, human intrinsic factor bound efficiently to concanavalin A, to some extent to lentil lectin, but only slightly to wheat germ agglutinin. The binding of IF-Cbl to the receptor was inhibited when the receptor was pre-incubated with soluble wheat germ agglutinin, with an inhibition constant estimated to be 1.9 mumol/l. After transfer of the purified receptor from SDS-PAGE to Immobilon, ligand blotting of the purified receptor with iodinated lectin showed that concanavalin A and lentil lectin bound to three (75, 56 and 43 kDa) components but that wheat germ agglutinin bound only to the 75 kDa component. These results showed that the alpha subunit of the receptor could bind to wheat germ agglutinin, resulting in an inhibition of its binding with intrinsic factor. Both binding sites of intrinsic factor and of wheat germ agglutinin could be located near to each other.

Effect of glycosidases and proteinases on cobalamin binding and physicochemical properties of purified saturated haptocorrin and intrinsic factor.

The effect of exoglycosidase, N-glycanase, trypsin and chymotrypsin was studied on the binding capacity and physicochemical properties of intrinsic factor and of haptocorrin using Superose 6 gel filtration. Intrinsic factor was purified as recently described by us. Haptocorrin was purified 6000-fold from human saliva using thermolabile affinity chromatography and high-performance cationic exchange chromatography with a specific activity of 20.6 nmol of cobalamin (Cbl) per mg protein and a yield of 44.7%. Exoglycosidases provoked a decrease of 54.3 and 78.2% of the Cbl binding capacity of haptocorrin and intrinsic factor, respectively. The sequential incubation of haptocorrin and intrinsic factor wit exoglycosidases and proteinases provoked a decrease of, respectively, 100 and 92.7% of their Cbl binding capacity, whereas the incubation with proteinase decreased the Cbl binding capacity of, respectively, 67.9 and 7.9%. The result of the incubation of [3H]intrinsic factor or [3H]haptocorrin with chymotrypsin and trypsin gave, respectively, no change in the elution position and a shift corresponding to a decrease of 50% of the estimated molecular mass. The estimated molecular mass of Cbl-intrinsic factor and of Cbl-haptocorrin decreased, respectively, to 57.1 kDa and to 88.1 kDa after incubation with exoglycosidases. It was concluded that (1) the carbohydrate core of intrinsic factor protects the whole protein whereas the carbohydrate core of haptocorrin protects only half part of the protein and (2) the carbohydrates are implicated in the formation of the cobalamin binding site of haptocorrin and intrinsic factor.