Structural comparisons between the soluble and the GPI-anchored forms of the Paramecium temperature-specific 156G surface antigen.

Biosynthetic labelling experiments performed on P primaurelia strain 156, expressing the temperature-specific G surface antigen, 156G SAg, demonstrated that the purified 156G SAg contained the components characteristic of a GPI-anchor. [3H]ethanolamine, [3H]myo-inositol, [32P]phosphoric acid and [3H]myristic acid could all be incorporated into the surface antigen. Myristic acid labelling was lost after treatment in vitro with Bacillus thuringiensis phosphatidylinositol-specific phospholipase C (PI-PLC). After complete digestion by pronase, a fragment containing the intact GPI-anchor of 156G surface antigen was isolated. This fragment was shown to be hydrophobic and glycosylated and to possess an epitope found specifically in the GPI component of GPI-anchored proteins. The role of the GPI-tail in anchoring the 156G surface antigen into the membrane was assessed by determining that purified 156G molecules with the GPI-anchor could be incorporated into lipid vesicles and red cell ghosts whereas the 156G molecules lacking the GPI-anchor, as result of treatment with B thuringiensis PI-PLC, could not. It has also been shown that the membrane-bound form and the soluble form, obtained after cleavage of the 156G SAg lipid moiety either by an endogenous PI-PLC or by a bacterial PI-PLC, displayed identical circular dichroic spectra.     

Problems associated with the labelling of RNA with radioactive precursors in vivo (author's transl)]

The radioactivity of RNA, DNA and proteins in the liver, muscles and cerebrum of 30-day-old rats after labelling with [3H]uridine, [14C]uridine, [3H]cytidine or [3H]orotic acid was measured. It was found that after administration of [3H]uridine, the proteins were 5 - 10 times more radioactive than the RNA. After administration of [14C]uridine, the proteins were 1 - 2 times more heavily labelled than the RNA. Hydrolysis of the proteins followed by chromatography of the amino acids revealed that the protein labelling was mostly due to [3H]glutamate. In the liver, [3H]orotic acid produced very specific labelling of the RNA. The radioactivity of the proteins is very slight. However, the specific labelling of the RNA in the muscles and cerebrum is not so pronounced with this precursor. [3H]Cytidine is an ideal precursor for RNA. The labelling of protein in all three organs examined is very slight, and furthermore, the specific activity of the RNA is 10 - 20 times higher than after labelling with uridine. We were also able to show that after labelling with radioactive uridine, the method of isolation of RNA by alkaline hydrolysis gives incorrect results, because [3H]amino acids interfere with the measurement of the specific activity of the RNA. The heavy labelling of proteins by [3H]-uridine must also be taken into account in histoautoradiography, because our experiments showed that in liver, the proteins in the cell nucleus are 3 times as radioactive as the nucleic acids. The particulate components of the cytoplasm are even 20 times more radioactive than the nucleic acids.     

Exposure of acetylcholine receptor to the lipid bilayer

All four subunits of the acetylcholine receptor in membrane fragments isolated from T. californica have been labeled with a photolabile hydrophobic probe, [3H]adamantanediazirine, which selectively labels regions of integral membrane proteins in contact with the hydrocarbon core of the lipid bilayer. As all of the homologous subunits are exposed to the lipid bilayer, it is probable that they each interact with the surrounding membrane in a similar fashion.     

Glycosylinositol-phosphoceramide in the free-living protozoan Paramecium primaurelia: modification of core glycans by mannosyl phosphate.

Glycolipids synthesized in a cell-free system prepared from the free-living protozoan Paramecium primaurelia and labelled with [3H]mannose and [3H]glucosamine using GDP-[3H]mannose and UDP-[3H]N-acetyl glucosamine, respectively, were identified and structurally characterized as glycosylinositol-phosphoceramides (GIP-ceramides). The ceramide-based lipid was also found in the GIP membrane anchor of the G surface antigen of P.primaurelia, strain 156. Using a combination of in vitro labelling with GDP-[3H]mannose and in vivo labelling with 33P, we found that the core glycans of the P.primaurelia GIP-ceramides were substituted with an acid-labile modification identified as mannosyl phosphate. The modification of the glycosylinositol-phospholipid core glycan by mannosyl phosphate has not been described to date in other organisms. The biosynthesis of GIP-ceramide intermediates in P.primaurelia was studied by a pulse-chase analysis. Their structural characterization is reported. We propose the following structure for the putative GIP-ceramide membrane anchor precursor of P.primaurelia surface proteins: ethanolamine phosphate-6Man-alpha 1-2Man-alpha 1-6Man-(mannosyl phosphate)-alpha 1-4glucosamine-inositol-phosphoceramide.     

Mechanism of local anesthetic effect on mitochondrial ATP synthase as deduced from photolabelling and inhibition studies with phenothiazine derivatives

The mode of interaction between mitochondrial ATP synthase and two phenothiazine derivatives, chlorpromazine (CPZ) and trifluoperazine (TFP), was studied as a model for the interaction of local anesthetic drugs with membrane proteins. Photolabelling experiments demonstrated that CPZ and TFP interact with various subunits of either the peripheral F1 moiety of the membrane-embedded F0 sector. Both drugs, however, labelled the membrane sector much more heavily. Qualitative differences in labelling were observed between CPZ and TFP, indicating non-identical sites of interaction. These diversities appeared related to the different hydrophobicities of the two drugs since: (a) TFP, which has a higher lipid/water partition coefficient, labelled the more hydrophobic subunits more markedly than CPZ; (b) reduced glutathione, a hydrophilic free radical scavenger that does not penetrate the membrane continuum, had a negligible effect on the labelling by TFP, whereas it reduced the labelling of various subunits by CPZ; (c) the labelling by [3H]TFP was poorly antagonized by cold CPZ, whereas it was almost totally prevented by fluphenazine, a phenothiazine similar to TFP in hydrophobic character. Consistently, double-inhibition experiments showed that TFP and fluphenazine are mutually exclusive inhibitors of mitochondrial ATP synthase, whereas TFP and CPZ are mutually nonexclusive. The nature of the phospholipid bilayer influenced neither the labelling nor the inhibition patterns. The complex of these data indicate that tertiary amine local anesthetics affect the activity of membrane proteins by interacting with a multiplicity of relatively aspecific hydrophobic sites located preferentially, but not exclusively, on the membrane-embedded domains. It is suggested that at least two phenothiazine derivatives of different hydrophobicities be used in photolabelling experiments, before any generalization is made, since the molecular targets of these drugs vary according to their hydrophobic character.     

Labelling of the intramembranous region of the major sialoglycoprotein of human erythrocytes with a photosensitive hydrophobic probe.

Human erythrocyte membranes were incubated with the photosensitive hydrophobic reagent 1-azido-r-iodo[3H]benzene and the mixture was irradiated. The major sialoglycoprotein was then isolated and the labelled polypeptide subjected to proteolytic dissection. Characterization of the purified tryptic and chymotryptic peptides show that the probe is covalently attached only to the transmembrane region of the protein. This labelling pattern is discussed in relation to the use of such reagents for the identification of segments of membrane proteins exposed to the hydrophobic millieu of the membrane.     

Photochemical labeling of membrane hydrophobic core of human erythrocytes using a new photoactivable reagent 2-[3H]diazofluorene

Photoactivable reagents have been useful for studying the structural aspects of membrane hydrophobic core. We have reported earlier (Anjaneyulu, P.S.R., and Lala, A. K. (1982) FEBS Lett. 146, 165-167) the use of diazofluorene as a probe for fluorescent photochemical labeling of hydrophobic core in artificial membranes. To quantitate and enhance the monitoring ability of this probe, we have synthesized 2-[3H]diazofluorene of high specific activity. This reagent rapidly partitions into phosphatidylcholine vesicles and selectively labels the fatty acyl chains of phosphatidylcholine. The insertion yield (13%) is not affected by the presence of scavengers like reduced glutathione. 2-[3H]Diazofluorene also readily partitions into erythrocyte membranes and on photolysis labels the membrane. The overall insertion was 48% with 9.7% in protein fraction and the rest in lipids. The distribution of radioactivity in labeled protein fraction was restricted to integral membrane proteins with Band 3 being the major protein labeled. There is little or no labeling associated with extrinsic proteins like spectrin. Further analysis of labeled Band 3 by treatment with chymotrypsin indicated that the labeling was restricted to the membrane spanning CH-17 and CH-35 fragments. No labeling of the cytoplasmic fragment of Band 3 could be observed. 2-[3H]Diazofluorene should prove useful for studying integral membrane proteins and their membrane-spanning regions.     

Characterization of [3H]palmitate- and [3H]ethanolamine-labelled proteins in the multicellular parasitic trematode Schistosoma mansoni.

Biosynthetic labelling experiments with cercariae and schistosomula of the multicellular parasitic trematode Schistosoma mansoni were performed to determine whether [3H]palmitate or [3H]ethanolamine was incorporated into proteins. Parasites incorporated [3H]palmitate into numerous proteins, as judged by SDS/polyacrylamide-gel electrophoresis and fluorography. The radiolabel was resistant to extraction with chloroform, but sensitive to alkaline hydrolysis, indicating the presence of an ester bond. Further investigation of the major 22 kDa [3H]palmitate-labelled species showed that the label could be recovered in a Pronase fragment which bound detergent and had an apparent molecular mass of 1200 Da as determined by gel filtration on Sephadex LH-20. Schistosomula incubated with [3H]ethanolamine for up to 24 h incorporated this precursor into several proteins; labelled Pronase fragments recovered from the three most intensely labelled proteins were hydrophilic and had a molecular mass of approx. 200 Da. Furthermore, reductive methylation of such fragments showed that the [3H]ethanolamine bears a free amino group, indicating the lack of an amide linkage. We also evaluated the effect of phosphatidylinositol-specific phospholipase C from Staphylococcus aureus: [3H]palmitate-labelled proteins of schistosomula and surface-iodinated proteins were resistant to hydrolysis with this enzyme. In conclusion, [3H]palmitate and [3H]ethanolamine are incorporated into distinct proteins of cercariae and schistosomula which do not bear glycophospholipid anchors. The [3H]ethanolamine-labelled proteins represent a novel variety of protein modification.     

Changes in plasma membrane glycoproteins of rat spermatozoa during maturation in the epididymis

Glycoproteins on the plasma membrane of testicular and cauda epididymidal spermatozoa have been labeled with galactose oxidase/NaB [3H]4 and sodium metaperiodate/NaB[3H]4, followed by analysis on SDS polyacrylamide gels. The major glycoprotein labeling on testicular spermatozoa has a molecular weight 110,000 whereas on cauda epididymidal spermatozoa greater than 90% of the radio-label is incorporated into proteins of molecular weight 32,000. These 32,000-mol wt X proteins are homologous with proteins of similar molecular weight purified from the epididymal secretion and which have been shown previously to be synthesized in the caput epididymidis under hormonal control. Immunofluorescence revealed that the 32,000-mol wt proteins are present on the flagellum of mature but not immature spermatozoa and that they have a patchy distribution suggesting that they are mobile within the plane of the membrane. The membrane-bound 32,000-mol wt proteins possess hydrophobic domains as revealed by charge-shift electrophoresis and they also label with a lipophilic photoaffinity probe suggesting that they are in contact with the lipid bilayer. The evidence indicates that there is a considerable reorganization of the molecular structure of the plasma membrane of spermatozoa during maturation in the epididymis and that some of the changes are brought about by a direct interaction with epididymal secretory proteins.     

Effects of monensin on posttranslational processing of myelin proteins

Rat brain slices were incubated with [3H]palmitic acid and [14C]glycine to label the lipid and protein moieties, respectively, of myelin proteolipid protein (PLP). The effects of monensin on posttranslational processing of proteins were examined by measuring the appearance of [14C]glycine- and [3H]palmitate-labeled proteins in myelin and myelin-like fractions. At 0.01 and 0.10 microM, monensin did not appreciably affect total lipid or protein synthesis; higher concentrations caused increased inhibition. Monensin at 0.10 microM markedly decreased the appearance of [14C]glycine-labeled PLP in myelin, but had little effect on the 14C basic proteins or the incorporation of [3H]palmitic acid into total or myelin PLP. The same relative effect was apparent at higher monensin concentrations. In the myelin-like fraction, monensin at 0.10 microM also depressed entry of [14C]glycine into protein comigrating with PLP, and again had no effect on incorporation of [3H]palmitic acid. In addition, monensin increased the [3H]palmitate label associated with two high-molecular-weight proteins in the myelin-like fraction with no concomitant increase in [14C]glycine label.     

Further studies on the size and composition of the chick embryo fibroblast cytosolic DNA complex

The chick embryo fibroblast cytosolic DNA complex shows anomalous elution behaviour on agarose gel column chromatography. The indicated molecular size varies between 5 X 10(5) dalton (higher exclusion limit gels) and 1.4 X 10(6) dalton (lower exclusion limit gels). Chromatography on lower exclusion limit gels shows the [3H]thymidine labelled (DNA) complex as a sharp peak, coincident with a peak of [3H]uridine and [3H]lysine labelling and similar pulse labelling patterns for the three precursors but with DNA labelling lagging behind RNA and protein. Both cultured and uncultured cell cytosols show an A260 peak coincident with the [3H]precursor labelling peaks.     

Phosphatidylethanolamine is the donor of the terminal phosphoethanolamine group in trypanosome glycosylphosphatidylinositols.

A variety of eukaryotic cell surface proteins, including the variant surface glycoproteins of African trypanosomes, rely on a covalently attached lipid, glycosylphosphatidylinositol (GPI), for membrane attachment. GPI anchors are synthesized in the endoplasmic reticulum by stepwise glycosylation of phosphatidylinositol (via UDP-GlcNAc and dolichol-P-mannose) followed by the addition of phosphoethanolamine. The experiments described in this paper are aimed at identifying the biosynthetic origin of the terminal phosphoethanolamine group. We show that trypanosome GPIs can be labelled via CDP-[3H]ethanolamine or [beta-32P]CDP-ethanolamine in a cell-free system, indicating that phosphoethanolamine is acquired en bloc. In pulse-chase experiments with CDP-[3H]ethanolamine we show that the GPI phosphoethanolamine is not derived directly from CDP-ethanolamine, but instead from a relatively stable metabolite, such as phosphatidylethanolamine (PE), generated from CDP-ethanolamine in the cell-free system. To test the possibility that PE is the immediate donor of the GPI phosphoethanolamine moiety, we describe metabolic labelling experiments with [3H]serine and show that GPIs can be labelled in the absence of detectable radiolabelled CDP-ethanolamine, presumably via [3H]PE generated from [3H]phosphatidylserine (PS). The data support the proposal that the terminal phosphoethanolamine group in trypanosome GPIs is derived from PE.     

Hydrophobic photolabeling in membranes: the human erythrocyte glucose transporter

Human erythrocyte membranes were labeled with a hydrophobic photoactivable reagent, 2-[3H]Diazofluorene. Electrophoretic analysis of the protein fraction showed that several membrane spanning proteins like Band 3 (the anion transporter), Band 4.5 (the glucose transporter), and the sialoglycoproteins PAS 1, 2, and 3 have been labeled. To isolate the diazofluorene-labeled glucose transporter, the membrane preparation was solubilized with Triton X-100 and passed through a DEAE-cellulose column. The flow-through fraction was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Radioactive analysis of the gel indicated that besides the Band 4.5, two more proteins corresponding to the Band 3 and Band 6 regions also coelute with the glucose transporter in the flow-through fraction. On the other hand, use of n-octyl glucoside gave a relatively better preparation. The 2-[3H]DAF-labeled glucose transporter isolated by the latter method on tryptic digestion indicated that the Mr 18,000 fragment corresponding to the C-terminal transmembrane fragment is labeled.     

Identification of hydrophobic regions of the calcium-transport ATPase from sarcoplasmic reticulum after photochemical labeling with adamantane diazirine

The Ca-ATPase from skeletal muscle sarcoplasmic reticulum was labeled with [3H]adamantane diazirine. Adamantane diazirine is a hydrophobic photoactivated probe that partitions into the cell membrane and can be used to identify regions of proteins that are embedded within the membrane. Digestion of the labeled protein with trypsin and separation of the labeled tryptic fragments by SDS-polyacrylamide-gel electrophoresis indicated that all of the major tryptic fragments were labeled by the probe. The presence of glutathione in the sample buffer during photolysis did not alter the pattern of labeling, indicating that adamantane diazirine labeled the Ca-ATPase from within the lipid bilayer. These results indicate that the Ca-ATPase polypeptide must cross the membrane at least 3 times.     

Electrostatic and hydrophobic interactions of the intermediate filament protein vimentin and its amino terminus with lipid bilayers

Immunofluorescence and electron microscopical studies on the intracellular distribution of intermediate filaments (IFs) have demonstrated a close proximity of these cytoskeletal structures to cellular membranes. Moreover, nonepithelial IF (protein)s have been shown to exhibit high affinities for lipids, especially for negatively charged and nonpolar lipids. Here, using hydrophobic labeling with the photoactivatable phosphatidylcholine analogue [3H]1-palmitoyl-2-[11-[4-(trifluoromethyldiazirinyl]undecanoyl+ ++]-sn- glycero-3-phosphorylcholine or with 1-azidopyrene at low and physiological ionic strength, it is demonstrated that the IF subunit protein vimentin can interact with the hydrophobic core of lipid bilayers, in addition to strong ionic relationships between both reactants. Whereas the presence of acidic phospholipids in the lipid vesicles was absolutely essential for efficient vimentin labeling, cholesterol played a synergistic role in this reaction. Proteolytic degradation of photolabeled vimentin localized the derivatization exclusively to the non-alpha-helical, highly positively charged N-terminal domain of the filament protein. Furthermore, circular dichroism studies performed on the isolated N terminus of vimentin revealed a significant increase in the alpha-helical content of the polypeptide upon its interaction with vesicles containing negatively charged phospholipids. These results indicate an amphiphilic character of the N terminus and suggest that the cationic arginine residues of the N-terminal domain react with the negatively charged head groups of acidic phospholipids prior or parallel to interaction of the polypeptide with hydrophobic regions of the lipid bilayer.     

Major membrane surface proteins of Mycoplasma hyopneumoniae selectively modified by covalently bound lipid.

Surface protein antigens of Mycoplasma hyopneumoniae were identified by direct antibody-surface binding or by radioimmunoprecipitation of surface 125I-labeled proteins with a series of monoclonal antibodies (MAbs). Surface proteins p70, p65, p50, and p44 were shown to be integral membrane components by selective partitioning into the hydrophobic phase during Triton X-114 (TX-114)-phase fractionation, whereas p41 was concomitantly identified as a surface protein exclusively partitioning into the aqueous phase. Radioimmunoprecipitation of TX-114-phase proteins from cells labeled with [35S]methionine, 14C-amino acids, or [3H] palmitic acid showed that proteins p65, p50, and p44 were abundant and (with one other hydrophobic protein, p60) were selectively labeled with lipid. Covalent lipid attachment was established by high-performance liquid chromatography identification of [3H]methyl palmitate after acid methanolysis of delipidated proteins. An additional, unidentified methanolysis product suggested conversion of palmitate to another form of lipid also attached to these proteins. Alkaline hydroxylamine treatment of labeled proteins indicated linkage of lipids by amide or stable O-linked ester bonds. Proteins p65, p50, and p44 were highly immunogenic in the natural host as measured by immunoblots of TX-114-phase proteins with antisera from swine inoculated with whole organisms. These proteins were antigenically and structurally unrelated, since hyperimmune mouse antibodies to individual gel-purified proteins were monospecific and gave distinct proteolytic epitope maps. Intraspecies size variants of one surface antigen of M. hyopneumoniae were revealed by a MAb to p70 (defined in strain J, ATCC 25934), which recognized a larger p73 component on strain VPP11 (ATCC 25617). In addition, MAb to internal, aqueous-phase protein p82 of strain J failed to bind an analogous antigen in strain VPP11. These studies establish that a highly restricted set of distinct, lipid-modified hydrophobic membrane proteins are major surface antigens of M. hyopneumoniae and that structural variants of surface antigens occur within this species.     

Cholinergic Binding to the Receptor Proteolipid from Torpedo Electroplax Separated by Ion Exchange Chromatography

Abstract

Proteolipids (i.e., hydrophobic proteins) have been extracted with chloroform-methanol (2:1) from lyophilized Torpedo electroplax, and fractionated on a DEAE-cellulose column. The elution system consisted of the same solvent and a gradient of the hydrophobic ion ptoluene sulfonate (0.1–100mM). The three fractions obtained (I, II and III) have different content of protein, lipid P, and reducing sugars. The amino acid composition shows a higher proportion of hydrophobic residues in I and more charged ones in fractions II and III. Polyacryl amide gel electrophoresis of fraction I shows a single major band of 39 kdaltons; in fractions II and III a major band of 42 kdaltons and fainter bands in the range 62–68 kdaltons are observed.

Fraction I has the highest specific binding for [³H]-acetylcholine (7.1 nmol/mg protein) and [³H]α-bungarotoxin (5.2 nmol/mg protein). The nicotinic nature of this proteolipid was demonstrated by blocking experiments. The Scatchard plot showed two affinity sites for [³H]-acetylcholine (Kd₁ = 3nM and Kd₂ = 1.1 μm). 4-(N-maleimido) pheny1 [³H]trimethylammonium specifically labeled the 39 kdaltons band.

The possible factors involved in the fractionation of proteolipids are discussed. The findings suggest that the 39 kdaltons polypeptide contains the receptor site for acetylcholine and that the other proteolipid components may play a different function in the membrane.     

An evaluation of techniques for labelling the surface proteins of cultured mammalian cells.

We have evaluated four techniques for labelling the surface proteins of cultured mammalian cells. The techniques are: (a) the lactoperoxidase system; (b) the pyridoxal phosphate-[3H]borohydride system; (c) the [3H]4,4'-diisothiocyano-2,2'-dihydrostilbene disulfonate sysem and (d) the galactose oxidase-[3H]borohydride system. The subcellular distribution of radiolabel produced by these technics has been evaluated by autoradiography at the light microscope level and by cellular fractionation. We find that while all four systems label the surface membranes in the majority of the cell population, they also heavily label internal sites in a small subpopulation of nonviable cells. The contribution of the internally labelled cells to further biochemical analysis may represent a severe problem in investigations which rely solely on surface labels for the study of plasma membrane organization.     

Modulation by phorbol 12-myristate 13-acetate of arachidonic acid release from rat basophilic leukemia cells stimulated with A23187

Rat basophilic leukemia (RBL-2H3) cells were cultured in medium containing [3H]arachidonic acid and labelling of the different lipid fractions was followed with time. After up to 4 h of culture, the label was found mostly in phosphatidylcholine. After 8 h, labelling of phosphatidylethanolamine gradually exceeded that of phosphatidylcholine, until at 24 h, approximate equilibrium labelling of the lipid fractions was attained and 45% of the label was found in phosphatidylethanolamine, 35% in phosphatidylcholine, 18% in the phosphatidylserine/inositide fraction and the remainder in the neutral lipid fraction. Stimulation of cells with A23187 after 30 min of labelling caused release of [3H]arachidonic acid which was accountable by a decrease in radioactivity of phosphatidylcholine, whereas stimulation of cells after 24 h of labelling caused the release of radioactive arachidonic acid, which was accompanied by a decrease of label in both phosphatidylcholine and phosphatidylethanolamine. Incubation of the labelled cells with phorbol 12-myristate 13-acetate prior to ionophore addition enhanced both the release of [3H]arachidonic acid and its metabolites and the decrease in label of the same phospholipids as those affected by ionophore alone. Under our conditions, the enhancement effects of phorbol ester were greatest after 2-5 min of preincubation, prior to ionophore addition. The results suggest that in basophilic leukemia cells, arachidonic acid release proceeds from several pools of phospholipids and that the activity of the phospholipase(s) involved is modulated by protein kinase C.     

Evidence for the organization of the transmembrane segments of (Na,K)-ATPase based on labeling lipid-embedded and surface domains of the alpha-subunit

The purpose of this work has been to examine the organization of the intramembranous portion of the alpha-subunit of membrane-bound (Na,K)-ATPase. Covalent labeling of the alpha-subunit and its tryptic fragments from within the lipid bilayer with [125I]iodonaphthylazide was combined with covalent labeling with 32P from [gamma-33P]ATP at the cytoplasmic surface and with [3H]N-(ouabain)-N'-(2-nitro-4-azidophenyl)ethylenediamine from the extra cellular surface. In control experiments using extensive proteolysis and reduced glutathione, it is confirmed that iodonaphthylazide labels segments of the protein within the lipid bilayer. The labeled segments of the alpha-subunit, produced by extensive proteolysis, are selectively extracted by organic solvents. Both at a low and at a high concentration of iodonaphthylazide, about 50% of label added to the medium is covalently attached to protein and lipid. At the low iodonaphthylazide concentration, the NH2-terminal Mr = 46,000 (46K) fragment of the alpha-subunit is preferentially labeled, while at the higher concentration of the 46K fragment, the 78K fragment, and the COOH-terminal 58K fragment are labeled. 32P from [gamma-32P]ATP is incorporated into the 46K fragment while [3H]N-(ouabain)-N'-(2-nitro-4-azido-phenyl)ethylenediamine from the extracellular surface labels all the major fragments, 78K, 58K, and 46K. The data provide evidence for a model of the path of the polypeptide chain with multiple traverses of the alpha-subunit across the bilayer and the NH2-terminal and three trypsin-sensitive bonds exposed at the cytoplasma surface.