Molecular properties of neurotensin receptors in rat brain. Identification of subunits by covalent labeling

Neurotensin binding sites in rat brain synaptic membranes were specifically and covalently labeled by two methods. In the first, a photoreactive and highly radioactive analogue of neurotensin, 125I-labeled azidobenzoyl[Trp11]neurotensin, was synthesized and used to photoaffinity label neurotensin receptors. In the second, the reversible association between neurotensin receptors and 125I-labeled[Trp11]neurotensin, a radioactive but nonphotoreactive analogue of neurotensin, was made irreversible by means of disuccinimidyl suberate, a bifunctional cross-linking reagent. Analysis of synaptic membranes by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography revealed that using both methods the same two protein bands with apparent molecular weights of 49,000 and 51,000 were specifically labeled. Identical results were obtained with or without reduction of the photolabeled membranes by beta-mercaptoethanol before electrophoresis. Variation of the ligand concentration did not modify the relative labeling intensities of the two bands, indicating that the high- and low-affinity neurotensin binding sites previously detected in rat brain synaptic membranes have similar molecular structures. These results indicate that neurotensin receptors in rat brain may be composed of two different protein subunits with similar molecular weight of about 50,000, that are linked together by noncovalent bonds.     

Radiolabeled Ligands Specific for the G Protein-Coupled State of Neurotensin Receptors

Abstract: Radiolabeled analogues of neuromedin N have been prepared by acylation of the α, ε1, and ε2 amino groups of [Lys²]neuromedin N (Lys-Lys-Pro-Tyr-Ile-Leu) either with the ¹²⁵I-labeled Bolton-Hunter reagent or with N -succinimidyl[2,3-³H]propionate. The binding properties of the purified analogues toward newborn mouse brain homogenate or toward membranes of cells transitorily (COS) or permanently (AA1) transfected with the cloned rat brain neurotensin receptor cDNA were evaluated and compared with those of radiolabeled neurotensin. The α-modified analogue of [Lys²]neuromedin N behaves exactly like neurotensin in these binding experiments, whereas the ε1- and ε2-modified analogues selectively recognize the fraction of neurotensin binding sites that is sensitive to GTPγS. The proportion of neurotensin receptors coupled to GTP binding proteins is ∼50% in membranes of newborn mouse brain or of AA1 cells that respond to neurotensin by an increase of the intracellular inositol trisphosphate concentration. By contrast, membranes of transitorily transfected COS cells that do not respond to neurotensin exhibit very low levels of GTP-sensitive receptors labeled with the ε1- or ε2-modified analogues. These radiolabeled peptides offer new tools to selectively detect active neurotensin receptors.     

Identification of the glucagon receptor by covalent labeling with a radiolabeled photoreactive glucagon analogue

The photoreactive 125I-labeled glucagon-NAPS [125I-labeled 2-[2-nitro-4-azidophenyl)sulfenyl]-Trp25-glucagon] was used to label the glucagon receptor sites in rat liver plasma membranes. The proteins labeled were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis with or without reduction with dithiothreitol. The photoaffinity peptide specifically labeled a number of bands with apparent molecular weights greater than 200000 and probably at least two protein bands in the molecular weight range 52000-70000. The relative amounts of radioactivity associated with these bands and their relative mobilities differed in samples from reduced and unreduced membranes. Their relative mobilities also differed with percent acrylamide cross-linking, suggesting a glycoprotein nature and the presence of intramolecular disulfide bonds. A nonspecifically labeled band with an apparent molecular weight of 27000-28000 also displayed a similar behavior. Photolabeling in the presence of 0.1 mM guanosine 5'-triphosphate (GTP) decreased the amount of radiolabeling of these bands, suggesting their involvement in the glucagon stimulation of adenylate cyclase. The photolabeled receptor in the membranes, solubilized with Lubrol-PX and fractionated on an Ultrogel AcA22 column, eluted with an apparent molecular weight of 200000-250000. Addition of GTP to the solubilized glucagon receptor of nonirradiated membranes caused complete dissociation of the complex. Gel electrophoresis of the partially purified radiolabeled receptor identified the same protein components observed in photolabeled membranes. These results indicate that the glucagon receptor is an oligomer probably composed of at least two different subunits that are linked together or greatly stabilized by disulfide bonds. They also show that 125I-labeled glucagon-NAPS can be used effectively to covalently label the putative glucagon receptor and thus aid in its further characterization.     

Functional Properties and Molecular Structure of Central and Peripheral Neurotensin Receptors

Abstract

Membranes prepared from mammalian brain or intestine contain two types of specific binding sites for neurotensin that differ by their affinity and by their sensitivity to sodium ions, GTP, and the antihistamine drug levocabastine. Only the high affinity sites are present in cell cultures and in soluble extracts of CHAPS-treated membranes. These sites represent functional neurotensin receptors coupled to GTP-binding proteins that regulate intracellular levels of cAMP, cGMP and inositol phosphates in neuroblastoma N1E115 cells. The molecular weight of neurotensin receptors in cells and membrane preparations of various origin is about 110,000.     

Purification of the neurotensin receptor from mouse brain by affinity chromatography

The neurotensin receptor was purified from newborn mouse brain by affinity chromatography. Active neurotensin binding sites were solubilized from brain homogenates using the nondenaturing detergent 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid (CHAPS) in the presence of cholesteryl hemisuccinate. Chromatography of the soluble extract on SP-Sephadex C-25 and hydroxylapatite eliminated 50% of proteins without loss of neurotensin binding activity. This prepurified material was loaded into an affinity column prepared by coupling neurotensin (2-13) to glutaraldehyde-activated Ultrogel AcA22. Nonspecifically adsorbed proteins were eliminated by extensive washing, and the receptor was eluted with a buffer containing 1 M NaCl, 0.1% CHAPS, and 0.02% cholesteryl hemisuccinate. After desalting, the purified receptor bound 125I-labeled neurotensin with a dissociation constant of 0.26 nM and retained its specificity towards a series of neurotensin analogues. The desalted NaCl eluate appeared on sodium dodecyl sulfate-polyacrylamide gel electrophoresis as a major band of molecular weight 100,000 which was identified as the receptor by affinity labeling with 125I-labeled neurotensin in the presence of disuccinimidyl suberate. The purity of the mouse brain receptor eluted from the affinity column was estimated to be 78%. Electroelution of the 100-kDa protein band gave an homogenous preparation of receptor. Very similar results were obtained with CHAPS-solubilized neurotensin receptors from rat and rabbit brain.     

Modulation of dopamine D1 receptor binding by neurotensin in the rat striatum

The effect of neurotensin on binding characteristics of dopamine D1 receptors was examined in the rat striatal membranes through radioreceptor assay. Neurotensin or its analogs were added to incubation medium of[³H]SCH 23390 saturation or dopamine/[³H]SCH 23390 inhibition experimental systems. Neurotensin did not modulate D1 antagonist binding but converted a part of D1 agonist high affinity binding sites to a low affinity state. Neurotensin_8–13 had the same potency as neurotensin itself, whereas neurotensin_1–8 had only weak activity in modulating D1 agonist binding. GTP and neurotensin had the same effect on D1 agonist binding. However, when both neurotensin and GTP were added, the result was the same as with either alone.

These data suggest that neurotensin modulates the functional state of D1 receptors probably via a GTP binding protein in the rat striatum.     

Photoreactive insulin analogues used to characterise the insulin receptor

Three photoreactive insulin analogues (“photoprobes”) have been prepared in which an aryl azide group was substituted at either the A1, B1 or B29 positions of the insulin molecule. When incubated with rat liver plasma membranes and irradiated all three photoprobes covalently labelled specific insulin binding sites within the membrane. SDS-polyacrylamide gel electrophoresis of plasma membranes covalently tagged with either of the three ¹²⁵I-photoprobes resolved one major specifically labelled polypeptide with an apparent molecular weight of 130,000. The labelled polypeptide migrated anomalously in SDS-polyacrylamide gels and a molecular weight of 90,000 for the polypeptide was determined from a ‘Ferguson’ plot using the combined results from gels of different acrylamide concentrations. Column chromatography of detergent solubilised photoprobe-labelled membranes indicated that the labelled polypeptide may be a subunit of a larger protein complex.     

Use of the heterobifunctional cross-linker m-maleimidobenzoyl N-hydroxysuccinimide ester to affinity label cholecystokinin binding proteins on rat pancreatic plasma membranes

The binding of 125I-cholecystokinin-33 (125I-CCK-33) to its receptors on rat pancreatic membranes was decreased by modification of membrane protein sulfhydryl groups. Sulfhydryl modifying reagents also caused an accelerated release of bound 125I-CCK-33 from its receptor. Because of the presence of an essential sulfhydryl group(s) in CCK receptor binding we studied the application of the heterobifunctional (SH,NH2) cross-linker, m-maleimidobenzoyl N-hydroxysuccinimide ester (MBS), to affinity label 125I-CCK-33 binding proteins on rat pancreatic plasma membranes. Analysis of the cross-linked products by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography revealed that this heterobifunctional cross-linker affinity labeled a major Mr = 80,000-95,000 protein previously identified as part of the CCK receptor on the basis of affinity labeling using homobifunctional and heterobifunctional photoreactive cross-linkers. Additional proteins of Mr greater than 200,000, and Mr = 130,000-140,000 were affinity labeled using MBS. The efficiency of the cross-linking reaction between 125I-CCK-33 and its membrane binding proteins with MBS was significantly greater than that obtained with NH2-directed homobifunctional reagents such as disuccinimidyl suberate. The efficiency of cross-linking could be dramatically improved by reduction of membrane proteins with low-molecular weight thiols prior to binding and cross-linking. The differential labeling patterns of the CCK binding proteins obtained with chemical cross-linkers of similar length but different chemical reactivity underscores the need for caution in predicting native receptor structure from affinity labeling data alone. Using the same pancreatic plasma membrane preparation and 125I-insulin, the Mr = 125,000 alpha-subunit of the insulin receptor was affinity labeled using MBS as cross-linker, demonstrating its utility in identifying other peptide hormone receptors.     

The effect of reducing agents on the structure of mammalian beta-adrenergic receptors

Under reducing conditions (5% beta-mercaptoethanol) the mammalian beta-adrenergic receptor binding site from both beta 1 (porcine heart membranes) and beta 2 receptors (hamster lung and rat erythrocyte membranes) appears to reside on peptides of Mr 62,000-65,000 as determined by photoaffinity labeling with p-azido-m-[125I]iodobenzylcarazolol and sodium dodecyl sulfate-polyacrylamide gel electrophoresis. When similar experiments are performed in these same systems under a variety of non-reducing conditions, there are minimal changes in the apparent molecular weight of both the beta 1- and beta 2-adrenergic receptor binding subunits and no specifically labeled higher molecular weight proteins are observed suggesting that there are no disulfide linked subunits in mammalian beta-adrenergic receptors.     

Differences in the Cyclic AMP-Dependent Phosphorylation of Plasma Membrane Proteins of Differentiated and Undifferentiated L6 Myogenic Cells

Differences in the cyclic AMP-dependent plasma membrane phosphorylation system of undifferentiated and differentiated L₆ myogenic cells have been detected. Endogenous plasma membrane protein phosphorylation in undifferentiated L₆ myoblasts was stimulated more than three fold by 5 × 10⁻⁵ M cyclic AMP, whereas no statistically significant cyclic AMP-dependent phosphorylation of endogenous plasma membrane proteins was observed in differentiated L₆ cells. In undifferentiated cells cyclic AMP promoted the phosphorylation of several proteins, the most prominent of which had a molecular weight of 110,000. In differentiated cells cyclic AMP did not selectively promote the phosphorylation of specific plasma membrane proteins. Both differentiated and undifferentiated L₆ cells, however, contain a cyclic AMP-dependent protein kinase capable of catalyzing the phosphorylation of exogenous substrates, such as histone f₂b. Therefore, the data show that differentiation in L₆ cells is associated with a selective change in the activity of a plasma membrane cyclic AMP-dependent protein kinase which employs endogenous membrane proteins as substrate.     

Purification of a High Molecular Weight Kininogen from Rat Plasma

A high molecular weight kininogen has been isolated from rat plasma and purified. At each preparative step the kininogen concentration and purity were monitored by assay on the perfused isolated rat uterus in terms of bradykinin equivalents formed per mg protein following incubation of the plasma fractions with rodent acid protease for 24 hours at 37 and pH 4.0. Kinin formation by crystalline trypsin and human pancreatic kallikrein also was compared. Citrated rat plasma first was precipitated with 43% ammonium sulfate. The kininogen fractions then were subjected to a series of gel filtration ion exchange chromatographic columns that included G-200 Sephadex, G-200: G-100 Sephadex interconnected columns, DEAE-A50 Sephadex, and hydroxylapatite. The kininogen fractions finally were subjected to preparative polyacrylamide gel electrophoresis, resulting in a final purification of 92.9-fold compared to the initial rat plasma. A single major kininogen protein band and a minor band of protein impurity were obtained on disc gel electrophoresis. Only the pancreatic kallikrein did not form kinin from this purified kininogen. The apparent molecular weight was estimated by SDS polyacrylamide gel technique to be 110,000.     

Periodate induced cross-linking of concanavalin A-reactive membrane proteins of rabbit thymocytes.

Purified plasma membranes of rabbit thymocytes are exposed to sodium periodate and galactose oxidase at conditions similar to those used to induce mitogenic transformation of lymphocytes. The membrane proteins are then fractionated by dodecyl sulfate poly-acrylamide gel electrophoresis. At concentrations of 0.005 M, Na IO4 cross-links 55,000 D and 110,000 D glycoproteins which are known to specifically bind concanavalin A. Galactose oxidase has a similar cross-linking effect, but, at the same time causes proteolytic degradation of membrane proteins. Our data indicate that oxidizing agents, like NaIO4 and galactose oxidase, can indeed cross-link receptors of the thymocyte plasma membrane as has often been proposed as a possible mechanism of their action.     

Cross-linking of endothelin 1 and endothelin 3 to rat brain membranes: identification of the putative receptor(s)

Affinity-labeling experiments with 125I-endothelin derivatives using bifunctional cross-linking reagents were carried out in an attempt to identify the polypeptide component(s) of the endothelin/sarafotoxin receptors in rat brain tissues. In rat cerebellum, cortex, and caudate putamen, endothelin 1 specifically labeled a major component with a molecular mass of around 53,000. In the same tissues endothelin 3 specifically labeled, in addition to the 53,000 band, a band of molecular mass of 38,000. This result clearly indicates that in the brain the endothelin binding site resides within a polypeptide of apparent Mr = 53,000. The possible presence of receptor subtypes is discussed with reference also to the reported identification of endothelin receptors in chick cardiac membrane and in rat mesangial cells.     

ATP-dependent Ca2+ uptake and Ca2+-dependent protein phosphorylation in basolateral liver plasma membranes

ATP-dependent Ca2+ uptake was measured in vesicles of rat liver cell basolateral plasma membranes. Nucleotide-dependent uptake was specific for ATP and observed at pH 7.0 and 7.4/7.5 but not at pH 8.0. ATP-dependent Ca2+ transport was only observed in the presence of Mg2+. Kinetic analysis of ATP-dependent transport revealed an apparent Km in the submicromolar region. Addition of calmodulin and trifluoperazine had no effect on ATP-dependent uptake. A Ca2+-dependent, phosphorylated intermediate with the apparent molecular weight of 135,000 could be demonstrated in the basolateral plasma membranes. Phosphorylated intermediates with apparent molecular weights of 200,000 and 110,000 were demonstrated in microsomes and appeared to contaminate 'basolateral' membrane protein phosphorylation. The results suggest that a 135,000 molecular weight protein is a Ca2+-ATPase and the enzymatic expression of the liver cell basolateral membrane Ca2+ pump.     

Characterization of Infant Rat Cerebral Cortical Membrane Proteins Phosphorylated In Vivo: Identification of the ACTH-Sensitive Phosphoprotein B-50

This study on the phosphorylation in vivo of membrane proteins in cerebral cortices of infant rats reports the identification of the adrenocorticotropin (ACTH)-sensitive phosphoprotein B-50 as one of the substrate proteins that are rapidly phosphorylated in vivo following intracisternal administration of 2 mCi [32P]orthophosphate. Rats were sacrificed 30 min after isotope injection. A fraction enriched in membranes, designated neural membranes (NM), was isolated from the cerebral cortices according to the procedure used for preparation of synaptic plasma membranes (SPM) from adult brain. This NM fraction was characterized by electron microscopy. The proteins of NM were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography. Numerous protein bands of NM in infant rat brain were phosphorylated in vivo. Attention was focussed on the 32P-labeled protein bands in the molecular weight range of 47K-67K. In this region one phosphoprotein band (MW 48K) was more highly labeled than the other bands. The electrophoretic behavior of three of these labeled bands, designated a, c, and e (MW 48K, 55K, and 62K, respectively) was compared with that of protein bands that were phosphorylated in vitro in cerebral membranes isolated from noninjected infant rats. The effects of ACTH1-24 and cyclic AMP in the in vitro system were also studied to probe for the presence of specific membrane proteins known to be sensitive to these modulators. On incubation of NM with [gamma-32P)ATP in the presence and absence of ACTH1-24 in vitro, phosphorylation of a 48K protein band was inhibited in a dose-dependent fashion by the neuropeptide. Two-dimensional electrophoretic separation of NM proteins labeled in vivo indicated that the 48K band had an isoelectric point of 4.5, identical to that of the ACTH-sensitive B-50 protein previously identified. Cyclic AMP stimulated phosphorylation in vitro of two protein bands (MW 55K and 59K) in NM preparations. This result indicates that the in vivo labeled band c may correspond to the cyclic AMP-sensitive 55K protein, whereas phosphoprotein band e, labeled in vivo, appears to be different from the cyclic AMP-sensitive 59K protein band. These observations indicate that neural membranes isolated from infant rat cerebral cortices contain a variety of proteins that can be phosphorylated in vivo. Several of these, for example, the 48K protein band, have the properties of synaptic plasma membrane proteins of adult rat brain that have been characterized by their sensitivity to neuromodulators in endogenous phosphorylating systems in vitro.     

Compared Binding Properties of 125I-Labeled Analogues of Neurotensin and Neuromedin N in Rat and Mouse Brain

Abstract: Neurotensin and neuromedin N are two structurally related peptides that are synthesized by a common precursor. The purpose of the present work was to characterize neuromedin N receptors in rat and mouse brain and to compare these receptors with those of neurotensin. A radiolabeled analogue of neuromedin N has been prepared by acylation of the N-terminal amino group of the peptide with the ¹²⁵I-labeled Bolton-Hunter reagent. This ¹²⁵I-labeled derivative of neuromedin N bound to newborn mouse brain homogenate with high affinity (K _d = 0.5 n M ). Cross-competition experiments between radiolabeled and unlabeled neurotensin and neuromedin N indicated that each peptide was able to displace completely and specifically the other peptide from its interaction with its receptor. Independently of the radioligand used, the affinity of neurotensin was always better than that of neuromedin N. Quantitative radioautographic studies demonstrated that the ratio of labeling intensities obtained with ¹²⁵I-labeled analogues of neurotensin and neuromedin N remained constant in all the brain areas. Our results do not support the existence of a specific neuromedin N receptor in rat and mouse brain and can be explained by the presence of a common receptor for both peptides.     

Identification and characterization of a heterogeneous population of growth hormone receptors in mouse hepatic membranes

Covalent cross-linking of radiolabeled mouse growth hormone (125I-mGH) with the homobifunctional cross-linking agent bis(sulfosuccinimidyl)suberate (BS3) to microsomal membranes prepared from late pregnant mouse liver resulted in the labeling of three specific mGH binding proteins (receptors) with apparent Mr = 125,000, 62,000, and 56,000, as measured by sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions. These same three specifically labeled proteins were present, but with slightly lower apparent molecular weights, when samples were electrophoresed in the absence of reductant. Cross-linking of 125I-mGH to plasma membrane-enriched fractions of late pregnant mouse liver resulted only in the specific labeling of the two lower molecular weight receptors. Removal of all N-linked carbohydrate with peptide: N-glycosidase F resulted in decreasing the apparent molecular weights of the three receptor forms to 110,000, 50,000, and 46,000 for the 125,000, 62,000, and 56,000 molecular weight forms of the receptor, respectively. Smaller decreases in the molecular weights of all three receptor forms were also apparent after treatment with neuraminidase. However, the differences seen in the intact forms of the growth hormone receptor were also present in the deglycosylated receptors. The relationship between the three forms of the growth hormone receptor was further investigated by comparing the fragments produced by proteolytic digestion of the cross-linked receptors with Staphylococcus aureus protease and endoproteinase Lys-C. The fragments produced from all three receptor forms had very similar molecular weights, although there were slight molecular weight differences in the fragments produced by endoproteinase Lys-C digestion. The overall similarity of the fragments produced by the proteolytic digestions suggests that the three forms of the receptor are related.     

Characterization and regulation by protein kinase C of renal glomerular atrial natriuretic peptide receptor-coupled guanylate cyclase

The nature and regulation of atrial natriuretic peptide (ANP)-sensitive guanylate cyclase in rat renal glomerular membranes was examined. By affinity crosslinking techniques, three bands with apparent molecular masses of 180, 130 and 64 kDa were specifically labeled with [125I]ANP. A specific antibody to the 180 kDa membrane guanylate cyclase of rat adrenocortical carcinoma recognized a 180 kDa band on Western blot analysis of solubilized, GTP-affinity purified glomerular membrane proteins. The same antibody completely inhibited ANP-stimulated guanylate cyclase activity in glomerular membrane fractions. Partially purified protein kinase C inhibited ANP-stimulated guanylate cyclase activity in glomerular membrane fractions. It is concluded that a 180 kDa ANP-sensitive guanylate cyclase is present in glomerular membranes, and that this enzyme is inhibited directly by protein kinase C.     

Immunocytochemical localization of a urate oxidase immunoreactive protein in the plasma membranes and membranes of the secretory/endocytic compartments of digestive gland cells of the mussel Mytilus galloprovincialis

The subcellular compartmentalization of urate oxidase (UOX) in the digestive glands of mussels, Mytilus galloprovincialis Lmk, was studied by means of immunoblotting and immunocytochemistry, using an antibody raised in rabbit against rat liver UOX. Western blot analysis of subcellular fractions revealed an immunoreactive polypeptide with a molecular weight similar to the corresponding mammalian hepatic protein. This crossreactive polypeptide of 32 kDa was particle-bound yet not peroxisome-associated. In paraffin sections the antiserum specifically labeled the plasma membrane of the digestive gland epithelial cells and discrete regions within the perinuclear and apical portions of the digestive tubules and duct cells. By electron microscopy gold particles representing antigenic sites were found on the microvilli and the lateral plasma membrane as well as the membranes of the secretory/ endocytic compartments, that is, the Golgi complex, secretory and some endocytic vesicle membranes. Since the peroxisomal UOX-antibody exhibits a comparable immunoreactivity towards a urate-transporter channel protein in rat kidney proximal tubules and has been used for its molecular cloning (Leal-Pinto et al., 1997, J. Biol. Chem. 272, 617-625), we suggest that the membrane protein identified in mussel digestive glands could represent a homologous urate-transporter protein.     

Isolation and characterization of the putative canalicular bile salt transport system of rat liver

Through labeling with the sodium salt of the photolabile bile salt derivative (7,7-azo-3 alpha,12 alpha-dihydroxy-5 beta-[3 beta-3H]cholan-24-oyl)- 2-aminoethanesulfonic acid, a bile salt-binding polypeptide with an apparent molecular weight of 100,000 was identified in isolated canalicular but not basolateral (sinusoidal) rat liver plasma membranes. This labeled polypeptide was isolated from octyl glucoside-solubilized canalicular membranes by DEAE-cellulose and subsequent wheat germ lectin Sepharose chromatography. The purified protein still contained covalently incorporated radioactive bile salt derivative and exhibited a single band with an apparent molecular weight of 100,000 on sodium dodecyl sulfate-gels. Antibodies were raised in rabbits and their monospecificity toward this canalicular polypeptide demonstrated by immunoblot analysis. No cross-reactivity was found with basolateral membrane proteins. The antibodies inhibited taurocholate uptake into isolated canalicular but not basolateral membrane vesicles. In addition, the antibodies also decreased efflux of taurocholate from canalicular vesicles. If the canalicular bile salt-binding polypeptide was immunoprecipitated from Triton X-100-solubilized canalicular membranes and subsequently deglycosylated with trifluoromethanesulfonic acid, the apparent molecular weight was decreased from 100,000 to 48,000 (sodium dodecyl sulfate-polyacrylamide gel electrophoresis). These studies confirm previous results in intact liver tissue and strongly indicate that a canalicular specific glycoprotein with an apparent molecular weight of 100,000 is directly involved in canalicular excretion of bile salts.