Rotational diffusion of the erythrocyte integral membrane protein band 3: effect of hemichrome binding.

Human erythrocyte band 3 was covalently labeled within the integral membrane domain by incubating intact erythrocytes with the phosphorescent probe eosinyl-5-maleimide. The rotational diffusion of band 3 in membranes prepared from these labeled cells was measured using the technique of time-resolved phosphorescence anisotropy. Three rotational correlation times ranging from 16 to 3800 microseconds were observed, suggesting that band 3 exists in different aggregate states within the plane of the membrane. The oxidizing agent phenylhydrazine was used to induce hemichrome formation within intact erythrocytes. The immobilization of band 3 in membranes prepared from these erythrocytes suggests that the binding of hemichromes induces clustering of band 3. The addition of purified hemichromes to erythrocyte ghosts leads to a similar effect. We have also examined the mobility of the cytoplasmic domain of band 3. This region was labeled indirectly using a phosphorescently labeled antibody which binds to an epitope within the cytoplasmic domain. We observed very rapid motion of the cytoplasmic region of band 3, which was only partially restricted upon hemichrome binding. This suggests that the integral and cytoplasmic domains of band 3 may be independently mobile.     

Ultrastructural localization of erythrocyte cytoskeletal and integral membrane proteins in Plasmodium falciparum-infected erythrocytes

The distributions of ankyrin, spectrin, band 3, and glycophorin A were examined in Plasmodium falciparum-infected erythrocytes by immunoelectron microscopy to determine whether movement of parasite proteins and membrane vesicles between the parasitophorous vacuole membrane and erythrocyte surface membrane involves internalization of host membrane skeleton proteins. Monospecific rabbit antisera to spectrin, band 3 and ankyrin and a mouse monoclonal antibody to glycophorin A reacted with these erythrocyte proteins in infected and uninfected human erythrocytes by immunoblotting. Cross-reacting malarial proteins were not detected. The rabbit sera also failed to immunoprecipitate [3H]isoleucine labeled malarial proteins from Triton X-100 and sodium dodecyl sulfate (SDS) extracts of infected erythrocytes. These three antibodies as well as the monoclonal antibody to glycophorin A bound to the membrane skeleton of infected and uninfected erythrocytes. The parasitophorous vacuole membrane was devoid of bound antibody, a result indicating that this membrane contains little, if any, of these host membrane proteins. With ring-, trophozoite- and schizont-infected erythrocytes, spectrin, band 3 and glycophorin A were absent from intracellular membranes including Maurer's clefts and other vesicles in the erythrocyte cytoplasm. In contrast, Maurer's clefts were specifically labeled by anti-ankyrin antibody. There was a slight, corresponding decrease in labeling of the membrane skeleton of infected erythrocytes. A second, morphologically distinct population of circular, vesicle-like membranes in the erythrocyte cytoplasm was not labeled with anti-ankyrin antibody. We conclude that membrane movement between the host erythrocyte surface membrane and parasitophorous vacuole membrane involves preferential sorting of ankyrin into a subpopulation of cytoplasmic membranes.     

Characterization of a modified red cell membrane protein expressed on erythrocytes infected with the human malaria parasite Plasmodium falciparum: possible role as a cytoadherent mediating protein

Infections with the human malaria Plasmodium falciparum are characterized by the retention of parasitized erythrocytes in tissue capillaries and venules. Erythrocytes containing trophozoites and schizonts attach to the endothelial cells that line these vessels by means of structurally identifiable excrescences present on the surface of the infected cell. Such excrescences, commonly called knobs, are visible by means of scanning or transmission electron microscopy. The biochemical mechanisms responsible for erythrocyte adherence to the endothelial cell are still undefined. In an attempt to identify the cytoadhesive molecule on the surface of the infected cell, we have prepared monoclonal antibodies to knob-bearing erythrocytes infected with the FCR-3 strain of P. falciparum. One of these monoclonal antibodies, designed 4A3, is an IgM that reacts (by means of immunofluorescence) with the surface of unfixed erythrocytes bearing mature parasites of the knobby line; it does not react with knobless lines or uninfected erythrocytes. By immunoelectron microscopy the monoclonal antibody 4A3 was localized to the knob region. In an in vitro cytoadherence assay, the monoclonal antibody partially blocked the binding of knob-bearing cells (FCR-3 strain) to formalin-fixed amelanotic melanoma cells. The monoclonal antibody was used to immunoprecipitate a protein from extracts of knobby erythrocytes that had been previously surface iodinated. By a two-dimensional peptide mapping technique, the antigen recognized by the monoclonal antibody was found to be structurally related to band 3 protein, the human erythrocyte anion transporter.     

Expression of the parasite protein Pc90 in plasma membranes of erythrocytes infected with Plasmodium chabaudi

Erythrocytes infected with the malaria parasite Plasmodium chabaudi contain the neo-protein Pc90 in their plasma membrane. We investigate origin, membrane disposition, and intraerythrocytic traffic of this Pc90. Metabolic labeling of P.-infected erythrocytes, combined with cell fractionation as well as Western blot analysis and immunoprecipitation using a Pc90-recognizing monoclonal antibody, show that Pc90 is synthesized by early to mid trophozoites and is transported without any apparent processing steps to the erythrocyte membrane. Based upon the inaccessibility of Pc90 from the outside in intact erythrocytes and the water solubility of membrane-associated Pc90, it is concluded that Pc90 is localized on the cytoplasmic face of the host erythrocyte membrane. Immunoelectron microscopy using a Pc90-specific monoclonal antibody and the occurrence of soluble Pc90 in host cell cytosol indicate that the Pc90 is transported in both a 'vesicle-bound' and a 'free' form through the erythrocyte cytoplasm.     

Identification and immunochemical characterization of the human erythrocyte membrane glycoproteins that carry the Xga antigen.

When the membrane components of Xg(a+) erythrocytes were separated by electrophoresis and immunoblotted with an anti-Xga of human origin, two diffuse bands of approximate Mr 22,000-25,000 and 26,500-29,000 were stained. These reactive components were not evident in membranes from proteinase-treated Xg(a+) erythrocytes. Neuraminidase treatment of erythrocytes before immunoblotting resulted in one diffuse Xga-reactive band, the leading edge of which had a slightly increased mobility corresponding to a decrease in Mr of approx. 1500. These results suggest that the membrane components that carry Xga are sialoglycoproteins. A genetic relationship exists between Xga and the antigen recognized by the monoclonal antibody 12E7. An immunochemical comparison of the structures that carry the Xga and 12E7 antigens demonstrated that they differ in Mr and in the way in which they are modified by neuraminidase. This is in accord with evidence that Xga and 12E7 are the products of two separate structural loci.     

Immunological evidence that band 3 is the major glucose transporter of the human erythrocyte membrane

We have previously reported that human erythrocyte band 3 contains 90-95% of the reconstitutable glucose transport activity of the erythrocyte membrane (Shelton, R.L. and Langdon, R.G. (1983) Biochim. Biophys. Acta 733, 25-33). We have now found that monoclonal and polyclonal antibodies to epitopes on band 3 specifically removed band 3 and more than 90% of the reconstitutable glucose transport activity from unfractionated octylglucoside extracts of erythrocyte membranes; nonimmune serum removed neither. Western blots of whole membrane extracts revealed that the polyclonal antibody to band 4.5 used to isolate cDNA clones presumed to code for the transporter (Mueckler, M., Caruso, C., Baldwin, C.A., Pancio, M., Blench, J., Morris, H.B., Allard, W.J., Lienhard, G.E. and Lodish, H.F. (1985) Science 229, 941-945) reacts strongly with six discrete bands in the 4.5 region. A monoclonal antibody to band 3 also reacts with a Mr 55,000 component of band 4.5. We conclude that band 3 contains the major glucose transporter of human erythrocytes, and that the transport activity in band 4.5 might be attributable to a band 3 fragment. Band 3 is probably a multifunctional transport protein responsible for transport of glucose, anions, and water.     

Degradation of the human erythrocyte membrane band 3 studied with monoclonal antibody directed against an epitope on the cytoplasmic fragment of band 3

The mouse hybridoma monoclonal antibody BIII.136 of the IgG2a class is specific for human erythrocyte band-3 protein. It was shown by means of immunoblotting and immunoprecipitation assays that the antibody recognized an epitope located in the cytoplasmic pole of the band-3 molecule within approximately 20 kDa from the N-terminal end. The N-terminal fragments of band-3 protein, migrating in SDS/polyacrylamide gel electrophoresis in the 60-kDa, 40-kDa and 20-kDa regions, were detected with the antibody in untreated red-cell membranes as products of autolysis of band-3 protein. A correlation was found between the amount of these fragments and erythrocyte age, which suggests that partial degradation of band 3 proceeds in vivo during senescence of erythrocytes. The further degradation of band-3 protein in vitro was not observed in intact erythrocytes stored at 4 degrees C, but progressed distinctly after hemolysis of red cells, during washing and storing the membranes.     

Abnormalities of the erythrocyte membrane in chronic myelogenous leukemia

Chronic myelogenous leukemia (CML) is a hematologic malignancy characterized by excessive growth of myeloid cells and their progenitors. The proportion of spectrin dimers compared to tetramers extracted from membranes at 4 degrees C, under low ionic strength conditions, increased in CML erythrocytes. These also displayed abnormal thermal sensitivity (between 45 and 46 instead of 49 degrees C). Crosslinking with the bifunctional reagent, dimethyl adipimidate (8.6 A) showed significant organizational modification of not only spectrin, but other cytoskeletal components such as ankyrin, bands 4.2 and 5. Enhanced concanavalin A (Con-A) agglutinability of CML erythrocytes also suggests altered topographic distribution of a functionally important membrane protein, band 3. The anion transport activities of erythrocytes from patients with CML and normal donors were comparable. In CML erythrocytes, significant reduction in the number of ankyrin-binding sites, present in the cytoplasmic domain of band 3, may lead to partial loss of cytoskeletal anchorage to the bilayer and account for their increased Con-A agglutinability and heat sensitivity and may lead to their premature removal from the circulation.     

Interaction of the aldolase and the membrane of human erythrocytes.

Up to 80% of cellular aldolase (EC 4.1.2.13) was retained in the membrane fraction isolated following hemolysis of human erythrocytes under appropriate conditions. Binding was reversed by increasing the pH and ionic strength. Millimolar levels of the substrate, fructose 1,6-bisphosphate, selectively eluted aldolase from the membrane, while related metabolites did not. Using the membrane as a high affinity adsorbant, electrophoretically pure aldolase of high specific activity was prepared in high yield. The reassociation of pure aldolase and membranes was characterized. The sole site of human erythrocyte aldolase binding was shown to be the cytoplasmic surface domain of band 3, the predominant membrane-spanning polypeptide. One aldolase molecule was bound per band 3 polypeptide. Upon binding to either whole membranes, solubilized band 3, or proteolytic fragments from the cytoplasmic surface pole of band 3, aldolase underwent a profound loss of catalytic activity, reversed by raising the substrate concentration.     

Association of spectrin with its membrane attachment site restricts lateral mobility of human erythrocyte integral membrane proteins

Interactions between spectrin and the inner surface of the human erythrocyte membrane have been implicated in the control of lateral mobility of the integral membrane proteins. We report here that incubation of “leaky” erythrocytes with a water-soluble proteolytic fragment containing the membrane attachment site for spectrin achieves a selective and controlled dissociation of spectrin from the membrane, and increases the rate of lateral mobility of fluorescein isothiocyanate-labeled integral membrane proteins (> 70% of label in band 3 and PAS-1). Mobility of membrane proteins is measured as an increase in the percentage of uniformly fluorescent cells with time after fusion of fluorescent with nonfluorescent erythrocytes by Sendai virus. The cells are permeable to macromolecules since virus-fused erythrocytes lose most of their hemoglobin. The membrane attachment site for spectrin has been solubilized by limited proteolysis of inside-out erythrocyte vesicles and has been purified (V). Bennett, J Biol Chem 253:2292 (1978). This 72,000-dalton fragment binds to spectrin in solution, competitively inhibits association of ³²P-spectrin with inside-out vesicles with a K_i of 10^?7M, and causes rapid dissociation of ³²P-spectrin from vesicles. Both acid-treated 72,000-dalton fragment and the 45,000 dalton-cytoplasmic portion of band 3, which also was isolated from the proteolytic digest, have no effect on spectrin binding, release, or membrane protein mobility. The enhancement of membrane protein lateral mobility by the same polypeptide that inhibits binding of spectrin to inverted vesicles and displaces spectrin from these vesicles provides direct evidence that the interaction of spectrin with protein components in the membrane restricts the lateral mobility of integral membrane proteins in the erythrocyte.     

Basis of unique red cell membrane properties in hereditary ovalocytosis.

Hereditary ovalocytes from a Mauritian subject are extremely rigid, with a shear elastic modulus about three times that of normal cells, and have increased resistance to invasion by the malaria parasite Plasmodium falciparum in vitro. The genetic anomaly resides in band 3; the protein gives rise to chymotryptic fragments with reduced mobility in SDS/polyacrylamide gel electrophoresis, but this is a result of anomalous binding of SDS and not a higher molecular weight. Analysis of the band 3 gene reveals (1) a point mutation (Lys56----Glu), which also occurs in a common asymptomatic band 3 (Memphis) variant and governs the electrophoretic properties, and (2) a deletion of nine amino acid residues, including a proline residue, encompassing the interface between the membrane-associated and the N-terminal cytoplasmic domains. The interaction of the mutant band 3 with ankyrin appears unperturbed. The fraction of band 3 capable of undergoing translation diffusion in the membrane is greatly reduced in the ovalocytes. Cells containing the asymptomatic band 3 variant were normal with respect to all the properties that we have studied. Possible mechanisms by which a structural change in band 3 at the membrane interface could regulate rigidity are examined.     

The parasitophorous vacuole membrane of Plasmodium falciparum: demonstration of vesicle formation using an immunoprobe

We have applied several immunolabeling techniques using a monoclonal antibody to a Plasmodium falciparum antigen to differentiate morphologically dissimilar membranous structures present in infected erythrocytes. Evidence is presented that cytoplasmic clefts, multimembranous structures and vesicles within the infected cell originate from the parasitophorous vacuole membrane by a process described as budding off. The parasitophorous vacuole membrane and related structures in infected, parasitized erythrocytes reacted with the cyanine dye Merocyanine 540, demonstrating that they are accessible to molecules from the extracellular environment. Immunogold labeling of freeze-fractured preparations and of thin sections of parasitized cells using pre- and post-embedding techniques revealed that each of the membranous structures carried a common parasite antigen, QF 116, which was identified by monoclonal antibody 8E7/55.     

Inhibition of phosphate transport across the human erythrocyte membrane by chemical modification of sulfhydryl groups.

Effects of sulfhydryl-reactive reagents on phosphate transport across human erythrocyte membranes were examined using 31P NMR. Phosphate transport was significantly inhibited in erythrocytes treated with sulfhydryl modifiers such as N-ethylmaleimide, diamide, and Cu2+/o-phenanthroline. Quantitation of sulfhydryl groups in band 3 showed that the inhibition is closely associated with the decrease of sulfhydryl groups. Data from erythrocytes treated with diamide or Cu2+/o-phenanthroline demonstrated that intermolecular cross-linking of band 3 by oxidation of a sulfhydryl group, perhaps Cys-201 or Cys-317, decreases the phosphate influx by about 10%. The inhibition was reversed by reduction using dithiothreitol. These results suggest that sulfhydryl groups in the cytoplasmic domain of band 3 may play an important role in the regulation of anion exchange across the membrane.     

Quantitative composition and characterization of the proteins in membrane vesicles released from erythrocytes by dimyristoylphosphatidylcholine. A membrane system without cytoskeleton

Membrane vesicles were prepared by incubation of human erythrocytes with dimyristoylphosphatidylcholine [3] and isolated by isopycnic centrifugation on Dextran density gradients. Protein analyses were carried out with crossed immunoelectrophoresis and dodecylsulfate polyacrylamide gel electrophoresis. The right-side-out-oriented membrane vesicles contained membrane and cytoplasmic proteins of the erythrocyte but lacked cytoskeletal components. Comparison of proteins in vesicles and erythrocyte membranes showed that acetylcholinesterase was enriched two to six times in the vesicles relative to both membrane-spanning proteins, band 3, and glycophorin. Two further, hitherto unidentified, sialic acid-containing membrane antigens were found in the vesicles. Both faced the outside of the membranes and were enriched two to seven times. Ankyrin was not present in the membrane vesicles and spectrin could not be detected by dodecylsulfate polyacrylamide gel electrophoresis. We suggest that the redistribution of proteins in the vesicles reflects differences in their interactions with other membrane components and their relative mobility within the erythrocyte membrane.     

Analysis of integral membrane protein contributions to the deformability and stability of the human erythrocyte membrane

Three major hypotheses have been proposed to explain the role of membrane-spanning proteins in establishing/maintaining membrane stability. These hypotheses ascribe the essential contribution of integral membrane proteins to (i) their ability to anchor the membrane skeleton to the lipid bilayer, (ii) their capacity to bind and stabilize membrane lipids, and (iii) their ability to influence and regulate local membrane curvature. In an effort to test these hypotheses in greater detail, we have modified both the membrane skeletal and lipid binding interactions of band 3 (the major membrane-spanning and skeletal binding protein of the human erythrocyte membrane) and have examined the impact of these modifications on erythrocyte membrane morphology, deformability, and stability. The desired changes in membrane skeletal and protein-lipid interactions were induced by 1) reaction of the cells with 4,4'-diisothiocyanostilbene-2,2'-disulfonate (DIDS), an inhibitor of band 3-mediated anion transport that dissociates band 3 into dimers (increasing its surface area in contact with lipid) and severs band 3 linkages to the membrane skeleton; 2) a fragment of ankyrin that ruptures the same ankyrin-band 3 bridge to the membrane skeleton, but drives the band 3 subunit equilibrium toward the tetramer (i.e. decreasing the band 3 surface area in contact with lipid); and 3) an antibody to the ankyrin-binding site on band 3 that promotes the same changes in band 3 skeletal and lipid interactions as the ankyrin fragment. We observed that although DIDS induced echinocytic morphological changes in the treated erythrocytes, it had little impact on either membrane deformability or stability. In contrast, resealing of either the ankyrin fragment or anti-band 3 IgG into erythrocytes caused spontaneous membrane fragmentation and loss of deformability/stability. Because these and other new observations cannot all be reconciled with any single hypothesis on membrane stability, we suggest that more than one hypothesis may be operative and provide an explanation of how each might individually contribute to net membrane stability.     

Localization of the C termini of the Rh (rhesus) polypeptides to the cytoplasmic face of the human erythrocyte membrane.

We have raised a rabbit antiserum to a synthetic peptide corresponding to the C terminus (residues 400-416) of the Rh30A polypeptide. The rabbit antiserum reacted with the Rh30B (D30) polypeptide in addition to the Rh30A (C/c and/or E/e) polypeptide(s), indicating that these proteins share homology at their C termini. The antiserum did not react with erythrocyte membranes from an individual with Rh(null) syndrome. The rabbit antiserum immunoprecipitated Rh polypeptides from erythrocyte membranes and alkali-stripped membranes, but not from intact erythrocytes. Treatment of intact red cells with carboxypeptidase Y did not affect the reactivity of the antiserum, whereas treatment of alkali-stripped and unsealed erythrocyte ghost membranes resulted in the loss of antibody binding. Carboxypeptidase A treatment of intact erythrocytes and alkali-stripped membranes had no effect on antibody binding, indicating that the C-terminal domains of the Rh polypeptides contain lysine, arginine, proline, or histidine residues. These results show that the C termini of the Rh polypeptides are located toward the cytoplasmic face of the erythrocyte membrane. Treatment of intact radioiodinated erythrocytes with bromelain followed by immunoprecipitation with monoclonal anti-D gave a band of M(r) 24,000-25,000, indicating that the Rh30B (D30) polypeptide is cleaved at an extracellular domain close to the N or C terminus, with loss of the major radioiodinated domain. Immunoblotting of bromelain treated D-positive erythrocyte membranes with the rabbit antiserum to the C-terminal peptide revealed a new band of M(r) 6000-6500, indicating that the extracellular bromelain cleavage site is located near the C terminus of the molecule. The band of M(r) 6000-6500 was not obtained in erythrocyte membranes derived from bromelain treated D-negative erythrocytes. Erythrocytes of the rare -D- phenotype appear to either totally lack, or have gross alterations in, the Cc/Ee polypeptide(s), since the bromelain treatment of these cells resulted in the total loss of staining in the M(r) 35,000-37,000 region and the concomitant appearance of the new band of M(r) 6000-6500.     

Interactions between protein 4.1 and band 3. An alternative binding site for an element of the membrane skeleton

Protein 4.1 from human erythrocytes formed a complex with band 3 in inside-out erythrocyte membrane vesicles and with soluble peptides derived from the cytoplasmic domain of band 3. Protein 4.1 labeled metabolically with 32P bound saturably to vesicles depleted of endogenous protein 4.1. The soluble cytoplasmic domain of band 3 (43K) competitively displaced approximately 60% of bound 32P-protein 4.1 from reconstituted membrane vesicles. Pretreatment of vesicles with anti-43K similarly inhibited the rebinding of protein 4.1. In solution, 125I-43K formed a complex with protein 4.1 that saturated at 1:1 stoichiometry and migrated as a discrete band when analyzed by nondenaturing polyacrylamide gel electrophoresis. In rate-zonal sedimentation in isotonic salt solutions, protein 4.1 and 43K sedimented as a sharp peak at 4.4 S. In experiments aimed at exploring the role of the protein 4.1-band 3 interaction in the organization of the membrane skeleton, the effect of spectrin was investigated. Spectrin and protein 4.1 formed a complex which co-sedimented in sucrose gradients, but the addition of 43K to preformed spectrin-protein 4.1 complexes resulted in disruption of the complex and co-sedimentation of most of the protein 4.1 with 43K. These results suggest that protein 4.1 can associate with band 3 in the erythrocyte membrane and that this association may modulate the attachment of the membrane skeleton to the membrane.     

The influence of melittin on the rotation of band 3 protein in the human erythrocyte membrane

The rotational mobility of band 3, a protein constituent of the human erythrocyte membrane, was measured by observing the flash-induced transient dichroism of the triplet probe eosin maleimide. In the presence of melittin, a pharmacologically active polypeptide from honey bee (Apis mellifera) venom, a dose-dependent loss of rotational mobility was detected. With acetylated melittin, the ability to immobilise is reduced. Succinylated melittin, however, is devoid of immobilising activity.The possible relevance of these findings to the normal mode of action of melittin was examined by comparing the relative abilities of the native, acetylated and succinylated melittins to lyse erythrocytes and synergise with phospholipase A₂, another constituent of bee venom. For both these properties, the order of effectiveness is native melittin > acetyl melittin > succinyl melittin = 0, the same as their order of effectiveness in immobilising band 3.A mechanism is proposed in which melittin is anchored in the membrane by its hydrophobic N-terminus, while its cationic C-terminal moiety binds to negatively charged residues on membrane proteins. This leads either directly or indirectly to protein aggregation and hence loss of mobility. From a detailed comparison of the different effects of the melittin derivatives, it is concluded that melittin may function in vivo by aggregating membrane proteins in order to allow phospholipase A₂ to gain access to the membrane bilayer and commence catalysis.     

Localization of the ankyrin-binding site on erythrocyte membrane protein, band 3

The predominant attachment site of the spectrin-based cytoskeleton to the erythrocyte membrane occurs via the interaction of ankyrin with the cytoplasmic domain of band 3 (cdb3). In order to further characterize this interaction, we have conducted experiments to localize the ankyrin-binding site on cdb3. Four monoclonal and three antipeptide polyclonal antibodies were raised against cdb3 and used in competition studies to identify regions of close association of cdb3 with ankyrin. Antibodies to regions of cdb3 near the cytoplasmic domain-membrane spanning domain junction had no effect on 125I-ankyrin binding. Likewise, an antibody to a highly conserved region between residues 142 and 154 did not inhibit ankyrin binding. However, antibodies at or near the cysteine 201-317 cluster and the proposed proline-rich hinge in the center of cdb3 were potent inhibitors of ankyrin association, as were antibodies to the acidic NH2 terminus. Additional evidence for interaction of ankyrin with the NH2-terminal region of cdb3 was obtained by demonstrating the ability of ankyrin to inhibit tyrosine phosphorylation of cdb3 at its NH2 terminus by a purified calf thymus tyrosine kinase. These studies reveal two regions of cdb3, distant in primary sequence, which interact with ankyrin. A specific conformation of cdb3 may be required to permit these regions to simultaneously associate with ankyrin and allow binding to occur.     

Plasmodium falciparum: analysis of the interaction of antigen Pf155/RESA with the erythrocyte membrane

The location of the Plasmodium falciparum vaccine candidate antigen Pf155/RESA in the membrane of infected erythrocytes was analzyed by means of selective surface radioiodination and immunofluorescence of surface-modified cells. The lack of radiolabel in Pf155/RESA as well as its localization by immunofluorescence similar to that of the N-terminal region of erythrocyte band 3 suggests that the antigen is associated with the cytoplasmic phase of the erythrocyte membrane. In concordance with this, Pf155/RESA was detected by immunofluorescence on the surface of inside out membrane vesicles from P. falciparum-infected erythrocytes. Pf155/RESA from spent culture medium also bound to inside out membrane vesicles of normal erythrocytes as well as to cytoskeletal shells of such vesicles, but failed to bind to sealed right-side out membrane vesicles. Depletion of spectrin from the vesicles abolished antigen binding, suggesting that Pf155/RESA association with the erythrocyte cytoskeleton is mediated by spectrin.