Separation and some properties of the major proteins of the human erythrocyte membrane

Membrane water channels and SH-groups

The transport of water across human red cell membranes is conmonly interpreted in terms of small aqueous channels. This interpretation is based largely on indirect evidence. In this report, two sets of experiments providing more direct evidence for this idea is presented. (1) The effect of various SH-reactive reagents on the movement of water was studied. Using these compounds we attempted to localize and characterize those membrane SH-groups which are important for water transport. (2) Experimental evidence which suggests that these channels are assembled from aggregates of specific membrane protein(s) is presented.

The structure of the major protein of the human erythrocyte membrane. Characterization of the intact protein and major fragments

Polypeptide 3, the major membrane-penetrating protein of the human erythrocyte membrane, was characterized, together with two major fragments derived by specific proteolysis of the native protein in the membrane. One fragment (fragment 3f) was obtained from thermolysin cleavage in the extracellular region of the protein, and the other (fragment T1) was derived from tryptic cleavage in the intracellular region of the protein. The results of N- and C-terminal group analysis suggest that fragment 3f contains the N-terminal region of polypeptide 3 and fragment T1 contains the C-terminal part of the molecule. The carbohydrate contents of the polypeptides suggest that carbohydrates are present in three regions of the molecule, much of this carbohydrate being present in the C-terminal part of the molecule. This region of the protein also contains the receptors for concanavalin and the lectins from Phaseolus vulgaris and Ricinis communis, and our results suggest that there is heterogeneity in the carbohydrate chains present in the C-terminal region of polypeptide 3. These data are related to the folding of polypeptide 3 in the erythrocyte membrane.

Labelling of membrane glycoprotein in erythrocytes infected with Plasmodium knowlesi

Normal rhesus monkey erythrocytes and erythrocytes infected by P. knowlesi were labelled with galactose oxidase (EC 1.1.3.9) and tritiated sodium borohydride. The glycoproteins of normal erythrocytes were not labelled unless the cells were pretreated with neuraminidase, when peaks of activity with apparent molecular weights of 170 000, 126 000, 90 000, 50 000, and 35 000 were observed. Schizont-infected erythrocytes showed an absence of glycoprotein labelling even after neuraminidase treatment. The results indicate that there is an alteration in the glycoproteins of schizont-infected erythrocytes, which may contribute to the increased permeability and the immunological alterations on the surface of these cells.

Changes in carbohydrate content of surface membranes of human erythrocytes during ageing

The carbohydrate content (neutral hexoses, fucose, sialic acid, and hexosamine) of human erythrocytes is significantly higher in young than in old erythrocytes. The differences observed in the carbohydrate content of surface glycopeptides, obtained after incubation with either Pronase or trypsin, is still more striking, suggesting important modifications of the carbohydrate moeties of the erythrocyte membrane during in vivo ageing.

Studies on the membrane glycoprotein defect of En(a-) erythrocytes. II. MN antigenic properties of En(a-) erythrocytes

En(a-) and EnaEn red cells from different sources were studied using biochemical and serological methods. The results suggest that Finnish En(a-) erythrocytes contain only "N" but no M antigenic properties. Data on the members of the English En(a-) family suggest that English En(a-) red cells exhibit a combination of three rare alterations. The English En(a-) individuals are apparently heterozygous for the defects En and Ms, to all appearances, converted to a M ("M") antigen. These extraordinary data are brought to light by investigations on the children of the propositus (G.P. and J.P.), whose red cells are M-N+S-s+"N"+"M"- and M-N+S+s+"N"+"M"+ respectively. Sodium dodecylsulphate polyacrylamide gel electrophoretic results indicate that the MN glycoprotein content is decreased by about 50% in all English S+s+EnaEn red cells, the Ss glycoprotein being normal. G.P. erythrocytes, however, have only about half of the normal MN and Ss glycoprotein content.

The major "intrinsic" membrane protein of human erythrocytes. Preparative isolation and immunoelectrophoretic analyses

(1) Preparative dodecyl sulfate gel electrophoresis of human erythrocyte membrane proteins has been used to isolate dodecylsulfate band 3 containing the M,N-glycoprotein and the major "intrinsic" membrane protein (Fairbanks, G., Steck, T.L. and Wallach, D.F.H. (1971) Biochemistry 10, 2606-2617; Bretscher, M.S. (1971) J. Mol. Biol. 59,351-357; Bretscher, M.S. (1971) Nat. New Biol. 231, 229-232 and Marchesi, V.T. and Andrews, E.P. (1972) Science 174, 1247-1248). Subsequent isoelectric focusing in polyacrylamide gels containing Triton X-100 separates these two entities and allows their simultaneous purification. (2) The proteins thus obtained retain their antigenic properties. They are pure according to electrophoretic and immunoelectrophoretic criteria. However, crossed immunoelectrophoresis yields evidence for molecular microheterogeneity of the major "intrinsic" protein. (3) Analyses utilizing crossed immunoelectrophoresis with antibodies absorbed with intact erythrocytes show that the major "intrinsic" protein possesses antigenic determinants on both membrane surfaces and therefore spans the erythrocyte membrane. All determinants of the M,N-glycoprotein detectable with our antibodies were found solely on the exterior membrane surface. (4) Neither the major "intrinsic" membrane protein nor the major M,N-glycoprotein bound significantly to concanavalin A in crossed immunoaffinoelectrophoresis.

A study of the relationship between inhibition of anion exchange and binding to the red blood cell membrane of 4,4'-diisothiocyano stilbene-2,2'-disulfonic acid (DIDS) and its dihydro derivative (H2DIDS)

DIDS (4,4'-diisothiocyano stilbene-2,2'-disulfonic acid) and H2DIDS (4,4'-diisothiocyano-1,2-diphenyl ethane-2,2'-disulfonic acid) binding to the human red cell membrane proteins were studied as a function of concentration, temperature and time. Most binding sites were common to both. The common sites were in band 3 of SDS polyacrylamide gel electropherograms (Steck, 1974. J. Cell Biol. 62:1), an unidentified adjacent band, and glycophorin. Reversible and irreversible binding occurred; both inhibited sulfate equilibrium exchange. The time courses of irreversible binding to band 3 and total binding to the membrane as a whole were biphasic. About 20% of H2DIDS and greater 60% of DIDS binding were rapid, independent of temperature. Slow H2-DIDS binding was monoexponential, activation enthalpy 23 kcal/mole. The stoichiometry of irreversible H2DIDS binding to band 3 was 1.1-1.2, concentration-dependent. Under the conditions studied (0-50 muM, hematocrit 10%, 5-37 degrees C) binding to band 3 was a constant fraction of total binding, 0.7 for H2DIDS and 0.8 for DIDS. Inhibition was a linear function of total binding, binding to band 3, and therefore also to nonband 3 sites, with either inhibitor during both phases, H2DIDS inhibition was complete at 1.9 X 10(6) or 1.2 X 10(6) molecules/cell total and band 3 binding respectively. For DIDS the corresponding figures were 1.3 X 10(6) and 1.1 X 10(6). It is shown how reagents of mixed function can react with biphasic kinetics. Binding to multiple contiguous sites may exhibit concentration-dependent stoichiometry. Under such conditions a linear inhibition-binding relationship is neither a necessary nor a sufficient condition for the identification of transport sites.

Studies on the membrane glycoprotein defect of En(a-) erythrocytes. I. Biochemical aspects

Discontinuous sodium dodecylsulphate-polyacrylamide gel electrophoresis, followed by periodic acid/Schiff or Coomassie staining and densitometry, spectrophotometric and gas-liquid chromatographic carbohydrate analyses as well as heterophile agglutinins are employed to study the nature of the membrane glycoprotein defect in En(a-) erythrocytes from Finland and England, heterozygous Ena red cells from Finland and the erythrocytes of two individuals from Switzerland. The results suggest that En(a-) cells lack the major membrane sialoglycoprotein, the so-called MN glycoprotein. Heterozygous Ena erythrocytes from Finland and those from Switzerland have only about half of the normal amount of MN glycoprotein. The molecular weight of the major Coomassie staining membrane protein (component III) is increased by approx. 5000 and 3000 daltons in En(a-) and the other red cells respectively. Some aspects of this membrane defect are discussed.

Hemolytic anemias. Failure of the red cell membrane

The normal erythrocyte membrane is composed of nearly equivalent amounts of lipid and protein. The lipid portion of the membrane has been well studied. Even though de novo synthesis of lipid does not occur in human red cells, many biochemical pathways exist which facilitate detoxification of lipid breakdown products and lipid renewal. Rare defects in these processes are associated with hemolytic disorders. Recent studies have revealed that the membrane proteins are diverse and suggest that protein dysfunction may also account for clinical disease. Protein and lipid are entwined in a physicochemical relationship which is probably best depicted by the classic lipid bilayer with interspersed proteins in both the inner and outer surfaces and also spanning the bilayer. Membrane failure results in hemolytic anemia. This failure can be intrinsic, caused by abnormal lipid or protein constituents; or extrinsic, with a normal membrane being unable to counteract physical, chemical or immunologic stress. Clinical examples of membrane failure and hemolytic anemia can be separated into three groups according to the predominant mechanism of the hemolysis: fragmentation, whole-cell lysis, and filtration and entrapment. Although these mechanisms can act separately or in concert, the final hemolytic destruction of the cell can usually be traced to a failure of membrane function.

Protein chromatography on adsorbents with hydrophobic and ionic groups. Chromatography of dodecyl sulphate-solubilized proteins of the human erythrocyte membrane on N-(3-carboxypropionyl)aminodecyl-sepharose

Human erythrocyte 'ghosts' were solubilized in 0.5% (w/v) sodium dodecyl sulphate at pH 4.0(I = 0.012 mol/I). At a loading of 1-2 mg of protein/ml of column volume, all of membrane proteins were adsorbed to a column of CPAD [N-(3-carboxypropionyl)-aminodecyl]-Sepharose at pH 4.0 (I = 0-012 mol/1) and room temperature (22 degrees C). Many proteins were subsequently desorbed by raising the pH or by including sodium dodecyl sulphate continuously in the eluting buffer. Experiments with a series of adsorbents homologous with CPAD-Sepharose, in which the length of the hydrocarbon chain was varied, provided strong evidence of hydrophobic interactions, in addition to ionic interactions, in the binding of these proteins to CPAD-Sepharose. Elution with increasing-pH gradients at different concentrations of sodium dodecyl sulphate showed that glycophorin (the major sialoglycoprotein) was eluted in the void volume, at recoveries close to 100%, when the detergent concentration was greater than or equal to 0.3% (w/v). Protein E, the major protein, was desorbed late in the pH gradient even at a high (0.5%, w/v) concentration of the detergent, and was always incompletely desorbed, the maximum recovery recorded being 40%. Spectrin (the high-molecular-weight polypeptide pair) did not behave in a well-defined manner, and was found widely distributed among the effluent fractions under all the conditions that were tested.

Abnormal carbohydrate composition of the major penetrating membrane protein of En(a-) human erythrocytes

The major penetrating membrane glycoprotein (band 3) was isolated from En(a-) and normal human erythrocytes. The two proteins differed only in carbohydrate composition. Band 3 from En(a-) erythrocytes contained greater amounts of galactose and N-acetyl-glucosamine. The loss of the sialoglycoprotein sialotetrasaccharides in the En(a-) cell is not compensated by the appearance of these units in band 3 of En(a-) erythrocytes.

Reducible components in the proteins of human erythrocyte membrane

In contrast to a previous report, no collagen or elastin-type cross-linked derived from lysine-aldehydes were detected in human erythrocyte membranes. The major reducible components of erythrocyte membranes were shown to be hexosyllysines. From their structure it is clear that these components cannot act as cross-links between the protein subunits of the membrane. The components were also shown to be present in varying proportions in human serum albumin and haemoglobin. Whether the hexose attachments have any physiological significance or are artefacts of the analytical procedure has not yet been demonstrated. One other major reducible component was present but, although unidentified, this compound was shown to be unrelated to any of the known lysine-aldehyde-derived cross-links of collagen and elastin. A minor acidic component was identified as glucosylvaline derived from the N-terminus of the beta chain of haemoglobin A1c and not a lysine-aldehyde precursor of the collagen cross-links.

Transport of pyruvate nad lactate into human erythrocytes. Evidence for the involvement of the chloride carrier and a chloride-independent carrier

The kinetics and activation energy of entry of pyruvate and lactate into the erythrocyte were studied at concentrations below 4 and 15mM respectively. The Km and Vmax. values for both substrates are reported, and it is shown that pyruvate inhibits competitively with respect to lactate and vice versa. In both cases the Km for the carboxylate as a substrate was the same as its Ki as an inhibitor. Alpha-Cyano-4-hydroxycinnamate and its analogues inhibited the uptake of both lactate and pyruvate competitively. Inhibition was also produced by treatment of cells with fluorodinitrobenzene but not with the thiol reagents or Pronase. At high concentrations of pyruvate or lactate (20mM), uptake of the carboxylate was accompanied by an efflux of Cl-ions. This efflux of Cl- was inhibited by alpha-cyano-4-hydroxycinnamate and picrate and could be totally abolished by very low (less than 10 muM) concentrations of the inhibitor of Cl- transport, 4,4'-di-isothiocyanostilbene-2,2'-disulphonic acid. This inhibitor titrated out the chlordie efflux induced by pyruvate, bicarbonate, formate and fluoride, in each case total inhibition becoming apparent when approximately 1.2x10(6) molecules of inhibitor were present per erythrocyte, that is, about one inhibitor molecule per molecule of the Cl- carrier. Evan when Cl- efflux was totally blocked pyruvate and lactate uptake occurred. Kinetic evidence is presented which suggests that the Cl- carrier can transport pyruvate and lactate with a high Km and high Vmax., but that an additional carrier with a low Km and a low Vmax. also exists. This carrier catalyses the exchange of small carboxylate anions with intracellular lactate, is competitively inhibited by alpha-cyano-4-hydroxycinnamate and non-competitively inhibited by picrate. The Cl- carrier shows a reverse pattern of inhibition. It is concluded that net efflux of lactic acid from the cell must occur on the Cl- carrier and involve exchange with HCO3 - followed by loss of CO2. The low Km carrier might be used in pyruvate/lactate or acetoacetate/beta-hydroxybutyrate exchanges involved in transferring reducing power across the cell membrane. The possibility that the Cl- carrier exists in cells other than the erythrocyte is discussed. It is concluded that its presence in other cell membranes together with a low intracellular Cl- concentration would explain why the pH in the cytoplasm is lower than that of the blood, and why permeable carboxylate anions do not accumulate within the cell when added from outside.

The major sialoglycoprotein of the human erythrocyte membrane. Release with a non-ionic detergent and purification

The major sialoglycoprotein of the human erythrocyte membrane has been selectively released by the non-ionic detergent Tween 20 and further purified in detergent-free buffers by hydroxyapatite chromatography and, finally, by hydrophobic interaction chromatography on pentyl-Sepharose. The purified glycoprotein shows one main zone, PAS-1, and up to three minor zones after staining both for protein and carbohydrate in polyacrylamide gel electrophoresis in the presence of dodecyl sulfate. The relative staining intensities are concentration dependent. When the purified glycoprotein has been heated to 100 degrees C in dodecyl sulfate, more stain appears in the most rapid zone, PAS-2, and less in the slower zones, indicating a disaggregation of oligomeric forms of this glycoprotein, including a dimer, PAS-1.

Isolation of the major intrinsic transmembrane protein of the human erythrocyte membrane

The major intrinsic protein of the human erythrocyte membrane commonly referred to as "Band 3", was isolated by a multi-step procedure. Extraction of ghost membranes in dilute solutions of lithium diiodosalicylate removed most of the proteins considered to be extrinsic to the membrane. The resulting membrane fragments were solubilized in sodium dodecyl sulfate, and the major sialoglycoprotein (glycophorin A) was removed by wheat germ agglutinin-Sepharose affinity chromatography. Gel filtration in sodium dodecyl sulfate was used as the final step to yield the band 3 polypeptide in electrophoretically homogeneous form.

Subunit structure of human erythrocyte glycophorin A

Glycophorin A is a sialoglycoprotein isolated from human erythrocyte membranes which seems to exist as stable dimeric complexes in the presence of sodium dodecyl sulfate. When analyzed by dodecyl sulfate acrylamide electrophoresis this molecule forms two PAS-stainable bands (PAS-U and PAS-2) which are reversibly interconvertible. This change in electrophoretic mobility is dependent on the concentration of dodecyl sulfate, the use of Trisbuffer systems, the protein concentration in the incubation mixture, and the duration and temperature of incubation before electrophoresis. Reducing agents do no influence the results. Chromatography of the sialoglycopeptides on Sepharose columns in dodecyl sulfate before and after heat treatment gave similar results. A small hydrophobic peptide (T-6) derived from glycophorin A was able to prevent reassociation of the monomeric subunits back to the higher molecular weight form. This peptide was able to bind to the subunit of glycophorin A, but not to the high molecular weight complex. These results are consistent with a model of glycophorin A composed of two subunits which can dissociate and reassociate in the presence of detergents. These subunits may interact via the hydrophobic portions of the polypeptide chains.

Proteolytic dissection of band 3, the predominant transmembrane polypeptide of the human erythrocyte membrane

Band 3 is the major, membrane-spanning, approximately90 000 dalton polypeptide of the human erythrocyte membrane. To facilitate the analysis of its structural integration into the membrane, we have cleaved this protein in situ into large fragments and ascertained their disposition. Digestion of intact cells with chymotrypsin yielded band 3 fragments with apparent molecular weights of 38 000 and 55 000. Both fragments resisted elution by NaOH and acetic acid, suggesting that they are anchored in the apolar core of the membrane. Both pieces communicate with the extracellular space, and the 55 000 dalton species extends to the cytoplasmic surface as well. Digestion of unsealed ghosts with chymotrypsin produced a hydrophobic 17 000 dalton species, a segment of the 55 000 dalton fragment, which spans and is firmly anchored in the core of the membrane. Trypsin and papain at low concentration generated integral band 3 fragments of 52 000 daltons and released major band 3 fragments of less than or equal to 41 000 daltons from the cytoplasmic side of the membrane. The latter water-soluble polypeptides remained associated in discrete complexes which retained the capacity to bind glyceraldehyde-3-phosphate dehydrogenase. An interchain disulfide bond, which can be induced only at the cytoplasmic surface, cross-linked intact band 3, and certain of its water-soluble fragments. Finally, fragments of 23 000 daltons were generated from the innersurface domain by reacting disulfide-linked band 3 dimers with cyanide or reduced polypeptides with 2-nitro-5-thiocyanobenzoate. A provisional ordering of these fragments is proposed.

Spatial arrangement of some erythrocyte membrane receptor sites

Alcian Blue proved to be a hemagglutinin specific for negatively charged receptors. Either sialylization or tryptic degradation of glycoproteins of human erythrocytes greatly diminished the agglutinability by Alcian Blue. Studies with group A antiserum, anti-AHP and the lectin of Lens culinaris demonstrated the masking efficiency of erythrocyte major glycoproteins. Previous proteolytic digestion enhanced the agglutinability of human red blood cells due to uncovering of what is considered glycolipidic group A receptors. The findings provide indirect evidence for the spatial arrangement of erythrocyte receptor sites.

Chemical modification of membrane proteins in relation to inhibition of anion exchange in human red blood cells

Mono-, di-, and trisulfonic acids, including 4,4'-diacetamido stilbene-2,2'-disulfonic acid (DAS) and 2-(4'-amino phenyl)-6-methylbenzene thiazol-3',7-disulfonic acid (APMB) produce a reversible inhibition of sulfate equilibrium exchange in human red cells. A study of the sidedness of the action of a number of these sulfonic acids in red cell ghosts revealed that some, like DAS, inhibit only at the outer membrane surface while others, like APMB, inhibit at either surface. This finding suggests that at least two different types of membrane sites are involved in the control of anion permeability. The nature of the anion permeability controlling sites in the outer cell surface was investigated by studying the effects of DAS on the inhibition by dinitrofluorobenzene (DNFB) of anion equilibrium exchange and on the binding of DNFB to the proteins of the red blood cell membrane. After exposure to DNFB in the presence of DAS for a certain period of time, there was a reduction of both the inhibitory effect of DNFB on sulfate exchange and the binding of DNFB to the protein in band 3 of SDS polyacrylamide gel electropherograms (nomenclature of Steck, J. Cell. Biol., 62: 1, '74). Since binding to other membrane proteins was not affected, this observation supports the assumption that the protein in band 3 plays some role in anion transport. In accordance with the absence of an inhibitory effect at the inner membrane surface, internal DAS does not affect DNFB binding to the protein in band 3. DAS protected the anion exchange system not only against inhibition by DNFB but also by m-isothiocyanato benzene sulfonic acid. In contrast to DAS, the equally inhibitory phlorizin does not reduce the rate of dinitrophenylation of the protein in band 3. This suggests that either not all inhibitors of anion exchange exert their action by a combination with sites on the protein in band 3 or that in spite of the described evidence this protein is not involved in the control of anion movements.

Quantitative immunoelectrophoresis of proteins in human erythrocyte membranes. Analysis of protein bands obtained by sodium dodecyl sulfate-polyacrylamide gel electrophoresis

1. We have defined conditions that permit quantitative immunoelectrophoresis in agarose gels of dodecyl sulfate-solubilized erythrocyte membrane proteins. 2. Using human serum albumin, transferrin, MN-glycoprotein (glycophorin) and crude spectrin as test proteins, we found that accurate analyses are possible if samples and gels are 1% in non-ionic detergent (Berol EMU-043) or Triton X-100) and if no more than 100 nmol free dodecyl sulfate is applied per sample. 3. Dodecyl sulfate treated membranes analyzed by crossed immunoelectrophoresis using rabbit antibodies against membrane material yielded optimal precipitation patterns in gels containing 1% of non-ionic detergent. 4. Crossed immunoelectrophoresis in the presence of 1% of Berol revealed precipitates when 10 protein bands defined and isolated by preparative dodecyl sulfate-polyacrylamide gel electrophoresis were run against anti-membrane antibodies. Seven of these bands showed more than one precipitation arc, indicating the presence of more than one antigenic component. 5. Crossed-line immunoelectrophoresis showed that dodecyl sulfate-polyacrylamide gel electrophoresis bands 1, 2 and 2.1 shared common antigenic components. The MN-glycoprotein was present in bands 3, 4A, 4B and 5, where antigenic components of the major intrinsic erythrocyte membrane protein, band 3, were also found. 6. After absorption of the anti-membrane antibody with intact erythrocytes, immunoelectrophoresis showed the disappearance of the MN-glycoprotein precipitates. An increase in the area below the precipitate corresponding to the major intrinsic protein (band 3) was also observed, indicating exposure of some antigens of this protein on the outer surface of intact cells. 7. After absorption of the antibody preparation with washed erythrocyte membranes, immunoprecipitates were not seen in any experiments, indicating that all antigenic determinants observed are exposed at one or both surfaces of the membrane. 8. Our analyses indicate that the peptide moieties of serum lipoproteins do not constitute a significant component of erythrocyte membranes.

Separation of ABH, I, Ss antigenic activity from the MN-active sialoglycoprotein of the human erythrocyte membrane

The erythrocyte sialoglycoprotein was purified from cells with complementary ABO(H), MN and Ss phenotypes. Serological examination of the sialoglycoprotein preparations demonstrated that this molecule does not carry the ABO(H), I or Ss antigens. The results also suggest that Ss activity may be associated with a minor erythorcyte glycoprotein. The characteristics of the ABO(H) and I antigens are consistent with earlier suggestions that these antigens are carried on complex glycolipids.

A critical evaluation of the proposal that serum apolipoproteins are the major constituents of the human erythrocyte membrane

1. The EDTA and Triton X-100 extracts of human erythrocyte ghosts gave no precipitin lines in double diffusion analyses with antibodies to either lipoprotein A, lipoprotein B, lipoprotein C, lipoprotein D, Lp(a) lipoprotein or arginine-rich apolipoprotein of normal human serum (for nomenclature for serum lipoprotein families and apolipoptoteins, see Alaupovic, P., Kostner, G., Lee, D. M., McConathy, W.J. and Magnani, HN. (1972) Expo. Annu. Biochem. Med. 31, 145-160 and Alaupovic, P., Lee, D.M. and McConathy, W.J., (1972) Biochim. Biophys. Acta 260, 689-707.) These membrane preparations also reacted negatively with commercially available antisera to alpha- and beta-lipoproteins. 2. The normal serum very low density, low density and high density lipoproteins formed no precipitin lines with antibodies to either intact or EDTA-extracted ghosts. 3. The serum apolipoproteins and their constitutive polypeptides (A-I, A-II, B, C-I, C-II, C-III, D and arginine-rich apolipoprotein) reacted negatively with antibodies to intact or EDTA-extracted ghosts. The EDTA and Triton X-100 extracts of erythrocyte ghosts gave no reaction with monospecific antibodies to serum apolipoproteins and their constitutive polypeptides. 4. Ghosts dissolved 2% sodium dodecyl sulfate gave positive immunoprecipitin lines with antisera to alpha- and beta-lipoproteins. However, the sodium dodecyl sulfate solution in concentrations greater than 0.1% also formed precipitin lines with antisera to the same lipoproteins. 5. These results do not support the suggestion (Langdon, R.G. (1974) Biochim. Biophys. Acta 342, 213-228) that serum apolipoptoteins are integral protein constituents of human erythrocyte ghosts. The immunoprecipitin lines observed in the latter study might have been due to the presence of trace amounts of serum lipoproteins loosely attached to the cellular surfaces or, more probably, resulted from nonspecific interactions between the proteins and the sodium dodecyl sulfate used as the solubilizing agent

Two-dimensional polyacrylamide-gel electrophoresis of the proteins and glycoproteins of the human erythrocyte membrane

A two-dimensional polyacrylamide gel electrophoresis technique has been developed, improving the analytical separation of some proteins and glycoproteins of the human erythrocyte membrane. Freshly prepared membranes are totally solubilized, subjected to dodecylsulfate--polyacrylamide gel electrophoresis in the first dimension, followed by electrophoresis in the second dimension, using a detergent-free polyacrylamide gradient gel. By this method the proteins of the human erythrocyte membrane could be resolved into a two-dimensional pattern, which has been shown to be highly reproducible with respect to various blood-groups and within one blood-group from specimen to specimen. The method enables especially the investigation of the hydrophobic and very likely integrated membrane proteins and glycoproteins. Thus, band III[Fairbanks, G., Steck, Th. & Wallach, D. F. H., Biochemistry, 10, 2606--2617 (1971)] could be shown to consist of five proteins, one of them being the major glycoprotein of the human erythrocyte membrand. The two spectrin bands differed considerably in their two-dimensional patterns. The value of the given method for the investigation of membrane defects, which may be linked with various diseases of human erythrocytes, could be demonstrated in the case of two patients suffering from congenital dyserythropoetic anaemia.

The major human erythrocyte membrane protein. Evidence for an S-shaped structure which traverses the membrane twice and contains a duplicated set of sites

The structure of the major human erythrocyte membrane protein (protein E) was investigated by studying the products of proteolysis of the native protein in the membrane. The distribution and location of the tyrosine residues labelled by radioiodination by lactoperoxidase was determined. Proteolysis of the extracellular region of the protein by thermolysin released four tyrosine-containing peptides, all of which were also found to remain in the major fragment that is retained in the membrane. The presence of these duplicated sites in the extracellular region of the protein was confirmed by limited trypsin digestion of the intracellular region of the protein. Two groups of fragments were obtained. Both groups contained a set of the extracellular labelled sites, but they differed in containing distinct groups of intracellular sites, showing that the two sets of extracellular sites are linked by an intracellular region of the protein. The polypeptide chain thus traverses the membrane twice. An S-shaped model which is consistent with these data is proposed.

Rapid preparative isolation of major erythrocyte membrane proteins using polyacrylamide gel electrophoresis in sodium dodecylsulfate

1. We describe a simple method for preparative, sodium dodecylsulfate/polyacrylamide gel electrophoresis of the major proteins in human erythrocyte membranes. 2. The method is based on extraction of prestained proteins from gel slabs. Three different fluorescent dyes (o-phthalaldehyde, fluorescamine and 1-dimethyl-aminonaphthalene-5-sulfonylchloride) have been used for pretaining. The method allows separation of up to 75 mg membrane protein and isolation of mg quantities of all major erythrocyte ghost proteins, while preserving the high resolution of analytical polyacrylamide gel electrophoresis. 3. Yield depends on extraction conditions and the molecular weight of the proteins being eluted. It ranges from 43-48% for protein 1 (apparent mol. wt approx. 310000) and 72-78% for protein 3(apparent mol. wt 87 000-93 000) to 87-93% for protein 6 (apparent mol. wt 35 000). 4. The labile behaviour of the high molecular "spectrin" bands (bands 1 and 2) is described. Extraction at room temperature tends to split these proteins into products of lower molecular weight. In contrast, the minor protein components 2.1 and 2.2 tend to aggregate yielding components 1 and 2. 5. N-terminal amino acid analyses have been performed on proteins 1, 2, 3, 4A, 4B, 5 and 6. Each of these bands contains several N-terminals, most of which appear constant. Some additional N-terminal amino acids vary from one donor to the next.

Proteolytic activity associated with human erythrocyte membranes. Self-digestion of isolated human erythrocyte membranes

At least two kinds of enzymes are active in the proteolytic self-digestion of erythrocyte membranes. The specific activities of these enzymes do not decrease with repeated washings of purified stroma. The effects of a variety of inhibitors on the membrane preparation's capacity to digest 125-I-labelled casein, covalently linked to latex beads, have been examined. Pepstatin-inhibitable enzyme, active at low pH, digests the membrane extensively to small polypeptide fragments. Spectrin, located at the internal part of the membrane, is readily degraded. Diisopropylfluorophosphate-inhibitable enzyme, active at pH 8-9, has only limited digestive capacity. Some of the membrane components, such as the small molecular weight glycoproteins, are resistant to digestion. The restricted capacity of digestion is due to the membrane molecular arrangement; increased disaggregation removes the restriction and increases the activity. Spectrin is not digested unless the membrane topography is disrupted by NP-40 neutral detergent. These observations suggest that the enzymes active at basic pH are located external to the cell. Intact cells do possess a limited capacity to degrade 125-I-labelled casein when their surfaces are brought into contact with substrate-coated beads.

A pulse radiolysis study of some free radical reactions with erythrocyte membranes

The reactions of the free radicals eaq- minus, OH and Br2- minus with haemoglobin-free erythrocyte ghost membranes have been studied by producing the radicals by pulse radiolysis and monitoring their reactions by optical spectroscopy. Hydrated electrons react rapidly with the membrane, but no attack at disulphide links was observed. Hydroxyl radical attack produced transient species absorbing weakly in the ultraviolet, which may arise from carbohydrate residues, such as N-acetyl neuraminic acid and N-acetyl glucosamine, on the membrane surface. No evidence was obtained for OH attack at ring-containing amino acid residues of the protein component. The Br2- minus radical, a more selective electrophile than OH, reacted only slowly with erythrocyte ghosts. Solubilization of the membranes with dodecylsulphate or digestion with alkali exposed protein containing tyrosine and tryptophan residues which reacted with Br2- minus. These results support other evidence for the absence of reactive protein at the membrane surface.

The N-terminal heterogeneity of edta-extractable erythrocyte membrane proteins

N-terminal analysis of the proteins extracted from ox (Bos taurus) erythrocyte membranes by dilute EDTA is used as a means of estimating the heterogeneity of the protein fractions. Dinitrophenylation and dansylation reveal up to 8 different N-terminal amino acids in fractions with electrophoresis after dodecylsulphate treatment shows as having far fewer polypeptide chains. These fractions are prepared by gel electrophoresis in the presence and absense of detergent. Molecular weight estimation by the Ferguson procedure of the components revealed in the EDTA extract by electrophoresis in the absence of detergent confirms the large size concluded from electrophoresis with dodecylsulphate. The N-terminal analyses indicate that the high molecular weight of the complexes, both in the presence and absence of detergent, is more probably due to their being complexes of several polypeptides than exceptionally long single polypeptide chains.

Circular dichroism of erythrocyte membrane glycoproteins

The circular dichroism spectra were obtained for purified equine, human and bovine membrane glycoproteins, which have 40, 55 and 70% carbohydrate, respectively. The spectra in aqueous buffer show similar shapes, maxima and minima but somewhat different peak amplitudes. Analysis of the spectra indicated that the glycoproteins can be pictured as existing primarily in an unordered form in dilute aqueous buffer with small amounts of alpha-helix (13-23%) present. In 2-chloroethanol, a helix-promoting solvent, the amount of alpha-helix is increased to 60-70%. The glycoproteins underwent denaturation in guanidine hydrochloride, although evidence of some residual structure did remain. The spectra of the glycoproteins change relatively little on going from aqueous buffer to dodecylsulfate solutions. Removal of 60% of the sialic acid does not induce significant conformational alterations. The anomalous behavior of the glycoproteins during molecular weight determinations does not appear to be related primarily to conformational restrictions on the polypeptide chain.

Isolation and partial characterization of two minor glycoproteins from human erythrocyte membranes

Two minor glycoproteins GP-II and GP-III, were isolated from human erythrocyte membranes and characterized chemically and immunologically. The chemical composition of GP-II and GP-III was similar: GP-II consisted of 81% protein and 19% carbohydrate of which 4.9% was hexose, 5.4% hexosamine and 7.8% sialic acid. GP-III consisted of 76% protein and 24% carbohydrate of which 7.6% was hexose, 7.2% hexosamine and 8.1% sialic acid. The amino acid composition of GP-II and GP-III was also similar. GP-II and GP-III, however, differed in chemical composition from the MN glycoprotein. GP-II and GP-III were associated with the blood group activities Ss, I and A, but not with the MN antigens. GP-III had higher blood group activities per mug of protein than did GP-II. The specific activities for the Ss blood group antigens were increased 3-10-fold by purificantion of GP-III from the aqueous phase of chloroform methanol extracts.

Sodium dodecyl sulfate in protein chemistry

This review summarizes in a brief manner the main aspects of the application of sodium dodecyl sulfate (SDS) to protein chemistry. The principal problems of SDS-polyacrylamide gel electrophoresis are described, as well as the anomalous behavior of protein-SDS complexes and the inactivation of enzymes due to variable binding of SDS to the polypeptides studied. The particular value of SDS in elucidating the protein composition of biological membranes and in membrane-reconstitution experiments is discussed.

Differences in the protein fluorecence of the two iron(III)-binding sites of ovotransferrin

1. Changes in the tryptophan fluorescence and the visible absorption spectrum resulting from the combination of apo-ovotransferrin with Fe3+, F,E2+, Cu2+, Zn2+, Mn2+, and Cd2+were measured. 2. As expected for a radiationless transfer of electronic excitation energy, only the ions Fe3+, Fe2+and Cu2+, which gave complexes with large extinctions between 300 and 370nm, resulted in large decreases in trytophan fluorescence. 3. The decrease in protein fluorescence was non-linear with increasing occupancy of the Fe3+ -and Cu2+ - binding sites. The decrease in fluorescence on binding of Fe3+ was biphasic and showed that the two metal-binding sites were being occupied sequentially at pH7.4-8.4. The first site reacted with Fe3+ instantaneously, the second was occupied over a minute. 5. The nonidentity of the two sites was also demonstrated by the preparation of a stable hybrid containing both Cu2+ and Zn2+.h Cu2+ and Zn2+

Reactions of erythrocyte glycoproteins and their degradation products with various anti-I sera

Three fractions of erythrocyte glycoproteins obtained from Sepharose 4-B chromatography were tested for I activity with ten serologically differentiated anti-I sera. The most active was fraction I, eluted at the void volume and containing the lowest amount of alkali-labile oligosaccharide chains. The desialization of glycoproteins increased their activity toward anti-I-s and anti-I-D sera, and did not change or decreased the activity toward anti-I-F sera. The most abundant fraction II (major sialoglycoprotein of erythrocyte membranes) showed no or only a very weak I activity, but I-active glycopeptides were isolated from products of digestion of fraction II with trypsin. The major product of digestion, sialoglycopeptide IIT-2 showed I activity only after alkaline elimination of alkali-labile oligosaccharide chains. The results indicate that I receptors are present in hindered form on apparently I-inactive components of erythrocyte membrane.