Human Erythrocyte Protein Research Articles

Asparaginyl deamidation-methylation of rat ventricular myosin light chains

Spontaneous asparaginyl deamidation can produce damage to cytoskeletal proteins, and may lead to their targeting for subsequent rapid intracellular breakdown or repair. To test if myofibrillar proteins are subject to spontaneous deamidation damage in vitro, purified rat ventricular myosin light chain 1 (MLC 1v) and phosphorylatable myosin light chain 2 (MPLC 2v) were incubated (37°C, 4 h, pH 2–11), and tested as substrates for human erythrocyte and rat cardiac protein carboxyl methyltransferase (PCMT). PCMT catalyzes the transfer of a methyl group from [ 3H-methyl] S-adenosyl methionine to deamidated asparaginyl residues and altered aspartyl residues on damaged proteins. MLC 1v and MPLC 2v underwent extensive incubation damage at neutral and alkaline pH. Sodium dodecyl sulfate polyacrylamide gel electrophoresis and fluorography revealed 3H-incorporation into MLC 1v, MPLC 2v, and a M r = 14 000 polypeptide. 3H-methylated, CNBr-cleavage fragments of PCMT-methylated light chains were then separated by reverse-phase high performance liquid chromatography, and sequenced by automated Edman degradation. The major 3H-labeled peptide of the M r = 14 000 protein proved homologous to residues 84 to 104 of rat MPLC 2v, with a proposed deamidation site at Asn 99-Ala 100. The major 3H-labeled peptide from MLC 1v, proved homologous to residues 73 to 111 of rat cardiac MLC 1v, with a proposed deamidation site at Asn 108-Ser 109. These results indicate that both myofibrillar protein subunits undergo selective non-enzymatic degradation at neutral and alkaline pH, resulting in the formation of methyl acceptor sites for human erythrocyte and rat cardiac PCMT. PCMT-catalyzed methylation of ventricular myosin light chains may be important in the repair, or subsequent proteolysis of these long-lived structural proteins of the myofibril.

Presence and localization of proteins immunologically related to erythrocyte protein 4.1 in human skin

Analogues of human erythrocyte protein 4.1 have been examined in the human skin by immunochemical techniques using anti-human erythrocyte protein 4.1 antibodies. Immunoblot analysis revealed that human epidermis contains 4.1-like proteins of 80 kDa and 78 kDa that cross react with anti-protein 4.1 antibodies. Analysis with immunofluorescence microscopy revealed that the plasma membrane of the human epidermal keratinocyte was stained intensively in the basal cells, whereas spinous cells were moderately stained. It is noted that eccrine sweat gland cells and ductal cells were also stained in the peripheral cytoplasma. Taken together, these results demonstrate that 4.1-like proteins are present in human epidermal keratinocytes, eccrine sweat gland cells and ductal cells. The present findings enable us to suggest that a membrane skeletal protein lattice might exist in these cells.

Effects of diethyl ether on membrane lipid ordering and on rotational dynamics of the anion exchange protein in intact human erythrocytes: correlations with anion exchange function

The roles of lipid ordering and protein dynamics on the function of the anion exchange protein (band 3) in intact human erythrocytes have been investigated. The effects of diethyl ether on the ordering of membrane lipids and on the rotational dynamics of band 3 were measured by EPR and saturation-transfer EPR spectroscopies, respectively, and correlated with the anion exchange function of band 3. With increasing concentration, diethyl ether monotonically decreased the ordering of membrane lipids near the polar head-group region, as reported by the lipid-soluble spin probe 5-doxylstearic acid, but produced comparatively little change in the ordering of lipids in the hydrophobic midzone, as reported by 16-doxylstearic acid. The rotational mobility of band 3, as reported by the affinity spin-label bis(sulfo-N-succinimidyl) doxyl-2-spiro-5'-azelate [Anjaneyulu et al. (1989) Biochemistry 28, 6583-6590], also increased monotonically with increasing ether concentration. This increase in rotational mobility was not due to a demonstrable change in its state of oligomerization, since band 3 was readily cross-linked by bis(sulfo-N-succinimidyl) suberate to covalent dimers in the presence or absence of ether. At concentrations up to 2 vol % ether, hemolysis of erythrocytes was negligible, and the spectroscopic changes observed were completely reversed following its removal. Km, Vmax, and Eact. for sulfate uptake into chloride-loaded erythrocytes were not significantly affected by addition of ether.(ABSTRACT TRUNCATED AT 250 WORDS)

Thiol-fatty acylation of the glucose transport protein of human erythrocytes

Incubation of intact human erythrocytes with [ 3H]palmitate labeled a protein with electrophoretic characteristics of the glucose transporter. This labeling occurred via a thioester linkage, since it was unaffected by organic solvent extraction, but was substantially removed as the hydroxamate upon treatment with neutral hydroxylamine. Immunoprecipitation of the labeled protein with a monoclonal antibody to the glucose transporter confirmed its identity.

Identification of the eosinyl-5-maleimide reaction site on the human erythrocyte anion-exchange protein: overlap with the reaction sites of other chemical probes

The anion-exchange protein (band 3) reaction site in human erythrocytes for the fluorescent/phosphorescent probe eosinyl-5-maleimide (EMA) has been identified. Proteolytic dissection of band 3 in situ indicated that EMA reacts with the membrane-spanning Mr 17K peptide produced by chymotrypsin cleavage of band 3 in intact erythrocytes followed by removal of the cytoplasmic domain by mild trypsin digestion of ghost membranes. Sequencing of the major eosin-labeled peptide obtained from HPLC purification of an extensive chymotrypsin digest of purified Mr 17K peptide allowed assignment of the covalent reaction site for EMA to lysine-430 of the human erythrocyte protein [Tanner et al. (1988) Biochem. J. 256, 703-712]. Hydropathy plots based upon the primary structure of the protein [Lux et al. (1989) Proc. Natl. Acad. Sci. U.S.A. 86, 9089-9093] suggest that this residue is in an extracellularly accessible loop connecting membrane-spanning segments 1 and 2 of native band 3 in the erythrocyte membrane. Inhibition of sequential labeling of intact erythrocytes by pairs of chemical probes including EMA, the anion transport inhibitor 4,4'-diisothiocyanodihydrostilbene-2,2'-disulfonate (H2-DIDS), and the reactively bifunctional spin-label bis(sulfo-N-succinimidyl) doxyl-2-spiro-5'-azelate (BSSDA) has also been investigated. Each of these reagents affinity labels band 3 when added separately to a suspension of intact human erythrocytes by formation of one or more stable covalent bonds. Prelabeling of intact erythrocytes with EMA reduced subsequent labeling of band 3 by H2-DIDS by approximately 95% and by BSSDA by 90%. Similarly, prelabeling with H2-DIDS reduced subsequent labeling of band 3 by EMA by over 90%, and BSSDA prelabeling reduced EMA labeling by approximately 95%. Therefore, though having widely divergent chemical structures and protein modification reactivities, each of these negatively charged reagents may be competing for reaction with spatially overlapping sites on band 3 which are accessible from the extracellular space.

Identification and preliminary characterization of two related proliferation-associated nuclear phosphoproteins.

Two nuclear phosphoproteins, pp35 and pp32, were purified from A20 cells, a murine B-lymphoblastoid cell line. Initially detected by cross-reactivity with antibodies to human erythrocyte protein 4.1, the 35- and 32-kDa proteins were purified by sequential fractionation of non-ionic detergent cell lysates on DEAE-cellulose, high performance liquid chromatography (HPLC)-anion-exchange chromatography, and HPLC hydroxylapatite chromatography. By two-dimensional peptide mapping, pp35 and pp32 are related but do not appear to represent sequential proteolytic products. Both pp35 and pp32 appear to be associated with cell proliferation. Antibodies specific for pp35 and pp32 show prominent intranuclear staining in A20 cells but only focal staining in normal murine lymphoid tissues. Quantitative immunoblotting showed that both pp35 and pp32 are, respectively, expressed at 5.9 x 10(4) and 7.0 x 10(4) copies/cell in small, dense resting B lymphocytes, increasing approximately 12- and 7-fold after polyclonal stimulation with lipopolysaccharide. When normalized to total cell protein, this represents specific inductions of approximately 4- and 2-fold. Expression of both pp35 and pp32 is constitutively high in populations of neoplastic B cell lines; moreover, both are expressed in the nuclei of intestinal crypt epithelial cells but not in other epithelial compartments in the same sections, suggesting that forms of pp35 and pp32 may be expressed in additional tissues and associated with proliferation.

Identification and reconstitution of the nucleoside transporter of CEM human leukemia cells

The major nucleoside transporter of the human T leukemia cell line CEM has been identified by photoaffinity labeling with the transport inhibitor nitrobenzylmercaptopurine riboside (NBMPR). The photolabeled protein migrates on SDS-PAGE gels as a broad band with a mean apparent molecular weight (75 000 ± 3000) significantly higher than that reported for the nucleoside transporter in human erythrocytes (55 000) (Young et al. (1983) J. Biol. Chem. 258, 2202–2208). However, after treatment with endoglycosidase F to remove carbohydrate, the NBMPR-binding protein in CEM cells migrates as a sharp peak with an apparent molecular weight (47 000 ± 3000) identical to that reported for the deglycosylated protein in human erythrocytes (Kwong et al. (1986) Biochem. J. 240, 349–356). It therefore appears that the difference in the apparent molecular weight of the NBMPR-sensitive nucleoside transporter between the CEM cell line and human erythrocytes is a result of differences in glycosylation. The NBMPR-binding protein from CEM cells has been solubilized with 1% octyl glucoside and reconstituted into phospholipid vesicles by a freeze-thaw sonication technique. Optimal reconstitution of uridine transport activity was achieved using a sonication interval of 5 to 10 s and lipid to protein ratios of 60:1 or greater. Under these conditions transport activity in the reconstituted vesicles was proportional to the protein concentration and was inhibited by NBMPR. Omission of lipid or protein, or substitution of a protein extract prepared from a nucleoside transport deficient mutant of the CEM cell line resulted in vesicles with no uridine transport activity. The initial rate of uridine transport, in the vesicles prepared with CEM protein, was saturable with a K m of 103 ± 11 μM and was inhibited by adenosine, thymidine and cytidine. The K m for uridine and the potency of the other nucleosides as inhibitors of uridine transport (adenosine > thymidine > cytidine) were similar to intact cells. Thus, although the nucleoside transporter of CEM cells has a higher molecular weight than the human erythrocyte transporter, it exhibits typical NBMPR-sensitive nucleoside transport activity both in the intact cell and when reconstituted into phospholipid vesicles.

Metabolic depletion effect on serine/threonine- and tyrosine-phosphorylations of membrane proteins in human erythrocytes

The response of serine/threonine-phosphorylation of the major transmembrane protein (band 3) in human erythrocytes to the metabolic state of the cells is different from that exhibited by the tyrosine-phosphorylation of the same protein. Precisely, both serine- and tyrosine-phosphorylation are decreased during metabolic depletion of the erythrocytes. However, the depletion-induced tyrosine-phosphorylation decrease of band 3 is not reversed by the subsequent metabolic repletion of the depleted cells, being accompanied by an irreversible inactivation of both membrane-bound and cytosolic tyrosine-protein kinase(s). By contrast, the depletion-induced phosphoserine-dephosphorylation is reversed by the following repletion, being accompanied by a reversible translocation of casein kinase(s) between cytosolic and membrane compartments. A possible functional correlation between the serine-phosphorylation state of band 3 protein and the band 3-mediated anion transport across the membrane is discussed.

Tyrosine phosphorylation of cytosolic proteins in human erythrocytes

Some cytosolic proteins of human erythrocytes can be phosphorylated on tyrosine residues by endogenous Tyr-protein kinase(s).Their phosphorylation is enhanced by addition of Tyr-protein kinase, purified from human erythrocyte cytosol.The most phosphorylatable is a 19 kDa protein. Its phosphorylation is more activated by Mn2+ than by Mg2+. It is inhibited by NaCl, 2,3-bisphosphoglycerate and by heparin.Similar response to the above effectors is exhibited by the phosphorylation of the other protein bands. However, the phosphorylation of a 73 kDa double band, which is negligible in the absence of added NaCl, is stimulated by this salt.

Fatty acid acylation of membrane skeletal proteins in human erythrocytes

Fatty acid acylation of membrane proteins was studied on human erythrocytes by measuring incorporation of [ 3H]palmitate at different specific radioactivities. A 55 kDa polypeptide within the band 4.5 region was the main acceptor protein for acylation by fatty acids (palmitate, stearate, oleate), while other polypeptides (80,65,48, 30 kDa) incorporated [ 3H]palmitate slowly, in substoichiometric amounts. Integral membrane proteins were preferentially fatty acid acylated. Skeletal membrane proteins were, however, poorly labeled. Neither purified ankyrin nor band 4.1 protein were fatty acid aeylated in human erythrocytes. On the other hand, label associated with high molecular weight skeletal proteins resisted low and high ionic strength extractions, and was extracted selectively by uran along with a small subpopulation of spectrin which was also tightly associated with the membrane.

Structure of the Promoter Region and Tissue Specificity of the Human Glycophorin C Gene

Glycophorin C (GPC) is an integral membrane protein of human erythrocytes which plays an important role in regulating the deformability and mechanical stability of red cells. Recently, the structural gene for this glycoprotein has been cloned (Colin, Y., Le Van Kim, C., Tsapis, A., Clerget, M., d'Auriol, L., London, J., Galibert, F., and Cartron, J. P. (1989) J. Biol. Chem. 264, 3773-3780), and we have now determined the sequence of the 1050 base pairs of DNA preceding the transcription initiation site mapped in erythroid cells. This region contains different potential regulatory cis-acting elements found in a variety of eukaryotic promoters (TATA box, CAAT box, Sp1-binding site) as well as sequences present in the promoter and enhancer regions of genes specific for the erythroid lineage (CACCC box and NF-E1-binding site). Northern blot analysis and immunological studies indicate that the GPC gene is expressed in a large number of cells and tissues. However, the level of transcription as well as the glycosylation of the mature GPC differ in erythroid and nonerythroid cells. Primer extension analysis and mapping of the 5' end GPC mRNA by the polymerase chain reaction indicate that different transcription sites are utilized for the expression of the GPC gene in erythroid and nonerythroid tissues and cell lines.

Functional expression of the human HepG2 and rat adipocyte glucose transporters in Xenopus oocytes: Comparison of kinetic parameters

Facilitated glucose transport is a ubiquitous characteristic of animal cells carried out by a family of membrane glycoproteins. Two members of this gene family are the well characterized human erythrocyte protein that has been cloned from the HepG2 cell line and the insulin-sensitive transporter that has been cloned from adipocytes and muscle tissue. In the present study the HepG2 and adipocyte glucose transporters were functionally expressed in Xenopus oocytes after injection of synthetic mRNAs produced by transcription in vitro from cloned cDNAs. Both 2-deoxyglucose uptake and 3-O-methylglucose transport were increased several-fold over basal levels in mRNA-injected oocytes. Increased uptake of 2-deoxyglucose was inhibited completely in the presence of cytochalasin B, and 3-O-methylglucose transport was blocked by D-glucose but not by L-glucose. The half-saturation constant and turnover number for 3-O-methylglucose transport at 22 degrees C via the HepG2 transporter were estimated to be 21 mM and 2.2 x 10(3) s-1 under equilibrium exchange conditions. The half-saturation constant for 3-O-methylglucose transport via the adipocyte transporter under the same conditions was estimated to be 1.8 mM. These data prove the functional identity of the cloned HepG2 and adipocyte cDNAs and indicate that the HepG2 and adipocyte transporters display similar kinetic behavior when expressed in the frog oocyte membrane as compared with their native membrane environments. Thus, the difference in the equilibrium exchange kinetic parameters for glucose transport in the erythrocyte and the adipocyte is a result of the expression of two distinct glucose transporter proteins.

The Gαz Gene Product in Human Erythrocytes: Identification as a 41-Kilodalton Protein

Abstract A cDNA encoding a previously unknown G protein alpha-subunit lacking the site for pertussis toxin-catalyzed ADP-ribosylation was recently cloned and its putative protein product named Gz (Fong, H. K. W., Yoshimoto, K. K., Eversole-Cire, P., and Simon, M. I. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 3066-3070) or Gx (Matsuoka, M., Itoh, H. Kozasa, T., and Kaziro, Y. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 5384-5388). A synthetic peptide corresponding to the deduced carboxyl-terminal decapeptide of this putative protein (alpha z) has been synthesized and used to prepare a polyclonal rabbit antiserum directed against the protein. The specificity and cross-reactivity of this antiserum was assessed using bacterially expressed recombinant G protein alpha-subunit fusion proteins (r alpha). The crude antiserum strongly recognizes r alpha z in immunoblots. Pretreatment of antiserum with antigen peptide greatly reduces the interaction of the antiserum with r alpha z. Affinity purified antiserum strongly recognizes expressed r alpha z, does not recognize r alpha s1, r alpha s1, r alpha o, or r alpha i3, and very weakly interacts with r alpha i1 and r alpha i2. In contrast, the alpha-subunits of purified bovine brain Gi1 and human erythrocyte Gi2 and Gi3 did not react with the alpha z-antiserum. Partially purified mixtures of human erythrocyte G proteins contain a 41-kDa protein that reacts specifically in immunoblots with both crude and affinity purified alpha z-specific antiserum. Quantitative immunoblotting using r alpha z as a standard indicates that there is 60-100 ng of alpha z/micrograms of 40/41-kDa alpha-subunit protein in partially purified human erythrocyte G protein preparations. We conclude that we have identified the alpha z gene product as a 41-kDa trace protein in human erythrocytes.

Primary structure of the cytoplasmic domain of human erythrocyte protein band 3. Comparison with its sequence in the mouse.

We report here the peptide profile of the human cytoplasmic domain of band 3 protein (CDB-3). The peptide alignment was designed allowing for maximal homology with the murine protein whose sequence was deduced from cDNA analysis by Kopito and Lodish (Kopito, R.R., Anderson, M. and Lodish, H.F. (1987) J. Biol. Chem. 262, 8035-8040). In the human protein, part of the amino acid sequence has been determined by Kaul et al. (Kaul, R.K., Murthy, P.S.N., Reddy, A.G., Steck. T.L. and Kohler, H. (1983) J. Biol. Chem. 258, 7981-7990). We have sequenced most of the fragment not described by these author. The homology with the murine protein is high (90%), except in a few peptides where it is only 50%. The actual miniaturization of the techniques allows for the determination of a clear peptide profile of human CDB-3 starting from 10 ml blood samples. Our characterization of the peptide profile of membrane proteins is the first step towards the identification of genetic mutations, which have to be looked for in hemolytic anemia when the presence of an abnormal membrane protein is suspected.

Inhibition of erythrocyte anion exchange by tolrestat, an inhibitor of aldose reductase

Tolrestat, and aldose-reductase inhibitor, was shown to be a rapid and potent inhibitor of chloride exchange on the band 3 protein of human erythrocytes. Tolrestat binds to a site distinct from the chloride transport site and binds to one half of the transporters at 5 × 10 −7 mol/L in the absence of chloride and at 3.6 × 10 −5 mol/L in physiologic chloride concentrations. Although these concentrations are 20- to 1,000-fold greater than the IC 50 for aldose-reductase inhibition by tolrestat, they are achieved during routine pharmacologic therapy in humans. Consequently, Cl HCO 3 exchange rates may be reduced, and there may be decreased CO 2 clearance from coronary and respiratory center capillary beds and inappropriate hyperpnea. There also may be transitory intracellular alkalinization in cells with a Cl HCO 3 exchanger in their plasma membrane.

Effects of lipid packing on polymorphic phase behavior and membrane properties.

The self assembly of phospholipid molecules in the bilayer form was considered in terms of equivalent molecular shapes representing intermolecular forces. The equivalent size of each phospholipid headgroup was approximated by the net atomic volume plus the volume of the associated water molecules, which was derived from water/hydrocarbon partitioning experiments. The equivalent lengths of unsaturated acyl chains were derived from the retention time data from chromatographic measurements. The spontaneous curvature of various phospholipid monolayers was calculated from their equivalent molecular shapes, and the energy required to flatten them to the bilayer plane was calculated, using the known bending modulus. With increasing bending energy, the mixtures showed increasing susceptibility to phospholipase A2, facilitated lipid transfer rate by phospholipid exchange proteins, permeability to carboxyfluorescein, incorporation of human erythrocyte proteins, and calcium transport by Ca-ATPase from sarcoplasmic reticulum in reconstituted vesicles. When the calculation was applied to known lipid compositions of nine cellular membranes, the protein/lipid ratio and phospholipid/cholesterol ratio were found to have a positive and a negative correlation, respectively, with the latent bending energy of the phospholipids. The energy expense in conforming to a bilayer phase may be an important physical parameter regarding the activity and the biogenesis of membranes.

Purification of Erythrocyte Dematin (protein 4.9) Reveals an Endogenous Protein Kinase That Modulates Actin-bundling Activity

A partially purified preparation of human erythrocyte protein 4.9, consisting of 48-, 52-, and 55-kilodalton polypeptides, is capable of bundling rabbit muscle actin in vitro (Siegel, D. L., and Branton, D. (1985) J. Cell Biol. 100, 775-785). Purification schemes, peptide mapping, antibody cross-reactivity, and chemical cross-linking techniques show that the 48- and 52-kDa polypeptides are sequence-related phosphorylated components, whereas the 55-kDa polypeptide is not. Purified protein 4.9 (dematin), consisting of 48- and 52-kDa polypeptides, effectively bundles actin in vitro; under similar conditions, the isolated 55-kDa polypeptide does not bundle actin. In fact, when added back to purified dematin, fractions containing the 55-kDa polypeptide can completely abolish dematin's actin-bundling activity. The basis for this inhibitory activity is an endogenous protein kinase that phosporylates both the 48- and 52-kDa isoforms of dematin, thus abolishing dematin's actin-bundling activity (Husain-Chishti, A., Levin, A., and Branton, D. (1988) Nature 334, 718-721). Although the endogenous kinase often co-purifies with the 55-kDa polypeptide, it can be separated from the 55-kDa polypeptide and has the characteristics of a catalytic subunit of a cyclic AMP-dependent protein kinase.

The rhesus rotavirus outer capsid protein VP4 functions as a hemagglutinin and is antigenically conserved when expressed by a baculovirus recombinant

Rhesus rotavirus (RRV) gene 4 was cloned into lambda bacteriophage, inserted into a polyhedrin promoter shuttle plasmid, and expressed in Sf9 cells by a recombinant baculovirus. The baculovirus-expressed VP4 protein made up approximately 5% of the Spodoptera frugiperda-infected cell protein. Monoclonal antibodies that neutralize the virus bound to the expressed VP4 polypeptide, indicating that the expressed VP4 protein was antigenically indistinguishable from viral VP4. In addition, we have determined that the baculovirus-expressed VP4 protein bound to erythrocytes and functions as the RRV hemagglutinin. The endogenous hemagglutinating activity of the VP4 protein, like the virus, was inhibited by guinea pig antirotavirus hyperimmune serum and by VP4-specific neutralizing monoclonal antibodies. The human erythrocyte protein, glycophorin, also inhibited hemagglutination by RRV or the expressed VP4 protein and appears to be the rotavirus erythrocyte receptor. The baculovirus-expressed VP4 protein was conserved functionally and antigenically in the absence of other outer or inner capsid rotavirus components and represents a logical candidate for future immunological studies.

Immunocytochemical localization of the glucose-transport protein in mammalian brain capillaries.

The endothelial cells of mammalian brain capillaries, which form the anatomical basis of the blood-brain barrier, have been investigated by immunocytochemical methods to determine the distribution of the glucose-transport protein. A monoclonal antibody raised against the intact human erythrocyte glucose-transport protein and polyclonal antibodies raised against a synthetic peptide corresponding to the C-terminal sequence of the human erythrocyte glucose-transport protein were used for immunofluorescent staining of isolated human and bovine cerebral cortex microvessels. The pattern of fluorescence with both antibodies demonstrated the antigen to be distributed throughout the plasma membrane of the capillary endothelial cells. These results provide further evidence for the homology between the human erythrocyte and brain capillary glucose-transport protein, and confirm its abundance in brain capillaries.

Erythrocyte adducin

Two major substrates for human erythrocyte protein kinase C (PK-C) of Mr 120,000 and 110,000, previously named PKC-1 and PKC-2 [Palfrey, H. C. & Waseem, A. (1985) J. Biol. Chem. 260, 16021-16029] have been found to be identical to CaM-BP 103/97 or 'adducin', recently described by K. Gardner and V. Bennett [(1986) J. Biol. Chem. 261, 1339-1348; (1987) Nature (Lond.) 328, 359-362]. These proteins have been purified from the membrane skeleton by high-salt extraction, ion-exchange and gel filtration chromatography. The two proteins co-fractionate in a ratio of approximately 1:1 under a number of conditions suggesting that they exist as a complex. Physicochemical data indicate that the native adducin complex is probably an asymmetric heterodimer of alpha and beta subunits. Adducin binds to a calmodulin (CaM) affinity matrix in a Ca2+-dependent manner and is specifically eluted with EGTA. Fingerprinting of the iodinated peptides derived from the alpha and beta subunits using three different proteases yields 16-37% overlapping peptides, indicating limited similarity between the two polypeptides. Affinity-purified polyclonal antibodies against each protein show little or no cross-reactivity with the other, indicating that the beta subunit is not derived from the alpha subunit or vice versa. Proteins reactive with both anti-(alpha-adducin) and anti-(beta-adducin) antibodies are found in erythrocytes from rat, rabbit, pig, ferret and duck. Immunoblots of adducin after non-ionic detergent extraction of ghosts reveal that a significant fraction of the protein may associate with non-skeleton membrane components. The phosphorylation of adducin is stimulated by both phorbol esters and cAMP analogues in intact erythrocytes. Fingerprinting suggests that protein kinase C preferentially phosphorylates four distinct sites on the two proteins. Phosphopeptide maps of alpha-adducin are virtually identical to those of beta-adducin after phorbol ester stimulation of intact cells, or after PK-C-catalyzed phosphorylation of the purified protein, indicating strong local similarities in the two proteins. Such maps also suggest that cAMP-dependent protein kinase (cAMP-PK) modifies adducin at some similar and some distinct sites as those modified by PK-C. In vitro phosphorylation of isolated adducin by purified PK-C results in rapid incorporation of phosphate to a final level of approximately 1.5 mol/mol in both alpha and beta subunits.(ABSTRACT TRUNCATED AT 400 WORDS)